NMR Spectroscopy of Polymers:

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.fw001

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Innovative Strategies for Complex

Macromolecules

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.fw001
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 ACS SYMPOSIUM SERIES 1077

NMR Spectroscopy of Polymers:

Innovative Strategies for Complex
Macromolecules
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.fw001

H. N. Cheng, Editor
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Southern Regional Research Center,

U.S. Department of Agriculture, Agricultural Research Service
New Orleans, Louisiana

Tetsuo Asakura, Editor

Tokyo University of Agriculture and Technology
Koganei, Tokyo

Alan D. English, Editor

DuPont Central Research and Development
Wilmington, Deleware

Sponsored by the
ACS Division of Polymer Chemistry, Inc.

American Chemical Society, Washington, DC

Distributed in print by Oxford University Press, Inc.

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Library of Congress Cataloging-in-Publication Data

NMR spectroscopy of polymers : innovative strategies for complex macromolecules /

H. N. Cheng, Tetsuo Asakura, Alan D. English, editor[s].
p. cm. -- (ACS symposium series ; 1077)
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.fw001

"Sponsored by the ACS Division of Polymer Chemistry, Inc., American Chemical Society,
Washington DC."
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Includes bibliographical references and index.

ISBN 978-0-8412-2667-8
1. Polymers--Spectra--Congresses. 2. Nuclear magnetic resonance spectroscopy--
Congresses. 3. Polymers--Analysis--Congresses. I. Cheng, H. N. II. Asakura, Tetsuo.
III. English, Alan D., 1947- IV. American Chemical Society. Division of Polymer Chemistry.
QC463.P5N58 2011
547.7046--dc23
2011036489

The paper used in this publication meets the minimum requirements of American National
Standard for Information SciencesPermanence of Paper for Printed Library Materials,
ANSI Z39.48n1984.

Copyright 2011 American Chemical Society

Distributed in print by Oxford University Press, Inc.

All Rights Reserved. Reprographic copying beyond that permitted by Sections 107 or 108
of the U.S. Copyright Act is allowed for internal use only, provided that a per-chapter fee of
$40.25 plus $0.75 per page is paid to the Copyright Clearance Center, Inc., 222 Rosewood
Drive, Danvers, MA 01923, USA. Republication or reproduction for sale of pages in this
book is permitted only under license from ACS. Direct these and other permission requests
to ACS Copyright Office, Publications Division, 1155 16th Street, N.W., Washington, DC
20036.

The citation of trade names and/or names of manufacturers in this publication is not to be
construed as an endorsement or as approval by ACS of the commercial products or services
referenced herein; nor should the mere reference herein to any drawing, specification,
chemical process, or other data be regarded as a license or as a conveyance of any right
or permission to the holder, reader, or any other person or corporation, to manufacture,
reproduce, use, or sell any patented invention or copyrighted work that may in any way be
related thereto. Registered names, trademarks, etc., used in this publication, even without
specific indication thereof, are not to be considered unprotected by law.

PRINTED IN THE UNITED STATES OF AMERICA

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Foreword
The ACS Symposium Series was first published in 1974 to provide a
mechanism for publishing symposia quickly in book form. The purpose of
the series is to publish timely, comprehensive books developed from the ACS
sponsored symposia based on current scientific research. Occasionally, books are
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.fw001

developed from symposia sponsored by other organizations when the topic is of

keen interest to the chemistry audience.
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Before agreeing to publish a book, the proposed table of contents is reviewed

for appropriate and comprehensive coverage and for interest to the audience. Some
papers may be excluded to better focus the book; others may be added to provide
comprehensiveness. When appropriate, overview or introductory chapters are
added. Drafts of chapters are peer-reviewed prior to final acceptance or rejection,
and manuscripts are prepared in camera-ready format.

As a rule, only original research papers and original review papers are
included in the volumes. Verbatim reproductions of previous published papers
are not accepted.

ACS Books Department

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Preface
Nuclear magnetic resonance (NMR) spectroscopy is a premiere technique
for the studies of polymers. Since 1950s, the literature on this topic has grown
rapidly. A large body of ingenious techniques has been developed that have
enriched the field and provided NMR practitioners with many valuable tools to use
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.pr001

in their work. The field now covers a broad spectrum of activities, from polymer
identification, quality control, and reaction monitoring to structure determination,
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

polymer morphology, compatibility, chain conformation, dynamics, diffusion,

and imaging.
For many years we have organized successful symposia devoted to NMR
Spectroscopy of Polymers at roughly every 3 years. The purpose is to provide
an international forum for the presentation of recent advances in NMR techniques
and characterization of natural and synthetic polymers, including agriculturally
based materials. We strive to invite speakers who are acknowledged leaders in
this field. Over the years, it has become the largest and the foremost symposia for
the NMR studies of polymers.
This book is based on such a symposium held at the Pacifichem meeting
on December 17-21, 2010 in Honolulu, Hawaii. At this meeting, there were 43
talks and 24 posters, representing the state-of-the-art research of polymer NMR
in solids, liquids, and imaging. Many exciting new techniques and findings were
reported.
In this symposium book, a total of 30 articles are included, covering the
following topics:
Fundamental and applied research
New methodology development
Polymer structure determination
Polymer dynamics and diffusion
Nanostructures and nanocomposites
Supramolecular assemblies
Blends, miscibility, and heterogeneity
Polymers under confinements and on surfaces
Multidimensional NMR
In addition, we are thankful to Professor Hans Spiess for an overview of solid
NMR and Professor Peter Rinaldi for a tutorial on solution NMR. It is our hope
that this book provides a good representation of what is happening at the forefront
of research in polymer NMR and also captures some of the excitements we saw at
the Pacifichem symposium.
This book is targeted for chemists, physicists, biochemists, and chemical
engineers as well as graduate students who are engaged in research and

xi
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 applications of polymer NMR. It can also be a useful reference book for people
who are interested in these topics.
We thank the authors for their timely contributions and their cooperation while
the manuscripts were being reviewed and revised. Thanks are also due to the
ACS Division of Polymer Chemistry, Inc. for sponsoring the 2010 symposium.
We also acknowledge the generous funding from the ACS Division of Polymer
Chemistry, Inc., Air Force Research Laboratories, Agilent Technologies, Bruker
Corporation, Chlorella Industry Co., Ltd., Japan Medical Materials Co., JOEL,
Ltd., MR Resources, National Science Foundation (NSF) Division of Materials
Research, New Era Enterprises, and REC Materials, Inc.

H. N. Cheng
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.pr001

Southern Regional Research Center

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

USDA Agricultural Research Service

1100 Robert E. Lee Blvd.
New Orleans, LA 70124, U.S.A.

Tetsuo Asakura
Department of Biotechnology
Tokyo University of Agriculture and Technology
2-24-16 Nakacho
Koganei, Tokyo 184-8588, JAPAN

Alan D. English
DuPont Central Research and Development Department
Experimental Station
Wilmington, DE 19880, U.S.A.

xii
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 H. W. Spiess - An Appreciation
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.pr002
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Hans Wolfgang Spiess studied at the University of Frankfurt and was awarded
his Ph.D. there in 1968. He worked as a postdoctoral fellow with R. K. Sheline
at Florida State University (19681970), K. H. Hausser at the MPI for Medical
Research, Heidelberg (19701975), and H. Sillescu at the University of Mainz
(19751983, including habilitation). He held professorships at the University of
Mnster and the University of Bayreuth (19811984). Since 1984, he has been
one of the directors of the Max Planck Institute in Polymer Research in Mainz.
Professor Spiess is an acknowledged leader in polymer NMR. He has
developed a large array of new NMR techniques for elucidating structure,
dynamics, phase behavior and order in synthetic macromolecules and
supramolecular systems. With brilliance and creativity, he applies these
techniques to study new polymer materials in order to relate their microscopic
and macroscopic behavior. His scientific depth and breadth can be illustrated by
the following examples:

- Two-and three-dimensional NMR methods for studying molecular

dynamics
- High-resolution multiple quantum NMR spectroscopy of solids for the
investigation of molecular structure and organization

xiii
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 - Spin diffusion techniques for determining the phase behavior and
compatibility in block copolymers, polymer blends, and core shell
particles
- NMR imaging and hyperpolarization by laser polarization, dynamic
nuclear polarization and parahydrogen induced polarization
- Investigation of polyelectrolytes and biomacromolecules in solution with
a broad variety of EPR methods including nm distance measurements
- Chain motion and its relationship to mechanical properties in crystalline
and amorphous polymers as well as dynamics at the glass transition

In addition to his prolific scientific work, Professor Spiess has served as

chairman of the European Polymer Federation (19911992) and chairman of the
Capital Investment Committee of the German Science Foundation (19941996).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.pr002

He was chairman of the Committee for Electronic Data Processing of the

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

MPG (19972006), President of the Groupement AMPERE (20002006),

and a Member of the Scientific Council of the Federal Republic of Germany
(19992005). He is the current President of the International Society of Magnetic
Resonance. He has also collaborated with the editors of this volume and others
to organize many international meetings dealing with NMR Spectroscopy of
Polymers to promote the development of this field.
He is the recipient of numerous awards and honorary lectureships. These
include the Leibniz-Prize of the DFG (1987), NMR Award of the Eastern
Analytical Symposium (2000), the AMPERE Prize (2002), the Liebig Medal
of the German Chemical Society (2002), the Presidential Medal of Cornell
University (2002), the Award of the Society of Polymer Science, Japan (2003),
the Zavoisky Prize (2010), and the Paul J. Flory Prize for polymer science (2010).
He holds honorary degrees (Dr. h.c.) of the Technical University of Cluj-Napoca,
Romania (1997) and the Adam Mickiewicz University, Poznan, Poland (1998).
Since 2011 he has been an Honorary Member of the Society of Polymer Science,
Japan.
After a distinguished career, Professor Spiess is retiring from the Max Planck
Institute. On behalf of all the authors in this symposium volume, we thank him for
his splendid work over the years and wish him health, happiness, and continued
vigor in years to come.

H. N. Cheng
Tetsuo Asakura
Alan D. English

xiv
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 1

Innovative NMR Strategies for Complex

Macromolecules
H. N. Cheng,*,1 Tetsuo Asakura,2 and Alan D. English3
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

1Southern Regional Research Center, USDA Agricultural Research Service,

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

1100 Robert E. Lee Blvd., New Orleans, LA 70124, U.S.A.

2Department of Biotechnology, Tokyo University of Agriculture and

Technology, 2-24-16 Nakacho, Koganei, Tokyo 184-8588, Japan

3DuPont Central Research and Development Department,

Experimental Station, Wilmington, DE 19880, U.S.A.

*E-mail: [email protected]

In recent years there has been an increasing research emphasis

on complex macromolecular systems. These include polymers
with precise control of structures, multicomponent systems with
higher degrees of organization, polymers involved in micelles,
interfaces, and confined environments, nanochemistry and
nanostructures, biopolymers and bio-inspired chemistry, and
application-driven polymer designs, such as fuel cells, batteries,
and ionic conductors; sensors and information processors; drug
delivery, biomedical devices, and imaging; stimulus-responsive
polymers, gels, and networks with defined function and control.
Successful NMR studies of these polymers require judicious
applications of existing techniques and development of new
or improved strategies and methodologies. In this article the
polymer/NMR literature in 2007-2011 is reviewed in view of
the recent trends in polymer research, with selected examples
taken from the literature and from the chapters included in this
book.

Introduction
Nuclear Magnetic Resonance (NMR) is a premiere technique for polymer
studies (15). The literature in the NMR of polymers continues to grow. Since

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 1990 there have been 3815 articles classified under NMR and polymer science,
and 3463 articles under NMR and material science in the ISI Web of Knowledge
(6). A summary with selected topics is shown in Table 1.

Table 1. Subject analysis of literature search for the category of NMR

since 1990 in the ISI World of Knowledge
topic record count total %
1990- 1995- 2001- 2007-
1994 2000 2006 2011a
chemistry 8284 10290 9681 6999 35254 50.0
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

instrumentation 120 139 122 88 469 0.7

polymer science 1122 1242 1000 451 3815 5.4
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

material science 647 1047 911 858 3463 4.9

biochemistry 3802 3940 3388 2294 13424 19.0
biophysics 1112 1262 1122 720 4216 6.0
food science 180 261 341 294 1076 1.5
plant sci. + agriculture 356 464 457 377 1654 2.3
total 18060 21013 18462 12960 70495 89.8
a Coverage in 2011 up to July 1, 2011 only.

A survey of these articles indicates several recurring research themes. In

addition to the utility of the wide array of NMR techniques to determine polymer
structures and solve polymeric problems, many researchers are concentrating
on increasingly complex macromolecular systems and on the development of
new or improved NMR strategies and methodologies. These trends are partly
stimulated by improved instrumentation, more sophisticated characterization
techniques, innovative synthetic methods, and application-driven polymer
designs. The complex macromolecular systems (7) include the following (not
mutually exclusive) areas: 1) polymers with precise control of structures,
2) multicomponent systems with higher degrees of organization, including
blends, composites, organic-inorganic hybrids, and supramolecular assemblies,
3) polymers involved in micelles, interfaces, and confined environments, 4)
nanochemistry and nanostructures, 5) heterogeneous materials, such as emulsions,
foams, and microporous materials, 6) biopolymers and bio-inspired chemistry,
and 7) application-driven polymer designs, such as polymers for fuel cells,
batteries, and ionic conductors; sensors and information processors; drug delivery
and biomedical devices, stimulus-responsive polymers, gels, and networks with
defined function and control.
It may be noted that this book provides an excellent collection of examples
of these complex macromolecular systems, together with the corresponding NMR
strategies and methodologies designed to study them. The purpose of this article

4
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 is to survey the NMR/polymer literature in the past 4-5 years in view of the
various research themes and polymer types and to highlight these developments
with selected examples taken from the recent literature and from appropriate
chapters of this book.

NMR Strategies and Methodologies

As macromolecular systems become more complex, NMR has been keeping
up with both established and novel strategies and methodologies. Several review
articles in the past 5 years provide good summaries of the latest developments.
The articles by Spiess (8) on solid state NMR and by Rinaldi (9) on solution NMR
in this book are especially outstanding. They are also helpful overviews for people
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

who are relatively new to the field.

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Other recent reviews of NMR of solid polymers include the articles by

Chen and Kurosu (10) on semicrystalline polymers, Paasch and Brunner (11) on
methods and applications that pertain to bioanalysis, Bockmann (12) on strategies
for structure determination of proteins, and Spiess (13) on the combined use
of solid state NMR and other complementary techniques. In addition, solid
state NMR of inorganic polymers has been reviewed by Borisov et al. (14),
and liquid crystals by Domenici (15). As for solution NMR of polymers, there
have been several recent reviews, particularly for biomacromolecules, e.g.,
Felli and Brutscher (16) and Foster et al. (17) on recently developed NMR
methodologies, Mittermaier and Kay (18) on biomacromolecular dynamics, Tate
(19) on anisotropic nuclear spin interactions for the morphological analysis of
proteins, Loria et al. (20) on characterization of enzyme motions by relaxation
dispersion, Markwick et al. (21) on molecular dynamics simulations for the
analysis of NMR data, and Bouvignies et al. (22) on the utility of residual dipolar
couplings for studies of proteins in solution.
Many specific NMR techniques and methodologies have been reviewed in
the appropriate chapters in this book and are described in the next section under
different research themes. In the following paragraphs, brief summaries are given
of selected techniques and methodologies that either have general utility or are not
adequately covered in the next section.
An important technique is NMR imaging. Two recent reviews by Britton (23)
and Villaraza et al. (24) have appeared. NMR imaging studies have been reported
of penetration of water into mesoporous matrices (25) and hydrogels (26), and of
styrene in polyethylene in supercritical CO2 (27). In this book Peng et al. (28) have
reported a new class of 19F polymeric imaging agents, which provide high contrast
because the body does not contain a significant concentration of fluorine to give
a background signal. Therien-Aubin et al. (29) have employed NMR imaging to
monitor the release process of pharmaceutical tablets, and Baianu and Prisecaru
(30) to study the hydration of wheat grains.
Another fruitful area of research is the use of NMR for diffusion
measurements. In his chapter Macdonald (31) gives a good introduction to the
pulsed field gradient experiment used for diffusion studies. Pulsed field gradient
results are reported by Macdonald (31), Hou et al. (32), and Jespersen et al. (33)
5
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 in this book. Two recent reviews, by Walderhaug et al. (34) and Cozzolino et
al. (35), have also provided helpful complementary information, particularly on
different types of applications.
129Xe NMR can be employed to measure pore size and volume in porous

materials. In his article, Yoshimizu (36) has studied the microvoids in several
glassy polymers and correlated the results with excess free volume in the polymers.
Other examples in the recent literature include analysis of porosity in membranes
and oligopeptides (37) and microporous polymer networks (38).
Latex state NMR is reviewed by Kawahara (39); high resolution NMR
spectra for elastomers have been obtained through Brownian motion of the latex
dispersion. Similar results on a polymer slurry are reported by Thakur et al. (40),
except that 1H and magic angle spinning have been used.
The combination of rheology and NMR is a relatively newer development.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

A recent review has described the basics of rheo-NMR and some applications
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

(41). The article by Bohme and Scheler (42) in this book examines the relaxation
behavior of polymer melts under shear. In the literature a paper has also reported
shear deformation of polymer melts observed via 1H NMR (43). The combination
of electrophoresis and NMR is discussed by Hou et al. (32). Electrophoretic NMR
has also been reported recently by Scheler (44, 45).
A technique that has gained a fair amount of traction is field cycling (46).
This is a good way to determine the frequency or field dependence of NMR
relaxation times. There are two means of achieving field cycling. A sample can
be moved mechanically to positions with different magnetic flux densities (called
"sample shuttling"). Alternatively, special electronics and magnet design permit
fast switching of magnetic fields between 0 and 0.5 T in a few milliseconds
(commonly known as "fast field cycling"). Some examples in the recent literature
include segmental reorientation dynamics in polybutadiene melts (47), dynamics
of polybutadiene in nanoscopic confinement (48), diffusion (49) and dipole-dipole
interactions (50) in polymer melts. More examples have been given in a recent
NMR field cycling relaxometry conference in Turin (51).
In solution NMR, of course, the utility of multidimensional NMR is
well known (9, 52). Isotopic labeling can also be helpful, especially in
biomacromolecules (17, 53). Several techniques have been developed to speed up
the time needed for NMR (5457). An area that continues to show vitality is the
combination of chromatography and NMR. In the literature several publications
couple NMR with size exclusion chromatography (5860), and NMR with liquid
chromatography at critical conditions (59, 6163). A high temperature 10-mm
cryoprobe has been found to give higher sensitivity for 13C NMR of polyolefins
at sample temperatures of 120-135C (64). Other recent interesting articles
include stop-flow NMR for fast reactions (65), high pressure NMR (66), and
use of specially designed labeled chain transfer agents to facilitate analysis of
polymers made with controlled free radical polymerization (67). Additional new
developments are given in the reviews by Spiess (8) and Rinaldi (9).

6
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Major Research Themes
Polymers in Solution and in the Melt

It is well known that solution NMR can be utilized to identify unknown

polymers, to determine homopolymer tacticity, copolymer composition and
sequence, branching and chain ends, and irregularities such as head-to-head and
tail-to-tail structures (25). The number of papers published on microstructures
of polymers is large and beyond the scope of this article. The tutorial chapter
by Rinaldi (9) gives an excellent review of advanced solution techniques being
used. In their article (52), Rinaldi et al. have applied advanced 2D solution
methods to analyze fluoropolymers. Another way to carry out NMR analysis
of microstructure is via statistical models; this is reviewed by Cheng and Miri
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

(68). Acrylic latex copolymers have been studied by solution NMR and statistical
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

analysis by Beshah and Antrim (69). Solution NMR has also been employed to
study poly(-L-lysine) by Maeda et al. (70), carbohydrate-protein interactions by
Baianu and Prisecaru (30), and silk fibroin in fluorinated solvents by Nagano et
al. (71).
A different way to study polymer microstructure and dynamics is to observe
them in the melt (7275). In this book Hunt et al. (76) have included a molten
inclusion compound in their studies. Hubner et al. (77) have implemented
the analyses of molten polyolefins in an industrial R&D setting where sample
high throughput is important. Bohme and Scheler (42) have studied molten
polymers with rheo-NMR. Ries et al. (78) have investigated the NMR rescaling
approximation and its application to the relaxation of polymer melts above their
glass transition temperatures. The relaxation in polymer melts has also been
reported by Chavez et al. in the literature (79).

Structure-Dynamics-Properties Relationships

This is a common theme that runs throughout many of the solid state NMR
articles in this book. As Spiess pointed out in his review (8), NMR provides
unique information on the local conformation of polymers as well as time scale
and amplitude of rotational motions in bulk. In particular, the articles by Jespersen
et al. (33), Asakawa et al. (80), Litvinov (81), Miyoshi (82), and Magusin et al.
(83) illustrate the valuable information obtainable in different polymeric systems.
In addition, the papers by White et al. (84) and Pramoda et al. (85) describe
experimental methods that can be applied to probe the dynamics of polymer mixing
and polymer-filler interactions, respectively.
There is a large body of literature on solid state NMR studies of the dynamics
of different polymers. For example, Saalwachter (86) has provided a review of 1H
multiple quantum method for studies of chain dynamics and structural constraints
in polymeric soft materials. The dynamics of elastomers has been studied by the
multiple quantum method and computer simulations (87), time-domain NMR (88),
and maximum entropy method (89).

7
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Polymers with Precise Control of Structure
Recent advances in polymer synthesis permit more precise and well
defined polymeric molecules to be prepared. NMR has been used for polymer
characterization in many of these cases. Recent examples of polymers analyzed
by NMR include those made with single-site catalysts (9093), precision
polymers made via acyclic diene metathesis/hydrogenation by Wagener et al.
(9496), polymers made with controlled or living polymerizations (67, 9799),
and polymers made via enzyme catalysis (100102). In this book Alam et al.
(103) have studied poly(ethylene-co-acrylic acid) with precise distribution of
pendant carboxylic acid group along the polymer backbone and its ionomer with
partial neutralization of the carboxylic acid with Zn2+ and Li+; the results have
been compared with ionomers with random spacing of the carboxylic groups.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

Another example is the computer-assisted approach devised by Cheng and Miri

(68) for the analysis of NMR spectral intensities for polyolefins made with
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

single-site metallocene catalysts.

Blends, Composites, and Nanostructures

This category includes polyblends, composites, organic-inorganic hybrids,
supramolecular assemblies, and nanostructures. Reviews have been reported
of solid state NMR studies of multicomponent organic/inorganic materials
by Geppi et al. (104), polymer/clay composites by Olejniczak et al. (105),
and polymeric and supramolecular systems by Brown (106). Reviews have
also been given by White and Wachowicz (107) on polyblends, and by White
et al. (84) on experimental methods that have been developed to probe the
dynamics and thermodynamics of mixing in amorphous blends. Examples
of polyblends are given in this book by White et al. (84) and Asano (108).
Yoshimizu (36) studied the free volume in a polymer blend with 129Xe NMR.
Examples of polyimide-graphite composites are shown by Pramoda et al. (85),
and organic-inorganic hybrids by Jespersen et al. (33).
Nanochemistry is a hot topic today. Silica nanoparticles are reported
in the papers by Jespersen et al. (33) and Thakur et al. (40). The former
paper uses NMR relaxation and PFG to measure the dynamics of this
nanoscale ionic material prepared from silica nanoparticles. The latter paper
employs magic angle spinning (MAS) to identify organic moieties adsorbed or
covalently bonded to silica nanoparticles. Asano (108) has studied poly(vinyl
isobutyl ether)/poly(-L-lysine)/saponite-clay nanocomposites. In the recent
literature, NMR studies have also been reported for several blends (109, 110),
nanocomposites (111114), polyelectrolyte complexes (115), and polyelectrolyte
multilayer systems (116).

Micelles, Interfaces, and Confined Environments

There is continued interest in polymer micelles, interfaces, and polymers
in confined environments. A review has appeared on NMR studies of block
copolymer micelles (117). In this book, Macdonald (31) introduces bicelles as
8
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 membrane mimetics and confinement media. He has studied the diffusion of
poly(ethylene glycol) confined in bicelles, using the pulsed field gradient (PFG)
NMR method. Hunt et al. (76) have observed the conformation and motions of
a hydrocarbon model compound in -cyclodextrin using chemical shifts and T1
information. In the literature, NMR has been used to study interfacial polymers in
carbon nanotube-based nanocomposites (118), surfactant layers in polymer-clay
nanocomposites (119), and polymer adsorbed on silica (120) and inorganic
membrane (121). Studies have also been made of ultra-thin polymer films on
a substrate; NMR measurements permit information to be obtained on polymer
dynamics (122) and polymer mixing (123) under confined environments.

Biopolymers, Biochemistry, and Bio-inspired Chemistry

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

As shown in Table 1, NMR has been utilized extensively for biochemistry

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

and biophysics, and (to less extents) plant science, food, and agriculture. The
natural polymers (such as proteins, polysaccharides, lipids, and related materials
found in nature) have been increasingly exploited for industrial and health related
applications. In this book microbial poly(-L-lysine) has been studied by Maeda
et al. (70), and blends and composites of poly(-L-lysine) by Asano (108).
Nagano et al. (71) have determined the structure of silk fibroin by NMR, and
designed and produced recombinant proteins from transgenic silkworm for use as
biomaterials. Nakazawa et al. (124) have conducted a detailed multinuclear NMR
study of amyloid -peptide (the major component of plaques in Alzheimers
disease), particularly its interaction with lipid bilayers and ganglioside. In the
literature, there are many papers on biomimetic or bio-inspired polymers, most
of which used NMR for characterization (e.g., refs. (125127)). Several articles
have reported the encapsulation of proteins with reverse micelles (128130) and
chaperonin (131) as a method to enhance biophysical studies of proteins.
NMR is a frequent tool to study polysaccharides. Recent articles on NMR
of polysaccharides include chondroitin sulfate (132), alginate (133), chitin and
chitosan (134), tamarind seed polysaccharide and larch arabinogalactan (135).
In this book, Baianu and Prisecaru (30) have applied solution NMR and T2 to
investigate interactions of wheat gliadin with sucrose in aqueous solutions and
wheat gluten-glucomannan interactions in hydrated gels.

Application-Driven Polymer Designs

Several topics are of current interest, such as polymers for energy applications,
for sensors and information processors, for drug delivery and biomedical devices,
and stimulus-responsive polymers, gels and networks with defined function
and control. For fuel cells, batteries, and ionic transducers, ionomers are often
involved. The benefit of NMR to characterize ionomers is illustrated by Hou
et al. (32) (using 2H and PFG NMR) and Alam et al.. (103) (using 1D and 2D
1H MAS). In the literature, reviews have been reported of the different NMR

techniques used for polyelectrolytes (34) and polymer electrolyte membranes for
fuel cells (136138). NMR has been employed to study polyelectrolyte systems
containing poly(ethylene oxide) and alumina (139).
9
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 An example of -conjugated polymers for possible electronic devices is given
by Asakawa et al. (80). A solution NMR study of a conjugated polymer has
been reported by Rahman et al. (140). A solid state NMR study of the structures
of materials involved in organic light-emitting diodes has been reported by Kaji
(141).
A study of coordination number of borate in poly(vinyl alcohol) gel is reported
by Yamada (142), using 1D and 2D 11B NMR. The utility of NMR to study phase
transitions in aqueous polymer solutions and gels has been reviewed by Spevacek
(143). T1 measurements have been engaged to investigate the phase structures
of poly(vinylidene fluoride) gels (144). A novel NMR approach for the studies of
gelation in flexible polymer systems has been devised by Saalwachter et al. (145).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

Acknowledgments
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Thanks are due to Suhad Wojkowski at USDA Southern Regional Research

Center for her help on the literature search.
Mention of trade names or commercial products in this publication is
solely for the purpose of providing specific information and does not imply
recommendation or endorsement by the U.S. Department of Agriculture. USDA
is an equal opportunity provider and employer. DuPont Company and Tokyo
University of Agriculture and Technology are also equal opportunity providers
and employers.

References
1. Recent books in the past 10 years include refs. 2-5. Recent reviews are cited
in the appropriate sections in this article.
2. Cheng, H. N.; English, A. D. (Eds.) NMR spectroscopy of polymers in
solution and in the solid state; ACS Symposium Series No. 834; American
Chemical Society: Washington, DC, 2002.
3. Mirau, P. A. A practical guide to understanding the NMR of polymers; Wiley:
Hoboken, NJ, 2004.
4. Kitayama, T.; Hatada, K. NMR spectroscopy of polymers; Springer: New
York, 2004.
5. Cheng, H. N. Structural studies of polymers by solution NMR; RAPRA
Review Reports, Vol. 11, Number 5, Rapra: Shewsbury, UK, 2001.
6. Thomson Reuters, New York, NY. http://www.webofknowledge.com/.
7. Interdisciplinary Globally-Leading Polymer Science & Engineering, NSF
Workshop, 2007. www.nsf.gov/mps/dmr/NSF_Polymer_Workshop.pdf.
8. Spiess, H. W. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 2.
9. Rinaldi, P. L. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 3.
10. Chen, Q; Kurosu, H. Ann. Rep. NMR Spectrosc. 2007, 61, 6124761281.
11. Paasch, S; Brunner, E. Anal. Bioanal. Chem. 2010, 398, 23512362.
12. Bockmann, A. Magn. Reson. Chem. 2007, 45, S24S31.
10
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 13. Spiess, H. W. Macromolecules 2010, 43, 54795491.
14. Borisov, A. S.; Hazendonk, P.; Hayes, P. G. J. Inorg. Organomet. Polym.
Mater. 2010, 20, 183212.
15. Domenici, V. Pure Appl. Chem. 2011, 83 (Special Issue), 6794.
16. Felli, I. C.; Brutscher, B. ChemPhysChem 2009, 10 (9-10), 13561368.
17. Foster, M. P.; McElroy, C. A.; Amero, C. D. Biochemistry 2007, 46 (2),
331340.
18. Mittermaier, A. K.; Kay, L. E. Trends Biochem. Sci. 2009, 4 (12), 601611.
19. Tate, S. I. Anal. Sci. 2008, 24 (1), 3950.
20. Loria, J. P.; Berlow, R. B.; Watt, E. D. Acct. Chem. Res. 2008, 41 (2),
214221.
21. Markwick, P. R. L.; Malliavin, T.; Nilges, M. PLOS Comput. Biol. 2008, 4
(9): Art No. e1000168 2008
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

22. Bouvignies, G.; Markwick, P. R. L.; Blackledge, M. ChemPhysChem 2007,

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

8, 19011909.
23. Britton, M. M. Chem. Soc. Rev. 2010, 39, 40364043.
24. Villaraza, A. J. L.; Bumb, A.; Brechbiel, M. W. Chem. Rev. 2010, 110,
29212959.
25. Le Feunteun, S.; Diat, O.; Guillermo, A.; Poulesquen, A.; Podor, R. Magn.
Reson. Imaging 2011, 29 (3), 443455.
26. Zainuddin; Strounina, E. V.; Hill, D. J. T.; Whittaker, A. K. Polym. Int. 2010,
59 (11), (Special Issue), 1520-1525.
27. Thurecht, K. J.; Hill, D.; Whittaker, A. K. Macromol. Chem. Phys. 2006,
207, 15391545.
28. Peng, H.; Thurecht, K.; Hsu, S.; Blakey, I.; Squires, O.; Kurniawan,
N.; Rose, S.; Whittaker, A. K. ACS Symposium Series 1077; American
Chemical Society: Washington, DC, 2011; Chapter 28.
29. Therien-Aubin, H.; Wang, Y. J.; Zhu, X. X. ACS Symposium Series 1077;
American Chemical Society: Washington, DC, 2011; Chapter 27.
30. Baianu, I. C.; Prisecaru, V. ACS Symposium Series 1077; American
Chemical Society: Washington, DC, 2011; Chapter 20.
31. Macdonald, P. M. ACS Symposium Series 1077; American Chemical
Society: Washington, DC, 2011; Chapter 14.
32. Hou, J.; Li, J.; Wilmsmeyer, K. G.; Zhang, Z.; Madsen, L. A. ACS
Symposium Series 1077; American Chemical Society: Washington, DC,
2011; Chapter 15.
33. Jespersen, M. L.; Mirau, P. A.; von Meerwall, E.; Vaia, R. A.; Rodriguez, R.;
Fernandes, N. J.; Giannelis, E. P. ACS Symposium Series 1077; American
Chemical Society: Washington, DC, 2011; Chapter 9.
34. Walderhaug, H.; Soderman, O.; Topgaard, D. Prog. Nucl. Magn. Reson.
Spectrosc. 2010, 56 (4), 406425.
35. Cozzolino, S.; Sanna, M. G.; Valentini, M. Magnetic Resonance Chem. 2008,
46, S16S23.
36. Yoshimizu, H. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 31.
37. Ripmeester, J.; Moudrakovski, I.; Soldatov, D.; Pawsey, S.; Anedda, R. ACS
Polym. Prepr. 2008, 49 (1), 653654.

11
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 38. Weber, J.; Schmidt, J.; Thomas, A.; Bohlmann, W. Langmuir 2010, 26 (19),
1565015656.
39. Kawahara, S. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 29.
40. Thakur, K. A. M.; McCormick, M.; Thompson, W. L.; Cao, C.; Schultz, W.
J. ACS Symposium Series 1077; American Chemical Society: Washington,
DC, 2011; Chapter 30.
41. Callaghan, P. T. Rheol. Acta 2008, 47 (3), 243255.
42. Bhme, U.; Scheler, U. ACS Symposium Series 1077; American Chemical
Society: Washington, DC, 2011; Chapter 26.
43. Douglass, B. S.; Cormier, R. J.; Callaghan, P. T. Phys. Rev. E 2007, 75(4),
Article No. 041802.
44. Bohme, U.; Scheler, U. ACS Polym. Prepr. 2008, 49 (1), 670671.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

45. Scheler, U. Curr. Opin. Colloid Interface Sci. 2009, 14 (3), 212215.
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

46. For example, Kimmich, R.; Anoardo, E. Prog. NMR Spectrosc. 2004, 44,
257-320.
47. Herrmann, A.; Novikov, V. N.; Rossler, E. A. Macromolecules 2009, 42 (6),
20632068.
48. Hofmann, M.; Herrmann, A.; Ok, S.; Franz, C.; Kruk, D.; Saalwachter, K.;
Steinhart, M.; Rossler, E. A. Macromolecules 2011, 44 (11), 40174021.
49. Kehr, M.; Fatkullin, N.; Kimmich, R. J. Chem. Phys. 2007, 126 (9), Article
No. 094903.
50. Fatkullin, N.; Gubaidullin, A.; Stapf, S. J. Chem. Phys. 2010, 132 (9), Article
No. 094903.
51. Seventh Conference on Field Cycling NMR Relaxometry, June 24, 2011.
www.ffcrelax.com/ffcrelax/home.php.
52. Rinaldi, P. L.; Baiagern, S.; Fox, P.; Howell, J. L.; Li, L.; Li, X.; Lyons, D. F.;
McCord, E. F.; Sahoo, S. K.; Twum, E. B.; Wyzgoski, F. J. ACS Symposium
Series 1077; American Chemical Society: Washington, DC, 2011; Chapter
21.
53. Liu, D. S.; Xu, R.; Cowburn, D. Methods Enzymol. 2009, 162, 151175.
54. Schanda, P. Prog. in NMR Spectrosc. 2009, 55 (3), 238265.
55. Vitorge, B.; Bodenhausen, G.; Pelupessy, P. J. Magnetic Resonance 2010,
207, 149152.
56. Springer F.; Steidle G.; Martirosian P.; Syha R.; Claussen C. D.; Schick F.
Invest. Radiol. 2011, online, PMID: 21577126.
57. Deschamps, M.; Campbell, I. D. J. Magnetic Resonance 2006, 178, 206211.
58. Cudaj, M.; Guthausen, G.; Hofe, T.; Wilhelm, M. Macromol. Rapid Comm.
2011, 32 (8), 665670.
59. Hiller, W.; Sinha, P.; Hehn, M.; Pasch, H.; Hofe, T. Macromolecules 2011,
44 (6), 13111318.
60. Pasch, H.; Heinz, L. C.; Macko, T.; Hiller, W. Pure Appl. Chem. 2008, 80
(8), 17471762.
61. Sinha, P.; Hiller, W.; Pasch, H. J. Separation Sci. 2010, 33(22), Special Issue,
3494-3500.
62. Hiller, W.; Sinha, P.; Pasch, H. Macromol. Chem. Phys. 2009, 210, 605613.

12
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 63. Hiller, W.; Sinha, P.; Pasch, H. Macromol. Chem. Phys. 2007, 208 (18),
19651978.
64. Zhou, Z; Kummerle, R.; Stevens, J. C.; Redwine, D.; He, Y. Y.; Qiu, X. H.;
Cong, R. J.; Klosin, J.; Montanez, N.; Roof, G. J. Magnetic Resonance 2009,
200 (2), 328333.
65. Christianson, M. D.; Landis, C. R. Concept Magn. Reson., Part A 2007, 30A
(4), 165183.
66. Whittaker, A.; Blakey, I.; Thurecht, K.; Squires, O., ; Varcoe, K. ACS Polym.
Prepr. 2008, 49 (1), 678679.
67. Pach, M.; Zehm, D.; Lange, M.; Dambowsky, I.; Weiss, J.; Laschewsky, A.
J. Amer. Chem. Soc. 2010, 132 (25), 87578765.
68. Cheng, H. N.; Miri, M. J. ACS Symposium Series 1077; American Chemical
Society: Washington, DC, 2011; Chapter 22.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

69. Beshah, K.; Antrim, R. F. ACS Symposium Series 1077; American Chemical
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Society: Washington, DC, 2011; Chapter 23.

70. Maeda, S.; Sasaki, C.; Kunimoto, K.-K. ACS Symposium Series 1077;
American Chemical Society: Washington, DC, 2011; Chapter 19.
71. Nagano, A.; Suzuki, Y.; Nakazawa, Y.; Gerig, J. T.; Asakura, T. ACS
Symposium Series 1077; American Chemical Society: Washington, DC,
2011; Chapter 17.
72. Schaler, K.; Ostas, E.; Schroter, K.; Thurn-Albrecht, T.; Binder, W. H.;
Saalwachter, K. Macromolecules 2011, 44 (8), 27432754.
73. Castignolles, P.; Graf, R.; Parkinson, M.; Wilhelm, M.; Gaborieau, M.
Polymer 2009, 50 (11), 23732383.
74. Hu, W.; Hagihara, H.; Miyoshi, T. Macromolecules 2007, 40, 35053509.
75. Kimmich, R.; Fatkullin, N. ACS Polym. Prepr. 2008, 49 (1), 689690.
76. Hunt, M. A.; Villar-Rodil, S.; Gomez-Fatou, M. A.; Shin, I. D.; Schilling, F.
C.; Tonelli, A. E. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 16.
77. Hubner, G.; Fonseca, I.; Parkinson, M. ACS Symposium Series 1077;
American Chemical Society: Washington, DC, 2011; Chapter 24.
78. Ries, M. E.; Bansal, A.; Brereton, M. G. ACS Symposium Series 1077;
American Chemical Society: Washington, DC, 2011; Chapter 25.
79. Chavez, F. V.; Saalwachter, K. Macromolecules 2011, 44 (6), 15601569.
80. Asakawa, N.; Inoue, Y.; Yamamoto, T.; Shimizu, R.; Tansho, M.; Yazawa,
K. ACS Symposium Series 1077; American Chemical Society: Washington,
DC, 2011; Chapter 10.
81. Litvinov, V. M. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 11.
82. Miyoshi, T. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 12.
83. Magusin, P. C. M. M.; Orza, R. A.; Litvinov, V. M.; van Duin, M.;
Saalwchter, K. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 13.
84. White, J. L.; Wachowicz, M.; Gill, L.; Damron, J.; Wolak-Dinsmore, J.
ACS Symposium Series 1077; American Chemical Society: Washington,
DC, 2011; Chapter 4.

13
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 85. Pramoda, K. P.; Mya, K. Y.; Lin, T. T.; Wang, J.; He, C. B. ACS Symposium
Series 1077; American Chemical Society: Washington, DC, 2011; Chapter
6.
86. Saalwachter, K. Prog. NMR Spectrosc. 2007, 51 (1), 135.
87. Saalwachter, K; Sommer, J. U. Macromol. Rapid Comm. 2007, 28,
14551465.
88. Chavez, F. V.; Saalwachter, K. Macromolecules 2011, 44 (6), 15491559.
89. Aguilera-Mercado, B. M.; Cohen, C.; Escobedo, F. A. Macromolecules 2009,
42 (22), 88898898.
90. Galland, G. B.; Da Silva, L. F.; Nicolini, A. J. Polym. Sci., Part A:
Polym.Chem. 2005, 43, 47444753.
91. Landis, C. R.; Christianson, M. D. Proc. Nat. Acad. Sci. U.S.A. 2006, 103,
1534915354.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

92. Cherian, A. E.; Lobkovsky, E. B.; Coates, G. W. Macromolecules 2005, 38,

Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

62596268.
93. Dijkstra, P. J.; Du, H.; Feijen, J. Polym. Chem. 2011, 2, 520527.
94. Boz, E.; Nemeth, A. J.; Wagener, K. B.; Jeon, K.; Smith, R.; Nazirov, F.;
Bockstaller, M. R.; Alamo, R. G. Macromolecules 2008, 41, 16471653.
95. Boz, E.; Nemeth, A. J.; Alamo, R. G.; Wagener, K. B. Adv. Synth. Catal.
2007, 349, 137141.
96. Boz, E.; Nemeth, A. J.; Ghiviriga, I.; Jeon, K.; Alamo, R. G.; Wagener, K.
B. Macromolecules 2007, 40, 65456551.
97. Moad, G.; Chong, Y. K.; Mulder, R.; Rizzardo, E.; Thang, S. H. ACS
Symposium Series 2009, 1024, 318.
98. Klumperman, B.; McLeary, J. B.; van den Dungen, E. T. A.; Pound, G.
Macromol. Symp. 2007, 248, 141149.
99. Brar, A. S.; Saini, T. J. Polymer Sci., Part A: Polym. Chem. 2006, 44,
29552971.
100. Puskas, J. E.; Seo, K. S.; Sen, M. Y. Eur. Polym. J. 2011, 47, 524534.
101. Padovani, M.; Hilker, I.; Duxbury, C. J.; Heise, A. Macromolecules 2008,
41, 24392444.
102. Mita, N.; Rawaki, S.; Uyama, H.; Kobayashi, S. Bull. Chem. Soc. Jpn. 2004,
77, 15231527.
103. Alam, T. M.; Jenkins, J. E.; Seitz, M. E.; Buitrago, C. F.; Winey, K. I.;
Opper, K.; Baughman, T. W.; Wagner, K. B. ACS Symposium Series 1077;
American Chemical Society: Washington, DC, 2011; Chapter 7.
104. Geppi, M; Borsacchi, S; Mollica, G; Veracini, C. A. Appl. Spectrosc. Rev.
2009, 44 (1), 189.
105. Olejniczak, S; Kazmierski, S; Pallathadka, PK; Potrzebowski, M. Polimery
2007, 52, 713721.
106. Brown, S. P. Macromol. Rapid Comm. 2009, 30 (9-10), 688716.
107. White, J. L.; Wachowicz, M. Ann. Rep. NMR Spectrosc. 2008, 64, 189209.
108. Asano, A. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 5.
109. Ediger, M.; Zhao, J. ACS Polym. Prepr. 2008, 49 (1), 706.
110. White, J.; Wachowicz, M.; Gill, L. ACS Polym. Prepr. 2008, 49 (1), 711.

14
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 111. Foston, M.; Leisen, J.; Beckham, H. ACS Polym. Prepr. 2008, 49 (1),
665666.
112. Papon, A.; Saalwachter, K.; Schaler, K.; Guy, L.; Lequeux, F.; Montes, H.
Macromolecules 2011, 44 (4), 913922.
113. Liu, H. H.; Chaudhary, D.; Yusa, S.; Tade, M. O. Carbohydrate Polym. 2011,
83 (4), 15911597.
114. Powers, D. S.; Vaia, R. A.; Koerner, H.; Serres, J.; Mirau, P. A.
Macromolecules 2008, 41, 42904295.
115. Causemann, S.; Schonhoff, M.; Eckert, H. Phys. Chem. Chem. Phys. 2011,
13 (19), 89678976.
116. Fortier-McGill, B.; Reven, L. Macromolecules 2009, 42, 247254.
117. Walderhaug, H.; Soderman, O. Curr. Opin. Colloid Interface Sci. 2009, 14
(3), 171177.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001

118. Mirau, P.; Lyons, M.; Powers, D.; Koerner, H.; Vaia, R. ACS Polym. Prepr.
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

2008, 49 (1), 704705.

119. Jeschke, G. ACS Polym. Prepr. 2008, 49 (1), 737738.
120. Metin, B.; Blum, F. D. J. Chem. Phys. 2006, 125, 0547079.
121. Okuom, M. O.; Metin, B.; Blum, F. D. Langmuir 2008, 24, 25392544.
122. Stapf, S.; Ayalur-Karunakaran, S.; Blumich, B. ACS Polym. Prepr. 2008, 49
(1), 661662.
123. VanderHart, D.; Prabhu, V.; Lavery, K.; Lin, E.; Dennis, C. ACS Polym.
Prepr. 2008, 49 (1), 739740.
124. Nakazawa, Y.; Suzuki, Y.; Saito, H.; Asakura, T. ACS Symposium Series
1077; American Chemical Society: Washington, DC, 2011; Chapter 18.
125. Laschewsky, A.; Marsat, J.-N.; Skrabania, S.; von Berlepsch, H.; Bottcher, C.
Macromol. Chem. Phys. 2010, 211, 215221.
126. Guan, Z. Polym. Int.l 2007, 56, 467473.
127. Kentgens, A. P. M.; van Eck, E.; Gowda, C.; Nolte, R.; Cornelissen, J.;
Rowan, A. E. ACS Polym. Prepr. 2008, 49 (1), 693694.
128. Flynn, P. F.; Simorellis, A. K.; Van Horn, W. D. Ann. Rep. NMR Spectrosc.
2007, 62, 179219.
129. Nucci, N. V.; Marques, B. S.; Bedard, S.; Dogan, J.; Gledhill, J. M.;
Moorman, V. R.; Peterson, R. W.; Valentine, K. G.; Wang, A. L.; Wand, A.
J. J. Biomol. NMR 2011, online, PMID: 21748265.
130. Peterson, R. W.; Pometun, M. S.; Shi, Z.; Wand, A. J. Protein Sci. 2005, 14,
29192921.
131. Tanaka, S.; Kawata, Y.; Otting, G.; Dixon, N. E.; Matsuzaki, K.; Hoshino, M.
Biochim. Biophys. Acta 2010, 1804, 866871.
132. Huckerby, T.; Lauder, R. ACS Polym. Prepr. 2008, 49 (1), 702703.
133. Bernin, D.; Goudappel, G. J.; van Ruijven, M.; Altskar, A.; Strom, A.;
Rudemo, M.; Hermansson, A. M.; Nyden, M. Soft Matter 2011, 7 (12),
57115716.
134. Kasaai, M. R. Carbohydrate Polym. 2010, 79, 801810.
135. Uccello-Barretta, G.; Nazzi, S.; Balzano, F.; Sanso, M. Int. J. Pharm. 2011,
406 (1-2), 7883.
136. Suarez, S; Greenbaum, S. Chem. Rec. 2010, 10 (6), 377393.
137. Ye, G.; Goward, G. ACS Polym. Prepr. 2008, 49 (1), 714715.

15
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 138. Bai, Z.; Shumaker, J. A.; Houtz, M. D.; Mirau, P. A.; Dang, T. D. Polymer
2009, 50, 14631469.
139. Ekanayake, P; Dissanayake, M. A. K. L. J. Solid State Electrochem. 2009,
13 (12), 18251829.
140. Rahman, M. H.; Chen, H. L.; Chen, S. A.; Chu, P. P. J. J. Chin. Chem. Soc.
2010, 57 (3) Special Issue, 490-495.
141. Kaji, H. ACS Polym. Prepr. 2008, 49 (1), 736.
142. Yamada, K. ACS Symposium Series 1077; American Chemical Society:
Washington, DC, 2011; Chapter 8.
143. Spevacek, J. Curr. Opin. Colloid Interface Sci. 2009, 14 (3), 184191.
144. Hubbard, H.; Ward, I. ACS Polym. Prepr. 2008, 49 (1), 709710.
145. Saalwachter, K.; Gottlieb, M.; Liu, R. G.; Oppermann, W. Macromolecules
2007, 40 (5), 15551561.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch001
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

16
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 2

Overview of NMR of Bulk Polymers

Hans Wolfgang Spiess*
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

Max-Planck-Institute for Polymer Research, P. O. Box 3148,

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

D-55021 Mainz, Germany

*E-mail: [email protected]

The favorable properties of bulk polymers are primarily due

to their molecular structure but also heavily depend on their
organization in bulk and their molecular dynamics. Solid
state NMR techniques provide unique information on the local
conformation of macromolecules as well as time scale and
amplitude of rotational motions in bulk. Moreover, translational
motions of the chains on mesoscopic length scales can be
elucidated by advanced NMR techniques. In order to obtain this
information, homo- and heteronuclear dipole-dipole couplings,
involving 1H and heteronuclei such as 13C and 15N, isotropic
as well as anisotropic chemical shifts, in particular of 13C,
and quadrupole coupling of 2H are employed. Experiments
can be performed on static samples and under magic angle
spinning (MAS). Double quantum and multidimensional NMR
techniques are especially informative. The methods will
be briefly introduced and illustrated by recent experimental
examples from different fields of polymer science, such as
chain microstructure, defects and dynamics, chain packing
in polymers for plastic electronics, proton motion in proton
conductors and packing of macrocycles in nanotubes.

Introduction
Precise knowledge of structure and dynamics of macromolecules of
well-defined architectures is key when tailoring them for specific functions. For
instance, such diverse technological challenges as efficient fuel cells, photonic
sensors and devices or gene delivery systems all require transport of electrones,
holes, protons or other ions. Likewise, the properties of conventional polymers

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 can be substantially improved by controlling, e.g., their microstructure and the
processing conditions. Properties of macromolecular systems critically depend on
the arrangement of the building blocks relative to each other and their mobility on
different length- and time scales. Therefore, success in polymer science requires
development of characterization techniques that are able to provide information
on these aspects. Scattering of light, X-rays and neutrons, the various forms of
microscopy as well as mechanical and dielectric relaxation are well-established
in polymer science (1).
NMR spectroscopy, on the other hand, is often considered just as an analytical
tool, monitoring the various synthetic steps, which eventually lead to the new
polymers or supramolecular systems. High-resolution NMR in one- and two
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

dimensions (2) provides a number of important parameters, which are specific to

macromolecules. Examples include stereochemical configuration, geometrical
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

isomerism, and regioregularity.

More recently, the rapid development of solid state NMR has substantially
broadened the application of NMR to polymers (3). In particular, molecular and
collective motions can be characterized in unprecedented detail. The new NMR
methods use concepts well established in X-ray and neutron scattering, yet on
much longer timescales. These techniques provide structural information with
atomic resolution even for systems that lack crystalline order in the traditional
sense (4). Moreover, solid state NMR concepts can also be applied to partially
mobile systems such as liquid crystals. Further examples are highly viscous
materials such as polymer melts in the vicinity of the glass transition or elastomers.
This chapter is intended as an overview rather than a detailed description of
the different possibilities NMR provides. It is based on previous reviews of the
author (1, 4); for further details we refer to recent reviews of the subject (57).

NMR Background
NMR spectroscopy is remarkably versatile and, therefore, is widely applied
in many fields of science, in particular physics, chemistry, biology, materials
science, and in medicine. This form of spectroscopy combines a number of
features that makes it an almost ideal tool: NMR is non-destructive, sample
preparation is easy, the possibility to observe different nuclei and isotopes
provides extreme structural selectivity. Moreover, dynamic features can be
studied over many decades of characteristic times, from picoseconds to minutes,
and length scales from interatomic distances in the 100 pm range up to a meter or
so in NMR imaging.
The wealth of information accessible by NMR spectroscopy results from the
fact that a variety of interactions of the nuclear spins with their surroundings can
be exploited. The structure of polymers and other supramolecular systems can
be elucidated along several routes. The chemical shift provides the basis of site
selectivity. Geometric parameters such as internuclear distances as well as dihedral
angles are encoded in the dipole-dipole and the J-coupling. Of the great variety
of molecular motions possible in polymers, rotations have the most pronounced
effects on NMR spectra and relaxation parameters, because the spin interactions
18
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 are angular-dependent. But conformational dynamics and translational motions
can be tackled as well. The anisotropic nuclear spin interactions also offer a means
to probe the alignment of residues in partially ordered systems such as drawn fibers
or oriented liquid crystals. Therefore, a brief outline of these anisotropic spin
interactions is given here.

Anisotropic Spin Interactions

As noted above, the information that can be extracted from solid state NMR
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

spectra is encoded via spin interactions such as the chemical shift, the quadrupolar
interaction, homo- and heteronuclear dipolar interactions, as well as J-couplings.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

For convenience, these couplings and the information that is accessible through
them are listed in Table 1.
A common characteristic of the relevant spin interactions is that they
are anisotropic and can be described by second-rank tensors. The resulting
orientation-dependent NMR frequency for each of the interactions is alike and of
the following form:

Here L is the Larmor frequency, iso is the isotropic chemical shift and
the other terms reflect the deviations due to angular-dependent contributions.
The strength of the anisotropic interaction is specified by , and (0 1) is
the asymmetry parameter which describes the deviation from axial symmetry.
The polar angles , , relate the orientation of the principal axes system of the
interaction tensor with the external magnetic field.
The most important interaction with respect to chemical information is the
chemical shift. It results from the shielding of the magnetic field at the position
of the nucleus by the electrons. Major advances have been achieved recently
in quantum chemical calculations of this parameter in bulk at different levels of
precision and calculational cost (8). The NMR frequency is thus shifted by an
isotropic contribution iso and by angular dependent terms. Assuming an equal
probability for all directions the powder average can be calculated and a broad
powder spectrum is obtained, which reflects the chemical-shift anisotropy. For
the case of an axially symmetric chemical-shift tensor, is zero and the angular-
dependent term simplifies to (3cos2-1)/2. Usually the chemical shift in polymers
spans about 10 ppm for 1H, 200 ppm for 13C, and 400 ppm for 15N nuclei, see Fig.
1.

19
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. Interactions of nuclei in NMR providing access to structure and
dynamics of polymers.
Interaction Electronic Geometry Nuclei Structure Dynamics
Structure
Chemical Yes Intrinsic and 1H, 13C, Confor- Conforma-
shift orientation 15N, 19F, mation, tional tran-
29Si, 31P through- sitions, ro-
space prox- tational mo-
imities tions
Dipole- No Internuclear 1H, 13C, Through- Transla-
dipole distance and 15N, 19F, space tional and
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

coupling orientation 29Si, 31P distances rotational

motions
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

J-Coupling Yes Intrinsic, 1H, 13C, Conforma- Conforma-

internuclear 15N, 19F, tion and tional tran-
distance and 29Si, 31P intergroup sitions, ro-
orientation binding tational mo-
tions
Quadrupole Yes Intrinsic and 2H, 14N, Electronic Rotational
Coupling orientation 17O, environ- motions
23Na, ment, chem-
27Al ical bonding

For abundant nuclei with spin , the spectrum is often dominated by

heteronuclear or homonuclear dipole-dipole interactions, i.e. the interactions
between the magnetic moments of two neighboring spins. In this case there is no
isotropic contribution and the interaction is axially symmetric, =0, so that Eq. (1)
simplifies correspondingly. For a two-spin system the interaction Hamiltonian,
which defines the frequencies, reads

where r12 is the magnitude of the vector connecting the two spins, is the angle
of this vector to the magnetic field. The Ii are spin operators and the i (i=1,2)
are the magnetogyric ratios of the spins. That is, the strength of the resulting
line splitting depends strongly on the distance between the two spins, so that
distance information can be extracted from such spectra. Homonuclear interaction
(equivalent spins with 1 = 2 = and heteronuclear interaction (non-equivalent
spins with 1 2) have to be distinguished. In the latter case, the flip-flop term
which is part of the product I1 I2 in Eq. (2) can be neglected. For a powder
sample one has again to take into account all angles and thus obtains the
so-called Pake spectrum with a considerable anisotropic line-broadening of up
to 50 kHz for homonuclear 1H-1H and up to 25 kHz for heteronuclear 13C-1H or
15N-1H dipolar interaction. Since the dipole-dipole coupling is a through-space

interaction, however, we have in principle to evaluate the sum over all possible
20
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 pair interactions. This and the presence of molecular motions lead to considerable
complications. Therefore, in practice a relatively structureless line-shape rather
than a Pake spectrum but is observed.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 1. Chemical shift ranges of organic compounds. (3)

In the case of a deuterated sample (2H, spin I=1), the spectra are usually
dominated by the quadrupole interaction, that is, the coupling of the nuclear
quadrupole moment with the electric field gradient of the C-2H bond. For
deuterons in C-2H bonds this can lead to a splitting of about 250 kHz. As in the
case of the dipole-dipole interaction, a Pake spectrum is obtained for a powder
sample. The z-principal axis of the quadrupolar interaction is oriented along
the bond axis. This makes deuteron NMR particularly useful for studies of
segmental orientations and molecular dynamics (reorientation), where the line
shapes strongly depend on the orientation of an oriented sample such as a polymer
fiber, or on the geometry of rotational motions generating averaged spectra, Fig.
2a. It is important to note that in the fast limit where the averaging is complete,
the anisotropic interaction can still be described by an expression as Eq. (1)
with averaged coupling parameters and angles. Even if an interaction is axially
symmetric in the static case, it can become asymmetric upon motional averaging
(3) and Fig. 2a.
In sufficiently mobile (i.e. liquid-like) systems where the residues undergo
rapid isotropic thermal motions, the anisotropy is typically averaged out
completely, leaving only the isotropic contributions.

21
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Manipulation of Spin Interactions

The rich information content of solid state spectra makes them difficult to
handle, in particular if more than one of the interactions introduced above has
to be taken into account. NMR methodology, however, provides the possibility
to decouple and recouple spin interactions nearly as desired (see (9) for a
comprehensive introduction). Moreover, the different sources of information
can be separated and correlated using the two-dimensional techniques discussed
below.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 2. Solid state NMR spectra; a) static, b) under MAS.

The most prominent example of a technique for decoupling or line-narrowing

is magic-angle spinning (MAS). Here, the angular dependent part of the
interactions is modulated by rapidly spinning the sample around an axis inclined at
an angle to the magnetic field. If the spinning axis is chosen along the so-called
magic-angle m=54.7, the relevant scaling factor (3cos2-1)/2 becomes zero
and the anisotropic part of the interaction vanishes, see Fig. 1. Today, fast MAS
with rotation frequencies reaching 70 kHz is possible. This paves the way for easy
access to high spectral resolution in solid state 1H-NMR and, more important,
largely simplifies the dipolar coupled network that prevents the dipole-dipole
coupling between 1H from being used for specific structural investigations in
non-spinning samples.
In addition, it is possible to manipulate the spin interactions by pulse
techniques (59) These act on the spin operators in the corresponding
Hamiltonians rather than on the geometric part. Depending on the applied
pulse sequence, a given spin interaction can be switched on and off in order to
22
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 discriminate the different contributions to the desired information. For instance,
heteronuclear dipole-dipole couplings are averaged out by MAS, yet a simple
train of 1800 pulses properly synchronized with the rotor phase reintroduces
the coupling for a well-defined period of time. This so-called Rotational-Echo
Double-Resonance (REDOR) technique can be seen as the basis of many more
sophisticated ways of reintroducing homo- and heteronuclear dipole-dipole as
well as anisotropic chemical shift interactions under MAS. Pictorially speaking
the mechanical rotation by MAS in real space is partially offset by counter
rotations in spin space.
Magic angle spinning modulates the spin interactions periodically. This
means it generates rotational echoes and data acquisition can be performed in two
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

ways: First, if only the echo-height is monitored (rotor synchronized acquisition)

a single line results in the spectrum for each resolved site and information about
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

the anisotropic coupling is lost. Secondly, if the whole echo-train is monitored,

a side band pattern results, which retains information about the anisotropic
coupling, yet with spectral resolution of the different sites. This is important
for precise structural elucidation, based on the dipole-dipole coupling as well as
using this interaction to study molecular dynamics resulting in reduction of the
coupling, Fig. 2b.

Double Quantum NMR

The basic concept for using the distance- and angular dependent dipole-dipole
coupling for structural studies is described in extended reviews (5, 7). In a two-
dimensional experiment, double quantum coherence (DQ) is created, e.g. between
two 1H with like or different chemical shifts Fig. 3a. During excitation of the
DQ coherence, the dipole-dipole coupling between the two spins, which is largely
reduced by MAS, is recoupled by a suitable pulse sequence. In the evolution time
of DQ coherence recoupling is turned off, such that the different residues can be
distinguished by their different chemical shifts. In the subsequent reconversion to
single quantum coherence needed for signal detection, recoupling is again applied.
Thus, a 2D spectrum, as shown in Fig. 3b is recorded, in which information about
internuclear distances is, first of all, encoded in the strength of the DQ peaks.
It should be appreciated that the technique is based on analyzing signals
resulting from a coherence originating from two spatially separated nuclei. Thus
the basic physics is analogous to coherent X-ray or neutron scattering, where
the coherent superposition of scattered waves is exploited to deduce the distance
between the scattering centers. In NMR, however, no translational symmetry is
necessary to determine distances in the range of 0.5 nm. In fact, proximities of
1H in the same or different moieties can be probed by DQ NMR by analyzing

the intensities of a so-called rotor-synchronized spectrum, which can typically

be recorded in about 10 min for 10 mg as-synthesized samples, without the need
of isotopic labeling. If the dipole-dipole coupling needs to be determined more
accurately, for more precise determination of internuclear distances or molecular
dynamics, see above, DQ sideband spectra, as displayed in Fig. 3c are recorded.
The measurement time is then considerably longer, typically overnight.

23
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 3. Principle of double quantum NMR. a) pulse sequence; b) double

quantum NMR spectra; c) sideband patterns.

Such DQ spectra can be recorded for both homonuclear 1H-1H, or

heteronuclear 1H-13C, 1H-15N coherences, exploiting the much higher site
selectivity of 13C and 15N chemical shifts. In the heteronuclear case polarization
transfer and recoupling take advantage of the popular REDOR technique
introduced above usually applied to isotopically labeled samples.

Two-Dimensional NMR Spectroscopy

Most of the advanced NMR techniques make use of two-dimensional (2D)

spectroscopy, because of the superb increase of resolution and ease of information
encoding. There is not enough space to explain these techniques in detail here, but
comprehensive books are available (3, 10) and only the basics are described here.
In general, a 2D NMR experiment is divided in several time periods that follow
each other. In order to record a 2D NMR spectrum, a two-dimensional data set
is acquired as a function of two time variables t1 and t2 as shown schematically
in Fig. 3a for the specific case of double quantum NMR. This is preceded by the
so-called preparation period in which coherences are excited by a suitable pulse
sequence, which in the simplest case is only one radio-frequency pulse. Unlike
conventional (1D) NMR spectroscopy, the excited signal is not directly acquired,
but is allowed to evolve in the so-called evolution period under influence of the
relevant spin interactions. The evolution time t1 is incremented in subsequent
experiments and provides the first time dimension of the experiment. After the
evolution period (and an optional mixing time), the remaining signal is directly
detected in the detection period for each time increment thus generating a 2D
data set. Two-dimensional Fourier transformation then gives the 2D spectrum.
Optionally, a so-called mixing period of length tm can be inserted between the
evolution and detection periods. During tm changes in the system can occur,
24
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 for instance, by molecular motions, spin interactions, relaxation or by spin
manipulation (for instance recoupling).
The different aspects of 2D NMR spectroscopy are reflected in the different
variants that can be distinguished. One variant, separation spectroscopy, is used
to separate different interactions by taking advantage of the spin manipulation
techniques For instance, during the evolution period the spin manipulation can be
made such that only the isotropic chemical shift is acquired while in the detection
period the full spectrum is acquired. This leaves the anisotropy to be studied
site-selectively.
Other 2D NMR techniques, so-called correlation techniques, aim at obtaining
new information by correlating different interactions. For instance, the 13C
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

chemical shift anisotropy can be correlated with the heteronuclear dipolar powder
pattern in order to obtain information on the relative orientation of the two
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

tensors. Moreover, since no signal is detected during the evolution time, it

can conveniently be used to provide the basis for recording a double quantum
spectrum which correlates single and double quantum coherences. Considering
the manifold of spin manipulation techniques there is a wealth of such 2D NMR
techniques that can be derived for different purposes.
Finally, introducing a mixing time tm, 2D exchange spectroscopy can be
performed, Fig. 4. The most important application of such exchange techniques,
with respect to polymer investigations, is the study of slow molecular dynamics.
In these experiments, reorientations due to molecular dynamics are allowed
to take place during the mixing time tm and lead to characteristic off-diagonal
patterns in the resulting 2D spectra. If the mixing time is increased in a series
of 2D experiments, slow dynamics in the range of milliseconds to seconds can
be investigated in detail. For instance, rotation of molecules by a well-defined
angle leads to an elliptical exchange ridge for a powder. This can be viewed as a
Lissajous-figure, from which the angle, by which the molecules have rotated, can
directly be read off by a ruler. Measuring time can be dramatically reduced in a 1D
variant under MAS. Likewise, exchange of magnetization can also occur by cross
relaxation or mutual spin-flips, the latter being designated as spin-diffusion. It can
be described by a diffusion equation and provides access to domain dimensions
in phase separated systems.

Applications

The chapters in this book describe NMR studies of polymers or

supramolecular structures in the solid-, the liquid crystalline-, or the liquid
state. Often one or the other aspect of the techniques introduced above is
used. Therefore, in this overview I have refrained from trying to describe a
comprehensive set of NMR applications in polymer science. Instead, I have just
included a few recent examples of our own work.

25
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 4. Principles of 2D exchange NMR for studying molecular motions.

Chain Microstructure, Defects, and Dynamics

The macroscopic properties of polyolefins strongly depend on the chain

microstructure. In recent years, new single site catalysts have enabled much
greater synthetic control over the polydispersity, type of branch, and branch
content. In polyethylene, the physical properties of both the solid and the melt
can be tuned by the presence of branches of various lengths in the polymer
backbone. Short-chain branches (SCB) can be introduced by copolymerization
of ethylene with short alkenes, or by isomerisation reactions during ethylene
homopolymerization Such branches form structural defects during crystallization
and thus strongly affect crystallization rates, ultimate crystallinity, and other
bulk mechanical properties. Similarly, long-chain branches (LCB) are formed
by macromonomer incorporation during polymerization. With these branches
typically being longer than the entanglement molecular weight their presence
strongly affects the processability of the bulk polymer. Long-chain branching
is known to influence the zero-shear viscosity even at concentrations of 2 LCB
per 100 000 CH2 groups. Thus, it is very important to quantify the degree of
chain branching and 13C NMR in solution seems to be the method of choice as
the chemical shifts of branch points as well as adjacent carbon positions can be
distinguished from the backbone resonances (2).
However, when applied to polyethylene a number of problems arise.
The most important being the low solubility of polyethylene, even at high
temperatures. The low concentration of 13C nuclei results in low NMR signals
requiring extended measurement times, especially for the quantification of very
low levels of branching. In specific cases up to two million transients had to be
acquired at high field in order to determine branch contents of 38 branches per
100 000 CH2. Another inherent limitation of solution-state NMR is the lack of
access to crosslinked and other non-soluble fractions of polyethylene. Solid state
26
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 type NMR, under magic-angle spinning (MAS), on the other hand, can be used to
overcome these limitations, despite the substantial loss of spectral resolution as
compared to solution NMR. Under optimal conditions time savings of about two
orders of magnitude are achieved, allowing quantification of 7-9 branches in 100
000 CH2 groups within 13 h (11), see Fig. 5a.
In commercial polymer samples the irregular distribution of the branching
sites along the main chain results in uncertainty of the branching influence on
morphology, chain dynamics, and other physical properties. Here polyethylene
samples with regularly spaced methyl branches obtained by Acyclic Diene
METathesis (ADMET) polycondensation (12) provide much more specific
information. By deuteration of the methyl branches, the molecular dynamics
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

of the defect site can be studied selectively via 2H NMR, whereas molecular
reorientations of the polymer chain between the branching sites can be monitored
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

separately via 13C chemical shift anisotropy. Combining these results with studies
of local conformations, a twist motion was identified, which is centered at the
branching sites for a spacing of more than twenty CH2 units between subsequent
branching points. In methyl branched PE samples with shorter spacing, e.g. 14
CH2 units between subsequent methyl branches, a collective dynamic process
emerges from this twist motion, which ultimately leads to the formation of a
rotator phase as indicated in Fig. 5b. Thus, localized and collective mobility
induced by the defects can clearly be distinguished. The superb spectral resolution
of a 850 MHz solid state NMR spectrometer was essential for obtaining these
results.

Dynamics in Polymeric Proton Conductors

The growing necessity for clean energy sources to substitute fossil energy has
created high demands for batteries and fuel cells. Therefore, various approaches
have been proposed aiming at developing new classes of proton conducting
membranes for high temperature applications (13). One promising approach to
the development of such a material is to combine the functions of the protogenic
group and the proton solvent in a single molecule. Such molecules must be
amphoteric in the sense that they behave as both a proton donor (acid) and proton
acceptor (base), and they must form dynamical hydrogen bond networks. The
first leads to the formation of a high concentration of intrinsic protonic defects
as a result of self-dissociation, and the latter may promote a high mobility of
these protonic charge carriers (excess and deficient protons). The mobility of
intrinsic defects is generally higher than that of extrinsic defects, which may be
introduced by acid or basic doping disturbing the local symmetry of the hydrogen
bond network.
Typical amphoteric liquids include phosphonic acid and diverse heterocycles
such as imidazole, pyrazole, benzimidazole and triazole. In the liquid state,
they all show relatively high proton conductivities with significant contributions
from structure diffusion, i.e. the motion of protonic defects (excess or deficient
protons) via intermolecular proton transfer, coupled to hydrogen bond breaking
and forming processes. A promising approach to eliminate the liquid electrolyte is
to attach the protic groups of the liquid electrolyte to the backbone of the polymer,
27
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 such that only chemical decomposition would result in a loss of ion carriers, a
prominent example being poly (vinyl phosphonic acid). From 1H, 2H, 13C, and
31P NMR combined with computer simulation, detailed information on the proton

mobility, water content, and the unwanted condensation of the phosphonic acid
groups can be obtained. The protons are highly mobile, but on the same time
scale no mobility associated with reorientation of the phosphonic acid groups
or the polymer backbone is observed. The 1H chemical shifts of P-OH protons
provide evidence for the presence of a complex hydrogen bond network, see
Fig. 6, which allows for proton transport via a Grotthus-type mechanism along
a given chain as well as between adjacent chains. The MD simulations further
show that proton vacancies can be trapped in the anhydride defect produced by
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

the condensation. More important, as shown in Fig. 6, the highest intrinsic proton
mobility with essentially isotropic geometry arises in systems like poly(vinyl
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

phosphonic acids), where the hydrogen bonded network is highly disordered,

characteristic of a hopping mechanism for proton conductivity (14).

Figure 5. a) Quantification of low branch content in optimized melt state

13C-NMR under MA. b) Local and Collective Motions in precise polyethylene
with CD3 Branches spaced every 15th and every 21st CH2 group.

28
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 6. Left: Distribution of the CP-OH angle within a given phosphonic
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

acid group of poly(vinyl phosphonic acid) PVPA, predicted from first-principles

molecular dynamics simulations. Right: Experimental 2H NMR lineshapes as
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

a function of temperature, illustrating motional narrowing due to the hydrogen

bond dynamics, exhibiting an effectively isotropic motion.

Hydrogen Bonding and Side Chain Conformation in Rigid-Rod Copolymers

In recent years the interest in polymer science has shifted considerably to

include liquid crystals (15) and supramolecular structures (16). As a recent
example, let us consider rod-coil copolymers such as oligo(p-benzamide)-
poly(ethylene glycol) (OPBA-PEG) copolymers with an oligomeric rod and
flexible PEG chains. They aggregate on a nanometer length scale, which is
important for many applications such as organic photovoltaics. However, this
aggregation behavior and the driving forces such as hydrogen bonding and -
interactions, as well as the role of the side groups, are not yet fully understood.
These non-covalent interactions can be studied by advanced solid state NMR,
combined with X-ray diffraction (XRD), differential scanning calorimetry (DSC),
and polarization optical microscopy (POM) (17). As shown in Fig. 7, longer
OPBAs form layered -sheet like aggregates stabilized by amide hydrogen
bonds, similar to polyamides with flexible side groups attached to the repeat
units rather than the end groups (18). These aggregates are remarkably stable
and apparently represent an equilibrium structure in both unsubstituted OPBAs
and OPBA-PEG rod coil copolymers. The binding of the PEG also introduces
a liquid-crystalline phase and the local structural order is improved in the
copolymer, if the sample is preorganized in that phase. Thus, by combining
different methods of structural investigation a model of local aggregation and
packing in both the liquid-crystalline and the solid state could be developed.

29
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

Figure 7. Scheme packing of rod-coil copolymers. The rigid rods are held
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

together by intermolecular hydrogen bonds and the space between the stacks is
filled by the PEG side chains adopting a Gaussian coil-like conformation.

Packing in Polymers with Ultrahigh Charge Carrier Mobility for

Field-Effect Transistors
Applications in plastic electronics like flexible displays or smart tags call for
organic semiconductors that can be easily processed into thin films and show field-
effect transistor (FET) charge carrier mobilities exceeding 0.1 cm2/Vs. To this end
a copolymer consisting of an alternating arrangement of cyclopentadithiophene
(CDT) as a donor and benzothiadiazole (BTZ) as an acceptor unit has recently
been examined (19), Fig. 8. The planar arrangement of the conjugated backbone
ensures a close -stacking which is required for efficient inter-molecular charge
carrier transport. Moreover, it was assumed that the close packing of the polymer
chains might originate from assisting donor-acceptor interactions. Indeed, for the
optimum molar mass, charge carrier motilities up to 3.3 cm2/Vs were achieved.
For the first time such a macromolecular system was recently examined by
solid state nuclear magnetic resonance (NMR) to provide a molecular scale picture
of the packing, Fig. 8. The 1H 2D DQ NMR spectra clearly reveal the relevant
packing contacts, confirming the expected - stacking for the polymer backbone.
The packing of the donor and the acceptor groups, however, was found to be more
delicate. Donor-acceptor groups are - stacked in a lamellar fashion and these
groups are ordered in an alternating way as shown in Figure 8d. Thereby, the
acceptor groups in one layer are located on top of the acceptor groups in adjacent
layers, however, not always in the exact same position, leading to a heterogeneous
packing. This model derived from NMR is consistent with the findings of X-ray
scattering. It also allows for optimal packing of the side chains that in the case
of long and bulky alkyl chains (C16) should be advantageous in order to avoid
steric clash. Conclusively, solid state NMR does not reveal a donor-acceptor
overlap within 0.4 nm. Thus, donor-acceptor interaction between the neighboring
CDT and BTZ groups located a adjacent chains apparently contributes little, if
any to the observed improvement in charge carrier mobility. NMR rather shows
the complexity of this remarkable CDT-BTZ copolymer system. This implies
that the enhanced packing order resulting in improved thin film crystallinity with
30
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 increasing Mn results primarily from -stacking interactions of the planar extended
conjugated backbones. (19).
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 8. Local packing and organization of the donor-acceptor groups in a

CDT-PTZ copolymer. (a) 2D contour plot of the 1H-1H DQ NMR spectrum.
(b) Color scheme used for assignments. (c) Expansion of the backbone region
showing the contacts between donor and acceptor groups. (d) Schematic drawing
illustrating the local packing of donor-acceptor groups in two neighboring
CDT-BTZ copolymer chain.

Empty Helical Nano-Channels from Low-Symmetry Macrocycles

Natural channel-forming structures are mandatory for connecting different

compartments within a living organism. For instance, trans-membrane proteins
function as ion-channels, transporters, or antibiotics. Recently, synthetic
nanochannels as large as 1.3 nm in diameter in a tight tubular supramolecular
superstructure have been generated from shape-persistent macrocycles. Solid
state NMR combined with computer simulation was able to provide the details of
the packing. The interplay of charge-transfer type interactions and steric effects
due to the side chains were unraveled (20). Moreover, it was shown that the
channels are empty and do not host solvent molecules or back-folded alkyl chains.
Fig. 9b displays the 2D NMR 13C{1H} hetero-nuclear correlation (HETCOR)
spectrum for the macrocycle in the LC phase. In this spectrum a remarkable
31
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 spectral resolution is observed, with 13C line widths in the order of ~100 Hz for
the core of the macrocycle, implying a very high degree of local order. From
the 2D NMR spectrum in Fig. 9b the complete assignment given in Fig. 9a is
obtained. It includes only one fourth of the total number of 13C signals possible
for the macrocycle indicating that the macrocycles are located in highly symmetric
environments, which can be envisaged in a helical arrangement of the macrocycles
in a column.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 9. Spectroscopic fingerprints of the helical assembly in an empty

nanochannel from solid state NMR. a) Assignment of the aromatic carbon
and proton resonances. b) 2D 13C{1H} HETCOR NMR spectrum in the
liquid crystalline phase. c)The helical packing with a pitch of 60 leads to
intermolecular correlation peaks AB and BB observed in the 1H-1H DQ- NMR
spectrum d).
The packing environment is also reflected in the 1H chemical shifts for the
core protons of the macrocycle. These differ substantially from each other (Fig.
9b), and also from the values found in solution, which is a clear indication of -
stacking. The specific pitch angle can be determined by combining 2D NMR
1H-1H DQ-SQ (Double-Quantum Single-Quantum) correlation spectroscopy with

ab initio calculations. This fixes the pitch angle between adjacent macrocycles to
~60 as illustrated by Fig. 9c and the inset of Fig. 9d. Within the stack every fourth
molecule is eclipsed, i.e. related by translation, resulting in a helical arrangement
(Fig. 10), as also observed in other self-assembled organic compounds. The
pitch angle of 60 is supported by additional NMR experiments and ab initio
calculations investigating the packing of pairs of macrocycles considering their
energetics and the 1H chemical shifts of neighboring macrocycles. These show
excellent agreement between experimental and calculated values. Thus, the
internal order of the channels can be molecularly controlled and adjusted for
future applications in recognition, stabilization, or organization of nanoparticles.
32
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Figure 10. a) Overall view of the helical packing arrangement of the macrocyclic
channel with a pitch angle of 60. b) Top view of the macrocycle. c) Packing
arrangement for the side chains illustrating the stabilizing effect of the outer
phenylene groups.

Conclusions and Outlook

Advances in synthesizing, characterizing, as well as understanding
macromolecular and supramolecular systems lead to an enormous variety and
complexity in the field of polymer science (21). The traditional separation in
terms of structure vs. dynamics, crystalline vs. amorphous, or experiment vs.
theory is increasingly overcome (1). As far as characterization of such materials
is concerned, no experimental or theoretical/simulation approach alone can
provide full information. Instead, a combination of techniques is called for and
conclusions should be backed by results provided by as many complementary
methods as possible, see Table 2, where NMR and scattering are compared.
Combining scattering or NMR spectroscopy with computer simulation is well
established today in the study of structure and dynamics of biomacromolecules.
The examples described here show the power of such an approach in the
supramolecular field involving the combination of spectroscopy, scattering and
computer simulation. Last, but not least, the development of NMR spectroscopy
is far from complete. In particular, in order to meet the ever-increasing demands
of miniaturization, the sensitivity of NMR spectroscopy has to be increased
substantially and several approaches in response to that challenge are underway
(22).

33
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 2. Comparison of scattering and NMR techniques as well as the
information provided about structure and dynamics of materials. WAXS:
Wide Angle X-ray scattering; SAXS: Small Angle X-ray Scattering; SANS:
Small Angle Neutron Scattering.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Acknowledgments
This paper is based on and concludes my work at the Max Planck Institute
for Polymer Research for more than 25 years. It has provided me with a unique
scientific environment, in which new ideas and approaches prospered. It gives me
great pleasure to thank my colleagues and co-workers for all their contributions.

References
1. Spiess, H. W. Macromolecules 2010, 43, 54795491.
2. Al Amri, A.; Sahoo, S. K.; Monwar, M.; McCord, E. F.; Rinaldi, P. L.
Macromolecules 2006, 39, 57685776.
3. Schmidt-Rohr, K.; Spiess, H. W. Multidimensional Solid-State NMR and
Polymers; Academic Press: New York, 1994.
4. Spiess, H. W. In Macromolecular Engineering: Precise Synthesis, Materials
Properties, Applications; Matyjaszewski, K., Gnanou, Y., Leibler, L., Eds.;
Wiley-VCH: Weinheim, 2007; pp 19371965
34
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 5. Brown, S. P. Macromol. Rapid Comm. 2009, 30, 688716.
6. Saalwaechter, K. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51, 135.
7. Schnell, I.; Spiess, H. W. J. Magn. Reson. 2001, 151, 153227.
8. Sebastiani, D. ChemPhysChem 2006, 33, 164175.
9. Hafner, S.; Spiess, H. W. Concepts Magn. Reson. 1998, 10, 99128.
10. Ernst, R. R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear Magnetic
Resonance in One and Two Dimensions; Clarendon Press: Oxford, 1987.
11. Klimke, K.; Parkinson, M.; Piel, C.; Kaminsky, W.; Spiess, H. W.;
Wilhelm, M. Macromol. Chem. Phys. 2006, 207, 382395.
12. Wei, Y.; Graf, R.; Sworen, J. C.; Cheng, C. Y.; Bowers, C. R.; Wagener, K.
B.; Spiess, H. W. Angew. Chem. Int. Ed. 2009, 48, 46174620.
Downloaded by STANFORD UNIV GREEN LIBR on June 19, 2012 | http://pubs.acs.org

13. Steininger, H.; Schuster, M.; Kreuer, K. D.; Kaltbeitzel, A.; Bingl, B.;
Meyer, W. H.; Schauff, S.; Brunklaus, G.; Maier, J.; Spiess, H. W. Phys.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch002

Chem. Chem. Phys. 2007, 9, 17641773.

14. Lee, Y. J.; Murakhtina, T.; Sebastiani, D.; Spiess, H. W. J. Am. Chem. Soc.
2007, 129, 1240612407.
15. Handbook of Liquid Crystals; Demus, D., Goodby, J., Gray, G. W., Spiess,
H. W., Vill, V., Eds.; Wiley-VCH: Weinheim, 1998.
16. Lehn, J. M. Science 2002, 295, 24002403.
17. Bohle, A.; Brunklaus, G.; Hansen, M. R.; Schleuss, T. W.; Kilbinger, A. F.
M.; Seltmann, J.; Spiess, H. W. Macromolecules 2010, 43, 49784985.
18. Clauss, J.; Schmidt-Rohr, K.; Adam, A.; Boeffel, C.; Spiess, H. W.
Macromolecules 1992, 25, 52085214.
19. Tsao, H. N.; Cho, D. M.; Park, I.; Hansen, M. R.; Mavrinskiy, A.; Yoon, D.
Y.; Graf, R.; Pisula, W.; Spiess, H. W.; Muellen, K. J. Am. Chem, Soc. 2011,
133, 26052612.
20. Fritzsche, M.; Bohle, A.; Dudenko, D.; Baumeister, U.; Sebastiani, D.;
Richardt, G.; Spiess, H. W.; Hansen, M. R.; Hger, S. Angew. Chem. Int.
Ed. 2011, 50, 30303033.
21. Ober, C. K.; Cheng, S. Z. D.; Hammond, P. T.; Muthukumar, M.;
Reichmanis, E.; Woolex, K. L.; Lodge, T. P. Macromolecules 2009, 42,
465471.
22. Spiess, H. W. Angew. Chem. Int. Ed. 2008, 47, 639642.

35
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 3

Tutorial on Solution NMR of Polymers

Peter L. Rinaldi*

Department of Chemistry, The University of Akron, Akron, Ohio 44325-3601

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

*E-mail: [email protected]
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

This paper is derived from a tutorial on the use of solution

NMR methods to characterize polymers. The presentation has
a dual purpose: 1) to orient new scientists who are just starting
to use NMR methods in polymer science; and 2) to introduce
new developments in NMR to experienced practitioners of
solution NMR of polymers. After a brief introduction to
using one-dimensional (1D-) NMR for quantitative analysis of
polymers, the presentation covers recent developments in pulse
sequences/software, hardware developments and methodology
that combines the two. For the benefit of novice NMR users, an
introduction to some basic NMR methodology is incorporated
into these discussions. The work cited includes NMR papers
on polymers as well as recent work on small-molecule NMR
that has potential for helping with polymer analysis.

Introduction
The most important aspect of polymer synthesis is to identify the structures
of the newly synthesized materials. Once a material with desirable properties
is prepared, it is important to know the structures present and their quantities,
so that the synthesis of the desired material can be reproduced. If properties of
synthesized materials are not reproducible, the polymer chemist must determine
whether the structure and composition of the material has changed to cause the
change in properties. Furthermore, systematic changes in polymer structure can
often be correlated with changes in mechanical and physical properties. This in
turn can lead to educated choices in designing the structures of new polymers with
desired properties. Finally, it is desirable to identify the nature and quantity of
adulterants which undesirably alter the properties of materials.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Solution state NMR is ideally suited for determining the structures and
compositions of organic materials. There are a wealth of pulsed 1D- and 2D-
NMR techniques for resolving the resonances of many unique atoms (especially
hydrogen and carbon) present in the mixtures of organic structure components
that make up a polymer. These methods can be used for determining how these
atomic components are linked to form molecules. In most cases, the structures
of unknown compounds can be determined by straightforward examination and
interpretation of NMR spectra.
Once the structures have been identified and NMR resonances have been
attributed to the structures present, NMR spectra can be used to quantitatively
determine the compositions of these structures in the polymer. In fact, NMR is
unique among spectroscopic methods in that a correctly acquired and processed
NMR spectrum has a simple one-to-one relationship between the strengths of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

signals and the number of atoms producing those signals. Other spectroscopic
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

methods require calibration procedures to determine extinction coefficients for

different functional groups before they can be used for quantitative analysis.
For the benefit of novice NMR users, this work contains a brief introduction to
polymer structure and some general comments on collection of quantitative 1D-
NMR spectra. It is followed by a description of some recently available (last 5
years) NMR methods which will have an impact on the field of solution NMR of
polymers. For more comprehensive descriptions of polymer NMR, the reader is
refered to some recent works (1, 2). For an excellent introduction to the use of
1D- and 2D-NMR for characterization of organic structures, the reader is referred
to an excellent textbook on the subject (3).

Polymer Structure

It is desirable to identify a number of structure elements in polymers. These

include: chain-ends, monomer composition, monomer sequence, sequences from
inverse monomer addition, stereosequence effects, branches and cross-link points
and polymer block junctions. It is also desirable to identify structure defects that
might be caused from side reactions during the polymerization, and to determine
the presence and amount of adulterants such as low molecular weight impurities
like solvents and unreacted monomer.

Chain-Ends

Equations (1)(6) outline the general mechanism for polymerization

including initiation, propagation and termination steps.

38
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Termination:

Chain transfer:
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

The relative rates of propagation and termination pathways determine the

molecular weights of the polymers produced. From a knowledge of the various
termination processes, the chemist can alter the reaction conditions to control the
polymers molecular weight. Additionally, many polymer degradation processes
start by reactions at the chain terminus. If the structures of the predominant
chain-ends are known, the reaction chemistry can be altered, or chain-end capping
reactions can be performed, to reduce the number of reactive chain-ends, leading
to a stable polymer.

Monomer Sequence

When a polymer is synthesized from two monomers, A and B,

the monomers can add in an alternating fashion (1), in a statistical
manner (2), in a manner to form blocks of A and blocks of B (3), or
such that sequences of B can be grafted to a chain of A sequences (4).

The prevalence of these various structures is dependent upon the relative

reactivities of A and B monomer with themselves or with each other, and on the
39
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 reaction conditions. The prevalence of these various structures has a dramatic
influence on polymer properties and is of intense interest to polymer chemists.
The relative reactivities of A/B monomers with themselves and with each
other can be derived from the analysis of n-ad sequences. Often, the resonances of
triads (AAA, AAB/BAA, BAB, ABA, ABB/BBA, BBB) can easily be resolved,
even at relatively low field. From an analysis of the relative compositions of these
triads, the predisposition of a particular copolymer to form alternating, random or
block structures can be quantitatively described (4). With the routine availability
of instruments with 1H resonance frequencies as high as 700 to 800 MHz, it is
sometimes possible to resolve resonances from heptads or even nonads. When it
is possible to resolve and measure the intensities of signals from different n-ads,
accurate information about relative reactivities is obtainable.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Stereosequence

When certain monomers such as 1-substituted ethylenes (e.g. styrene)

polymerize, they create stereogenic centers on alternating carbons along the
polymer backbone. In most cases the absolute stereochemistry is not important,
as a racemic mixture of polymer chains is created. However, the relative
stereochemistry of adjoining stereogenic centers is important, and has a dramatic
effect on polymer properties.
If the backbone of a vinyl polymer is drawn in a planar zig-zag orientation
as in structures 57, two relative orientations are seen: the meso (m) structure
has two adjoining R substituents on the same face of the polymer as in the
two dyads in 5; and the racemic (r) structure has two adjoining R substituents
on opposite faces of the polymer as in the two dyads in 6. If triads are
considered, isotactic polymers have the same orientations of all R substituents
to produce polymers with all m dyad placements. Syndiotactic polymers have
alternating orientations of R substituents to produce polymers with all r dyad
placements. Atactic polymers have random orientations of R substituents to
produce polymers with equal numbers of randomly placed m and r dyads.

40
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Since the properties of polymers are related in part to their tacticity,
measurement of m and r dyad composition is crucial. NMR is ideally suited to
this measurement, as the NMR resonance positions are influenced by relative
stereochemistry (5).

Other Defects

NMR is ideally suited for determining the structures and quantities of polymer
defects, whether they are incorporated intentionally or accidentally. These include
branch points, junctions between two polymer blocks, cross-link junctions and
other defects. Finally, adulterants such as residual solvents or monomer, and
additives such as antioxidants can usually be detected and identified.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

Structure proofs can usually be accomplished by comparison of resonance

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

positions with those in the spectra of authentic material. When authentic materials
are difficult to obtain or are not available, combinations of 2D-NMR methods
can be used to determine atomic attachments, prove structure elements and
simultaneously assign resonances.

Quantitative One-Dimensional NMR

Once the presence of the above structure elements is determined and their
resonances identified, it is desirable to quantitatively evaluate the amounts of these
structures present in the polymer. This is usually accomplished with 1D-NMR
experiments. While NMR is unique among spectroscopic methods in that peak
areas can be directly related to the number of atoms of each type, it is important
to properly obtain NMR spectra in order to get reliable and accurate analysis
of composition and purity. Accurate quantitative analysis of polymers requires
attention to the details at the data acquisition stage as well as the data processing
stage.

Data Acquisition

The most important factor in generating good quality 1D-NMR spectra for
quantitative analysis it to measure the T1 relaxtion times for the nuclei whose
signals are to be measured. This is accomplished using an inversion recovery
experiment (relaxation delay-180o--90o-acquire, where is a variable relaxation
delay) experiment. By fitting the exponentially decaying peak intensity variation
vs. from a series of spectra obtained with varying delays, the exponential
recovery time constant T1 for each peak in the spectrum is determined (1). In
setting up a quantitative 1D-NMR experiment, the relaxation delay should be 5-10
times the longest T1 among the signals of interest.
If decoupling is performed (as is the case when 13C NMR is used) then in
most cases, the nuclear Overhauser effect (NOE) resulting from saturation of the
decoupled nuclei must be suppressed, since signals of different 13C atoms are
differentially enhanced depending upon the unique relaxation characteristics of
41
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 each nucleus. To accomplish NOE suppression, a gated decoupling experiment
must be used, where the 1H decoupler is turned off except during the acquisition
time, and a 90o excitation pulse is used. Under these circumstances complete
suppression of the NOE requires a relaxation delay (during which the 1H decoupler
is turned off) of at least 10xT1. In rare instances, when signals of very similar atoms
are compared, spectra with NOE enhancement can be used if the enhancement is
the same for all signals to be compared.
Data accumulation should be performed with an acquisition time such that
the NMR signal decays (at least 3xT2*, the apparent transverse relaxation of the
NMR signal). In selecting the spectral window, the effective excitation window of
the pulse should be considered. A square pulse produces an excitation window in
which peak height vs. offset from the transmitter (center of the spectral window)
is defined by a sin(x)/x function. A good rule of thumb is that for a 5s pulse only
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

the middle 50 kHz of the spectral window experiences uniform excitation. Outside
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

of this window, the peak heights will be attenuated by incomplete flipping of the
spins. If the pulse width is doubled (10 s) then the useable spectral window will
be halved (25 kHz).
A second factor to consider when setting the spectral window is hardware
filtering of the audio signal. Older NMR instruments mix the rf signal down to
the audio frequency range where it can be digitized by older analog-to-digital
converters (ADCs). To prevent aliasing of high frequency signal components
from outside of the spectral window (thus maintaining good signal-to-noise) audio
filters are used. These hardware filters have non-ideal response and can attenuate
signals outside of the middle half of the spectral window; attenuation of signals
near the edges of the spectral window is often 50% or more, depending on the age
of the instrument. Thus, only the middle half of the spectral window is useable for
quantitative analysis.
Instruments manufactured in the past five years use digital manipulation over
a much larger portion of the signal receiver path. In these newer instruments,
the NMR signal is digitized in the rf range, and mathematical filtering is used
to prevent aliasing of signals from outside of the desired spectral window.
Mathematical filters have more desirable brick wall behavior; completely
attenuating signals from outside of the spectral window, while leaving signals
within the spectral window unperturbed. In these newer digital instruments,
signals in the middle 90% or more of the spectral window can be used for
quantitative analysis.
Once the spectral window is determined, the delay between the end of
the pulse and start of data acquisition should be adjusted such that in the final
spectrum all peaks can be phased correctly with adjustment of only the zero order
phase correction. The spectrum should be properly phased without the use of first
order phase correction. The optimum preacquisition delay is spectral window
dependent, so that once analysis parameters are determined, it is unnecessary to
make further adjustments and the parameters can be used for subsequent analyses.
If the spectrum is collected with an incorrect preacquistion delay, then a sinusoidal
roll will be introduced into the baseline. The intensity and number of cycles will
be dependent on the extent of the misset, and the spectral window. The problems
introduced will be more severe for data collected with larger spectral windows.

42
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Finally, for good quantitative analysis, spectra with good signal-to-noise
(S:N) are required. Analytical criteria specify that a minimum S:N of 10:1 is
required before quantitative measurements can be performed. However, for
accurate analyses, S:N levels an order of magnitude or greater should be used.

Data Processing
In order to accurately measure a peak integral, the peak must be accurately
represented in the digital domain. Typically, it is desirable to have approximately
10 points to define the top half of the peak shape. A good rule of thumb is to
zero fill a properly collected time domain signal (free induction decay, FID) to 2
or 4 times the original dataset size. To improve S:N and to further contribute to
sufficient digital resolution of the peaks, it is also useful to apply an exponentially
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

decaying function to the decaying FID. Optimum S:N is achieved when the decay
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

of weighting function matches the decay of the FID. Often, in real spectra there
will be a mixture of broad and sharp peaks from polymer backbone and side-chain
resonances, respectively. Therefore, it is usually desirable to examine the spectral
with different weighting functions.
A properly phased spectrum is essential for good quantitative analysis. As
mentioned above, if the data are properly collected, it should be possible to phase
the spectrum solely by adjustment of the zero order phase parameter. Often the
operator will phase spectra by zooming in on the peak and attempting to make the
peak symmetrical. If the peak is inherently asymmetric due to a signal component
producing a shoulder on the main peak, or due to an asymmetric shape from
incorrect shimming, then integration error will result. There are many published
protocols for phasing. Some of these involve phase adjustments to produce a
maximum peak integral and numerically monitoring the baseline on either side
of the peak. When phasing, it is useful to examine the true baseline regions on
either side of the peak and to adjust the phase so the extensions of these baseline
components to the peak center, intersect at the same vertical point.
Distortions in the baseline should be removed with the operators favorite
baseline correction algorithm. For properly collected data, the distortions should
not be too severe and a spline fit is effective. Other more involved procedures
involve fitting the baseline to an nth order polynomial, or to a sum of sinusoidal
functions. In all cases the baseline (peak-free regions of the spectrum) needs to
be identified. This is often accomplished by dividing the integral regions of the
spectrum (e.g. such that odd regions are peak-free). When defining the baseline
regions it is important to consider the fact that Lorenzian peaks typically found in
NMR spectra extend for quite a distance from the frequency of the peak maximum.
Approximately 99% of the peak area is found over a region 20 times the peak
width at half height (1/2); 98% of the peak area is found within a region with 10
times 1/2; and only 95% of the peak area is found within a region with 5 times
1/2. While it might be desirable to integrate the spectra with narrower integral
regions surrounding the peak maxima, it is necessary to select baseline regions of
the spectrum that do not fall within 10-20 peak widths of the intense signals in the
spectrum. If this is not done, baseline correction will shift the baseline offsets due
to the presence of real signals, and lead to inaccurate peak integrals.
43
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Accurate quantitative analysis requires that all of the resonance assignments
be known and properly counted. Some particular problems include contributions
from 13C satellites and the outer wings of AB patterns in 1H and 19F NMR spectra.
Failure to count these signals properly can cause errors in quantitative analysis,
especially if the intensities of weak and strong signals are being compared.
Figure 1 shows an excellent example of how both of these factors can
cause complications in quantitative analysis. This region of the 19F spectrum of
poly(hexafluoropropylene oxide) (pHFPO) contains two AB patterns, labeled a2
and a2 from the two diastereotopic CF2 fluorines shown on the two structures
at the bottom of the figure. The structures differ by inversion of the central
CF(CF3)CF2 monomer unit. The central parts of these AB pattern are off scale.
They are components of the AB patterns from the chain-ends groups of pHFPO
containing CF2 signals from normal (a2) and inverse addition (a2) of the second
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

monomer unit in the chain. The two weak signals at the outer edges of this
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

spectral region are the outer satellites of the AB pattern from the a2 fluorines.
The measured couplings in this AB pattern are 2JFF = 285 Hz, which is typical
of a two-bond 19F-19F homonuclear coupling. The corresponding outer parts of
the AB pattern from the a2 fluorines are too weak to detect; they are ca. 10 fold
weaker due to the lower concentration of structures from monomer inversion.

Figure 1. Section of the 470 MHz 19F 1D -NMR spectrum from pHFPO showing
the region containing the a2 and a2 CF2 resonances as labeled on the structures
at the bottom of the figure. The ratio of the two structures is ca. 100:12.

Satellites of the a2 resonance from 1.1% 19F-13C and 2.2% 19F-C-13C are seen
with couplings of 1JCF = 270 and 2JCF = 40 Hz, respectively. The center of the
doublet from 1JCF is shifted upfield by ca. 0.1 ppm by a 13C istotope shift. The
center of the doublet from 2JCF is shifted upfield by ca. 0.01 ppm by a 13C istotope
44
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 shift. Additionally, the latter doublet (from 2JCF ) is twice as large as the former
doublet (from 1JCF ) because the probability of coupling to 13C is twice as large
(13Cs can be present at two sites, CF2 and CF3, which are two bonds away from
the a2 fluorines). These signals from coupling to 13C could easily be misinterpreted
as weak resonances from other minor chain-end structures, especially because
they are asymmetrically spaced around the central 19F resonance. Corresponding
13C satellite signals from the weaker a2 resonances are not seen in the spectrum

because these structures are present in ca. 10-fold lower concentration.

Also, if these satellites had fallen beneath weaks signals of minor structure
components, they could easily lead to enormous errors in compositional
calculations.
If very accurate compositional calculations are needed, one should compare
integrals which are a similar multiple of the peak width. Usually, 10 peak widths is
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

acceptable. However, one must also consider overlap of neighboring resonances

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

and the effect of J-couplings on the width of multiplets in deciding integral reset
positions. When doing compositional calculation with a high degree of accuracy,
one must also decide whether to incorporate 13C satellites, and to be consistent in
their inclusion or omission.
Once the peak areas are determined, there are several strategies for calculating
compositions of various monomer- or stereo-sequences. The traditional strategy
involves determining the relationship between peak areas and various n-ad
structure elements (monomer sequences or stereosequences). A set of linear
equations is obtained which can be solved for n-ad compositions. In many cases
some of the resonances are associated with odd-numbered n-ads such as triads,
and other resonances are associated with even number m-ads such as dyads and
tetrads. These can be related to each other through necessary relationships, which
also provide additional linear relationships (2).
On occasion problems with incorrectly counting structure units occur when
more than one n-ad is considered in the compositional analysis. Researchers
from Dow proposed an alternative method for extracting polymer compositions
from NMR data (6). Their method involves examining resonances and attributing
them to the central monomer of odd-numbered n-ads, thus counting atoms only
once. From these assignments, a set of linear equations can be obtained that
relate peak integrals to a set of structures. The compositions can be solved by
numerically fitting calculated peak areas to those experimentally determined.
Recently, Momose et al. (7) described the use of multivariate analysis as a method
for studying the compositions of copolymers and homopolymer blends.

New Experiments and Pulse Sequences

Over the past decade, 2D-NMR methods have become routine for studying
the structures of complex polymers and producing the unequivocal resonance
assignments that are needed for quantitative analyses. Complete spectral
analysis usually involves collection of two or more 2D-NMR spectra and using
complementary information from the spectra to construct an entire structure.

45
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 2D-NMR Methodologies for Determining Polymer Structure

Various methodologies have been developed for determining polymer

structure. Reference (3) serves as an excellent initial source for learning how to
determine organic structure and resonance assignments from NMR data. Two
commom methodologies for extracting structures from multidimensional NMR
spectra involve the use homonuclear or heteronuclear 2D-NMR methods.

Homonuclear 2D-NMR

In the case where sample quantity and/or sensitivity is an issue, methodologies

involving 1H - 1H (or 19F - 19F) correlation 2D-NMR experiments are useful. Such
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

situations include cases where sub-milligram quantites of material are available,

or where very weak signals from structures such as chain ends are to be studied.
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

In these homonuclear experiments the natural abundance of both nuclei is 100%,

and high sensitivity detection of nuclei with large magnetic moments (1H or 19F)
is performed.

Figure 2. Schematic representation of homonuclear 2D-NMR experiments and

the structure information available from them: a) COSY, b) DQCOSY, c) TOCSY,
and d) NOESY. Reprinted with permission from reference (1).

The structure information available from homonuclear 2D-NMR experiments

is illustrated schematically in Figure 2. The atomic connectivities in the structure
in Figure 2 are highlighted by colored boxes that match the color codes of the
experiment names which provide that structure connectivity information.
The COSY (correlation spectroscopy, highlighted in yellow) 2D-NMR
experiment produces a spectrum schematically illustrated by Figure 2a with the
46
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 identical homonuclear chemicals shifts plotted along the two axes. Peaks are
found along a diagonal at the shifts of the resonances in the spectrum. If there is
germinal or vicinal coupling between protons, cross-peaks appear off the diagonal
at the intersection of the two shifts indicating atomic connectivity highlighted in
the yellow box (i.e. the vicinal relationship between HA and HB) on the chemical
structure in the figure. Similar information from the COSY spectrum identifies
the vicinal relationship between HB and HC (not shown on the figure).
The DQCOSY (double quantum filtered COSY, also highlighted in yellow)
NMR experiment produces a 2D spectrum (Figure 2b) similar to that in COSY,
but with signals from singlets (resonance not coupled) filtered from the spectrum.
These signals do not produce useful information in COSY spectra and only serve to
hide cross-peaks that fall near the diagonal, so their absence from the 2D spectrum
is an advantage.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

TOCSY (total correlation spectroscopy, highlighted in blue) 2D-NMR spectra

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

(Figure 2c) have an appearance similar to COSY spectra, but with correlations
among resonances of all protons that are part of the same spin system as illustrated
by the structure elements highlight in blue on the structure in the figure. Note
crosspeaks between HA and HC in this spectrum, even though they are not coupled
to each other. These TOCSY crosspeaks are produced because they are part of the
same spin system, where HA is coupled to HB and HB in turn is coupled to HC.
NOESY (NOE spectroscopy) 2D-NMR experiments produce spectra (Figure
2d) with cross-peak patterns similar to those found in COSY spectra, but with
cross-peaks appearing at the intersection of the chemical shifts of protons that are
close (usually within 0.5 nm) in space. They provide structure information about
the spacial proximity of groups (regardless of coupling) like those shown by the
red arrow on the structure in the figure.
From a set of homonuclear experiments, it is usually possible to determine
chemical structures. However, these experiments suffer from the limited chemical
shift dispersion of 1H (10 ppm). If many structures are present with slight
environmental differences, such as those found from stereosequence effects, it
can be difficult to resolve all of the cross-peaks needed to determine structure.
Furthermore, the distinction/resolution between cross-peaks is complicated
by the fact that the cross-peak pattern are a cluster of (n x m) peaks where n and m
are the multiplicities of the coupled peaks in the 1D-NMR spectrum. For example,
the cross-peaks between HA and HB resonances in Figure 2a are a 2x4 set of peaks.
The projection of these cross-peaks produce a 2-line pattern at the chemical shift
of the HA resonance, which is a doublet; and a 4-line pattern at the chemical shift
of HB, which is a doublet of doublets. If two sets of cross-peaks occur at similar
shifts, it is often difficult to distinguish between them.
Finally, the individual peaks within a cross-peak pattern are a series of positive
and negative contours, determined by the relative signs of the J-couplings. In
low resolution 2D-NMR spectra, overlap between adjoining anti-phase peaks can
lead to signal cancellation, especially if the J-couplings are small compared to the
digital resolution and/or the peak width. In these circumstances it is easy to miss
cross-peaks, leading to misassignments and incorrect interpretation of the spectra.
This is especially true with spectra of polymers, which contain broader lines than
those found in the spectra of small molecules.

47
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 For these reasons, when adequate signal is present, it is often preferable to
study structures using a methodology involving two or more heteronuclear 2D-
NMR experiments.

Heteronuclear 2D-NMR

In the case where ample quantities of material are available (1-100 mg

depending upon the abundance of the structure to be detected and the field
strength of the instrument used) heteronuclear 2D-NMR experiments can be
performed. These experiments most commonly involve sensitive detection of a
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

nucleus such as 1H or 19F, which has a large magnetic moment and is also present
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

in high natural abundance. However, the shifts of these nuclei are correlated with
a low abundance nucleus such as 13C (1.1% abundance), so that NMR signal is
detected from one molecule in 100. These experiments have the advantage that
the chemical shift dispersion in the indirectly detected dimension is determined
by the shift range of the heteronucleus (200 ppm in the case of 13C). Also, the
correlation is a (1 x n) pattern as there is usually no coupling in the 13C dimension
(unless there is a third NMR-active nucleus in the structure). Thus, it is possible
to resolve many resonances from chemically similar structure fragements.
Figure 3 schematically illustrates the most useful heteronuclear 2D-NMR
spectra. As in the previous section, the connectivities on the structure in the
figure are highlighted with colored boxes to match the colors of the boxes around
the experiment names from which the atomic connectivities are derived. The
HSQC (heteronuclear single quantum coherence, colored in yellow) and HMQC
(heteronuclear multiple quantum coherence) experiments use one-bond C-H
couplings to produce spectra (Figure 3a) with cross-peaks indentifying direct
C-H attactment, as illustrated by the fragments highlighted by the yellow boxes
on the structure in the figure. The HMQC experiment was originally used for this
purpose, but has been replaced by the HSQC experiment which produces more
intense signals and less complicated cross-peak patterns. Variations of the HSQC
experiment that use adiabatic 180o pulses usually give better signal to noise levels
than spectra from HSQC experiments using simple 180o pulses (8).
The HSQC experiment is usually used in combination with data from the
HMBC (heteronuclear multiple bond correlation, highlighted by the blue boxes)
experiment which produces spectra similar to that illustrated by Figure 3b. The
HMBC spectrum shows correlations between 13C and 1H atoms coupled through
2- and 3-bonds. Structure fragments highlighted in blue on the structure in the
figure are identified. Complications arise in distinguishing between 2- and 3-bond
C-H correlations. One method to enhance one or the other set of cross-peaks
is to perform the experiment twice with different delays optimized for 2- and
3-bond couplings (80 and 50 ms usually work for the former and the latter). Some
judgement of relative cross-peak intensity in the two spectra is necessary, and the
results are often ambiguous.

48
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 3. Schematic illustration of heteronuclear correlated 2D-NMR spectra

from the hypothetical structure fragment shown at the top of the figure: a) HSQC
or HMQC, b) HMBC, c) H2BC, d) HSQC-TOCSY. Reproduced with permission
from reference (1).

The advantage of this experiment is that correlations across heteratoms and

quaternary carbons can be detected (e.g. in this case a correlations can be seen
between the resonances of the methoxyl protons and CC). Distinction between 2-
and 3-bond correlations in the HMBC spectra is usually the Achilles heel of this
heteronuclear 2D-NMR methodology. Many variations of the HMBC experiment
have been devised to help differentiate between between 2-bond and 3-bond
correlations. Most of these work nicely for small molecules, but because polymers
give spectra with broader lines from shorter relaxation times, these experiments
rarely work well for polymers. The exception is the H2BC (heteronuclear
two-bond correlation) experiment which produces spectra illustrated by Figure
3c (9). This experiment uses a combination 1JCH and 3JHH to produce correlations
among the atoms in the structure fragment highlighted by the red box on the
structure in the figure. In this example a 2-bond correlation is seen between CA
and HB. The H2BC cross-peak patterns are considerably simpler that those in the
HMBC spectra because 13C decoupling can be used during the acquisition time
in the former experiment. Furthermore, the cross-peaks are independent of 2JCH;
49
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 two bond C-H couplings are often small leading to inefficient coherence transfer
between nuclei in the HMBC experiment. When polymers are studied, for the
same experiment time, signals in H2BC spectra are considerably more intense
than signals in HMBC spectra. The drawback of H2BC is that correlations to
quaternary carbons are not detected.
In the HSQC-TOCSY spectrum (Figure 3d), at the shift of each proton bearing
13C, a C-H correlation is detected at the intersection of the chemical shifts of

directly bound carbon and hydrogen (e.g. between CA and HA). In addition, at this
13C shift, additional TOCSY correlations are observed to the shifts of all the other

protons (HB and HC in this case) that are part of the consecutively coupled proton
spin system. The experiment is usually run to produce positive contours (red spots)
to indicate correlations between the resonances of directly bound carbons and
hydrogens, and negative contours (blue spots) to indicate correlations between the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

resonances of the initial C-H carbon and protons not directly bound to that carbon
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

but part of the same coupled proton spin system.

Figure 4 shows selected regions from the heteronuclear 2D-NMR spectra of
poly(lactate-co-glycolate)-b-poly(ethylene glycol) capped with folate end groups
on the polyester chain-end (PLGA-PEG-FOL). These data show the regions which
support the assignments of resonances from the lactate portion of polymer, as
shown on the structure in the figure. Figure 4a contains the HSQC data showing
a one-bond correlation for the methyl group. Figure 4c shows the portion of the
HMBC spectrum containing correlations from lactate methyl protons to the lactate
methine and ester carbonyl carbons (correlations A1 and A2, respectively), two-
and three-bonds away, respectively. On its own, this spectrum does not permit
distinction between the two- and three-bond correlations. Figure 4b shows the
relevant region from the H2BC spectrum; it identifies the A1 cross-peak as the
one associated with the two-bond HMBC correlation.

Figure 4. Polylactic acid regions from the 2D-NMR spectra of

PLGA-PEG-Folate: a) HSQC, b) H2BC, and c) HMBC of PLGA-PEG-folate.

50
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Quantitative NMR

The spectral dispersion provided by 2D-NMR methods is enormously

powerful for resolving and assigning resonances in complicated polymer
structures. However, once resonance assignments are made, it is desirable
and often necessary to use this information to perform quantitative analysis of
polymer composition. This can be a problem if resonances are not resolved in
standard 1D-NMR spectra. There are alternatives for getting resolved peaks for
quantitative analysis.

2D-NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

The utilization of 2D-NMR peak integrals for quantitative analysis is

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

complicated by several factors that produce non-linear instrument response

which hampers quantification. These include: 1) resonance offset effects on
both transmitter and decoupler channels; 2) inconsistent polarization transfer
efficiency due to ranges of J-couplings; 3) variations in 1H T1 causing inconsistent
return of the magnetization to its equilibrium state during the relaxation delays;
4) differential polarization transfer due to variations in T2s; 5) effects from
inadequate digital resolution; and 6) multiplicity effects (CH, CH2 and CH3
signals do not provide identical responses).
Lemaster et al. (10) first used quantitative 2D-NMR analysis for the study
polymers by examining the spectra of polyester urethanes from 1,4-butanediol/
4,4-diphenylmethane diisocyanate and adipic acid. They used 2D-HSQC and 3D-
TOCSY-HSQC spectra to obtain resonance assignments. They then used HSQC
spectra under conditions so that the resonance offset effects were insignificant.
They determined relative 1H T1s and corrected for shorter T1s of the backbone
protons relative to the chain-end groups. They compared the 2D-NMR integrals of
CH2 groups which were expected to have similar 1JCH couplings. They determined
that inconsistent T2s did not contribute to the integration errors. Digital resolution
problems were eliminated by collecting higher resolution 2D-NMR data, and by
using signal processing methods to extend the acquired signal (especially in the
evolution time dimension). Only signals of CH2 groups were compared. After
accounting for these issues, they obtained excellent results when they were able to
determine the monomer distribution ratio and the ratio of backbone to chain-end
units.
Most often, inadequate resolution in 1D-NMR spectra occurs when comparing
resonances of similar structures, as these are the resonances most likely to overlap.
This was the case in research reported by Qiu et al. where they were concerned
with the composition of poly(ethylene-co-octene) (6). 1D- and 2D-NMR data
from the -CH2 region of one of their samples of poly(ethylene-co-1-octene)s
(polyEO) containing 64 mole% 1-octene are shown in Figure 5.
The quantitative 13C 1D-NMR spectrum is shown across the top of the
corresponding region from the HSQC spectrum collected under conditions
where the 2D-peak volumes are expected to be quantitative. Since the -CH2
resonances are being compared, identical JCH couplings are expected, identical

51
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 relaxation is expected, effects of resonance offset are identical in both the 1H
and 13C dimensions, and multiplicity effects are identical. 1H T1 could not
be measured directly because the -protons are not resolved; however, T1s
for these protons (0.2s) could be measured indirectly through detection of the
resolved - 13C resonances. The bar graph at the bottom compares the 1D-(blue)
and 2D-(purple) peak integrals for the three resolved resonances. Excellent
comparisons are seen, indicating the utility of 2D-NMR spectra for quantitative
analysis.

Diffusion
NMR diffusion measurements have played an increasingly important role in
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

characterizing components in polymer systems. Although the first NMR diffusion

measurement using pulsed magnetic field gradients was described by Stejskal
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

and Tanner (11), the development of a 2D-pulsed field gradient NMR method,
DOSY (diffusion ordered spectroscopy), really sparked applications of diffusion
measurements (12). A good summary of applications and methodology can be
found in reference (13).

New Hardware
Over the past decade, hardware improvements have been made in several
areas. Modern instruments provide precise and accurate control of the rf
amplitude, phase and frequency. After decades of incremental improvements in
instrument sensitivity, new detection circuitry has provided a quantum jump in
signal-to-noise level. Incorporation of multiple receivers permits simultaneous
detection of two or more spectra.

Control of Rf
Modern instruments provide unimaginable control of the rf signals to permit
flexible and arbitrary schemes for manipulating the states of nuclear spins. Many
of these schemes reduce the demands of collecting high resolution 2D-NMR data
by accomplishing selective excitation of the information containing regions of the
spectrum, while ignoring other uninteresting regions of the spectrum. There are
a number of options for accomplishing this; only a few of these will be discussed
here.

Selective Excitation

The COSY pulse sequence is the most basic 2D-NMR experiment, consisting
of only two rf pulses separated by an evolution period. It is simple to perform,
tolerant of miscalibrations, has good sensitivity, and requires very little expertise to
perform. When dealing with hydrocarbon-based structures, it is the first 2D-NMR
experiment performed and it generally provides useful structure information.
52
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 5. Backbone -CH2 region from the quantitative 1D-13C and

2D-HSQC NMR spectra poly(ethylene-co-octene) containing predominantly
racemic relative orientations (top of figure). A bar graph is shown comparing
compositions of monomer sequences OOOO, EOOO and EOOE determined from
1D- (blue) and 2D- (purple) integrals (bottom part of figure). Reproduced with
permission from reference (6).
When working with fluoropolymers, the huge 19F spectral window (200 ppm
compared with 10 ppm for 1H) is an advantage in that the resonance positions are
well dispersed and sensitive to the chemical environment of the fluorine atoms.
53
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 When COSY experiments are perfomed with 19F, many FIDs with different values
of the t1 evolution period must be collected to digitize the indirectly detected
dimension of the 2D-NMR spectrum.
Figure 6a shows a small fraction of the entire 19F-19F COSY spectrum
(ca. 1% of the total spectrum area) from tetrameric(hexafluoropropylene
oxide) T(HFPO), which is a model compound for Krytox fluoropolymer, a
poly(hexafluoropropylene oxide) P(HFPO). To obtain this data, a spectrum with
several thousand t1 increments was collected over the course of a weekend.
The experiment was long because many FIDs were needed, not because signal
averaging was required to obtain good signal-to-noise. Only the region shown
was needed to obtain the desired correlation information between the CF2-O-CF
structure fragments illustrated by the red arrows on the structure in the figure.
Despite the long experiment, spectral detail in the multiplets is not resolved.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

Figure 6b shows the entire 2D-NMR spectrum from a selective COSY (14)
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

experiment in which shaped pulses were used to produce a spectrum containing

only CF resonances in the f1 dimension and CF2 resonances in the f2 dimension,
corresponding to the same spectral window shown in Figure 6a. The experiment
required less than 30 minutes of instrument time in which several hundred FIDs
were collected (15, 16). The experiment required hardware to produce arbitrary
rf pulse shapes with arbitrary frequency offsets, not just the simple square pulses
centered at the frequency in the center of the spectral window as are commonly
used in NMR. Note that the selective COSY spectrum contains far more detail than
is seen in the corresponding region from the standard COSY spectrum.
Figure 6c contains a spectrum obtained from a selective COSY spectrum
collected with inversion pulses centered in the middle of the evolution period, in
order to remove homonuclear couplings. The flexible hardware permits resolution
of unprecedented spectral detail. The cross-peaks all appear as singlets in the
f1 dimension; multiplet pattern containing coupling information are retained in
the f2 dimension where collection of high resolution data has little impact on the
total experiment time. Two sets of four cross-peak patterns are observed; one
set showing the B2-C1 correlation between fluorine resonances from the central
CF2OCF groups, and a second set showing A2-B1 correlations between the fluorine
resonances of the CF2OCF fragment on the left part of the structure at the top of the
figure. Four cross-peak patterns are observed for each C-F group because the three
stereogenic centers in the structure produce four diastereomers and their mirror
images.

Homonuclear Decoupling

Once multidimensional NMR experiments provide resolution and resonance

assignments, the next immediate need is to measure how much of each structure
component is present using peak area integration. 13C 1D-NMR experiments
are ideally suited for this purpose since all of the protons can be decoupled with
broadband irradiation, to produce a spectrum with a single line for each unique
carbon atom. Since the natural abundance of 13C is only 1.1 %, homonuclear

54
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 coupling between 13C atoms does not complicate the spectrum because coupled
13C-13C spin pairs occur in only 1/104 molecules.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 6. 2D-COSY spectra of hexafluoropropylene oxide tetramer: a) segment

from a standard COSY spectrum; b) full spectrum from a selective COSY
experiment; and c) full spectrum from a selective COSY experiment with
homonuclear decoupling in the f1 dimension.

55
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 When doing NMR of abundant spins such as 1H or 19F, homonuclear coupling
produces spectra with complex multiplets. In complex molecules, these multiplets
result in overlapping patterns which cant be resolved from one another, and cant
be easily integrated. Although many attempts have been made, until now it has not
been generally possible to accomplish broadband decoupling of homonuclear spins
to produce 1D-NMR spectra with integrable singlets. Until recently, it has only
been possible to obtain spectra with single frequency irradiation of the resonance of
a single atom, or at best to perform homonuclear decoupling of a narrow multiplet.
Recently, Espindola et al. (17) described the use of state-of-the-art rf hardware
to produce shaped inversion pulses. These capabilities were used to decouple an
arbitrary combination of nuclei with resonances in various spectral windows, and
produce a spectrum of singlets with broadband decoupling of the rest of the nuclei
in the spectrum.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 7. Selected (CF containing) regions from the 470 MHz 19F 1D-NMR
spectra of hexafluoropropylene oxide tetramer: (bottom) standard spectrum;
(top) spectrum with broadband decoupling of CF2 resonances.

This is illustrated with the spectra of T(HFPO) shown in Figure 7, where the
C-F containing regions of the 470 MHz 19F spectra are shown. In the bottom
spectrum, coupled C-F multiplets from the middle (B and C) monomer units
shown in the structure at the top of the figure (Bn and Cn designate the fluorines of
CFn groups in monomer units B and C in the structure). These resonances were
resolved and assigned with the aid of the selective COSY spectra shown above.
It is impossible to resolve and integrate the 8 separate sets of resonances due
56
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 to large couplings with the CF2 groups. In structure such as this, the four-bond
couplings are usually large (20-40 Hz), compared with the three-bond couplings
to the CF3 groups.
The top spectrum was collected with shaped inversion pulses between data
point samplings in the FID, such that the CF2 atoms (resonances between -75
and -90 ppm) are inverted. As a consequence, homonuclear couplings that evolve
during the first half of the dwell time (period between sampling of two data points
in the FID) are refocused during the second half of each dwell time. From the
homonuclear decoupled spectrum, shown in the top of the figure, it is possible to
separately integrate most of the resonances. At worst, it is possible to perform
peak deconvolution on some of the overlapping resonances to obtain quantitative
information.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Sensitive Detection

As with most spectroscopic techniques, in NMR there has been a constant

struggle to improve the receptivity of instruments so that spectra can be obtained
with smaller quantities of material. Over the past 60 years, this improvement has
been incremental, and resulted from improvements in instrument stability, magnet
homogeneity, magnetic field strength, rf electronics, solid state circuitry and probe
design. The exception has been the introduction of Fourier transform techniques
in the late 1960s and early 1970s, which lead to a quantum leap in receptivity.
The cumulative effect of these many incremental improvements has been a 105-106
fold gain in receptivity such that with the most sensitivie NMR instruments it is
now possible to get 1H NMR spectra from 10-100 ng of material compared with
100mg of material required to obtain spectra from instruments available in the
early 1960s.
After many years of instrument improvements, the major sources of noise
in the NMR experiment are now the detection coil and the sample itself. In the
past decade, probes with the detection coil cooled to cryogenic temperatures
have become commercially available. In these probes the detection coil is
typically cooled to 200K, while the sample located a few mm away, remains at
ambient temperature. Such probes provide a factor of 4-5 gain in receptivity
compared to older style probes in which the coil remains at ambient temperature.
Most of these commercial probes are designed to operate with the sample near
ambient temperature, using 5mm sample tubes for sensitive detection of 1H,
while decoupling 13C and 15N. These probes are primarily designed and used
for biological applications. For polymer applications we are not usually sample
limited, so that it is desirable to study larger samples. Also, often polymers
require high temperature to dissolve or swell the material for high resolution
solution NMR studies.
Zhou et al. (18) describe a new commercial cryoprobe designed for 13C
detection using 10 mm sample tubes, heated at temperatures up to 420oK, while the
coil temperature remains at 20oK. Under these conditions, they observed almost
6-fold increase in receptivity compared to a standard 10 mm probe designed for

57
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 13Cdetection. This new cryoprobe permitted them to study the minor structure
components in poly(propylene-co-1-octene) using 13C NMR.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 8. 100 MHz 13C NMR spectra of poly(propylene-co-1-octene):

a) INADEQUATE spectrum showing correlations of the major structure
components; b) vertical amplification of the spectrum in (a) revealing additional
correlations from minor structure components; and c) 1D-NMR spectrum.
Reproduced with permission from reference (18).

In this copolymer, the major structure component is polypropylene; the

next major component is derived from insertions of 1-octene units within the
polypropylene backbone; and finally the smallest component is derived from
inverse addition of propylene units. The study of this polymers structure
58
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 requires sensitive detection of the weakest signal components. The 100 MHz
13C 1D-NMR spectrum is shown in Figure 8c; the three signals that are off scale

(labeled (CH2)PP, (CH2)PO, (CH)P and (CH3)P) are derived from CH, CH2 and
CH3 carbons of normal polypropylene segments; the set of signals that fall near
half scale (labeled 1B6 6B6 and (CH)O) are from 1-octene units incorporated
into the polymer; and the weak signals that are labeled with capital letters (C, D
and F-N) are derived from inverse addition of propylene units. The latter labels
indicate resonance assignments corresponding to the labels on the structure at the
top of Figure 8.
The 2D-INADEQUATE experiment provides correlations among the
resonances of carbons related by 13C-13C attachments in a structure. Because
these spins occur together in only 1/104 molecules, obtaining an INADEQUATE
spectrum is 100 times more difficult than performing a simple 13C 1D-NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

experiment. Their plot of an INADEQUATE spectrum in Figure 8a, showing

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

correlations identifying the major structure fragments in the polymer, represents

the capabilities of most state-of-the-art NMR spectrometers. The vertical
amplification of this spectrum, shown in Figure 8b, reveals additional correlations
identifying low concentration structures from inverse propylene addition. These
structures are present in 3-5 times lower concentration.
Their 2D-INADEQUATE spectrum took 2.6 days to collect using a
cryoprobe with a 10 mm coil. They estimated that it would take 2.6 months
to collect a similar spectrum on a 400 MHz spectrometer using a standard probe
with a detection coil at ambient sample temperature. If quantitative 13C 1D-NMR
experiments are to be performed with senstitive detection of weak signals,
cryoprobes like the one described here could boost an instruments productivity
by 30-fold when long-term signal averaging is necessary. Alternatively, it might
be possible to detect weak signals in a weekend using this cryoprobe, when
another lab with a standard probe would never be able to dedicate 3 months to
collection of a single spectrum.

Multiple Receiver Systems

Multiple receiver systems have been used to speed up collection of image

data in magnetic resonance imaging (MRI) applications. These multiple receiver
systems are beginning to appear in labs doing high resolution NMR in chemistry
related applications.
In one paper, representative of one application area, Kupce et al. (19) were
able to nest 1H{13C}-HSQC and 19F{13C}-HSQC 2D-NMR experiments and to
use an instrument with two receiver channels to simultaneously detect the 1H and
19F signals (Figure 9). Such work had not been reported with a polymer sample so

the illustration here is with data from a small molecule. Two separate experiments
were performed simultaneously to double the productivity of the instrument.
The reader is encouraged to look at the cited literature in this paper to learn of
other possible variations. These experiments can increase the productivity of
instruments that are dedicated to routine screening of samples.

59
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 9. Nested HSQC pulse sequences (left) to produce 1H (black) and 19F
detected HSQC spectra (right). Reproduced with permission from reference (19).

Alternative Lock Channel

Another valuable use of instruments with a second receiver systems is as an

alternative to the field-frequency locking system. In order for the NMR signal to
add coherently, the peak frequencies must remain fixed from one transient (and
one spectrum) to the next. To compensate for magnetic field drift, instruments
have a third complete rf channel to monitor the frequency of a third nucleus and
to adjust Bo to maintain a constant resonance frequency. In almost all cases, this
third nucleus is 2H from a deuterated solvent. Deuteration of the solvent also has
the fringe benefit of reducing the solvent signal strength, especially in 1H NMR.
In solution NMR studies of polymers, it is often necessary to use exotic
solvents to dissolve the material, or to heat and swell the polymer. For example,
chlorinated solvents such as tetrachloroethylene or dichlorobenzenes are often
used to study polyolefins, because they permit heating the sample to 120-140oC.
At these temperatures, the polymers swell and produce high resolution spectra.
Often deuterated forms of these solvents are expensive or unavailable and
alternatives are needed to provide the needed lock signal. These range from
incorporating a small amount of a deuterated cosolvent, to addition of a capillary
containing a high boiling deuterated solvent to the center of the NMR tube.
Kupce and Freeman (20) used their second receiver system to monitor the
frequency of an independent NMR signal and to calculate a correction factor to be
applied to the spectrum during the data processing stage. Their results are shown
in Figure 10. In Figure 10a and b, the results from plotting a 13C NMR signal from
a 0.1 M solution of sucrose are shown without and with compensation for field
drift, respectively. Figure 10c and d show the effects of field drift on the HMBC
spectrum without and with compensation, respectively.

60
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 10. Results from using a second receiver as an alternate lock channel
while collecting HMBC spectra of 0.1 M sucrose in D2O: a) 13C signal monitored
without lock and b) the same signal as in (a), but with digital compensation for
field drift; c) HMBC spectrum without lock and d) same spectrum as in (c), but
with compensation for field drift. Reproduced with permission from reference
(20).

61
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Summary
This paper presents a small sampling of the new NMR capabilities that can
potentially impact polymer NMR spectroscopy. There are many advances in NMR
hardware, software and methodology that could not be mentioned here. Space
limitations require that the list stop at some point. If anything, the NMR field
seems to be advancing at a more rapid pace than ever before. Advancements
continue to contribute to improving the sensitivity of the NMR experiment so that
it is possible to work with smaller samples (or detect lower occurrence structure
components). New NMR methodologies help to provide information about new
structure components so that detailed structure information can be resolved. New
computers and software automate experiment setup so that once methodology is
developed, it can be applied by users who are not expert in NMR. Development of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

powerful computers and software will make it easier to extract information from
complicated spectra or groups of spectra. For new workers in this field, the times
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

are as exciting as ever.

Acknowledgments
The author wishes to thank the National Science Foundation (DMR-0905120)
for financial support during the preparation of this work. The author also wishes
to thank current and former research group members, University of Akron NMR
Center staff members (past and present) and collaborators who are too numerous
to mention here. They have contributed in countless ways to exciting research and
learning experiences over the past few decades.

References
1. Rinaldi, P. L.; Li, L.; Li, X.; Paudel, L.; Twum, E. B. Solution NMR of
Polymers. In Comprehensive Encyclopedia of Polymer; Spiess, H., Ed.;
Elsevier: New York, 2011; Chapter 2.06 (in press).
2. Mirau, P. A. A Practical Guide to Understanding the NMR of Polymers;
Wiley-Interscience: New York, 2004.
3. Silverstein, R. M.; Webster, F. X.; Kiemle, D. J. Spectrometric Identification
of Organic Compounds, 7th ed.; John Wiley & Sons: Hoboken, NJ, 2005.
4. Cheng, H. N.; Bennett, M. A. Anal. Chem. 1984, 56, 23202327.
5. Miri, M. J.; Pritchard, B. P.; Cheng, H. N. J. Mol. Model. 2010, DOI 10.1007/
s00894-010-0880-8.
6. Qui, X.; Redwine, D.; Gobbi, G.; Nuamthanom, A.; Rinaldi, P. L.
Macromolecules 2007, 40, 68796884.
7. Momose, H.; Hattori, K.; Hirano, T.; Ute, K. Polymer 2009, 50, 38193821.
8. Hu, J.; Krishnamurthy, K. Magn. Reson. Chem. 2008, 46, 683.
9. Nyber, N. T.; Duus, J. O.; Sorensen, O. W. J. Am. Chem. Soc. 2005, 127,
6154.
10. LeMaster, D. M.; Hernandez, G. Macromolecules 2000, 33, 35693576.
11. Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1968, 49, 17681777.
12. Morris, K. F.; Johnson, C. S., Jr. J. Am. Chem. Soc. 1992, 114, 31393141.
62
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 13. Antalek, B. Concepts Magn. Reson. 2002, 14, 225258.
14. Huth, J.; Kurur, N.; Bodenhausen, G. J. Magn. Reson., Ser. A 1996, 118,
286290.
15. Li, X.; McCord, E. F.; Baiagern, S.; Fox, P.; Howel, J. L.; Sahoo, S. K.;
Rinaldi, P. L. Magn. Reson. Chem. In press, DOI 10.1002/mrc.2763.
16. Rinaldi, P. L.; Baiagern, S.; Fox, P.; Howel, J. L.; Li, L.; Li, X.; Lyons, D.;
McCord, E. F.; Sahoo, S. K.; Twum, E. B.; Wyzgoski, F. J. Advanced Solution
2D-NMR of Fluoropolymers; Cheng, H. N., Asakura, T., English, A., Eds.;
ACS Symposium Series 1077; American Chemical Society: Washington,
DC, 2011; Chapter 21.
17. Espindola, A. P. D. M.; Crouch, R.; De Bergh, J. R.; Ready, J. M.;
MacMillan, J. B. J. Am. Chem. Soc. 2009, 131, 1599415995.
18. Zhou, Z.; Kummerle, R.; Stevens, J. C.; Redwine, D.; He, Y.; Qiu, X.;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch003

Cong, R.; Klosin, J.; Monanez, N.; Roof, G. J. Magn. Reson. 2009, 200,
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

328333.
19. Kupce, E.; Cheatham, S.; Freeman, R. Magn. Reson. Chem. 2007, 45,
378380.
20. Kupce, E.; Freeman, R. J. Magn. Reson. 2010, 206, 147153.

63
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 4

New Insights into Amorphous Macromolecules

and Their Mixtures from Advanced Magnetic
Resonance Experiments
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Jeffery L. White,*,1 Marcin Wachowicz,1 Lance Gill,1 Joshua Damron,1

and Justyna Wolak-Dinsmore2
1Department of Chemistry, Oklahoma State University,
Stillwater, OK 74078
2Department of Chemistry, North Carolina State University,

Raleigh, NC 27695-8204
(Current address: Liposciences, Inc., Raleigh, NC)
*E-mail: [email protected]

Amorphous blends of high molecular weight macromolecules

present some uniquely challenging problems in structure-
function investigations, and pose some interesting fundamental
questions regarding dynamics and thermodynamics. In the
absence of ordered morphologies or extensive isotopic labeling,
traditional diffraction and spectroscopic techniques may not
provide component-specific information. In this contribution,
we review experimental methods which have been developed to
specifically probe the dynamics and thermodynamics of mixing
in amorphous blends. Our approach is based on advanced
NMR techniques that provide chain-specific data over a wide
temperature range without introduction of any isotopic labeling
or probe molecules.

Introduction
The science related to polymer blends has historically enjoyed a productive
marriage between intellectual pursuit and economic utility, as recently discussed
in a national workshop hosted by the National Science Foundation (1). The
notion that the polymer blend area is a mature science can be challenged by
simply assessing the lack of understanding surrounding mixtures of amorphous

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 macromolecules of even the most simple chemical structure. As the polymer
science community, both from an intellectual and engineering motivation, pursues
increasingly complex systems, tailored composites, hybrid and hierarchical
structures, self-assembly, and biologically inspired materials as outlined in a
recent Macromolecules Perspectives article (2), we find that the same key
questions that characterize problems in mixtures of amorphous polymers continue
to arise in these emerging interdisciplinary areas. For example, simply measuring
with confidence the length scales within which two polymers are mixed in a
binary mixture of two non-crystalline macromolecules can be very challenging,
even for very simple polymer chain architectures. Also, what causes amorphous
mixtures to have a homogeneous, intimately mixed arrangement of its constituent
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

chains versus a heterogeneous, phase separated morphology? If intimate mixing

of the two polymer species occurs, does this mean that the chains assume new
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

identities relative to their pure component characteristics, and if so, in what

ways? Can we synthetically manipulate the macroscopic properties of the
end-result blend by systematic variation of chemical structure? Amorphous
polymer blends, which for the purposes of this contribution mean mixtures of
high polymers with no crystalline regions in the final blend, represent the first step
in generating complex polymer systems. As such, fundamental understanding of
how they behave, and connections between microscopic structure versus end-use
properties is paramount to continued progress in key areas of macromolecular
science, ranging from synthetic to engineered to biological materials. In addition,
the science of soft matter is growing as an interdisciplinary focal point for
researchers from physical, biological, and materials science areas (3), revealing
an increasingly apparent need for rigorous experimental techniques capable of
providing component-specific information in heterogeneous amorphous systems.
We note in passing that much of the same language that has pervaded the polymer
science community is now beginning to emerge in protein and enzyme science
(4).
In this review, we will discuss experimental strategies developed to
address key questions originally motivated by our interest in understanding
miscibility in technologically important polyolefin blends, which are composed
of completely saturated (i.e., only sp3 carbons and hydrogens) high-molecular
weight macromolecules. Since polyolefin blends have no chemical heterogeneity,
they are very difficult to study with chain-level specificity in the absence of
extensive isotopic labeling. In addition, they often are composed of chains with
similar dipole moments, or no permanent dipole at all, and as amorphous mixtures,
exhibit broad and diffuse glass transitions. Therefore, traditional methods like
dielectric spectroscopy and differential scanning calorimetry are not generally
applicable to amorphous polyolefin mixtures. As various polymer combinations
were examined, we found that the selectivity inherent to our experimental
approach based on variable-temperature CODEX5 solid-state NMR experiments
addressed many general questions in the current polymer blend literature related
to thermodynamics of mixing, differential chain dynamics, time-temperature
superposition, dynamic heterogeneity, and glass transition time-scales. Guiding
principles for the strategies discussed here include: [1] Non-invasive access to
quantitative chain-specific information before and after formation of mixtures;

68
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 [2] Data outputs in a format accessible to the general polymer science (i.e.,
non-NMR) community; [3] Experimental measures of average/mean chain
behavior as well as heterogeneity/distributions in chain behavior; [4] Quantitative
conclusions related to the thermodynamics of macromolecular mixing; [5]
General applicability to amorphous materials. These will be discussed in the
ensuing sections.

Experimental Section
Complete details for the synthesis, acquisition, and characterization of
the polymers have been described (69). In general, all polymers used in
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

our studies have molecular weights in the range 30,000 1,000,000, i.e.,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

synthetic high polymers well above the entanglement molecular weight.

The individual systems will be described in the text when appropriate, but
include polymers and various binary blends composed of atactic polypropylene
(aPP), head-to-head polypropylene (hhPP: made via anionic polymerization
of dienes and hydrogenation), polyethylene-co-butene (PEB: monodisperse
copolymer made via anionic polymerization and hydrogenation), polyisobutylene
(PIB-a commercial material), polyvinylethylene (PVE: 88% 1,2-polybutadiene
enchainment), and polyisoprene (PI: 97% cis-1,4 enchainment). Representative
structures for the repeat units are shown in Scheme 1.
All systems were studied in bulk, i.e., without any solvents. Miscible blends
(either equimolar or 50/50 wt%) were made by casting from solutions in a good
solvent, and were exposed to vacuum pumping for several days prior to use.
Characterization by TGA and 1H NMR revealed that no residual solvent existed
in any blend.

Scheme 1

In this review, we will primarily focus on an experimental strategy based

on variable-temperature CODEX (Centerband-Only Detection of EXchange)

69
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 experiments. However, several other experiments (including spin-diffusion
measurements to verify miscibility, GPC, and DSC) were completed prior to
CODEX analysis, as previously described (79). The CODEX experiment,
first described by Schmidt-Rohr and coworkers in 1999 (5), is ideally suited for
studying slow motions in macromolecules. Stated simply, it is a one-dimensional
chemical exchange experiment that relies on incomplete refocusing of a chemical
shift anisotropy (CSA) echo based on molecular reorientation during a mixing
time. Complete details have been explained previously (10, 11), but the process is
schematically represented by the pulse sequence shown in Figure 1 below which
is applied under magic-angle spinning (MAS) conditions.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 1. CODEX experiment pulse sequence applied under conditions of MAS.

The value of the exchange mixing time tm ranges from 0.05 to 0.20 seconds for all
data reported here. The total CSA evolution time corresponding to the sum of the
first and second recoupling periods was 2Ntr = (2)(4)(0.22 ms) = 1.76 ms. The tz
period, to ensure a constant time experiment, was 1 millisecond.

It is important to note that the any dynamics which occur during the mixing
time are events which are detected in real time, albeit indirectly based on their
attenuation of the chemical shift echo. Although we have typically used mixing
times tm values ranging from 0.050 0.20 seconds, in theory this value could
range widely, with an upper limit defined by spin-lattice relaxation. All 13C and
1H measurements were collected on a Bruker DSX-300 with field strength =

7.05T. The probe temperature was calibrated using PbNO3 to within 1 K. All
CODEX exchange data was acquired with an actively-controlled MAS speed
(typically 4-5 kHz), a 1-ms cross-polarization contact time, rotor synchronization,
and as a precaution, CODEX measurements were altered between the CODEX
and reference signal every 256 scans to eliminate spectrometer drift. In practice,
one acquires a reference spectrum by interchanging the constant 1-ms tz time
with the mixing time. Since this time value is too short relative to slow chain
dynamics, no motion can occur during such a short mixing period. This reference
experiment yields a signal of maximum amplitude (So), which is then compared to
the experiment where the much longer mixing time tm is used (S). The normalized
difference of the two experiments is referred to as the exchange intensity E =
(So S)/So, or S/So, which quantitatively reflects slow segmental dynamics.
The exchange intensity E is a direct function of the length of the mixing time
tm, the characteric correlation time constant for whatever type of motion occurs
(c), and the temperature T. The complete equations describing the quantitative

70
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 relationship between E and these variables have been described in detail in
previous publications (711). In all cases, 13C spectra were analyzed.
We are cognizant of concerns arising from possible inhomogeneous polymer
chain sampling based on experiments that begin with a cross-polarization step.
While our primary interests revolve around very slow backbone chain dynamics at
temperatures slightly below, near, and above the glass transition temperature (but
much lower than terminal dynamics region), we do not wish to preferentially select
chain segments that are much more rigid than the bulk. Rather, homogeneous
sampling of all polymer chains in mixtures is desired. To address this concern
in the context of amorphous polymer blends, we devised a modified version of
the experiment employing only direct carbon polarization as the initial step in the
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

experiment. Based on quantitative comparisons of the modified direct polarization

versus CP-based CODEX results over a wide temperature range (including Tg)
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

for atactic polypropylene (aPP), we demonstrated that results representative of all

polymer chains in the sample are obtained in both experimental approaches (12).
Therefore, we have confidence that CP-based CODEX-based exchange methods
provide chain-level information representative of the bulk mixing and miscibility
in amorphous macromolecules.

Results and Discussion

Using the experimental approach described above, one can extract exchange
intensities E(T) versus temperature for chain specific locations, i.e., backbone CH2
versus side-group CH3. Whenever possible, signals from carbons in the backbone
are used as reporters for slow segmental dynamics. In other cases, a methyl side
group might be advantageous in that the signal is well-resolved from other signals
once mixtures are formed. We will discuss examples from each approach in the
following sections.
Figure 2 shows systematic comparisons of CODEX exchange intensities
measured from backbone CH2 versus side-group CH3 in the two pure polymers
(PEB and aPP) over the entire temperature range for which a measurable
exchange signal is detected. This is an important control experiment, since
it eliminates any uncertainty associated with additional side group dynamics
that might influence the interpretation of the CODEX results in the blend, and
their relevance to slow segmental dynamics. The onset of detectable exchange
intensity for either functional group, as well as the temperature of the maximum
E(T) value, is identical within each polymer. The absolute value of the maximum
E(T) is markedly different for the backbone CH2 vs. side-chain CH3 signals in
the PEB-66 polymer, indicative of additional ethyl branch motions which further
reduce the magnitude of the chemical shift anisotropy for that pendant methyl
group relative to backbone moeties, thereby decreasing E(T) values compared
to the backbone CH2. The exchange intensity for this CH3 group in PEB-66 did
increase, as expected, in experiments with longer recoupling times (not shown
here). Since only a single carbon-carbon bond separates the CH3 group from the
main-chain in aPP, this dramatic difference in exchange intensity relative to the
backbone is not observed. Two important points from this control experiment are:
71
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (1) the temperature of the maximum E(T) value is independent of which group is
measured, which means that the CH3 signals can accurately report conformational
exchange in the blend, an advantage given that they are better resolved than their
respective CH or CH2 backbone counterparts and can be deconvoluted accurately;
(2) the onset of detectable E(T) signal in the CODEX experiment agrees with
DSC data, in that the first one or two data points on the low temperature side
of each curve coincide with the DSC Tg (PEB66 = 219 K and aPP = 262 K).
Finally, in all cases a decrease in E(T) amplitude occurs when temperatures are
high enough that thermally activated conformational rearrangements occur with
a frequency larger than the magnitude of the CSA; this eliminates the possibility
of detecting any signal differences between the exchange and reference spectrum
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

since there is no longer orientation-dependent shielding information preserved in

the system due to motional averaging.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 2. Normalized exchange intensities E(T) for methyl and methylene signals
of pure PEB66 (a copolymer of ethylene and 66 wt% 1-butene) and pure aPP.
The solid lines are Arrhenius model fits (see below) to the data, whereas the
dash-dot lines are drawn through the methylene experimental points only to
guide the eye. For reference, the Tg ranges via DSC are shown at the top of the
figure. (Reproduced from reference (8))

From these example results, we observe that the CODEX experiment

under MAS conditions provides dynamic information for slow chain
reorientation/segmental motions on timescales similar to that probed by DSC
methods, the latter being most familiar to practicing polymer scientists. However,
as we will demonstrate in subsequent sections, and as has been described in detail
in references 7-9, the overall information content and the chain specificity in
mixed systems are much higher with the CODEX approach.
Figure 3 shows temperature dependent exhange intensities E for
the same polymers described in Figure 2, and their miscible blend. We
will use this miscible aPP/PEB-66 blend to illustrate the wide range of
information (both model-independent and model-dependent) accessible using
72
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the variable-temperature CODEX approach, and then conclude with brief
representative results from other amorphous polymer blends. The data in Figure 3
represent the outcome of three independent temperature dependent experiments;
one each for the two pure polymers and a single experiment for the blend from
which polymer specific exchange intensities were extracted at each temperature.
This latter point is important. In contrast to other experimental approaches, a
single experiment on a mixed polymer system can provide data for all components
in the mixture, which eliminates sample preparation uncertainties/reproducibility
that often plague sequential methods where only one component may be analyzed
in any one experiment (e.g., selective deuteration, dielectric spectroscopy, etc.)
Prior to any consideration of a quantitative model to fit the results, several
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

key points may be discerned via simple inspection of the raw data in Figure 3,
comparing the response of the 13C signals in the CODEX experiment for each
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

pure polymer to the response of that same polymer in the blend. First, the PEB66
exchange intensity curve shifts to higher temperature in the blend relative to pure
chains (PEB66 is the lower Tg component). Similarly, the exchange intensity
curve shifts to lower temperature for the aPP in the blend compared to its pure
response. For the PEB66, the E(T) maximum shifts from 227 K to 247 K upon
blend formation, while the aPP E(T) maximum changes from the pure value of
273 K to 253 K in the blend. Although each curve exhibits an identical 20 K
change, they do not converge to an identical common value (5-7 K difference
in E(T) maxima) even though the chains are intimately mixed. While omitted
from the figure in order to maintain clarity, the E(T) versus T curve for the
backbone CH2 peak of aPP in the blend has a maximum at the same 252-253 K
position as the CH3 peak shown in Figure 3. Secondly, the breadth of each E(T)
curve increases for either component in the blend relative to the pure polymer,
especially for aPP. Finally, the absolute value of E(T) at each temperature across
the detectable range decreases in the blend relative to the unmixed result for
both polymer components. Although each temperature dependent exchange
intensity curve decreases in intensity and increases in breadth for the polymers
in the blend compared to the pure polymers, the overall integrated area under
the curve fits (vide infra) remains constant for each polymer, within the error
of the data analysis. While the intermediate temperature values for the E(T)
curves are reminiscent of DSC results on blends, the ability to extract these
specific details for each polymer in an amorphous mixture by simultaneous
detection of the two unique E(T) curves is difficult using traditional thermal
analysis methods. Typically, DSC traces on multicomponent miscible blends
are broad and featureless, and one cannot discern individual behaviors for the
polymer components. We conclude from these points that the overall dynamic
heterogeneity for both polymer chains increases in the blend, and in addition,
it is also known that the CODEX exchange intensity decreases with increasing
number of sites involved in the exchange process for a fixed recoupling and
mixing time (11). The details specifying exactly how the dynamic heterogeneity
increases for both chains will be discussed in the following sections, and the
reader can consult references 7-9 (and references therein) for additional details.

73
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 3. Normalized exchange intensities E(T) for pure PEB66 ( ), PEB66 in

the blend (), pure aPP (), and aPP in the blend (). The smooth lines are fits to
the data using an Arrhenius model as described in the text. Note shifts of equal
but opposite magnitude for each component upon blend formation, but a lack
of complete convergence to a common temperature for the exchange intensity
maximum of each polymer in the blend. The 5K/min scan DSC Tg range for each
polymer and the blend is plotted for reference on the top of the figure, with the
box length representing the beginning and end of the endotherm. (Reproduced
from reference (8))

Historically, the polymer science community espouses a miscible blend is

characterized by identical Tgs for each polymer component in the miscible blend.
While this can be true, it should not be taken as a criterion for miscibility, since
Figure 3 above and Figure 8 (see below for further discussion of the PI/PVE
miscible blend) clearly demonstrate that unique glass transitions (or effective
Tgs) can occur for constituent polymers in a miscible blend. Our results are
in agreement with recent publications by Lodge discussing inequivalent Tgs for
the polymers in miscible blends (13, 14).
The fits to the experimental data shown in Figures 2 and 3 were obtained
by combining isotropic rotational diffusion with a temperature-dependent discrete
log Gaussian correlation time distribution/Arrhenius model (1519). The absolute
value of the exchange intensity E(T) at each temperature, for a fixed recoupling
and mixing time, depends on the correlation time constant characteristic of the
motion modulating the chemical shift anisotropy as well as the distribution of
correlation time constants for all of the segments in the amorphous polymer or
polymer mixture. We use a linear model that relates the width of the correlation
time distribution to increasing temperature (20). Key results that can be obtained
from the analyses include central correlation time constants for slow backbone
reorientations in the pure and mixed systems (beginning from slightly below the
glass transition to higher temperatures, but still in segmental dynamics regime), the
74
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 width of the correlation time constant distribution in both pure and mixed systems,
and activation energies for segmental dynamics in the neat and blended polymers.
We recognize that other temperature dependent models may be more familiar
to the polymer scientist. Figure 4 shows a comparison of the correlation time
distribution/log Gaussian/Arrhenius model discussed above with a KWW/WLF
analysis for the exchange intensity data from pure aPP and aPP in the blend; this
is the same aPP raw data shown in Figure 3. The KWW/WLF fitting parameters
are reported in the captions to Figures 4 and 5; we observe excellent agreement
between the two models in terms of correlation time values over the temperature
range of our data. Such low values upon blend formation are consistent
with increased dynamic heterogeneity in aPP,11 as is apparent from direct
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

inspection of the E(T) exchange curve intensities in Figure 3 and 4. Detailed

comparisons of KWW values to corresponding values of the correlation time
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

distribution widths g() are provided in the Supporting Information to Reference

(8). That equivalent correlation times for the center of the distribution (c) are
obtained using either model is more clearly evident by the representation of their
temperature dependence for both polymer components in Figure 5. One observes
that the temperature dependence of the slow chain dynamics for PEB66 and aPP
are very different. Figure 5 also shows that while the magnitude of the change in
c values upon blend formation differs between the two polymers significantly at
any temperature (ca. 5 decade decrease in aPP versus ca. 2.5-3 decade increase
in PEB66), the two polymer components in the blend have identical values of c
near 250 K.

Figure 4. Comparison of fits to the aPP exchange data obtained using two
different models: (1) Arrhenius temperature dependence of correlation times with
variable-width log-Gaussian distribution, and (2) WLF temperature dependence
combined with KWW distribution. The WLF/KWW parameters for pure aPP were
C1 = 15.5, C2 = 41 K, (Tg) = 100 s, Tg = 259 K, = 0.8, whereas for aPP in
blend Tg = 237 K and = 0.2 was used. (Reproduced from reference (8))

75
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 5. Temperature dependence of correlation times obtained using Arrhenius

and WLF models for the aPP/PEB66 miscible blend. The WLF parameters for
pure aPP were: C1 = 15.5, C2 = 41 K, (Tg) = 100 s, Tg = 259 K, whereas for
aPP in blend Tg = 237 K was used. For pure PEB66 we used C1 = 15.5, C2
= 55 K, (Tg) = 100 s, Tg = 210 K, and the best fit for PEB66 in blend was
obtained with C2 = 68 K and Tg = 224 K. An Arrhenius model can be treated
as linear approximation to the WLF curve, since both Arrhenius and WLF
models give similar results over the temperature range for slow motions near
Tg. (Reproduced from reference (8))

Figure 5 also clearly demonstrates that quantitative Tg timescales are revealed

by these experiments, both for the pure polymers and the polymers in the blends.
The correlation time constants for Tg in pure PEB and pure aPP both appear to
be between 10 and 100 seconds (PEB near 10 and aPP nearer 100), but most
importantly the data indicates that the change in Tg timescales, as well as their
absolute values, for either component once the miscible blend is formed are quite
different outside of a very narrow temperature range (250-260K).
Figure 6 shows calculated correlation time distributions from the fits to data
in Figures 2 and 3 for the aPP component, demonstrating how the correlation
time distribution function g() can influence E(T) (21). We note how much
broader the correlation time distributions become upon formation of the blend,
and also, the increased distribution width near the Tg value (low T) for a
pure polymer. By comparing Figures 5 and 6, we immediately observe that
while central correlation time constants converge for each component in the
miscible blend, albeit changing by largely different values, that the correlation
time distributions actually diverge. Figure 6 shows that aPP, i.e. the high Tg
component, becomes much more dynamically heterogeneous relative to its pure
state, which is a common result for binary blends we have examined to date.
Indeed, the total dynamic heterogeneity in the blend is much larger than the
sum of the two unmixed polymer components, an important concept as one
considers the thermodynamics of mixing in amorphous macromolecules (vide
76
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 infra). The reader may consult references 7-9 to see more detailed representations
of correlation time distributions, and different schematics depicting their width as
a function of temperature for pure and mixed states.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 6. Temperature dependence of correlation time distribution widths for

pure versus blended aPP and PEB66. (Reproduced from reference (8))

For brevity, only the variable temperature CODEX exchange curves are
shown for two additional blend systems in Figures 7 (PIB/hhPP) and 8 (PI/PVE),
along with fits to the raw data using the techniques described above. PIB/hhPP
is unique among miscible blends we have studied to date in that the CODEX
exchange curves do converge to the same value (as defined by onset of detectable
exchange intensity, or more definitively by the maximum in the curve). If one
determines the central correlation time constant for the temperature corresponding
to the maximum in the E(T) curve, an identical value is observed for each polymer
component in the blend. However, the change in segmental correlation time
constants upon formation of the miscible blend is again much larger (by several
orders of magnitude) for the high-Tg component hhPP than the low-Tg PIB
component. Interestingly, the temperature where maximum exchange intensity
occurs, as indicated by the vertical arrows in Figure 7, is 5-7 degrees lower than
predicted by Gordon-Taylor/Fox equation mixing rules. In other words, it does not
agree with composition-weighted average values, and qualitatively suggests an
entropically driven mixing. Quantitative calculations based on central correlation
time constants determined experimentally for the pure and mixed polymers, using
the approach described here, confirm that positive configurational entropies of
mixing exist for this blend, as well as other miscible polyolefin blends we have
examined to date.

77
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 7. Normalized exchange intensities E(tm) for a) pure PIB measured at the
quaternary backbone carbon () and PIB in the blend (o), and b) pure hhPP ()
and hhPP in the blend (). Note the common exchange maximum temperature
T=246 K for each polymer in the blend (left arrow) and the Fox equation
prediction (right arrow). The smooth lines are fits to the data as described in the
text. (Reproduced from reference (7))

Figure 8 shows exchange curves for the PI/PVE miscible blend system. Note
the apparent similarities between the general shape/location of the four curves and
those in Figure 3, in that the curves for the polymers in the miscible blend do
not converge to the same values (unlike the PIB/hhPP system in Figure 7). In
total, these three results suggest that one cannot assume how slow segmental chain
dynamics, which are the macromolecular dynamics most important for mechanical
properties, change when a miscible blend is formed. Stated differently, one cannot
assume that (a) a common glass transition will occur for the polymers in a miscible
blend, and (b) composition-weighted averaged chain behavior takes place.

78
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 8. Normalized 200-ms exchange intensity E(T) curves for pure PI and
PVE and the same components measured independently in the blend. All data
shown here reflect only backbone aliphatic carbons, not side chains or olefinic
groups. The solid lines are fits to the data using an Arrhenius model with a
variable-width log-Gaussian correlation time distribution. The dashed lines are
fits to the data using the WLF/KWW model. The thick gray lines at the top of the
figure indicate the 10 K/min DSC Tg range for the two pure polymers and the
miscible blend. (Reproduced from reference (9))

Complete quantitative analysis, using both Arrhenius and WLF/KWW

models, of the data in Figures 7 and 8, including effective Tgs, central correlation
time constants, correlation time distributions, and activation energies are reported
in references (7) and (9). It is worth mentioning here that the PI/PVE system was
an important control case as this blend has been studied by multiple investigators
using a wide variety of techniques (2227). However, as ours was the first
method to get chain-selective information non-invasively from a single set of
experiments on a single blend sample, we were pleased to find good agreement
between quantitative data from our approach and selective deuteration techniques
involving multiple blend preparations, as previously described (26, 27). In
addition to central correlation time constants, correlation time distributions, and
activation energies, segmental friction coefficients were determined for each
component in the PI/PVE system using a KWW/WLF treatment of the central
correlation time data, where the characteristic segment length was taken as equal
to the Kuhn length, or 1.1 nm.

79
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 During this review, we have deliberately pointed out model-independent and
model-dependent conclusions accessible via the variable-temperature CODEX
approach. Given that many of the key quantitative results come from fitting the
raw E(T) data, it is appropriate to consider a mechanism by which the central
correlation time constants c can be evaluated independently.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 9. Exchange time constants extracted from an exponential fit to the

rising exchange intensity curves are indicated on each of the Figures 9a-d for
PVE, atactic PP (PP), head-to-head PP (hhPP), and polyisobutylene (PIB),
respectively. For each of these same four polymers, at the same temperature as
indicated on each plot, the respective central correlation time constants obtained
from full analysis of the variable temperature exchange intensity curves of the
type represented by the data in Figures 3, 4, 7, and 8, are: (a) c = 14 ms (b) c =
24 ms (c) c = 20 ms (d) c = 25 ms. The temperatures indicated in a-d are equal
or very nearly equal to the exchange intensity maximum temperature for each
polymer, resulting in similar values of c. (Reproduced from reference (9)).

In Figure 9, we show results from such an independent evaluation, in which

cs were determined at temperatures near the maxima in the E(T) curves, using
variable mixing time experiments for four different pure polymers including
PVE, atactic polypropylene (PP), head-to-head polypropylene (hhPP), and
polyisobutylene (PIB) at the indicated temperatures. The correlation time
constants, extracted from an exponential fit to the rising intensity curve, are
indicated on each of the Figures 9a-d, while values extracted from fitting the
80
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 CODEX data are listed in the Figure 9 caption. (In figures where there are multiple
data points, either multiple carbon positions in the polymer chain were probed, or
as in Figure 9d, two different recoupling times in the CODEX experiment were
compared.) The results in Figure 9 indicate that the correlation time constants
obtained by full analysis of the raw data of the type shown in Figures 3, 4, 7, and
8 are accurate, with the largest difference for the polymers shown equal to 12%
for PVE, and significantly less in the other three cases. While the time required to
generate the experimental data in Figure 3, 7, or 8 is not trivial, such an approach
is significantly shorter (by a factor of 5 to 10) than the time required to obtain
full variable mixing time exchange curves, like those shown in Figure 9, over
wide temperature ranges. This control experiment, as well as comparisons of
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

PI and PVE data to previously reported correlation time values (9), indicate the
quantitative robustness of our approach for macromolecular mixtures.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Given that we can calculate central correlation time constants characteristic of

slow segmental reorientation for pure and mixed polymers over a wide temperature
range, we can use the Adams-Gibbs formalism to quantitatively assess entropy
changes that occur upon formation of a miscible blend (28). The relationship
between configurational entropy Sc and c is:

where we equate c the value of the correlation time at the center of the distribution
for either a pure polymer or the same polymer in a blend. Using data of the type
shown in Figure 5, and assuming o ranging from 10-12 s to 10-15 seconds and c
as constant for each polymer in pure versus mixed state at a fixed temperature
(usually taken as the temperature corresponding to maximum exchange intensity),
we have determined that the total change in configurational entropy for PIB/hhPP
and aPP/PEB-66 blends ranges from +10 to +20% upon formation of the miscible
blend (and relative to the unmixed components). Therefore, one must conclude
that an increase in the number of surface contacts either between dissimilar
polymer chains or within chains themselves (an enthalpic model) in miscible
polyolefin blends simply does not exist relative to the unmixed pure polymers.
We view this as a conservative limit, since it does not quantitatively capture the
large increases in the distribution of correlation times, or dynamic heterogeneity,
associated with forming the miscible blend. The PI/PVE system is somewhat
different in that it possesses sp2 functional groups, and therefore free electron
density which increases polarizability and induced association. Even so, Figure
10 clearly demonstrates that there is a slight positive entropy of mixing associated
with forming the miscible blend, relative to the sum of the configurational entropy
for the two unmixed components. The ability to non-invasively, selectively, and
quantitatively determine this kind of data for amorphous polymers in a miscible
mixture is a key advantage of this experimental approach.

81
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

Figure 10. Calculated values of the ratio of the total configurational entropy in
the miscible blend to the sum of the pure component entropies (Sc Blend / Sc
Pure), expressed in percent, versus the position on the exchange intensity curves
shown in Figure 8 moving from low to high temperature. Each trace represents a
calculation across the entire temperature range of the curve.

Conclusions
A chain-specific experimental approach based on variable-temperature
solid-state CODEX NMR experiments reveals that the effective glass transitions
for each chain type in miscible blends may be inequivalent, and slow segmental
dynamics for each polymer in the blend are characterized by unique central
correlation times and unique correlation time distributions. Quantitative
analyses of the raw data indicate that good agreement exists between effective
Tgs, central correlation time constants, correlation time distributions, and
friction coefficients extracted from this approach versus those obtained by
other well-documented methods. Results from an isotropic rotation diffusion
model with Arrhenius/log-Gaussian or WLF/KWW treatments of temperature
dependence show clear sensitivity to changes that occur upon blend formation
relative to the unmixed components. Positive configurational entropies of
mixing are detected experimentally. That such quantitative information may be
obtained for either polymer component in an amorphous mixture, without isotopic
labeling, electric dipole moment constraints, or introduction of probe molecules,
is a unique advantage of this experimental strategy and illustrates applicability
to a wide range of multicomponent macromolecular systems beyond miscible
blends, including polymer nanocomposites, organic/inorganic hybrids, biological
macromolecules, and block copolymers.

82
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Acknowledgments
The principal investigator gratefully acknowledges support from the National
Science Foundation Division of Materials Research through grant DMR-0756291
and DMR-0611474.

References
1. National Science Foundation Report. Interdisciplinary Globally-Leading
Polymer Science and Engineering, 2007, available via the Web at
http://www.nsf.gov/mps/dmr/reports.jsp.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

2. Ober, C. K.; Cheng, S. D.; Hammond, P. T.; Muthukumar, M.;

Reichmanis, E.; Wooley, K. L.; Lodge, T. P. Macromolecules 2009, 42, 465.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

3. Donald, A. M. MRS Bulletin 2010, 35, 702.

4. Frauenfelder, H.; Chen, G.; Berendzen, J.; Fenimore, P. W.; Jansson, H.;
McMahon, B. H.; Stroe, I. R.; Swenson, J.; Young, R. D. Proc. Nat. Acad.
Sci. U.S.A. 2009, 106, 5129.
5. deAzevedo, E.; Hu, W. G.; Bonagamba, T. J; Schmidt-Rohr, K. J. Am. Chem.
Soc. 1999, 121, 84118412.
6. Wolak, J. E.; Jia, X.; White, J. L. J. Am. Chem. Soc. 2003, 125, 13660.
7. Wachowicz, M.; White, J. L. Macromolecules 2007, 40, 5433.
8. Wachowicz, M.; Wolak, J. E.; Gill, L.; White, J. L. Macromolecules 2008,
41, 2832.
9. Wachowicz, M.; Gill, L.; Damron, J.; White, J. L. Macromolecules 2010, 43,
3903.
10. deAzevedo, E.; Hu, W. G.; Bonagamba, T. J; Schmidt-Rohr, K. J. Chem.
Phys. 2000, 112, 89889001.
11. deAzevedo, E. R.; Tozoni, J. R.; Schmidt-Rohr, K.; Bonagamba, T. J. J.
Chem. Phys. 2005, 122, 154506.
12. Wachowicz, M.; Gill, L.; White, J. L. Macromolecules 2009, 42, 553.
13. Lodge, T. P.; McLeish, T. C. B. Macromolecules 2000, 33, 5278.
14. Haley, J. C.; Lodge, T. P. J. Rheology 2005, 49, 1277.
15. Wefing, S.; Kaufmann, S.; Spiess, H. W. J. Chem. Phys. 1988, 89, 1234.
16. Wind, M.; Brombacher, L.; Heuer, A.; Graf, R.; Spiess, H. W. Solid State
Nucl Magn. Reson. 2005, 27, 132.
17. Kaufmann, S.; Wefing, S.; Schaefer, D.; Spiess, H. W. J. Chem. Phys. 1990,
93, 197.
18. Saalwachter, K.; Fischbach, I. J. Magn. Reson. 2002, 157, 17.
19. OConnor, R. D.; Ginsburg, E. J.; Blum, F. D. J. Chem. Phys. 2000, 112,
7247.
20. deAzevedo, E. R.; Reichert, D.; Vidoto, E. L. G.; Dahmouche, K.;
Judeinstein, P.; Bonagamba, T. J. Chem Mater 2003, 15, 2070.
21. See references 79 for discrete versus continuous plots illustrating this
relationship.
22. Hefner, S.; Mirau, P. A. Macromolecules 1994, 27, 7283.
23. Roovers, J.; Toporowski, P. M. Macromolecules 1992, 25, 1096.
24. Roovers, J.; Toporowski, P. M. Macromolecules 1992, 25, 3454.
83
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 25. Alegria, A.; Colmonero, J.; Ngai, K. L.; Roland, C. M. Macromolecules
1994, 27, 4486.
26. Chung, G. C.; Kornfield, J. A.; Smith, S. D. Macromolecules 1994, 27, 964.
27. Chung, G. C.; Kornfield, J. A.; Smith, S. D. Macromolecules 1994, 27, 5729.
28. Adam, G.; Gibbs, J. H. J. Chem. Phys. 1965, 43, 139.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch004

84
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 5

Solid-State NMR Investigations of

Semi-Crystalline PVIBE/-PL and PK/PA
Blends: Crystallite Size, Type, and Morphology
Related to Physical Properties
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Atsushi Asano*

Department of Applied Chemistry, National Defense Academy,

1-10-20 Hashirimizu, Yokosuka, 239-8686 Japan
*E-mail: [email protected]

Crystallinity and the size of crystalline phase of poly(-L-lysine)

(-PL) were determined by solid-state 13C NMR . The decrease
in the melting temperature with increasing -PL content
ascribed to the crystalline phase of -PL in poly(vinyl isobutyl
ether) (PVIBE)/-PL blends was interpreted by the decrease
in total lamellar thickness in the crystalline phase of -PL.
Furthermore, the independence of the melting temperature with
-PL content for PVIBE/-PL/saponite-clay nanocomposites
was also explained in a similar way. For polyketone/nylon
6 (PK/PA) blends, the high impact resistance under the wet
condition was explained by the characteristic morphology of
PK crystalline phase, the increased elasticity due to water
absorption in the amorphous phase of PA, and conversion of
crystallite to the crystallite morphology of PA (as shown by
solid-state 15N NMR).

Introduction
Polymer blends and nanocomposites have been widely used for engineering
purposes. Their favorable mechanical and physical properties are closely related to
their morphology (especially in semi-crystalline polymers), the domain size of the
crystalline phase, and the crystalline type. Hence, the morphology of crystalline
phase and the degree of crystallinity are among the important information

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 that we have to know to understand the physical and chemical properties of
semi-crystalline polymers, their polymer blends, and their nanocomposites. To
obtain the degree of crystallinity, we can choose any appropriate method, such as
differential scanning calorimetry (DSC), X-ray diffraction, Raman and infrared
(IR) spectroscopies, and nuclear magnetic resonance (NMR) (1, 2). Among these
methods, NMR can detect the crystalline phase of a component polymer even
in a blend or a nanocomposite distinguishably and also estimate correctly the
amount of microcrystalline phase (35). It is valuable to analyze the domain
size, type, and morphology of crystalline and non-crystalline phases (including
inter-phase and amorphous phase) separately to reveal the relationship between
morphological information and physical properties, such as melting temperature,
impact strength, modulus, and elasticity.
In this review, I shall first discuss the determination of crystallinity via
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

NMR parameters, especially for poly(-L-lysine) (-PL). There are several ways
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

in NMR to study the crystalline phase, such as 13C spin-lattice relaxation time
(T1C), 1H spin-spin relaxation time (T2H), and 13C NMR spectral editing. We shall
compare these parameters with the results from X-ray diffraction (XRD) (3).
Secondly, I shall discuss the effects of inorganic clay fillers on crystallinity and
melting temperatures of -PL and poly(vinyl isobutyl ether) (PVIBE) blends, and
their nanocomposites with saponite clay (6, 7). The melting temperature of -PL
measured from differential scanning calorimetry (DSC) showed a dependence
on composition in the PVIBE/-PL blends, but compositional independence
in the PVIBE/-PL/saponite nanocomposites. These interesting phenomena
were explained by the domain size of the crystalline phase in -PL, i.e., the
Gibbs-Thomson effect (1). Finally, I shall discuss the analysis of the morphology
of polyketone/polyamide (PK/PA) blends. The high impact property of the PK/PA
blend with water absorption is due to the effects of change of PA crystalline
phase type and the increase of elasticity of PA amorphous phase under wet
conditions. Furthermore, transmission electron microscopy (TEM) shows a
characteristic crystalline lamella-like picture of PK. The 1H spin-diffusion study
via well-resolved 13C NMR signals indicates that both the domain sizes of PK
and PA are not affected by the moisture absorption (810).

Experimental Section
Materials
Poly(-L-lysine) (-PL: repeat unit -NH(CH2)4CH(NH2)CO-) was provided
by Chisso Corporation as a solid powder, with an average degree of polymerization
approximately 30. -PL is known as a biodegradable hydrophilic semi-crystalline
polymer and nontoxic toward humans and environment (11). Semi-crystalline
poly(vinyl isobutyl ether) (PVIBE: repeat unit -CH2CH(OCH2CH(CH3)2-) was
obtained from Scientific Polymer Products, Inc, and its Mw is 600,000. Polyketone
(PK, an ethylene/propylene/CO copolymer -(CH2CH2CO)m-(CH2CH(CH3)CO)n-)
is a commercial material (Carilon D26HM100), provided by Shell Co.
Polyamide-6 (PA; repeating unit -NH(CH2)6CO-) is a commercial material
(Amilan CM1017), provided by Toray Ind. Both PK and PA are semi-crystalline
86
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 polymers. Saponite clay, which is synthetic and a smectite clay, was obtained
from Kunimine Industries Co., Ltd. It has a relatively small aspect ratio when
compared to montmorillonite and has no paramagnetic center such as Fe3+.
The PVIBE/-PL blend and PVIBE/-PL/saponite nanocomposite films
were prepared by casting 90:10 (v/v) chloroform/methanol mixed solutions that
contained those polymers at a concentration of 15 (w/v) % on a Teflon plate at
313K, which were further dried under vacuum at 313K for 1 or 2 days. The film
thus obtained was opaque and elastic. In order to exfoliate the saponite clay layers,
the saponite clay was dissolved in -PL/water solution at 3 wt% concentration
prior to mixing with PVIBE (-PL/saponite-clay=1/0.03). The -PL/saponite-clay
mixed powder was obtained by drying the -PL water solution.
The PK/PA blends were obtained by mechanical mixing. We used a simple
terminology for the PK/PA blend: PK60PA40 means a blend of 60 weight %
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

PK (volume fraction 0.578) and 40 weight % of PA (volume fraction 0.422).

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

The PK/PA blends were moisture-conditioned by holding them at 50% or 95%

relative humidity (RH) and a temperature of 296 K for three weeks or longer. Dry
test specimens were prepared by drying them in vacuum at 373K for 72 h. The
moisture absorption rate was measured by the Karl Fisher method.

Measurements and Instruments

Nuclear Magnetic Resonance (NMR)

High-resolution solid-state 13C NMR spectra were obtained using a

Varian 400WB NMR spectrometer or a Bruker DMX500 spectrometer. The
ramped-amplitude cross-polarization (CP) (12, 13) and magic-angle spinning
(MAS) with 1H high-power dipolar decoupling technique was used, employing
a rotor with 4.0 mm diameter for both NMR systems. The radio-frequency field
strength for 1H decoupling was 120 kHz for the Varian and 56 kHz for the Bruker
instrument, and the two-pulse-phase-modulation (TPPM) method (14) was used.
The MAS frequencies used were 10 to 16 kHz. 13C chemical shifts were measured
relative to tetramethylsilane (TMS) using the benzene carbon signal at 132.07
ppm for solid hexamethylbenzene or 29.47 ppm for solid adamantane as an
external reference. High-resolution solid-state 15N NMR spectra were measured
using the Varian NMR 400WB spectrometer at a MAS rate of 5 kHz and a 1H
decoupling frequency of 65 kHz with a 7.5 mm diameter rotor. 15N chemical
shifts were measured relative to the glycine signal as an external reference at 0
ppm.
The 1H spin-lattice relaxation times in the laboratory frame (T1H) were
indirectly measured from well-resolved 13C signals enhanced by 500 s to 2
ms CP contact times. These parameters were chosen to obtain the maximum
enhanced signal, applied after the inversion-recovery pulse sequence for the
1H nucleus. T1C decay was indirectly measured from protons through the CP

enhancement as proposed by Torchia (15).

87
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Differential Scannning Calorimetry (DSC)

DSC curves were observed using a Perkin-Elmer 7 instrument with the

temperature increasing at a rate of 2 Kmin-1 from 233 K to 473 K.

Transmission Electron Microscopy (TEM)

TEM images were collected using a Hitachi H-800 instrument with an

accelerating voltage of 200 kV. Electron staining was done with phosphotungstic
acid. Three-dimensional TEM images were examined using a FEI Tecnai G2 F20
TEM with a tilting angle that ranged from -70 to +70 for every 2 step.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

X-ray Diffraction (XRD)

XRD patterns were collected using MAC Science M21X diffractometer with
Cu K radiation of = 0.154 nm at 293 K. For the XRD experiments, we prepared
a sample of -PL which was mostly non-crystalline. A -PL sample sealed in a
glass tube with nitrogen atmosphere was heat treated at 458 K for 20 min: the
crystalline phase was completely melted. After the heat treatment, the glass tube
was immediately immersed in liquid nitrogen to quench the amorphous phase.

Results and Discussion

Crystallinity and Crystal Size (Gibbs-Thomson Relationship)
Figure 1 shows the observed 13C CPMAS NMR spectra of (a) PVIBE, (b)
PVIBE/-PL=10/3, (c) PVIBE/-PL/saponite=10/3/0.09, (d) -PL, and (e) -PL/
saponite-clay=1/0.03. The expanded spectra for the CH-O-CH2 region of PVIBE
and CH region of -PL are also shown in (b) and (c). Furthermore, the spectra
for the carbonyl carbon region of -PL are depicted in the left-hand side of (b)
to (e) and the peak assignments are also shown in the figure. Although the peak
intensities are enhanced individually by CP, we determined the real composition
of PVIBE/-PL blends and nanocomposites via 1H dipolar-decoupling (DD)-MAS
spectra. The spectrum of the -PL/saponite-clay=1/0.03 (e) shows that the relative
ratio of broad lines is much more than that of -PL (d). For example, the CH2 peaks
at 20 to 40 ppm region, which consists of apparently four narrow peaks plus broad
lines, show much broader features. This suggests that the degree of crystallinity
is decreased by adding saponite clay. Similarly, the spectrum of PVIBE/-PL/
saponite=10/3/0.09 nanocomposite (c) shows more broadening as compared to that
of PVIBE/-PL=10/3 (b). The broadening can be particularly seen at the CH peak
of -PL from 50 to 65 ppm. This is ascribed to the effect of adding clay, too. For
the carbonyl (C=O) peak at ca. 178 ppm, the tendency is clearly observed; a sharp
doublet peak of the crystalline (CR) phase (d), which is observed on top of the
broad peak attributed to the non-crystalline (NC) phase, becomes small by adding
saponite clay (e). Interestingly such a significant broadening does not appear in
88
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the PVIBE/-PL=10/3 blend (b); however, the sharp doublet peak was obscured
after blending with both PVIBE and saponite clay (c). However, the CH-O-CH2
region of PVIBE from 65 to 85 ppm does not show such a broadening by adding
saponite clay, as shown in (b) and (c).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 1. Observed 13C CPMAS NMR spectra of (a) PVIBE, (b)

PVIBE/-PL=10/3 blend, (c) PVIBE/-PL/saponite=10/3/0.09 nanocomposite,
(d) -PL, and (e) -PL/saponite-clay=1/0.03 powder, respectively.

To investigate the degree of crystallinity, we examined the 13C CPMAS NMR

spectra separated into contributions from the CR and NC phases on the basis of the
differences in the intrinsic 1H spin-lattice relaxation time in the rotating frame that
characterize CR and NC phases (4, 16). This method provides the same quality
for the degree of crystallinity as well as XRD measurement for each component
even in polymer blends. Furthermore, it can be examined at room temperature
basically and usually. For the T1C measurement, it is not easy to select appropriate
89
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 temperature for both component polymers. The differences among NMR, XRD,
and T1C measurements for detection of the crystallinity are briefly discussed later.
Figure 2(A) shows an example of the spectral separation to the CR (dotted line,
sharp component) and the NC (broken line, broad component) signals for the CH
peak of pure -PL (3) as shown in Figure 1(d). The dotted and broken lines are
the results of linear combinations of the spectra obtained from the different spin-
locking times of 1 s and 5 ms prior to the CP contact time of 800 s, respectively.
In order to obtain each line, we have attempted to null the NC and CR signal
contributions, respectively (4, 16). The solid line spectrum is obtained by the sum
of the spectra of the CR and the NC signals.
It is noteworthy that both CR and NC spectra depend on the difference in
chain order only and not on molecular motion. Since the peak intensities decrease
during CP irradiation depending on the contact time, it is necessary to obtain the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

efficiency factors of CP enhancement to estimate the correct crystallinity using

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

both CR and NC integrals. We examined the CP dependency with various CP

contact times from 25 s to 7 ms to obtain the relative CP efficiencies for both CR
and NC phases at a CP contact time of 800 s.
The estimated degree of the crystallinity for pure PVIBE was ca. 20 % and for
pure -PL ca. 54 %. These values have an experimental uncertainty of 5 %. The
obtained crystallinity for PVIBE in the PVIBE/-PL blends was approximately a
half of that of the pure one (around 10%). However, the values in the PVIBE/
-PL/saponite nanocomposites were not affected and remained unchanged at ca.
20 %. This is also confirmed by comparing the expanded spectra of CH-O-CH2
region between Figures 1(b) and 1(c). The degree of crystallinity of -PL in the
blends remained unchanged at 54 5% except for PVIBE/-PL=10/1 blend where
the value was ca. 30% (6, 7). In the case of nanocomposites, the crystallinity of
-PL was changed to be around 45%, except for PVIBE/-PL/saponite=10/1/0.03
where the value was 36% (7).
Here, we discuss briefly the difference in sensitivity to both CR and NC phases
towards spectral separation, T1C curve, and XRD (3). Figure 2(B) shows the T1C
curves of the CH peak of pure -PL. The insert is the decay curve obtained from
the whole integral. The solid circles are obtained from the integrals of the CR phase
and the open circles from those of the NC phase. The data were obtained from the
integrals of the peaks depicted by the dotted line (CR) and by the broken line (NC),
respectively, in Figure 2(A). The sum of both decays is equal to the curve of the
insert. The T1C decay in the insert shows double exponentials and the fraction of
the long component was obtained to be 62 11% by the least square fit with a
conventional double-exponential function. The T1C decay curve of the CR phase
shows single exponential () and that of the NC phase non-single exponential ().
For the CR decay, we obtained the T1C value of 25 2 s by a least-square fit with
a conventional single-exponential curve. This value is comparable to the value of
the long T1C component obtained from the decay of the insert (22 3 s) but not
completely the same: the value obtained from the CR phase is longer than that
of the whole integral. This indicates that the molecular motion of the intrinsic
CR phase is much slower than that estimated from the long T1C component, and
the long T1C component includes the inter-phase, where the molecular motion is
comparable to that of the CR phase. Thus, the value obtained becomes slightly

90
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 shorter than that obtained from the CR phase. Actually, the T1C curve of the
NC phase shows a non-single exponential decay. The dashed and dotted lines
in Figure 2(B) represent the decays of short and long components in the insert,
respectively. The combination of both lines reproduces the T1C curve of the NC
phase () well, suggesting that the broad NC component includes both inter-phase
and amorphous regions. It is also important to discuss herewith the accuracy
of crystallinity determined from NMR. The XRD experiment can estimate the
crystallinity reflecting the chain order but not mobility. Figure 2(C) shows the
XRD patterns of whole -PL (regular solid line), mostly CR phase (bold solid
line), and almost all NC phase (dotted line). The XRD pattern of the mostly CR
phase was obtained from the linear combination of both XRD patterns of whole
-PL and almost all NC phase. The estimated value from XRD is ca. 58 % with an
uncertainly of 5-10% (3). This value is in excellent agreement with that estimated
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

from spectral separation, 54%, within experimental error. These observations

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

indicate that spectral separation spectral separation provides the same quality and
accuracy for the crystallinity as well as XRD observation. Of course, it is also true
that T1C provides accurate crystallinity if the appropriate temperature is chosen
because mobility is largely affected by temperature chosen. Furthermore, it has
a potential to provide morphological information ascribed to mobility difference.
Actually, the combination study of the spectral separation and T1C experiments
gives us information on the fraction of inter-phase in the material (3).

Figure 2. (A) Observed and expanded 13C CPMAS NMR spectra of CH carbon
of -PL: dotted line and broken lines represent the CR and the NC phases,
respectively. (B) The observed 13C spin-lattice relaxation curve obtained from
the CR phase () and the NC phase (). The insert is obtained from the entire
integral. (C) Observed XRD patterns of -PL (regular solid line),-PL mainly
consisted of NC phase (dotted line), and the CR phase (bold solid line).
91
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 3. DSC curves of PVIBE/-PL blends (A) and PVIBE/-PL/saponite-clay

nanocomposites (B).

A discussion is needed of the relationship between melting temperature

and the domain size of crystalline phase. For PVIBE/-PL blends and
PVIBE/-PL/saponite nanocomposites, we found an interesting DSC phenomenon
shown in Figure 3. Figure 3(A) shows DSC curves of pure PVIBE, PVIBE/-PL
blends, and pure -PL. Similarly, Figure 3(B) depicts DSC curves of the melting
temperature, Tm, region of -PL for the PVIBE/-PL/saponite nanocomposites.
From Figure 3(A), we recognize that the Tm transition of -PL shifts towards lower
temperatures with increasing PVIBE composition although Tm for PVIBE remains
unchanged with -PL addition. On the other hand, the Tm transition of -PL
in the PVIBE/-PL/saponite nanocomposites do not show such a composition
dependence as shown in Figure 3(B). For the PVIBE/-PL/saponite=10/3/0.09 to
10/5/0.15 regions, the Tm transition does not change with composition and shows
a constant value of 439 K. This value is comparable to that of PVIBE/-PL=10/3
blend. With increasing -PL composition, this value decreases to 433 K at
PVIBE/-PL/saponite=10/1/0.03. This value is somewhat higher than that of
PVIBE/-PL=10/1 blend. Here, if the diluent effect is predominant, the shift
tendency of Tm towards lower temperature with composition is reasonable for
the change of PVIBE/-PL/saponite-clay=10/2/0.06 to 10/1/0.03. However, the
constancy of Tm values for PVIBE/-PL/saponite-clay=10/3/0.09 to 10/5/0.15
cannot be interpreted as well. On the basis of the Gibbs-Thomson effect (1), the
same lamellar thickness exhibits the same Tm value. The DSC results shown in
Figure 3 imply that the change of Tm is attributable to the lamellar thickness in
the crystalline domain of -PL but not to the diluent effect by adding PVIBE.
To investigate the domain size of crystalline phase, we employed the 1H spin-
lattice relaxation decay analysis. The crystalline domain size can be inferred from
92
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the 1H spin-diffusion rate between PVIBE and -PL. Figure 4 shows the T1H curves
observed from PVIBE (CR: , NC: ) and -PL (CR: , NC: 4) in PVIBE/
-PL=10/1 blend. For pure PVIBE and pure -PL, the observed T1H curve of
the CR phase agrees with that of the NC phase due to the fast 1H spin diffusion
between the CR and NC domains; the dotted straight lines represent the respective
relaxation curves of pure PVIBE (0.8 s) and pure -PL (2.8 s), respectively. This
figure indicates that PVIBE and -PL are heterogeneous on a scale of 20-100 nm,
because both T1H curves are not consistent with each other. Furthermore, the T1H
curve of the CR () for -PL is different from that of the NC (4), while that
of CR for PVIBE () is the same as that of NC (). This observation indicates
that insufficient 1H spin diffusion occurs between PVIBE and -PL during the T1H
measuring period, and similarly even between the CR and NC phases of -PL for
the PVIBE/-PL=10/1 blend. This is because the molecular motion of NC chains
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

of -PL becomes much faster after blending with a large amount of mobile PVIBE
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

enough to hinder the 1H spin diffusion between NC and CR domains even in the
same -PL. This indicates that NC chains of PVIBE coexist with and penetrate
into the NC domain of -PL intimately much more than the other blends, and then
PVIBE chains are close proximity of the CR chains of -PL. Thus, the 1H spins
of PVIBE exchange smoothly with 1H spins in the CR domain of -PL as well
as in the NC domain of -PL. This results in the three-spin system. Actually,
the observed T1H curves show the typical curvature in the case of insufficient
1H spin-diffusion rate for a three-spin system (17): T1H curves are one concave

(PVIBE) and two convex (-PL). For the other compositions, the T1H curves show
the typical two-spin system decay: the CR curve is in good agreement with the
NC curve for each polymer (6, 7).
The proton magnetizations Mi(t) of i spin for two- and three-spin systems are
expressed by equations (1) and (2), respectively as follows:

where and . Symbols A and B denote 1H spins of

PVIBE and -PL, respectively. The sum of fA and fB equals to 1.

where , , and
. Parameters f, K, and k represent 1H
molar fraction,
intrinsic 1H relaxation rate, and 1H spin-diffusion rate, respectively. Numbers 1,
2, and 3 express 1H spins of PVIBE, NC of -PL, and CR of -PL, respectively.
The sum of f1, f2 and f3 equals to 1.
The simulated best-fit curves are depicted as solid, dashed, and dotted lines
in Figure 4. The simulated lines are in good agreement with the observed data
points. The obtained values are K1 = 1.12 s1, K2 = 0.39 s1, K3 = 0.25 s1, k12 = 0.57
93
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 s1, k23 = 2.0 s1, and k13 = 0.33 s1. The uncertainty is approximately 10 %. The
repeating unit length, L, can be estimated from the relation of L = /(fPVIBE
f-PL): D is the 1H spin-diffusion coefficient for PVIBE/-PL system (380 nm2s-1)
and obtained from the 1H spin-spin relaxation time (1820).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 4. Observed T1H relaxation curves of PVIBE (CR: , NC: ) and -PL
(CR:, NC: 4) for the PVIBE/-PL=10/1 blend. Each solid line represents the
calculated curve from equation (2). The T1H relaxation curves for pure PVIBE
and -PL are represented by the dotted lines.

The obtained k, the estimated L, the degree of crystallinity, xC, of -PL, and the
domain size which is related to the total lamellar thickness, LC, of the crystalline
phase of -PL are listed in Table 1. There exists a good relationship between the
decrease of LC values and the depression of the melting point with decreasing -
PL for blends. On the other hand, the LC values for PVIBE/-PL/saponite=10/
3/0.09 to 10/5/0.15 exhibit the same value of 30-33 nm within an error of 10%.
For further PVIBE-rich compositions, LC values decreases with PVIBE as well
as those of blends. The LC values for PVIBE/-PL/saponite =10/3/0.09 to 10/5/
0.15 are comparable to that for the PVIBE/-PL=10/3 blend (35 nm). It is also
interesting and surprising that the Tm values of the PVIBE/-PL/saponite=10/3/
0.09 to 10/5/0.15 nanocomposites show the same value of 439 K with each other.
Furthermore, this value is in excellent agreement with that of the PVIBE/-PL=10/
3 blend at 440K. For PVIBE/-PL=10/2 or 10/1 compositions, similar relationship
also remains; i.e., Tm value depends on the LC value. These results clearly prove
that the Tm value is governed by the lamellar thickness in crystalline domain and
not by the diluent effect in PVIBE/-PL.

94
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. 1H spin-diffusion rates, k, the estimated repeat lengths, L, of both
PVIBE and -PL domains, and the degree of crystallinity, xC, of -PL

composition k / s-1 L / nm L f-PL / nm xC / % LC / nm

10/1 0.33 569 41 30 12
10/2 0.37 309 42 58 24
10/3 0.20 320 61 57 35
10/4 0.12 350 83 55 46
10/5 0.10 345 97 59 57
10/1/0.03 0.25 654 47 36 17
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

10/2/0.06 0.38 305 41 46 19

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

10/3/0.09 0.15 369 70 47 33

10/4/0.12 0.18 286 68 44 30
10/5/0.15 0.21 238 67 49 33

L is estimated from k and the formula of L = /(fPVIBE f-PL). LC is caluculated

from L f-PL xC/100.

The crystallinity of -PL in the PVIBE/-PL/saponite=10/5/0.15-10/2/0.06

nanocomposites (around 45%) is lower than that of PVIBE/-PL blends (around
55 %). In order to clarify the effect of adding clay on crystallite growth, I
employed the Gibbs-Thomson relation between the lamellar thickness and Tm.
According to the Gibbs-Thomson relation (1), the depression of Tm is inversely
proportional to the CR thickness. The estimated total lamellar thickness, d, is not
suitable for applying the relation directly because we never know the intrinsic
and single lamellar thickness by this NMR simple estimation. However, if the
intrinsic lamellar thickness does not change but the amount of lamellar decreases
with decreasing of the domain size to keep the constancy of crystallinity from
10/5 to 10/2 compositions, the Tm values do not change. Thus, the change of Tm
should be explained by a diluent effect of adding PVIBE. However, the constancy
of Tm for PVIBE/-PL/saponite=10/5/0.15 to 10/3/0.09 cannot be interpreted by
the diluent effect. It is reasonable to consider, instead, that the thickness of the
single lamellar layer decreases with decreasing of the crystalline domain size
of -PL but the amount increases in the crystalline phase to keep the constant
crystallinity: namely, the estimated value d will correspond to an integral multiple
of the intrinsic and single lamellar thickness, x. Here, I adapt the value, d, to the
Gibbs-Thomson relation instead of the intrinsic and single lamellar thickness of x
to elucidate the effect of adding saponite clay on the crystalline growth of -PL.

95
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 5. Plot of Tm and the total lamellar thickness of -PL, d, for the
PVIBE/-PL blends. The solid curve is obtained from equation (3).
Figure 5 shows the plot of Tm against the total thickness of the CR phase
of -PL, d, in the PVIBE/-PL blends. The Gibbs-Thomson effect is given as
following equation.

Here, Tm0, , Hm0, and x are, respectively, the melting temperature of a

complete crystal of -PL, surface energy at the lamella interface, equilibrium
enthalpy per a unit cell, and the thickness of CR. The solid line is the least-square
fitted curve to the data points by substituting x to d. The fitted curve is in excellent
agreement with the observed data points. The estimated Tm0 and are 446.8 0.2
K and 220 10 Knm, respectively. By using these parameters, the thickness of
CR for pure -PL can be estimated to be ca. 120 nm and that of -PL/saponite to
be ca. 46 nm. This suggests that adding clay has a large ability of depression of
crystalline size for -PL. The degree of crystallinity of pure -PL is approximately
54 % but that of -PL/saponite is 42 %. The decrease ratio of crystalline size
(1.0 46/120 = 0.62) is much larger than the depression ratio of crystallinity (1.0
42/54 = 0.22). This indicates that the crystalline phase of the -PL/saponite
probably consists mainly of microcrystallites.

Morphology and Impact Resistence

The impact resistence of PK/PA blends becomes much better in the
wet condition, particularly for the range of PK/PA=10/90 to 70/30 (810).
Surprisingly, the impact energy under dry condition is around 25 kJ m-2, but at >
1.5% humidity the value exceeds 160 kJ m-2. This large increase of impact energy
occurs for the PK/PA blends with a PA content of over 30 weight % only, and it
is interesting that the wet pure PA sample does not show such an improvement,
though it is seen even for the PK10PA90 blend. The improvement of impact
96
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 resistence under the wet condition is limited to the compositions of 80/20 and
90/10. Furthermore, no improvement is seen for either pure PA or pure PK.
Therefore, the pronounced improvement in impact resistance should be related to
both morphology and humidity.
Figure 6 shows sliced sections of three-dimensional (3D) TEM images of
PK10PA90, PK60PA40, PK80PA20, and PK90PA10 blends. The white color
represents mainly the amorphous region and the black color the crystalline
region. It is clear in the image of PK60PA40 that there is a lamellar-like
structure shown in black ascribed to the crystalline region. Furthermore, there
are many co-continuous spherical amorphous regions. For both pure PK and
pure PA, the lamellar-like structure is not clearly observed in the TEM image.
However, the lamellar-like structure is detected even in the PK90PA10 blend
in the 3D-TEM image, albeit not clearly. It is confirmed from electron energy
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

loss spectroscopy (EELS) image (9, 10) that this region mainly comes from PK;
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

hence the lamellar-like structure is assigned to the crystalline lamellar of PK. For
PK80PA20 and PK90PA10 blends, both 3D-TEM images show a characteristic
spherical amorphous domain filled with lamellar structure. For PK10PA90
blend, the 3D-TEM image shows poor contrast but there is a lamellar network
in the deep white region (perhaps amorphous). Considering the PK loading, it
is inconceivable that the entire lamellar network is attributed to the PK chains.
From small angle X-ray scattering (SAXS) study (9, 10), the long period of
lamellae for PK-rich blends is estimated to be ca. 14 nm but that for PA-rich
blends decreases from 14 to 9 nm gradually (21). The long periods of both pure
PK and pure PA are ca. 14 nm and ca. 8 nm, respectively. Therefore, the lamellar
network for the PA-rich blends consists of a mixture network of both PK and PA.
Furthermore, the impact strength of both PK80PA20 and PK90PA10 does not
show the pronounced improvement under the wet condition. This suggests that
the existences of PK lamellae and the co-continuous spherical amorphous region
are strongly related to the improvement in impact resistance.

Figure 6. The 3D-TEM images in are views selected at arbitrary angles for
PK10PA90, PK60PA40, PK80PA20, and PK90PA10 blends.

In order to investigate the structure of the crystalline phase, domain size, and
the effect of humidity on polymer chain, 13C CPMAS NMR spectra were obtained
for both dry and wet conditions. Figure 7 shows the observed and enlarged solid-
state 13C CPMAS NMR spectra of the PK60PA40 blend in the aliphatic (a) and
carbonyl (b) carbon regions. The solid and broken lines represent the 13C CPMAS
97
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 NMR spectra for dry and wet conditions, respectively. By comparing both dry
and wet spectra, it is clear that the signal intensity under the wet condition is
lower than that under the dry condition. The relative intensity of the amorphous
peak of PA decreases more markedly in the presence of moisture, while those of
PK remains unchanged. This indicates that the CP efficiency of the amorphous
phase of PA declines selectively. When water absorption occurs predominantly
in the amorphous phase, the polymer chains display higher mobility than in the
crystalline phase. Since CP works effectively between the 13C and 1H spin pairs
at a lower level of mobility, the CP efficiency in the mobile spins becomes worse.
Similarly, the carbonyl (C=O) peak of PK does not change after water absorption,
while the C=O signal of PA decreases. This observation suggests that water is
absorbed predominantly in the amorphous phase of PA but not in the crystalline
phase of PA and PK and the amorphous phase of PK.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 7. Observed and enlarged 13C CPMAS SSNMR spectra for (a) the
aliphatic and (b) the carbonyl carbon regions of PK60PA40 blend under dry and
wet (broken line) conditions.

98
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 PA is known to have the two crystalline phases, and crystallites (2226).
In order to clarify the change in the crystalline phase of PA after blending with
PK, we examined the 15N CPMAS NMR spectra of PK60PA40 blend and pure
PA. Figure 8 shows the 15N CPMAS NMR spectra of (a) the PK60PA40 blend and
(b) pure PA under the wet condition.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 8. Observed 15N CPMAS NMR spectra of (a) PK60PA40 blend and (b)
pure PA.

For the wet condition, the amorphous phase of PA is water-absorbed, so that

the CP efficiency of the amorphous phase is depressed. Thus the amorphous peak
is obscured in Figure 8. It is well known that the 15N NMR signal is sensitive
to the difference in the polymer chain direction between the and crystalline
phases of PA and shows a different chemical shift, which is due to the difference
in the circumstances of hydrogen bonding (2426). The signal ascribed to the
crystalline phase of PA resonates at 84 ppm and that to the crystalline phase at
89 ppm (25). Obviously, the crystalline phase of PA in the PK60PA40 blend
is dominant compared with that of pure PA: rapid cooling of PA induced the
growth of the crystalline phase in pure PA, although the crystalline phase
is thermally stable usually. This indicates that PK induces the growth of the
crystalline phase of PA in the blend rather than crystallites. The relative fraction
of the crystalline phase of PA changes from 40% to 70% after blending with

99
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 PK. This observation suggests that PK induces the thermally stable crystalline
phase of PA in the blend. Presumably, the lamellar-like structure of PK will
be stabilized by the alternation of the crystalline phase of PA between parallel
() and anti-parallel () chain directions. It is also revealed that PK causes the
crystallization temperature of PA to decrease; a higher crystallization temperature
induces crystallites (9). The change in the crystalline form of PA is related to
the complicated morphology detected in the TEM image and to the high impact
properties under the wet condition.
Figure 9 shows the observed T1H curves of both PK and PA in the PK60PA40
blend for both dry and wet conditions. These T1H curves were obtained from the
C=O carbon peaks of PK and PA, respectively. Because both C=O carbons are
close to the protons of the main chain aliphatic carbons or NH, 1H spin diffusion
occurs efficiently even though protons are not bonded directly to carbon within the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

CP contact time of 2 ms. The T1H curves of both pure PK and pure PA (dashed
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

lines) show single exponential decay, resulting in semi-log plots in a straight line.
The T1H values for pure PK and pure PA under the dry condition are 2.06 0.03
s (1/T1H = K = 0.485 s1) and 1.06 0.01 s (K = 0.943 s1), respectively. The
values under the wet condition are 1.50 0.01 s (K = 0.667 s1) for PK and 0.86
0.01 s (K = 1.163 s1) for PA. The T1H rates, K, under the wet condition are
faster than those under the dry condition. This indicates that the molecular motion
of both polymers is affected and is increased by water absorption. For PA, water
is absorbed predominantly in the amorphous phase and the humidity is 1.9%; for
the PA-rich alloys, the humidity is 1.7%. On the other hand, PK absorbs water
without selectivity, and the amount of absorbed water is only 0.6%. This difference
in water absorption shows that the 13C CPMAS NMR spectrum of PK remains
unchanged under the wet condition, but that there is sufficient water absorption to
change the molecular motion and to affect the T1H of PK.
To obtain the information on domain size, the observed data points are fitted
by equation (1) with fPA of 0.548. The solid lines, which are obtained by the least-
square fit to the data, are in excellent agreement with the observed T1H decay curves
( and ) in the PK60PA40 blend. The K values obtained for the dry condition
are 0.53 0.02 s-1 (1.90 s) for PK and 0.81 0.02 s-1 (1.21 s) for PA, respectively:
the values in parenthesis are 1/K. The 1H spin-diffusion rate, k, was 0.8 0.3 s-1.
The values obtained for the wet condition are KPK = 0.66 0.04 s-1 (1.51 s), KPA
= 0.95 0.04 s-1 (1.05 s), and k = 0.7 0.5 s-1, respectively. The initial relaxation
rate of PA, KPA, in the blend is more affected by blending with PK than that of PK.
The KPA value is slower than that of pure PA by ca. 0.2 s1, whereas that of PK in
the blend is comparable to that of pure PK. This observation suggests that PA is
dispersed and located in the PK matrix. It also indicates that the molecular motion
of PA is greatly affected after blending with PK. Since the molecular motion of
the rigid PK lamellar structure is not influenced by PA, the initial relaxation rate
does not change after blending, whereas that of PA is altered. On the other hand,
the 1H spin-diffusion rate, k, is comparable to the initial relaxation rate of both
polymers and intermediate between KPK and KPA. This value indicates that 1H spin
diffusion is slow, but that the two polymers are not completely phase-separated on
a scale of several tens to several hundreds of nm. The maximum diffusive path

100
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 length, (i.e., average value of r), for the three-dimensional diffusion model
is estimated by ; D is the diffusion coefficient for proton spin energy. From
the TEM images that show that the direction of the lamellae is not uniform, and we
employed the to estimate the domain size. The effective D value (27, 28)
has been estimated to be ca. 700 nm2s1 from both diffusion coefficients DPK = 840
nm2s1 and DPA = 600 nm2s1, which are obtained from the 1H spin-spin relaxation
rate (20). The estimated path length, r, is about 77 nm. Thus, the average domain
size is approximately 150 nm which is twice as large as r. This value is consistent
with the averaged overall size of the elliptical (or spherical or cylindrical) domains
measured from the 3D TEM.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 9. Semi-logarithmic plots of normalized T1H decay curves for PK ()

and PA () in the PK60PA40 blend under the dry and wet conditions. Dashed
lines represent T1H decay curves for pure PA and pure PK. The solid lines were
obtained with equation 1.
Thus, the domain size under the wet condition did not change as compared
to that under the dry condition. From 3D TEM image, however, it was suggested
that the creation of the co-continuous spherical amorphous region contributes the
high impact resistance, and PK crystalline lamellae are evidently formed only by
blending with PA. From the NMR study, it was futher revealed that the crystalline
phase is converted to the crystalline phase for PA by blending with PK, and water
absorption occurs selectively in the amorphous region of PA. These observations
indicate that the water-absorbing amorphous phase of PA plays an important role in
enabling PK/PA blends to exhibit high impact resistance and toughness. Actually,
from SAXS the long period of PK lamellae becomes wider by approximately 1
nm under the wet condition through the swelling of PA amorphous phase (9, 21).
Hence, the combination of both rigidity of PK lamellae and elasticity of water-
101
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 absorbing amorphous phase of PA is needed to obtain high impact properties,
modulus, and toughness.

Conclusion
In this article, I first showed the relationship between the domain
size of crystalline phase and Tm of -PL for both PVIBE/-PL blends and
PVIBE/-PL/saponite nanocomposites. The Tm values are associated with the
total lamellar thickness estimated from an analysis of the effect of 1H spin
diffusion on T1H. Even so, it is disputable whether the calculated value is really
related to the single lamellar thickness. As I mentioned above, there are two
possibilities for the lamellar thickness in the case of similar crystallinity. One is
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

that the lamellar thickness does not change but the amount of lamellar decreases
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

with decreasing domain size. Another one is that the thickness of the single
lamellar layer decreases with decreasing crystalline domain size of -PL but the
amount increases in the crystalline phase to keep the crystallinity constant. The
mostly constant crystallinity has been estimated, as listed in Table 1 for 10/2 to
10/5 compositions. In the latter case, Tm changes as expected with the estimated
LC value and obeys the Gibbs-Thomson relation. If the actual morphology is
similar to the former condition, the decreasing Tm can be explained by the diluent
effect of adding PVIBE. In actual fact, we cannot distinguish between the diluent
effect and the Gibbs-Thomson effect from the LC estimation for the blends only.
However, it is also true that the diluent effect cannot explain the constancy of Tm
observed for the nanocomposites. The decrease of the lamellar thickness gives
unambiguous solution of the shift and the constancy of Tm values observed for
both blends and nanocomposites, if the estimated LC value is comparable to the
summation of a single lamellar layer thickness.
Furthermore, the relationship between the crystalline structure and impact
property for the PK/PA blends is elucidated. The high impact resistance
and toughness of PK/PA blends are the result of the interactions among the
water-absorbing amorphous phase of PA, the existence of lamellar-like structure
of PK, and the conversion of PA crystallite types.

Acknowledgments
I am very grateful to collaborators at National Defense Academy, NISSAN
ARC LTD., and Industrial Technology Center of Okayama Prefecture for their
helpful and valuable cooperation with this study.

References
1. Strobl, G. R. The physics of polymers; Springer-Verlag: Berlin, 1996;
Chapter 4.
2. Kitamaru, R.; Kaji, K. Kobunshi Jikken-gaku. In Kobunshi no Kotai Kouzou
II (Solid structure of polymer II); Kyoritsu Shuppan Co., Ltd.: Tokyo, 1984;
Vol. 17, Chapter 6 (in Japanese).
102
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 3. Asano, A.; Tanaka, C.; Murata, Y. Polymer 2007, 48, 38093816.
4. VanderHart, D. L.; Asano, A.; Gilman, J. W. Chem. Mater. 2001, 13,
37813795.
5. VanderHart, D. L.; Asano, A.; Gilman, J. W. Chem. Mater. 2001, 13,
37963809.
6. Asano, A.; Murata, Y.; Kurotsu, T. e-J. Soft Mater. 2007, 3, 18.
7. Asano, A.; Tanak, C.; Kurotsu, T. Macromolecules 2008, 41, 94699473.
8. Asano, A.; Nishioka, M.; Kato, A.; Takahashi, Y.; Sawabe, H.; Arao, M.;
Sato, S.; Sato, H.; Izumi, T.; Drozdova, O.; Ishikawa, D.; Hasegawa, T.;
Okamura, T.; Nagata, K.; Hikasa, S.; Iwabuki, H. Polym. Prep. 2008, 49,
682683.
9. Asano, A.; Nishioka, M.; Takahashi, Y.; Kato, A.; Hikasa, S.; Iwabuki, H.;
Nagata, K.; Sato, H.; Hasegawa, T.; Sawabe, H.; Arao, M.; Suda, T.;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch005

Isoda, A.; Mukai, M.; Ishikawa, D.; Izumi, T. Macromolecules 2009, 42,
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

95069514.
10. Kato, A.; Nishioka, M.; Tkahashi, Y.; Suda, T.; Sawabe, H.; Isoda, A.;
Drozdova, O.; Hasegawa, T.; Izumi, T.; Nagata, K.; Hikasa, S.; Iwabuki, H.;
Asano, A. J. Appl. Polym. Sci. 2010, 116, 30563069.
11. Shih, I. L.; Shen, M. H.; Van, Y. T. Bioresource Tech. 2006, 97, 11481159.
12. Bennet, A. E.; Rienstra, C. M.; Auger, M.; Lakshmi, K. V.; Griffin, R. G. J.
Chem. Phys. 1996, 103, 69516958.
13. Peersen, O. B.; Wu, X.; Kustanovich, I.; Smith, S. O. J. Magn. Reson. 1993,
A104, 334339.
14. Metz, G.; Wu, X.; Smith, S. O. J. Magn. Reson. 1994, A110, 219227.
15. Torchia, D. A. J. Magn. Reson. 1978, 30, 613616.
16. VanderHart, D. L.; Prez, E. Macromolecules 1986, 19, 19021909.
17. Asano, A.; Takegoshi, K.; Hikichi, K. Polymer 1994, 35, 56305636.
18. VanderHart, D. L.; McFadden , G. B. Solid State Nucl. Magn. Reson. 1996,
7, 4566.
19. Demco, D. E.; Johansson, A.; Tegenfeldt, J. Solid State Nucl. Magn. Reson.
1995, 4, 1338.
20. Assink, R. A. Macromolecules 1978, 11, 12331237.
21. Takahasi, Y.; Nishioka, M.; Kato, A.; Hikasa, S.; Iwabuki, H.; Nagata, K.;
Asano, A. Kobunshi Ronbunshu 2009, 66, 577584.
22. Weeding, T. L.; Veeman, W. S.; Gaur, H. A.; Huysmans, G. B.
Macromolecules 1988, 21, 20282032.
23. Okada, A.; Kawasumi, M.; Tajima, I.; Kurauchi, T.; Kamigaito, O. J. Appl.
Polym. Sci. 1989, 37, 13631371.
24. Hatfield, G. R.; Glans, J. H.; Hammond, W. B. Macromolecules 1990, 23,
16541658.
25. Mathias, L. J.; Davis, R. D.; Jarrett, W. L. Macromolecules 1999, 32,
79587960.
26. Andreis, M.; Koenig, J. L. Adv. Polym. Sci. 1995, 124, 191237.
27. Clauss, J.; Schmidt-Rohr, K.; Spiess, H. W. Acta Polym. 1993, 44, 117.
28. Mellinger, F.; Wilhelm, M.; Spiess, H. W. Macromolecules 1999, 32,
46864691.

103
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 6

Investigation of Filler-Matrix Interactions and

Confinement Effect in Polymer Nanocomposites
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

K. P. Pramoda,*,1 K. Y. Mya,1 T. T. Lin,1 J. Wang,1 and C. B. He1,2

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

1Institute
of Materials Research and Engineering (IMRE),
(A*STAR) Agency for Science, Technology and Research,
3 Research Link, Singapore 117602
2Department of Materials Science and Engineering, National University of

Singapore, 7 Engineering Drive 1, Singapore 117574

*E-mail: [email protected]

Although nanocomposites are known for decades, the

molecular interactions under the confined environments of such
composites are not well understood. In this article we report
the influence of two types of fillers, viz., graphite (G) and
graphene oxide (GO) nanosheets, on polyimide (PI) matrix with
varying filler content. Solid state NMR and thermo-mechanical
analysis were deployed to characterize the morphology of
hybrid materials and the interactions between the polymer
and the filler. CP-MAS and relaxation experiments were
performed to probe the interaction between inorganic and
organic phases on a molecular level and to elucidate polymer
dynamics. NMR relaxation measurements further helped in
elucidating the polymer dynamics and were supplemented by
thermo-mechanical relaxation. From the relaxation data it is
clear that GO interacts better than G with the polyimide matrix.
This is further confirmed by the higher modulus observed for
PI-GO nanocomposites than for PI-G.

Introduction
Nanocomposites are a newly emerged class of composite materials that
incorporate relatively small percentages of nanometer-sized filler particles.
Nanocomposites with carbon as filler materials exhibit improved electrical,

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 thermal and mechanical properties (1, 2). Among the various other carbon
additives, graphite is much more attractive to researchers as it is comparable
to carbon nanotube (CNT) in properties. Moreover, it has a layered structure
(similar to clay) constituted by a large number of graphene sheets held by van
der Waals forces (3, 4). Therefore, pretreatment and modification of graphite is
necessary to achieve the nanosheet distribution of graphene. Surface treatment
of graphene makes it more compatible to polymer matrix which helps good
dispersion in a polymer matrix, resulting in nanocomposites with excellent
mechanical and electrical properties. Polymer nanocomposites find applications
in automobile, food packaging and biomedical industries due to improvements
in thermal, physical and gas permeability properties of the polymers attributed
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

to the nanoscale filler materials. For silicates, these enhancements are achieved
with less than 10% wt addition of the exfoliated nanoscale dispersion of 1 nm
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

thick layers, with diameters between 20 and 500 nm. This is in stark contrast to
conventional polymer fillers, such as talc, mica, silica, and carbon black, which
require high concentrations (>30 wt %) and may cause deterioration of fracture
toughness and in processibility.
Daniel et al (5) reported a 15% improvement in elastic modulus of 1 wt %
expanded graphite (EG)-epoxy nanocomposite over pure epoxy and attributed it to
the in-situ formation of graphite nanosheets as well as to their uniform dispersion
and exfoliation in the epoxy matrix. High density polyethylene reinforced with EG
and untreated graphite by a melt compounding process also shows an improvement
in the electrical and mechanical properties of nanocomposite (6). Poly(methyl
methacrylate)-EG composites prepared by in-situ polymerization exhibits better
mechanical and electrical properties than those prepared using solution blending
methods (7). Although epoxy and thermoplastic nanocomposites with graphite/
graphene oxide nanosheet fillers have been studied, to our knowledge, there have
been no reports available on polyimide (PI)-based graphite nanocomposites.
Polyimides (PIs) are widely used in defense and aerospace applications as
well as in the electronics industry for a variety of applications because of their high
temperature resistance, low dielectric constant, inertness to solvent, and long-term
stability (8). Polyimide nanocomposites with various fillers such as CNT, clay, and
POSS (Poly Oligomer SilSesquioxane) have been studied (912). In this article we
report on the influence of two types of fillers, namely, graphite (G) and graphene
oxide (GO) nanosheets, on polyimide matrix with varying filler content. Solid
state nuclear magnetic resonance (SSNMR) spectroscopy and thermo-mechanical
analysis were deployed to characterize the morphology of hybrid materials and
the interaction between the polymer and the filler. The chemical oxidations of
graphite to graphite oxide are confirmed by XRD. The interaction between filler
and the polymer matrix is further studied by XPS analysis. In addition, dynamic
mechanical thermal analysis (DMTA), and thermomechanical analysis (TMA) are
applied to characterize the glass transition temperature (Tg), storage modulus (E),
and coefficient of thermal expansion (CTE).

106
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Experimental Section
Materials and Synthesis of Polyimide Composites

Biphenyltetracarboxylic dianhydride, phenylene diamine, and N-methyl-2-

pyrrolidinone (NMP) were purchased from Sigma-Aldrich. All the monomers
were purified before use. Graphite, donated generously by Asbury Carbons
(Asbury, NJ, 3775 grade) was used as received and also oxidized by a modified
Hummers method (13). The poly(amic acid) (PAA) precursor solution was
prepared through polycondensation reaction between dianhydride and diamine
in NMP. The resulting PAA solutions were controlled to have solid contents of
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

10-15wt%. In a separate flask, the filler (G/GO) was first uniformly dispersed in
NMP using an ultrasonic bath to obtain their respective nanosheet. After that, the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

required amount of G-NMP/GO-NMP was added to the PAA solution and stirred
continuously. The solution was then cast on a glass plate using automatic film
applicator and thermally treated at 100C for 1 h, 200C for 1 h, and 300C for 4
h, respectively. The films were then removed from the glass substrate with the
aid of warm water and dried at 50C in a vacuum oven for 2 days. The thickness
of the films was ~100 m. Hereafter, the PI composite film with G filler was
denoted as PI-G and the film with GO filler PI-GO.

Characterization

X-ray diffraction (XRD) patterns of the samples were recorded by using a

Bruker GADDS diffractometer, with an area detector operating under 40 kV and
40mA, using CuK radiation ( = 0.154 18 nm). X-ray photoelectron spectroscopy
(XPS) data were recorded using a VG Scientific EscaLab 220 IXL spectrometer
with an Al K X-ray source (h = 1 486.6 eV). The spectra were corrected for
carbon shift binding energy, Cs = 284.5 eV in order to determine the accurate
binding energies. SSNMR measurements were carried out with Varian 400 MHz
(9.4 T) wide bore spectrometer and the samples were spun at the magic angle
at the rate of 15 kHz using a 4 mm double resonance probe. The spectrometer
was operated at resonance frequencies of 100 and 400 MHz to obtain 13C and 1H
spectra, respectively. The spectra were referenced using tetramethylsilane (TMS).
Thermomechanical behavior of the composites was analyzed with the help of
dynamic mechanical thermal analyzer (DMTA) and thermo mechanical analyzer
(TMA) from TA Instruments. The modeling and simulation on these complexes
were done using Materials Studio software and DMol3 (14). The binding energy
was calculated using the density functional theory (DFT) calculation with the
following equation:

107
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Results and Discussion
Chemically oxidized graphite (G) is known as graphene oxide (GO) which
forms a stable dispersion in water or other organic solvents leading to GO
nanosheets. The extent of oxidation and the surface structure of GO are the key
parameters to improve their compatibility with the polymer. The oxidation of
graphite may affect its chemical structure and was confirmed by XRD and FTIR
(15). Figure 1 represents the XRD pattern of G and GO together with its polymer
composites. A strong peak is observed at 2 = 26.6 for G and a scattering signal
at 2 = 9 for GO. The peak at 26.6 represents the intact crystal lattice in G and
the peak at 9 for GO corresponds to a layer spacing of 10 (16). The absence
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

of the peak at 26.6 (2) in GO implies that the crystal lattice structure of G has
been disrupted. On the other hand, a small peak that appears at 26.6 (2) in the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

PI-G composites shows the intercalation of graphite in the PI matrix. However,

the absence of the peak at 9 (2) that is the characteristic of GO in the PI-GO
composites indicates that the packing of the GO has been entirely disrupted by
PI-chains.

Figure 1. XRD spectra for G and GO along with the neat PI, PI-G and PI-GO
systems, respectively.

The interaction between filler and the PI matrix plays a critical role in the
property enhancement in the nanocomposites and therefore it is studied by XPS
and SSNMR analyses. XPS provides information on the specific interaction in
polymer nanocomposites as the binding energy (BE) of a core-level electron
depends on its chemical environment (17). SSNMR reveals the structure,
conformation and dynamics of polymer chains in the composites, and thereby
the interactions between the polymer and the filler. However, there have been
only a few papers on the application of SSNMR to the structure and dynamics
of organic-inorganic hybrid nanomaterials. VanderHarts research group at
108
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 NIST has studied Nylon-6/clay nanocomposites, including clay dispersion,
crystal stratification, and stability of organic modifiers, using 1H T1 relaxation
times. Nanodispersion in polystyrene-montmorillonite nanocomposites was also
investigated by Gilman and coworkers (1820).
Figure 2 shows the 1H NMR spectra of the PI nanocomposites. It is observed
that the peak shifts upfield from 7.73 ppm in neat PI to 7.16 ppm along with the
presence of shoulder peaks (3.01 and 0.84 ppm) in PI-GO system. It is clearly
indicated that the interaction between filler and the matrix depends on the structure
and relaxation of polymer chains. In addition, the incorporation of GO shows a
better intensified peak compared to G, indicating better compatibility between the
filler and the matrix in PI-GO system.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

Figure 2. 1H SS NMR spectra for neat PI, and 1 wt% of PI-G, PI-GO.

Relaxation measurements on these samples can be informative as well. The

T1 relaxation times for 1H and 13C are tabulated in Table 1. The T1 values for the
composites are reduced compared to neat PI. The T1 values for both 1H and 13C
are reduced in the following order PI > PI-G > PI-GO. This trend suggests strong
filler-matrix interaction between PI chains and GO nanosheets compared to PI-G.
This was further verified by thermomechanical properties such as modulus and
CTE enhancement (discussed later) in PI-GO composites as compared to PI-G.
The N 1s binding energies (BEs) for PI composites are displayed in the Table
1. As seen in the Table, the peak at 400.8 eV in neat PI is shifted by 0.8 eV for 1
wt% GO (400 eV) and the shift is only 0.4 eV in the case of 1 wt% G (400.4 eV).
The magnitude of the changes in BEs of N 1s correlates well to the strength of
interaction, indicating that the interaction between PI and GO is stronger than that
in PI-G composite. The observed strong interaction in PI-GO system is possibly
due to the presence of functional groups such as OH, COOH, and epoxy, which
becomes the preferable site for the interaction with the linear chain packing of PIs.
109
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 However, the observed decrease in the BE values are not prominent, mainly due
to the lower sensitivity of N1s in the XPS measurement.

Table 1. N1s XPS and 1H SS NMR fit data along with their relaxation time
BE N1s XPS (eV) T1 (s)
13C 1H

PI 400.8 38 1.5
PI-G-1% 400.4 32 1.0
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

PI-GO-1% 400.0 22 0.3

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

Thermomechanical properties of PI composites were determined by DMTA

and TMA analyses. Figure 3 shows the DMTA curves for PI-G and PI-GO
composites with 1 wt% of G and GO, respectively. The storage modulus (E)
increased with increasing filler (G/GO) content as seen in Figure 3. Table 2
summarizes the storage modulus (E) at 100C along with and transition
temperature (T and T). T or Tg ~ 340C; the presence of filler did not affect
much the Tg of the PI matrix. The observed modulus enhancement was greater
with GO addition as compared to PI-G system. A ca. 38% improvement in E
for 1 wt% of G loading was observed versus an increase of ca. 106% with GO
loading for PI-GO composites. The extent of increase in E was greater in the
PI-GO system, which is possibly due to the better interaction between the GO
and the PI, leading to both physical and chemical cross-linking between filler
and the PI matrix, as evidenced by the 1H-NMR spectra. In comparison to our
system, PI-clay composites (10) showed a modulus increase of ~42% with 2 wt%
clay loading relative to that of neat PI. Similarly, Zhu et al (11) also reported
that the modulus of the PI-multi-walled carbon nanotube (MWCNT) composites
exhibited 40% increase with 5 wt% of MWCNTs content.

Table 2. Storage modulus, (E@100C), T, T(Tg), and CTE values

T (C) T(Tg) E (GPa) CTE (100-250C)
(C) @ 100C ppm/C
PI 191 341 8.41 15.70
PI-G-1% 187 340 11.6 3.98
PI-GO-1% 186 344 17.3 1.67

110
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

Figure 3. Temperature dependant storage moduli and tangent delta for the neat
PI and its composites.

The influences of G/GO on the CTE of the PI composites were also

investigated, and the results are shown in Table 2. It was found that the CTE
decreased when G and GO were incorporated into the PI matrix. The CTE of
the PI composite films is significantly reduced by about 75% and 90%, for 1%
loading of G and GO, respectively. Since CTE depends on the rigidity of PI-chain
and the interaction between PI and the fillers, the observed decreases in the CTE
for PI-G and PI-GO composite systems suggested that the interaction between
PI-chain and the G/GO nanosheet were strong which hindered the motion of
the PI-chain. On the other hand, the relatively lower CTE observed in PI-GO
when compared with PI-G could be attributed to the strong interaction between
the GO with PI-chain, which has been confirmed by XPS and NMR studies.
Furthermore, our DFT calculation also suggests that the binding energy for the
composite PI-GO was found to be 10.7 kcal/mol with an orbital cutoff 3.7
while it was only 3.7 kcal/mol for PI-G with same value of cutoff (i.e. 3.7 ).
The theoretical calculations also supported our experimental findings; i.e., the
interactions between the PI-chains and the GO nanosheets are much stronger than
that in the PI-G.

111
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Conclusions
Spectroscopic investigation was carried out to determine the relaxation
behavior of PI composites. The relaxation times obtained from NMR analysis
indicate that the interaction between PI chain and GO nanosheet is stronger than
that in PI-G composites. In agreement, the DFT calculations revealed that the
chemical compatibility achieved between PI and filler were mainly with GO
rather than G. Dynamic mechanical thermal analysis shows that the incorporation
of GO nanosheets significantly enhance the storage moduli of PI nanocomposites
compared to that of G nanosheets. The moduli improved ~106% with only 1 wt%
GO nanosheets while it is only ~38% with G nanosheets. The extent of thermal
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

expansion coefficient drop in the case of GO incorporation is greater than that of

G.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

Acknowledgments
The authors thank IMRE and Agency for Science, Technology, and Research
(A*STAR), Singapore for financial support.

References
1. Polymer Nanocomposites: Synthesis, Characterization and Modeling;
Krishnamoorti, R., Vaia, R. A., Eds.; ACS Symposium Series 804; American
Chemical Society: Washington, DC, 2002.
2. Pinnavaia, T. J.; Beall, B. W. Polymer-Clay Nanocomposites; Wiley: New
York, 2001.
3. Chung, D. D. L. J. Mater. Sci. 2004, 39, 26452661.
4. Jang, B. Z.; Zhamu, A. J. Mater. Sci. 2008, 43, 50925101.
5. Yashmin, A.; Luo, J. J.; Daniel, I. M. Compos. Sci. Technol. 2006, 66,
11821189.
6. Zheng, W.; Lu, X.; Wong, S. C. J. Appl. Polym. Sci. 2004, 91, 27812788.
7. Ramanathan, T.; Abdala, A. A.; Stankovich, S.; Dikin, D. A.; Herrera-
Alonso, M.; Piner, R. D.; Adamson, D. H.; Schniepp, H. C.; Chen, X.;
Ruoff, R. S.; Nguyen, S. T.; Aksay, I. A.; Prudhomme, R. K.; Brinson, L.
C. Nat. Nanotechnol. 2008, 3, 327331.
8. Ghosh, K. L.; Mittal, K. L. Polyimides: Fundamentals and Applications;
Marcel Dekker: New York, 1996.
9. Wahab, M. A.; Khine, M. Y.; He, C. B. J. Polym. Sci., Part A: Polym. Chem.
2008, 46, 58875896.
10. Agag, T.; Koga, T.; Takeichi, T. Polymer 2001, 42, 33993408.
11. Zhu, B.-K.; Xie, S.-H.; Xu, Z.-K.; Xu, Y.-Y. Compos. Sci. Technol. 2006,
66, 548554.
12. Mya, K. Y.; Wang, K.; Chen, L.; Lin, T. T.; Pallathadka, P. K.; Pan, J.; He, C.
B. Macromol. Chem. Phys. 2008, 209, 643650.
13. Hummers, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80, 1339.
14. Delley, B. J. Chem. Phys. 1990, 92, 508517.
112
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 15. Pramoda, K. P.; Hussain, H.; Koh, H. M.; Tan, H. R.; He, C. B. J. Polym.
Sci., Part A: Polym. Chem. 2010, 48, 42624267.
16. Jeong, H. K.; Lee, Y. P.; Lahaye, R. J. W. E.; Park, M. H.; An, K. H.; Kim, I.
J.; Yang, C. W.; Park, C. Y.; Ruoff, R. S.; Lee, Y. H. J. Am. Chem. Soc. 2008,
130, 13621366.
17. Beamson ,G.; Briggs, D. High Resolution XPS of Organic Polymers; John
Wiley and Sons: Chichester 1992.
18. Bourbigot, S.; VanderHart, D. L.; Gilman, J. W.; Awad, W. H.; et al. J. Polym.
Sci., Part B: Polym. Phys. 2003, 41, 31883213.
19. VanderHart, D. L.; Asano, A.; Gilman, J. W. Chem. Mater. 2001, 13,
37813795.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 19, 2012 | http://pubs.acs.org

20. VanderHart, D. L.; Asano, A.; Gilman, J. W. Macromolecules 2001, 34,

38193822.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch006

113
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 7

1H MAS NMR Spectroscopy of Polyethylene

Acrylic Acid Copolymers and Ionomers
Todd M. Alam,*,1 Janelle E. Jenkins,1 Michelle E. Seitz,2
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

C. Francisco Buitrago,3 Karen I. Winey,2 Kathleen L. Opper,4 Travis

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

W. Baughman,4 and Kenneth B. Wagener4

1Electronicand Nanostructured Materials, Sandia National Laboratories,
Albuquerque, NM 87185
2Department of Materials Science and Department of

Chemical and Biomolecular Engineering, University of

Pennsylvania, Philadelphia, PA 19104
3Department of Chemical and Biomolecular Engineering, University of

Pennsylvania, Philadelphia, PA 19104

4Department of Chemistry, University of Florida, Gainesville, FL 32611
*E-mail: [email protected]

High speed solid-state 1H MAS NMR spectroscopy has been

used to characterize a series of poly(ethylene-co-acrylic acid)
P(E-AA) materials where the distributions of the pendant
carboxylic acid group along the polymer backbone are precisely
controlled. Ionomers obtained from partial neutralization of
the carboxylic acid within these P(E-AA) materials with Zn2+
or Li+ were also investigated. Using a combination of 1D 1H
MAS NMR and 2D 1H MAS NMR double quantum (DQ)
and NOESY correlation experiments, details about the local
P(E-AA) structure were obtained. The influence of precise
versus random spacing of the carboxylic acid and the impact of
Li- and Zn-neutralization on the polymer structure is discussed.

Introduction
The control and tailoring of polymer microstructure to suit the final application
of the material remains an important objective. In addition, the ability to precisely
locate ions within the polymer structure and morphology is predicted to provide

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 control over the performance of ionomer membranes. Advances in the acyclic
diene metathesis (ADMET) and ring-opening metathesis polymerization (ROMP)
chemistry have now allowed production of polymers where the functional groups
are precisely and pseudo-randomly placed along the polyethylene chain. Recent
examples include the introduction of alkyl chains, halogens, phosphonic acids, and
carboxylic acid groups with very controlled chain spacing (14). Understanding
the local structure and morphology variations that occur with the inclusion of these
functional groups is an area of active research. Characterization of these precise
polymer materials has included the use of solid-state NMR spectroscopy; with
13C, 19F and 31P magic angle spinning (MAS) NMR (2, 3, 5), and static 2H NMR

(4) investigations exploring the impact of random versus precise functional group
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

incorporation on polymer crystallization.

In this chapter, we present then high speed 1H MAS NMR characterization
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

of a series of poly(ethylene-co-acrylic acid) copolymers, P(E-AA), and both

Zn- and Li-neutralized P(E-AA) ionomers. The term ionomer has found use in
literature to describe polymers that have ionic functional groups in humidified
environments. For example, Nafion in considered an ionomer in both the acidic
and neutralized salt form, because the sulfonic acid readily ionizes with water. In
the present discussion we will refer to the non-neutralized P(E-AA) materials as
acid copolymers, and the Li- or Zn-neutralized P(E-AA) systems as ionomers.
This distinction arises because the P(E-AA) copolymer acts as a weak acid, with
the carboxylic acids not being extensively ionized in a humid environment.
The 1H MAS NMR spectra of polymers below Tg are commonly broad
unresolved lines due to the presence of strong 1H-1H dipolar coupling (6,
7). Utilizing high speed (> 30 kHz) MAS spinning, this dipolar coupling is
significantly reduced producing spectra with increased spectral resolution (8).
Even though the 1H MAS NMR spectra for the P(E-AA) materials detailed here
are relatively simple, details about the local hydrogen bonding environment and
spatial connectivity between different functional groups can readily be obtained
using variable temperature (VT) and multi-dimensional NMR spectroscopy
correlation experiments as described below.

Experimental Section
Material Preparation

The synthesis and characterization of the linear poly(ethylene-co-acrylic

acid), P(E-AA), materials have been previously described (1). Polymers with
precisely spaced carboxylic acid groups were prepared using the ADMET
chemistry. These polymer samples are designated as p9AA, p15AA and p21AA,
and reflect samples where the carboxylic acid groups are precisely (p) located
every 9th, 15th and 21st carbon along the backbone, respectively. The relative
acid concentrations are 22, 13.3, and 9.5 mol%, respectively. A ROMP synthesis
method was used to produce materials of equivalent acid concentrations, but
with the carboxylic acid groups pseudo-randomly distributed along the polymer
backbone. The designation r15AA, r21AA and r44AA reflect the random (r)
116
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 nature and the average number of carbons between acid groups. The structures of
these random and precise materials are shown in Scheme 1.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Scheme 1. Structures of poly(ethylene-co-acrylic) acid copolymers and the

corresponding Li- and Zn-neutralized ionomers.

The Zn2+ neutralization of the P(E-AA) copolymers has been previously

described (9). These polymer samples are designated as p21AA-56%Zn,
p21AA-116%Zn, p15AA-82%Zn and p9AA-66%Zn to reflect the initial precise
copolymer and the extent of neutralization. The Li+ neutralized material was
prepared by dissolving the acid copolymer in a 1:4 mixture of 1,4-dioxane
and 1-butanol at 90 oC, adding the appropriate amount of lithium acetate salt,
followed by filtration of the resultant precipitant. These Li-neutralized materials
are designated as p22AA-43%Li, p15AA-45%Li and r15AA-31%Li. The extent
of Zn2+ or Li+ neutralization achieved was determined using inductively coupled
plasma elemental analysis performed by Galbraith Laboratories (Knoxville, TN).
The generalized structure for the exchanged P(E-AA) copolymers is shown in
Scheme 1. The materials utilized for the NMR studies had previously been hot
pressed at 150 oC (~ 0.4 mm film thickness) for X-ray scattering analysis. These
polymer films were used as received unless otherwise noted.

Solid-State 1H NMR Spectroscopy

The solid-state 1D 1H magic angle spinning (MAS) NMR spectra were

obtained on a Bruker AVANCE-III spectrometer operating at 600.13 MHz using
a 2.5 mm broadband MAS probe, using N2 for spinning. A rotor-synchronized
Hahn spin echo pulse sequence was employed (Figure 1a), with 2.5 s /2 pulse,
16 scan averages, and a 5 s recycle delay. The rotor spinning speed for analysis
was 30 kHz, unless specifically noted. Spin regulation was maintained at 1Hz
through the experiments. It is known that significant frictional heating occurs
at high MAS speeds. The actual sample temperature was calibrated using the
207Pb chemical shift change of a secondary Pb(NO3)2 sample (10, 11), with all

temperatures reported in this chapter reflecting this correction. The 1H MAS NMR
chemical shifts were referenced to the secondary external standard adamantane,
= +1.63 ppm with respect to TMS = 0.0 ppm.
117
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The 2D 1H MAS NMR double quantum (DQ) to single quantum (SQ)
correlation experiments utilized the chemical shift anisotropy (CSA) and off-set
compensated Back-to-Back (BaBa) multiple pulse sequence for the excitation
and reconversion of the multiple quantum coherences as shown in Figure 1b
(8). The 0 2 0 1 coherence pathway was selected using a 64 step
phase cycle, 128-256 non-rotor synchronized t1 increments, a 2.5 s /2 pulse
length, 8-64 scan averages, and an excitation/reconversion length exc of 66.66
s, and a dephasing delay of 0 = 10 s. Phase sensitive detection in the F1
dimension was obtained using the States-TPPI method (12). For this excitation
time the observation of signal in the DQ NMR spectra reveals the existence
of dipolar coupling between a pair of nuclei with 1H-1H distances < 5 . The
solid-state 2D 1H NOESY MAS NMR correlation experiments utilized a standard
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

non-rotor-synchronized sequence (Figure 1c) with mixing times mix ranging

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

to 1 ms. Typically 8 to 16 scan averages were employed with 512 to 1024 t1

increments. Spectral deconvolutions were performed using the DMFIT software
package (13). The average 1H-1H dipolar coupling constants were obtained
from simulation of the 1H MAS NMR spectra using in house software written in
MATLAB 2010a (The Mathworks, Inc.).

Figure 1. Solid-state 1H MAS NMR pulse sequences utilized: a) 1D

rotor-synchronized Hahn spin echo, b) 2D DQ-SQ MAS NMR correlation
experiment using the DQ BaBa excitation/reconversion sequence, and c) 2D 1H
MAS NMR NOESY. In these sequences the rotor period is defined as R.

Ab Initio Calculations

The small acid-, ester- and Li-containing clusters were optimized in

the Gaussian 09 software (14) (Gaussian Inc., Wallingford CT) using the
6-311++G(2d,2p) basis set (15, 16), and density functional theory (DFT) utilizing
118
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Beckes three parameter exchange functional (17), and the LYP correlation
function (B3LYP). The optimizations utilized a surrounding PCM solvent with
a dielectric of = 2.3 to represent the surrounding continuum of a PE polymer.
The integral equation formalism (IEF-PCM) model was implemented for these
calculations (18). The chemical shielding tensors, , were calculated using
the Gaussian 09 program utilizing the gauge-including atomic orbital (GIAO)
method at the DFT level (19). All NMR shielding calculations for the small acid
clusters used the same exchange and correlation functionals and basis sets as for
the structure optimization. The NMR chemical shift is defined with respect to the
chemical shielding of the TMS reference species by
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

where positive values represent environments that are deshielded and resonate
at a higher frequency.

Results and Discussion

1H MAS NMR

The solid state 1H MAS NMR spectra for the different random and precise
P(E-AA) materials are shown in Figure 2. As predicted for such a simple
copolymer, two major resonances were observed corresponding to protons in
a methylene (CH2) environment ( ~ +1.5 ppm) and to protons in carboxylic
acid (COOH) environment ( ~ +12 ppm). The high chemical shifts suggest
that the carboxylic acids are strongly hydrogen bonded and were assigned to
acid-acid dimers as shown in Scheme 2. The relative intensity of the COOH
resonance scales correctly with the overall acrylic acid mol % (see Figure 2).
For example, the p9AA (22 mol %) copolymer has the largest relative COOH
intensity. The methine CH resonance ( ~+2.2 ppm), and the CH2 groups to
the COOH functionality ( ~ 1.5 to 1.7 ppm) were not resolved below Tg under
the current MAS conditions. At very high temperatures in the melt (> Tg or Tm)
the methine resonance was observable in the 1D 1H MAS NMR spectra. The 1H
NMR chemical shifts and full width at half maximum (FWHM) line widths for
the observed resonances are given in Table 1.
Minor proton environments were also observed in several of these P(E-AA)
materials, and include small amounts of adsorbed water ( ~ +4.8 ppm), and
a carboxylic ester resonance at ~ +3.4 ppm (see r15AA in Figure 2). The
assignment of this ester resonances is based on the presence of additional minor
OCH3 carbon species, (13C) ~ +51 ppm, and an additional minor C=O species,
= +174 ppm, observed in both the solution and the solid state 13C NMR. It should
be noted that this lot of the r15AA polymer had some residual adsorbed methanol
remaining from polymer precipitation and processing, which was finally removed
after extensive vacuum pumping at 80 oC (see upper spectral inset, Figure 2). For
the final (vacuum dried) r15AA sample, the carboxylic ester resonance represents
a minor proton species and corresponds to ~1 to 3% (assuming a CH3 integration)
of the entire 1H concentration. For samples containing observable carboxylic
119
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 ester, a second carboxylic acid resonance (COOHb) was also observed at ~ +7
ppm. The smaller 1H chemical shift of this acid species reflects a reduction in
the hydrogen bond strength, and is assigned to carboxylic acids that are free or
complexed to the carboxylic acid ester, instead of the strong hydrogen bonded
acid-acid dimer structure. Additional discussion concerning the local COOHb
proton environment is provided in the section on 2D correlation experiments
below. The relative integration of the COOHa and COOHb carboxylic acid
resonances are given by the a/b ratio in Table 1, and range between 10 and 3 for
the r15AA and r21AA samples, respectively.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Figure 2. The solid-state 1H MAS NMR spectra (r = 30 kHz, 332 K) for

the precise and pseudo-random P(E-AA) copolymers. The NMR spectra are
dominated by protons in the methylene and carboxylic acid environments.

Scheme 2. Proposed carboxylic acid dimer structure and carboxylic acid

carboxylic acid ester structure based on small cluster optimizations using ab
initio DFT B3LYP/6-311++(2d,2p) methods. Simulation details are given in the
experimental section.
While the methylene can be deconvoluted into a broad and narrow component
resonance under high speed (30 kHz) MAS, there are no clear 1H NMR spectral
120
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 signatures that allow quantification of the amorphous and crystalline morphologies
known to be present within these pseudo-random and precise P(E-AA) acidic
copolymers. The different contribution to the methylene line width are better
resolved by analysis of static 1H spectral line shapes and DQ buildups, but are
not discussed here. The crystalline and amorphous morphologies are clearly
identifiable in the solid-state 13C MAS NMR spectra, vary with the degree of
neutralization and preparation method, but will be presented and discussed in a
subsequent publication (Jenkins and Alam, unpublished results). The changes
in the crystallinity does not appear to significantly impact the observed 1H
NMR chemical shifts. There are some generalized observations concerning the
methylene line width that can be made. For the precise copolymer (Figure 2, left
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

side) there is an increase in the CH2 line width as the spacing between carboxylic
acids decreases. This inhomogeneous line width most likely reflects differences
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

in the relative concentrations of the crystalline and amorphous component,

along with the increased disorder in the CH2 environment for the copolymers
with higher acid concentrations. This variation in the methylene line width was
not as pronounced in the random copolymers (Figure 2, right side), where the
relative contributions to narrow and broad line width components do not change
significantly with increased spacing between acid groups. This is also consistent
with the invariance in the amorphous/crystalline morphology ratio observed for
these random materials using 13C MAS NMR (Jenkins and Alam, unpublished
results).
It has been shown that 1H MAS NMR can provide a wealth of information
concerning hydrogen bonding (20), and in general, the chemical shift of the OH
species gets larger with increasing hydrogen bond strength (21). The variation of
the 1H NMR chemical shifts for the COOH environment with temperature in the
P(E-AA) acidic copolymers is shown in Figure 3a. The > +12 ppm resonances
have been assigned to acid protons in a strong hydrogen bonding environment
such as that present in the acid-acid dimer structure (Scheme 2). For the random
materials (r15AA and r21AA), the observation of distinct COOHa and COOHb
resonances over the temperature range investigated (270 to 340 K) demonstrates
that the proton exchange rate between these two different carboxylic acid hydrogen
environments is significantly slower than the NMR time scale of the chemical shift
difference (1/exchange < 3500 Hz).
For the precise and random P(E-AA) acidic copolymers studied, the high
frequency COOHa resonance shows the largest temperature variation (Figure 3a)
ranging from = +13.3 ppm to +12.2 ppm ( = 1.1 ppm), and is very similar for
all materials. The exception to this trend is the p15AA acidic copolymer which
has a shifted chemical shift variation ranging from = +12.9 to = +11.9 ppm.
The low frequency (weakly hydrogen bonded) COOHb environment also exhibits
temperature variation (data not shown), but with a greatly reduced variation ( =
0.3 ppm). Several of these acidic copolymers have transitions (Tg and Tm) in the
temperature range studied by this NMR study (i.e. p9AA, p21AA and r15AA);
which do not appear to have an impact of magnitude of the COOH temperature
variation. The presence of semi-crystalline components appears to have little
impact on the hydrogen bond strength. The consistent 1H NMR chemical shift
temperature behavior between these different acidic copolymers demonstrates

121
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 that the local COOH environments are not significantly changed by the spacing
between AA groups, or by the random or precise arrangement of these carboxylic
acid functional groups. The lower 1H NMR chemical shift of the p15AA acidic
copolymer argues that for this material the hydrogen bonding is weaker, again
consistent with the lowest Tg observed (see Table 1). This reduced hydrogen
bond strength most likely results from the energetic competition between acid
dimer formation and the chain packing in the amorphous phase produced for this
acid group spacing.

Table 1. The 1H MAS NMR (332 K) chemical shifts and line widths for the
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

random and precise PE-AA copolymers, and the Zn2+ and Li+ ionomers
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Mol% Tg(K) (ppm)

AA [Tm] [FWHM]a
COOHa COOHb CH2
Precise
p9AA 22 295c 12.3 [778] -- 2.5 [1403, 23.2%]b
1.5 [1070,76.8% ] b
p15AA 13.3 269c 12.3 [577] -- 1.9 [1660, 36.6%] b
1.1 [670,63.4%] b
p21AA 9.5 318c,d 12.4 [650] -- 1.7 [1564, 56.5%] b
1.6 [333, 44.5%] b
Random
r15AA(dried) 13.3 347 c,d 12.3 [455] 7.3 [365] 1.65 [1940, 55.3%]
[a/b ~ 10]e b
1.54 [410, 44/7%] b
r21AA 9.5 358 c,d 12.5 [750] 7.23 [360] 1.51 [1790, 43.9%]
[a/b ~ 3.2]e b
1.48 [460, 56.1%] b
r44AA 4.5 371 c,d 12.2 [925] -- 1.7 [2130, 45%] b
1.5 [385, 55%]
Zn-Neutralized
p9AA-66%Zn 22 --i 12.2 1.53 [1110]
[1260]
p15AA-82%Zn 13.3 -- i 12.8 [970] 1.62 [975]
p21AA-56%Zn 9.5 326f 12.4 1.7 [3560, 11.5%]
[1275] 1.6 [705, 88.5%]
p21AA- 9.5 327g -- 1.55 [2365, 40%]
116%Zn 1.55 [745, 60%]
Li- Neutralized
Continued on next page.

122
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. (Continued). The 1H MAS NMR (332 K) chemical shifts and line
widths for the random and precise PE-AA copolymers, and the Zn2+ and
Li+ ionomers
Mol% Tg(K) (ppm)
AA [Tm] [FWHM]a
COOHa COOHb CH2
-- i --
p9AA-43%Li 22 14.5 1.59 [1415]
[1700]
13.2
[1710]
-- i --
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

p15AA-45%Li 13.3 14.5 1.61 [995]

[1700]
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

r15AA-31%Li 13.3 -- i 14.5 -- 1.53 [775]

[1725]
a = chemical shift ( 0.05 ppm), FWHM = full width at half maximum (5 Hz ). b The
relative % contributions of the different line width components. c Reference (1). d Tm,
since this copolymer contained significant crystalline component. e a/b = ratio of integrals
between the COOHa and COOHb species. f Reference (9). g Obtained from static 1H
NMR line width analysis. i Not measured.

A linear correlation between the solid-state 1H NMR chemical shift and the
HO hydrogen bond length has been described for inorganic crystals (22)

While this relationship was developed for crystalline materials, it can still
provide some quantitative information about the hydrogen bonding in these
P(E-AA) copolymers and ionomers. For the acidic copolymer, the observed
1H NMR chemical shifts of the COOH environment vary between = +13.3

and +12.2 ppm (Figure 3a), with Equation (2) predicting a hydrogen bond
length between 1.63 to 1.68 . The lengthening of the hydrogen bond at higher
temperatures (approaching or above Tg) is consistent with increased local motions
of the polymer chains that disrupt hydrogen bonding. Interestingly, these 1H NMR
chemical shift-predicted bond lengths are very close to the 1.66 predicted from
ab initio modeling of small carboxylic acid dimer (Scheme 2). Direct Gaussian
GIAO modeling of the acidic dimmers predicts a ~ +13.7 ppm. Details for
these simulations are provided in the experimental section. These NMR results
demonstrate that the majority of the carboxylic acid protons in the P(E-AA) acidic
copolymers reside in strongly hydrogen bonded-type dimer complexes. Only in
the acidic copolymers, where an observable concentration of carboxylic acid ester
was observed (see discussion above), are free or weakly complexed carboxylic
acids resonances present. The ab initio predicted structure of the acid-ester dimer
complex (Scheme 2) predicts an increase in the hydrogen bond length to 1.81 ,
corresponding to a predicted = 9.75 ppm (Gaussian GIAO = 10.6 ppm). While
this decrease in the 1H NMR chemical shift is in the correct direction, it is not as
123
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 large as that observed experimentally ( ~ +7.3 ppm), arguing that the hydrogen
bonding in the COOHb species is disrupted to a larger extent than predicted using
this simple model. The predicted 1H NMR chemical shift for a free, uncomplexed
COOH species is = +6.6 ppm in a continuum dielectric of = 2.3 (polyethylene)
using Gaussian GIAO methods.
This lack of free acid groups in the P(E-AA) copolymers and Zn- and
Li-neutralized ionomers (see those results below) is consistent with previous
FTIR studies on poly(ethylene-ran-methacrylic acid) ionomers (23). It should
also be noted that at a sample temperature of 270K (the lower limit of our fast
MAS VT capabilities) the temperature changes in the chemical shift had not yet
reached a plateau, suggesting that the dynamics of these carboxylic acid protons
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

that influence the chemical shift are still dominant at this temperature.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Figure 3. The temperature variation of the COOH proton chemical shift for a)
P(E-AA) acidic copolymers and b) Zn-neutralized P(E-AA) ionomers. The Tg and
Tm values (Table 1) for different P(E-AA) materials in this temperature range are
shown for comparison.

The solid-state 1H MAS NMR spectra for the partially Zn- and Li-neutralized
P(E-AA) copolymers are shown in Figure 4. The chemical shift and FWHM line
width for the different exchanged materials are presented in Table 1. Similar to
the acidic copolymers, the CH2 protons give rise to the dominant resonance, with
the COOH resonance being reduced in intensity with increasing neutralization.
For example, the inset (Figure 4) shows an expansion of the COOH resonance
for the p21AA copolymer with different levels of Zn2+ neutralization, clearly
revealing the loss of the COOH proton with Zn2+ incorporation into the material.
For the p21AA-116%Zn material there is no visible COOH 1H NMR signal
remaining showing complete removal of the carboxylic acid proton. For the
Zn-neutralized materials the COOH chemical shifts are unchanged with respect
to the un-neutralized materials, suggesting that the COOH groups retain strong
dimer-type hydrogen bonding and that the neutralization with Zn2+ does not
drastically impact the COOH environments. In contrast, the 1H NMR chemical
124
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 shift of the COOH protons in the Li-neutralized ionomers are larger (higher
frequency) arguing an increase in the hydrogen bond strength following partial
Li-neutralization. Using Equation (1), the observed = +14.5 and +13.2 ppm
predicts a shortened hydrogen bond distance between 1.56 and 1.63 .
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Figure 4. The 1H MAS NMR spectra (r = 30 kHz) for the Zn- and Li-neutralized
precise and pseudo-random P(E-AA) copolymers at 332K. The inset shows an
expansion of the COOH proton region for the p21AA Zn neutralized materials:
0% (blue), 56% (red) and 116% (black) demonstrates the loss of this proton with
increasing exchange.

This is quantitatively captured in the ab initio cluster calculations for the

Li-containing dimer and tetramer clusters shown in Scheme 3. For the partially
neutralized Li-acid dimer the predicted hydrogen bond length was 1.60
(predicted = +13.8 ppm). The preferred oxygen coordination of Li+ is 4, such
that the tetramer cluster containing the partially Li+ neutralized carboxylic acid
groups is perhaps a better description of the Li-acid clusters being formed. This
tetramer cluster has a range of hydrogen bond lengths (1.71, 1.62 and 1.51 ),
with the average 1.61 hydrogen bond distance predicting a = +13.6 ppm.
It should be noted that these ab initio clusters are representative of the type of
changes that are occurring in these ionomers, and are probably simplistic in the
description of the overall structure. For example, in the Zn-neutralized ionomers,
large ionic aggregates containing numerous Zn-acid environments have been
proposed to described the overall polymer structure, (9) and that the NMR spectra
would be an average over these different configurations. Clearly the 1H MAS
NMR chemical shift remains a sensitive probe to changes in the hydrogen bond
strength in these P(E-AA) ionomers during neutralization.
125
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Scheme 3. Proposed carboxylic acid Li+ neutralized dimer structure and the
carboxylic acid Li+ neutralized tetramer structure based on small optimized
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

cluster models obtained using ab initio DFT B3LYP/6-311++(2d,2p) methods.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

The temperature variation of the 1H NMR chemical shifts for the COOHa
resonance in the Zn-neutralized P(E-AA) ionomers is shown in Figure 3b. The
p21AA-116%Zn ionomer has no observable COOH resonance, and is therefore
these values are not included. At low temperatures the chemical shifts are similar
to those observed in the acidic copolymers, but show a marked deviation at
higher temperatures. This may reflect changes that are occurring during the
melting transition, which is in contrast to the temperature invariance of the
acidic copolymers during transition. This would suggest differences in the
polymer structure between the Zn-neutralized ionomers and the original acidic
copolymers. The temperature variation of COOH environment in the Li P(E-AA)
ionomers was very complex and revealed multiple COOH domains with differing
concentration of coordinated Li+. This differential segregation was explored
using 1H-7Li REDOR MAS NMR spectroscopy and will be presented elsewhere
(Alam, unpublished results).

2D MAS NMR Correlation Experiments

Additional details about the local structural environment are obtained using
2D NMR correlation experiments. Figure 5a shows the 2D 1H MAS NMR
DQ-SQ correlation spectrum for the p9AA acid copolymer. Similar results
were observed for the p21AA copolymer. These 2D DQ experiments provide
a sensitive tool to measure spatial proximity of protons within these materials,
with the intensity of the DQ peaks being a function of the distance between
interacting spins (~ 1/r6). The 2D DQ-SQ NMR experiments reveal the expected
interactions for the proposed structures in these P(E-AA) copolymers. There is
a strong autocorrelation peak on the diagonal (dashed line) for the CH2 proton
environments, consistent with strong dipolar couplings (close spatial interactions)
within the polymer backbone. In these experiments, a lack of an autocorrelation
resonance would show that protons of the same chemical shift (i.e. the same
environment) are not dipolar coupled to each other, or that local dynamics are
large enough to average the 1H-1H dipolar coupling. For the p9AA copolymer,
the observation of a COOH autocorrelation resonance proves close proton-proton
distances between acid groups, and supports the argument for an acid dimer
126
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 structure in these materials (ab initio predicted 1H-1H distance 2.4 ). Due
to exchange effects for the acid proton, (see previous temperature variation of
chemical shift discussion) we are unable to directly measure the 1H-1H distance
using the DQ NMR experiments. A water autocorrelation resonance was also
observed, and is consistent with either water-water clustering in the material,
or strongly absorbed water molecules with very limited mobility (24). The off
diagonal resonances in the 2D NMR DQ-SQ correlation experiments reveal
spatial contact between different types of protons, including the predicted
COOH-CH2 cross peak. By increasing the sample temperature to 332 K (above
Tg) the COOH-COOH cross peak completely disappears in the 2D NMR DQ-SQ
spectrum, consistent with the motional averaging of the 1H-1H dipolar coupling
due to increased polymer backbone dynamics. Figure 5b shows the 2D 1H MAS
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

NMR NOESY correlation experiment, which reveals similar spatial interaction

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

for the off-diagonal resonances. No information about the spatial interaction

between similar proton environments (autocorrelation type resonances) can be
extracted from these NOESY experiments as the diagonal resonances in these 2D
experiments represent magnetization that has not exchanged.

Figure 5. The a) 2D 1H MAS NMR DQ-SQ correlation spectra obtained at 273K

(r = 30 kHz) and the b) 2D 1H MAS NOESY (mix = 10 ms) at 332K (r = 30
kHz) for the precise p9AA copolymer. Several different cross peaks resulting
from through space conductivities were observed, including the carboxylic
acid auto-correlation resonance in the 2D DQ spectra. Additional discussion
provided in the text.
The 2D 1H MAS NMR DQ-SQ correlation experiments (data not shown)
for the pseudo-random P(E-AA) acid copolymers are very similar, with the
exception of those containing significant concentrations of the carboxylic ester
proton species. Additional information concerning the local structure in these
ester-containing materials can be realized by inspection of the 2D 1H MAS
NMR NOESY spectra shown in Figure 6. The protons in the carboxylic ester
environments ( ~ 3.8 ppm) are more mobile (have a longer T2), such that the
2D NMR NOESY spectrum tends to emphasizes these spectral features at mix =
127
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 10ms, allowing strong correlations to be observed even for these minor species.
Interestingly, Figure 6 shows that the strongly bonded carboxylic acid protons
(COOHa) only have spatial contacts to the CH2 protons (similar to that in Figure
5b), and does not appear to be spatially located near either the weakly hydrogen
bonded carboxylic acid (COOHb) or the carboxylic ester. Similarly, the protons
in the COOHb species only show through space contact with the ester protons,
consistent with the previous assignment of a weakly bonded interaction. These
results suggest that these different types of carboxylic acids are not within the
same cluster.

Dynamic Heterogeneity
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

The 1H MAS NMR spectra reveal that strong 1H-1H dipolar interaction are
present within both the P(E-AA) copolymers and ionomers. Below Tg, relatively
broad CH2 resonances with spinning sidebands (SSB) were observed due to strong
1H-1H dipolar couplings not completely removed by fast MAS spinning (30 kHz)

or local chain dynamics. These SSB occur at integer values of the spinning speed
around the dominant CH2 resonance, as shown in Figure 7a. The previous sections
have concentrated on analysis of the isotropic resonance region to extract chemical
shift information, and have ignored the SSB patterns present within the NMR
spectra. Simulations of the SSB patterns in the 1H MAS NMR spectra can provide
an effective or average dipolar coupling values (Deff).

Figure 6. The 2D 1H MAS NOESY (mix = 10 ms) at 332K (r = 30 kHz) for

the r15AA acid copolymer. The strongly hydrogen bonded carboxylic acid
(COOHa) environment reveals only through space proton-proton contacts with
the backbone CH2 protons, while the more weakly hydrogen bonded carboxylic
acid (COOHb) has dominant through space interactions with the methyl ester
protons (OCH3). No contact between the COOHa and COOHb protons is
observed. Additional discussion provided in the text.
128
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Figure 7. The influence of inhomogeneous T2 relaxation on: a) The observed

1D 1H MAS NMR line shape for the p15AA-45%Li ionomer as a function of
the echo time. b) Expansion of the -1 and -2 SSB spectral region showing this
heterogeneous loss, c) Change in overall signal intensity for the COOH and
CH2 resonances, d) Variation of the FWHM for the isotropic resonance and e)
Change in the effective dipolar coupling constant extracted from simulation of
the 1D 1H MAS NMR spectra in (a).

The polymer dynamics on the other hand produce averaging of the dipolar
coupling, which is readily reflected in changes to the SSB patterns in the
MAS NMR spectra. While it is tempting to discuss the local structure and
environment of these P(E-AA) copolymers as being uniform, it is important to
remember that these co-polymers are heterogeneous in both their structure and
dynamics. For example, heterogeneous dynamics were readily observed during
1H MAS NMR echo spin-spin T2 relaxation experiments, where changes in the

MAS NMR spectra result with increasing echo times. For example, Figure 7a
shows the T2-Hahn echo spectra for the pAA15-45%Li ionomer sample as a
function of rotor-synchronized echo delays (Nc = number of rotor cycles, r =
spinning speed). Even though the overall signal intensity for the CH2 and COOH
resonances reveals a single exponential decay (Figure 7c), close inspection of the
1H MAS NMR spectra (Figure 7a) reveal heterogeneous changes in the spectra.

Most importantly are the change in relative intensity of the 1 and 2 spinning
sideband (SSB) shown in Figure 7b, where the spectra have been normalized to
the isotropic peak intensity. The SSB intensities decrease with increasing echo
time as a result of differential T2 relaxation. Simulations of these SSB changes
correspond to a decrease in the effective dipolar coupling Deff (Figure 7d). These
spectral changes occur because the polymer fractions with smaller Deff (fewer
SSB) have longer T2 relaxation times, with this signal from these fractions being
preferentially retained at longer echo periods. A reduction in the line width of the
129
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 isotropic resonance as a function of echo time (Figure 7e) was also observed. The
MAS line width of the isotropic resonance is known to be impacted by dipolar
coupling involving higher order spin correlation, which are diminished (narrower
line widths) for those polymer environment that are more dynamic, and have a
longer T2 relaxation. These heterogeneous dynamics and the averaging inherent
in NMR measurements should always be considered when comparing information
obtained using other structural refinement methods (dielectric, neutron scattering,
etc.), or to structural details extracted from other NMR experiments employing
multiple pulse sequences where heterogeneous relaxation may be involved.

Conclusions
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

High speed solid-state 1H MAS NMR has been presented for a series of
precise and pseudo-random P(E-AA) acidic copolymers and the related Li- and
Zn-neutralized ionomers. Even though the 1H NMR spectra were relatively
simple, a wealth of information concerning the local carboxylic acid hydrogen
bonding environment in these materials was obtained. These NMR results
demonstrate the formation of strong acid-acid dimers, with very little free acid
present within the P(E-AA) materials. While the Zn-neutralization does not
produce a large change in the acid-acid hydrogen bonding environment, the
introduction of Li+ into the P(E-AA) ionomers actually increases the strength of
the acid hydrogen bonding.

Acknowledgments
Sandia National Laboratories is a multi-program laboratory operated by
Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Company,
for the U.S. Department of Energys National Nuclear Security Administration
under contract DE-AC04-94AL85000.
The NMR portion of this research was funded entirely through Sandias
LDRD program. KIW acknowledges funding from the National Science
Foundation Polymers Program, Grant DMR 0549116.

References
1. Baughman, T. W.; Chan, C. D.; Winey, K. I.; Wagener, K. B. Macromolecules
2007, 40, 65646571.
2. Alamo, R. G.; Jeon, K.; Smith, R. L.; Boz, E.; Wagener, K. B.;
Bockstaller, M. R. Macromolecules 2008, 41, 71417151.
3. Opper, K. L.; Fassbender, B.; Brunklaus, G.; Spiess, H. W.; Wagener, K. B.
Macromolecules 2009, 42, 44074409.
4. Wei, Y.; Graf, R.; Sworen, J. C.; Cheng, C.-Y.; Bowers, C. R.; Wagener, K.
B.; Spiess, H. W. Angew. Chem. Int. Ed. 2009, 48, 46174620.
5. Boz, E.; Wagener, K. B.; Ghosal, A.; Fu, R.; Alamo, R. G. Macromolecules
2006, 39, 44374447.
130
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 6. Mirau, P. A. A Practical Guide to Understanding the NMR of Polymers; John
Wiley & Sons, Inc.: Hoboken, 2005; p 418.
7. Schmidt-Rohr, K.; Spiess, H. W. Multidimensional Solid-State NMR and
Polymers; Academic Press Inc.: San Diego, 1994.
8. Schnell, I.; Spiess, H. W. J. Magn. Reson. 2001, 151, 153227.
9. Seitz, M. E.; Chan, C. D.; Opper, K. L.; Baughman, T. W.; Wagener, K. B.;
Winey, K. I. J. Am. Chem. Soc. 2010, 132, 81658174.
10. Bielecki, A.; Burum, D. P. J. Magn. Reson, Ser. A 1995, 116 (2), 21520.
11. Takahashi, T.; Kawashima, H.; Sugisawa, H.; Baba, T. Solid State Nucl.
Magn. Reson. 1999, 15 (2), 11923.
12. Marion, D.; Ikura, M.; Tschudin, R.; Bax, A. J. Magn. Reson. 1989, 85 (2),
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

3939.
13. Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calv, S.; Alonso, B.;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch007

Durand, J. O.; Bujoli, B.; Gan, Z.; Hoatson, G. Magn. Reson. Chem. 2002,
40, 7076.
14. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J.
R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.
Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino,
G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd,
E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A.Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenburg, S. Dapprich, A. D.
Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,
Gaussian 09, Revision A.02.; Gaussian, Inc.: Wallingford, CT, 2009.
15. McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72 (10), 56395648.
16. Krishnan, R.; Binkley, J. S.; Pople, J. A. J. Chem. Phys. 1980, 72 (1),
650654.
17. Becke, A. D. J. Chem. Phys. 1992, 98 (7), 56485652.
18. Scalmani, G.; Frisch, M. J. J. Chem. Phys. 2010, 132 (1-15), 114110.
19. Cheeseman, J. R.; Trucks, G. W.; Keith, T. A.; Frisch, M. J. J. Chem. Phys.
1996, 104 (14), 54975509.
20. Brown, S. P. Macromol. Rapid Commun. 2009, 30, 688716.
21. Lutz, H. D. J. Mol. Sci. 2003, 646, 227236.
22. Jeffrey, G. A.; Yeon, Y. Acta Cryst. 1986, B42, 410413.
23. Walters, R. M.; Sohn, K. E.; Winey, K. I.; Composto, R. J. J. Polym. Sci.,
Part B: Polym. Phys. 2002, 41, 28332841.
24. Alam, T. M.; Nyman, M.; McIntyre, S. K. J. Phys. Chem. A 2007, 111,
17921799.

131
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 8

A Solid-State NMR Study of Boric Acid Doped

in Poly(vinyl alcohol)
Kazuhiko Yamada*
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Department of Chemistry and Materials Science, Graduate School

of Science and Engineering, Tokyo Institute of Technology,
2-12-1-E4-5 Ookayama Meguro-ku, Tokyo, Japan
*E-mail: [email protected]

A multinuclear solid-state NMR study of boric acid doped

in poly(vinyl alcohol) (PVA) is presented. The experiments
used for PVA films include 1H windowed Phase-Modulated
Lee-Goldburg (wPMLG), and 11B Magic-Angle Spinning
(MAS), 11B stationary NMR, 11B Multiple-Quantum MAS, and
1H-11B CP and HETCOR at 14.1 and/or 21.6 T. The present

analysis strongly suggests the possibility that the boric acid

serves as a crosslinker with a coordination number other than
four.

Introduction
Polyvinyl alcohol (PVA) is industrially important and is used in a variety
of materials including vinylon fiber, textile sizing agent, adhesive agents, and
paper coatings (13). The versatility of PVA is due to its water solubility, and its
emulsifying, film forming, and adhesive properties. Usually a small amount of a
boric acid is added to modify the physical properties of PVA, such as mechanical
strength and water resistance. It is believed that the doped boric acid plays an
important role in crosslinking. Thus far, several schematic representations for the
crosslinked structures have been proposed, most of which are known as the di-diol
model (4, 5). In this model, a boron site involved in the formation of crosslinks
has a coordination number of four.
Nevertheless, the detailed structure for the boric acid crosslink is not
understood completely probably because it is difficult to obtain the molecular
structure of a crosslinker in a polymer that contains both crystalline and

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 non-crystalline regions. For example, X-ray diffraction patterns reflect the
crystalline parts only. Solid-state NMR is known as one of the most powerful
techniques to study non-crystalline samples (68). Moreover, it has a clear
advantage of nuclear selectivity, which means that it is possible to obtain detailed
information on the structural and electronic properties of the nucleus of interest. In
the present case, boron, the key atom for crosslinking, contains two NMR-active
nuclei: 10B ( = 2.8746 107 rad/T s, I = 3, natural abundance = 19.6 %) and
11B ( = 8.5843 107 rad/T s I = 3/2, natural abundance = 80.4 %), and PVA

contains 1H and 13C. Thus, through 1H, 11B, or 13C NMR, the interactions between
boric acid and PVA can be investigated at the molecular level. In this paper, a
study has been made of the molecular properties of boric acid doped in a PVA
film obtained by 1H and 11B solid-state NMR, and a possible conformation of the
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

boron crosslinker is proposed.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Experimental Section
All polyvinyl alcohol (PVA) films used in the present work were kindly
provided by KURARAY Co., Ltd. (Tokyo, Japan). For the preparation of PVA
film containing boric acid, PVA was dissolved in aqueous solutions of boric acid,
which was dried and heat-treated at 200C to produce the polymer film. The final
content of the boric acid in the PVA film was found to be less than 1.0 wt%.
For reference, a PVA film without boric acid was also prepared with the same
procedure without boric acid. Prior to NMR measurements, all PVA films were
dried at approximately 90C under reduced pressures for several hours in order
to remove residual water. Borax and boric acid, which were used as reference
compounds, were purchased from Wako Pure Chemical Industries, Ltd. (Tokyo,
Japan) and were employed without further purification.
All NMR experiments were performed on JEOL ECA 500 and 930
spectrometers (9), operating at frequencies of 160.5 and 298.1 MHz, respectively.
PVA films or polycrystalline borax and boric acid were packed into 4-mm rotors
of zirconium oxide. An external sample of aqueous solutions of a boric acid
was used for chemical shift referencing and pulse-length determination. The
recycle delay was typically 10 180 s. The 1H decoupling power was around 80
100 kHz. Magic angle spinning (MAS) frequencies were set to be 10.00 kHz.
1H and 11B T1 relaxation times were measured by using the saturation recovery

method. 11B 3QMQMAS experiments were carried out on the JEOL ECA 500
spectrometer. The z-filter pulse sequence proposed by Amoureux et al. (10) was
employed. For 1H-11B CP experiments, pre-saturation pulses were added to avoid
unwanted signals from the direct polarizations. The contact time was set to be 2
ms. For 1H windowed Phase-Modulated Lee-Goldburg (wPMLG) experiments
(11), the optimized pulse length was found to be 2.9 s and the 1H offset position
was -3ppm. For 1H-11B dipolar HETCOR experiments (12), the CP condition
was adjusted for the boron site with a coordination number of three, which will be
discussed in the text. All the NMR spectra were processed by Delta (JEOL USA,
Inc.) software. Spectral simulations were performed on a personal computer
using the program written by the author on MATLAB (The Math Works, Inc.).
134
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Results and Discussion
For convenience, the results will be presented in four sections: (1)
peak assignments in 11B MAS spectra of PVA film and extraction of 11B
NMR parameters, (2) quantitative analysis of boron sites as a crosslinker, (3)
investigation of the relationship between boric acids and polymer chains at the
molecular level, and (4) a proposal for the possible conformation of boric acid
doped in a PVA film.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 1. Experimental 11B MAS spectra of the boric acid/PVA film, measured at
(a) 11.7 and (b) 21.6 T.

11B MAS NMR of Boric Acid in PVA Films

Figure 1 shows the experimental 11B MAS NMR spectra of boric acid/PVA
films observed at (a) 11.7 and (b) 21.6 T with sample spinning frequencies of 10.00
135
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 kHz. At 11.7 T, the low-field peak exhibits the typical lineshape arising from
second-order quadrupolar interactions, and the high-field peak appears as a singlet.
At 21.6 T, the two 11B sites can be clearly separated because in quadrupole nuclei
the line width is inversely proportional to the strength of the applied magnetic field
(13). Unfortunately, the spectrum measured at 21.6 T exhibits shapeless peaks so
that it is difficult to analyze it. Although the spectral width is large, the spectrum
obtained at 11.7 T is used for analysis. For the assignments of the two peaks
and extraction of the 11B NMR parameters, we first study borax as a reference
compound.
The experimental 11B MAS NMR spectrum of borax observed at 11.7 T is
shown in Figure 2. As in boric acid/PVA film, there are two boron sites in the
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

spectrum; a broad line shape at the low-field and a sharp one at the high-field.
From the result of X-ray diffraction (14), this compound contains two
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

chemically different boron; one associated with a coordination number of three

and the other with four. In the rest of this paper, these boron sites are called B3
and B4, respectively. It was reported from the pioneer work of single-crystal 11B
NMR experiments (15) that the broad (low-field) and sharp (high-field) peaks
were assigned to be B3 and B4, respectively. This assignment is quite reasonable
in that the line width depends on the magnitudes of quadrupolar interactions,
reflecting the local electric-field-gradient around the boron nucleus. For example,
the electric-field-gradient of B4 is small due to the tetrahedral structure so that it
exhibits a small magnitude of the quadrupolar interactions, thereby a narrow line
shape.

Figure 2. Experimental 11B MAS spectrum of borax, measured at 11.7 T.

The analysis of the 1D MAS spectrum of borax yielded the following
parameters for B3: iso = 19 2 ppm, CQ = 2.5 0.1 MHz, Q = 0.09 0.06.
The results are reasonably consistent with those reported for B3 of borax (15).
Unfortunately, the spectral analysis of B4 provides no information on quadrupole
136
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 NMR parameters due to the shapeless peak. At this point, it is difficult to judge
whether B4 contains non-quadrupolar interaction, or just a small one. Yet, it
was found that the /2 pulse width for B4 of the PVA film was different from
that determined in boric acid solution, which implies that B4 contains some
quadrupolar interactions. Note that, in the literature, a single-crystal of borax was
used so that the 11B NMR parameters could be accurately determined.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 3. Stationary NMR spectra of (a) sodium nitride (23Na) and (b) borax
(11B).
To obtain 11B quadrupole NMR parameters for B4 of borax, the first order
effect of the quadrupolar interactions was used instead of the second order.
This method was reported by Mao et al (16). The procedure is based on the
detection of satellite transition by FTNMR, which is useful for analyzing such
a shapeless peak. As a test case, the stationary 23Na NMR spectrum of sodium
137
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 nitride is shown in Figure 3(a). The spectral width was set to be 8000 ppm.
When quadrupolar interactions exist, the satellite transitions may be observed.
Because of the wide spectral width, all NMR transitions cannot be excited. As a
result, only line singularities such as peaks and shoulders remain in the spectrum
and broad parts of the line shapes are lost. By measuring the frequency between
the remaining peaks or shoulders, the NMR parameters can be successfully
determined. Such analyses of the 1D stationary NMR spectra of sodium nitride
yielded the following 23Na NMR parameters: CQ = 1.095 0.005 MHz, Q =
0.106 0.005, which are in good agreement with literature values (17). The same
spectral analysis is applied for B4 of borax, which is shown in Figure 3(b), and
the following 11B NMR parameters were obtained: CQ = 0.5412 0.0008 MHz,
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

Q = 0.637 0.008. The isotropic chemical shift, iso, was determined to be 2

1ppm from Figure 2. In the 11B NMR spectrum of Figure 3(b), the spectral width
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

was set to be 6000 ppm, and the large broad peaks observed in the middle were
associated with B3.

Figure 4. Dependence of signal intensities on the pulse length for borax,

measured at 11.7 T.

The above procedure has been applied to obtain 11B NMR parameters for B3
and B4 of the boric acid/PVA film. The 11B NMR parameters for B3 are: iso =
192 ppm, CQ = 2.4 0.1 MHz, Q = 0.09 0.08, which are similar to borax.
Unfortunately, the NMR signals of the satellites were too small to be observed for
B4. Therefore, it was assumed that the 11B NMR parameters are the same as those
for B4 of borax.

138
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 5. (a) A pulse sequence for direct polarization (b) Plot of signal intensity
against the pulse length.

Quantitative Analysis of Boric Acid Doped in PVA Films

For quadrupole nuclei, the strength of the RF field and its pulse length
affect the line shapes as well as relative intensities (18). Therefore, information
on relative populations is sometimes difficult to obtain. Figure 4 shows
the relationship between signal intensities of B3 and B4 in borax, and the
corresponding pulse lengths. The pulse lengths were varied from 0.5 to 10.0 s
with 0.1s steps. The /2 pulses for B3 and B4 were 1.7 and 2.7 s, respectively.
The magnetization of the 11B nuclei of B3 seems to process faster than that of
B4. In order to obtain accurate quantitative information, such effect must be
considered in the spectral simulation. Let us consider the direct polarization
sequence given in Figure 5(a). The initial value of the FID is equal to the area of
the NMR peak. This initial point can be calculated by using the density matrix
at the end of the pulse width, tp. If there is quadrupolar interaction, the total
Hamiltonian during tp must be the sum of Hamiltonians of the effect of RF field,
139
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 HRF, and the quadrupolar interaction, HQ. To calculate the expected value, we
need to carry out diagonalization. The details are well explained in the literature
by Man (19). Figure 5(b) shows the calculated signal intensity of a central
transition line for I = 3/2 system against the pulse length for several different
magnitudes of quadrupolar interactions. In the calculations, the pulse power was
fixed to be 50 kHz, and the magnitudes of quadrupolar interactions were varied
from 0 to 500 kHz, as described in the figure. When there was no quadrupolar
interaction, the /2 pulse length became 5s, which corresponds to the situation
of I = 1/2 in the solution state. As the quadrupole interactions became larger, the
/2 pulse lengths tended to be shorter with lower intensities. When the magnitude
of the quadrupolar interaction was much larger than that of the pulse power, the
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

/2 pulse length became one-half of that for non-quadrupole interaction, which

is 2.5 s, with the reduced intensities of 2/5. These observations are in good
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

agreement with the experimental results. Similar calculation has been done for
the relative populations of the boric acid/PVA film.

Figure 6. Experimental and calculated 11B MAS spectra of the PVA film.
140
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 6 shows experimental and calculated 1D 11B NMR spectra of the boric
acid/PVA film measured at 11.7 T. The calculated spectrum was the sum of those
of B3 and B4. In the spectral simulation, the 11B NMR parameters obtained in the
above were fixed and the intensity of each sub-spectrum was varied until the total
theoretical spectrum fitted to the experimental one. From the present analysis, the
ratio of B3 to B4 was found to be around 14. Contrary to the di-diol model in
which the coordination number of boron crosslinks is four, the population of B4
was found to be minor. The crystal structure of a boric acid (20) shows that the
boron site has the B3 structure. Thus, it may be expected that crystalline boric acid
(with B3 structure) is present in the PVA polymer. There is also the possibility that
B3 plays an important role in crosslinking because of its high abundance.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 7. (a) 2D experimental 11B 3QMQMAS spectrum of the boric acid/PVA

film. (b) 11B stationary NMR spectra of (upper) the boric acid/PVA film and
(lower) the crystalline boric acid.

141
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Interactions between Boric Acid and Polymer Chains

Next we explore the relationship between doped boric acid and the polymer
chains at the molecular levels, focusing in particular on the majority component,
B3. Figure 7(a) shows a contour plot of the 11B z-filter 3QMQMAS spectrum
of the PVA film measured at 11.7 T. Both the F1 and F2 projections of the 2D
MQMAS spectra are given on the side and at the top, respectively. Obviously, one
major peak can be observed, suggesting that B3 is a single component. Figure 7(b)
shows experimental 11B stationary spectra of the boric acid/PVA film (upper) and
crystalline boric acid (lower), obtained at 11.7 T. Relatively similar line shapes are
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

obtained for both NMR spectra. Apparently, B4 is too small to be detected. From
Figure 7, it is expected that the boric acid/PVA film contains a single boron site, B3,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

in which the chemical environment may be close to that of the polycrystalline boric
acid. However, it was also found that T1 relaxation time for the boric acid/PVA
film, which was determined to be less than 2 s, was much shorter than that for the
polycrystalline boric acid (36 s). This is probably because, in polymers, there are
additional relaxation mechanisms, or the electronic properties around the boron
nuclei may be different.
Horii and co-workers previously reported the 1H chemical shifts of PVA by
analyzing high-resolution 1H NMR spectra in details (21). According to their
results, there are three kinds of hydroxyl groups involved in intramolecular and
intermolecular hydrogen and non-hydrogen bonding. The chemical shifts of these
protons are reported to be in the range of 2.85-5.80 ppm. The chemical shifts of
CH2 and CH groups are found at 1.40-1.60 and 4.20 ppm, respectively. Figure 8
shows 1H windowed Phase-Modulated Lee-Goldburg (wPMLG) spectra of PVA
films with (a) and without (b) boric acids. The present 1H NMR spectrum of
the PVA film could be assigned on the basis of the results reported by Horii and
co-workers (16). It is important to point out that there is no signal corresponding
to polycrystalline boric acid, whose chemical shift was 5.5-7.0 ppm, observed
under the same experimental conditions of 1H wPMLG experiments (data not
shown). As shown in Figure 8, there is no difference in lineshapes between the two
NMR spectra, indicating that the chemical shift of B3 may be overlapped on the
major peaks of the CH and CH2 groups, or that the PVA polymer does not contain
polycrystalline boric acid.
The adequate 1H-11B CP conditions were evaluated in order to investigate the
chemical environment around B3 in the PVA film. Figure 9 shows experimental
1H-11B CP MAS spectra of boric acid/PVA films in which the 11B CP power was

varied from 45 to 90% at 5% step while the 1H CP power was fixed. The adequate
CP conditions were found to be quite different between B3 and B4 due to the fact
that they depend on the magnitude of quadrupolar interactions. The maximal peaks
for B3 and B4 were at 60 and 90%, respectively. For reference, the adequate CP
condition for the polycrystalline boric acid was found to be at 40%, suggesting
that the proton environment around B3 is different from that of the crystalline
counterpart.

142
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 8. 1H wPMLG NMR spectra of PVA films (a) with and (b) without boric
acids.
It is useful to carry out heteronuclear correlation experiments in order to
understand the relationship between the doped boric acid and the polymer chains.
Figure 10 shows a contour plot of the 1H-11B HETCOR spectrum of the PVA film
observed at 11.7 T. Both the F1 (1H) and F2 (11B) projections of the 2D spectra
are also given on the side and at the top, respectively. In the experiment, the 11B
CP power was set to be 60%, which is the appropriate CP condition for B3. In
this setting, little cross peak could be observed for B4. It can be seen that there
are strong cross peaks between B3 and the CH and CH2 groups. From a careful
analysis of the F1 projection, i.e., the high-resolution 1H NMR spectrum in which
there is a small shoulder at the peak arising from the CH group, it is deduced
that there are small cross peaks between B3 and the OH groups. Importantly,
no cross peak between B3 and the polycrystalline boric acid is confirmed. In
summary, B3 is not attributed to the polycrystalline boric acid itself, but it is
surrounded by the PVA main and side chains. It should be noted that a cross peak
143
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 appears via dipole-dipole interactions, which implies that the two atoms may not
be connected by chemical bonds, but are spatially close to each other.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

Figure 9. 1H-11B CP profiles of boric acid/PVA films in which the 11B CP power
was varied from 45 to 90 % with 5% step while the 1H CP power was fixed.

Figure 10. 2D experimental 1H-11B HETCOR spectrum of boric acid doped

in PVA.

144
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Possible Conformation of Boric Acid Doped in PVA Films
From the authors previous work and experience, the mechanical strength of
a PVA film is known to depend critically on the amount of boric acid. The higher
the concentration of the doped boric acid, the greater is the mechanical strength of
the PVA film. Therefore, it is deduced that a doped boric acid somehow forms a
chemical bond between the polymer and the boric acid. Assuming that boric acid
and PVA are linked by hydrogen bonds, the present results indicate that the boric
acid crosslinker has a coordination number of three instead of four.

Conclusion
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

This work aims to investigate the molecular properties of boric acid doped
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

in PVA film via multinuclear solid-state NMR. Contrary to expectations from

the previously proposed the di-diol model, the major boron site was found to
have a coordination number of three. The chemical environment of the boron
site is quite different from that of the polycrystalline boric acid. The boron is
surrounded by polymer main chain and side chains. Because the amount of doped
boric acid is related to the polymer properties such as mechanical strengths and
water-resistance, it can be assumed that the doped boric acid are linked to the
polymer, e.g., through hydrogen bonds. To the best of the authors knowledge, this
is the first report of boron/PVA crosslinker with a coordination number of three.
Finally, it has been demonstrated that solid-state NMR is useful for studying the
molecular properties of crosslinkers in crystalline and non-crystalline polymers,
and it will be applied in this laboratory to a variety of crosslinkers in rubbers and
gels in the near future.

Acknowledgments
Polyvinyl alcohol was kindly supplied by KURARAY Co., Ltd. (Okayama,
Japan). This work was supported by the Ministry of Education, Culture,
Sports, Science and Technology (MEXT) Grant-in-Aid for Young Scientists
(B) (22750009). I thank Dr. Tadashi Shimizu and Kenzo Deguchi for their
technical assistance with the solid-state NMR experiments at National Institute
for Materials Science. Finally I wish to thank Takahiro Ishii, Hideki Kamada,
Hiroshi Kawai, and Toshiaki Kobayashi for helpful comments on the fabrication
and properties of PVA films.

References
1. Sakurada, I. International Fiber Science and Technology Series 6, Polyvinyl
Alcohol Fibers; Marcel Dekker Inc.: New York, 1985.
2. Finch, C. A. Polyvinyl AlcoholDevelopments; John Wiley & Sons Ltd.:
Oxford, England, 1992.
3. Goodship, V.; Jacobs, D. Polyvinyl Alcohol: Materials, Processing and
Applications; Smithers Rapra Technology, 2009.
145
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 4. Kurokawa, H.; Shibayama, M.; Ishimaru, T.; Nomura, S. Polymer 1992, 33,
2182.
5. Deuel, H.; Neukom, A. Makromol. Chem. 1949, 3, 113.
6. Schmidr-Rohr, K.; Spiess, H. W. Multidimensional Solid-state NMR and
Polymers; Academic Press Limited, New York, 1994.
7. Saito, H.; Ando, I.; Naito, A. Solid-state NMR Spectroscopy for Biopolymers:
Principles and Applications; Springer: Amsterdam, Netherlands, 2006.
8. Ando, I., Asakura, T., Eds.; Studies in physical and theoretical chemistry 84,
Solid State NMR of Polymers; Elsevier Science: Amsterdam, 1998.
9. Tsukuba Magnet Laboratory, National Institute for Materials Science,
Tsukuba city, Japan. http://www.nims.go.jp/TML/.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

10. Amoureux, J. P.; Fernandez, C.; Steuernagel, S. J. Magn. Reson. 1996, 123,
116.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch008

11. Vinogradov, E; Madhu, P. K.; Vega, S. Chem. Phys. Lett. 2002, 354, 193.
12. Caravatti, P; Braunschweiler, L; Ernst, R. R. Chem. Phys. Lett. 1983, 100,
305.
13. Slichter, C. P. Principles of Magnetic Resonance, Springer Series in Solid-
state Sciences; Springer: Berlin, 1996.
14. Gainsford, G. J.; Kemmitt, T.; Higham, C. Acta Cryst. 2008, E64, i24.
15. J. D. Cuthbert, J. D.; Petch, H. E. J. Chem. Phys. 1963, 38, 1912.
16. Mao, D; Bray, P. J. Solid State Nucl. Magn. Reson. 1992, 1, 255.
17. Weiss, A. Z. Naturforsch. 1960, 15a, 536.
18. Man, P. P. J. Magn. Reson. 1988, 77, 148.
19. Man, P. P. Molec. Phys. 1993, 78, 307.
20. Zachariasen, W. H. Z. Kristallogr. 1934, 88, 150.
21. Masuda, K; Kaji, H; Horii, F. Polym. J. 2001, 33, 190.

146
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 9

NMR Characterization of Canopy Dynamics in

Nanoscale Ionic Materials
Michael L. Jespersen,1,2 Peter A. Mirau,*,1 Ernst von Meerwall,3
Richard A. Vaia,1 Robert Rodriguez,4,5 Nikhil J. Fernandes,4
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009

and Emmanuel P. Giannelis4

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

1Materials and Manufacturing Directorate, Air Force Research Laboratory,

Wright-Patterson AFB, OH 45433
2UES, Inc., Dayton, OH 45432
3Department of Physics, University of Akron, Akron, OH 44325
4Department of Materials Science and Engineering and School of Applied

and Engineering Physics, Cornell University, Ithaca, NY 14853

5Current address: Intel Corp., 2501 NW 229th Ave., Hillsoboro, OR 97124
*E-mail: [email protected]

Nanoscale ionic materials (NIMs) are organic-inorganic hybrids

in which a core nanoparticle is functionalized with a covalently
attached corona and an ionically tethered polymer canopy.
NIMs exhibit liquid-like character under ambient conditions
in the absence of solvent and are of interest for a variety
of applications. We have used nuclear magnetic resonance
(NMR) relaxation and pulsed-field gradient (PFG) diffusion
experiments to measure the canopy dynamics of NIMs prepared
from 18-nm silica nanoparticles. NMR studies show that the
fast (ns) local dynamics of the canopy are insensitive to the
presence of the silica nanoparticles. Canopy diffusion in the
NIMs is slowed relative to the neat copolymer, but not all
canopy molecules are slowed equally due to crowding at the
nanoparticle surface, resulting in a strongly bound fraction at
the surface and a weakly bound outer sphere. Electrostatic
interactions with other ionic (Na+) species alter the dynamics
by screening interactions with the nanoparticle.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Introduction
Nanoscale ionic materials (NIMs) are an emerging class of functionalized
nanoparticles that have recently generated a great deal of interest due to their
unique characteristics (13). NIMs are organic-inorganic hybrids consisting of a
core nanoparticle functionalized with a covalently bound ionic corona and a bulky
counter-ion as the canopy (Figure 1). NIMs are nanoparticles stabilized by an
organic ionic liquid coating and either behave as liquids at room temperature or
undergo reversible, macroscopic solid-to-liquid transitions near room temperature
in the absence of solvent. The library of NIMs reported to date has grown to
include NIMs based on metal oxides (SiO2 (1, 4, 5), Fe2O3 (5), TiO2 (2), and ZnO
(6)), metals (Au (79), Pt (7, 9), Pd (7), and Rh (7)), quantum dots (10), carbon
nanotubes (11, 12), nanorods (13) and fullerols (14).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009

The room temperature liquid character of NIMs allows for the design of
liquids that retain the unique size-dependent properties of the core nanoparticles
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

(e.g., magnetic fluids with high viscosity, etc.). Exploiting the versatility of
the NIMs platform in widespread applications (9, 15, 16) will depend on the
degree to which NIMs properties can be tuned by modifying their structure and
composition (core shape and size, canopy composition and structure, and ionic
content). Flexibility in NIMs design also presents an opportunity to investigate
the influence of structure and composition on the dynamics of the canopy, which
will likely have a strong impact on the macroscopic properties of NIMs.
In this study we report the NMR relaxation and diffusion studies of silica-
based NIMs with a polymer canopy in order to determine the relationship between
chemical structure and dynamics. The results show how the properties of the
canopy relate to the macroscopic properties.

Experimental Section
Materials
Silica nanoparticles (LUDOX HS-30 colloidal silica, 30 wt. % suspension
in H2O, 18-nm diameter, Aldrich Chemical Co.), 3-(Trihydroxysilyl)-1-
propanesulfonic acid (30-35 wt. % solution in H2O, Gelest, Inc.), and Jeffamine
M-2070 Polyetheramine (Mn=2263, Mw=2334, PDI=1.03, Huntsman Corporation
) were used as received. Deionized water (18.2 M-cm) was purified using a
Barnstead Nanopure system.

NIMs Preparation
Silica nanoparticles were functionalized according to a previously reported
procedure using (trihydroxysilyl)-1-propanesulfonic acid (SIT, 30-35 wt. %
in deionized water) (1). The functionalized nanoparticles were purified by
dialysis to remove excess SIT and then stirred in the presence of an ion exchange
resin (HCR-W2) for at least 48 hours in order to exchange sodium atoms for
protons. NIMs were prepared by titration of the bound sulfonic acid groups with
M-2070 in deionized water. The equivalence point of the titration represents
150
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 a 1:1 ratio of canopy to bound corona. The canopy:corona ratio can be tuned
by adding the appropriate amount of M-2070 relative to that required to reach
stoichiometric equivalence. Samples were prepared at 100% (NIMs100) and
60% (NIMs60) neutralization of the anionic surface of the silica nanoparticles.
Once the appropriate amount of M-2070 had been added to generate the targeted
corona:canopy ratio, solvent was removed under vacuum at 35-45 C for several
days, resulting in a pale yellow product that flows as a viscous liquid at room
temperature.

NMR Characterization
13C
relaxation experiments were conducted using a Tecmag Apollo 500 MHz
NMR spectrometer equipped with a DOTY Scientific magic-angle spinning probe.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009

Diffusion coefficients were measured at 50.5 C using proton pulsed-gradient spin-

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

echo NMR as previously described (17) using the stimulated echo sequence with
pulsed magnetic field gradients of magnitude G = 652 Gauss/cm.
The magnetization decay as a function of gradient field strength was used to
determine the diffusion coefficient. In the case of a single diffusing species, the
signal decay is given by (18, 19)

where is the gyromagnetic ratio of the nucleus of interest, is the time of the
gradient pulse, is the diffusion delay time in the pulse sequence, G is the gradient
strength, and the diffusion coefficient D is related to the hydrodynamic radius rH
by the Stoke-Einstein equation (19, 20).

In those cases where single-exponential decay is not observed (vide infra), we

have chosen to fit the data to a stretched exponential function, given by

where is a fitting parameter relating to the deviation from single-exponential

behavior. In cases where the stretched exponential did not give a good fit, the data
was fit to a double exponential function, given by

where the subscripts f and s refer to fast and slow diffusion processes.

Results and Discussion

The dynamics of the NIMs corona-canopy interface were evaluated from the
13C spin-lattice relaxation times and the proton diffusion coefficients measured
151
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 by pulsed-field gradient (PFG) NMR. The NIMs of interest consist of an 18-nm
silica nanoparticle core functionalized with a monolayer of trihydroxysilylpropyl
sulfonic acid (SIT). The anionic SIT corona is paired with an amine-terminated
(cationic) ethylene oxide/propylene oxide diblock copolymer canopy (Figure 1).
NIMs have been described as monolithic hybrid materials with liquid properties,
where each nanostructure carries its share of the solvent (1, 21). For this reason,
a canopy: corona ratio of 1:1 was chosen as the most reasonable starting point for
examining the molecular-level dynamics of NIMs using NMR.
The first step in characterizing the dynamics by NMR relaxation is
measurement of the T1 relaxation across a range of temperatures and identification
of the 13C T1 minimum. The T1 minimum occurs when the molecular motions are
near the spectrometer frequency (125 MHz) (22). The dynamics of the free canopy
(M-2070) and the NIMs canopy in neat, dry samples were measured from the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009

carbon T1 of the methylene carbon peak at 74 ppm, which is attributed primarily

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

to the ethylene oxide segments. Figure 2 compares the temperature dependence

of the carbon T1s for the bulk M-2070 and NIMs samples prepared with 100%
and 60% neutralization of the sulfonic acid groups. The most remarkable feature
of this plot is that the relaxation times versus temperature are identical and that
the T1 minimum occurs at the same temperature for the NIMs100 and NIMs60
samples as for the bulk M-2070.

Figure 1. NIMs diagram showing the core, corona and canopy.

A more quantitative understanding of canopy dynamics in NIMs can be

obtained by fitting the temperature dependence of the relaxation to a specific
model. A number of models for the C-H autocorrelation function were explored
to determine the rotational correlation times from the T1 relaxation data. We
observed that the T1 minimum (0.1 s) calculated from the isotropic reorientation
model for the C-H autocorrelation function did not adequately predict the
152
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 experimentally observed T1 minimum (0.26 s) for either the neat canopy or the
NIMs. Given that the canopy molecules are smaller than an entangled polymer
in the melt, we have chosen to use the Lipari-Szabo (LS) model (23, 24) for the
autocorrelation function rather than a more sophisticated model like the modified
KWW model (25) commonly used for high molecular weight polymers. The
key assumption of the LS model is that the overall molecular reorientation is
decoupled from faster internal motions, and the autocorrelation function C(t) is
given by the product of the correlation functions for overall reorientation C0(t)
and the internal motion Ci(t) as

The spectral densities in the LS model are given by

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

where S2 is the generalized order parameter and where i is

the correlation time for rapid librational motions and r is the rotational correlation
time.
The LS model with a generalized order parameter of 0.4 gave a good fit to the
T1 minimum (268 K) and was used to calculate correlation times as a function of
temperature. The key conclusions from the T1 fits are that the carbon relaxation
is due to a combination of rapid (10 ps) librational motions and slower (0.1-2 ns)
reorientation. The local dynamics do not appear to depend on the presence of the
18-nm silica nanoparticle for NIMs100 and NIMs60.
The silica nanoparticles in the NIMs are functionalized at nearly every
surface hydroxyl group on the silica nanoparticle, resulting in an estimated corona
density of 4.5/nm2. At this density, the average distance between between acid
groups is less than the radius of gyration of the canopy (1.25 nm) calculated for a
freely jointed chain (26). The resultant crowding could result in stretched chain
configurations extending away from the surface, analogous to a brush. At lower
canopy densities than those studied here, configurations where the chain folds
back onto the nanoparticle surface, allowing interactions between the oxygen
atoms in the ethylene oxide/propylene oxide monomers and the sulfonic acid
groups or water molecules strongly bound to the sulfonic acids, are more likely.
We are currently investigating systems with the possibility for these interactions,
which we would expect to result in slower local dynamics.
One of the key assumptions in NIMs design is that each functionalized
nanoparticle carries its share of the solvent (i.e., the canopy) because the charge on
the ionic terminal functionality of the corona is balanced by a strongly associated,
oppositely-charged canopy molecule (1, 21). In the case of strong association
of the canopy with the nanoparticle surface, significant differences between the
diffusion coefficients measured for the bulk M-2070 and NIMs canopy would be
expected. To evaluate this hypothesis, we have measured the canopy diffusion
153
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 coefficients using the stimulated-echo PFG pulse sequence with corrections for
gradient artifacts (27).
The first measure of the interaction between the anionic terminal functionality
on the nanoparticle and the cationic corona is to compare the diffusion coefficients
in solutions of the canopy and the NIMs. Figure 3 compares the diffusion curves
for 8% (w/w) solutions of the canopy and the NIMs in D2O at 50C. The most
notable feature is that the diffusion coefficients are very similar for the two
samples. In both cases, the data was fit to a stretched exponential function and
the results are listed in Table I (28).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 2. The carbon T1 vs. temperature for M-2070 (), NIMs100 (), and
NIMs60 ().

The magnetization decay for M-2070 in D2O appears as a single exponential

(=0.98) within experimental error, as expected for a nearly monodisperse
polymer (PDI=1.03) (29). The measured diffusion coefficient for M-2070 closely
agrees with the value of 2.1 x 10-6 cm2/s calculated from the Stokes-Einstein
relation using the viscosity of D2O at 50 C and a hydrodynamic radius of 1.25
nm, calculated assuming a freely jointed chain (26). We note that rH =1.25
nm compares favorably to the value determined from dynamic light scattering
experiments (1.15 nm). By comparison, the decay curve for the NIMs100 canopy
shows only a 15% decrease in the diffusion coefficient compared to dissolved
M-2070. We can estimate the diffusion coefficient expected if the canopy
molecules were bound to the nanoparticles by calculating the hydrodynamic radius
for the silica nanoparticle (9 nm), the corona (0.5 nm) and the canopy (2 x 1.25
nm). This gives a hard-sphere radius of 12 nm, from which we would expect the
154
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 diffusion coefficient to decrease by a factor of 9.6 relative to a dissolved canopy
molecule in D2O. The fact that we observe only a 15% decrease in the diffusion
coefficient shows that the NIMs canopy does not undergo hard-sphere diffusion
in solution. Rather, the canopy is exchanging between free molecules and other
nanoparticles in D2O on the time scale (ms) of the diffusion measurements.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 3. PFG NMR diffusion for M-2070 () and NIMs100 () in D2O.

(Reproduced with permission from reference (28). Copyright 2010 American
Chemical Society.)

Table I. Self-diffusion coefficients and fitting parameters for the canopy

and NIMs in D2O at 50C
Sample D (cm2/s)
M-2070 1.84 10-6 0.98
NIMs100 1.56 10-6 0.73

Figure 4 compares the self-diffusion decay curves for neat, dry samples of
the free canopy, NIMs100, and NIMs100 spiked with NaCl (SIT:Na=1:1). We
observe curvature on the semilog plots and good fits to the stretched exponential
function for free canopy and Na+-spiked NIMs100. The data for NIMs 100 is fit
by a double exponential function, and the diffusion coefficients and fit parameters
are listed in Table II.

155
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 4. PFG diffusion for bulk M-2070 (), NIMs100 () and NIMs100 spiked
with NaCl ().

Table II. Self-diffusion coefficients for neat M-2070, NIMs100 and NIMs100
spiked with NaCl
Sample D (cm2/s)
Bulk M-2070 5.310-8 0.83
3.210-8 (fast) --
NIMs100
1.410-9 (slow)
NIMs100/NaCl 1.310-8 0.90

Even though we observed single exponential decay for M-2070 in D2O

solution, Figure 4 and Table II show that a stretched exponential function (=0.83)
is required to fit the data for the neat canopy. Since the solution experiments
eliminate polydispersity as an explanation for curvature in the signal decay, we
believe the curvature results from self-association of M-2070 in the neat state,
which does not occur when the molecule is dissolved in D2O at 50 C (Figure 3).
We note that M-2070 is similar in structure to Pluronic block copolymers, which
are known to self-associate in solution, giving rise to distributions in diffusion
coefficients (30). We have also observed self-association of M-2070 in dilute
solutions of by GPC and in both dilute solutions of M-2070 and neat M-2070 by
small angle x-ray scattering (not shown).

156
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 5. Diagram showing the crowding of polymer chains at the surface of

the nanoparticle.

Canopy diffusion in NIMs100 is significantly slower than in bulk M-2070

and is best fit with a double exponential function. The double-exponential nature
of the curve demonstrates that there are two populations of canopy polymers, one
of which is not exchanging on the time scale (ms) of the diffusion experiment.
We believe this can be explained by polymer crowding at the particle surface
because the radius of gyration of the M-2070 is larger than the spacing between
anionic groups on the nanoparticle surface, as shown schematically in Figure 5.
A balance of electrostatic attractions between sulfonic acid and amine groups
and entropic repulsion arising from stretched chain conformations would result
in a diffuse electrostatic coupling zone extending away from the anionic corona.
This hypothesis is further supported by the observation that the slow diffusion
coefficient (1.4 x 10-9 cm2/s) gives a hard-sphere radius of 14 nm, which closely
agrees with the estimated hard-sphere radius (12 nm), defined by the silica
nanoparticle diameter plus the SIT layer and two times the hydrodynamic radius
of the canopy. The outer sphere canopy molecules are attracted to the surface
by Coulombic interactions, but they are unable to reach the surface due to the
molecular crowding. We note that the fast diffusion in the NIMs100 sample
is similar to that of bulk M-2070. We previously reported that the diffusion
157
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 coefficients are not sensitive to the diffusion delay time, which demonstrates the
diffusion is not restricted to the surface of the nanoparticle (28).
Canopy exchange is slowed in neat NIMs100, but it is greatly increased in
the presence of NaCl. This is illustrated in Figure 4 () for the sample spiked
with NaCl at a level of one sodium atom per acid group. Under these conditions,
the diffusion coefficient is an average of that observed for hard-sphere diffusion
and for the bulk canopy. The diffusion coefficient measured for the NaCl-spiked
sample is similar to that reported for another NIMs sample which we have
subsequently determined also contained approximately one sodium ion per
acid group (28). This behavior suggests that the presence of small amounts of
contaminant ions can have a dramatic effect on the canopy dynamics in NIMs
and that the diffusion properties can potentially be tuned by controlling the
concentration and valence of additional ions in the NIMs.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Conclusion
In summary, we have studied the dynamics and diffusion of the organic
canopy in SiO2 NIMs using a combination of NMR relaxation and pulsed-field
gradient diffusion. The local molecular relaxation times are similar for the bulk
canopy and the completely neutralized NIMs, suggesting that the dynamics are
liquid-like at the surface of the nanoparticle. The diffusion experiments reveal two
different populations of canopy polymers with different translational diffusion
characteristics. The canopy polymers strongly tethered to the surface exhibit
hard-sphere-like diffusion, whereas the outer sphere layer of the canopy exhibits
bulk-like M-2070 diffusion. These data are explained with a model in which
canopy crowding at the nanoparticle surface prevents all of the amine-terminated
canopy from interacting directly with the acid groups on the surface. We are
currently investigating the effects of canopy molecular weight, canopy:corona
ratios, nanoparticle size, and contaminant ion concentrations on the dynamics and
rheology of NIMs.

Acknowledgments
Funding provided by the Air Force Office of Scientific Research is gratefully
acknowledged. The diffusion portion of this work was supported by the National
Science Foundation under Grant No. DMR 04 55117. This publication is based
on work supported by Award No. KUS-C1-018-02, made by King Abdullah
University of Science and Technology (KAUST). A portion of this research was
carried out while M. Jespersen was a National Research Council Associate at
the Air Force Research Laboratory. Jeffamine M-2070 was generously donated
by Huntsman Corporation (Houston, TX). The authors would like to thank
George Fultz and Timothy Reid (University of Dayton Research Institute) for
viscosity and ICP-MS data supporting this research. M. Tchoul and H. Koerner
(AFRL/RX) contributed GPC and SAXS data, respectively, in support of this
study. The authors also thankfully acknowledge Dr. Rajiv Berry and Phuong Ngo
(AFRL/RX) for helpful discussions regarding this work.
158
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 References
1. Rodriguez, R.; Herrera, R.; Archer, L. A.; Giannelis, E. P. Adv. Mater. 2008,
20, 43534358.
2. Bourlinos, A. B.; Chowdhury, S. R.; Herrera, R.; Jiang, D. D.; Zhang, Q.;
Archer, L. A.; Giannelis, E. P. Adv. Funct. Mater. 2005, 15, 12851290.
3. Rodriguez, R.; Herrera, R.; Bourlinos, A. B.; Li, R.; Amassian, A.; Archer, L.
A.; Giannelis, E. P. Appl. Organomet. Chem. 2010, 24, 581589.
4. Bourlinos, A. B.; Chowdhury, S. R.; Jiang, D. D.; An, Y.-U.; Zhang, Q.;
Archer, L. A.; Giannelis, E. P. Small 2005, 1, 8082.
5. Bourlinos, A. B.; Herrera, R.; Chalkias, N.; Jiang, D. D.; Zhang, Q.;
Archer, L. A.; Giannelis, E. P. Adv. Mater. 2005, 17, 234237.
6. Bourlinos, A. B.; Stassinopoulos, A.; Anglos, D.; Herrera, R.;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009

Anastasiadis, S. H.; Petridis, D.; Giannelis, E. P. Small 2006, 2, 513516.

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

7. Warren, S. C.; Banholzer, M. J.; Slaughter, L. S.; Giannelis, E. P.; DiSalvo, F.

J.; Wiesner, U. B. J. Am. Chem. Soc. 2006, 128, 1207412075.
8. Zheng, Y.; Zhang, J.; Lan, L.; Yu, P.; Rodriguez, R.; Herrera, R.; Wang, D.;
Giannelis, E. P. ChemPhysChem 2010, 11, 6164.
9. Voevodin, A. A.; Vaia, R. A.; Patton, S. T.; Diamanti, S.; Pender, M.;
Yoonessi, M.; Brubaker, J.; Hu, J.-J.; Sanders, J. H.; Phillips, B. S.;
MacCuspie, R. I. Small 2007, 3, 19571963.
10. Sun, L.; Fang, J.; Reed, J. C.; Estevez, L.; Bartnik, A. C.; Hyun, B.-R.;
Wise, F. W.; Malliaras, G. G.; Giannelis, E. P. Small 2010, 6, 638641.
11. Bourlinos, A. B.; Georgakilas, V.; Boukos, N.; Dallas, P.; Trapalis, C.;
Giannelis, E. P. Carbon 2007, 45, 15831595.
12. Bourlinos, A. B.; Georgakilas, V.; Tzitzios, V.; Boukos, N.; Herrera, R.;
Giannelis, E. P. Small 2006, 2, 11181191.
13. Bhattacharjee, R. R.; Li, R.; Estevez, L.; Smilgies, D.-M.; Amassian, A.;
Giannelis, E. P. J. Mater. Chem. 2009, 19, 87288731.
14. Fernandes, N.; Dallas, P.; Rodriguez, R.; Bourlinos, A. B.; Georgakilas, V.;
Eiannelis, E. P. Nanoscale 2010, 2, 16531656.
15. Eastman, J. A.; Choi, S. U. S.; Li, S.; Yu, W.; Thompson, L. J. Appl. Phys.
Lett. 2001, 78, 718720.
16. Park, S. I.; Lim, J. H.; Kim, J. H.; Yun, H. I.; Roh, J. S.; Kim, C. G.; Kim, C.
O. Phys. Status Solidi B 2004, 241, 16621664.
17. Iannacchione, G.; Von Meerwall, E. J. Polym. Sci., Part B: Polym. Phys.
1991, 29, 659668.
18. Price, W. S. Concepts Magn. Reson. 1998, 104, 299336.
19. Johnson, C. S. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203256.
20. Price, W. S. Concepts Magn. Reson. 1998, 104, 197237.
21. Agarwal, P.; Qi, H.; Archer, L. A. Nano Lett. 2010, 10, 111115.
22. Mirau, P. A Practical Guide to the NMR of Polymers; John Wiley & Sons:
Hoboken, 2004; p 418.
23. Lipari, G.; Szabo, A. J. Am. Chem. Soc. 1982, 104, 45464559.
24. Lipari, G.; Szabo, A. J. Am. Chem. Soc. 1982, 104, 45594570.
25. Qiu, X.; Moe, N. E.; Ediger, M. D.; Fetters, L. J. J. Chem. Phys. 2000, 113,
29182926.

159
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 26. Hiemenz, P. C. Polymer Chemistry, The Basic Concepts; Marcel Dekker,
Inc.: New York, 1984.
27. Von Meerwall, E.; Kmat, M. J. Magn. Reson. 1989, 83, 309323.
28. Jespersen, M. L.; Mirau, P. A.; von Meerwall, E.; Vaia, R. A.; Rodriguez, R.;
Giannelis, E. P. ACS Nano 2010, 4, 37353742.
29. Hakansson, B.; Nyden, M.; Soderman, O. Colloid Polym. Sci. 2000, 278,
399405.
30. Nilsson, M.; Hakansson, B.; Soderman, O.; Topgaard, D. Macromolecules
2007, 40, 82508258.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch009
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

160
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 10

Dynamics of Disordered Structure of

-Conjugated Polymers Investigated by
Solid-State NMR
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

N. Asakawa,*,1 Y. Inoue,2 T. Yamamoto,3 R. Shimizu,4 M. Tansho,4

and K. Yazawa2
1Department of Chemistry and Chemical Biology, Graduate

School of Engineering, Gunma University, 1-5-1 Tenjincho, Kiryu,

Gunma 376-8515, Japan
2Department of Biomolecular Engineering, Graduate School of Bioscience

and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta-cho,

Midori-ku, Yokohama, Kanagawa 226-8501, Japan
3Chemical Resources Laboratory, Tokyo Insitute of Technology, 4259

Nagatsuta-cho, Midori-ku, Yokohama, Kangawa 226-8503, Japan

4National Institute for Materials Science, 3-13 Sakura, Tsukuba,

Ibaraki 305-0003, Japan

*E-mail: [email protected]

Stochastically excitable threshold units using functional

materials will potentially be among the key devices for the
production of noise-driven bio-inspired sensors and information
processors with ultra-low energy consumption. In particular, a
noise generator and threshold unit will be able to be fabricated
utilizing the fluctuations found in materials that include
structural, electric dipole, magnetic, and spin. This article deals
with studies of polymer dynamics mainly by 13C solid-state
NMR and structural fluctuations due to twist dynamics of
-conjugated polymers near the order-disorder phase transition
or twist glass transition. The twist dynamics of -conjugated
polymers will be important in the design and production
of future electronic devices such as bio-inspired stochastic
information processors.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Introduction
Biological systems employ a mechanism of sensoring and information
transmission/processing with ultra low energy consumption, which are found in
their sensory (13) and central (46) nervous systems. This situation contrasts
sharply with conventional digital computers, where energy consumption is quite
high. The low energy consumption of biological information processing is due to
its mechanism of environmental noise utilization.
Conventional digital technology has been developed based on noise
suppression. Because of this design, the system consumes high energy, say 3
or 5 volts and 10-100 watts per CPU in order to ensure error-less, deterministic
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

operation with Boolean logic. On the other hand, it is known that, for instance,
a human brain consumes at most 10 W (7) for information processing although
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

the number of elements (neurons) are at least hundreds of times larger than a
typical central processing unit (CPU). This is due to noise tolerant mechanism of
information transmission and processing.
Sensory nerves of several biological systems including crickets (1), crayfish
(2), and paddlefish (3) exploit the phenomenon of stochastic resonance (SR)
for weak signal detection within a noisy environment (4). Counterintuitively,
the signal/noise ratio (SNR) of weak sensory signal is enhanced by internal
and/or external noise to these systems. Thus. biological systems enhance their
performance with progressive utilization of noise. Similar phenomena can be
found in several layers of biological hierarchy, including membrane proteins
(8), cells (9), nervous tissues (10), brains (11), and individuals (12). Recently,
the concept of neuronal computations with stochastic network in the brain
has also been accepted (13). If one seeks to bio-mimick artificial agents with
high adaptability to unpredictable, abrupt environmental change, it would be
difficult to do with conventional digital technology, where operations based on
IfThenElse are prerequisite. In such a case, one needs to establish novel
information sensors and processors with ultra-low energy consumption that
employs biological mechanisms of progressive utilization of noise.
Recently, Asakawa et al. (14) demonstrated that a delayed feedback network
of stochastic threshold units have the ability of noise-driven autonomous switching
depending on a sensory signal. Below we listed the four important characteristics
that can be realized with such a device element.

i) nonlinear response to external field,

ii) noise generation,
iii) one-directional signal transmission, and
iv) dynamic modulation of inter-elemental connectivity.

Historically, polymeric materials had not been used for active electronic
device materials mainly because many polymers show unstable electric properties
due to poor heat resistance or heat deflection, large structural fluctuation, or
chemical instability. Of course, recent developments in organic light emitting
diodes(OLED) (15), organic field effect transistors(OFET) (16), and organic
photovoltaic cells(OPV) (17) have been remarkable, and these applications

162
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 related to polymers will form a mainstream of polymer electronics. Even so,
unstable properties, or (more correctly) large time and/or spatial fluctuations of
physical properties, e.g., carrier mobility and electric conductivity, can be an
important determinant in producing noise-driven bio-inspired devices that we are
interested in.
Poly(alkylthiophene)s [P3ATs] are a class of -conjugated polymers soluble
in ordinary organic solvents. Previous x-ray diffraction (18) and differential
scanning calorimetry (DSC) measurements (19, 20) indicate that P3ATs show an
order-disorder phase transition. By tuning the length of alkyl side-group, one can
observe a transition temperature near room temperature. It is also well known
that P3ATs without doping show nonlinear electrical conductivity. Since it is
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

theoretically predicted that static (21) and dynamic (22) disorder affect the carrier
mobility of -conjugated polymers, the dynamics associated with the transition
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

could be the origin of random electric properties and would have the potential
to be utilized for a stochastic threshold unit or a molecular noise generator. An
interacting ensemble of such stochastic threshold units can be thought of as a
complex system that generates such emergent properties as synchronization
(23, 24) and chaotic state (25) (Fig.1). We believe that such cooperative
dynamics produced by an ensemble of stochastic elements will be able to be used
for noise-driven information sensoring and processing in the future.
In this article, we shall show our recent works (2628) concerning dynamics
of P3ATs investigated by DSC, Fourier transform infrared (FTIR), and 13C solid-
state nuclear magnetic resonance (NMR) spectroscopies. The knowledge obtained
here will be useful for material selection/screening in the fabrication of stochastic
threshold units using -conjugated polymers.

Results and Discussions

Molecular Dynamics of Regioregulated P3AT
We have recently investigated the phase diagram for P3ATs (2628). There
are various thermodynamic and non-equilibrium states including crystalline
(C), glassy crystalline (GC), plastic crystalline (PC), liquid glass (G), and
isotropic lipuid (I). The complex diagram can be understood using the concept of
frustration against crystallization (29). In this article, we focus on the structure
and dynamics of poly(3-butylthiophene) [P3BT] and poly(3-hexylthiophene)
[P3HT]. P3BT shows the richest phase diagram among P3AT, and P3HT is the
most widely used polymer in P3ATs mainly because it shows the largest carrier
mobility among P3ATs (30).

DSC of Regioregulated P3BT and P3HT

Figure 2A shows the DSC chart of P3BT. We found an endothermal peak at
333 K as well as a melting peak (520 K). The peak at 333 K is probably different
from the heat capacity jump at 303K for the glass transition of the quenched sample
by liquid nitrogen (Figure 2B). (Here, glass transition means conventional liquid-
glass transition.) Thus, the peak at 333 K is attributable to some sort of solid-solid
163
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 transition. The endothermal peak at 333 K could not be observed in the second
DSC heating scan after cooling at a rate of 50 K/min., inferring that the kinetics of
the transition is quite slow. An x-ray diffraction study of P3BT is available (31),
where the authors concluded that there is a side-group transition from a mixture
of end-to-end (phase I) and interdigitation (phase II) packings to the pure phase I
state. Nevertheless, on the DSC of P3HT, we could not detect anything other than
the melting peak around 500 K (Figure 2C).
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 1. An example of Langevin dynamics simulation for an ensemble of

stochastic threshold units. Gray dots stands for the timing when the unit outputs
a firing pulse. A) Schematic drawing of a neural network consisting of 100
excitatory units (shown as E) and 100 inhibitory neurons (I). All the
excitatory units are connected to one another and the same is true for the
inhibitory units. B) oscillation emergent from an ensemble of units. External
noise input or internal noise is required for the network to emerge from the
cooperative dynamics. C) Typical bifurcation phenomenon from oscillation
(limit cycle) to chaos when inter-unit interaction is modified through sensory
input by abrupt environmental change.

164
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 2. Differential scanning calorimetry (DSC) charts for powdered samples

for regioregulated P3ATs. (A) the first heating scan for pristine P3BT, (B) the
second heating scan for a quenched sample by liquid nitrogen after the first
heating scan in DSC measurements, and (C) the first heating scan for P3HT.

Here, three questions remain. First, why did the - stacking peak in the
x-ray powder pattern smear out below 323 K, while the same peak was visible
above 323 K? Secondly, what is the driving force of transition? Thirdly, why
is the difference between P3BT and P3HT? We performed FTIR and solid-state
NMR measurements in order to address these problems.

FTIR Measurements for Regioregulated P3BT and P3HT

Figure 3 shows variable temperature FTIR measurements for P3BT and P3HT.
The spectra are for the out-of-plane deformation mode for the C4-H bond in the
main chain. For P3BT, there are two peaks, one at 825 cm-1 and the other at
810 cm-1. With increasing temperature, these signals are merged to the 820 cm-1
band (Fig.3A, right). At higher temperatures, the absorption of the 820 cm-1 band
is reduced and that of the 837 cm-1 band for melting is increased. In total, we
observed four peaks. At lower temperatures, the 825 cm-1 band is constant down
to 173 K, and the 820 cm-1 band is gradually shifted to 810 cm-1. From these
experiments, we assigned the four peaks as follows: 837 cm-1 (liquid), 825 cm-
1(unknown from IR measurements), 820 cm-1 (metastable crystalline state), and

810 cm-1 (crystalline state, energetically stabler than that of 820 cm-1).

165
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 3. Variable temperature FTIR spectra for thin films of regioregulated

P3ATs, P3BT (A) and P3HT (B).

For P3HT, the IR spectrum was similar to but simpler than that for P3BT.
Below 300 K (=Tcp, the transition temperature between the crystalline and plastic
crystalline states; vide infra for the detailed definition), two main peaks were
observed at 816 and 820 cm-1. From curve fitting, a small signal at 826 cm-1 was
also detected as a minor component. Above 300K, the 820 cm-1 peak is dominant
up to the melting temperature (500 K).

13C CPMAS NMR Measurements of Regioregulated P3BT

In order to investigate the transition around 333 K, we performed 13C cross-

polarization magic-angle sample spinning (CPMAS) experiments for powdered
samples of P3BT and P3HT.
The chemical structure of 3-butylthiophene unit and 13C CPMAS spectra of
P3BT are shown in Figure 4A and B, respectively. The clear shoulder signal of the
C4 methyl carbon of P3BT can be seen around 16.0 ppm below 333 K, whereas the
spectra at the higher temperatures show a unique component (14.6 ppm), meaning
that at least two chemically inequivalent methyl carbons in P3BT exist below 333
K (Fig.4C). Because the methylene carbons show no peak splitting, only the end
of the butyl chain shows two or more distinct states.
In Figure 4D, the signals of the thiophene ring for P3BT appear at 120145
ppm as noted in previous studies of oligothiophene (32) and regiorandom poly(3-
166
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 octylthiophene) (33). Broadening and shift of C2 and C5 peaks with heating were
also observed. We shall discuss the broadening and shift later in conjunction with
spin-lattice relaxation times. Results of FTIR and CPMAS NMR show that the
structural change of P3BT occurs markedly around 333 K. At temperatures greater
than 333 K, the state of the main chain is attributed almost uniquely to 820 cm-
1 in FTIR. The side chain is also at a unique state, as shown by the results of

CPMAS NMR. Below 333 K, the main chain consists of mainly two states: the
main component attributed to 825 cm-1 and the other to 810 cm-1. From these, the
fact that the methyl moiety of the side chain also shows at least two components
is probably related to the main-chain states.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 4. Variable temperature 67.8MHz 13C CPMAS NMR spectra for a

powdered sample of P3BT. A) chemical structure of 3-butylthiophene unit, B) full
range spectra, C) expanded spectra for alkyl carbon, and D) expanded spectra
for thiophene ring carbons.

Let us recall the three aforementioned questions. What explains the absence of
the (010) peak (- stacking) and the weakness of (100) peaks (inter-chain packing
parallel to the planar -conjugation) below 323 K in the WAXD measurement (31)?
Even if both phases (phase I and phase II) coexist, the respective scattering peaks
of x-ray diffraction should be observed. Furthermore, the peak shifts in the IR
spectrum from 825 and 810 cm-1 to 820 cm-1 with heating do not agree with the
argument by Causin et al. (31), i.e., the transition from phase I and phase II to
phase I, because no common IR absorption peak attributable to phase I is visible
below and above the transition temperature. Furthermore, what is the driving force
of the transition? What is the difference between P3BT and P3HT? Clarifying
167
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 these issues would be the key to rationalizing the data obtained from DSC, FTIR,
CPMAS NMR, and WAXD. Here, we specifically examine the dynamics of the
twisting motion of the main chain in the crystalline state.

13C Spin-Lattice Relaxation Time Measurements for Regioregulated P3BT

We performed 13C spin-lattice relaxation time measurements for P3BT using

the Torchia method (34) at various temperatures to investigate the molecular
dynamics. An Arrhenius plot of the spin-lattice relaxation rate (R1=T1-1) for each
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

carbon is shown in Figure 5.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

For the side group (Figure 5A), each R1 value was obtained using a simple
exponential fitting curve. The figure shows the subtle change of slopes around
333 K, but the decrease of R1 was observed throughout the measured temperature
range with heating (from 303 to 373 K). This tendency is visible in the extremely
narrowed regime according to the classical Bloembergen-Pound-Purcell (BPP)
theory (35), indicating that the side chains behave similarly to liquid. The subtle
change of slopes was inferred to result from the conformational change of the
main chain because it would be difficult to believe that further changes in average
conformation should occur in the side group in the liquid state.
On the other hand, the relaxation of main-chain carbons showed no single
exponential decay in the measured temperature range. For that reason, we tried
to fit the decay curves using a Kohlraush-Williams-Watts (KWW) function (36).
In general, the KWW function is used to express magnetization near and below
the glass-transition temperature (Tg ) because the distribution of the relaxation
rate gave rise to nonexponential recoveries (37).The temperature dependences
of R1 for each aromatic carbon are shown in Figure 5B. This figure shows that
the main chain must be in the slow motion regime because of the increase of
R1 with heating. More noteworthy is the fact that discontinuous slope changes
are apparent around 333 K. In many cases, similar tendency is apparent in 2H
NMR measurements, at the glass transition for many glass formers (3840). In
the case of NMR spin-lattice relaxation measurements, we can distinguish three
temperature regions for the arguments of glass transition. (i) T<Tg: the R1 of
glass formers is dominated by the slow process (Johari-Goldstein process (41)),
which follows the Arrhenius-type thermal activation process. (ii) Tg <T<1.2Tg :
Both the process and the Johari-Goldstein process affect R1, where the process
is described as stretched exponential (KWW) functions. (iii) T>1.2Tg : R1 is
dominated only by the process. Below 333 K, we observed the Arrhenius-type
thermally activated process for all carbons. That tendency is similar to the
Johari-Goldstein process. Above 333 K, although the measured temperature range
is insufficient to consider the condition of (ii) or (iii) because of the limitation of
the available temperature range of our instrument, discontinuous slope changes
can be a fingerprint for the effect of process.

168
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 5. Temperature dependence of 13C spin-lattice relaxation time at 67.8

MHz for alkyl side group carbons (A) and for thiophene ring carbons (B). Insets
are typical decay curves measured using Torchias method.

Figure 5B illustrates another point. For the CPMAS 13C measurements,

relaxation occurs through two paths: one is due to the local field fluctuation
by 13C-1H magnetic dipolar coupling and the other is due to fluctuation of 13C
chemical-shift anisotropy (CSA). C4 carbon that is directly connected to a proton
has larger dipolar coupling effect than those of unconnected carbons. Below
333 K, only the C4 carbon shows larger R1 than other carbons, indicating that
the relaxation is caused mainly by 13C-1H magnetic dipolar coupling. That is,
the thiophene ring does not undergo reorientations sufficient to fluctuate the
chemical-shift anisotropy. Above 333 K, the other carbons (C2,C3,C5) showed
individually specific slopes, indicating that the reorientation of the thiophene
ring brings about fluctuations in CSA as well. In view of both IR and NMR
detectability, the most plausible reorientation is twist motion. Similar low-energy
excitations of twist motion were observed in structural phase transitions in
oligophenylenes using Raman (42) and NMR (43, 44).
The above arguments indicate that the glass transition with respect to the
thiophene twisting in the crystalline state occurs around 333 K. We define the
transition as a twist glass transition. Around the twist glass transition, it is thought
169
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 that the translational symmetry with respect to the molecular axis remains almost
unchanged. Although the side group is in the liquid state, there exist residual
thiophene -stacking and then P3BT assumes a kind of crystalline state. But since
the -stacking would not be strong, the planarity of -conjugation would be far
from the perfect. If we think of the thiophene twisting as rotational degree of
freedom, the crystalline phase with thiophene twisting is twist plastic crystalline
phase (PC), and the corresponding non-equilbrium state with frozen twist below
the transition temperature is twist glassy crystalline (GC). Therefore, twist glass
transition is called PC-GC transition and the transition temperature is expressed
as Tgp. Accordingly, the four peaks observed in FTIR are assigned a follows; 825
cm-1 (GC), 820 cm-1 (PC), 837 cm-1 (I), and 810 cm-1 (C), respectively.
Back to the 13C spin-lattice relaxation of P3BT, the temperature dependence
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

of the stretching parameter does not accord with the results of other glass
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

formers. In general, for T>Tg, is nearly unity because the motional correlation
time is much shorter than the spin-lattice relaxation time, which means that
ergodicity is achieved. Around Tg, the value of begins to decrease with
decreasing temperature because of the T1 distribution caused by the slowing
of the process. Then, the value of increases again slightly with further
cooling (40, 45, 46). In these cases, the single form of the spectral density
function J() is used to determine T1 because it allows the specific examination
of intramolecular motions if we ignore intermolecular interactions. In our case,
because of the crystalline state of strong packing, we cannot ignore the fluctuation
of intermolecular interaction, especially above the transition temperature where
twist motion exists. For that reason, two kinds of spectral-density functions
are needed Jintra() and Jinter(). Because two interactions, magnetic dipolar
coupling and chemical shift anisotropy, should affect each J() in a specific
ratio depending on the second moment of interactions, the decay curve cannot
be expressed by a single stretched-exponential function, indicating that cannot
express the T1 distribution directly at higher temperatures. At T < 333 K, where
a twist motion is absent, the magnetization curve can be expressed using a single
stretched exponential function. As mentioned above, however, at least two states
exist, which are attributed to 825 and 810 cm-1 in the FTIR spectrum in the
temperature region. Therefore, the value of does lose the conventional meaning
of the distribution parameter in the case of P3BT.

13C CPMAS NMR of Regioregulated P3HT

Combining the results of FTIR for P3BT and P3HT and solid-state NMR of
P3BT, we assigned the absorption bands in FTIR spectra for P3HT thus: 816 cm-1
(C), 820 cm-1 (PC), 826 cm-1 (GC), and 832 cm-1 (I), respectively. The gradual
transition was observed in FTIR (Fig.3B, left) and C and GC phases disappears
around 300K, which is expressed as Tcp, the C-PC transition temperature. Since
the fraction of GC phase is quite low, it would be appropriate to express the C-PC
transition rather than GC-PC transition. In this section, in order to investigate
the structure and molecular dynamics of P3HT around Tcp observed in the FTIR
study (27), we first performed 13C CPMAS measurements for P3HT at various
temperatures. The chemical structure of the 3-hexylthiophene unit and the spectra
170
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 are shown in Figure 6A and 6B, respectively. The peaks higher than 40 ppm
correspond to the aliphatic carbon atoms of the side group (Figure 6C) and four
aromatic peaks at 120-140 ppm correspond to the four carbons of the thiophene
unit(Figure 6D). The assignments of each carbon were made by comparing with
the solution NMR of P3HT (47). Notably the signals narrow with increasing
temperature for both the main chain and the side group. The narrowing probably
indicates the onset or acceleration of molecular motion of the P3HT main chain and
side group. Figure 6E and 6F show the change in the full width at half-maximum
(FWHM) of each peak for the side group and the main chain, respectively.
For the side group, the FWHM decreased gradually with heating, and the slope
turned particularly steeper between 200 and 250 K. After that, they showed a weak
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

temperature dependence (note the plateau) between 270 and 300 K, and finally
decreased again above 300 K. Interestingly, this transition temperature region for
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

the steeper slope between 200 and 250 K corresponds to that of the FTIR study
(27), where the C4-H out-of-plane vibration band gradually shifted with heating
from two peaks, 826 and 816 cm-1, to a single peak, 820 cm-1, which corresponds to
GC-PC (although the fraction is quite small) and C-PC transition, respectively. In
the FTIR study, we focused only on the behavior of the thiophene ring moiety, that
is, the change in absorption of the C4-H out-of-plane region around the transition
temperature (27). The CPMAS spectra, however, imply that the conformation or
the dynamics of the side group changes around the transition temperature. There
was no noticeable peak shift in the IR spectra from trans to gauche in the CH2
asymmetric band in the measured temperature range (28). Thus, the FTIR data
shows that the alkyl side group keeps the gauche conformation even below Tcp,
although the line width of the CPMAS spectra changed in this temperature region.
The 13C chemical shift of CPMAS spectra for the alkyl group is consistent with
the IR results. The methylene peaks (C1-3) at 30.5 ppm maintain its height even
below Tcp. From the chemical shift of the methylene peak, the side group mainly
takes on the gauche conformation even below Tcp. In fact, CPMAS spectra below
Tcp showed small shoulders around 25 and 34 ppm in the C5 and C1-3 signals,
respectively, indicating at least two or more chemically inequivalent methylene
carbons in the side group below Tcp. Our FTIR study showed that several phases,
PC, GC, C, and amorphous, coexist in this temperature region. These various
phases may affect the chemical shift of the side group. Therefore, CPMAS spectra
below Tcp should have a complicated shape. Unfortunately, the signal of C4
carbon which appears around 33 ppm overlaps with the all-trans signal of other
methylene carbons. This makes it difficult to assign the conformation at the lower
temperature accurately. However, the fact that the signal at 30 ppm is clearly
apparent and still being a main component even down to 173 K means that at least
a certain amount of the gauche conformation exists below Tcp. It has often been
claimed that the side groups of P3ATs take a perfect all trans conformation at lower
temperatures. However, it is not true in the case of P3HT.
Consequently, one can say that the increase of FWHM below Tcp originates
from the slowing of the side group dynamics. At the transition temperature,
the side group froze without a conformational change similar to a glass
transition. As mentioned above, P3ATs with longer alkyl side group, such as

171
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 poly(3-dodecylthiophene) [P3DDT], show a melting behavior of the side group,
and it leads to polymorphism (48, 49).
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 6. Variable temperature 125.8 MHz 13C CPMAS NMR spectra for a
powdered sample of P3HT. A) chemical structure of 3-hexylthiophene unit, B)
full range spectra, C) expanded spectra for alkyl carbon, D) expanded spectra
for thiophene ring carbons, E) full width at half maximum (FWHM) for alkyl
carbons, and F) FWHM for thiophene carbons.

Nonetheless, the side group transition in P3HT is not an order-disorder

transition but a glass transition because of its short side group. Pankaj et al.
(50) investigated side chain dynamics in a series of regiorandom P3ATs by
dynamic mechanical analysis, and a process of the alkyl side group for P3HT
was observed at 186 K. The fact that the temperature at which the narrowing
of the CPMAS spectrum started is almost consistent with their results confirms
the glass transition of the alkyl side group in the crystalline state. Similar to the
behavior of the side group, the line width for the main chain also narrowed with
heating (see Figure 6D and 6F). Below 300 K, similar temperature dependence of
FWHM to the side group was observed, although slope changes around 190-210
K are not so clear as observed in the side group. The narrowing with heating
below 300K should be attributed to an inhomogeneity-homogeneity change in the
172
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 conformation or to acceleration of the thiophene ring twisting motion. Indeed,
the IR band of the C4-H out-of-plane vibration below 300 K (=Tcp) was able to be
decomposed to four components, PC, C, GC, and amorphous, while it could be
decomposed to two states, PC and amorphous, above Tcp (27). Therefore, both the
decrease of conformational distribution and the increase of molecular dynamics
must have contributed to the narrowing of CPMAS signals.

13C Spin-Lattice Relaxation Time Measurements of Regioregulated P3HT

Next, we performed 13C spin-lattice relaxation-time measurements for P3HT
(28). The relaxation rates (R1 ) for each carbon at various temperatures are plotted
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

in Figure 7. For the side groups (Figure 7A), each R1 value was obtained using
a KWW function. Whereas below 230 K the relaxation rates were distributed,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

the value of showed nearly unity above 250 K (see the relaxation curves in
the inset of Figure 7A); i.e., the obtained fitting curve was a simple exponential.
It suggests that an inhomogeneity-homogeneity transition was occurred with
heating. The temperature range where the value is turned into unity corresponds
to the peak narrowing of the CPMAS spectrum, suggesting side group transition.
However, a clear transition of R1 was not observed probably due to the noise in
R1 measurement, which is very susceptible to plots with low amplitude in the
relaxation curve. With further heating, R1 reached the maximum value around
270-300 K, except for the methyl carbon of the side group, which monotonously
decreased with heating, meaning that the correlation time, c, for the inner side
group reached the inverse of Larmor frequency, 125 MHz. At a temperature
greater than 300 K, the molecular motion of these carbons entered the extremely
narrowing regime (35), and eventually the value of R1 decreased with heating.
Interestingly, the onset temperature of extremely narrowing regime is consistent
with Tcp, indicating that rapid alkyl group motion occurs in the PC state. As
mentioned above, the decay of magnetization of the side group carbon obeyed
Debye type relaxation in this temperature range. In this case, we were able to
determine the activation energy, Ea, of 8.2-15.0 kJ/mol. This is consistent with
the recent result of quasi-elastic neutron scattering by Obrzut et al. (51).
For the main chain carbons, the decay of magnetization obeyed the stretched
exponential fitting curve in all measured temperature ranges as shown by
the representative curves in the inset of Figure 7B. This distribution of R1 is
reasonable for the similar reasons as P3BT. Obtained value of R1 are shown in
Figure 7B. Two step jumps with heating are visible around 240 and 310 K. The
abrupt slope change in the lower temperature region seems to coincide with the
FWHM change observed in the alkyl side group. We can describe a plausible
scenario of the transition; the alkyl chain begins to undergo molecular motion
around 200 K, which is the onset of slope change in the FWHM upon heating,
and then the thiophene begins to twist around 240 K in response to the side group
motion accelerated enough to stimulate the main chain motion. Around 270-310
K, R1 show almost constant values. This plateau of R1 is probably related to the
critical behavior of C-PC transition. At a temperature greater than 310 K, R1
increases again with heating. The steep rise of R1 was observed in the case of the
twist glass transition for P3BT as well (26). We believe that all thiophene rings
173
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 undergo twisting, which means that they are in the PC state. However, various
factors derived from the same reasons as P3BT prevent an estimation of exact
activation energy for the thiophene twist. As mentioned above, R1 remains almost
constant around 270-310 K. Similar phenomenon has been reported for the phase
transition of the second kind between a high-temperature normal phase and an
incommensurately modulated phase, Ti, particularly named partial slow motion
limit in some insulating crystals (52). Just above the critical temperature, R1
is temperature independent due to the critical slowing down of order parameter
dynamics (52). The fact that the temperature independent region is observed just
above Tcp for P3HT might indicate that the transition might be a second-order
phase transition as well. The absence of remarkable latent heat in DSC (26)
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

also supports the second-order phase transition. However, the phenomenon that
violates the general property of second-order transition has also been observed;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

i.e., PC can exist even below Tgp (=Tcp) as the supercooling liquid (27). Further
investigations such as the frequency dependence of T1 NMR measurements are
necessary to determine the classification of the phase transition occurring in
P3HT.

Comparison between P3BT and P3HT

This section is devoted to the aforementioned third question. The temperature
dependent R1 also provides information about the difference of dominant structures
between P3BT and P3HT around Tgp or Tcp. For P3BT, only the C4 carbon,
which is directly connected to a proton, showed larger R1 than other carbons below
the twist glass transition temperature (T < Tgp=340K) (26), indicating that the
relaxation was mainly due to fluctuation of 13C-1H magnetic dipolar coupling. The
thiophene ring would not undergo reorientations sufficient to fluctuate the chemical
shift anisotropy (CSA). Above the Tgp, on the other hand, the other carbons (C2,
C3, C5) showed specific slopes individually, where reorientation of the thiophene
ring, i.e., twisting, brought about fluctuations in CSA as well. For P3HT, however,
all carbons showed similar R1 in the measured temperature range, suggesting that
the relaxation occurred not only by the fluctuation of magnetic dipolar coupling but
also by that of chemical shift anisotropy even below Tcp. This difference between
P3BT and P3HT seems to be consistent with our argument based on the IR results;
GC and C, in which thiophene does not undergo twisting motion, are dominant
in P3BT, whereas PC is dominant in P3HT below Tgp or Tcp. This tendency can
be explained by the degree of frustration against crystallization due to interchain
interaction (27). In the case of P3BT, locally favored structures (cooperative
rearrangement region-like structure) exist with strong frustration. Consequently,
GC and C are preferentially realized with respect to a decrease of temperature.
For P3HT, in contrast, PC is the dominant structure instead of GC or C because
of its weak frustration. The relationship between the molecular mobilities of the
side group and main chain supports this hypothesis. The C-PC transition occurred
in P3HT when the hexyl group entered the extreme narrowing regime, which is
confirmed by a negative slope of R1 vs T -1. In other words, - stacking is easily
weakened by the rapid side groups motion. On the other hand, P3BT kept GC
and C states even though the butyl group was in the extremely narrowing regime.
174
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The stronger intermolecular interaction in P3BT yields a more rigid crystal as
well as a large amount of GC. The temperature dependence of optical absorption
measurements for P3HT support these argument (28).
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 7. Temperature dependence of 13C spin-lattice relaxation time at 125.8

MHz for alkyl side group carbons (A) and for thiophene ring carbons (B). Insets
are typical decay curves measured using Torchias method.

Conclusion
In this article, we summarized our recent results concerning the phase
diagram for P3ATs. There are various thermodynamic and non-equilibrium
states including crystalline, glassy crystalline, plastic crystalline, liquid glass,
and isotropic liquid. Figure 8 illustrates the phase diagrams for P3BT and P3HT
including the thermodynamical phases and non-equilibrium states. This complex
situation is due to frustration against crystallization in P3ATs. The strength of the
frustration depends on the length of alkyl side-groups. For example, RR-P3BT is
GC rich, while RR-P3HT is PC rich for a wide temperature range.
175
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

Figure 8. Proposed phase diagram for P3BT and P3HT. The horizontal axis
stands for the fraction of thermodynamical phasesand non-equilibrium states
estimated by the FTIR measurements and the fractions are presumably connected
to the degree of frustration against crystallization of P3AT.

Acknowledgments
One of the authors (N.A.) is supported by the Ministry of Education, Culture,
Sports, Science and Technology (Japan) through a Grant-in-Aid for science
research (21350125).

References
1. Levin, J. E.; Miller, J. P. Nature (London) 1996, 380, 165168.
2. Douglas, J. K.; Wilkens, L.; Pantazelou, E.; Moss, F. Nature (London) 1993,
365, 337340.
3. Russell, D. F.; Wilkens, L. A.; Moss, F. Nature (London) 1999, 402, 291294.
4. Gluckman, B. J.; Netoff, T. I.; Neel, E. J.; Ditto, W. L.; Spano, M. L.;
Schiff, S. J. Phys. Rev. Lett. 1996, 77, 40984101.
5. T. Mori, T.; S. Kai, S. Phys. Rev. Lett. 2002, 88, 218101-1218101-4.
6. K. Kitajo, K.; Nozaki, D.; Ward, L. M.; Yamamoto, Y. Phys. Rev. Lett. 2003,
90, 218103-1218103-4.
7. Merkle, R. C. Foresight Update 1989, August, 6.
8. Berzukov, S. M.; Vodyanoy, I. Nature 1995, 378, 362364.
9. Ueda, M.; Ueda, M.; Takagi, H.; Sato, M. J.; Yanagida, T.; Yamashita, I.;
Setsune, K. Physica A 2008, 387, 44754481.
10. Gluckman, B. J.; Netoff, T. I.; Neel, E. J.; Ditto, W. L.; Spano, M. L.;
Schiff, S. J. Phys. Rev. Lett. 1996, 77, 40984101.
11. Hidaka, I.; Nozaki, D.; Yamamoto, Y. Phys. Rev. Lett. 2000, 85, 37403743.
176
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 12. Priplata, A.; Niemi, J.; Salen, M.; Harry, J.; Lipsitz, L. A.; Collins, J. J. Phys.
Rev. Lett. 2002, 89, 238101-1238101-4.
13. Destexhe, A.; Contreras, D. Science 2006, 314, 8590.
14. Asakawa, N; Hotta, Y.; Kanki, T.; Kawai, T.; Tabata, H. Phys. Rev. E 2009,
79, 021902-1021902-8.
15. Wessling, B. In Handbook of Conducting Polymers, 3rd ed.: Processing and
Applications; Skotheim, T. A.,Reynolds, J. R., Eds.; CRC Press: Boca Raton,
FL, 2007; pp 1-3 - 1-75.
16. Tessler, N.; Veres, J.; Globerman, O.; Rappport, N.; Preezant, Y.; Roichman,
Y.; Solomesch, O.; Tal, S.; Gershman, E.; Adler, M.; Zolotarev, V.; Gorelik,
V.; Eichen, Y. In Handbook of Conducting Polymers, 3rd ed.: Processing
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

and Applications; Skotheim, T. A., Reynolds, J. R., Eds.; CRC Press: Boca
Raton, FL, 2007; pp 7-1 7-42.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

17. Mozer, A. J.; Sariciftci, N. S. In Handbook of Conducting Polymers, 3rd ed.:

Processing and Applications; Skotheim, T. A., Reynolds, J. R., Eds.; CRC
Press: Boca Raton, FL, 2007; pp 10-1 10-37.
18. uny, W.; Pro, A. Synth. Metals 1996, 79, 3741.
19. Bolognesi, A.; Porzio, W.; Provasoli, F. Makromol. Chem. 1993, 194,
817827.
20. Bolognesi, A.; Botta, C.; Mercogliano, C.; Marinelli, M.; Porzio, W.;
Angiolini, L.; Salatelli, E. Polym. Adv. Tech. 2003, 14, 537543.
21. Yu, Z. G.; Smith, D. L.; Saxena, A.; Martin, R. L.; Bishop, A. R. Phys. Rev.
Lett. 2000, 84, 721724.
22. Troisi, A.; Orlandi, G. Phys. Rev. Lett. 2006, 96, 086601-1086601-4.
23. Gerstner, W.; van Hemmen, J. L. Phys. Rev. Lett. 1993, 71, 312315.
24. Hotta, Y.; Kanki, T.; Asakawa, N.; Tabata, H.; Kawai, T. Appl. Phys. Exp.
2008, 1, 088002-1088002-3.
25. Kanamaru, T.; Sekine, M. Phys. Rev. E 2003, 67, 031916-1031916-8.
26. Yazawa, K.; Inoue, Y.; Yamamoto, T.; Asakawa, N. Phys. Rev. B 2006, 74,
094204-0094204-12.
27. Yazawa, K.; Inoue, Y.; Yamamoto, T.; Asakawa, N. J. Phys. Chem. B 2008,
112, 1158011585.
28. Yazawa, K.; Inoue, Y.; Shimizu, T.; Tansho, M.; Asakawa, N. J. Phys. Chem.
B 2010, 114, 12411248.
29. Shintani, H.; Tanaka, H. Nat. Phys. 2006, 2, 200206.
30. Babel, A.; Jenekhe, S. A. Synth. Metals 2005, 148, 169173.
31. Causin, V.; Marega, C.; Marigo, A.; Valentini, L.; Kenny, J. M.
Macromolecules 2005, 38, 409415.
32. Barbarella, G; Casarini, D.; Zambianchi, M.; Favaretto, L.; Rossini, S. Adv.
Mater. 1996, 8, 6973.
33. Bolognesi, A.; Porzio, W.; Provasoli, A.; Botta, C.; Comotti, A.; Sozzani, P.;
Simonutti, R. Macromol. Chem. Phys. 2001, 202, 25862591.
34. Torchia, D. A. J. Magn. Reson. 1978, 30, 613616.
35. Bloembergen, N.; Purcell, E. M.; Pound, R. V. Phys. Rev. 1948, 73, 679712.
36. Ngai, K. L. Solid State Ionics 1981, 5, 2733.
37. Schnauss, W.; Fujara, F.; Hartmann, K.; Sillescu, H. Chem. Phys. Lett. 1990,
166, 381384.

177
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 38. Bhmer, R.; Diezemann, G.; Hinze, G.; Rssler, E. Prog. Nucl. Magn.
Reson. Spectrosc. 2001, 39, 191267.
39. Rssler, E.; Warschewske, U.; Eiermann, P.; Sokolov, A. P.; Quitmann, D. J.
Non-Cryst. Solids 1994, 172-174, 113125.
40. D, A.; Hinze, G.; Diezemann, G.; Bhmer, R. Acta Polym. 1998, 49,
5658.
41. Johari, G. P.; Goldstein, M. J. Chem. Phys. 1970, 53, 23722388.
42. Guha, S.; Graupner, W.; Resel, R.; Chandrasekhar, M.; Chandrasekhar, H.
R.; Glaser, R.; Leising, G. Phys. Rev. Lett. 1999, 82, 36253628.
43. Kohda, K.; Nakamura, N.; Chihira, H. J. Phys. Soc. Jpn. 1982, 51,
39363941.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

44. Liu, S.-B.; Conradi, M. S. Phys. Rev. Lett. 1985, 54, 12871290.
45. Hinze, G.; Sillescu, H. J. Chem. Phys. 1996, 104, 314319.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch010

46. Bhmer, R.; Hinze, G.; Jrg, T.; Qi, F.; Sillescu, H. J. Phys.: Condens.
Matter 2000, 12, A383A390.
47. Ferrari, M.; Mucci, A.; Schenetti, L.; Malmusi, L. Magn. Reson. Chem.
1995, 33, 657663.
48. Prosa, T. J.; Winokur, M. J.; McCullough, R. D. Macromolecules 1996, 29,
36543657.
49. Meille, S. V.; Winokur, M. J.; McCullough, R. D. Macromolecules 1997, 30,
78987905.
50. Pankaj, S.; Hempel, M.; Beiner, M. Macromolecules 2009, 42, 716724.
51. Obzrut, J.; Page, K. A. Phys. Rev. B 2009, 80, 195211-1195211-7.
52. Decker, F.; Petersson, J.; Irsch, M.; Michel, D. Phys. Rev. B 2001, 65,
014110-1014110-11.

178
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 11

Molecular Mobility and Phase Composition

in Polyolefins: From Fundamental to Applied
Research
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

V. M. Litvinov*

DSM Resolve, P.O. Box 18, 6160 MD Geleen, The Netherlands

*E-mail: [email protected]

Molecular mobility, phase composition and morphology

of polyethylenes (PE), polypropylenes (PP) and random
poly(ethylene-co-propylene) (PPR) were studied by solid-state
low-field 1H NMR, X-ray, DSC and microscopy methods.
Influence of the following factors on the phase composition
and molecular mobility is discussed: (1) the effect of the
amount and the type of comonomers in PE and PP chains,
(2) thermal aging, and (3) deformation. Small changes in the
chemical composition, thermal history and mechanical load can
significantly influence molecular mobility in the amorphous
phase. A series of studies by the author shows that solid-state
NMR provides a unique and complementary tool to traditional
methods in obtaining information about physical structures and
local dynamics in polyolefins. This information is useful to
achieve a better understanding of deformation behavior of this
class of polymers.

Introduction
The physical and mechanical properties of semi-crystalline polymers are
significantly influenced by morphology, phase composition and molecular
mobility in different phases. Therefore, a quantitative characterization of
these meso-/nano-scopic characteristics of semi-crystalline polymers is of great
importance to advance our understanding of their properties. In this respect,
the phase composition is probably one of the most important morphological
parameters, mainly because the amorphous and crystalline phases exhibit vastly

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 different behavior and their relative contributions to the material properties
should be accurately known. Numerous 1H, 2H and 13C solid-state NMR
studies of polyolefins have provided detailed information on phase structure
and morphology, especially with respect to crystalline polymorphs, phase
composition, chain conformation, and molecular motions. In this paper, a series
of studies of polyolefins using low-field 1H NMR is reviewed. Several examples
of methods and applications are provided, which help improve our understanding
of the physical structures of polyolefins and the role of the physical structures in
the different applications.

Results and Discussion

Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

Methods for Crystallinity Determination

The most common methods for crystallinity determination are X-ray
diffraction, density measurement, and differential scanning calorimetry. Each
of these methods is based on a different physical property and gives rise to a
different definition of the crystalline phase. In general, crystallinity determination
using different methods does not always yield the same result on even the same
sample (1) because of the following reasons: (i) The complex morphology of
semi-crystalline polymers requires different sets of assumptions for the analysis of
the data recorded by different techniques. (ii) The discrimination of the crystalline
and amorphous phases is made on the basis of different characteristics, such as
the enthalpy of melting (DSC), long range periodicity (WAXD), and the specific
volume (density analysis). (iii) The two-phase model, which is traditionally
used for determining crystallinity, is rather simplified for the description of
semi-crystalline polymers due to the presence of a crystal-amorphous interface or
rigid amorphous fraction, which can be detected either as crystalline or amorphous
fraction depending on the method used. In other words, the borderline is difficult
to draw between crystalline and amorphous phases for different methods.
Several experimental methods, such as neutron scattering, dielectric
relaxation, calorimetry, and solid-state NMR, have shown that an intermediate
layer separates crystalline and amorphous phases, and the properties of this layer,
i.e., crystal-amorphous interface (1, 4, 5) or rigid amorphous fraction (6, 7), are
intermediate between those for crystalline and amorphous phases. A three-phase
model, in which the interface is taken into account, could provide better
description of the experimental data. Therefore, the term phase composition is
perhaps more appropriate than simply crystallinity to emphasize the multi-phase
nature of semi-crystalline polymers. The interfacial layer is usually thought to be
a transitional region located in between the crystals and the mobile amorphous
regions. The interface has distinct chain dynamic and is characterized by a degree
of order perpendicular to the lamellae surface but disorder in the lateral direction
(8).
Solid-state NMR is one of the most informative techniques for the
characterization of molecular mobility and molecular scale heterogeneity in
materials. Over the years, different solid-state NMR methods with relatively
basic or sophisticated features were used for the investigation of morphology
180
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and molecular mobility in polyolefins. 13C NMR spectroscopic methods provide
detailed information about molecular mobility in the different phases of PE
because of high phase selectivity, which exploits differences in chemical shifts
for crystalline and amorphous phases (912). However, quantitative studies of
phase composition, domain sizes, and molecular mobility in PE by 13C NMR
could suffer from a lack of accuracy due to very long 13C T1 value for PE crystals.
For the characterization of a large series of samples, one should choose the most
robust, convenient and accurate NMR method. 1H NMR T2 relaxometry was
widely used for this purpose. The high sensitivity of 1H NMR makes this method
very attractive for the characterization of crystallization kinetics (13), premelting
behavior (14), and quality control (15). Wide-line 1H NMR and relaxation
experiments were frequently applied to study the effect of the chemical structure,
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

molar mass and temperature on the phase composition and molecular mobility
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

(1620).

Phase Composition Analysis by NMR T2 Relaxometry

When the term crystallinity is used in connection with low-resolution NMR
technique, this refers to the rigid fraction at temperatures well above Tg, i.e.,
the fraction of material possessing the lowest molecular mobility; this term is
also distinguished from semi-rigid (interfacial) and soft non-crystalline phases
(either oriented or non oriented) in a material. Characterization of the three-phase
structure is important from a theoretical viewpoint, and it is a key factor in
determining the overall crystalline structure, the morphology and (thus) the
mechanical properties.

Optimum Analysis Temperature for Phase Composition by NMR T2 Relaxometry

At room temperature, a significant fraction of the amorphous phase in i-PP

and HDPE is rigid. With increasing temperature, molecular mobility begins to
increase in the inner/softer part of the amorphous phase towards the lamellae
surface resulting in decrease in the amount of the rigid fraction, which is
composed of the crystalline phase and the rigid fraction of the amorphous phase
(Figure 1). To observe distinct differences in molecular mobility of the crystalline
and amorphous phases, and consequently in the T2 relaxation, the temperature of
the sample should substantially exceed the dynamic glass transition temperature
(Tgd) at the time scale of the NMR experiment, i.e., in the range of 10-4 - 10-5 sec.
Since the T2 relaxation experiments should be performed at temperatures well
above Tg, the sample exposure to elevated temperatures can cause irreversible
changes in the phase composition and in molecular mobility. Therefore, the
temperature for the experiments should not be too high to prevent annealing of
the sample during the NMR experiment.
The amount of the rigid fraction in HDPE (21) and i-PP (22) is almost constant
in the temperature range from ~70C to 100 C, and its value is close to the
crystallinity measured by DSC and X-ray. In this temperature range, the difference
in molecular mobility in rigid, semi-rigid and soft fractions is high enough for (i)
181
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 accurate deconvolution of the decay of the transverse magnetization relaxation
(T2 decay) into the relaxation components originating from different phases, and
(ii) annealing can be avoided during NMR experiments at these temperatures.
Therefore, this temperature range is optimal for phase composition analysis in
HDPE and i-PP.

Analysis of the Decay of the Transverse Magnetization Relaxation

Precautions should be taken for accurate measurements of the T2 relaxation

function of materials which are composed of rigid and soft phases, like semi-
crystalline polymers above Tg of the amorphous phase (2326). The T2 decay the
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

time dependence of the Mxy component of the macroscopic magnetization vector

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

- is a weighted sum of T2 decays from phases with distinct difference in molecular

mobility. The relative fraction of the T2 relaxation components is proportional to
the content of hydrogen in these phases. The T2 relaxation time of these relaxation
components allows the determination of relative difference in molecular mobility
in different phases and in a series of samples of the same chemical origin. The
longer the T2 relaxation time, the larger the frequency (and/or the amplitude) of
molecular motions is.

Figure 1. Schematic drawing of the effect of temperature on the amount of rigid,

semi-rigid ands soft fractions of melt-crystallized PE, and molecular mobility in
these fractions as determined by the analysis of T2 decay.

182
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Analysis of the phase composition in complex polymeric materials
using T2 relaxation data is usually complicated due to several reasons: (a)
multiphase/component composition; (b) morphological heterogeneities of
materials, such as distribution of domain sizes resulting in a distribution of
frequency of molecular motions; (c) spatial heterogeneity of materials on
the micrometer scale, which is formed during processing, i.e., heterogeneous
distribution of components, differences in morphology through the sample
volume because of variation in temperature gradient and flow induced orientation
in different parts of a sample, for example in skin layer versus core part; (d)
molecular-scale and morphological heterogeneity, which is caused by the
chemical heterogeneity, such as molar mass distribution and a variation in
the chain composition, chain sequences and tacticity along a single chain; (e)
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

complexity of molecular motions causing a complex shape of the decay of the

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

transverse magnetization relaxation. Because of these reasons, a thorough study

of the T2 relaxation as a function of temperature both for a polymeric material
and for the separate components that were used for its preparation, is desired for
reliable data analysis.
The theoretical description of the T2 decay shape for semicrystalline polymers
above the Tgd is still under debate and a purely phenomenological analysis of
the decay shape with two- or more components is commonly used. Different
functions were used for describing the T2 decay for semicrystalline polyolefins
(20, 2631). The shape of the relaxation component for the crystalline phase
was described either by an Abragamian or a Gaussian function. A Gaussian,
Weibull or an exponential function was used to describe the T2 relaxation of the
crystal-amorphous interface. The relaxation of the soft amorphous fraction was
described either with a single exponential or a sum of two exponential functions. It
should be noted that the phase composition, as characterized by the NMR method,
is affected to some extent by the temperature of the experiment (25), and by a fitting
function used for the deconvolution of the T2 decay into the separate components
(13, 20, 30, 32). Nevertheless, the amount of rigid phase, which is determined
from the T2 relaxation analysis, corresponds rather well with crystallinity values
obtained by other methods.

Temperature Dependence of Phase Composition and Molecular Mobility in

Polyolefins

1. HDPE, Ethylene-Octene Copolymer (m-PE) and i-PP

The three-phase model provides the most appropriate description of the phase
composition in HDPE (21) and i-PP (22) at temperatures of ~70 - 100 C. Above
these temperatures, the phase composition is affected by annealing and partial
melting. Crystallinity values, as determined by the NMR method, are in good
agreement with those determined for the same samples using SAXS. Crystallinity
of m-PE is significantly lower than that of HDPE (33). Above room temperature,

183
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 crystallinity of m-PE with 30 wt% octane chain units gradually decreases with
increasing temperature and approaches zero at 70C.

2. Random Poly(Ethylene-co- propylene) (PPR)

The phase structure of PPR is more complex than that of i-PP due to the
presence of small amount of ethylene chain units in the main chain, as well
as atactic propylene chain units. Four different phases/fractions with different
molecular mobility are determined at 110C by the T2 relaxation method (26): (i)
crystalline phase, (ii) semi-rigid crystal-amorphous interface, (iii) soft fraction of
the amorphous phase, and (iv) small amount of the amorphous fraction, which
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

has molecular mobility similar to that of rubbers.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

3. Thermoplastic Vulcanizates (TPV) Composed of i-PP, EPDM, and Paraffinic

Oil

TPVs are rubbery materials, which are processable as thermoplastics but

exhibits properties similar to those of vulcanized rubbers at usage temperatures.
TPVs are produced by dynamic vulcanization of blends containing a thermoplast
and an elastomer. At 110 C, the NMR relaxation method allows us to perform
a rather accurate determination of the amount of different phases/components of
TPV namely, crystalline and amorphous phases of PP, EPDM and paraffinic
oil and molecular mobility in these phases (34). Crystallinity of PP is hardly
affected by the TPVs composition and is close to 70 wt %. A small fraction of
oil molecules plasticizes the amorphous phase of PP and the plasticization effect
is proportional to oil : PP mass ratio in the TPV composition. Results suggest
that a thin interfacial layer is formed by propylene-rich chain fragments of EPDM
of EPDM/oil rubbery phase and by PP chains of the PP phase. The interface is
the source of physical junctions, which are formed at the shell of EPDM rubber
particles and at the surface of PP particles occluded in the EPDM phase.
NMR allows quantitative determination of the network density in the rubbery
phase of the TPVs (34). The network structure is formed by chemical crosslinks,
physical junctions at the EPDM/PP interface, temporary and trapped chain
entanglements. The total network density in EPDM rubbery phase is affected by
the following changes in the TPV composition: (i) the network density increases
with increasing amount of the crosslinker per weight unit of EPDM; (ii) decreases
with increasing oil content because of chain disentanglements; and (iii) increases
with increasing PP/EPDM weight ratio due to an increase in the density of
physical junctions at the EPDM/PP interface. The EPDM network density largely
influences the translational and rotational mobility of oil molecules; i.e. the higher
the crosslink density, the more the mobility of oil molecules is hindered. Some
relationships between the TPV composition, network density and mechanical
properties are established (34).

184
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Domain Size Determination by NMR Spin-Diffusion Experiments

Traditionally, transmission electron micrograph (TEM) and small-angle

X-ray scattering (SAXS) are two major techniques used to determine the lamellar
thickness. The former offers the advantage of direct access to the morphology.
SAXS is a well-developed method to quantitatively determine the thicknesses
of alternating layers of the crystalline and amorphous regions of the lamellae
morphology. The NMR experiments permit unambiguous identification of the
three-phase morphology due to a large difference in the relaxation behavior of the
chain units in mobile amorphous fraction, semi-rigid crystal-amorphous interface
and crystalline phase (21, 22). The domain thickness in HDPE and i-PP, which
is determined by NMR in the temperature range from ~70 to 100C, is in good
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

agreement to those measured by SAXS and TEM on the same sample. The
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

thickness of the interface is nearly constant in the studied temperature range and
equals 1.1 1.3 nm. This value is in the same range as previously estimated for
PE value by Monte Carlo simulations (35). It should be noted that the interface
thickness is close to the length of the statistical segment of PE and i-PP chains,
which consists of 6 7 carboncarbon bonds and equals to ~0.8 nm for fully
extended polyolefinic chain. Therefore, distinct differences in molecular mobility
in different fractions of HDPE are apparently caused by short-range correlations
of chain motion due to the short length of the statistical segment.

The Effect of Annealing on Phase Composition, Molecular Mobility, and

Domain Sizes

Annealing at Temperatures Close to Melting Temperature Range

Chain rearrangements upon prolonged annealing of HDPE and i-PP at

elevated temperatures result in an increase in the lamellar thickness at the
expense of the interfacial layer and the soft amorphous phase (21, 22). Annealing
also causes a decrease in molecular mobility in the crystalline phase due to
perfecting of the crystalline order. Lamellae thickening and a slight increase in the
crystallinity would cause additional slippage of chain entanglements towards the
inner part of inter-lamellar amorphous regions causing additional immobilization
of the soft fraction of the amorphous phase. The following mechanism of
morphological changes during annealing at elevated temperatures is suggested.
Upon approaching the melting temperature, molecular mobility increases both
in the amorphous phase, the crystal-amorphous interface, and in the crystalline
phase. An increase in molecular mobility is accompanied by partial melting of
small crystals and less ordered fragments of lamellae. An increased mobility in
the amorphous phase and chain diffusion in and out of the crystals (c-relaxation)
facilitates structural reorganizations, which occur in the amorphous layer adjacent
to the lamella surface, causing a continuous shift of the interface towards the inner
part of the amorphous regions and thus reducing the thickness of the amorphous
layer and increasing the lamellae thickness. The thickness of rigid (crystalline)
domains increases linearly as a function of the logarithm of the annealing time,

185
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 which is in good agreement with theoretical predictions and experimental results
for other polymers (36, 37).

Aging of i-PP at Temperatures Slightly Exceeding Tg

A significant change in impact strength, flexural modulus and dimensions

of injection molded part, which are made from i-PP, occurs upon prolonged
aging at ambient temperature (38). The exact origin of these changes is not well
understood. NMR T2 relaxation experiments has helped in understanding this
phenomenon (38). Molecular mobility of the rigid fractions, which at 28C is
composed of crystalline phase and rigid amorphous fraction, is hardly affected
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

by aging for 1000 hours. The amount of the rigid fraction increases by ~12
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

wt%. Aging causes a significant decrease in molecular mobility in the soft

amorphous fraction. This decrease can be caused by two different phenomena:
(i) the formation of small crystals in inter-spherolitic amorphous domains during
the secondary crystallization, and (ii) an increase in the density of the amorphous
phase due to relaxation of residual stresses. The results obtained suggest that
aging-induced changes in the macroscopic properties of injection molded articles
largely originate from a decrease in molecular mobility in the amorphous phase.

Long Term Aging at Elevated Temperatures and Hydrostatic Pressure

One of the important applications of random propylene copolymers (PPR)

is the production of hot water pipes. The pipes can be used under hydrostatic
pressure and at elevated temperatures up to 70C continuously for 50 years and
for a short time at 80C. Breakage of pipes can occur if aging time at a certain
temperature exceeds that of the specification. It has been shown that 1H NMR
T2 relaxation analysis is the most sensitive tool for determining changes in PPR
samples that are caused by storage of PPR pipes at these conditions (26). The long
term aging causes the following changes in the phase composition and molecular
mobility. The crystallinity increases by 5% with increasing storage temperature.
The amount of semi-rigid and soft amorphous fractions slightly decreases upon
storage at the expense of the highly mobile chain fragments in the amorphous
phase, which have a higher concentration of ethylene comonomer units than chain
fragments in the soft amorphous fraction.
The following processes, which occur in PPR due to exposure to hydrostatic
pressure at elevated temperatures, are suggested (26). (i) Creep upon hydrostatic
pressure causing chain elongation in the amorphous phase. (ii) The increase in
crystallinity, which can occur due to the attachment of chain fragments in the
amorphous phase to existing crystals and/or due to the formation of new crystals.
In both cases, mobility in the amorphous phase decreases due to a decrease in
the chain length between adjacent crystals, i.e., the length of tie molecules. (iii)
A better phase separation of ethylene-rich chain fragments from the PP matrix.
The better phase separation results in an additional decrease in molecular mobility
in the PPR matrix, since more mobile chain fragments with ethylene chain units,

186
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 if they are mixed on the molecular scale with PP chain fragments, cause some
increase in molecular mobility of the PP matrix due to intermolecular coupling of
chain motions. All these changes cause embrittlement of pipes followed by their
breakage. It was shown that the combination of DSC and NMR relaxometry is a
powerful tool for the investigation of the molecular origin of damages in broken
PPR pipes and fittings, and for estimating their usage temperature and time.

The Effect of Deformation on Phase Composition, Molecular Mobility, and

Domain Sizes

Ethylene Copolymers
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

Changes in phase composition and molecular mobility were studied for

vulcanized semi-crystalline ethylene-octene copolymer (m-PE) and amorphous
EPDM rubber using a specially designed device, which made it possible to
perform NMR experiments under fixed uniaxial compression and to record
applied force (33). These two samples differ as regards network structure, i.e.,
EPDM has a larger density of chemical crosslinks, whereas the total network
density in m-PE is significantly larger because small crystallites act as physical
network junctions. These differences in the network structure largely influence the
elastic properties. Despite the larger total network density in m-PE, strain-induced
orientation of EPDM chains in compressed sample is significantly greater,
suggesting that the density of chemical crosslinks largely determines the elastic
behavior. Crystallinity of m-PE does not increase upon compression. It appears
that compression causes rearrangements of crystallites due to chain detachment
from the crystal surface and its attachment to neighboring crystals. It can be
concluded that physical network junctions, which originate from chains anchoring
to crystallites, would not be efficient in bearing the force, and the applied force is
mainly carried by chemical crosslinks and trapped chain entanglements. Thus,
an increase in the density of chemical crosslinks is required for improvement
the elastic recovery (the compression set) of m-PE. It has been demonstrated
that 1H NMR T2 relaxation experiments under uniaxial compression can provide
useful information to help in a better understanding the viscoelastic behavior of
heterogeneous elastomeric materials, such as semi-crystalline polymers, filled
elastomers, thermoplastic elastomers and thermoplastic vulcanizates.

Isotactic Polypropylene

Changes in phase composition, molecular mobility and domain thickness in

uniaxially stretched i-PP have been investigated as a function of temperature, draw
ratio (), drawing temperature and drawing rate (39). The NMR experiments have
been performed on samples after strain recovery. For the drawing temperature
range between 25 and 110C, typical ductile deformation behavior with a yield
point, neck formation and propagation, and strain hardening have been found.

187
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The Effect of Drawing Rate

Deformation-induced changes in the phase composition and molecular

mobility are hardly affected by changes in drawing speed which is varied from 1
to 100 mm/min. This suggests that chain rearrangements, which lead to changes
in morphology and physical structures upon drawing, are fast as compared to the
total time scale of deformation used, or the effect of drawing speed diminishes
during strain recovery.

The Effect of Drawing Ratio

Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

Drawing to = 7, which corresponds to beginning of the strain hardening,

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

causes a small increase in the amount of the rigid fraction, which indicates the
presence of highly strained tie molecules in the amorphous phase. Upon increase
in from 10 to 16, a further increase in the amount of the rigid fraction is
observed. This increase can be attributed to strain-induced crystallization and/or
large immobilization of chains in the amorphous phase. At = 7, molecular
mobility in the semi-rigid and soft fractions is already largely decreased, and it
continues to decrease up to the highest draw ratio.

The Effect of Drawing Temperature

Drawing temperature has a large effect on the strain-induced transformation

of the spherulitic morphology of i-PP to fibrillar morphology. The amount of
the rigid fraction increases with increasing drawing temperature for all draw
ratios. Molecular mobility in crystalline and amorphous phases is lower at higher
deformation temperatures. Relative differences in imperfections in the crystalline
structure can be identified by comparing T2 relaxation time for the rigid fraction -
T2r. At the same , T2r is longer at lower deformation temperatures. This indicates
larger imperfections in the crystalline phase of i-PP, which is stretched at 25C, as
compared to the undeformed sample. At drawing temperature of 25C, the long
period and the thickness of the rigid domains slightly decrease at the expense of
the thickness of the semi-rigid and soft domains, which suggests defragmentation
and disordering of crystals. At higher drawing temperatures, an increase in the
long period and lamellae thickness is observed upon increasing the draw ratio and
drawing temperature, whereas the thickness of amorphous layer, which separate
adjacent crystals, slightly decreases. As far as molecular mobility is concerned,
it decreases with increasing draw ratio both in the rigid and soft phases. The
decrease is more pronounced at higher deformation temperatures, which points
out a more perfect structure organization due to partial melting followed by
recrystallization at higher deformation temperatures. The observed changes in
the phase composition and molecular mobility suggest that faster rate of chain
motions in the crystalline phase (c-relaxation) facilitates transformation of the
spherulitic morphology to a fibrillar one. Moreover, partial melting followed by
strain-induced crystallization can possibly occur at higher temperatures.

188
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Conclusions
A series of studies of polyolefins shows that solid-state low-field 1H NMR
provides a powerful and complementary tool to traditional methods to obtain
information about physical structures, the phase composition, molecular mobility
in different phases. Even small changes in the chemical composition, thermal
history and a mechanical load can significantly influence molecular mobility in
the amorphous phase. This information is useful to achieve a better understanding
of yield and deformation behavior of polyolefins. If more detailed information
about physical phases and their chemical origin is required, high-field 13C NMR
spectroscopy should be used .
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

Acknowledgments
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

The studies of the author were sponsored by DSM and SABIC-Europe. The
author thanks D. Demco, C. Hedesiu, K. Remerie, M. Soliman, G. Vanden Poel, R.
Kleppinger, R. Deblieck, W. Gijsbers, B. Blmich and V. Mathot for collaboration
on topics presented in this paper.

References
1. Isasi, J. R.; Mandelkern, L.; Galante, M. J.; Alamo, R. G. J. Polym. Sci.,
Polym. Phys. 1999, 37, 323.
2. Baker, A. M. E.; Windle, A. H. Polymer 2001, 42, 667.
3. Hiss, R.; Hobeika, S.; Lynn, C.; Strobl, G. Macromolecules 1999, 32, 4390.
4. Iwata, K. Polymer 2002, 43, 6609.
5. Sajkiewicz, P.; Hashimoto, T.; Saijo, K.; Gradys, A. Polymer 2005, 46, 513.
6. Schick, C.; Wurm, A.; Mohammed, A. Thermochim. Acta 2003, 396, 119.
7. Androsch, R.; Wunderlich, B. Polymer 2005, 46, 12556.
8. Gautman, S.; Balijepalli, S.; Rutledge, G. C. Macromolecules 2000, 33, 9136.
9. Schmidt-Rohr, K.; Spiess, H. W. Macromolecules 1991, 24, 5288.
10. Hillebrand, L.; Schmidt, A.; Bolz, A.; Hess, M.; Veeman, W.; Meier, R. J.;
Van der Velden, G. Macromolecules 1998, 31, 5010.
11. Kuwabara, K.; Kaji, H.; Tsuji, M.; Horii, F. Macromolecules 2000, 33, 7093.
12. Litvinov, V. M.; Mathot, V. B. F. Solid State Nucl. Magn. Reson. 2002, 22,
218.
13. Kristiansen, P. E.; Hansen, E. W.; Pedersen, B. J. Phys. Chem. B 1999, 103,
3552.
14. Kristiansen, P. E.; Hansen, E. W.; Pedersen, B. Polymer 2000, 41, 311.
15. Tanzer, C. I.; Roy, A. K. Proc. SPE ANTEC 95: Boston, MA, USA 1995, 2,
2700.
16. Bergmann, K. J. Polym. Sci., Polym. Phys. 1978, 16, 1611.
17. Fedotov, V. D.; Abdrashitiva, N. A. Polym. Sci. USSR 1985, 27, 287.
18. Kakudate, T.; Kakizaki, M.; Hideshima, T. J. Polym. Sci.: Polym. Phys.
1985, 23, 787.
19. Eckman, R. R.; Henrichs, P. M.; Peacock, A. J. Macromolecules 1997, 30,
2474.
189
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 20. Hansen, E. W.; Kristiansen, P. E.; Pedersen, B. J. Phys. Chem. B 1998, 102,
5444.
21. Hedesiu, C.; Kleppinger, R.; Demco, D. E.; Buda, A.; Blmich, B.;
Remerie, K.; Litvinov, V. M. Polymer 2007, 48, 763.
22. Hedesiu, C.; Demco, D. E.; Kleppinger, R.; Vanden Poel, G.; Gijsbers, W.;
Blmich, B.; Remerie, K.; Litvinov, V. M. Macromolecules 2007, 40, 3977.
23. Zhang, L.; Hansen, E. W.; Helland, I.; Hinrichsen, E.; Larsen, A.; Roots, J.
Macromolecules 2009, 42, 5189.
24. Mauri, M.; Thomann, Y.; Schneider, H.; Saalwchter, K. Solid State Nucl.
Magn. Reson. 2008, 34, 125.
25. Litvinov, V. M.; Penning, J. P. Macromol. Chem. Phys. 2004, 205, 1721.
26. Litvinov, V. M.; Soliman, M. Polymer 2005, 46, 3077.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

27. Tanaka, H.; Inoue, Y. Polym. International. 1993, 31, 9.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch011

28. Dadauli, D.; Harris, R. K.; Kenwright, A. M.; Say, B. J.; Snnetiolu, M.
M. Polymer 1994, 35, 4083.
29. Dujourdy, L.; Bazile, J. P.; Cohen-Addad, J. P. Polym. Int. 1999, 48, 558.
30. Schreurs, S.; Franois, J. P.; Adriaensens, P.; Gelan, J. J. Phys. Chem. B
1999, 103, 1393.
31. Weglarz, W. P.; Peemoeller, H.; Rudin, A. J. Polym. Sci., Part B: Polym.
Phys. 2000, 38, 2487.
32. Kenwright, A. M.; Say, B. J. In NMR Spectroscopy of Polymers; Ibbett, R.
N., Ed.; Blackie Academic & Professional: London, 1993; p. 231.
33. Litvinov, V. M. Macromolecules 2001, 34, 8468.
34. Litvinov, V. M. Macromolecules 2006, 39, 8727.
35. Int Veld, P. J.; Htter, M.; Rutledge, G. C. Macromolecules 2006, 39, 439.
36. Peterlin, A. Polymer 1967, 6, 25.
37. Sanchez, I. C.; Peterlin, A.; Egy, R. K.; McCrackin, F. L. J. Appl. Phys.
1974, 45, 4216.
38. Hedesiu, C.; Demco, D. E.; Kleppinger, R.; Vanden Poel, G.; Remerie, K.;
Litvinov, V. M.; Blmich, B.; Steenbakkers, R. Macromol. Mater. Eng.
2008, 293, 847.
39. Hedesiu, C.; Demco, D. E.; Remerie, K.; Blmich, B.; Litvinov, V. M.
Macromol. Chem. Phys. 2008, 2009, 734.

190
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 12

Molecular Dynamics and Structure of the

Crystalline Region of Isotactic-Polyolefins
Characterized by Solid-State NMR
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Toshikazu Miyoshi*

Department of Polymer Science, The University of Akron,

Akron, OH 44325, U.S.A.
*E-mail: [email protected]

A review is given of recent solidstate NMR work on structure

and molecular dynamics of polyolefins such as isotactic
poly(1butene) (iPB1) and isotactic polypropylene (iPP). In
iPB1, meltcrystallization leads to metastable form II, and
subsequent solidsolid transition results in stable form I. 1H13C
wide line separation (WISE) and center bands only detection
of exchange (CODEX) NMR techniques provided geometric
and timekinetic parameters of molecular dynamics for overall
and side chains in both forms at natural abundance. The 1H
lineshape analysis using a twospin approximation and 13C T1
results indicate that form II chains perform uniaxially rotational
diffusions accompanying sidechain conformational transitions
in the fast motional limit (correlation time < 10-7 s). After
irreversible solid transition, 13C CODEX results indicate that
crystalline stems and sidechain conformations are completely
fixed, up to the melting points (correlation time > 10 s). The
effects of the unusual molecular dynamics on the mechanical
properties are discussed. The second part of this article deals
with a reinvestigation (by high resolution 13C NMR) of packing
structures of the form of iPP. TwoPulse PhaseModulation
decoupling during a rather long acquisition time results
in significantly narrowed signals corresponding to ordered
packing. The enhanced spectral resolution allows us to properly
evaluate orderdisorder phenomenon in polymer crystals. From
CODEX experiments, lamellar thickness and packing order

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 apparently suppress helical jumps in the crystalline regions.
These experiments provide valuable information on chainlevel
structures and dynamics, particularly on the variations in
lamellae thickness as a function of supercooling.

Introduction
Understanding relationship between structure and molecular dynamics is one
of the particularly important subjects in polymer science. There are continuous
interests to understand molecular dynamics in the crystalline (1), amorphous (2),
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

and interfacial regions as well as in the limited dimensions such as thin films
(3) and nanopores (4). Molecular dynamics in several crystalline polymers has
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

been well characterized by solidstate NMR (1). Dynamic geometry and time
kinetics of molecular dynamics in the crystalline regions are significantly affected
by conformation, packing, and, interstem interaction. SchmidtRohr and Hu
categorized polymer crystals as either mobile or fixed crystals (1). The flexible
polymer chains such as polyethylene (PE) (5), isotacticpoly(propylene) (iPP) (6),
poly(oxy methylene) (7), and poly(ethylene oxide) (8), etc. are categorized in
mobile, which commonly shows helical jump motions that combine the dynamics
of well defined discrete jump motions and translations along chain axis. On the
other hand, Nylon and poly(ethylene terephtalate) (PET), etc. are examples of
fixed crystals. English et al (9) demonstrated using 2H line shape that Nylon
essentially does not perform overall molecular motions in the crystalline region
up to melting temperature (Tm).
Polyolefins consists of hydrocarbons and do not have specific functional
groups. They are ideal semicrystalline polymers to further investigate i) the
relationship between structure and molecular dynamics, and ii) the roles of
molecular dynamics for bulk property and structural organization of polymer
crystals.
In this work, a review is conducted of our recent works on detailed dynamic
analysis of iPB1 (10, 11) and iPP (12) by using center bands only detection of
exchange (CODEX) (13) and wide line separation (WISE) NMR (1416). On the
basis of the experimental data, useful information has been obtained on molecular
dynamics, structure, and bulk material properties.

Experimental Section
Samples

iPB1 with an average molecular weight of Mw = 186,000 and a polydispersity

of Mw/Mn = 3.3 and isotacticity, (mmmm) = 92 % was purchased from
Polysciences, Inc. iPP with Mw of 360,000 and Mw/Mn of 3.3 and isotacticity,
(mmmm) of 97 % was purchased from Polysciences, Inc. The sample was melted
and crystallized at a required crystallization temperature, Tc under nitrogen
atmosphere. Equilibrium melting temperature, Tm0, of iPB and iPP were 135
2 and 184 4 C, respectively.
192
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Measurements

NMR

The solidstate 13C NMR experiments were carried out on a BRUKER

AVANCE 300 spectrometer, equipped with 4 and 7 mm VT CPMAS NMR probe.
The 1H and 13C carrier frequencies are 300.1 and 75.6 MHz, respectively. The
MAS frequency was set to 4000 and 6800 3 Hz. The 90 pulses for 1H and 13C
were 4.55.0 s. The recycle delay and crosspolarization (CP) time were 2 s
and 1 ms, respectively. High power 1H TwoPulse PhaseModulation (TPPM)
decoupling with a field strength of 65 kHz was used during acquisition time of 80
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

ms. For iPP, severe decoupling conditions with a field strength of 110 kHz and
acquisition time of 160 ms were used. Phase sensitive 1H13C 2D WISE NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

spectrum was obtained in TPPI mode. The data matrix had 512 points along the t2
dimension and 128 points along the t1 dimension with a dwell time of 2 and 7 s.
A short CP of 50 s was utilized. Continuous wave (CW) decoupling with a field
strength of 55 kHz was applied in t1 dimension to suppress 1H13C heteronuclear
interaction (15, 16). The CODEX experiments utilize the recoupling of chemical
shift anisotropy (CSA) interaction by 180 pulses trains in the two evolution
periods sandwiching a mixing period, tmix (13). The MAS frequency was 4000
3 Hz. The 1H rf field strength for CW decoupling during 13C 180 pulse with
pulse length of 15 s was set to 100 kHz. All other rf parameters were the same
as for the CPMAS experiments. The reference and exchange experiments were
obtained alternatively by every 128 transients, to suppress drift of the NMR
spectrometer. The T1H filter was incorporated into CODEX pulse program, for
suppression of the amorphous signal contribution to CODEX results at ambient
temperature and 10 C. Totally, each spectrum was obtained by accumulating
1024 transients. The total experimental time for mixingtime dependence up
to 4 s was approximately 24 hs. The total experimental time of a typical Ntr
experiment was about 12 hs.

SAXS

SAXS studies were carried out using a CuK radiation generated by a Rigaku
Ultrax 4153A 172B Xray diffractometer, and a pointfocusing SAXS camera.
The camera length used was 740 mm and the images were recorded using an image
plate (IP) with an exposure time of 2.5 h. Digitized data were then read from the IP
using the image plate reader. Using IP, very small change in SAXS patterns could
be obtained with a very short exposure time. The corrected pattern of an empty
sample holder was subtracted from each pattern. To calculate he long period and
lamellar thickness, a correlation-function method by Rigaku Raxis software was
used.

193
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 DSC

All the samples were measured with an updated computer interfaced

PerkinElmer DSC7. Both temperature and heatflow levels were corrected
by standard materials. Measurements of the melting points were performed at
a heating rate of 10 C /min. To prevent thermal degradation, nitrogen gas was
circulated around the sample pan.

Results and Discussion

Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

iPB1
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

iPB1 is an important industrial material due to outstanding mechanical

strengths relative to chemically similar polyolfins such as PE, iPP, and
isotacticpoly(4methyl1pentene) (iP4M1P) (17). iPB1 crystallizes from melt
under stationary conditions, kinetically favoring the metastable tetragonal crystal
with 113 helices (form II) (18, 19). This form II spontaneously transforms into
the stable trigonal crystal of 31 helices (form I) (20) at ambient temperature via
solidsolid transition (21). This solid transition leads to superior mechanical
property without changing lamellae thickness and crystallinity (17, 22). However,
origin of the mechanical property is not fully understood.
Figures 1(a) and (b) show the 1H MAS spectra of a form IIrich sample at
100 C at a MAS frequency of 4 kHz. The spectrum is dominated by the isotropic
component with intense SSB patterns. The isotropic signal corresponds to the
amorphous component and SSB envelop originates from form II. Very intense SSB
patterns reflect fast and anisotropic complex dynamics in polymer backbones (23),
conformational transitions in the ethyl group (24), and methyl rotations (25). The
13C relaxation time, T1 is sensitive to high frequency motions with a minimum

value of ~ 0.1 s when c of molecular dynamics approaches to an inverse of ~

75 MHz. In fact, the mainchain (m) CH2, side chain (s) CH2, and CH have
T1 values of 0.25, 0.32, and 0.38 s, respectively, at 100 C. This simple T1 result
suggests the dynamic frequency for the coupled side group and backbone motions
is in the order of ~ 10 MHz or higher. Such high frequency dynamics are rare in
the solid state and are rather close to the frequency range in the melted state.
1H13C WISE NMR spectrum was obtained to further investigate functional

mobility at a MAS frequency of 6.8 kHz at 100 C. Figure 1(c) and (d) shows
1H slice data of mCH2 and sCH2 protons obtained through highly resolved 13C

signals. A short CP (CP time = 50 s) and 13C CW decoupling in t1 dimension

were applied to selectively observe the crystalline signals and to quench spin
diffusions between the different functional groups in the crystalline region and
suppress heteronuclear dipolar interactions, respectively. The slice data clearly
indicate that isotropic components corresponding to the amorphous ones are not
polarized to 13C signals.

194
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

Figure 1. (a) 1H MAS NMR spectrum, (b) amplified vertically by 20 times at a

MAS frequency of 4 kHz at 100 C; and c) 1H slice date of 1H13C WISE NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

spectra of (c) the main chain (mCH2) and (d) side chains CH2 (sCH2) protons
through high resolved corresponding 13C signals at a MAS frequency of 6.8 kHz
at 100 C. The (c) red and (d) blue spectra showing best fitted curves to the
experimental data with bestfit parameters of (c) =0.44 0.02 and =0
(DII/2 = 9.0 0.5 kHz) and (d) =0.61 0.02 and = 0 (DII/2 = 12.5
0.5 kHz), respectively. (see color insert)

Schnell et al (26) simulated third spin effects on the 1H SSB signals at n

order depending on geometry and distances. They realized that if a third spin
significantly interacts with the isolated spin pairs, asymmetric line shapes shall
appear, depending on the relative geometry, the mutual distances of three spins,
and the MAS frequency. Our experimental data indicate that all SSB signals at
n order are symmetric and have similar line widths as the center band. Thus, it
can be safely assumed that the two spin approximation holds for CH2 geminal
protons. Under two spin approximation, the dipolar patterns were calculated under
an assumption of averaged asymmetry parameter, = 0. The calculated line
shapes are well consistent with the experimental data.
Surprisingly, the sCH2 geminal protons show much larger dipolar coupling
constant, DII / 2 = 12.5 0.5 kHz than 9.0 0.5 kHz for mCH2. These
results were unexpected for the dynamics of the polymer with a side chain, since
sidechain protons are influenced by the complex dynamics of both the main
and side chains. Generally, two independent dynamic processes with different
dynamic axes result in a smaller dipolar coupling constant in the side group than
that in the main chain. In fact, Mirau et al (16) detected smaller linewidths for
the side group than that in main chain of poly(ethyl methacrylate) (PEMA) at a
temperature above Tg. Our experimental results in form II are opposite to those
expected in the sidechain polymers in glassy polymers but shall be explained by
the assumption of an anisotropic dynamical process.
In the crystalline region, parallel packing of the stems restricts the overall
dynamics to uniaxial motions around the stem axes. For several conformational
disordered (CONDIS) crystals such as PE in the hexagonal phase (27),
1,4transpolybutadiene in the high temperature phase (28), and poly(diethyl
siloxane) (PDES) (29), it has been reported by SSNMR that stems commonly
show uniaxial diffusions around the stem axes. Under such conditions, is 0 and
195
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 absolute averaged anisotropy parameter, is described in terms of a following
equation:

where is the anisotropy parameter of CH2 geminal protons under no motion (20.6
kHz), is an angle between dipolar vector of the germinal protons and dynamic
axis (24).
Figure 2 illustrates a part of 113 helix of form II and possible dynamic models.
The mCH2 protons were colored by red. The dipolar vector of germinal protons
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

of mCH2 in 113 helix is inclined by an angle of 82 with respect to the stem

axis. Under uniaxial diffusion in a fast limit, the value for mCH2 geminal
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

protons is calculated to be 0.47. This value is consistent with the experimentally

obtained value of 0.44 0.02. The small reduction in the observed value may
be due to small librations around the CC bonds of main chains and/or slight
conformational deviations from an ideal 113 helix.

Figure 2. Schematic illustration of dynamic effects of overall chain diffusions

and conformational transitions on m and sCH2 dipolar vectors which being
colored by red and blue, respectively, of form II of iPB1. The sidechain
dynamics being defined by a jump angle , where being a difference angle
between two interchanging conformations defined by dihedral angles, 1 and
2. (see color insert)

Former high resolution 13C NMR (24) indicated that ethyl groups in form II
adopt two conformations and perform conformational transitions at temperatures
above 110 C. The complex dynamics of sidechain and overall dynamics would
contribute to the value for the sCH2 geminal protons (0.61 0.02) (Figure
196
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 2). Two independent dynamic effects on dipolar patterns were numerically
calculated as functions of dihedral angle, and conformational transition angle, ,
where corresponds to a difference between dihedral angles in two interchanging
conformations, 1 2, where lower numbers 1 and 2 denote two different
conformations. Here, the population ratios of two conformations obtained in
former high resolution NMR at 110 C were used (24). The calculated result
on value is illustrated in Figure 3. The overall dynamic effect of sCH2
protons in different conformations with a dihedral angle, , on is shown
along a horizontal axis. The conformational transition effect with a jump angle,
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

, on is shown as vertical axis, The calculated vales of = 0.4, 0.5,

and 0.6 are shown as dotted solid black, and red curves, respectively, in Figure
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

3. It is clearly seen that only = 88 and 268 10 and their transitions with
= 180 20 or no dynamics (0 20) satisfy the experimental value. Other
conformations and jump angles show much smaller values than the experimental
result ( = 0.61). The obtained structures and dynamics ( = 88 and 268
10, and = 0 and 180 20) also satisfy = 0. However, the lack of lateral
dynamics does not average out higher spin interactions within the stem. In view
of multispin interactions, it is concluded that chains in form II of iPB1 performs
uniaxial diffusions accompanying the conformational transition ( = 180 20)
of the side chain between two conformations at = 88 and 268 10 in the fast
motional limit (10). The obtained molecular dynamics clearly indicate that form
II is a CONDIS crystal.

Figure 3. Effects of overall motions and side-chain conformational transition

with a jump angle, on value of sCH2 geminal protons at a dihedral
angle of under two spin approximation. corresponding to a difference
angle between dihedral angle in two interchanging conformations. The dotted
and solid black and red curves corresponding to = 0.4, 0.5, and 0.6,
197
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 respectively. Reproduced with permission from ref (10). Copyright 2010 ACS.
(see color insert)

Figure 4(a) shows 1H MAS NMR spectrum for form Irich sample of
iPB1 enlarged vertically by 20 times at 118 C. The spectrum consists of two
components of the broad crystalline and narrow amorphous signals. The observed
1H line shape in crystalline regions show a typical rigid spectral pattern which is

largely different from those of form II. Figure 4(b) and (c) shows 1H slice data
of 1H13C WISE spectra in the main and side chains, respectively. The obtained
line shapes were fitted by Gaussian functions. The best fitted results indicate that
1H full line width at half height (flwhh) is 57, 57, 54, and 37 kHz, for mCH2,
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

CH, and sCH2, and CH3 protons, respectively. This result indicates that overall
and sidechain conformation in form I do not perform large amplitude motions
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

at frequency over ~ mid 10 kHz even at high temperature of 118 C close to

equilibrium melting temperature (Tm0 = 135 C).

Figure 4. (a) 1H MAS NMR spectrum amplified vertically by 20 times and 1H

slice date of 1H13C MAS WISE NMR spectra of (b) the side chains and (c) main
chains of the form I richsample at 118 C. The dotted lines showing 1H slice
date through 13C CH and CH3 signals, and solid lines showing 1H ones through
13C sCH2 and mCH2 signals.

The CODEX experiments were carried out to investigate molecular dynamics

of from I in slow dynamic range. If there is molecular motion during tmix, the
orientationdependent frequency before and after tmix will be different, and the
magnetization will not completely be refocused. The resulting dephasing leads
to a decay of the signalintensity in the exchange spectrum (S). If there is no
molecular motion during tmix, the evolutions in the two evolution periods will
cancel each other, and there will be no decay of the signal intensity. To remove
relaxation effects during tmix and Ntr, a reference spectrum is acquired. The signal
intensity in the reference spectrum (S0) is not sensitive to exchange process but is
only dominated by T1, T2, and pulse length errors. The motional correlation time
and information about the motional geometry can be obtained by plotting the ratio
(S/S0) versus tmix and (S/S0) versus Ntr, respectively.
Figure 5(a) shows CODEX reference, exchange, and different spectra of iPB1
form I with a tmix = 200 ms and Ntr = 2 ms at ambient temperature. S - S0 is
nearly zero in all the functional signals. This means that there is no slow molecular
dynamics which reorients CSA interactions in all functional carbons during tmix at
Ntr= 2 ms.
198
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 5(b) and (c) shows tmix dependence of (S/S0)* intensity ratios of
mCH2, CH, CH3 at ambient temperature and 118 C, respectively, where *
means pure CODEX (S/S0) decay after spindiffusion corrections. The (S/S0)*
intensity ratios of mCH2 and CH carbons do not show essential decay up to tmix
= 4 s with a fixed Ntr = 2 ms at ambient temperature and 118 C. Here, it is noted
that 118C is a few degree lower than onset of melting. The (S/S0)* ratios of CH3
signal at tmix 2 s are fluctuated and shows large experimental errors. This is
caused not by molecular dynamics but by short 13C T1 value. The solid curve to
the experimental data of CH signal shows <c> = 860 5000 s at 118 C. This
long value is out of dynamic window in our experiments ( 10 s). Thus obtained
<c> is not reliable.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Figure 5. (a) 13C CODEX reference, exchange, and difference spectra of iPB1
form I at a mixing time of 200 ms and Ntr =2.0 ms at ambient temperature. (b)
and (c) CODEX tmix dependence of (S/S0)* and (d) Ntr dependence of (S/S0)
intensity ratios of mCH2 (), CH (), and CH3 (*) with Ntr = 2 ms at (b) ambient
temperature, and (c) and (d) 118C. Reproduced from ref (11).
Figure 5(d) shows Ntr dependence of (S/S0) intensity rations of mCH2, CH,
and CH3 13C signals with a tmix = 200 ms at 118 C. The (S/S0) intensity ratios
of all the signals do not decay with increasing Ntr up to 4.5 ms even at high
temperature of 118 C. The solid line with a jump angle of 0 is well consistent
with the experimental data within the experimental errors. Therefore, both
tmix and Ntr dependences of CODEX results indicate that there is no evidence
supporting overall and sidechain motions which induce reorientations of CSA
principle axes of all the functional signals of iPB1 crystal in slow dynamic range
even at very high temperatures close to melting point. Thus, polymer chains in
forms I and II show markedly different mobility. According to crystallographic
data (14), solidsolid transition leads to 15 % contraction in ab plane and 12 %
199
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 extension along c axis (chain axis). Such densifications lead to huge dynamic
differences in geometry and time kinetic parameters in different polymorphs of
the same polymer.
Here, we consider how molecular dynamics in forms I and II contribute
to structural organization and superior mechanical property even at very high
temperatures of iPB1. Extremely high mobility in form II is highly related to
crystallization process. For example, crystallization of form II of iPB1 completes
within 3 days at 110 C (T = 25C)) The crystallized sample shows lamellae
thickness, <l>, of ~ 28 nm at T = 25C (11). For comparison, iPP which is an
example of mobile crystals shows <l> of ~ 23 nm at similar supercooling (T
= 23C) (12). This crystallization takes about 12 days. Namely, extremely fast
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

dynamics in form II shows much faster crystallization and makes much thicker
lamellae than mobility in a typical mobile crystal.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Irreversible solidsolid transitions not only lead to drastic changes of stem

mobility from extremely fast to slowlimit dynamics. Slowlimit dynamics of
chains in form I is largely different from mobility of similar polyolefins such as
PE (5), iPP (6), and iP4M1P (30). In these polymers, it is suggested that dissipation
of mechanical property is related to chain dynamics in the crystalline regions (5).
Thus, it is indicated that immobility of chains in form I crystal plays a critical
role for superior mechanical property. Additionally, solidsolid transition process
preserves very thick lamellae and high crystallinity which extremely fast chain
dynamics in form II makes (22). One step crystallization into fixed crystal leads to
thin lamellae with a low crystallinity (11). Therefore, it is reasonably concluded
that molecular dynamics in two step processes of i) crystallization into form II
(extremely mobile crystal) and ii) subsequently irreversible solidsolid transitions
into form I (fixed crystal) play critical roles for superior mechanical strength of
iPB1 even at very high temperatures.

iPP
We have shown that extremely high mobility in form II of iPB1 results in
significant thickness at very high temperatures. This result simply suggests that
crystalline mobility plays a critical role in lamellar thickness. Hereinafter, we
investigate contributions of other structural factors to lamellar thickness.
Stereoregular polymers adopt one of four types of helical structure consisting
of two independently structural factors of handness (right (R) or left (L)) and
orientation (upward (u) or downward (d)) in the crystalline region (31), namely,
Ru, Rd, Ld, and Lu, each of which shows an identical rotational energy in isolated
states. The chain conformation and packing in form of iPP have been well
characterized by XRD. form shows two limit packing structures as shown in
Figure 6 (a) and (b) (32). The chains in two limiting structures adopt 31 helices
but different orientations along c axis. There is no upward and downward disorder
of stem orientations in full limit order. On the other hand, the disorder statistically
occurs at each site in full limit disorder. Full limit disorder appears at a low Tc. The
ordered packing fraction increases with increasing crystallization temperatures Tc,
and reaches to 90100 % at Tc = 150 C (3335). Previous high resolution 13C
NMR also supported these XRD results (3638). However, spectral resolutions in
200
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 previous NMR was insufficient to properly evaluate orderdisorder phenomenon
due to strong 1H13C dipolar interactions.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Figure 6. Unit-cell structures in (a) full limit order and (b) full limit disorder for
form of iPP. (c) The 13C T1H filter CPMAS NMR spectra for form of iPP
depending on supercooling at ambient temperature and signal assignments.
Reproduced with permission from ref (12). Copyright 2010 ACS. (see color
insert)
Figure 6(c) shows 13C T1H filtered CPMAS NMR spectra for pure form
crystallized at various T. TPPM decoupling with a filed strength of 110 kHz
during a long acquisition time of 160 ms were applied to fully detect free
induction decays of the crystal signals. The structure less and inhomogeneously
broadened 13C signals were observed at T = 84 C. These line shapes correspond
to packing structures in full limit disorder. With decreasing T, newly narrowed
doublet signals corresponding to those in full limit order appear and grow on the
disordered line shapes. The observed spectral changing features are similar to
the formers. However, current result shows much higher spectral resolution, and
clearly indicates that the doublet signals reach to maximum intensities at T = 34
C and disordered and ordered structures coexist even at very high temperatures
less than T = 34 C. Considering CP and relaxation effects, improved 13C signals
determine the maximum ordered packing fraction, forder = 62 % at T = 34 C (12).
This result is mostly different from former XRD (3335) and NMR results (forder
= 90 100 %) (38). The obtained discrepancy is attributed to characterization
tools. XRD analysis uses very minor differences in diffraction as an evidence
201
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of orderdisorder phenomenon. This minor difference is affected by amorphous
halo, background, and neighboring peaks. Additionally, threedimensional
periodicity and sizes also influence the results. For these reasons XRD might
overestimate forder.
Current NMR results clearly indicate that polymer crystals include a large
amount of structural disorders even at a very low supercooling. This result
suggests that a host of factors (viz., packing energy, kinetics, chemical and
physical features, such as molecular weights, stereoregularity, regiodefects,
entanglements, and chain folding) influence the packing structures of form of
iPP under bulk crystallization (12).
The relationship between molecular dynamics and structures were further
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

investigated. As reported before (6), iPP shows helical jump motions in the
crystalline regions. Figure 7 shows the temperature dependence of <c> of helical
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

jumps of form of iPP obtained by CODEX, crystallized at T = 34 (<l> = 10

nm ) and 84 C (<l> = 23 nm ). The <c > of helical jumps in the ordered packing
of iPP_T34 is 1270 20 ms at 95 C and is 37fold longer than that in the
disordered packing of iPP_T84 (<c> = 34 1 ms ) at the same temperature.
The <c> vs. 1/T plot exhibits Arrhenius behavior with activation energies of Ea
= 102 5 and 72 4 kJ/mol for the ordered packing form of iPP_T34 (This
sample crystallize at T= 34 C) and for the disordered packing of iPP_T84,
respectively (12). These experimental results clearly indicate that both <c> and
Ea highly depend on packing order and lamellae thickness. If chain mobility alone
plays a critical role in increasing lamellae thickness, annealing would lead to very
thick lamellae without structural change. Thus, we investigate the relationship
between <l> and forder of iPP samples annealed at various temperatures.

Figure 7. Arrhenius plot of <c> for helical jump motions of stems in ordered
packing of iPP_T 34 () and in disordered packing of iPP_T 84 () with best

202
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 fitted values of Ea = 102 5 and 72 4 kJ/ mol, respectively. Reproduced with
permission from ref (12). Copyright 2010 ACS.
Figure 8 shows the <l> vs. forder curve during annealing at various
temperatures (Ta). The starting conditions of <l> and forder are 9 nm and 0 %,
respectively. With increasing Ta, the lamellar thickness increases by only 3 nm
without changing forder up to 140 C. At Ta = 150 C, forder suddenly increases to
25 %, and <l> also increases. Further increase of Ta induces further increases
in forder, accompanying large <l> enhancement. Figure 8 clearly indicates that
<l> significantly increases in the forder region of > 50 %. Actually, replacement
of upward and downward orientations (stem orientations) is impossible in the
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

limited spaces of solid crystals. This structural change requires partial melt
and recrystallization. Therefore, it is reasonable to consider thin lamellae that
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

partially melt, and after recrystallization which accompanies rearrangement of

polymer chains packing ordering, <l> significantly increases.

Figure 8. <l> versus forder of form of iPP under (b) annealing at 110 170 C.
Reproduced with permission from ref (12). Copyright 2010 ACS.
Here, we extract chainlevel information from the packing structures
under the hypothesis of adjacent reentry structure. In the case of stereoregular
polymers, chain folding is allowed between antichiral isomorphous or isochiral
antimorphous chains. This selection rule leads to a strong relationship between
packing structures and chainfolding directions. In the case of full limit order,
chain folding is possible only between alternating R and L stems along a*axis
(Figure, 9 (a)). On the other hand, there is no restriction on chainfolding
directions in full limit disorder (Figure 9(b)). Therefore, two limiting structures
on packing levels mean markedly different trajectories on chain levels. Thus,
Ta dependence of forder represents a degree of ordering of chainlevel structures
under hypothesis of adjacent reentry structures. Actual lamellae thickening
203
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 process at different Ta accompanies changes of chainfolding directions inside of
crystals. This fact indicates that not only molecular dynamics but also chainlevel
structures contribute to Ta dependence of lamellar thickness in the polymer
crystals. The same discussion would hold for isothermal crystallization (1/T
dependence of <l>).
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Figure 9. Possible models for of polymer chain trajectory of iPP in the

crystalline region in (a) full limit order and (b) full limit disorder. Reproduced
with permission from ref (12). Copyright 2010 ACS.

Conclusion
In the first part, extremely fast and slowlimit dynamics of polymer backbone
are shown to be sources for unusual mechanical property of iPB1. This experiment
clearly indicates that available dynamic spaces induce different dynamic geometry
and time kinetics in different polymorphs of the same polymer. Extremely fast
dynamics result in very thick lamellae within a relatively short crystallization time
at high Tc. After irreversible solidsolid transition, crystalline stems in form I do
not perform any overall and sidechain dynamics up to melting points in slow
motional limit (<c> > 10 s). The huge contrasts in these dynamics play a critical
role in the structural organization and the resulting material property of iPB1.
In the second part, we revisited packing order of form of iPP by high
resolution 13C NMR. Strong dipolar decoupling significantly sharpens 13C
signals of the ordered packing structures. Improved resolution of 13C crystalline
signals properly accesses orderdisorder phenomenon of the packing structures
depending on Tc and Ta. The experiments provide not only molecular dynamics
204
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 but also structural ordering information at molecular levels and the effects on the
lamellae thickening process.

Acknowledgments
We really appreciate Dr. H. N. Cheng for carefully checking our manuscript.

References
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

1. Hu, WG.; ShimidtRohr, K. Acta Polym. 1999, 50, 271285.

2. Ediger, M. D. Spatially Heterogeneous Dynamics in Supercooled Liquid.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

Ann. Rev. Phys. Chem. 2000, 51, 99128.

3. Serghei, A.; Kremer, F. Phys. Rev. Lett. 2003, 91, 16702, 14.
4. Uemura, T.; Horike, S.; Kitagawa, K.; Mizuno, M; Endo, K.; Bracco, K.;
Comotti, A.; Sozzani, P.; Nagaoka, M.; Kitagawa, S. J. Am. Chem. Soc.
2008, 138, 67816788.
5. Hu, W. G.; Boeffel, C.; SchmidtRohr, K. Macromolecules 1999, 32,
16111619.
6. Schaefer, D.; Spiess, H. W.; Suter, U. W.; Fleming, W. W. Macromolecules
1990, 23, 34313439.
7. Kentgens, A. P. M.; de Boer, E.; Veeman, W. S. J. Chem. Phys. 1987, 87,
68596865.
8. SchmidtRohr, K.; Spiess, H. W. Multidimensional Solidstate NMR and
Polymers, 1994 Academic Press.
9. Hirschinger, J.; Miura, H.; Gardner, K. H.; English, A. D. Macromolecules
1990, 23, 21532169.
10. Miyoshi, T.; Mamun, A.; Reichert, D. Macomoelcules 2010, 43, 39863989.
11. Miyoshi, T.; Mamun, A. Polym. J., in press.
12. Miyoshi, T.; Mamun, A.; Hu, W. J. Phys. Chem. B. 2010, 114, 92100.
13. deAzevedo, E. R.; Hu, W. G.; Bonagamba, T. J.; SchmidtRohr, K. J. Am.
Chem. Soc. 1999, 121, 84118412.
14. SchmidtRohr, K.; Clauss, J.; Spiess, H. W. Macromolecules 1992, 25,
32733277.
15. Tekely, P.; Palmas, P.; Mutzenhardt, P. Macromolecules 1992, 26,
73637365.
16. Qiu, X.; Mirau, A. P. J. Magn. Reson. 2000, 142, 183189.
17. Lugiani, L.; Seppl, J.; Lfgren, B. Prog. Polym. Sci. 1988, 13, 3762.
18. TurnerJones, A. J. Polym. Sci. Part B. 1963, 1, 455456.
19. Petraccone, V.; Pirozzi, B.; Frasci, A.; Corradini, P. Eur. Polym. J. 1976, 12,
323327.
20. Natta, G.; Corradini, P.; Bassi, I. W. Nuov. Chim. Suppl. 1960, 15, 5267.
21. Maruyama, M.; Sakamoto, Y.; Nozaki, K.; Yamamoto, T.; Kajioka, H.;
Toda, A.; Yamada, K. Polymer 2010, 51, 55325538.
22. Azzuri, F.; Flores, A.; Alfonso, G. C.; Sics, I.; Hsiao, B. S.; Balta; Calleja, F.
J. Polymer 2003, 44, 16411645.
205
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 23. Beckham, H. W.; SchmidtRohr, K.; Spiess, H. W. ACS Symp. Ser. 1995,
598, 243253.
24. Miyoshi, T.; Hayashi, S.; Imashiro, F.; Kaito, A. Macromolecules 2002, 35,
60606063.
25. Maring, D.; Meurer, B.; Weill, G. J. Polym. Sci., Part B: Polym. Phys. 1995,
33, 12351247.
26. Schnell, I.; Spiess, H. W. J. Magn. Reson. 2001, 151, 153227.
27. de Langen, M.; Prins, K. O. Chem. Phys. Lett. 1999, 299, 195200.
28. Mller, M. Makromol. Chem. Rapid Commun. 1988, 9, 107114.
29. Litvinov, V. M.; Macho, V.; Spiess, H. W. Acta Polym. 1997, 48, 471477.
30. Miyoshi, T.; Pascui, O.; Reichert, D. Macromolcules 2004, 37, 64606471.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

31. Petraccone, V.; Pirozzi, B.; Meille, S. V. Polymer 1986, 27, 16651668.
32. Hikosaka, M.; Seto, T. Polym. J. 1973, 5, 111127.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch012

33. Radhakrishnan, J.; Ichikawa, K.; Yamada, K.; Toda, A.; Hikosaka, M.
Polymer 1998, 39, 29952997.
34. Auriemma, F.; de Ballesteros, O. R.; De Rosa, C.; Corradini, P.
Macromolecules 2000, 33, 87648774.
35. Nakamura, K.; Shimizu, S.; Umemoto, S.; Thierry, A.; Lotz, B.; Okui, N.
Polym. J. 2008, 40, 915922.
36. Bunn, A.; Cudby, M. E. A.; Harris, R. K.; Packer, K. J.; Say, B. J. Polymer
1982, 23, 694698.
37. Saito, S.; Moteki, Y.; Nakagawa, M.; Horii, F.; Kitamaru, R. Macomolecules
1990, 23, 32563260.
38. Caldas, V.; Brown, G. R.; Nohr, R. S.; Macdonald, J. G. J. Polym. Sci., Part
B: Polym. Phys. 1996, 34, 20852098.

206
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 13

Chain Mobility in Crosslinked EPDM Rubbers.

Comparison of 1H NMR T2 Relaxometry and
Double-Quantum 1H NMR
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Pieter C. M. M. Magusin,*,1 Ramona A. Orza,1,2 Victor M. Litvinov,2,3

Martin van Duin,4 and Kay Saalwchter5
1Eindhoven Univ. of Technology, P.O. Box 513, 5600 MB Eindhoven,
The Netherlands
2Dutch Polymer Institute, P.O. Box 902, 5600 AX Eindhoven,

The Netherlands
3DSM Resolve, P.O. Box 18, 6160 MD Geleen, The Netherlands
4Lanxess Elastomers Global R&D, P.O. Box 1130, 6160 BC Geleen,

The Netherlands
5Martin-Luther-Universitt Halle-Wittenberg, Institut fr Physik,

Friedemann-Bach-Platz 6, D-06108 Halle, Germany

*E-mail: [email protected]

1H NMR transverse relaxation and double-quantum (DQ)

build up have been compared for a series of EPDM grades
crosslinked with different amounts of sulfur or peroxide. The
modulus at 100% elongation of the same EPDM samples
correlates linearly to the effective transverse-relaxation and
DQ build-up rate, suggesting that both rates are proportional to
the total network density caused by chain entanglements and
crosslinks. For none of the rates, however, the linear trends
of sulfur- and peroxide-crosslinked EPDM coincide. This
suggests that sulfur- and peroxide-crosslinked EPDM networks
are qualitatively different, which is confirmed by different
distributions of residual dipolar coupling constants.

Introduction
Elastomers, like ethylene-propylene copolymers (EPM) or ethylene-
propylene-diene terpolymers (EPDM), are crosslinked to improve their

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 performance, such as elasticity, tensile strength and resistance to solvents.
Traditionally, sulfur vulcanization is applied, which, however, suffers from S-S
bond cleavage at elevated temperatures. Peroxide cure results in more thermally
stable networks (1). The mechanical properties of the rubber networks are not
only determined by the chemical crosslinks but also by the physical entanglements
between the polymer chains. Macroscopic characterization techniques, like
mechanical measurements, yield overall information of the crosslink density.
Solid-state 1H NMR gives a more detailed picture of the network density and
the heterogeneity thereof in different parts of the rubber network. The higher the
density of crosslinks and entanglements, the more restricted the polymer-chain
mobility and thus the higher the residual proton dipole coupling, as reflected
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

in the 1H NMR lineshape, transverse (T2) relaxation (2, 3) and the build up of
double-quantum coherenc (4, 5) For rigid non-oriented polymers the orientation
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

dependence of the dipolar 1H-1H coupling leads to broad 1H NMR lineshapes. In

mobile polymers the dipole coupling is partly averaged by chain motions, which
are fast compared to the static dipole coupling constant D0 (~104 Hz),. Since in
a real mobile polymer the cumulative rotational effect of chain motions is not
isotropic, a residual dipole coupling (RDC) with reduced coupling constant D
remains. This is reflected in the NMR spectrum by a narrower lineshape. The less
restricted the chain motion, the narrower the lineshape.
Although, in principle, the linewidth in a 1H NMR spectrum can be used
as a source of information about chain mobility, it tends to be polluted by other
linebroadening mechanisms, like chemical-shift heterogeneity and magnetic
susceptibility. For mobile polymers with narrow 1H NMR lines, such as
weakly crosslinked elastomers above Tg, this inhomogeneous broadening can
significantly spoil the interpretation. Therefore as an improved approach, 1H
NMR transverse relaxation, reflected in Hahn-echo decays, is often investigated
for characterization of chain mobility in polymer networks (2, 68). By reflecting
the overall anisotropy of sub-millisecond polymer motions, the T2 relaxation
time for elastomer networks is sensitive to the conformational mean position of
the network chains, which is affected by the presence of chemical and physical
network junctions. Long network chains undergo less restricted motions, resulting
in strongly averaged dipole couplings and thus long T2 values. Short network
chains have short T2 relaxation times.
For ideal polymer networks, a direct relation exists between the crosslink
density and proton transvere relaxation far above the glass transition Tg, where
chain mobility is fast and only restricted by the network junctions (6, 7, 9, 10).
In practice, however, one has to deal with network defects, such as dangling
chain-ends, chain loops and sol. Careful analysis of the T2 decay curves often
allows separating the relaxation behavior of the network defects from that of
network chains. The theory of the transverse relaxation in crosslinked rubbers
is based on the submolecule concept of network chains with both ends fixed
in a laboratory system of coordinates regardless of the origin of the junctions.
By assumption, the network chain consists of a number of statistical segments
between the network junctions. The Kuhn and Grn model of freely jointed
chains is used to calculate the conformational mean of the chain function in
elastomers with low crosslink density (network chains obeying the Gaussian

208
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 chain statistics). The weight average molar mass of polymer chains between
neighboring network junctions in crosslinked elastomers can be calculated from
the transverse relaxation rate at the high temperature plateau.
1H NMR T2 relaxation of a polymer reflects the overall anisotropy of chain

motions that are fast on the sub-millisecond timescale. It is also particularly

sensitive to intermediate motions at the millisecond timescale. If intermediate
motions are significantly present in the broad range of timescales governing
polymer dynamics, the validity of standard theories relating network density via
the anisotropy of sub-millisecond motions to T2 relaxation can be questioned
(11). As further improvement, Double-Quantum (DQ) NMR has therefore been
advocated (5). The same residual dipolar coupling that underlie T2 relaxation
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

can also be used to generate double-quantum coherences in hydrogen spin pairs.

By a strategic combination of NMR signals, the DQ NMR technique is able
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

to distinguish between a coherent build-up of double-quantum coherence, and

incoherent loss of transverse magnetization. The DQ build-up rate measured in
this way only reflects the residual dipole coupling averaged by the fast chain
motions. The effect of intermediate motions ends up in the incoherent part
and is not included in analysis in terms of network density. The advantage is
that the technique zooms in on the anisotropy of fast motions, which is more
correctly analyzed by use of the submolecule theories for rubber networks.
In the previous investigation (1214) of chemical structures and density of
crosslinks in peroxide-cured EPDM the main focus was on MAS 1H and 13C NMR
spectroscopy and static 1H NMR T2 relaxometry. In the present paper we want
to compare the outcome of T2 relaxometry and DQ build-up measurements for a
series of sulfur- and peroxide-crosslinked EPDM grades, and get a more detailed
picture of different network distributions caused by the different crosslinking
method.

Experimental Section
Materials

Amorphous EPM and EPDM co- and terpolymers with 0, 2, 4.5 and 9 wt.%
ethylidene norbonene (ENB) and respective ethylene content of 49, 54, 52, 48
wt.% were obtained from DSM Elastomers (K3200, V2727, K4802 and K4703).
For the systematic T2 relaxometry study of peroxide crosslinking illustrated in Fig.
1, the EP(D)M grades were cured with 1.25, 2.5, and 5 parts per hundred rubber
parts (phr) of bis (t-butylperoxy-i-propyl)benzene), commercially available as
Perkadox Px-14-40 (AKZO). The combined T2 relaxometry and double-quantum
build-up study described in this chapter was applied to five sulfur-vulcanized
EPDM systems with 9 wt.% ENB and varied sulfur levels of 0.7, 0.8, 1.2, 5
and 4.5 phr (as well as accelerating additives), and three peroxide-cured EPDM
systems with peroxide levels of 1.25, 2.5 and 5 phr and co-varying ENB content
of 0, 2 and 9 wt.% ENB, respectively. Further sample-preparation details can be
found elsewhere (2, 13).

209
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 1H NMR

1H NMR Transverse Relaxometry

1HNNMR transverse relaxation was measured by use of the Hahn-echo pulse

sequence on a Bruker Minispec MQ spectrometer at a proton frequency of 20
MHz. All decays of the transverse magnetization were obtained and analyzed
as described previously (6, 7). Briefly, the Hahn-echo decays were decomposed
into three relaxation components with in total 6 fit parameters. To avoid overfitting
artefacts only the weight-average relaxation rate R2 = 1/T2 of the three components
was used for the analysis. According to submolecule model of crosslinked rubbers,
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

the effective relaxation rate R2 in a polymer network is proportional to the average

density of network junctions resulting from quasi-permanent chain entanglements
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

and chemical crosslinks. For EPDM the proportionality constant, which depends
on the molecular structure and the flexibility of the polymer chains, equals 0.285
mmol s kg-1.

Multiple-Quantum 1H NMR

1Hmultiple-quantum NMR measurements were performed on a Bruker

Minispec mp20 at 90 C. The adjustment procedures and pulse sequences applied
are specified in references (5) and (15).

Results and Discussion

1H NMR Transverse Relaxometry

Fig. 1a shows the effective transverse relaxation rate R2 = 1/T2 of

ethylene-propylene copolymer (EPM) and three ethylene-propylene-diene
terpolymer (EPDM) grades, respectively, containing 2, 4.5 and 9 wt% of
5-ethylidene-2-norbonene (ENB) as the third monomer, crosslinked with a
varied amount of peroxide. The transverse relaxation rate is proportional to the
underlying total rubber network density, i.e. the sum of physical entanglements
and chemical crosslinks. Thus, for the given proportionality constant determined
by the molecular structure and flexibility of the polymer chains, the relaxation
rates can be directly translated into network densities (Fig. 1a, vertical axis
righthand side). Up to 5 phr peroxide the network density of the crosslinked
EP(D)M grades depend linearly on the peroxide level. Extrapolation to zero
peroxide yields similar chain entanglement densities in the range 0.20 - 0.24
mol/kg, This is consistent with the outcome of previous NMR and other studies
(2, 16). The higher network density after peroxide crosslinking is caused by
the formation of chemical crosslinks. The presence of ENB monomers in the
EPDM chain leads to enhanced crosslink formation. These ENB monomers are
involved in crosslinking reactions via addition of macro-radicals to the pendent
ENB unsaturation and via combination of ENB-derived allyl radicals (14).
210
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

Figure 1. (a) 1H NMR transverse relaxation rates R2 = 1/T2 and network density
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

of EP(D)M grades with 0, 2, 4.5 and 9 wt.% ENB crosslinked with varied initial
peroxide content. The three peroxide-cured EP(D)M systems that have also
been studied with DQ NMR are encircled. (b) Correlation of rheometer torque
difference of the EPM and EPDM grades with total network crosslink density
derived from 1H NMR relaxometry.

Figure 1b shows the excellent correlation between separately measured

rheometer torque difference for the complete set of peroxide-cured EPM and
EPDM grades and the total network density determined from NMR. Apparently,
the network density calculated from T2 relaxometry has a real physical meaning
and is relevant for the macroscopic properties of the crosslinked EPDM rubbers.

Multiple-Quantum 1H NMR

The left side of Fig. 2 shows DQ build-up curves of EPDM with 9 wt.%
ENB and crosslinked with sulfur amounts of 0.7, 1.5 and 4.5 phr. The results are
compared with the DQ build-up curves of three peroxide-cured EP(D)M systems
with 0, 2 and 9 wt.% ENB and respective peroxide levels of 1.25, 2.5 and 5 phr
(Fig. 2, right). The three peroxide-cured systems represent a typical selection
across the set of EP(D)Ms already characterized with T2 relaxometry (encircled
points in Fig. 1a). Both the DQ build up SDQ and the reference decay Sref were
recorded. Fig. 2 shows the SDQ, Sref and the sum S = SDQ+Sref as a function of
evolution time DQ (sum of excitation and reconversion time). For comparison
the curve fits to the Hahn-echo decays of the same samples are also shown. Sref
and SDQ show the typical decay - and build-up behavior, respectively. For an
isotropic ensemble of isolated spin pairs without relaxation and spin diffusion,
Sref and SDQ should approach each other at sufficient long evolution time. This
is indeed observed for the sulfur-vulcanized systems, but not for peroxide-cured
EPDM, where Sref is always bigger than SDQ. Apparently the latter contains a
comparably large fraction (~10 % monomer units, vide infra) of mobile chain
fragments, such as long network chains or dangling ends. The DQ build-up
rate of these fragments is low compared to the overall loss of coherence due to T2
relaxation. As a result, these mobile chain fragments will only contribute to Sref
and hardly to SDQ, which causes Sref to be systematically larger than SDQ even at
211
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 long DQ evolution time. The loss of overall coherence is reflected by the decay
of S. The observed sum-intensity decays (~5 ms) are always slower than the
corresponding Hahn-echo decays (~1 ms).
The difference between Sref and SDQ at long time is a signature of incoherent
components, of which the loss of overall coherence is faster than the build-up
of DQ coherence. To remove these components from the NMR data we have
therefore fitted a bi-exponential model to the difference intensity Sref SDQ over
the range 5 18 ms, and used this to normalize the DQ build up according to:
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Figure 2. DQ build-up intensity SDQ, reference intensity Sref (magnetization not

converted into DQ coherence) and their sum S versus the evolution time DQ, as
compared to Hahn-echo decay curves, and bi-exponential curves fitted to the tail
of Sref between 5 and 18 ms.The results shown here are (left) for EPDM with 9
wt.% ENB vulcanized with 0.7, 1.5 and 4.5 phr sulfur and (right) for EP(D)M
with 0, 2 and 9 wt.% ENB cured with 1.25, 2.5 and 5 phr peroxide, respectively.
212
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The bi-exponential correction is negligible for the sulfur-vulcanized EPDM
(Fig. 2, left), and corresponds to the removal of ~10 % of the initial intensity for the
peroxide-cured EPDM, with a tendency to decrease at increasing peroxide-level
(Fig. 2, right).
The normalized DQ build-up curves for the same sulfur- and peroxide-
crosslinked EPDM rubbers are depicted up to 5 ms in Fig. 3. As expected, the DQ
build-up becomes faster at higher sulfur and peroxide/ENB levels. The respective
shapes reflect the underlying distribution of residual dipole coupling constants.
Before giving a detailed analysis in terms of specific distribution models, we note
an interesting shape similarity among the build-up curves of the sulfur-vulcanized
EPDM systems when plotted along a logarithmic time axis (Fig. 4a). Up to
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

1.5-phr sulfur the curves coincide upon scaling the evolution time (Fig. 4c). The
same is true for peroxide-cured EPDM up to peroxide levels of 2.5 phr (Fig. 4b
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

and 4d).

Figure 3. Normalized DQ build-up curves with curve fits based on a Gaussian

and bi-Gaussian distribution of residual dipole coupling (RDC) constants for the
same sulfur- and peroxide-crosslinked EP(D)M systems as in Fig. 2. Simulated
curves for a bi-Gaussian distribution of RDC constants fit better to the observed
DQ build ups than those for a Gaussian RDC distribution.
The faster the DQ build-up, the bigger the required scaling factor, which thus
represents a model-free relative build-up rate RDQ. Arbitrarily taking the curve
of 1.25-phr peroxide-cured EPM as the reference, we have determined the time-
scaling factors of the other curves with respect to this curve. There is a systematic
213
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 shape difference between sulfur- and peroxide-cured EP(D)M, which cannot be
removed by time scaling (Fig. 5). This makes comparison of the relative build-up
rates between sulfur- and peroxide-cured EPDM slightly ambiguous, but a rough
comparison is still possible.
RDQ depends similarly on the sulfur and peroxide levels as the (weighted-
average) Hahn-echo decay rate R2 = 1/T2 (Fig. 6a and 6b). As shown in Fig. 1a,
the effective relaxation rate R2 of peroxide-cured EP(D)M varies linearly with the
peroxide level below 5 phr. In contrast, for sulfur-cured EPDM R2 and RDQ follow
a kind of saturation curve. This indicates that at sulfur levels > 4 phr the content
of third monomers becomes limiting for the crosslink density, which agrees with
earlier observations (2).
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

Since R2 and RDQ appear to follow different slopes for sulfur- and peroxide-
cured EPDM, one could wonder which of the two, R2 or RDQ, correlates best with
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

the macroscopic mechanical properties. The ideal behavior would be an NMR

parameter predicting, e.g., the modulus, independent of the applied curing method.
Fig. 6c and 6d shows plots of the modulus at 100 % elongation (M100) versus
either R2 or RDQ. For each of the two crosslinking types there is a linear correlation
between the modulus and R2 and RDQ, respectively. However, for neither R2 nor
RDQ the crosslinking-type dependence disappears. This confirms that the R2 and
RDQ difference between sulfur- and peroxide-crosslinked networks is not so much
caused by NMR artifacts related to the selected NMR parameter, but may actually
be related to the even qualitatively different types of networks formed.

Figure 4. Normalized quantum build-up curves of (a,c) sulfur-vulcanized and

(b,d) peroxide-cured EPDM crosslinked at varied levels of sulfur and peroxide,
respectively, versus (a,b) evolution time and (c,d) scaled evolution time. The
five sulfur-vulcanized EPDM rubbers all contained 9 wt.% ENB. The three
peroxide-cured EPDM systems are the same as in Fig. 2.

214
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 5. Normalized quantum build-up curves of EPDM vulcanized with 0.7 phr
sulfur and EPM cured with 1.25 phr peroxide versus the scaled DQ evolution
time.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Figure 6. (a) Effective relaxation rates R2 and (b) relative DQ build-up rates
RDQ for the same sulfur- and peroxide-crosslinked EP(D)M rubbers as in Fig.
4. (c,d) correlation between the modulus at 100% elongation M100 and (c) R2
and (d) RDQ. For clarity the data are labeled with the ENB content of the specific
EP(D)M grade, i.e. 9 wt.% for the five sulfur-vulcanized systems, and 0, 2 and 9
wt.% for the three peroxide-cured systems. The curved lines connecting the data
for the sulfur-vulcanized systems in (a) and (b) are guides for the eye only.
The similar dependence of R2 and RDQ on the sulfur- or peroxide levels
is also clear from the linear correlations between the two (Fig. 7). Sulfur-
and peroxide-crosslinked EPDM samples show different slopes, which may
215
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 reflect the difference in the underlying network topology. Peroxide can generate
radicals anywhere along the EPDM chains, whereas sulfur requires the residual
unsaturation of the third monomer to generate crosslinks. Interestingly, the
linear correlations between R2 and RDQ for sulfur-vulcanized and peroxide-cured
EPDM cross each other at RDQ = 0 and a positive value R2* ~ 300 s-1. This may
indicate that the entanglements, which determine R2 and RDQ in the absence of
chemical crosslinks, may be different. Alternatively, R2* = 1/T2BWR may represent
Bloch-Wangsness-Redfield type of relaxation contributions (17, 18) associated
with incoherent loss of transverse magnzetization:
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

where 1/T2coh denotes the decay rate of the Hahn echo under the influence of the
coherent dipolar dephasing.

Figure 7. Correlation between effective relaxation rate R2 and RDQ for the same
sulfur- and peroxide-cured EP(D)M samples as in Fig. 4.

After this essentially model-free analysis we have also analyzed the DQ build-
up curves in terms of mono- and bi-modal Gaussian distributions:

with n = 1 or 2 , respectively, and in the latter case f1 + f2 = 1, Dc(1), Dc(2), 1 and

2 the center values and widths of the two Gaussian components describing the
distribution of residual dipolar coupling (RDC) constants. This distribution can
be convoluted with the Gaussian initial behavior of the DQ build-up curves for an
isotropic ensemble of isolated spin pairs with identical residual couplings D:

216
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The resulting analytical expression for the DQ build-up curve is (15):
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Curves based on a single Gaussian distribution of RDC constants with

the maximum at 0 fit well to the initial part of the DQ build-up curves of
the peroxide-cured EPDM (Fig. 3, right hand side). At increasing peroxide
content the width of the half-Gaussian RDC distribution increases (Fig. 8b),
but the maximum stays at 0. This indicates that the chains between the network
junctions become shorter at increasing peroxide content, but there is always a
significant fraction with long chain segments (Dres=0) even at high peroxide
content. Such network structure could indicate a tendency of peroxide to form
heterogeneous networks with parts of the EPDM chains non-crosslinked. One
of the chemical pathways in the peroxide crosslinking of EPDM is the addition
of EPDM macro-radicals to the residual unsaturation of ENB monomers in the
polymer chain. This is followed by a hydrogen transfer, generating a new EPDM
macro-radical which can add to the unsaturation of another ENB monomer, or
combine with another EPDM macro-radical (13). The net result is a cluster of
two or four crosslinks. In addition, a side reaction of peroxide curing is main
chain degradation yielding a rubber network with dangling ends. Other types of
chain reactions, such as formation of multifunctional crosslinks by double-bond
polymerization, may also play a role during peroxide curing. These various
reactions lead to a broad distribution of chain mobility, and thus to a broad RDC
distribution. Similar observations were in fact previously reported for peroxide
cross-linked natural and butadiene rubbers (19).
In contrast to the peroxide-cured samples, mono-modal curves based on Eq.
5 do not fit well to the DQ build-up curves of sulfur-crosslinked EPDM (Fig. 3,
left). Bi-modal distributions consisting of a broad Gaussian centered at 0 and a
narrow Gaussian with variable center and width fit better. In this case both the
center and the width of the dominant narrow Gaussian component increases at
increasing sulfur content (Fig. 8c). The good fit with the bimodal model should not
be interpreted as indication of pronounced inhomogeneity, as the RDC distribution
is in fact dominated by the narrower component. (Note the logarithmic scale in
Fig. 8.) As support for this argument, sulfur vulcanization of natural rubber,
i.e., poly(isoprene), also yields homogeneous networks (19). Sulfur vulcanization
requires the residual unsaturation of the ENB monomer which is known to be

217
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 incorporated randomly into the EPDM chains. Sulfur vulcanization thus produces
crosslinks more homogeneously distributed along the chain.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Figure 8. (a,b) (single-)Gaussian and (c,d) bi-Gaussian RDC distributions

yielding DQ build-up curves which fit best to the experimentaldata of the (a,c)
sulfur- and (b,d) peroxide-crosslinked EPDM grades.

Conclusion
We have compared 1H NMR transverse relaxation and double-quantum
(DQ) build up for sulfur- and peroxide-crosslinked EPDM grades with varied
sulfur- and peroxide content. The network density derived from transverse
relaxation correlates well with macroscopic properties such as the rheometer
torque difference and the modulus at 100% elongation. For peroxide-cured
EPDM DQ NMR indicates a heterogeneous network with a significant amount
of dangling ends, whereas sulfur-vulcanized EPDM shows the features of a
more homogeneous polymer network. For both types of crosslinking, the DQ
build up is faster at increasing curative content, which is consistent with an
increased immobilization of the polmer chains. The DQ build-up curves of
the sulfur-crosslinked EPDM samples have similar shapes, and can be made to
coincide by time scaling. The DQ build-up curves of the peroxide-cured EPDM
218
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 grades have a different shape. Whereas the peroxide-crosslinked EPDM grades
show a linear increase of the effective relaxation rates R2 and relative DQ build-up
rates up to 4.5 phr peroxide, the corresponding rates for sulfur-vulcanized EPDM
grades level off already at initial sulfur content > 1.5 phr. Mechanical properties,
like the modulus at 100% elongation, correlate linearly with both the effective
relaxation rate R2 and the relative DQ build up rate RDQ, but for neither of the
two rates the linear trends for sulfur- and peroxide EPDM grades coincide. This
is probably a consequence of the underlying different types of rubber networks.
A difference between sulfur- and peroxide-crosslinking is that the first requires
the unsaturation of the ENB monomers which are randomly distributed along the
EPDM chain, whereas peroxide radicals create combination crosslinks anywhere
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

along the chain, as well as addition crosslinks and other possible chain reactions.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

Acknowledgments
This work was part of the research program of the Dutch Polymer Institute
(DPI Project No. 511).

References
1. van Duin, M.; Dikland, H. G. Rubber Chem. Technol. 2003, 7, 132144.
2. Litvinov, V. M.; Barendswaard, W.; van Duin, M. Rubber Chem. Technol.
1998, 71, 105118.
3. Fulber, C.; Demco, D. E.; Weintraub, O.; Blmich, B. Macromol. Chem.
Phys. 1996, 197, 581.
4. Schneider, M.; Gasper, L.; Demco, D. E.; Blmich, B. J. Chem. Phys. 1999,
111, 402415.
5. Saalwchter, K.; Ziegler, P.; Spyckerelle, O.; Haidar, B.; Vidal, A.;
Sommer, J.-U. J. Chem. Phys. 2003, 119, 34683482.
6. Litvinov, V. M.; Dias, A. A. Macromolecules 2001, 34, 40514060.
7. Litvinov, V. M. Spectroscopy of Rubbers and Rubbery Materials; RAPRA:
Shawbury, U.K., 2002; p 353.
8. Rastogi, S.; Lippits, D. R.; Peters, G. W. M.; Graf, R.; Yao, Y.; Spiess, H. W.
Nat. Mater. 2005, 4, 635641.
9. Cohen-Addad, J. P. Prog. Nucl. Magn. Reson. Spectrosc. 1993, 25, 1316.
10. Brereton, M. G. Macromolecules 1991, 24, 20682074.
11. Saalwchter, K. Macromolecules 2005, 38, 15081512.
12. Orza, R. A.; Magusin, P. C. M. M.; Litvinov, V. M.; van Duin, M.;
Michels, M. A. J. Macromol. Symp. 2005, 230, 144148.
13. Orza, R. A.; Magusin, P. C. M. M.; Litvinov, V. M.; van Duin, M.;
Michels, M. A. J. Macromolecules 2007, 40, 89999008.
14. Orza, R. A.; Magusin, P. C. M. M.; Litvinov, V. M.; van Duin, M.;
Michels, M. A. J. Macromolecules 2009, 42, 89148924.
15. Saalwchter, K. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51, 135.
16. Richter, D.; Farago, B.; Butera, R.; Fetters, L. J.; Huang, J. S.; Ewen, B.
Macromolecules 1993, 26, 795804.
219
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 17. Wangsness, R. K.; Bloch, F. Phys. Rev. 1953, 89, 728739.
18. Redfield, A. G. IBM J. Res. Dev. 1957, 1, 1931.
19. Valentn, J. L.; Posadas, P.; Fernndez-Torres, A.; Malmierca, M. A.;
Gonzlez, L.; Chass, W.; Saalwchter, K. Macromolecules 2010, 43,
42104222.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch013

220
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 14

Diffusion NMR of Polymers in Bicelles

Peter M. Macdonald*
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

Department of Chemical and Physical Sciences,

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

University of Toronto Mississauga,

Mississauga, Ontario, Canada L5L 1C6
*E-mail: [email protected]

This mini-review describes diffusion NMR studies of

polymers in bicelles using the pulsed field gradient (PFG)
NMR technique. Bicelles are first introduced and their
morphology and applications as membrane mimetics and
confinement media are briefly covered. The PFG NMR
diffusion method is then introduced, emphasizing its specific
advantages and limitations when applied to diffusion
measurements in macroscopically-oriented lamellar systems
such as magnetically-aligned bicelles. The utility of PFG
NMR diffusion measurements in bicellar model membrane
systems for examining lateral diffusion of membrane-bound,
surface-grafted poly(ethylene glycol) (PEG) molecular species
is demonstrated, focusing on those features of lateral diffusion
and / or bicelle morphology that such studies illuminate.
Further, PFG NMR diffusion studies of various molecular
weight soluble PEG species confined within the parallel-plate
geometry of magnetically aligned bicelles are reviewed, again
focusing on revelations concerning bicelle morphology. The
discussion concludes with an outline of future prospects for
diffusion NMR studies in bicelles.

Introduction
The sensitivity of the NMR experiment to diffusion was recognized virtually
from the inception of NMR spectroscopy (1). In the succeeding decades, diffusion
NMR, i.e., the use of NMR spectroscopy to detect and characterize diffusion, has
matured into a sophisticated suite of techniques and been applied widely in a host

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of physical, chemical, biological and medical investigations. Much of the impetus
for these developments stems from the recognition that knowledge of diffusion
properties often can offer key insights into the structural features of complex, or
otherwise intractable, systems.
In this mini-review the focus is diffusion NMR applications in bicelles; a
novel biomembrane mimetic system increasingly employed in NMR studies of
membrane-associating molecules (27) and as an orienting medium for NMR
structural studies of soluble proteins and nucleic acids (811). Like any good
membrane mimetic, bicelles provide an environment mirroring the lipid bilayer
structural foundation of biological membranes and thus encourage the adoption
by membrane-bound species of their proper functional form. However, it is the
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

propensity of bicelles to align in a magnetic field, and the enhanced NMR spectral
resolution that this engenders, that is the ultimate source of their ever increasing
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

popularity; a popularity evident from the ongoing effusion of reviews of bicelle

properties and applications (1223). Notwithstanding their popularity, certain
key questions regarding fundamental bicelle properties remain unresolved. And
because of their broad utility, new bicelle compositions continue to be introduced
(24), while novel bicelle properties continue to emerge (2527).
Here, after reviewing some fundamental properties of bicelles, an overview
of the diffusion NMR experiment is provided, followed by an examination of its
particulars as applied to magnetically-aligned bicelle systems in our laboratory.
Of special interest will be the insights gleaned from diffusion NMR of polymers
incorporated into bicelles: insights regarding bicelle morphology, the resolution
of controversies thus provided and the description of new properties. Of course,
diffusion NMR is but one technique and the contributions of the multitude of other
techniques that have informed the present understanding of bicellar systems will
be acknowledged as they come to bear on the subject at hand.

Bicelle Fundamentals
Bicelles are self-assembled mixtures of short-chain and long-chain
amphiphiles (28, 29), typified by the canonical combination of
dihexanoylphosphatidylcholine (DHPC) and dimyristoylphosphatidylcholine
(DMPC) first reported by Sanders and Schwonek (30). As shown in Figure 1,
DMPC, being roughly cylindrical in shape, on its own in water assembles into
a bilayer structure such that its hydrophobic tetradecyl chains are sequestered
away from water in the bilayer interior, while its polar phosphocholine headgroup
occupies the interface between the hydrophobic interior and the aqueous bathing
medium. To eliminate the exposure of hydrophobic acyl chains at the edge of the
bilayer to water, the DMPC bilayer assembly vesiculates into a macroscopically
spherical morphology. DHPC, in contrast, being roughly conical in shape due to
its shorter acyl chains, prefers a highly curved, micellar assembly. Mixtures of
DMPC and DHPC, however, assemble into single bilayer-thickness DMPC-rich
lamellar sheets stabilized by a DHPC-rich coating along the edge regions. Since
the DMPC tetradecyl acyl chains at the edge regions are masked from water by
the coating of DHPC, there is no drive to vesiculate, and the resulting morphology
222
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 is macroscopically planar. Because bicelles are self-assembled structures, their
morphology is subject to both compositional and environmental influences, and
exhibits a plasticity that is at once both beguiling and bedeviling.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 1. Schematic of the morphological evolution of bicellar self-assemblies

as a function of the molar ratio q = DMPC/DHPC. Pure DHPC (black) forms
micelles, while pure DMPC (gray) forms lamellae. Mixtures of the two form
disks at low q, or perforated lamellae at higher q, with DHPC occupying curved
edge regions while DMPC forms a planar bilayer.

The current consensus understanding of bicelle morphology is the result

of studies using a host of physical techniques including electron microscopy
(EM) (3133), fluorescence spectroscopy (34), NMR (3538), small angle X-ray
scattering (SAXS) (39) and small angle neutron scattering (SANS) (19, 25, 26,
3942). A major influence on this morphology is the molar ratio of long-chain to
short-chain amphiphiles, i.e., DMPC / DHPC, typically given the designation q.
As illustrated in Figure 1, at low q bicelles adopt a discoidal morphology with
DHPC populating the highly curved outer edge regions encircling a DMPC-rich
planar bilayer disc, as per the ideal bicelle model (43), and in close analogy to
the discoidal amphiphilic self-assemblies described and characterized by Reeves
and co-workers in the 1970s (see (44) for a review of this early work). With
increasing q, the morphology evolves towards a perforated lamellar structure
having dimensions far larger than envisaged by the ideal bicelle model and
wherein DHPC populates the highly curved inner edges of toroidal perforations
decorating the DMPC-rich bilayer lamella.
From an NMR perspective, bicelles should be regarded as soft materials,
self-assembled as the result of an accumulation of otherwise weak interactions,
223
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and retaining, therefore, considerable freedom of molecular vibrational, rotational
and translational motion. Because of the constraints imposed by the planar
bilayer structure, the motional averaging is anisotropic, so that an NMR
spectrum obtained from bicelles will be influenced by residual anisotropic,
i.e., orientation-dependent, interactions such as those arising from chemical
shift, or dipolar or quadrupolar interactions. Both the size of any such residual
orientation-dependent interactions and the distribution of orientations across the
sample will influence the NMR spectrum. When there is a random distribution
of bicelle orientations across the sample a so-called powder spectrum results,
as shown in Figure 2A for the case of the 31P NMR spectrum of DMPC /
DHPC bicelles at a temperature below the DMPC gel-to-liquid-crystalline phase
transition temperature (Tm ~ 250C) where bicelles cannot magnetically align. The
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

width of the powder spectrum = - reflects the residual 31P chemical shift
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

anisotropy difference between the extrema of the parallel, , and perpendicular,

, orientations of the phospholipid long axis (and bilayer normal) relative to
the direction of the magnetic field. The line shape observed reflects the powder
distribution of bicelle orientations between those extrema. Because DHPC resides
in highly curved edge regions, it experiences effectively isotropic motional
averaging and hence exhibits an isotropically narrow 31P NMR resonance.
Magnetic alignment of bicelles becomes possible once the temperature is
raised above the Tm of DMPC so that the self assembly becomes flexible and able to
reorganize. Such spontaneous magnetic alignment of amphiphilic self-assembled
bilayers was first reported by Lawson and Flautt in 1967 (45). It is the interaction
between the magnetic field and the magnetic susceptibility anisotropy of the
self-assembly that drives the magnetic alignment (46). Individual amphiphiles
generally possess only a small negative magnetic susceptibility anisotropy, but
when assembled into a bilayer, and aided by cooperative interactions between
neighbouring assemblies, the volume magnetic susceptibility anisotropy is
sufficient to overcome thermal fluctuations.
For DMPC / DHPC bicelles the spontaneous, so-called negative, alignment is
such that the normal to the bilayer plane lies perpendicular to the magnetic field
direction. As shown in Figure 2B, the resulting 31P NMR resonance now consists
of a narrow resonance having a frequency corresponding to , reflecting the
narrow distribution of alignments about the direction perpendicular to the magnetic
field.
In the presence of certain membrane-associating species having intrinsic
positive magnetic susceptibility anisotropy the direction of bicelle alignment is
such that the bilayer normal lies parallel to the magnetic field the so-called
positive alignment. The pioneering work of Reeves and co-workers demonstrated
that simple variations in the counterion composition of the electric double layer
could induce a flip from the negative to the positive alignment (44). For
bicelles, such a flip to the positive alignment was first reported by Prosser
and co-workers (4750) who added lanthanides which bind to the bicelle
surface. Subsequently, the same positive alignment was shown to occur upon
incorporating phosphatidylcholines containing biphenyl groups (5153), or
peptides having phenyl ring side chains (54), or lipophilic aromatic compounds
(55). As illustrated in Figure 2C, the resulting 31P NMR resonance now consists

224
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of a narrowed resonance having a frequency corresponding to , reflecting the
narrow distribution of alignments about the direction parallel to the magnetic
field.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 2. Magnetic alignment of bicelles and corresponding 31P NMR spectra.

(A) At temperatures below the Tm of DMPC the bicelles are randomly aligned.
(B) At T > Tm, spontaneous negative magnetic alignment can occur. (C) At T >
Tm upon addition of lanthanides, like Yb3+, positive magnetic alignment results.
The paramagnetic Yb3+ ions also cause NMR line broadening. Spectrum (B) was
adapted from Soong and Macdonald (159). Spectra (A) and (C) were provided
courtesy of Hannah Morales (personal communication).

It is this ability to align in a magnetic field, and the consequently enhanced

resolution relative to powder-type NMR spectra, while still manifesting
orientation-dependent interactions, that have made bicelles so popular as
biomembrane mimetics for solid state (5676) and solution state (7782) NMR
225
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 structural studies of membrane proteins and membrane associating molecules
(8395), and as an orienting medium for solution state NMR structural studies of
soluble proteins and nucleic acids (96108).
This very popularity has engendered continuing re-examinations of bicelle
morphology, new aspects of which continue to emerge, along with reports of novel
bicelle compositions with enhanced properties. For example, both SANS (25,
26) and cryo-TEM (32) methods indicate that the morphology of neutral DMPC
/ DHPC bicelles is actually a ribbon-like aggregate which converts to a smectic
phase upon addition of a negatively-charged amphiphile like phosphatidyglycerol.
A further example concerns the location of DHPC, originally believed to be
confined to highly-curved edge regions, but now recognized to be miscible with
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

DMPC and to reside partly within the planar bilayer region (109). Most recently,
bicelles having DHPC replaced with the detergent Triton X100 (24) have been
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

reported to yield assemblies with enhanced structural features.

The remainder of this mini-review will highlight the contributions of
diffusion NMR studies from our laboratory to the description of bicelle structure,
to the resolution of certain controversies regarding their morphology, and our use
of bicelles as a platform for diffusion NMR studies of lateral diffusion within
and between lipid bilayers. It will commence with a review of diffusion NMR
fundamentals, as follows.

Diffusion NMR Fundamentals

Self-diffusion is the thermally-generated centre-of-mass translational motion
occurring in the absence of a concentration gradient. For isotropic diffusion of a
spherical object of hydrodynamic radius RH in an isotropic medium of viscosity ,
the diffusion coefficient is given by the Stokes-Einstein equation,

where kB is the Boltzmann constant and T is the absolute temperature.

Most generally, however, diffusion will be anisotropic so that the diffusion
coefficient is properly described in terms of a second-rank symmetric tensor,

requiring the determination of six independent tensor components Dij to

completely specify the diffusion coefficient (110).
For the special case of lateral diffusion within a planar bilayer membrane,
or diffusion confined between two parallel plates, only two independent diffusion
tensor components need be considered: D and D corresponding to diffusion
parallel and perpendicular, respectively, to the direction of the bilayer normal (111,
226
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 112), as shown schematically in Figure 3. Consequently, the diffusion tensor is
given by,

where D is the diffusion coefficient within, or along the direction of, the plane of
the lipid bilayer membrane.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 3. The diffusion tensor in the membrane frame of reference, relative to the
laboratory frame of reference as defined by the magnetic field gradient direction,
here assumed to lie parallel to the static magnetic field.
Techniques for measuring membrane diffusion abound, with fluorescence-
based methods such as fluorescence recovery after photobleaching (FRAP),
fluorescence correlation spectroscopy (FCS) and single particle tracking (SPT)
being the most sensitive and widely employed. Obviously, fluorescence
techniques require the presence of an observable fluorophore attached to the
species of interest. NMR methods for measuring membrane diffusion do not
necessarily require specific isotropic enrichment with an NMR sensitive nucleus.
They may be divided broadly into those based on exchange in the presence of
orientation-dependent interactions and those based on the application of field
gradients to confer spatial sensitivity. The latter will be the specific focus of the
following discussion.
Pulsed field gradient (PFG) diffusion NMR was introduced more than four
decades ago by Stejskal and Tanner in their now classic article (110). Their
original spin echo PFG NMR experiment is often replaced by the stimulated echo
227
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (STE) PFG NMR sequence (113) shown schematically in Figure 4, where the
diffusion time, , is limited only by the longitudinal relaxation time, T1, rather
than the transverse relaxation time, T2, which limits the diffusion time in the spin
echo case: an advantage when, as is often the case, T1 > T2. During the evolution
period between the first and the second 900 radio frequency (rf) pulse the nuclear
spins precess in the transverse plane, acquiring a phase angle proportional to their
relevant interaction Hamiltonian, i.e., chemical shift, scalar coupling, dipolar
coupling, or quadrupolar coupling. The second 900 rf pulse places the spin
magnetization along the z-axis where it may be stored for a duration 1 < T1. The
third 900 rf pulse returns the spin magnetization to the transverse plane where a
stimulated echo forms after a time 2. The experiment is rendered sensitive to
diffusion by applying gradient pulses of amplitude g (T m-1) and duration (s)
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

during the evolution periods following the first and third rf pulse. The gradient
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

pulse induces an additional phase shift of angle i = gzi, where (rad T-1 s-1) is
the magnetogyric ratio and zi is the position of an individual nuclear spin along
the direction of the applied field gradient, assumed to lie parallel to the static
magnetic field. Thus, the first gradient pulse encodes the nuclear spins according
to their position. If no diffusion occurs during the diffusion time , then the
second gradient pulse exactly decodes, i.e., reverses, the position-dependent
phase shift due to the first gradient pulse and the stimulated echo forms with its
amplitude unchanged. If diffusion occurs then the stimulated echo amplitude is
reduced as per,

where D is the diffusion coefficient along the direction of the applied field gradient.
Experimentally, either the gradient pulse duration , or its amplitude g, or the
diffusion time , is varied progressively. For the case of simple Gaussian diffusion,
a semi-log plot of I/Io versus k=(g)2(-/3) produces a linear decay, assuming
constant 1 and 2, with a slope proportional to the diffusion coefficient.
An in-depth analysis of the theory underlying the PFG NMR experiment is
beyond the scope of this mini-review, but the reader interested in greater detail
may refer to several authoritative reviews (114117). Since its introduction, the
basic PFG NMR experiment has been greatly refined as techniques have been
introduced to surmount problems such as the presence of background gradients
(118), or convection (119), or the need for solvent suppression (120). In our
laboratory, for example, we find it particularly useful to employ a slice-selection
strategy (121123) to avoid the effects of non-linear field gradients.
For simple Gaussian diffusion the observed diffusive decay described above
is linear with k while the observed diffusion coefficient is independent of the
diffusion time . There exist many situations, however, in which the diffusive
decay is non-linear while the measured diffusion coefficient depends on the
diffusion time. Such non-Gaussian diffusion behavior typically involves complex
heterogeneous systems wherein centre-of-mass diffusion is confined due to
morphological constraints (124126). Indeed, it is in exactly such situations
228
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 that the PFG NMR diffusion experiment can prove advantageous in extracting
morphological details.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 4. The stimulated echo pulsed field gradient NMR pulse sequence
introduced by Tanner (113) for measurement of molecular diffusion. Black bars
represent radio-frequency pulses, while grey rectangles represent field gradient
pulses. See the text for further details.

Diffusion NMR in Lipid Bilayer Membranes

When applied to lipid bilayer membranes the PFG NMR experiment
encounters two fundamental difficulties. First, due to residual anisotropic
interactions, and dipolar interactions especially, the NMR resonances in lipid
bilayers are typically broad, which produces rapid loss of coherence during the
evolution times in a PFG NMR sequence and consequent loss of signal. Second,
in a normal spherical lipid bilayer vesicle there is a powder distribution of local
bilayer normal orientations and this complicates considerably the extraction of a
diffusion coefficient from any observed intensity decay.
One means to overcome the broad resonance problem is to combine magic
angle spinning (MAS) with field gradients applied co-directional with the spinning
axis (127). MAS produces line narrowing while the gradient configuration ensures
that all nuclear spins experience an identical gradients strength throughout a rotor
cycle. Gawrisch and co-workers have employed this approach to considerable
effect in their studies of the lateral diffusion of various membrane-associating
compounds (128134). In such measurements cognizance must be taken of
the need to correct the diffusive decay for the powder distribution of bilayer
orientations within the sample, and a further correction made for the radius of
curvature of the bilayers as it affects the apparent diffusion coefficient (128).
One means to overcome the distribution of bilayer orientations problem is to
macroscopically align lipid bilayers between planar solid supports. Lindblom and
Wennerstrm (135) first used this strategy in lateral diffusion measurements, while
Lindblom and radd have been its principal proponents ever since, reporting, for
229
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 example, the characterization of domain composition and properties in various
binary and ternary lipid mixtures (112, 136140). For such planar supported lipid
bilayers there is a single bilayer orientation across the entire sample. The diffusion
coefficient extracted from a plot of Ln I/Io versus k depends, however, on the
angle defining the orientation of the bilayer normal relative to the direction of
the applied field gradient as per (111, 112)
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

where the geometry is shown in Figure 3. In order to minimize the line broadening
problem, the planar aligned lipid bilayer is placed such that the bilayer normal
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

is at an orientation of 54.740 relative to the direction of the static magnetic field

(assumed to be co-linear with the field gradient direction). At the so-called magic
angle =54.740 residual line broadening effects, including in particular for 1H
NMR measurements dipolar broadening, disappear. Since for molecular species
confined within a planar lipid bilayer milieu D >> D, the cosine term in the
above equation can be ignored and the experiment measures a scaled lateral
diffusion coefficient Dzzlab =0.667 D.
Another solution to the distribution of bilayer orientations problem involves
the use of magnetically-aligned bicelles. As may be ascertained from inspection
of Figure 2, for negatively magnetically-aligned bicelles, where =900, it follows
that Dzzlab =0.667 D, so the experiment measures directly the lateral diffusion
coefficient of any NMR-observable membrane-associated species. On the other
hand, for positively magnetically-aligned bicelles, where =00, it follows that
Dzzlab = D, so the experiment measures only transbilayer diffusion, i.e., of
soluble species. The latter is a function, principally, of the number and size of
any DHPC-rich perforations in the bicellar lamellae, as will be discussed in a
subsequent section.
Neither orientation, however, eliminates the line width problem. In fact, for
the STE PFG 1H NMR sequence, the residual dipolar broadening in magnetically-
aligned bicelles is sufficient to render most lipid resonances undetectable due to
loss of coherence during the evolution periods. Nevertheless, instances abound in
which, due to inherently near-isotropic motional averaging, the resulting narrow
resonances can be used for lateral diffusion measurements.
The first demonstration of this possibility involved a PEGylated lipid
incorporated into negatively magnetically-aligned bicelles (141). PEGylated
lipids consist of a water-soluble poly(ethylene glycol), i.e., PEG, group
covalently attached to the polar head group of a lipid moiety such as
dimyristoylphosphatidylethanolamine (DMPE). The DMPE intercalates into a
lipid bilayer, leaving the PEG group displayed at the aqueous interface, thereby
effectively grafting the polymer to the lipid bilayer surface. PEGylated lipids
are used in the fabrication of stealth liposomes for drug delivery applications
(142), where the steric barrier created by the PEG coating at the liposomal
surface delays liposome clearance by the reticulo-endothelial system. In bicelles,
PEGylated lipids enhance bicelle stability, again by virtue of steric stabilization
230
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (143). Marsh et al (144) have recently reviewed the voluminous research into
the physicochemical properties of these enormously useful bio-compatibilizing
agents. Of particular virtue for the purposes of diffusion NMR measurements
is the fact that essentially all of the protons of the ethylene oxide (EO) units of
which PEG is composed are equivalent, while the EO units undergo rapid and
facile rotational isomerizations such that, overall, even rather large PEG species,
even when attached at a surface, exhibit a single, nearly isotropically narrow 1H
NMR resonance.
A typical series of STE PFG 1H NMR spectra as a function of k for DMPC
/ DHPC negatively magnetically aligned bicelles containing DMPE-PEG 2000 is
shown in Figure 5. Proton resonances from the DMPC and DHPC are more-or-
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

less absent due to their short transverse relaxation time relative to the transverse
evolution times. The remaining significant proton resonances are those from water
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

and the EO protons of the PEGylated lipid. The water resonance decays rapidly
with increasing k as expected for the rapid diffusion of such a small molecule.
The EO proton resonance decays far more slowly as expected for a larger lipid
molecule associated with the lipid bilayer.
The diffusion coefficients are extracted from the slope in plots of Ln I/Io versus
k=(g)2(-/3) of the type shown in Figure 6. The water diffusion coefficient for
the case shown in Figure 5 is 1.6 x 10-9 m2s-1 indicating significantly (30%) slower
diffusion than that of bulk water at the same temperature (145) the difference
being attributable to the population of lipid bilayer surface-bound water in fast
exchange with bulk water. Note that the water diffusion being measured here is
along the direction of the two-dimensional channels lying between the parallel
plates formed by the negatively magnetically aligned bicelles.
Figure 6 compares specifically the diffusive decays for DMPE-PEG 2000 and
free, i.e., non-hydrophobically modified, PEG 2000, both incorporated into DMPC
/ DHPC bicelles which have been negatively magnetically aligned.
DMPE-PEG 2000 becomes intercalated into the lipid bilayers while free
PEG 2000 resides within the aqueous interstices separating the bicelles. For
DMPE-PEG 2000, the intensity decay over most of the range of k values is linear,
i.e., monoexponential, indicating normal Guassian diffusion. (There is a small
contribution at lower k values from an overlapping proton resonance that decays
rapidly and is assigned to the choline methyl resonance of DHPC.) The lateral
diffusion coefficient for 1 mole% (relative to DMPC) DMPE-PEG 2000 was 1.35
x 10-11 m2s-1. Increasing the DMPE-PEG 2000 content to 2 mole% reduced its
lateral diffusion coefficient to 1.15 x 10-11 m2s-1. By comparison, the diffusion of
free PEG 2000 present in the aqueous interstices between these bicelles, although
hindered relative to PEG 2000 in bulk solution, was a factor of 30 times faster
than that of DMPE-PEG 2000 and was essentially concentration independent.

Diffusion NMR and Bicelle Morphology

Given the basic themes described above, that diffusion NMR in magnetically
aligned bicelle systems provides useful and facile membrane lateral diffusion
measurements, the focus of the remainder of this article will concern the insights
231
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 provided by such techniques regarding lateral diffusion of PEGylated lipids and
bicelle morphology.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 5. STE PFG 1H NMR spectral series obtained at 35 0C with q = 4.5

DMPC/DHPC negatively magnetically aligned bicelles containing 1 mol%
DMPE-PEG 2000 as a function of increasing gradient pulse duration in the STE
PFG NMR pulse sequence. The HDO resonance decays rapidly as expected given
the fast diffusion of water. The ethylene oxide (EO) resonance of DMPE-PEG
2000 decays much more slowly as expected for a large molecule bound to the
lipid bilayer regions of the bicelle. Other resonances are largely absent due to
their fast transverse relaxation. Adapted from Soong and Macdonald (141).

PEGylated Lipid Diffusion Obeys the Free Area Model

The free area model of lateral diffusion in membranes posits that, when the
diffusing species is of a size comparable to that of the solvent, in this case the
solvent being the lipids of which the lipid bilayer is composed, lateral diffusion
is dictated by the availability of free area within the lipid bilayer relative to the
cross-sectional area occupied by the diffusant within the lipid bilayer (146148).
The fact that the lateral diffusion coefficient of DMPE-PEG 2000 at 1 mole%
(D = 1.35 x 10-11 m2s-1) is virtually identical to values reported for DMPC at
similar temperatures using FRAP (149) demonstrates the validity of this notion,
since DMPE-PEG 2000 and DMPC will occupy virtually identical cross-sectional
areas. The size of the extra-membranous PEG moiety is entirely secondary
because it experiences only the aqueous bathing medium where the viscosity is a
fraction of that of the lipid bilayer proper. Non-hydrophobically anchored PEG
2000 diffusing between bicelles diffuses at least an order of magnitude faster.
232
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 6. STE intensity decays as a function of DMPE-PEG 2000 concentration

in STE PFG 1H NMR spectra (35 0C) of magnetically aligned DMPC / DHPC
(q = 4.5) bicelles plus various levels of either DMPE-PEG 2000 or free (i.e.
non-hydrophobically-modified) PEG 2000. Open circles: 1 mole% DMPE-PEG
2000, DPEG = 1.35 10-11 m2s-1. Open squares: 2 mole% DMPE-PEG 2000,
DPEG = 1.15 10-11 m2s-1. Closed circles: 1 mole% free PEG 2000, DPEG = 5.00
10-10 m2s-1. Closed squares: 2 mole% free PEG 2000, DPEG = 5.23 10-10
m2s-1. Adapted from Soong and Macdonald (141).

Membrane Crowding Slows Lateral Diffusion

The slower lateral diffusion measured for 2 mole% relative to 1 mole%

DMPE-PEG 2000 can be attributed directly to the onset of surface crowding
of the PEG groups. Polymer theory indicates that for unperturbed surface
grafted polymers, a situation pertinent to the PEG group of DMPE-PEG 2000
incorporated into bicelles at the 1 mole% level, the shape of the polymer may
be modeled as a half-sphere or mushroom extending out from the surface a
distance equivalent to the Flory radius, RF, and covering an area A=RF2 (150).
The Flory radius is calculated using the Flory equation, RF=N3/5a where N is the
number of monomer units and a is the length of one such monomer (151). For
DMPE-PEG 2000, where N=45 and a=3.5 (152), RF=35 and A=3850 2.
Hence, assuming DMPC occupies 60 2, complete surface coverage with PEG is
achieved at 1.6 mole% DMPE-PEG 2000. Thus, at 2 mole% DMPE-PEG 2000
overlap and entanglement of PEG groups has commenced, leading to crowding
and slower lateral diffusion. The effect is exacerbated by further increasing the
233
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 DMPE-PEG 2000 concentration or by increasing the size of the PEG group of
DMPE-PEG (153) Surface crowding is one reason postulated to explain the
observation that lateral diffusion is generally slower in real biological membranes
relative to model lipid bilayer membranes (154158).

The Bicelle Disk versus Perforated Lamellae Question

For some time controversy percolated regarding whether bicelles should be
regarded as disks encircled by a bikini of DHPC, or as lamellae perforated by
toroidal holes lined with DHPC. The PFG NMR lateral diffusion measurements
on DMPE-PEG 2000 demonstrate unequivocally that, at least under the particular
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

conditions of q = DMPC / DHPC = 4.5 and T = 35 0C, bicelle morphology is that

of a perforated lamella. Specifically, the fact that the semi-log diffusive intensity
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

decays for the PEGylated lipid as shown in Figure 5 are linear with k and are
diffusion time-independent is indicative of normal Gaussian diffusion, as opposed
to, for example, restricted diffusion. This permits a straightforward calculation of
the root-mean-square (rms) displacement undergone by the PEGylated lipid during
the experimental diffusion time via the Einstein equation, <z2>1/2=4D yielding
rms displacements on the order of 6 m for the case D = 1.35 x 10-10 m2s-1 and
= 600 ms. This distance far exceeds the radial dimensions of 300 predicted
for ideal diskoidal bicelles (43). Further, if truly diskoidal, given the operative
diffusion coefficient and diffusion time, such bicelles should have lead to restricted
diffusion behaviour in the PFG NMR diffusion experiment characterized by a slope
of zero at large k values in plots such as those in Figure 5. One must conclude that
the perforated lamellae model pertains.

DHPC Is Not Strictly Segregated to Regions of High Curvature

As originally conceived, in DMPC / DHPC bicelles DHPC was regarded
as remaining completely segregated into regions of high curvature due to its
immiscibility with DMPC. Recent 31P NMR results have called this view
into question, providing evidence that in fact DHPC does migrate from
highly-curved regions into DMPC-rich planar regions at elevated temperatures
(109). Corroborative evidence for the mixed bicelle model proposed by these
researchers is provided by diffusion NMR experiments.
Examples of the type of 31P NMR spectra obtained in our laboratory, which
conform with those reported by Triba et al (109) and prompted their mixed bicelle
model proposal, are shown in Figure 7 for the case of positively magnetically
aligned bicelles as a function of q (at constant temperature 35 0C) and of
temperature (at constant q = 4.5) (159). An essential observation is that the narrow
resonance assigned to DHPC migrates towards the broader resonance assigned to
DMPC as either temperature or q increases. The interpretation is that DHPC is in
fast exchange between highly-curved regions (isotropic chemical shift) and planar
regions (anisotropic chemical shift), so that the chemical shift observed for DHPC
reflects its equilibrium distribution between the two environments. Given certain
conservative assumptions (109, 159), the 31P NMR data may be used to calculate
an effective ratio of planar-to-edge phospholipid populations, q*=(q+*)/(1-*),
234
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 where q = DMPC/DHPC and * = DHPC / DMPC, with DHPC and DMPC being
the observed chemical shifts for DHPC and DMPC, respectively.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 7. 31P NMR spectra of (left) positively magnetically aligned bicelles

composed of 100 / 5 / 1 (mol/mol/mol) DMPC / DMPG / DMPE-PEG2000 +
DHPC in the ratio q = (DMPC+DMPG+DMPE-PEG2000) / DHPC = 4.5 as
a function of temperature, and (right) at 35 0C as a function of q as indicated.
All samples contained 25% w/w lipid/water and included ytterbium in the ratio
Yb3+ / P = 1 / 75 to achieve positive magnetic alignment. Adapted from Soong
and Macdonald (159).
If the mixed bicelle model is correct, then the number and/or size of
DHPC-dependent perforations in the lamellae must decrease as the equilibrium
distribution of DHPC shifts away from the edge, and towards the planar,
environment. One means to test this prediction is to conduct diffusion NMR
measurements of transbilayer diffusion in the same positively magnetically
aligned bicelles. Transbilayer diffusion of a small molecule such as water will
depend principally on the total surface area of perforations in the lamellar sheets
which otherwise act as barriers to diffusion. This suggests a simple relationship
between the reduced transbilayer water diffusion coefficient D/Do, (Do being the
bulk water diffusion coefficient at the corresponding temperature), and q*,

where fpore is the surface fraction of pores, Alam and Aperf are the respective areas
of lamellar and perforation regions, while B is a model dependent scaling factor
relating the number of DHPC occupying perforations to the fractional area of
perforations.
235
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 8. Reduced transbilayer water diffusion coefficient D/D0 as a function of

the proportion of phospholipids q* resident in planar versus edge regions of the
bicelles. D/D0 was measured in positively magnetically aligned bicelles as a
function of temperature at constant q = 4.5 (closed circles), and as a function of
q at a constant temperature of 35 0C (open squares). q* was calculated from 31P
NMR spectra as described in the text and is assumed to be inversely proportional
to the fractional surface area of bicellar perforations. For the line of best 1:1 fit
shown in the figure, B=0.16. Adapted from Soong and Macdonald (159).

As shown in Figure 8, there is excellent correspondence between the decrease

in transbilayer water diffusion observed via diffusion NMR and the shift of the
DHPC population away from curved regions and into planar regions as quantified
via 31P NMR. Note that transbilayer water diffusion actually decreases with
increasing temperature, an effect that is counter-intuitive, but readily explicable in
terms of the corresponding decrease in the total area of transbilayer perforations
if the mixed bicelle model interpretation of the 31P NMR observations is correct.
The fact that both the temperature dependence and, separately, the q dependence
of D/Do obey the same q* dependence strongly suggests that the common
element is indeed the fraction of DHPC available to form perforations, and
strongly supports the mixed bicelle model of Triba et al (109).
An interesting aspect of the interdependence of q* and D/Do concerns the
fate of the water displaced as the DHPC-rich perforations decrease in size and/or
number with increasing temperature (160). Specifically, one may interpret the
reduced transbilayer water diffusion coefficient as directly reflecting the fractional
surface area of lamellar perforation or pores, fpores. Thus, Figure 8 indicates that the
fractional surface area of pores decreases from roughly 60% at a temperature near
236
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the Tm of DMPC to as small as 20% at 40 0C. The volume of water thus displaced
may be estimated using

where dB is the bilayer thickness at a given temperature as determined from SANS

(160).
The same SANS measurements (160) show that in such bicelles, as in
lipid bilayers in general, dB decreases with increasing temperature due to the
corresponding increase in the probability of trans-gauche isomerizations along
the hydrocarbon chains of the phospholipids. However, SANS measurements also
demonstrate that, unlike lipid bilayers in general, for bicelles the interlamellar
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

spacing, d, increases with increasing temperature. It follows that the interstitial

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

volume between adjacent bicelles must be increasing with increasing temperature.

The change in interstitial volume may be estimated via:

The obvious question becomes whether the increase in interstitial volume

matches the volume of water displaced by annealing of lamellar perforations with
increasing temperature. In fact, we found that Vbilayer / Vintersticial = 0.93 using
values of fpore derived from diffusion NMR and values of d and dB obtained from
SANS (160).
A schematic summarizing the temperature dependence of bicelle morphology,
as derived from the combination of SANS, 31P NMR and diffusion NMR results,
is provided in Figure 9. At a lower temperature, i.e., near the Tm of DMPC,
DHPC resides almost exclusively within, and lines the inner highly-curved
edges of, toroidal perforations decorating the DMPC-rich lamellar bicelles. With
increasing temperature DHPC becomes more readily miscible with DMPC,
and is able to migrate into the DMPC-rich lamellar regions. Consequently, the
number and/or size of the toroidal perforations decrease, so that transbilayer
water diffusion decreases correspondingly. Further, the water displaced by this
temperature-dependent annealing of the perforations migrates into the interstitial
spaces, thereby increasing the interlamellar spacing.
Water being a small molecule, its transbilayer diffusion depends principally
on the fractional area of lamellar perforations. Hence, such measurements say
little regarding the specific pore size or shape, assuming the pores are large
relative to water. One possible means to investigate pore size would be to examine
transbilayer diffusion of various size probe molecules in positively magnetically
aligned bicelles. A good candidate for such probes would be PEG.

PEG Diffusion Confined Between Bicellar Parallel Plates

As a prelude to probing pore sizes in bicelles via measurements of transbilayer
PEG diffusion, we first inspected PEG diffusion in negatively aligned bicelles
(161). In this situation the adjacent bicellar lamellae will form a set of confining
two-dimensional parallel-plate channels along which PEG may diffuse. Figure 10
A shows the PEG diffusion coefficients obtained via diffusion NMR for a series of
237
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 11 different molecular weight PEG varying from 200 to 20,000 Daltons, comparing
results for PEG free in solution and confined between bicellar lamellae. For the
case of a flexible random coil polymer free in solution in a good solvent, such
as the case of PEG in water, D is expected to vary as 1/Mw. Thus, in a plot of
Log D versus Log N, where N is the degree of polymerization, i.e., the number
of EO units, the slope should be close to -0.5. As shown by the solid line, where
the slope = -0.473, this expectation is fulfilled for free PEG. From such diffusion
coefficients one calculates the hydrodynamic radius RH via the Stokes-Einstein
equation and, subsequently, the critical overlap concentration for a given polymer
molecular weight (162). The PEG concentration used here (3.33 mg /ml) was well
below the overlap concentration in all cases, so issues of polymer entanglement
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

will not complicate the interpretation of the diffusion data.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 9. Schematic cross-section through a set of stacked perforated lamellae of

DMPC (white) /DHPC (grey) bicellar mixtures, illustrating temperature effects
on morphology. Adapted from Nieh et al (160).

238
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 As show in Figure 10 A, at the low end of the molecular weight range,
bicellar-confined PEG diffusion parallels the size effects seen for free PEG, but
is reduced by roughly 30% relative to free PEG; an effect that can be attributed
to the approximately 30% viscosity increase in the interlamellar space. With
increasing PEG size, there is a progressively greater hindrance of PEG diffusion
in the confined relative to the free situation. This additional slowing of diffusion
with larger PEG must be attributed to hindrance due to confinement effects.
For diffusion confined to a two-dimensional parallel-plate geometry, the
important scalar is the ratio RH/H, where H is the separation between the confining
plates. For bicelles, H is the interlamellar spacing, i.e., the width of the aqueous
interstices between adjacent bicellar lamellae. This may be determined via
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

SANS and has been reported to equal 60 for bicelles of composition virtually
identical to those investigated here (41, 163). The free PEG diffusion data
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

mentioned above serve to demonstrate that the ratio RH/H will always be less
than unity for the range of PEG sizes investigated here. De Gennes scaling
arguments (164166) predict D/Do~(RH/H)-2/3 for strong confinement regimes,
defined by De Gennes as RH/H >1. In Figure 10 B, the normalized diffusion
coefficient, D/Do, where Do is the diffusion coefficient of free PEG in aqueous
solution, is plotted as a function of RH/H in a log-log format. It is evident that
strong confinement effects commence once RH/H > 0.4, i.e., far sooner than
predicted by scaling theory, although the -2/3 exponent is born out. This reflects
one of the challenges of scaling arguments; that of defining the cross-over point
between bulk and confined behavior. Describing confinement effects on polymer
diffusion in terms of the expected behavior of an equivalently-sized hard sphere,
as modeled by Pawar and Anderson (167) and shown as the dashed line in Figure
10 B, fails utterly to capture the quantitative details of the diffusion of the flexible
PEG in the two-dimensional confinement imposed by the bicellar lamellae.
Newer simulations using a Brownian dynamics combined with hydrodynamic
interactions approach (168) predict that strong confinement commences when
RH/H > 0.5 and that in this regime D/Do~(RH/H)-2/3. This agrees well with the
diffusion behavior observed in our laboratory for PEG confined between bicelles
(161), as well as that of double-stranded DNA confined between parallel plates as
reported by others (169).

Probing Bicellar Pore Size via Transbilayer PEG Diffusion NMR

The basic notion here is that smaller PEG will succeed in entering the
lamellar perforations in bicelles and thus be able to diffuse in the transbilayer
direction, while larger PEG will not. Consequently, the relationship between the
PEG molecular weight and its transbilayer diffusion coefficient should reflect the
size of the pores through which it must diffuse.

239
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 10. (A) Log - log plot of the PEG diffusion coefficient versus the degree of
polymerization N. Closed circles: free PEG in aqueous solution (3.33 mg/ml, 35
0C). Open circles: PEG confined between the lamellae of negatively magnetically

aligned q = (DMPC+DMPG) / DHPC = 4.5 bicelles containing 5 mol %

DMPG (25 wt% lipid, 35 0C) in the presence of 3.33 mg ml-1 PEG . The solid
line is from a regression analysis of the free PEG data and has a slope equal
to -0.473. The dashed line has the same slope as the solid line but was scaled
lower by 30% to account for the increased aqueous viscosity within the bicelle
interlamellar aqueous space. (B) Log - log plot of the reduced diffusivity D/Do,
where D is the diffusion coefficient of the bicelle-confined PEG and Do is that
of the same molecular weight PEG free in aqueous solution, versus the ratio of
the hydration radius to the confinement dimension RH/H. RH was calculated
from the diffusion coefficient of PEG free in solution as per the Stokes-Einstein
equation. H was take to be 60 (41, 163). The dashed line shows the results for
the Pawar-Anderson equation (167) simulating a hard sphere confined between
two walls. The solid line shows the de Gennes scaling prediction (166) wherein
D/Do ~ (RH/H)-2/3. Adapted from Soong and Macdonald (161).
240
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 11. Normalized diffusive intensity decays from STE PFG 1H NMR
spectra (35 0C) of various molecular weight PEG (as indicated) confined within
positively magnetically aligned bicelles. Symbols represent different diffusion
times: 210 ms (closed triangles), 410 ms (open circles), 610 ms (closed squares),
810 ms (open diamonds), 1010 ms (closed inverted triangles). The solid lines are
the fits used to obtain the diffusion coefficient. Adapted from Soong et al (170).

Figure 11 shows the diffusive decays obtained via STE PFG 1H NMR of
various size PEG in positively magnetically aligned bicelles (170). The bicelles
act as a series of stacked, equally-spaced semi-permeable barriers to diffusion
so that, in general, the apparent diffusion coefficient should be diffusion-time
dependent, i.e., different diffusion coefficients would be measured for different
experimental diffusion times. Only in Tanners asymptotic limit that the rms
displacement exceeds many multiples of the barrier spacing (roughly 100 in
this case (41)) does the apparent diffusion coefficient become diffusion-time
independent (171). As shown in Figure 11 this is indeed the case, in that for
each size PEG the decays overlap for all different experimental diffusion times.
Hence, the diffusion coefficient is obtained by simply fitting the slope in the usual
fashion. The rms displacement calculated for PEG 4600, where D 1x10-11 m2s-1,
is several microns, equivalent to several hundred barrier spacings, confirming that
Tanners asymptotic limit indeed should apply.
The profound decrease in the transbilayer diffusion coefficient with increasing
PEG size is illustrated in Figure 12, where the reduced transbilayer diffusion
coefficient, D/Do, is plotted as a function of the corresponding hydrodynamic
radius RH. Do is the diffusion coefficient of the particular PEG free in solution
(see Figure 10 A) from which RH is derived via the Stokes-Einstein equation.

241
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

Figure 12. Reduced transbilayer diffusion coefficients D/Do for various

molecular weight PEG as a function of the corresponding Stokes-Einstein radius
RH. Data are shown for three q values: 3.5 (triangles), 4.5 (circles) and 5.5
(squares). Solid curves show the best-fit Davidson and Deen model (172, 173)
predictions for the mean pore radii <Rp> as described in the text. Adapted
from Soong et al (170).
To obtain pore size information from data such as those in Figure 12 requires
an appropriate model for diffusion of flexible polymers through pores. In this
regard, the Davidson and Deen hydrodynamic model is particularly useful
(172174), in that both steric exclusion from the pore and increased hydrodynamic
drag within the pore are taken into consideration,

where is the partition coefficient accounting for the energy cost of inserting a
polymer into a pore, K-1 quantifies the increased hydrodynamic drag experienced
by the polymer inside the pore, and fpore is the surface area fraction of pores as
derived from transbilayer water diffusion measurements such as shown in Figure
8. Details regarding the derivation of the expressions for and K-1 are provided in
the original references by Davidson and Deen (172, 173) and the comprehensive
review by Deen (174). The key point is their dependence on the ratio RH/Rp, i.e.,
the size of the polymer to the pore radius.
Fitting the Davidson Deen model to the transbilayer diffusion of PEG in
positively aligned bicelles, as shown in Figure 12, yields values for the average
pore radius for three different values of the ratio q = DMPC/DHPC. For the case
q = 3.5, the best fit value of Rp is 165 , while for both q = 4.5 and 5.5, the best
fit value yielded Rp equal to 125 . These experimentally-derived values for Rp
242
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 are consistently larger than those derived from static models, whether a diskoidal
or lamellar morphology is assumed (109, 159) and the trend with changing q is
diametrically opposite. One possible factor contributing to this discrepancy is
that real bicelles undergo fluctuations, both in-plane and out-of-plane, and such
fluctuations can contribute to transbilayer diffusion (175). Static geometric models
do not capture the effects of such fluctuations.

Conclusions and Future Propsects

Bicelles are a model system for membrane diffusion studies, while
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

simultaneously providing a insights into bicelle morphology. Diffusion

measurements have resolved certain controversies, but new aspects of bicelle
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

morphological plasticity continue to emerge, more complex lipid compositions

than the canonical DMPC/DHPC mixtures continue to be a focus of interest, and
always one wishes to incorporate and investigate proteins. Thus, it seems likely
that diffusion NMR in bicelles will continue to yield results of significance.
Attaching a PEG moiety to the diffusant of interest is one means by which to
obtain a readily visible NMR resonance, which is a prerequisite for straightforward
bicelle diffusion measurements via PFG NMR diffusion. PEGylation of virtually
any chemical species of interest is now possible. The high internal mobility
of the PEG group, even when effectively grafted to the membrane surface, is
the property which renders PEG so useful for such measurements. But many
membrane associating species, proteins in particular, exhibit regions of high
internal mobility and thus may prove amenable to diffusion measurements even
without PEGylation.
Bicelle morphology is still far from fully understood, as witness the nematic
ribbon structure observed via SANS and PFG NMR for neutral, i.e., uncharged,
bicelles, where diffusion is confined to an apparently one-dimensional path (176).
The physical origin of this morphology is still obscure.
New bicelle compositions with improved alignment properties have
been reported recently (24) but morpholigcal details are so far lacking. One
composition of interest that has not yet been described in bicelles is that containing
raft-forming lipids, which form domains and are of intense interest biologically.
Indeed, domain formation in bicelles, other than segregation of DMPC and
DHPC, has yet to be demonstrated convincingly. No doubt this has to do with
the technical challenge of mixing together lipids which, by definition, prefer to
demix from the common membranous milieu.
As for membrane-associating proteins, it seems likely that diffusion NMR
in bicelles should be able readily to differentiate oligomerization states of
membrane surface-associating peptides, where surface association may catalyze
aggregation. As for the state of oligomerization of transmembrane helical
proteins, however important biologically, diffusion NMR seems unlikely to
provide useful measurements given the expected weak dependence of the lateral
diffusion coefficient on the membrane cross-sectional area in the hydrodynamic
size regime relevant to membrane proteins.

243
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Finally, PFG NMR diffusion techniques may prove useful for the study of
ligand-membrane protein association studies. Membrane proteins, such as G
protein coupled receptors, are common drug targets. Thus, on might imagine
that magnetically aligned bicelles containing reconstituted membrane proteins
could serve as a platform for diffusion NMR studies of drug-membrane protein
interactions.

Acknowledgments
The research described here was supported by grants from the Natural
Sciences and Engineering Research Council (NSERC) of Canada. The author
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

wishes to thank Hannah Morales for providing certain NMR spectra presented
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

here.

References
1. Hahn, E. L. Phys. Rev. 1950, 80, 580594.
2. Sanders, C. R., II; Landis, G. C. J. Am. Chem. Soc. 1994, 116, 64706471.
3. Sanders, C. R., II; Landis, G. C. Biochemistry 1995, 34, 40304040.
4. Struppe, J.; Komives, E. A.; Taylor, S. S.; Vold, R. R. Biochemistry 1998,
37, 1552315527.
5. Whiles, J. A.; Brasseur, R.; Glover, K. J.; Melacini, G.; Komives, E. A.;
Vold, R. R. Biophys. J. 2001, 80, 280293.
6. Glover, K. J.; Whiles, J. A.; Wood, M. J.; Melacini, G.; Komives, E. A.;
Vold, R. R. Biochemistry 2001, 40, 1313713142.
7. Marcotte, I.; Dufourc, E. J.; Ouellet, M.; Auger, M. Biophys. J. 2003, 85,
328339.
8. Tjandra, N; Bax, A. Science 1997, 278, 1111114.
9. Vold, R. R.; Deese, A. J.; Prosser, R. S. J. Biomol. NMR 1997, 9, 329335.
10. Boyd, J.; Redfield, C. J. Am. Chem. Soc. 1999, 121, 74417442.
11. Cavagnero, S.; Dyson, H. J.; Wright, P. E. J. Biomol. NMR 1999, 13,
387391.
12. Sanders, C. R.; Hare, B. J.; Howard, K. P.; Prestegard, J. H. Progr. Nucl.
Magn. Reson. Spectrosc. 1994, 26, 421444.
13. Sanders, C. R.; Oxenoid, K. Biochim. Biophys. Acta 2000, 1508, 129145.
14. Whiles, J. A.; Deems, R.; Vold, R. R.; Dennis, E. A. Bioorg. Chem. 2002,
30, 431442.
15. Nevzorov, A. A.; Mesleh, M. F.; Opella, S. J. Magn. Reson. Chem. 2004,
42, 162171.
16. Sanders, C. R.; Hoffmann, A. K.; Gray, D. N.; Keyes, M. H.; Ellis, C. D.
ChemBioChem 2004, 5, 423426.
17. De Angelis, A. A.; Jones, D. H.; Grant, C. V.; Park, S. H.; Mesleh, M. F.;
Opella, S. J. Methods Enzymol. 2005, 394, 350382.
18. Marcotte, I.; Auger, M. Concepts Magn. Reson., Part A 2005, 24A, 1737.
19. Katsaras, J.; Harroun, T. A.; Pencer, J.; Nieh, M. P. Naturwissenschaften
2005, 92, 355366.
244
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 20. Sanders, C. R.; Snnichsen, F. Magn. Reson. Chem. 2006, 44, S24S40.
21. Prosser, R. S.; Evanics, F.; Kitevski, J. L.; Al-Abdul-Wahid, M. S.
Biochemistry 2006, 45, 84538465.
22. Poget, S. F.; Girvin, M. E. Biochim. Biophys. Acta 2007, 1768, 30983106.
23. Kim, H. J.; Howell, S. C.; Van Horn, W. D.; Jeon, Y. H.; Sanders, C. R. Progr.
Nucl. Magn. Reson. Spectrosc. 2009, 55, 335360.
24. Park, S. H.; Opella, S. J. J. Am. Chem. Soc. 2010, 132, 1255212553.
25. Nieh, M. P.; Raghunathan, V. A.; Glinka, C. J.; Harroun, T. A.; Pabst, G.;
Katsaras, J. Langmuir 2004, 20, 78937897.
26. Nieh, M.-P.; Raghunathan, V. A.; Glinka, C. J.; Harroun, T. A.; Katsaras, J.
Macromol. Symp. 2005, 219, 135145.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

27. Nieh, M. P.; Raghunathan, V. A.; Pabst, G.; Harroun, T.; Nagashima, K.;
Morales, H.; Katsaras, J.; Macdonald, P. M. Langmuir 2011, 27, 48384847.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

28. Ram, P.; Prestegard, J. H. Biochim. Biophys. Acta 1988, 940, 289294.
29. Sanders, C. R.; Prestegard, J. H. Biophys. J. 1990, 58, 447460.
30. Sanders, C. R.; Schwonek, J. P. Biochemistry 1992, 31, 88988905.
31. Arnold, A.; Labrot, T.; Oda, R.; Dufourc, E. J. Biophys. J. 2002, 83,
26672680.
32. Van Dam, L.; Karlsson, G.; Edwards, K. Biochim. Biophys. Acta 2004, 1664,
241256.
33. Van Dam, L.; Karlsson, G.; Edwards, K. Langmuir 2006, 22, 32803285.
34. Rowe, B. A.; Neal, S. L. Langmuir 2003, 19, 20392048.
35. Ottinger, M.; Bax, A. J. Biomol. NMR 1998, 12, 361372.
36. Raffard, G.; Steinbruckner, S.; Arnold, A.; Davis, J. H.; Dufourc, E. J.
Langmuir 2000, 16, 76557662.
37. Sternin, E.; Nizza, D.; Gawrisch, K. Langmuir 2001, 17, 26102616.
38. Boltze, J.; Fujisawa, T.; Nagao, T.; Norisada, K.; Saito, H.; Naito, A. Chem.
Phys. Lett. 2000, 329, 215220.
39. Luchette, P. A.; Vetman, T. N.; Prosser, R. S.; Hancock, R. E.; Nieh, M. P.;
Glinka, C. J.; Krueger, S.; Katsaras, J. Biochim. Biophys. Acta 2001, 1513,
8394.
40. Nieh, M. P.; Ginka, C. J.; Krueger, S.; Prosser, R. S.; Katsaras, J. Langmuir
2001, 17, 26292638.
41. Nieh, M. P.; Ginka, C. J.; Krueger, S.; Prosser, R. S.; Katsaras, J. Biophys. J.
2002, 82, 24872498.
42. Harroun, T. A.; Koslowsky, M.; Nieh, M.-P.; de Lannoy, C. F.;
Raghunathan, V. A.; Katsaras, J. Langmuir 2005, 21, 53565361.
43. Vold, R. R.; Prosser, R. S. J. Magn. Reson., Ser. B 1996, 113, 267271.
44. Forrest, B. J.; Reeves, L. W. Chem. Rev. 1981, 81, 114.
45. Lawson, K. D.; Flautt, T. J. J. Am. Chem. Soc. 1967, 89, 54895491.
46. Boroske, E.; Helfrich, W. Biophys. J. 1978, 24, 863868.
47. Prosser, R. S.; Hunt, S. A.; DiNatale, J. A.; Vold, R. R. J. Am. Chem. Soc.
1996, 118, 269270.
48. Prosser, R. S.; Hwang, J. S.; Vold, R. R. Biophys. J. 1998, 74, 24052418.
49. Prosser, R. S.; Bryant, H; Bryant, R. G.; Vold, R. R. J. Magn. Reson. 1999,
141, 256260.
50. Prosser, R. S.; Shiyanovskaya, I. V. Concepts Magn. Reson. 2001, 13, 1931.

245
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 51. Cho, G.; Fung, B. M.; Reddy, V. B. J. Am. Chem. Soc. 2001, 123,
15371538.
52. Tan, C.; Fung, B. M.; Cho, G. J. Am. Chem. Soc. 2002, 124, 1182711832.
53. Loudet, C.; Manet, S.; Gineste, S.; Oda, R.; Achard, M.-F.; Dufourc, E. J.
Biophys. J. 2007, 92, 39493959.
54. Picard, F.; Paquet, M. J.; Levesque, J.; Blanger, A.; Auger, M. Biophys. J.
1999, 77, 888902.
55. Sanders, C. R.; Schaff, J. E.; Prestegard, J. H. Biophys. J. 1993, 64,
10691080.
56. Sanders, C. R.; Landis, G. C. Biochemistry 1995, 34, 40304040.
57. Drechsler, A.; Separovic, F. IUBMB Life 2003, 55, 515523.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

58. Williamson, P. T. F.; Zandomeneghi, G.; Barrantes, F. J.; Watts, A.; Meier, B.
H. Mol. Membr. Biol. 2005, 22, 485496.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

59. Triba, M. N.; Zoonens, M.; Popot, J. L.; Devaux, P. F.; Warschawski, D. E.
Eur. Biophys. J. 2006, 35, 268275.
60. De Angelis, A. A.; Howell, S. C.; Nevzorov, A. A.; Opella, S. J. J. Am. Chem.
Soc. 2006, 128, 1225612267.
61. Park, S. H.; De Angelis, A. A.; Nevzorov, A. A.; Wu, C. H.; Opella, S. J.
Biophys. J. 2006, 91, 30323042.
62. De Angelis, A. A.; Opella, S. J. Nat. Protoc. 2007, 2, 23322338.
63. Nevzorov, A. A.; Park, S. H.; Opella, S. J. J. Biomol. NMR 2007, 37,
113116.
64. Durr, U. H. N.; Yamamoto, K.; Im, S. C.; Waskell, L.; Ramamoorthy, A. J.
Am. Chem. Soc. 2007, 129, 66706671.
65. Park, S. H.; Opella, S. J. Protein Sci. 2007, 16, 22052215.
66. Muller, S. D.; De Angelis, A. A.; Walther, T. H.; Grage, S. L.; Lange, C.;
Opella, S. J.; Ulrich, A. S. Biochim. Biophys. Acta 2007, 1768, 30713079.
67. Mahalakshmi, R.; Franzin, C. M.; Choi, J.; Marassi, F. M. Biochim. Biophys.
Acta 2007, 1768, 32163224.
68. Durr, U. H. N.; Waskell, L.; Ramamoorthy, A. Biochim. Biophys. Acta 2007,
1768, 32353259.
69. McKibbin, C.; Farmer, N. A.; Jeans, C.; Reeves, P. J.; Khorana, H. G.;
Wallace, B. A.; Edwards, P. C.; Villa, C.; Booth, P. J. J. Mol. Biol. 2007,
374, 13191332.
70. Mahalakshmi, R.; Marassi, F. M. Biochemistry 2008, 47, 65316538.
71. Park, S. H.; Loudet, C.; Marassi, F. M.; Dufourc, E. J.; Opella, S. J. J. Magn.
Reson. 2008, 193, 133138.
72. Xu, C. Q.; Gagnon, E.; Call, M. E.; Schnell, J. R.; Schwieters, C. D.;
Carman, C. V.; Chou, J. J.; Wucherpfennig, K. W. Cell 2008, 135, 702713.
73. Daily, A. E.; Greathouse, D. V.; van der Wel, P. C. A.; Koeppe, R. E. Biophys.
J. 2008, 94, 480491.
74. Diller, A.; Loudet, C.; Aussenac, F.; Raffard, G.; Fournier, S.; Laguerre, M.;
Grelard, A.; Opella, S. J.; Marassi, F. M.; Dufourc, E. J. Biochimie 2009, 91,
744751.
75. Naito, A. Solid State NMR 2009, 36, 6776.
76. Cui, T. X.; Canlas, C. G.; Xu, Y.; Tang, P. Biochim. Biophys. Acta 2010,
1798, 161166.

246
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 77. Bocharov, E. V.; Pustovalova, Y. E.; Pavlov, K. V.; Volynsky, P. E.;
Goncharuk, M. V.; Ermolyuk, Y. S.; Karpunin, D. V.; SchulgaA.A.;
Kirpichnikov, M. P.; Efremov, R. G.; Maslennikov, I. V.; Arseniev, A. S. J.
Biol. Chem. 2007, 282, 1625616266.
78. Poget, S. F.; Girvin, M. E. Biochim. Biophys. Acta 2007, 1768, 30983106.
79. Wang, G. S. Curr. Protein Pept. Sci. 2008, 9, 5069.
80. Lee, D.; Walter, K. F. A.; Bruckner, A. K.; Hilty, C.; Becker, S.;
Griesinger, C. J. Am. Chem. Soc. 2008, 130, 1382213823.
81. Hiller, S.; Wagner, G. Curr. Opin. Struct. Biol. 2009, 19, 396401.
82. Kim, H. J.; Howell, S. C.; Van Horn, W. D.; Jeon, Y. H.; Sanders, C. R. Prog.
Nucl. Magn. Reson. Spectrosc. 2009, 55, 335360.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

83. Sanders, C. R.; Landis, G. C. J. Am. Chem. Soc. 1994, 116, 64706471.
84. Struppe, J.; Komives, E. A.; Taylor, S. S.; Vold, R. R. Biochemistry 1998,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

37, 1552315527.
85. Losonczi, J. A.; Tian, F.; Prestegard, J. H. Biochemistry 2000, 39, 38043816.
86. Yu, K.; Kang, S.; Kim, S. D.; Ryu, P. D.; Kim, Y. J. Biomol. Struct. 2001,
18, 595606.
87. Whiles, J. A.; Brasseur, R.; Glover, K. J.; Melacini, G.; Komives, E. A.;
Vold, R. R. Biophys. J. 2001, 80, 280293.
88. Glover, K. J.; Whiles, J. A.; Wood, M. J.; Melacini, G.; Komives, E. A.;
Vold, R. R. Biochemistry 2001, 40, 1313713142.
89. Marcotte, I.; Dufourc, E. J.; Ouellet, M.; Auger, M. Biophys. J. 2003, 85,
328339.
90. Lindberg, M.; Biverstahl, H.; Graslund, A.; Maler, L. Eur. J. Biochem. 2003,
270, 30553063.
91. Biverstahl, H.; Andersson, A.; Graslund, A; Maler, L. Biochemistry 2004,
43, 1494014947.
92. Ellena, J. F.; Moulthrop, J.; Wu, J.; Rauch, M.; Jaysinghne, S.; Castle, J. D.;
Cafiso, D. S. Biophys. J. 2004, 87, 32213233.
93. Anderluh, G.; Razpotnik, A.; Podlesek, Z.; Macek, P.; Separovic, F.;
Norton, R. S. J. Mol. Biol. 2005, 347, 2739.
94. Dvinskikh, S. V.; Durr, U. H. N.; Yamamoto, K.; Ramamoorthy, A. J. Am.
Chem. Soc. 2007, 129, 794802.
95. Al-Abdul-Wahid, M. S.; Neale, C.; Pomes, R.; Prosser, R. S. J. Am. Chem.
Soc. 2009, 131, 64526459.
96. Tjandra, N; Bax, A. Science 1997, 278, 11111114.
97. Vold, R. R.; Deese, A. J.; Prosser, R. S. J. Biomol. NMR 1997, 9, 329335.
98. Bax, A.; Tjandra, N. J. Biomol. NMR 1997, 10, 289292.
99. Clore, G. M.; Gronenborn, A. M.; Bax, A. J. Magn. Reson. 1998, 133,
216221.
100. Ottiger, M.; Bax, A. J. Biomol. NMR 1999, 13, 187191.
101. Fischer, M. W. F.; Losonczi, J. A.; Weaver, J. L.; Prestegard, J. H.
Biochemistry 1999, 38, 90139022.
102. Markus, M. A.; Gerstner, R. B.; Draper, D. E.; Torchia, D. A. J. Mol. Biol.
1999, 292, 375387.
103. Boyd, J.; Redfield, C. J. Am. Chem. Soc. 1999, 121, 74417442.

247
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 104. Cavagnero, S.; Dyson, H. J.; Wright, P. E. J. Biomol. NMR 1999, 13,
387391.
105. Cornilescu, G.; Bax, A. J. Am. Chem. Soc. 2000, 122, 1014310154.
106. Schwalbe, H.; Grimshaw, S. B.; Spencer, A.; Buck, M.; Boyd, J.; Dobson, C.
M.; Redfield, C.; Smith, L. J. Protein Sci. 2001, 10, 677688.
107. Ohnishi, S.; Shortle, D. Proteins: Struct., Funct., Genet. 2003, 50, 546551.
108. Luhrs, T. T.; Zahn, R.; Wuthrich, K. J. Mol. Biol. 2006, 357, 833841.
109. Triba, M. N.; Warschawski, D. E.; Devaux, P. F. Biophys. J. 2005, 88,
18871901.
110. Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288295.
111. Callaghan, P. T.; Sderman, O. J. Phys. Chem. 1983, 87, 17371744.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

112. Lindblom, G.; Ordd, G. Prog. Nucl. Magn. Reson. Spectrosc. 1994, 26,
483515.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

113. Tanner, J. E. J. Chem. Phys. 1970, 52, 25232526.

114. Stilbs, P. Prog. Nucl. Magn. Reson. Spectrosc. 1987, 19, 145.
115. Krger, J.; Pfeifer, H.; Heink, W. Adv. Magn. Opt. Reson. 1988, 12, 189.
116. Price, W. S. Concepts Magn. Reson. 1997, 9, 299336.
117. Price, W. S. Concepts Magn. Reson. 1998, 10, 197237.
118. Cotts, R. M.; Hoch, M. J. R.; Sun, T.; Marker, J. T. J. Magn. Reson. 1989,
83, 252266.
119. Momot, K. I.; Kuchel, P. W. J. Magn. Reson. 2005, 174, 229236.
120. Price, W. S.; Elwinger, F.; Vigouroux, C.; Stilbs, P. Magn. Reson. Chem.
2002, 40, 391395.
121. Antalek, B. Concepts Magn. Reson. 2002, 14, 225258.
122. Antalek, B.; Hewitt, J. M.; Windig, W.; Yacobucci, P. D.; Mourey, T.; Le, K.
Magn. Reson. Chem. 2002, 40, S60S71.
123. Park, K. D.; Lee, Y. J. Magn. Reson. Chem. 2006, 44, 887891.
124. Krger, J. Diffus. Fundam. 2005, 2, 781.
125. Mitra, P. P.; Sen, P. B.; Schwatrz, L. M. Phys. Rev. B 1993, 47, 85658574.
126. Mitra, P. P.; Sen, P. B. Phys. Rev. B 1992, 45, 143156.
127. Maas, W. E.; Laukien, F. H.; Cory, D. G. J. Am. Chem. Soc. 1996, 118,
1308513086.
128. Gaede, H. C.; Gawrisch, K. Biophys J. 2003, 85, 17341740.
129. Scheidt, H. A.; Pampel, A.; Nissler, L.; Gebhardt, R.; Huster, D. Biochim.
Biophys. Acta 2004, 1663, 97107.
130. Kimura, T.; Cheng, K.; Rice, K. C.; Gawrisch, K. Biophys. J. 2009, 96,
49164924.
131. Gaede, H. C.; Yau, W.-Y.; Gawrisch, K. J. Phys. Chem. B 2005, 109,
1301413023.
132. Scheidt, H. A.; Huster, D.; Gawrisch, K. Biophys. J. 2005, 89, 25042512.
133. Polozov, I. V.; Gawrisch, K. Biophys. J. 2004, 87, 17411751.
134. Polozov, I. V.; Bezrukov, L.; Gawrisch, K.; Zimmerberg, J. Nat. Chem. Biol.
2008, 4, 248255.
135. Lindblom, G.; Wennerstrm, H. Biophys. Chem. 1977, 6, 167171.
136. Ordd, G.; Lindblom, G. Magn. Reson. Chem. 2004, 42, 123131.
137. Ordd, G.; Lindblom, G. Spectroscopy 2005, 19, 191198.

248
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 138. Lindblom, G.; Grbner, G. Curr. Opin. Colloid Interface Sci. 2006, 11,
2429.
139. Lindblom, G.; Ordd, G. J. Dispersion Sci. Technol. 2007, 28, 5561.
140. Lindblom, G.; Ordd, G. Biochim. Biophys. Acta 2009, 1788, 234244.
141. Soong, R.; Macdonald, P. M. Biophys. J. 2005, 88, 255268.
142. Lasic, D. D.; Needham, D. Chem. Rev. 1995, 95, 26012628.
143. King, V.; Parker, M.; Howard, K. P. J. Magn. Reson. 2000, 142, 177182.
144. Marsh, D.; Bartucci, R.; Sportelli, L. Biochim. Biophys. Acta 2003, 1615,
3359.
145. Mills, R. J. Phys. Chem. 1973, 77, 685688.
146. Vaz, W. L. C.; Clegg, R. M.; Hallmann, D. Biochemistry 1985, 24, 781768.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

147. Almeida, P. F. F.; Vaz, W. L. C.; Thompson, T. E. Biochemistry 1992, 31,

67396747.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

148. Almeida, P. F. F.; Vaz, W. L. C.; Thompson, T. E. Biochemistry 1992, 31,

71987210.
149. Tocanne, J.-F.; Dupou-Czanne, L.; Lopez, A. Progr. Lipid Res. 1994, 33,
203237.
150. de Gennes, P. G. Macromolecules 1980, 13, 10691075.
151. Flory, P. Principles of Polymer Chemistry; Cornell University Press: Ithica,
NY, 1971.
152. Hristova, K.; Needham, D. Macromolecules 1995, 28, 9911002.
153. Soong, R.; Macdonald, P. M. Biochim. Biophys. Acta 2007, 1768,
18051814.
154. Jovin, T. M.; Vaz, W. L. C. Methods Enzymol. 1989, 172, 471513.
155. Tocanne, J.-F.; Dupou-Czanne, L.; Lopez, A. Progr. Lipid Res. 1994, 33,
203237.
156. Haustein, E.; Schwille, P. Annu. Rev. Biophys. Biomolec. Struct. 2007, 36,
151169.
157. Saxton, M. J.; Jacobson, K. Annu. Rev. Biophys. Biomol. Struct. 1997, 26,
373399.
158. Kusumi, A.; Nakada, C.; Richie, K.; Murase, K.; Suzuki, K.; Murakoshi, H.;
Kasai, R.; Kondo, J.; Fujiwara, T. Annu. Rev. Biophys. Biomol. Struct. 2005,
34, 351378.
159. Soong, R.; Macdonald, P. M. Langmuir 2009, 25, 380390.
160. Nieh, M. P.; Raghunathan, V. A.; Pabst, G.; Harroun, T.; Nagashima, K.;
Morales, H.; Katsaras, J.; Macdonald, P. Langmuir 2011, 27, 46384647.
161. Soong, R.; Macdonald, P. M. Langmuir 2008, 24, 518527.
162. Doi, M.; Edwards, S. F. In The Theory of Polymer Dynamics; Oxford
University Press: New York, 1986; p 141.
163. Nieh, M.-P.; Ginka, C. J.; Krueger, S; Prosser, R. S.; Katsaras, J. Biophys. J.
2002, 82, 24872498.
164. Brochard, F.; de Gennes, P. G. J. Chem. Phys. 1977, 67, 5256.
165. Daoud, M.; de Gennes, P. G. J. Phys. (Paris) 1977, 38, 8593.
166. de Gennes, P.-G. In Scaling Concepts in Polymer Physics; Cornell University
Press: Ithica, NY, 1979.
167. Pawar, Y.; Anderson, J. L. Ind. Eng. Chem. Res. 1993, 32, 743746.

249
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 168. Jendrejack, R. M.; Schwartz, D. C.; Graham, M. D.; de Pablo, J. J. J. Chem.
Phys. 2003, 119, 11651173.
169. Chen, Y.-L.; Graham, M. D.; de Pablo, J. J.; Randall, G. C.; Gupta, M.;
Doyle, P. S. Phys. Rev. E 2004, 70, 06090110609014.
170. Soong, R.; Majonis, D.; Macdonald, P. M. Biophys. J. 2009, 97, 796805.
171. Tanner, J. E. J. Chem. Phys. 1978, 69, 17481754.
172. Davidson, M. G.; Suter, U.; Deen, W. M. Macromolecules 1987, 20,
11411146.
173. Davidson, M. G.; Deen, W. M. Macromolecules 1988, 21, 34743481.
174. Deen, W. M. AIChE J. 1987, 33, 14091425.
175. Lutti, A.; Callaghan, P. T. Appl. Magn. Reson. 2008, 33, 293310.
Downloaded by PENNSYLVANIA STATE UNIV on June 20, 2012 | http://pubs.acs.org

176. Soong, R.; Nieh, M.-P.; Nicholson, E.; Katsaras, J.; Macdonald, P. M.
Langmuir 2010, 26, 26302638.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch014

250
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 15

Understanding Anisotropy, Transport, and Ion

Associations Inside Ionic Polymers
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

Jianbo Hou, Jing Li, Kyle G. Wilmsmeyer, Zhiyang Zhang,

and Louis A. Madsen*
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Department of Chemistry and Macromolecules Interfaces and Institute,

Virginia Tech, Blacksburg, VA 24061
*E-mail: [email protected]; [email protected]

Anisotropy, transport and ion associations critically determine

the performance of many ionic polymer-based materials and
devices, from fuel cells to batteries to ionic transducers. Our
group is pursuing a range of NMR studies combined with
other structural and morphological information to understand
these complex materials. We observe uniformly aligned
hydrophilic channels in different perfluorosulfonate ionomers
and aromatic-based multi-block copolymers, quantified
by 2H NMR spectroscopy and pulsed-field-gradient NMR
diffusometry. Nafion 112 exhibits biaxial alignment with its
principal and secondary axes along two orthogonal in-plane
directions. Both Nafion 212 (NRE212) and block copolymers
show uniaxial alignment with the symmetry axis perpendicular
to the membrane plane. Our further discovery of the linear
coupling between diffusion anisotropy and oritentational order
parameter for mechanically stretched Nafion 117 membranes
strongly indicates that channel dimensions, domain structure,
and defect character are unperturbed by the macroscopic
mechanical deformation. When combining ionic liquids
(ILs) with ionic polymers, ion associations result in up to 4X
faster cation than anion diffusion at lower hydration levels
and 3X slower cation diffusion at higher hydration levels.
Conversely, we use ILs as probes to map local structures
in ionic polymers. Ion diffusion coefficients exhibit strong
dependencies on diffusion time, signifying the presence of
sub-micron domain boundaries. Finally, we will discuss

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 application of electrophoretic NMR, which renders a promising
opportunity to study ion transport and associations under
electrochemical cell conditions.

Introduction

Ionic polymers manifest themselves as promising candidates for a series

of functional materials and devices, from fuel cells to water reverse osmosis
membranes to artificial muscle actuators (15). Many of these useful ionic
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

polymer membranes consist of hydrophilic and hydrophobic moieties that will

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

nanophase separate to form interconnected hydrophilic channels that allow

ion conduction and water transport (69). As illustrated in figure 1, structural
features such as orientation, size and connectivity of ionic domains and channels
within the polymer network will strongly impact water and ion transport (1015).
Understanding the role of these key factors in the transport process will give
insightful guidance on and direct targeted design of new materials.
Our group has endeavored to understand anisotropy, transport and ion
associations inside ionic polymers by combining various NMR methods (1520).
In general, our strategy relies on probing multi-scale (~10 nm 10 m) structural
features (domain sizes, periodicity, confinement, defects) in soft materials by
tracking the position and reorientation of diffusing species, such as H2O and
ions (16, 17, 21). Sample modulations, such as hydration level, molecular
weight variation and mechanical deformation strongly impact these structural
characteristics (15, 18, 2225). Through systematic NMR studies on different
samples, as well as combining NMR with other techniques like X-ray scattering
and TEM (24, 26, 27), it is our goal to bridge the gap between microscopic
and macroscopic worlds to obtain a better understanding of these complex
materials. Herein, we summarize our recent work on relevant topics using
pulsed-field-gradient NMR diffusometry and 2H NMR spectroscopy. Diffusion
and anisotropy measurements yield quantitative information regarding structural
anisotropy and motions of ions and molecules inside ionic polymers. 2H NMR
spectroscopy further provides useful information on the orientational ordering
inherited by the probe molecules (D2O, CD3OD) through quadrupolar splitting.
The combination of these measurements allow quantitative assessment of average
alignment and alignment mode of hydrophilic channels inside the polymer. When
combining ionic liquids with ionic polymers, ion associations due to strong ionic
interactions (molecular packing, electrostatic interactions) will be modulated
by variation of hydration levels. The presence of these ionic aggregates results
in up to 4X faster cation than anion diffusion inside the ionic polymer at lower
hydration levels and 3X slower cation diffusion at higher hydration levels (28)
Further implementation of electrophoretic NMR (ENMR), which measures ion
motions in the presence of an electric field, unfolds a promising prospect for
studying and understanding ion transport and ion associations (2931).

252
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Figure 1. Structural features on different length scales (a-c) inside ionic

polymers. (a) Chemical structure of Nafion. x:y represents the ratio of
hydrophilic to hydrophobic part. The short side chain is smaller than 1 nm. (b)
Ion clusters aggregate and form ion channels that are a few nms in size. (c)
Submicron hydrophilic domain bundles (blue part) are distributed within the
scaffold of polymer matrix. (d) Distribution of ionic domains impacts proton
transport inside ionic polymer membranes for fuel cell applications.

Experimental Methods
Membrane Preparation
Nafion 112, Nafion 117 (extruded) and Nafion 212 (NRE212) (dispersion-
cast) membranes, all with equivalent weight of 1100 (grams of dry membrane
per mole of sulfonate group), were obtained from E.I. DuPont (Wilmington,
Delaware) in acid form. Deuterium oxide was obtained from Cambridge Isotope
Laboratories, Inc. (Andover, Massachusetts) at 99.9% purity. The multi-block
copolymers named as BPSH-BPS (A: B) were obtained from Prof. James E.
McGraths group and the synthetic procedures have been reported elsewhere
(1). Similar to Nafion, these multi-block copolymers contain two parts, with A
and B representing the block masses of hydrophilic and hydrophobic segments
individually, and are coupled with the linkage group of decafluorobiphenyl
253
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (DFBP). For the materials presented in this work, the block masses are equal
(A = B). As reported (1), all these block copolymers were redissolved in
N-methyl-2-pyrrolidinone (NMP, Fisher), followed by solution casting onto a
clean glass substrate. The prepared films were then cast under an infrared lamp
with temperature controlled between 45-55 oC for 2 days. Membranes were
further dried in a vacuum oven at 110 oC for a whole day to remove the residual
solvent. For acidification, the membranes were boiled in 0.5 M sulfuric acid for
2h, then rinsed and boiled in deionized water. Dry membrane thicknesses were
30-40 m. Depending on membrane type and uptake, this value increased to
60-100 m upon water swelling.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

Ionic Liquid and Water Uptake Determination

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Water and ion transport in the following four ionic liquids (ILs)
were studied: 1-ethyl-3-methyl imidazolium trifluoromethanesulfonate
([C2mim][TfO]), 1-butyl-3-methyl imidazolium trifluoromethanesulfonate
([C4mim][TfO]), 1-ethyl-3-methyl imidazolium tetrafluoroborate ([C2mim][BF4])
and 1-butyl-3-methyl imidazolium tetrafluoroborate ([C4mim][BF4]). All these
ILs were purchased from Solvent Innovation GMBH (Cologne, Germany) with
purity >99%. ILs were dried in vacuum at 70oC for 48h to remove residual water
prior to diffusion measurements, and diffusion coefficients and NMR spectra were
checked for stability over time to verify that water absorption was insignificant.
Extruded Nafion 117 (N117) membranes were cut into pieces of 5 mm x 5 mm
in size, stacked together to a total mass of ~60 mg and dried in a vacuum oven
for 12h at room temperature to determine the dry membrane mass (massdry). The
membranes were then soaked with IL-D2O mixtures at different temperatures to
achieve different uptakes. The wet membranes were blotted to remove any free
surface liquid (ILs and water) and transferred to a sealed Teflon cell to equilibrate
for later diffusion measurement. All diffusion measurements were performed at
25oC after sample equilibration. To vary water content, we allowed the samples
to dry in open air while the mass of IL plus the mass of the membranes (massIL)
remained constant due to its negligible vapor pressure. Masses of wet membranes
(masswet) including water and IL were determined gravimetrically after the NMR
experiments. IL uptake and water mole ratio (water) were calculated respectively
using equations (1) and (2):

Diffusion Measured by Pulsed-Field-Gradient NMR

We applied the robust and simple pulsed-gradient stimulated echo (PGSTE)
sequence for all diffusion measurements. We measured 1H and 19F diffusion for
254
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 water and ILs using a Bruker Avance III WB 400 MHz (9.4 T) NMR equipped with
a Micro5 triple-axis-gradient microimaging probe and 8 mm double resonance (1H/
2H) RF coil. The triple axis gradients each having a maximum value of 300 G/cm

allowed for measurement of diffusion along three orthogonal directions relative to

membranes, denoted as X, Y (in-plane) and Z (through plane) (18). Verification of
orientations of the membrane stacks in the magnetic field was via using a Y-Z image
slice collected with a RARE pulse sequence (Rapid Acquisition with Relaxation
Enhancement) (32). The PGSTE sequence used a /2 pulse time of 32 s, gradient
pulse durations ranging from 1 5 ms, diffusion times ranging from 10
600 ms and 16 gradient steps with appropriate selection of maximum gradient
strength to result in 50% - 90% of NMR signal attenuation. Due to differences
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

in signal intensity, the number of scans varied from 4 to 512 to produce sufficient
signal-to-noise ratio for each data point. All parameters for the gradient have been
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

calibrated and optimized as reported earlier (15, 18).

2H NMR Spectroscopy
2H NMR experiments were performed additionally to observe orientational

ordering in ionic polymers. This method can assist in determining the alignment
modes of materials with anisotropic structures (1416). Single pulse experiments
(/2 = 20 s) were performed with repetition time of 0.5 s and number of scans
ranging from 256-1024, depending on D2O uptake. The studied materials were
soaked in D2O (99.9%, Cambridge Isotope Labs) with the measured uptake
ranging from 6 - 20 wt %. Custom-built Teflon cells with special configurations
allowed orientation of membrane stacks either vertically or horizontally with
respect to the magnetic field. These cells were placed inside the above described
imaging probe and rf coil. Relevant detailed procedures are summarized in our
previous report (15). Deuterium quadrupole splittings Q were obtained by
fitting each spectrum with two Lorentzian peaks using NutsPro software (Acorn
NMR Inc., Livermore, CA).

Results and Discussion

Diffusion Anisotropy and Channels Alignment in Ionic Polymers
Figure 2a. compares diffusion anisotropy among Nafion 112, NRE212 and
BPSH-BPS(10k-10k) multi-block copolymer at a similar water uptake. In general,
both Nafion112 and NRE212 show relatively weak diffusion anisotropy as we
observe only slightly faster diffusion in plane. In contrast, for the BPSH-BPS(10k-
10k) block copolymer, we observe equal in-plane diffusion (Dx = Dy) which is
much faster than through plane diffusion Dz. This results in substantially larger
diffusion anisotropy as compared with Nafion membranes. We further orient the
block copolymer membranes in three orthogonal directions along the magnetic
field for 2H NMR studies, as shown in Figure 2b. Experimental results suggest
uniaxial alignment in the block copolymer membranes. We observe maximum
peak splittings when the membrane plane is perpendicular to the magnetic field and
half maximum peak splittings when membrane plane is parallel to the magnetic
255
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 field. This behavior is a result of the P2(cos) dependence of the quadrupole
splitting, as described in our previous report (14, 15). Both diffusion anisotropy
and 2H NMR results well correlate with the image of lamellar structure revealed
by the TEM as shown in Figure 3a. (KD is the microtoming direction and A is the
membrane plane normal)
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Figure 2. (a) Comparison of diffusion anisotropy for Nafion 112, NRE 212 and
BPSH-BPS(10k:10k) block copolymer at a similar water uptake. Both Nafion 112
and NRE212 show weak diffusion anisotropy. In contrast, the block copolymer
exhibits large diffusion anisotropy where diffusion in plane is 3X faster than
through plane. (b) eExample 2H NMR spectra for the block copolymer.
Membranes are oriented in three orthogonal directions along the magnetic field
B0. Peak splittings show the maximum and half maximum values when membrane
plane is perpendicular and parallel to B0, respectively.

256
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 For Nafion112, in-plane diffusion Dx > Dy and Dy is close to through plane
diffusion Dz. With regard to NRE212, diffusion along two in-plane directions is
similar (Dx Dy) and faster than diffusion through plane, Dz. These results, in
correlation with 2H NMR measurements (spectra not shown), confirm the biaxial
and uniaxial aligned channels inside Nafion 112 and NRE212, respectively.
Nafion112 (Figure 3b) exhibits biaxial alignment with its principal axis along the
in-plane direction (x direction) and the secondary axis along the other orthogonal
in-plane direction (y direction). In contrast, NRE212 (Figure 3c, side view) shows
uniaxial alignment. To account for the faster diffusion in plane, the symmetry
axis, which represents the average alignment of domain bundle assemblies (blue
features), is perpendicular to the membrane plane. Figure 3d is the top view
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

of Figure 3c, where the local domain bundle axes have an average alignment
(symmetry) axis coinciding with the membrane plane normal, but the bundle axes
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

are randomly distributed within the plane.

Figure 3. (a) TEM image of BPSH-BPS (10k-10k) multiblock copolymer. KD

is along the microtoming direction and A indicates the air-membrane plane
interface normal. (b)Biaxially aligned ion channels (blue cylinders) in (extruded)
Nafion112. The principal and the secondary axes are along x and y directions,
respectively. (c) Side view of uniaxially aligned domain assemblies inside NRE
212. (d) Top view of figure (c). X-Y plane is perpendicular to the symmetry axis
direction which represents the average alignment axis of domain bundles.

257
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 We have further investigated the impact of mechanical stretching on
channel dimension, domain orientation, and defect structures in ionic polymers.
We measure water transport and anisotropy in a series of drawn Nafion 117
membranes, defining the draw ratios as L = final length (l) / initial length (l0)
(19). We notice conservation of the diffusion tensor trace (independent of
water uptake) and linear coupling between order parameter (by 2H splitting) and
diffusion anisotropy. This evidence strongly demonstrates that these domains
of channels behave like liquid crystals (3336), simply reorienting along the
uniaxial stretching direction without perturbing their dimensions and the nature
of defect structures (character, density). Based on the absolute value of order
parameter derived from X-ray data, the linear coupling between order parameter
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

and diffusion anisotropy results in a molecular aspect ratio of 1.8, which agrees
well with that of the diffusing water molecules (37).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Ion Associations of Ionic Liquid Inside Ionic Polymers

Figure 4 summarizes the diffusion ratio of Dcation/Danion for ILs inside
ionic polymer membranes as a function of hydration level. We observe
striking acceleration in anion diffusion than cation diffusion for BF4-based ILs
([C2mim][BF4], [C4mim][BF4]) at high hydration level. However, such an effect
is relatively smaller for TfO-based ILs ([C2mim][TfO], [C4mim] [TfO]). We
attribute these results to 1) the interactions of polymer-fixed sulfonate groups
and cations 2) difference in basicity and ion pairing between [TfO] and [BF4].
SO3- groups on the polymer side chain can attract cations to hinder their motions.
[TfO] is more likely to pair up with cations since it is a stronger Lewis base
than [BF4]. Thus, [BF4] can move relatively free and lead to enhanced anion
diffusion for BF4-based ILs.
On the other side, cations diffuse much faster than anions at low water content
for [C2mim]-based ILs ([C2mim][TfO], [C2mim][BF4]), where Dcation/Danion
reaches 2.5 and 3.1 for [C2mim][TfO] and [C2mim][BF4], respectively. Such an
observation contradicts the general prediction that cation should diffuse slower
within the SO3-matrix due to the strong coulombic attraction from the side chain.
Considering that our measured cation and anion diffusion coefficients are both
time independent, we rule out the difference in global viscosity experienced by
different ions. In comparison with the diffusion ratio (Dcation/Danion ~1.3-1.5) for
free state ILs, those novel phenomena strongly suggest the formation of ionic
aggregates with specific features. We further propose that anionic aggregates are
prevalent inside ionic polymers at low hydration level. Charge neutrality requires
charge balance by means of more isolated cations which results in enhanced cation
diffusion on average. For the ease of illustration, we use a simple aggregation
model as shown in Figure 5 to explain this rationale. We only consider four types
of ions: single, dipole, triple and quadrupole. Based on this model, anionic triple
ions dominate at low hydration and lead to more isolated (single) cations with
enhanced cation diffusion. Dipoles and quadrupoles do not produce imbalanced
average D ratios since cations and anions are symmetrically distributed, and
cationic triple ions will be less likely to exist as they contribute negatively to the
observed D+Average. In other words, we conclude that anionic aggregates with
258
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 reduced diffusion are prevalent at low hydration, while the presence of more
isolated cations results in faster cation diffusion on average.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Figure 4. Summary of diffusion ratio Dcation/Danion as a function of water content

water for various types and uptakes of ILs inside Nafion membranes. Anion
diffusion dramatically accelerates at higher hydration levels for BF4-based
ILs, and Danion/Dcation reaches 3 and 4 for [C2mim][BF4] and [C4mim][BF4],
respectively. At low hydration, cation diffusion becomes substantially faster
than anion diffusion for [C2mim][TfO] and [C2mim][BF4], where Dcation/Danion
reach 2.5 and 3.1 respectively.

Mapping Local Domain Information

In our recent work, we attempt to use ILs to probe structural features such
as domain structure and connectivity within the polymer network. Ionic liquids
become ideal candidates for such an investigation due to the following reasons:
1) ILs find promising applications in polymer-based materials and devices, like
batteries and soft mechanical actuator. 2) topological features strongly impact ion
transport inside polymers, thus affecting the performance of materials and devices
3) these ionic probes can sense a shorter diffusion length scale on average due
to their slow motions (~ 10-13 m2/s) inside polymers. As a result, they will more
effectively report on the local structural heterogeneity rather than the average
global structural information. Thus, one might obtain more detailed information
that correlates ion motions with structural characteristics. Our preliminary results
on ion diffusion coefficients (both Dcation and Danion) exhibit strong dependence
on both gradient pulse duration () and diffusion time (), indicating the presence
of local barriers sensed by diffusing ions. In agreement with experimental
observations, further theoretical analysis and computational results explain the
259
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 finite gradient pulse effect and time-dependent diffusion behaviors. We estimate
that the relevant length scale that characterizes the ionic domain size is a few
hundred nm based on an empirical model for diffusion in a porous network. This
study has relevance for fundamental understanding of IL-polymer interactions,
ionic polymer morphology, improvement of IL diffusion studies, and the design
of IL-polymer composite actuators.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Figure 5. Ion associations of ILs inside the ~ 2 nm hydrophilic channels of

Nafion membrane. At low water content, (a) negatively charged triple ions are
prevalent due to strong electrostatic attractions among charged species, leaving
more isolated cations and resulting in enhanced cation diffusion on average. At
high water content (b), water dramatically reduces electrostatic interactions
among cations and anions and leads to ion disassociation. Anions (especially
[BF4]) are released from local electrostatic networks and move relatively freely,
while sulfonate groups fixed to polymer side chains attract cations and thus
slow their average translational motion.

Probing Ion Motions Using Electrophoretic NMR

We also attempt to study motions of ions under an applied electric field
to quantify the degree of ion associations. Electric-field driven, also known as
electrophoretic NMR (ENMR), demonstrates to be a powerful tool for studying
motions of ions and aggregates in solutions (38). In contrast to a normal diffusion
spectrum, coherent motion of ions driven by the electric field will result in a net
phase shift rather than signal attenuation. The phase shift can be calculated via
equation (3):

Where is the nuclei gyromagnetic ratio, g is gradient strength, is gradient

duration time, is diffusion time, E is electric field strength and is ion
electrophoretic mobility. For example, Figure 6 shows a series of 19F ENMR
spectra for [BF4] in a dilute solution. From the bottom up, spectral phase shift
gradually increases with electric current, which also signifies an increment of
electric field strength. According to equation (3), for a given set of , g, ,
260
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and E values, one can estimate the ion mobility by examing the spectral phase
shift. In this case, our measured ion mobility for [BF4] is 6x10-8 m2s-1V-1. In
further combination with the Nernst-Einstein equation, which relates molecular
diffusion with mobility, these measurements will provide insight into quantifying
the degree of ion associations.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Figure 6. Application of electrophoretic NMR allows measurement of ion

mobility. Coherent ion motions result in an NMR spectral phase shift as shown
above. From bottom to top, the increment in phase shift is associated with the
increment of electric field strength, as indicated by the electric current value. In
a dilute solution, the measured ion mobility for [BF4] is 6x10-8 m2s-1V-1.

Conclusion
We have quantified the alignment of hydrophilic channels in a series of
ionic polymers, from Nafion 112 to Nafion 212 to aromatic-based multi-block
copolymers using 2H NMR spectroscopy and pulsed-field-gradient NMR
diffusometry. We conclude that Nafion 112 exhibits biaxial alignment mode
whereas Nafion 212 (NRE212) and block copolymers both uniaxially align along
their symmetry axes, which are perpendicular to the membrane plane. We further
revealed the linear coupling between diffusion anisotropy and oritentational order
parameter for mechanically stretched Nafion 117 membranes. These results
strongly demonstrate that macroscopic mechanical deformation has no impact
on channel dimensions, domain structure, and defect character. Combining
ionic liquids (ILs) with ionic polymers, we observe that ion associations result
in up to 4X faster cation than anion diffusion at lower hydration levels and 3X
slower cation diffusion at higher hydration levels. When we use ILs as probes to
map local structures in ionic polymers, ion diffusion coefficients exhibit strong
261
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 dependencies on diffusion time, signifying the presence of sub-micron domain
boundaries. Finally, we also show that measurement of coherent ion motions via
electrophoretic NMR renders a promising prospect to study ion transport and
quantify ion associations.

Acknowledgments
We would like to graciously thank Professor James E. McGrath and Dr.
Hae-Seung Lee at Virginia Tech for providing block copolymer samples, Professor
Robert B. Moore and Dr. Abhishek Roy for providing N112, NRE212 and N117
membrane samples and Prof. Qiming Zhang at Pennsylvania State University for
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

providing the IL samples and relevant information and discussions. We also thank
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

Dong Wang and Prof. James R. Heflin for helpful discussion on ion associations
and actuator bending test. This material is based upon work supported by the
National Science Foundation under award number DMR 0844933 and CBET
0756439. Any opinions, findings and conclusions or recommendations expressed
in this material are those of the author(s) and do not necessarily reflect the
views of the National Science Foundation (NSF). Experiments at PAL were
supported in part by the Ministry of Education, Science and Technology of
Korea and POSTECH. This material is based upon work supported in part by the
U.S. Army Research Office under Grant W911NF-07-1-0452 Ionic Liquids in
Electro-Active Devices (ILEAD) MURI. Acknowledgment is made to the Donors
of the American Chemical Society Petroleum Research Fund for partial support
of this research and to Virginia Tech for startup funds.

References
1. Lee, H. S.; Roy, A.; Lane, O.; Dunn, S.; McGrath, J. E. Polymer 2008, 49,
715.
2. Duncan, A. J.; Leo, D. J.; Long, T. E. Macromolecules 2008, 41, 7765.
3. Petersen, R. J. J. Membrane Sci. 1993, 83, 81.
4. Grot, W. Chem. Ing. Tech .1978, 50, 299.
5. Park, J. K.; Moore, R. B. ACS Appl. Mater. Interfaces 2009, 1, 697.
6. Gierke, T. D.; Munn, G. E.; Wilson, F. C. J. Polym. Sci., Part B: Polm. Phys.
1981, 19, 1687.
7. Schmidt-Rohr, K.; Chen, Q. Nat. Mater. 2008, 7, 75.
8. Mauritz, K. A.; Moore, R. B. Chem. Rev .2004, 104, 4535.
9. Roy, A.; Lee, H. S.; McGrath, J. E. Polymer 2008, 49, 5037.
10. Park, M. J.; Balsara, N. P. Macromolecules 2010, 43, 292.
11. Ohkubo, T.; Kidena, K.; Ohira, A. Macromolecules 2008, 41, 8688.
12. Rubatat, L.; Rollet, A. L.; Gebel, G.; Diat, O. Macromolecules 2002, 35,
4050.
13. Li, J.; Wilmsmeyer, K. G.; Hou, J. B.; Madsen, L. A. Soft Matter 2009, 5,
2596.
14. Li, J.; Wilmsmeyer, K. G.; Madsen, L. A. Macromolecules 2008, 41, 4555.
15. Li, J.; Wilmsmeyer, K. G.; Madsen, L. A. Macromolecules 2009, 42, 255.
262
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 16. Deloche, B.; Samulski, E. T. Bull. Am. Phys. Soc. 1981, 26, 327.
17. Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 288.
18. Hou, J. B.; Li, J.; Madsen, L. A. Macromolecules 2010, 43, 347.
19. Li, J.; Park, J. K.; Moore, R. B.; Madsen, L. A. Nat. Mater. 2011 Submitted.
20. Rankothge, M.; Haryadi; Moran, G.; Hook, J.; Vangorkom, L. Solid State
Ionics 1994, 67, 241.
21. Price, W. S. Concept Magn. Reson. 1997, 9, 299.
22. Zawodzinski, T. A.; Springer, T. E.; Davey, J.; Jestel, R.; Lopez, C.;
Valerio, J.; Gottesfeld, S. J. Electrochem. Soc. 1993, 140, 1981.
23. Lee, M.; Park, J. K.; Lee, H. S.; Lane, O.; Moore, R. B.; McGrath, J. E.;
Baird, D. G. Polymer 2009, 50, 6129.
Downloaded by STANFORD UNIV GREEN LIBR on June 21, 2012 | http://pubs.acs.org

24. Lin, J.; Wu, P. H.; Wycisk, R.; Pintauro, P. N.; Shi, Z. Q. Macromolecules
2008, 41, 4284.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch015

25. van der Heijden, P. C.; Rubatat, L.; Diat, O. Macromolecules 2004, 37, 5327.
26. Cable, K. M.; Maurtiz, K. A.; Moore, R. B. Chem. Mater. 1995, 7, 1601.
27. Weber, A. Z.; Newman, J. J. Electrochem. Soc. 2004, 151A311.
28. Hou, J. B.; Zhang, Z. Y.; Madsen, L. A. J. Phys. Chem. B 2011, 115, 4576.
29. Hallberg, F.; Weise, C. F.; Yushmanov, P. V.; Pettersson, E. T.; Stilbs, P.;
Furo, I. J. Am. Chem. Soc. 2008, 130, 7550.
30. He, Q. H.; Hinton, D. P.; Johnson, C. S. J. Magn. Reson. 1991, 91, 654.
31. Saarinen, T. R.; Johnson, C. S. J. Am. Chem. Soc. 1988, 110, 3332.
32. Hennig, J.; Nauerth, A.; Friedburg, H. Magn. Reson. Med. 1986, 3, 823.
33. Basser, P. J.; Mattiello, J.; Lebihan, D. J. Magn. Reson., Ser. B 1994, 103,
247.
34. Yin, Y. G.; Zhao, C. H.; Kuroki, S.; Ando, I. Macromolecules 2002, 35, 2335.
35. Kirkwood, J. G. J. Polym. Sci. 1954, 12, 1.
36. Doi, M.; Edwards, S. F. The Theory of Polymer Dynamics; Oxford University
Press: Oxford, U.K., 1988.
37. Hess, S.; Frenkel, D.; Allen, M. P. Mol. Phys. 1991, 74, 765.
38. Hallberg, F.; Furo, I.; Stilbs, P. J. Am. Chem. Soc. 2009, 131, 13900.

263
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 16

13C-NMR
Observed Conformations and
Motions of Neat Liquid and Crystalline
n-Hexatriacontane and as a Guest in the
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Narrow Channels of Its Inclusion Compound

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Formed with -Cyclodextrin

M. A. Hunt,1 S. Villar-Rodil,2 M. A. Gomez-Fatou,3 I. D. Shin,4 F.
C. Schilling,5 and A. E. Tonelli*,6
1Oak
Ridge National Laboratory, Oak Ridge, TN 37831
2National
Institute of Coal (INCAR), CSIC, Oviedo, Spain
3Department of Polymer Physics and Engineering, Institute of Science and

Technology of Polymers (ICTP), C.S.I.C., Madrid, Spain

4College of Pharmacy & Health Sciences, Campbell University,

Buies Creek, NC 27506

537433 South Ocotillo Canyon Dr., Tucson, AZ 85739
6Fiber & Polymer Science, North Carolina State University,

Campus Box 8301, Raleigh, NC 27695-8301

*E-mail: [email protected]

A non-covalently bonded inclusion compound (IC) was formed

between a 36 carbon guest n-alkane, hexatriacontane (HTC),
and the host -cyclodextrin (-CD) and observed by solid-state
13C-NMR, as were neat HTC in both its liquid melt and in

its crystallline solid. Based on the number and frequencies of

observed 13C resonances, HTC in its neat crystals is restricted
to the fully extended all trans conformation, while in the melt
HTC chains are experiencing rapid inter-conversions between
all possible conformations containing trans and gauche bonds.
The spin-lattice relax-ation times, T1(13C), observed for interior
CH2 carbons in crystalline HTC are ~500 s and for the molten
liquid are 1-3 s. In the crystal HTC chains experience virtually
no ~100 MHz motions, while molten HTC chains are efficiently
moving at both this and much higher frequencies, leading to

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 an over two orders of magnitude decrease in HTC spin-lattice
relaxation times. While the narrow channels (~0.5 nm) of
its -CD-IC largely restrict the HTC chains to the all-trans
conformation, the T1(13C) relaxation times of its interior CH2
carbons range from 1-4 s at temperatures from -30 to 85 C. In
other words, even though the conformations of HTC chains in
the narrow -CD-IC channels are severely restricted compared
to those of neat molten HTC chains, they are also experiencing
efficient ~100 MHz motions that lead to virtually identical
T1(13C)s in both environments. Here we attempt to identify
similarities and differences between the types, length-scales,
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

and cooperativities of the motions experienced by HTC chains

in the neat melt and in the narrow crystalline channels of its
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

-CD-IC.

Introduction
The conformations and motions experienced by polymer chains in various
environments are intimately related to the physical behaviors manifested by
the materials made from them. Because NMR spectroscopy can be used to
simultaneously probe both the conformations and motions of polymers, their
connections to polymer properties can be drawn by NMR examination of samples
whose constituent chains reside in distinct and well defined environments, such
as those in mobile liquid and rigid crystalline polymers.
In addition, through formation of non-covalently bonded inclusion
compounds (ICs) between guest polymer chains and certain small molecule hosts,
polymer chains can be separated, extended, and confined to occupy the narrow
channels of the resultant IC or clathrate. This is illustrated in Figure 1 with the
IC-host -cyclodextrin (-CD), a cyclic oligosaccharide containing 6 glucose
units and resulting from the enzymatic degradation of starch. The diameter of the
-CD cavity and the resultant channel in its columnar polymer--CD-IC crystal
is very narrow at ~0.5 nm, similar to the inter-chain separation found in bulk
polymer crystals. The extensions and geometrical constraints experienced by
individual chains are similar in both of these solid-state environments. However,
what may distinguish them are differences between the motional constraints
imposed by regularly and closely packed chains in the bulk crystal and the host
channels in -CD-ICs, which are necessarily highly cooperative in the first case
and may not be in the second.
Using 13C-NMR, here we probe and compare the conformations and motions
of the 36 carbon n-alkane hexatriacontane (HTC), in its bulk crystal, in the melt,
and as a guest in the narrow channels of its IC formed with -CD. The observed
resonance frequencies, (13C), are used to obtain conformational information,
while spin-lattice relaxation times, T1(13C), are measured to characterize the
natures of the ~100 MHz motions experienced by HTC in each of these
environments.

266
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Figure 1. Structure of -CD and a schematic of a polymer--CD-IC.

Experimental Section
Materials

-CD was purchased from Cerestar in powder form. Hexatriacontane (HTC,

n-C36H74) and a xylene mixture were obtained from Aldrich and used without
further purification. Formation of the HTC--CD-IC was described previously
(1) and is schematically illustrated in Figure 2. Approximately 62 mg of HTC
was placed in the bottom of a test tube. Then 1 g of -CD was carefully placed
in the test tube above the HTC. Approximately 6 mL of water was poured slowly
down the side of the test tube. The test tube was sealed with a rubber septum and
placed in an oil bath at 90 C. The HTC melted and formed a thin layer on top of
the water, while the -CD dissolved in the water. The HTC--CD-IC precipitate
began to form immediately. After three days at 90 C, the mixture was slowly
cooled to room temperature by turning off the bath heater. The precipitate was
then washed with 100 mL of xylenes at 100 C and water at room temperature and
allowed to dry overnight at 50 C.

Methods

A Bruker AVANCE 500 MHz NMR Spectrometer with an Oxford Narrow

Bore Magnet was used to measure 13C T1s of molten HTC using the inversion
recovery pulse sequence. The carbon frequency was 125.77 MHz and DMSO-d6
was used for spin-locking at approximately 85 C. Chemical shifts were referenced
to DMSO assuming it resonates at 39.51 ppm vs TMS.

267
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Figure 2. Formation of HTC--CD-IC.

High-resolution solid-state 13C-NMR experiments were performed at ICTP
(Madrid, Spain) on a Bruker AvanceTM 400 spectrometer (Bruker Analytik GmbH
Karlsrube, Germany) equipped with a Bruker UltrashieldTM 9.4T (13C frequency of
100.62 MHz), 8.9 cm vertical-bore super-conducting magnet. Cross-polarization
and magic angle spinning (CP-MAS) NMR spectra were acquired wth a standard
Bruker broad band MAS probe, spinning at 10 KHz. In all cases, high-power
proton decoupling was used. The spectra were acquired with 1 ms CP contact
time, 5 s recycle delay, and 1920 transients. The NMR spectra were processed
and analyzed with the software package XWIN-NMRTM by Bruker. All free-
induction decays were subjected to standard Fourier transformation with 10 Hz
line broadening and phasing. The chemical shifts were externally referenced to
adamantane.
Solid-state spin-lattice relaxation times, T1(13C), were measured for
crystalline HTC and the HTC--CD-IC by inversion recovery experiments using
the Torchia pulse sequence (2). Typically 800 transients were used for each point
and 20-30 points were used for each measurement, covering the decay of the
signal until it reached 30-40% of its maximum value. The NMR signal intensities
of carbon nuclei observed for different time delays were used to calculate the
spin-lattice relaxation times by fitting the signal intensity data to a single or
double exponential decay function. The criterion for an acceptable fit was set at
95% of residuals within the range 0.03.

Results and Discussion

HTC Conformations
Figure 3 presents the 125.8 MHz 13C- NMR spectrum of neat liquid HTC
measured at 85 C, where 5 distinct resonances are observed. Their assignment
268
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 may be made by reference to Figure 4, which outlines the substituent effects (3)
experienced by the HTC nuclei. The most intense resonance, at 28.75 ppm vs
TMS, is produced by the internal methylene carbons C5-32, while the much smaller
resonance slightly upfield at 28.37 ppm belongs to the C4,33 nuclei. Even though
each of the C4-33 carbons have 2, 2, and 2 carbon substituents, carbons C4 and
C33 experience increased shielding from one of their -substituents (3), C1 or C36,
respectively, because the C2C3 and C34C35 bonds are expected (4, 5) to have a
slightly higher gauche population than the other internal CC bonds.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Figure 3. 125.8 MHz 13C- NMR spectrum of neat liquid HTC measured at 85 C
(ppm vs TMS). It should be noted that the chemical shift scale above assumes that
(DMSO) vs TMS observed at 85 C remains at 39.51ppm, the value observed
at room temperature.

The small resonance ~2.5 ppm downfield from that of the internal C5-32
carbons belongs to C3,34, which experience only a single shielding -substituent.
Finally the two small resonances at 21.60 and 12.81 ppm belong to the C2,35
methylene and C1,36 methyl carbons, respectively, which have 2,1,1 and
1,1,1 substituents.

C1,36, C2,35, C3,34, and C4-33 have ,,; 2,,; 2,2,; and 2,2,2 substituents,
respectively. ,-substituents deshield ~ 9-10 ppm, while -substituents produce
a maximum shielding of ~ -5 ppm (3) (See Figure 4).

269
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 4. HTC conformations and 13C nuclear shielding. 13C nuclei separated by
three bonds in the shielding (~ -5 ppm) gauche (left) and non-shielding (0 ppm)
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

trans (right) conformations.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Figure 5. 100.6 MHz CP-MAS/DD 13C- NMR spectra of neat crystalline HTC at
25 C and HTC--CD-IC at 25 and 85 C.
270
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 5 provides a comparison of the expanded CP-MAS spectra of neat
crystalline HTC recorded at 25 and HTC--CD-IC recorded at 25 and 85 C,
showing only HTC resonances. All resonances observed in neat crystalline HTC
(Figure 5) appear downfield from their positions in the liquid spectrum (Figure
3). The downfield shifts are either ~ 2 or ~ 4 ppm, depending on whether or not
a particular carbon has a single (C1,2,3,34,35,36) or a pair (C4-33) of -substituents,
reflecting the fact that the internal CC bonds in liquid HTC are ~40% gauche
(4) which provides a shielding of (1 or 2)x0.4x(-5 ppm) ~ -(2 or 4 ppm).
The only notable differences between the CP-MAS spectra of HTC in its neat
crystal and in its -CD-IC is the splitting of the C2,35 methylene resonances at 24.8
ppm in crystalline HTC into two resonances at ~24.8 and ~22.8 ppm and the small
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

resonance at 30.5-31.0 ppm, some -2 to -2.5 ppm upfield from the main resonance
of the internal methylenes at ~33 ppm in the spectrum of HTC--CD-IC. We assign
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

this latter peak to C5,32, because they are to C2,35. In neat crystalline HTC, all
internal CC bonds are trans, so the two resonances observed for the C2,35 and C5,32
methylenes of HTC residing in the narrow -CD-IC channels strongly suggest two
conformational populations for the C3C4 and C33C34 bonds: i) rigidly trans (24.8
and 33 ppm) and ii) trans and gauche conformations inter-converting rapidly
(22.8 and 30.5-31 ppm) on the ~100 MHz frequency scale, like the internal CC
bonds in liquid HTC. This explains the two distinct conformational environments
and resultant resonance frequencies for the C2,5,32,35 carbons.
It is important to note that the HTC methyl carbon region is absent in all three
expanded spectra in Figure 5, but a single methyl resonance at ~15 ppm appears
in each full spectrum. This is consistent with C2C3 and C34C35 bonds that are
rigidly trans, similar to all remaining CC bonds between C4 and C33. Apparently,
only the C3C4 and C33C34 bonds are exhibiting significant contents of rapidly
inter-converting gauche and trans conformations, even though the C2C3 and
C34C35 bonds might also have been expected to, since they are even closer to the
HTC termini.

HTC Dynamics

The spin-lattice relaxation times, T1( 13C), observed for molten HTC are
presented in Table 1, and those measured for neat crystalline HTC and HTC
residing in the narrow channels of its -CD-IC are compared in Table 2. The
T1(13C)s observed for interior CH2 carbons in neat crystalline HTC are ~500
s and for the molten liquid HTC are 1-2 s, which are consistent with previous
reports (6, 7). The interior CH2 carbons of HTC chains in the narrow channels
(~0.5 nm) of its -CD-IC exhibit T1(13C) relaxation times of 1-4 sec. over the
temperature range -30 to 85 C. In the crystal HTC chains experience virtually
no ~100 MHz motions, while molten HTC chains are efficiently moving at this
frequency, leading to more than a two orders of magnitude decrease in HTC
spin-lattice relaxation times. Even though the conformations of HTC chains in
the narrow -CD-IC channels are severely restricted compared to those of neat
molten HTC chains, they are also experiencing efficient ~100 MHz motions that
lead to virtually identical T1(13C)s in both environments.

271
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. Spin-lattice relaxation times, T1( 13C), for liquid HTC at 85 C
HTC Carbon = (13C), ppm T1( 13C), s
C1,36 = 12.81 3.4
C2,35 = 21.60 2.6
C4,33 = 28.36 1.7
C5-32 = 28.75 1.2
C3,34 = 30.95 2.1
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Table 2. Spin-lattice relaxation times, T1( 13C), observed for HTC (C5-32) in
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

the neat liquid and crystal and in the channels of its -CD-IC
Sample T1(13C), s Environment Conformation
IC (-30 C) 1.2 -CD-IC-Channel nearly all trans
IC (25 C) 1,3.6
IC (85 C) 2.6
Neat (25 C) 505 Bulk crystal all trans
Neat (85 C) 2.1 Bulk liquid random coil

Thus, we are faced with identifying similarities and differences between the
types, length-scales, and cooperativities of HTC chain motions in the very different
environments provided by its neat melt and the narrow crystalline channels of its
-CD-IC, as indicated in Figure 6.

Figure 6. Comparison of conformations and environments of HTC chains in

their neat randomly-coiling liquid and confined to the narrow channels of their
-CD-IC.

In Table 3 the spin-lattice relaxation times of neat -CD and for HTC and
-CD observed in their IC are presented. First, it is obvious that in the cage
272
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 crystalline structure of neat -CD, the constituent carbons in each glucose ring
evidence T1(13C)s an order of magnitude longer than those observed in the
columnar crystal structure of the HTC--CD-IC. It is also clear that the T1(13C)s
of the ring carbons (C1-5) in -CD are also ~ an order of magnitude longer than
those of the guest HTC chains residing in the columnar IC channels formed by
the stacks of host -CD. As a result, we cannot attribute the very short T1(13C)s
observed for the included guest HTC chains to facile ~100 MHz motions of the
host -CD protons. The near coincidence of T1(13C)s observed for HTC in the
melt and in the narrow channels of its -CD-IC must be a consequence solely of
the ability of HTC chains to undergo facile ~100 MHz motions in both of these
very disparate environments.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Lyerla et al. (7) examined the 13C spin-lattice relaxation times for a series
of liquid n-alkanes and a single branched alkane, 2-methylnonadecane. From
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

observations made at 15.08 MHz, they were able to determine the corresponding
effective correlation times, eff, for rotation of their CH bond vectors,

where NH is the number of protons directly attached to a carbon nucleus and rCH is
the length of the CH bond. They noted that i) for a particular carbon (C1, C2, C3,
etc.) eff progressively increases as the chain length increases, ii) effs increase from
the chain ends toward the interior of each n-alkane, and iii) the ratio of correlation
times for the terminal methyl groups to those of the internal methylene carbons also
increases as the length of the n-alkane increases. This suggested that the rotational
motion of CH bond vectors in hydrocarbons can be analyzed as a sum of rigid
body rotation of the entire n-alkane, with a rate of 0-1, plus internal motion due to
rotations about individual CC bonds i, with a rate i-1.
The methyl branch in 2-methylnonadecane caused its C1-5 carbons to have
resonances distinct from those of the C15-19 carbons at the unsubstituted terminus.
However, the T1(13C) observed and the correlation time derived for C5 were the
same as those observed and derived for the C6-15 carbons in 2-methylnonadecane
and the C5-16 carbons in the linear isomer eicosane. As a consequence, Lyerla et al.
(7) concluded that the segmental motion, which determines the correlation time
of a carbon in a (liquid) long chain n-alkane, involves ca.5-6 carbons on each side
of a given carbon. From this we can at least suggest that the efficient ~100 MHz
motions responsible for the short T1(13C)s observed for interior HTC carbons in the
melt involve somewhat cooperative inter-conversions between the conformations
of 10-12 carbon atom chain segments.
Clearly distinguishable from these longer-range cooperative motions are the
much faster, more localized, and probably less correlated t g conformational
inter-conversions occurring about the -CC- bonds in n-alkanes Cn (n 4) at
frequencies much greater than 100 MHz. These result in the rapid establishment
of equilibrium bond conformational populations that are reflected in the
resonance frequencies observed in liquid n-alkanes, which are described above
by the conformationally sensitive -gauche effect, and are ineffective in the
spin-relaxation of their carbon nuclei.

273
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 3. Solid-State 13C NMR spin-lattice relaxation times for neat -CD
and HTC and -CD in HTC--CD-IC
T1, s
-CD
Sample C1 C4 C2,C3,C5 C6 CH2(HTC)
Cage -CD (25C) 200, 13 190, 12 190, 12 103, 7
HTC/IC (25 C) 37 27 21 4, 0.4 1, 3.6
HTC/IC (85 C) 21 17 13 0.6 2.6
HTC (crystal at 25 505
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

C)
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

HTC (melt at 85 C) 1.2

Turning now to the ~100 MHz motions of HTC chains isolated in the narrow
channels of its -CD-IC, two things are clear from the observed splitting of the
C2,35 methylene resonance at 24.8 ppm in crystalline HTC into two resonances at
~24.8 and ~22.8 ppm in HTC--CD-IC (also C5,32 at ~33 and ~31 ppm) and the two
distinct T1s observed for the internal carbons at 25 C (See Table 3). First, the short
T1s observed for the internal carbons of HTC included in the -CD-IC channels are
not dominated by rigid-body rotation of all-trans HTC chains in the host -CD-IC
channels. Second, roughly half the C3C4 and C33C34 bonds are experiencing t
g inter-conversions that are much too rapid to cause spin-lattice relaxation of
HTC carbon nuclei at the ~100 MHz frequency employed here to observe them,
while the other half are rigidly trans.
Consistent with the distinct T1s observed for terminal and interior carbons in
crystalline n-alkanes (6), we observe T1s of 1.8, 30, and 505 s for C1,36, C2,35, and
C3-34, respectively, in neat crystalline HTC and 3.3 and 1.3 s for the C1,36, and C2-35
HTC carbons in the -CD-IC. When compared to the single T1s observed for all
n-alkane carbons in their high temperature rotator phases (6), it is clear that neither
in the neat crystal or -CD-IC are the short HTC T1s a result of rigid body rotation.
In addition, the two T1s observed at 25 C for the internal HTC carbons in
its -CD-IC suggest that the HTC chains are moving under the constraints of two
distinct channel environments. In Figure 7 note that the difference in diameters
of the head and tail portions of -CDs results in a gradual undulation of their
columnar channels that is similar to stacked hour-glasses. Thus, the motions of
those portions of HTC chains in the head-to-head channel regions would not be as
constrained as those in the tail-to-tail regions, possibly providing sufficient room
for facile ~100 MHz motions important to spin-lattice relaxation. These are likely
the regions where C3C4 and C33C34 bonds experience the much more rapid t
g conformational inter-conversions that lead to resonance frequencies for some
C2 and C35 carbons that are shifted upfield by conformationally averaged -gauche
shielding from those in a rigid trans arrangement with C5 and C32.
To assist our analysis and discussion of the ~100 MHz motions that efficiently
relax the HTC carbon nuclei included in the narrow channels of the HTC--CD-IC,

274
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 we briefly summarize a fully atomistic molecular dynamics (MD) modeling of 30
repeat unit (60 carbon) polyethylene (PE-60) chain confined to a single -CD- IC
channel, and conducted by Pozuelo et al. (8). Both rigid and flexible -CD-IC
channels were considered during 1ns MD tra-jectories run at 500K. At the end
of both trajectories, each -CH2CH2- bond had a trans (0 7-11) population of
over 99%. These small torsional fluctuations occured much more rapidly than
~100 MHz, and could not produce efficient T1(13C) relaxation of the HTC carbons
when confined in the narrow -CD-IC channels. In addition, virtually no rigid-
body rotational or translational motions were observed for the included PE-60
chain, consistent with the previously mentioned observation of distinct T1s for the
terminal and interior HTC carbons in its -CD-IC.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

Figure 7. The columnar structure of polymer--CD-ICs. The black areas indicate

the channels occupied by the included guest polymer chains.

Though an order of magnitude less frequently observed than in a single free

PE-60 chain, t g conformational inter-conversions were observed about ~ 5
and 0 of the PE-60 bonds when confined to flexible and rigid -CD-IC channels,
respectively. Extrapolation of these PE-60--CD-IC MD simulation results,
suggests that nearly each of the 59 PE-60 -CC- bonds [(5 bonds/ns)x 10ns = 50
bonds] would undergo a t g conformational inter-conversion over a longer
10ns trajectory, with a corresponding frequency of ~ 100 MHz, while maintaining
an overall nearly all trans average PE conformation. This might explain the short
T1s observed for the predominantly all trans HTC chains in the HTC--CD-IC,
but do the motions of these highly constrained conformational inter-conversions
also explain the similar short T1s observed for HTC in the melt?

275
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 According to the analysis of the spin-lattice relaxations observed in liquid
n-alkanes by Lyerla et al. (7), apparently not, because the segmental motion,
which determines the correlation time of a carbon in a (liquid) long chain
n-alkane, involves ca.5-6 carbons on each side of a given carbon. While the
~100 MHz chain motions effective in relaxing the carbon nuclei in liquid HTC
and in its -CD-IC are both conformational in origin, they likely differ both in
their length-scales and cooperativities. In the -CD-IC, t g transitions are
likely restricted to shorter-range ttt gtg (kink) and possibly ttttt gtttg
(jog) transitions in the somewhat flexible -CD channels (9), and, because t g
transitions require surmounting a higher intrachain barrier than g t transitions
and g conformations are rare there, they are also likely more cooperative. Liquid
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

HTC chains are not, however, constrained by -CD-IC channels, but must only
avoid segmental overlaps during and after their conformational transitions, and
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

so their ~100 MHz motions are likely less cooperative and apparently longer in
range.
Before concluding, it should be noted that the conformations and motions of
n-alkane guests included in the narrow channels of their inclusion compounds or
clathrates made with other hosts, such as urea (U), cyclotriphosphazenes (TPP),
perhydrotriphenylene (PHTP), etc. generally appear to be distinct (1013) from
those discussed here for HTC in its -CD-IC. This may be due to the unique
covalently-bonded cyclic nature of host CDs (See Figure 1). The constraints
experienced by guest polymers, that are threaded through host CDs and that are
eventually completely included in the channels formed by the uniaxially stacked
CDs when they crystallize to form a polymer-CD-IC, are expected to be more
severe than those produced by the channels formed from U, TPP, and PHTP
hosts, whose constraining channel walls lack the strength of the covalent bonding
in individual CDs.

Summary and Conclusions

In the rigid crystals of HTC there are apparently almost no ~100 MHz
motions able to relax the HTC 13C nuclei. In the randomly-coiling liquid there
are relatively long-range conformational motions, that are able to efficiently relax
the HTC 13C nuclei. Though predominantly restricted to the nearly fully extended
all trans conformation in the narrow (~5) -CD-IC channels, apparently here
there are also facile ~100 MHz conformational (t g) motions occurring that
very efficiently relax the HTC 13C nuclei. However, in the -CD-IC channels the
motions of guest HTC chains are more cooperative in nature and of a shorter-range
than in the neat liquid.

References
1. Hunt, M. A.; Tonelli, A. E.; Balik, C. M. J. Phys. Chem. B 2007, 111, 3853.
2. Torchia, D. A. J. Magn. Reson. 1978, 30, 613.
3. Tonelli, A. E. NMR Spectroscopy and Polymer Microstructure:The
Conformational Connection; VCH Publishers: New York, 1989.
276
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 4. Flory, P. J. Statistical Mechanics of Chain Molecules; Wiley-Interscience:
New York, 1969; Chapter 2.
5. Cheng, H. N.; Bovey, F. A. Org. Magn. Reson. 1978, 11, 457.
6. Russell, K. E.; Wu, G.; Blake, S.; Heyding, R. D. Polymer 1992, 33, 951.
7. Lyerla, J. R.; McIntyre, H. M.; Torchia, D. A. Macromolecules 1974, 7, 11.
8. Pozuelo, J.; Mendicutti, F.; Saiz, E. Polymer 2002, 43, 523.
9. Tonelli, A. E. Macromolecules 1990, 23, 3134.
10. Sozzani, P.; Bovey, F. A.; Schilling, F. C. Macromolecules 1991, 24, 6764.
11. Schilling, F. C.; Amundson, K. R.; Sozzani, P. Macromolecules 1994, 27,
6498.
12. Nakaoki, T.; Nagano, H.; Yangida, T. J. Mol. Struct. 2004, 699, 1.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

13. Becker, J.; Comotti, A.; Simonutti, R.; Sozzani, P.; Saalwachter, K. J. Phys.
Chem. B 2005, 109, 23285.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch016

277
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 17

NMR Characterization and Product Design of

Novel Silk-Based Biomaterials
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

Aya Nagano,1 Yu Suzuki,1 Yasumoto Nakazawa,2 J. T. Gerig,3

and Tetsuo Asakura*,1
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

1Department of Biotechnology, Tokyo University of Agriculture

and Technology, Koganei, Tokyo, 184-8588 Japan
2Science Museum, Tokyo University of Agriculture and Technology,

Koganei, Tokyo, 184-8588 Japan

3Department of Chemistry, University of California,

Santa Barbara, CA 93106 U.S.A.

*E-mail: [email protected]

The atomic-level structure of the crystalline region of B. mori

silk fibroin was studied using solid state NMR and statistical
mechanical calculations. Whereas lamellar structures were
proposed for both (AG)15 and (AGSGAG)5 repetitive sequences
of the crystalline region, the -turn in (AGSGAG)5 was found
to be more distorted compared to that in (AG)15. In view of
this structural information, silk-like mineralized composite
materials were designed and characterized by solid state
NMR. Furthermore, recombinant proteins were designed and
produced in E. coli and transgenic silkworm. The films of these
recombinant proteins exhibit high calcium binding activity and
rapid mineralization.

I. Introduction
Silks are fibrous proteins with properties that have intrigued scientists ranging
from polymer chemists to biomedical researchers (1). Silkworm (Bombyx mori or
B. mori) produces silk fibroin filaments with outstanding strength, toughness, and
other desirable properties, despite being spun from an aqueous solution at ambient
temperature and pressure. In addition, silk fibroin has good biocompatibility,
resistance to thermal and enzymatic degradation, and mechanical resilience, and

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 it can be fashioned into a wide range of forms including nano- and microparticles,
films, sponges, and composites. This portfolio of properties suggests its
considerable potential for applications in medicine and surgery (2).
Silk fibroin is composed of amorphous domains (mainly consisting of
hydrophilic amino acids) and crystalline domains (mostly hydrophobic amino
acids). The primary structure of crystalline domains contains multiple repetitions
of (G-A-G-A-G-S) sequence which make up approximately one-half of the total
silk fibroin (3). These repetitive sequences are responsible for the formation
of antiparallel -sheets in the spun fibers. The -sheet formation ultimately
contributes to the stability and the remarkable mechanical properties of silk fibers.
The objectives of this work are to investigate the molecular-level
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

conformation of silk fibroin crystalline region and design silk-like biomaterials

based on the structural information obtained. In Section II, the conformational
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

analysis of the crystalline region of silk fibroin is described. The design of novel
silk-based biomaterials and their characterization are delineated in Section III.
The applications of these novel biomaterials are described in Section IV. Finally,
in Section V the behavior of silk fibroin in fluorinated alcohols as studied by
solution NMR is summarized.

II. Structural Studies

Silk fibroin has two distinct structures in the solid state, silk I (before spinning)
and silk II (after spinning). The conformation of silk I has been shown by solid
state NMR and X-ray diffraction to have a repeated type II -turn structure (4,
5). As for the structure of silk II, Marsh et al. (6) were the first to propose
an antiparallel -sheet model based on fiber diffraction studies of native B. mori
silk fiber. Even though the local protein conformation is still the basic -sheet
as proposed by Marsh et al., a refined silk II model incorporates stacking of the
-sheet planes in two different arrangements (7).
Solid state NMR was used to clarify the silk II structure of both B. mori silk
fibroin and (AG)15 (810). Broad and asymmetric Ala C peaks were observed
in the 13C CP/MAS NMR spectra, indicating a heterogeneous structure. The
relative proportions of the various heterogeneous components were determined
from the relative peak intensities of the deconvoluted spectra. For example, the
deconvoluted Ala C peaks yielded 27% distorted -turn (16.7 ppm), 46% -sheet
(alternating Ala residues, 19.1 ppm), and 27% -sheet (parallel Ala residues, 22.4
ppm) (9).
In an investigation of the heterogeneous components, the local structure
for each Ala residue was determined from 13C CP/MAS NMR spectra for
model peptides of the crystalline region of silk fibroin, (AG)15 and (AGSGAG)5
(11, 12). Figure 1 shows the difference in Ala C spectra of 10 different
[3-13C]Ala-(AGSGAG)5 peptides differing in their 13C labeling positions. It
shows that the peak patterns change largely depending on the labeled position of
the Ala residue, indicating varying local structures along the peptide chain. The
fraction corresponding to the 16.7 ppm component was determined to be 75% for
the C carbon of Ala (1) in the peptide, indicating that the Ala residue at the start
282
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of the sequence has a high probability of forming a random coil. The fraction
was markedly lower (26%) at Ala (5) but rose to a second maximum (44%) at
Ala (19).
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 1. Fraction (%) of the 16.7 ppm peak calculated from the deconvolution
of Ala C peaks in the 13C CP/MAS NMR spectra of 10 [3-13C](AGSGAG)5 with
different [3-13C]Ala labeling sites as a function of the 13C labeled positions.

For an investigation of the entire conformation of (AGSGAG)5 in silk II

form, it is necessary to determine the local conformations of Ser residues as
well as Ala residues. In our laboratory the information was obtained from the
internuclear distances between [1-13C]Gly and [15N]Gly nuclei in the sequence,
[1-13C]Gly-Ser-[15N]Gly, indirectly because the distance reflected the local
conformation of the Ser residue between two Gly residues. REDOR (13) was
used for this purpose.(Figure 2) As a starting point, the theoretical REDOR
curve was calculated by assuming the distance between [1-13C]Gly and [15N]Gly
nuclei in the sequence, [1-13C]Gly-Ser-[15N]Gly, to be 4.75 , similar to the
antiparallel -sheet structure obtained from the crystal structure of B. mori silk
fibroin determined by Takahashi et al. (7) However, the curve could not reproduce
the observed REDOR curve, indicating that the conformation of the Ser residue
was a mixture of random coil and antiparallel -sheet structures, as expected. The
atomic distance for random coil conformation was calculated to be 3.80 by
averaging the distances between the carbonyl carbon of the first Gly residue and
the amide nitrogen of the third Gly residue in the Gly-Ser-Gly sequences selected
from PDB data base. With the assumption of the presence of both antiparallel
-sheets and random coil, the REDOR curve was re-calculated. The theoretical
REDOR curve of the mixture, 30% random coil and 70% antiparallel -sheet,
gave the best-fit to the experimental REDOR plot of Ser15th. Thus, the fraction of
random coil for Ser15th residue was determined to be 30 % and the same fitting
process was performed for the remaining four peptides.

283
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

Figure 2. Observed and calculated REDOR plots for the [1-13C]Gly14-Ser15-

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

[15N]Gly16-labeled (AGSGAG)5 peptides (Ser15th). Solid lines represent the

REDOR curves calculated by changing the fractions of the random coil (3.80 )
and -sheet (4.75 ) forms. The broken line corresponds to the -sheet (4.75 ).

The fractions of the distorted -turn and/or random coil components were
determined from the relative intensities of [3-13C]AlaC peaks at 16.7 ppm for
Ala residues and for Ser residues from the REDOR curves of the [1-13C]Gly-Ser-
[15N]Gly labeled versions. Figure 3 shows these data plotted against the residue
position within (AG)15 (gray) and (AGSGAG)5 (black). These changes strongly
indicate -sheet -turn -sheet repeated lamellar structure. A comparison of
the plots for (AG)15 and (AGSGAG)5 shows that on the whole the fractions of the
distorted -turn and/or random coil component were higher in (AGSGAG)5 except
for the two Ala residues at the N-terminal sites. The way the fractions varied with
position was slightly different in the two peptides.

Figure 3. Observed relative intensities (black) of distorted -turn and/or random

coil structural component of (AGSGAG)5 obtained from the peak at 16.7 ppm
of Ala C and REDOR bimodal fitting for Ser residues are shown at different
position. The observed relative intensities (grey) of distorted -turn and/or
random coil structure component of (AG)15 are also shown for comparison.

284
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 We used statistical mechanical calculation (27) to investigate possible lamellar
structures in (AG)15 (14) and (AGSGAG)5 (12). The calculations were initially
based on the following assumptions: (1) after -turn formation along the chain
there was at least one pair of intramolecular hydrogen-bonded strands forming a
-sheet structure; (2) the direction of the -turn formation along the chain was
always from the N-terminal to the C-terminal; (3) there was one or two turns in
the peptide. The occurrence probability, p(i), of the i-th structure for (AG)15 and
(AGSGAG)5 was calculated as follows:
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

where E(i) is the potential energy of the i-th structure, k is the Boltzmanns
constant, and T is the absolute temperature. The relative potential energy of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

the i-th structure, E(i), was estimated from the stabilization energy, EH (due
to the formation of inter-residue hydrogen bonds for each residue with -sheet
structure), ET (due to the formation of turn structure), and ES (due to Ser
residues incorporated in -sheet).

where nH(i), nT(i) and nS(i) are the numbers of hydrogen bonds, turns in the i-th
structure, and Ser residues incorporated in -sheet, respectively.
The fraction of non--sheet conformations calculated with optimized EH,
ET, and ES were plotted against the residue position. The calculated plot
agreed well with the observed data for (AG)15. However, the agreement between
the observed and calculated plots for (AGSGAG)5 was not satisfactory when
compared to (AG)15 as shown in Figure 4(a). Therefore, the calculations were
repeated by changing an assumption slightly for (AGSGAG)5; the residues
incorporated in the -sheet structure next to the -turn positions were allowed
to form one hydrogen bond instead of two. Other assumptions were kept
the same. With optimized EH, ET and ES, the fraction of non--sheet
conformations calculated was plotted against the residue position. A considerably
better agreement between the observed and calculated plots was obtained when
compared with the first calculation (Figure 4(b)).
The most probable lamellar structure for (AGSGAG)5 is shown in Figure 5.
A comparison of the plots for (AG)15 and (AGSGAG)5 shows that on the whole
the fractions of the distorted -turn and/or random coil component were higher in
(AGSGAG)5 except for the two Ala residues at the N-terminal sites. This result
indicates that the presence of repeated Ser residues in the peptide chain may
result in some disturbance to the -sheet distorted -turn lamella structure. In this
connection it is interesting to note that Fraser et al. (15, 16) reported that unit cell
dimensions (a, b, c) of (AG)n and (AGAGSG)m are 9.42, 6.95, 8.87 and 9.39,
6.85, 9.05 respectively. In the c-axis, the direction perpendicular to the plane of
the pleated sheet, the unit cell dimension in (AGAGSG)m was 9.05 , markedly
larger than the 8.87 in (AG)n. The data strongly support our hypothesis that
the side chains of Ser residues in (AGSGAG)5 stick out perpendicular to the
plane of the -sheet, producing a larger c-spacing compared to (AG)15. We
suggest that this larger separation allows looser, less regularly located turns in
285
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the lamella structure. Gong et al. (17) proposed a tentative structural model for
cross- protofibrils prepared from regenerated B. mori silk fibroin solution on
the basis of X-ray diffraction, TEM, and AFM images. The nanofibrils were
approximately 5 nm in width while the GS or GA at the C-end of a typical
segment GAGAGSGAGAGS gave rise to a single turn between a -strand dimer.
The laminated -sheet length proposed by them is reasonably consistent with our
models which consist of turns and -strand with approximately 8-12 amino acids
at the central sequence of the proposed lamella structure with two turns for the
peptides, (AG)15 and (AGSGAG)5.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 4. Observed (black) and calculated relative intensities (grey) of the -turn
and/or random coil structure component with different position. (a) this plot
represents the first calculation, where the agreement between the observed and
calculated plots is not satisfactory compared to the same analysis for (AG)15.
(b) this plot gives the revised calculation where the assumptions are slightly
changed: the residues incorporated into -sheet structure next to the -turn
positions form only one hydrogen bond instead of two. The agreement between
the observed and the calculated plots is satisfactory.

Figure 5. The most probable lamellar structure for (AGSGAG)5, obtained by the
combined use of 13C CP/MAS NMR and statistical mechanical calculations.
286
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 III. Design and Modifications
The mineralization process is naturally induced in native bone through
self-assembly at the charged acidic domains of non-collagenous proteins, which
provide adequate conditions for calcium-binding, nucleation and development
of ordered hydroxyapatite crystals (1821). Peptide-mediated initiation of
nanocrystals offers a new platform for obtaining self-assembling of peptides
with tailor-made sequences that promote biomineralization. The concept of
molecular self-assembly has become the driving force for the development of new
biomaterials (2235). Scientists have identified key parts of molecules (mainly
proteins) responsible for the self-assembly process, analyzed them, and developed
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

simpler systems. Earlier research has revealed the tendency of copolypeptides

with alternating hydrophilic and hydrophobic residues to form water-soluble
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

-sheet structures through self-assembly (3639). These studies have shown

that alternating amphiphilic-peptide sequences such as poly-(Lys-Phe),
poly-(Glu-Ala), poly-(Tyr-Glu), poly-(Val-Lys), and poly-(Lys-Leu) produce
-sheet secondary structures and aggregates, depending on salt concentration and
pH (3639).
In our efforts, we sought to design silk-like amphiphilic proteins for bone
repair with glutamic acid or aspartic acid components (4042). B. mori silk
crystalline region has a lamellar structure as indicated in Section II; it seems
useful to design biomaterials based on that structure. The design concept also
originated from a desire to utilize silk with superior mechanical properties
and biocompatibility as bone repair scaffold and to develop the hydrophobic
(AGSGAG) region in silk as a part of self-assembling amphiphilic moiety. As
mineralization site, we adopted a hydrophilic sequence with contiguous acidic
(glutamic and aspartic acids) from non-collagenous proteins. The proteins were
then produced with the hydrophobic region of silk and the hydrophilic region
with mineralization characteristics.
Thus, we first started with silk-like peptides based on Ala-Gly repeated
sequences with a lamellar structure and Asp as a Ca binding site (Figure 6). We
further modified the design of the lamellar structure by introducing a Ser residue
between (GlyAla)3 and (AlaGly)3 sequences.
1. Structure 41SDSDS - (GlyAla)3Ser(AlaGly)3Asp(GlyAla)3
Ser(AlaGly)3Asp - (GlyAla)3Ser(AlaGly)3. There are three labeled versions with
13C labeling in different positions.

2. Structure 14DS16 - ThrSer-[(AlaGly)3Asp(GlyAla)3Ser]16.

The lamellar structures of the two novel artificial silk-like materials
(41SDSDS and the much larger 14DS16) were characterized using 13C CP/MAS
and spin-diffusion solid-state NMR. The -sheet fraction in Ala residues increased
with increased distance from the Asp residue in the turn part. The Ser residue
assumed almost 100% -sheet structure probably because it formed an extra
hydrogen bond that stabilized the stem part of (AlaGly)n. Position-selective
and sensitive information useful to characterize the detailed lamellar structure
with heterogeneous local conformation was obtained by selective labeling of the
peptide with 13C and determining 13C conformation-dependent NMR chemical
shifts. If the Asp residue is located at the center of the turn, one-half of these

287
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Ala residues forms -sheet with intermolecular hydrogen bonding, but the other
half takes on a distorted structure under the influence of the adjacent Asp residue.
The high fraction of random coils strongly suggested that Asp (14) is at the
center of the turn as discussed by Wang et al (43). In contrast, the fraction of
-sheet increased to 67% for Ala (10) and 68% for Ala (8) residues. Similarly,
the sheet increased to 70% for Ala (18) and 79% for Ala (20) residue. Thus,
apart from proximity to the turn position (Asp (14)) the fraction of -sheet was
high. The fraction of -sheet was slightly higher for Ala (16), Ala (18) and Ala
(20) compared with Ala (12), Ala (10), and Ala (8) indicating that the proportion
of -sheet slightly increased for the Ala residues located in the central parts of the
stem region of the lamellar structure (Figure 7). As shown in Figure 7, the Ser C
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

(54.6 ppm) and C (63.7 ppm) peaks indicated that the Ser residues in 41SDSDS
take on almost solely -sheet structure (44). The deuterium solid-state NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

study of [3,3-2H2]Ser B. mori silk fibroin fiber suggested that the hydroxyl groups
of Ser interact with carbonyl groups on adjacent chains and thereby contribute to
the intermolecular hydrogen-bonding network of the fiber (45). Figures 7(d) and
6(e) show the 13C CP/MAS NMR spectra of nonlabeled 41SDSDS and 14DS16
respectively. The two spectra are very similar; for example, the chemical shifts
of the main peaks are identical. Thus, 14DS16 is thought to take on a structure
similar to that of 41SDSDS; Asp residues are exposed on the surface of the
lamellar structure and are not incorporated into the -sheet structure. The Ser
residues in the 14DS16 may similarly contribute an additional intermolecular
hydrogen bond, thereby further stabilizing the -sheet structure of the stem part
of 41SDSDS. Our results from 13C CP/MAS NMR indicated that there are still
around 30% distorted -turn and/or random coil in the Ala residues adjacent to
the Ser residues.

Figure 6. Design of lamellar structure of the silk-like peptide based on Ala-Gly

repeated sequences with antiparallel -sheet structure, where Asp (D) is a Ca
binding site in the turn part and a Ser (S) residue is located between (GlyAla)3
and (AlaGly)3 sequences to strengthen the stem part of the lamella.

We designed another silk-like protein, (Glu)n(Ala-Gly-Ser-Gly-Ala-Gly)4;

(E)n(AGSGAG)4 where n = 4-8. First, peptides with a combination of hydrophilic
and hydrophobic sequences mimicking the primary structure of Bombyx mori
288
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 silk fibroin were synthesized and studied in the solid state by NMR using 13C
selective labeling coupled with 13C conformation-dependent chemical shifts
and 2D solid-state spin diffusion NMR (41). The hydrophilic sequence was
poly(L-glutamic acid) (E)n, and the hydrophobic part was the consensus sequence
of the crystalline fraction of B. mori silk fibroin, (AGSGAG)4. The balance of
hydrophilic and hydrophobic characters of the peptide was controlled by changing
the relative length, n, of (E)n from 4 to 8. When n = 4 and 5, the structure of
the hydrophobic sequence is basically Silk I (the structure of B. mori silk fibroin
before spinning in the solid state), and the polyglutamate sequences take on the
random coil form (Figure 8). On the other hand, when n = 6-8, the structure of
the polyglutamate sequence changes gradually from random coil to -sheet, and
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

the hydrophobic sequence adopts a mixture of -sheet and random coil/distorted

-turn forms, although the fraction of the latter form decreases gradually by
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

increasing the number n from 6 to 8.

Figure 7. 13C CP/MAS NMR spectra of (a) [1-13C]Ala12,[3-13C]-Ala16-, (b)

[1-13C]Ala10,[3-13C]Ala18-, (c) [1-13C]Ala8,[3-13C]Ala20-labeled 41SDSDS, with
expanded spectrum of Ala C=O and Ala C(d) nonlabeled 41SDSDS, and (e)
silk-like protein analogue 14DS16.

IV. Applications
The protein 14DS16, analogous to 41SDSDS but considerably larger, was
produced by genetic engineering and overexpression in E. coli as a possible
biomaterial. Mineralization of films of 14DS16 in SBF (simulated body fluid) (46)
was investigated. Films of this protein treated with SBF were rapidly mineralized
with hydroxyapatite (Figure 9). For both Ala residues located closest to the Asp
289
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (14) residue, distorted -turn and/or random coil accounted for one-half of the
conformations. Thus, hydroxyapatite formation is promoted by the structure
designed through analogy with lamellar silk crystalline region.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 8. 13C CP/MAS NMR spectra of natural abundance (E)n(AGSGAG)4 (n

= 4-8). The spectra from (a) to (e) correspond to the samples from n = 4 to n
= 8, respectively.

Genetically modified silk fibroin containing a poly-glutamic acid site,

[(AGSGAG)4E8AS]4, for mineralization was produced as fibers by transgenic
silkworms through systematic transformation of the silkworms (42). The Ca
binding activity and mineralization of the transgenic silk fibroin were examined
in vitro, showing that this transgenic silk fibroin had relatively high Ca binding
activity compared with native silk fibroin. Porous silk scaffolds were prepared
with the transgenic and native silk fibroins. Healing of femoral epicondyle defects
in rabbit femurs treated with the scaffolds was examined by observing changes
in images of the defects using micro-computed tomography(Figure 10). Earlier
mineralization and bone formation were observed with scaffolds of transgenic
silk fibroin compared with those of native silk fibroin. Thus, this study shows the
feasibility of using genetically modified silk fibroin from transgenic silkworms as
a mineralization-accelerating material for bone repair.

290
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 9. (a) SEM photomicrographs of a film surface of 14DS16 7 days after

immersing the films in SBF solution, (b) EDX spectrum of 14DS16 film pretreated
with calcium chloride and incubated in SBF for 7 days, (c) High-resolution SEM
photomicrographs showing mineralization on the surface of a film of 14DS16
pretreated with calcium chloride and incubated for 7 days in SBF, (d) SEM
photomicrographs with arrows indicating small but regular crystals with their
flat surfaces in contact with the recombinant protein film.

V. Effects of Fluorinated Solvents

The regenerated silk fibers prepared from aqueous silk solution do not have
as good mechanical properties as the native silk fibroin. However, the fluorinated
organic solvents 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) and hexafluoroacetone
trihydrate (HFA) have been successfully used to produce silk fibroin fibers with
high strength (4749). To understand the reasons for the increased strength
of regenerated silk fibers prepared from these two solvents, we analyzed and
compared the properties of the native silk fibroin and (AGSGAG)2, a model for
the crystalline part of fibroin, in the two solvents by 1H/13C chemical shifts and
1H-19F intermolecular NOEs (50, 51).

291
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 10. (a) Micro-CT image of rabbit femurs in not-implated group (control),
native silk fibroin scaffold-implanted group, and transgenic silk fibroin scaffold
with (E)8(AGSGAG)4 - implanted group, respectively, and quantitative results of
(b) bone mineral content and (c) bone volume per total volume.

The 13C and 1H chemical shifts obtained from 1H-13C HSQC spectra of silk
fibroin in HFIP, HFA, and water indicated that silk fibroin formed helix-like
structures in the fluorinated alcohols. A similar tendency was observed for
(AGSGAG)2 although the chemical shift change was smaller. In addition,
intramolecular 1H-1H NOE data for (AGSGAG)2 imply the presence of helical
structures in the middle part of the peptide in HFIP but not in HFA, although an
equilibrating ensemble of conformations are likely present in both solvents.
Nuclear spin dipole-dipole interactions between the spins of (AGSGAG)2
and the fluorines of the solvent can produce relaxation of the protons on the
peptide. These solute-solvent interactions lead to intermolecular NOEs and
are characterized by a cross relaxation parameter () (52). Figure 11 shows
the observed 1H-19F NOE spectra of (AGSGAG)2 in HFIP and HFA. All the
detectable intermolecular NOEs arising from peptide in HFIP are positive; all
NOEs in HFA are negative.
The simplest interpretation of intermolecular NOEs assumes that solute and
solvent molecules can be represented by hard spheres and that interactions of the
spheres depend only on their mutual translational diffusion (53, 54). A numerical
procedure that involves the treatment of Ayant et al. (54, 55) can be used to
incorporate the shape of the solute molecule into predictions of . Input to these
calculations includes experimental translational diffusion coefficients, the radius
of the sphere that represents a solvent molecule, the effective rotational correlation
time that characterizes 1H-1H intramolecular dipolar interactions within the solute
and a correlation time for internal rotation of methyl groups. The latter correlation
times can be estimated by considering the observed spin-lattice relaxation times
of the protons of the solute (56).
292
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Figure 11. 1H-19F intermolecular NOE spectra of (AGSGAG)2 in HFIP (A) and
in HFA(B). Upper spectra in each set are the observed 1D 1H spectrum while the
lower spectrum is the 1H-19F NOE spectrum at a mixing time of 500 ms for the
HFIP system and 300 ms for the HFA system.

Cross relaxation terms (HF) for all conformations of (AGSGAG)2 in

HFIP and in HFA developed by our analysis of intramolecular 1H-1H distance
constraints were calculated using the diffusion coefficients and the other
parameters mentioned. The average of these calculated HF values for each proton
of the peptide in each solvent is compared to the corresponding experimental value
in Figure 12. We find that HF predicted in this way for (AGSGAG)2 dissolved
in HFIP are in good agreement with experimental data for most protons of the
peptide. This result indicates that the model used for these predictions reasonably
describes the interactions between HFIP and the peptide. Thus, these interactions
can be characterized as random collisions between peptide and HFIP, and the
dynamics of these collisions are sufficiently characterized by the experimental
bulk diffusion coefficients. In contrast, observed HF for (AGSGAG)2 in HFA
are negative for all residues while calculated HF were positive. Peptide-solvent
interactions are thus more complex in HFA than what seems to be the case in
HFIP.
In the processes for producing regenerated silk fibers from a silk solution,
the solution typically travels through a stainless steel spinneret into methanol
to achieve coagulation. Coagulation involves removal of solvents from around
the dissolved silk fibroin producing a conformational transition of the crystalline
regions of silk fibroin to -sheet structures. The regenerated fiber prepared from
the HFIP solution shows slightly greater tensile strength when the draw ratio is
1:3 than that of native silk fiber, but the strength of the regenerated fiber with
293
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 draw ratio 1:3 from the HFA solution is about 40% smaller than that of native
silk fiber. The X-ray diffraction patterns and Ala C signals in solid state 13C
CP/MAS NMR spectra revealed that this difference in the tensile strength of the
regenerated silk fibers between two solvents comes from the difference in the
long-range orientation of the crystalline regions (57). According to the results,
the displacement of HFA molecules during coagulation may be less complete due
to the stronger interactions of HFA with silk fibroin. A less extensive -sheet
aggregation during coagulation could result, leading to lower tensile strength of
fibers from HFA solutions. Also, the silk model peptide conformation possesses
a somewhat helical structure in HFIP compared to that in HFA, implying that silk
fibroin with more extensive helical conformations in HFIP tends to align more
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

like native silk solution in silk gland and favor the extensive -sheet aggregation.
Silk fibroin dissolved in HFIP can be fashioned into a wide range of forms
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

when applied to biomaterials, e.g., sponges and scaffolds for bone tissues, films
for wound dressings, and scaffolds for cornia. The knowledge obtained in this
work should be useful in the development of better and improved silk-based
biomaterials.

Figure 12. Comparison of observed HF (solid line) and calculated HF (dotted

line). The calculated data are the average of calculations for the conformations
obtained by DYANA calculation. Error bars represent standard deviations of
calculated data for 10 conformations for HFIP sample and 7 conformations for
HFA sample found in the structure determination.

294
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 References
1. Asakura, T.; Kaplan, D. Encylopedia of Agricultural Science; Academic
Press: New York, 1994; p 1.
2. Hakimi, O.; Knight, D. P.; Vollrath, F.; Vadgama, P. Compos Part B-Eng
2007, 38 (3), 324337.
3. Zhou, C. Z.; Confalonieri, F.; Jacquet, M.; Perasso, R.; Li, Z. G.; Janin, J.
Proteins: Struct., Funct., Genet. 2001, 44 (2), 119122.
4. Asakura, T.; Ohgo, K.; Komatsu, K.; Kanenari, M.; Okuyama, K.
Macromolecules 2005, 38 (17), 73977403.
5. Asakura, T.; Ashida, J.; Yamane, T.; Kameda, T.; Nakazawa, Y.; Ohgo, K.;
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

Komatsu, K. J. Mol. Biol. 2001, 306 (2), 291305.

6. Marsh, R. E.; Corey, R. B.; Pauling, L. Biochim. Biophys. Acta 1955, 16 (1),
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

134.
7. Takahashi, Y.; Gehoh, M.; Yuzuriha, K. Int. J. Biol. Macromol. 1999, 24
(2-3), 127138.
8. Asakura, T.; Yao, J.; Yamane, T.; Umemura, K.; Ulrich, A. S. J. Am. Chem.
Soc. 2002, 124 (30), 87948795.
9. Yao, J. M.; Ohgo, K.; Sugino, R.; Kishore, R.; Asakura, T.
Biomacromolecules 2004, 5 (5), 17631769.
10. Asakura, T.; Yao, J. M. Protein Sci. 2002, 11 (11), 27062713.
11. Asakura, T.; Nakazawa, Y.; Ohnishi, E.; Moro, F. Protein Sci. 2005, 14 (10),
26542657.
12. Suzuki, Y.; Aoki, A.; Nakazawa, Y.; Knight, D. P.; Asakura, T.
Macromolecules 2010, 43, 94349440.
13. Gullion, T.; Schaefer, J. J. Magn. Reson. 1989, 81 (1), 196200.
14. Asakura, T.; Sato, H.; Moro, F.; Nakazawa, Y.; Aoki, A. J. Am. Chem. Soc.
2007, 129 (17), 57035709.
15. Fraser, R. D.; MacRae, T. P.; Stewart, F. H. J. Mol. Biol. 1966, 19 (2),
580582.
16. Fraser, R. D.; Macrae, T. P.; Stewart, F. H.; Suzuki, E. J. Mol. Biol. 1965,
11, 706712.
17. Gong, Z. G.; Huang, L.; Yang, Y. H.; Chen, X.; Shao, Z. Z. Chem. Commun.
2009 (48), 75067508.
18. Tye, C. E.; Rattray, K. R.; Warner, K. J.; Gordon, J. A. R.; Sodek, J.;
Hunter, G. K.; Goldberg, H. A. J. Biol. Chem. 2003, 278 (10), 79497955.
19. Hunter, G. K.; Goldberg, H. A. Biochem. J. 1994, 302 (Pt 1), 175179.
20. Harris, N. L.; Rattray, K. R.; Tye, C. E.; Underhill, T. M.; Somerman, M. J.;
Derrico, J. A.; Chambers, A. F.; Hunter, G. K.; Goldberg, H. A. Bone 2000,
27 (6), 795802.
21. Ganss, B.; Kim, R. H.; Sodek, J. Crit. Rev. Oral Biol. Med. 1999, 10 (1),
7998.
22. Gajjeraman, S.; Narayanan, K.; Hao, J.; Qin, C.; George, A. J. Biol. Chem.
2007, 282 (2), 11931204.
23. Palmer, L. C.; Newcomb, C. J.; Kaltz, S. R.; Spoerke, E. D.; Stupp, S. I.
Chem. Rev. 2008, 108 (11), 47544783.
24. Cui, F. Z.; Li, Y.; Ge, J. Mater. Sci. Eng., R 2007, 57 (1-6), 127.

295
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 25. Bouligand, Y. C. R. Palevol 2004, 3 (6-7), 617628.
26. Semino, C. E. J. Dent. Res. 2008, 87 (7), 606616.
27. Addadi, L.; Weiner, S.; Geva, M. Z. Kardiol. 2001, 90 (3), 9298.
28. He, G.; Dahl, T.; Veis, A.; George, A. Nat. Mater. 2003, 2 (8), 552558.
29. Ball, P. Nature 1994, 367, 323324.
30. Domurado, D.; Vert, M. J. Biomater. Sci., Polym. Ed. 2007, 18 (3), 287301.
31. Rijcken, C. J. F.; Soga, O.; Hennink, W. E.; Nostrum, C. F. J. Controlled
Release 2007, 120 (3), 131148.
32. Khokhlov, A. R.; Khalatur, P. G. Conformation-dependent design of
sequences in copolymers; Springer Verlag: 2006.
33. Firth, A.; Aggeli, A.; Burke, J. L.; Yang, X.; Kirkham, J. Nanomedicine 2006,
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org

1 (2), 189199.
34. Kirkham, J.; Firth, A.; Vernals, D.; Boden, N.; Robinson, C.; Shore, R. C.;
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

Brookes, S. J.; Aggeli, A. J. Dent. Res. 2007, 86 (5), 426430.

35. Gajjeraman, S.; He, G.; Narayanan, K.; George, A. Adv. Funct. Mater. 2008,
18 (24), 39723980.
36. Peggion, E.; Cosani, A.; Terbojevich, M.; Borin, G. Pept. Sci. 1972, 11 (3),
633643.
37. Rippon, W. B.; Chen, H. H.; Walton, A. G. J. Mol. Biol. 1973, 75 (2),
369375.
38. Seipke, G.; Arfmann, H.-A.; Wagner, K. G. Biopolymers 1980, 19 (1),
189201.
39. Brack, A.; Orgel, L. E. Nature 1975, 256 (5516), 383387.
40. Asakura, T.; Sato, H.; Moro, F.; Yang, M.; Nakazawa, Y.; Collins, A. M.;
Knight, D. Macromolecules 2007, 40 (25), 89838990.
41. Nagano, A.; Kikuchi, Y.; Sato, H.; Nakazawa, Y.; Asakura, T.
Macromolecules 2009, 42 (22), 89508958.
42. Nagano, A.; Tanioka, Y.; Sakurai, N.; Sezutsu, H.; Kuboyama, N.; Kiba, H.;
Tanimoto, Y.; Nishiyama, N.; Asakura, T. Acta Biomater. 2011, 7 (3),
11921201.
43. Wang, J.; Parkhe, A. D.; Tirrell, D. A.; Thompson, L. K. Macromolecules
1996, 29 (5), 15481553.
44. Asakura, T.; Iwadate, M.; Demura, M.; Williamson, M. P. Int. J. Biol.
Macromol. 1999, 24 (2-3), 167171.
45. Kameda, T.; Ohkawa, Y.; Yoshizawa, K.; Naito, J.; Ulrich, A. S.; Asakura, T.
Macromolecules 1999, 32 (21), 71667171.
46. Kokubo, T.; Kushitani, H.; Sakka, S.; Kitsugi, T.; T. Yamamuro, T. J. Biomed.
Mater. Res. 1990, 24, 721734.
47. Trabbic, K. A.; Yager, P. Macromolecules 1998, 31 (2), 462471.
48. Yao, J. M.; Masuda, H.; Zhao, C. H.; Asakura, T. Macromolecules 2002, 35
(1), 69.
49. Zhu, Z. H.; Imada, T.; Asakura, T. Mater. Chem. Phys. 2009, 117 (2-3),
430433.
50. Ha, S. W.; Asakura, T.; Kishore, R. Biomacromolecules 2006, 7 (1), 1823.
51. Suzuki, Y.; Gerig, I. T.; Asakura, T. Macromolecules 2010, 43 (5),
23642370.
52. Gerig, J. T.; Strickler, M. A. Biopolymers 2002, 64, 227235.

296
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 53. Hennel, J. W.; Klinowski, J. Fundamentals of Nuclear Magnetic Resonance;
Longman: Essex, U.K., 1993.
54. Ayant, Y.; Belorizky, E.; Fries, P.; Rosset, J. J. Phys. (Paris) 1977, 38 (3),
325337.
55. Gerig, J. T. J. Org. Chem. 2003, 68 (13), 52445248.
56. Neuman, R. C., Jr.; Gerig, J. T. Magn. Reson. Chem. 2009, 47, 925931.
57. Zhu, Z.; Kikuchi, Y.; Kojima, K.; Tamura, T.; Kuwabara, N.; Nakamura, T.;
Asakura, T. J. Biomater. Sci., Polym. Ed. 2010, 21 (3), 395411.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch017

297
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 18

The Interaction of A(1-40) Peptide with

Lipid Bilayers and Ganglioside As Studied by
Multinuclear Solid-State NMR
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

Yasumoto Nakazawa,1 Yu Suzuki,1 Hazime Sait,2

and Tetsuo Asakura*,1
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

1Institute of Engineering, Tokyo University of Agriculture and Technology,

2-24-16, Nakacho, Koganei, Tokyo, 184-8588 Japan
2Department of Life Science, Himeji Institute of Technology,

University of Hyogo, Kamigori, Hyogo, 678-1297 Japan

*E-mail: [email protected]

Amyloid -peptide (A) is a major component of plaques in

Alzheimers disease, and formation of senile plaques has been
suggested to originate from regions of neuronal membrane
rich in gangliosides. We have analyzed the interaction of
A with lipid bilayers by multinuclear NMR using 15N, 31P
and 13C nuclei. The result of the present study implies that
the fibrillogenic seed nucleus involves an interaction of His
residues with the sialic acid moiety of GM1. Moreover,
A(1-40) with ganglioside GM1 perturbs the bilayer structure
to form a non-lamellar structure such as hexagonal H11 lipids
and also produces single vesicles or micelles, as shown by
angular-dependent 31P NMR experiments. In conclusion, the
A peptide penetrates into the lipid bilayer, takes on an -helical
form, and produces non-lamellar lipid and micelles.

Introduction
Alzheimers disease (AD) is a devastating neurodegenerative disease
affecting up to 15 million individuals worldwide. The brains of Alzheimers
disease patients are characterized by fibrillar (1) amyloid plaques that are
associated with progressive deposits (1). The principal component of amyloid
deposits is a 39-42 amino acid residue peptide, A (2, 3), a product of proteolytic

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 processing of a much larger amyloid precursor protein (APP) encoded by a gene
on chromosome 21 (4). Several spectroscopic investigations have clarified the
structure of the fibrils, which form a cross- sheet structure (57). Moreover,
a link between A and AD neuropathological lesions is demonstrated by the
toxicity of aggregated A to neuronal cells in culture (8) and in aged brain (9).
However, the molecular mechanisms of the neurotoxic action of A remain
unknown. A growing number of observations indicate that A may alter the
physicochemical properties of neuronal membranes, including membrane fluidity
(10) and permeability to ions and nonelectrolytes (11). These findings strongly
suggest that at least some of the pathophysiological effects of A may be mediated
by A-membrane interactions. Indeed, a number of studies have shown that A
is able to perturb lipid bilayers (1114). Interest in studies of the interaction
between A and constituents of brain membranes has been further stimulated by
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

the identification of an A-GM1 ganglioside-bound form in AD brain (1517).

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 1. Molecular structure of GM1 ganglioside (top) and asialo-GM1

ganglioside (bottom).

Gangliosides (18, 19) are sialic acid-containing glycosphingolipids and

consist of two main components: a hydrophobic ceramide unit, which anchors the
ganglioside to the plasma membrane, and a hydrophilic oligosaccharide chain,
to which one or more sialic acid groups are attached. Gangliosides are abundant
components of neuronal membranes and are involved in important neurobiological
events such as neurodifferentiation, synaptogenesis and synaptic transmission
(20). A permeabilizes ganglioside-containing membranes and thus disturbs ion
homeostasis (21, 22). Different research groups have reported extensive studies
of A-ganglioside interactions using different methodologies (16, 2326), but no
conclusive results have been obtained. GM1 ganglioside-mediated accumulation
of A protein was proposed together with the effect of cholesterol (26).

300
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 In our efforts to clarify the interaction between -amyloid peptide and
lipid bilayers including ganglioside, we have shown two things. First, we
have demonstrated using solution state NMR of 15N-labelled A(140) and
A(142) that ganglioside GM1 and asialo-GM1(Figure 1) contain micelles
(27). Secondly, we have recorded 1H decoupled 31P solid-state NMR spectra
of dimyristoylphospatidylcholine (DMPC) and DMPC/GM1 ganglioside in a
mechanically oriented system in the presence of A(1-40) in order to clarify the
behavior of A in the presence of GM1 ganglioside. Finally, we have studied
13C-labeled A peptide by CP-MAS, DD-MAS NMR, as well as low-power

solution NMR to determine conformational features of the A peptide in the

presence of lipid bilayers.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

Materials and Methods

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Solution NMR Experiments

Uniformly 15N-labelled A(140) and A(142) were purchased from
rPeptide (Athens, GA, U.S.A.), and gangliosides were purchased from Axxora
Ltd (Exeter, U.K.). The purity of the gangliosides was quoted as >98% by the
manufacturer, but was not checked. All other reagents were from SigmaAldrich
Co. (St. Louis, MO, USA). The heparin used was the sodium salt (H4784).
A(140) solutions were prepared as described earlier (28). Briefly, the
peptide was dissolved at a concentration of 200 M in 10 mM NaOH with 1
min of sonication, and immediately frozen if required. Subsequently, the pH
was adjusted to 7.2 with a minimal amount of 0.1 M HCl, and 2H2O was added
to make an approx. 200 M peptide solution containing 10% 2H2O. Solutions
prepared in this way were stable and showed no sign of aggregation for at least
1 week. In contrast, solutions made up in a buffer aggregated much faster, often
almost entirely precipitated after 24 h. A(142) was pre-treated by dissolution in
hexafluoropropan-2-ol and freeze-drying, before dissolving in 10 mM ammonium
hydroxide, followed by adjustment of the pH to 7.2. These solutions started
to aggregate and form fibrils immediately and were only usable for 23 days.
Solutions of the commercial material directly into NaOH resulted in no NMR
signal, indicating significant aggregation with the use of this method.
Stock solutions of gangliosides or heparin were prepared at high
concentrations (approx. 100200 mM), adjusted to pH 7.2 and added directly
to the NMR tube. All NMR experiments were carried out on a Bruker
DRX-500 spectrometer equipped with a cryoprobe, and operated at 13C. 15N
HSQC (heteronuclear single-quantum correlation) experiments used gradient
selection for water suppression and water flipback pulses to minimize loss of
magnetization through exchange and relaxation processes. The three-dimensional
TOCSY-HSQC spectrum incorporated solvent suppression using gradients, with
a 35 ms spin-lock at 8.3 kHz decoupler power. Processing of NMR data used
FELIX (Accelrys, San Diego, CA, U.S.A.), and titration data were analysed
using home-written scripts. Cross-peak intensities were measured within FELIX,
transferred to a text file, and fitted to an exponential decay using a Marquardt
non-linear least-squares fitting based originally on a Numerical Recipes algorithm.
301
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Angular Dependent NMR Experiments

DMPC and GM1 ganglioside were purchased from Sigma (St. Louis, MO,
USA). A(1-40) and 13C labeled DAEFRHDSGYEVHHQKL[1-13C]V18FF[3-
13C]A21EDVGS-NK[2-13C]G29AIIGLMVGGVV were synthesized in a stepwise

fashion on Fmoc-Val-PEG-PS resin (PE Biosystems, Tokyo, Japan) by a

PioneerTM peptide synthesizer (PE Biosystems) using Fmoc chemistry. After
synthesis, the peptides were cleaved from the resin by treatment with a mixture
of TFA(trifluoroacetic acid), phenol, triisopropylsilane and water (88:5:2:5 vol%)
for 2 hours at room temperature. The precipitated peptide was washed repeatedly
with cold diethylether. Then, the crude peptide was purified by preparative liquid
chromatography.
13C- labeled A peptide in the presence of DMPC bilayer was prepared by
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

drying the DMPC under a stream of nitrogen gas from methanol/chloroform

solution, and then in vacuo, followed by hydration with 60% water by weight
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

of lipids wherein the content of DMPC lipids. Repeated freezing and thawing
were performed after hydration. For the preparation of oriented samples, DMPC
(or DMPC/GM1, 10:1) was dissolved in methanol/chloroform (1:1) and A was
dissolved in benzene. These solutions were mixed and spread onto 0.1 mm thin-
glass plates (59 mm) and then dried under vacuum for 24h. The dry plates and
a small amount of water were then stacked in a 10NMR tube (20 mm length).
Water was added to achieve 60% hydration. The sample was then sealed with a
Teflon cap and epoxy resin and incubated at 45 for 4 days.
High power proton-decoupled 31P NMR spectra were performed on a
Chemagnetics CMX 400 spectrometer operating at a resonance frequency of
397.79 MHz for 1H and 161.03 MHz for 31P. A static double-resonance probe
equipped with a goniometer was used. 31P NMR spectra were recorded for the
oriented samples. 31P NMR spectra of sonicated lipid bilayer were also recorded
in a solution NMR spectrometer. For the 31P NMR experiments, typical NMR
parameters were 7-8 s (90) pulse length, 3 s recycle delay and 1000~2000
scans. The 31P chemical shifts in ppm were referenced to H3PO4.
High resolution 13C solid- and liquid-state NMR spectra were recorded on a
Chemagnetics CMX-400 and JEOL -500 NMR spectrometer respectively. For
the former, 13C CP-MAS and DD-MAS (dipolar decoupled magic angle spinning)
NMR with a single pulse excitation method were recorded. The /2 pulse for
carbon and proton nuclei was 5 s and a 1H decoupling frequency of 60 kHz
was used. Contact time was 2msec and spinning rate was 8kHz. Free induction
decays were usually acquired with 512 data points. The 13C chemical shifts were
referenced to adamantane [(28.8 ppm from tetramethylsilane (TMS)] and then
expressed as relative shifts from the value of TMS.

Results and Discussion

Structural Change of A(1-40)-Ganglioside System with Solution NMR

The HSQC spectrum is shown in Figure 2. Almost all backbone signals are
resolved. At 13C, all signals are found except for the N-terminal residue and His6.
302
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 His14 gives a weak signal, as expected because of rapid amide exchange under
these conditions (29). The reason for the absence of His6 in our HSQC spectra is
less clear, but it was also not observable by Hou et al. (28). On the basis of the
chemical shifts, we concur with other authors (28, 30) that the peptide is a random
coil in aqueous solution. The majority of HSQC peaks in A(142) has been
assigned in the same way, but the assignment is somewhat less complete due to
its rapid aggregation, which limits the time available for three-dimensional NMR
experiments. The signals from A(142) have very similar chemical shifts to
those of A(140) except at the C-terminus, implying that there are no significant
conformational differences between them.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 2. HSQC spectrum of 15N-labelled A-(140) in water, pH 7.2, 13C. The

small unlabelled peak at (1H=7.2, 15N=119 p.p.m.) is probably an arginine side
chain NH. Single-letter amino acid codes are used.

GM1 forms micelles with a critical micelle concentration in the low

micromolar range (31). Thus, at concentrations >100 M, as used here, it is
essentially 100% micellar. Upon titration of GM1 micelles into A(140),
chemical shift changes are seen in the NMR spectrum, as shown in Figure 3 (left).
The chemical shift changes are very small, but are reproducible and specific, since
many residues have essentially no change in shift. Smaller chemical shift changes
have been demonstrated to be biologically relevant on many previous occasions
(32, 33). Almost all of the chemical shift changes occur in the N-terminal half of
the peptide, and are close to potentially positively charged residues: Glu3Arg5
303
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (close to Arg5 and His6) and Val12Leu17 (close to His13, His14 and Lys16).
However, the pKa values of the three histidine residues are all approx. 6.5 (34).
Hence, at the pH of our measurements, 7.2, most of the histidine residues should
be unprotonated.
We therefore expect that the only residues significantly positively charged at
this pH should be Arg5 and Lys16, together with Lys28. This makes it unlikely
that the chemical shift changes are due only to Coulombic interactions with the
single negative charge in the GM1 headgroup, the sialic acid.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 3. Chemical shift changes in A-(140) upon the addition of GM1(Left)

and asialo-GM1(Right) to 200 M peptide in water. Each box shows the
results for a different residue, with 1H shifts horizontal (increasing left to right,
total range 0.02 p.p.m (Left) and 0.015ppm (Right) and 15N shifts vertical
[increasing from bottom to top, total range 0.1 p.p.m (Left) and 0.075ppm
(Right)]. This representation therefore resembles the change seen in an HSQC
spectrum, except that the directions of the axes are reversed. The start of the
titration is indicated with a filled circle, and subsequent titrations contain 1, 2, 4
and 8 equivalents of GM1. The size of the circles approximates the experimental
uncertainty. No data are shown for His6 because it could not be observed.
Single-letter amino acid codes are used.

In order to confirm this, a further titration has been carried out at an

approximate physiological salt concentration (150 mM NaCl), which should
markedly reduce purely Coulombic interactions in water. The results show
reduced chemical shift changes, indicating a loss of affinity (data not shown).
Residues 1417 still shift, but the changes seen in residues 313 are much
reduced. This implies that, although some of the binding and the chemical shift
304
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 changes are primarily Coulombic in origin, others (and in particular the changes
seen in residues 1517) are much less so. We note that other authors have
concluded that binding to GM1 is not primarily Coulombic, although again there
are clearly Coulombic interactions (22, 35, 36).
A further titration of A(140) has been carried out, using micelles of asialo-
GM1. The results are shown in Figure 3 (right), in which the chemical shift ranges
used in the plot are 75% of those used for Figure 3 (left); it can be seen that
chemical shift changes with asialo-GM1 are similar to those seen for GM1 but are
reduced in magnitude for equivalent concentrations by approx. 25%. A reduction
is expected in shift magnitude, because asialo-GM1 is known to bind less tightly
than GM1 to A. The extent of the binding has been reported differently before.
Choo-Smith et al. (35) reported no binding at all to asialo-GM1, while a factor
of 2 was also reported (37). Our data agree better with the latter result (37). The
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

overall similarity of the chemical shift changes for GM1 and asialo-GM1 implies
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

that the binding interactions are similar, although with some differences in the
region of His13Gln15. This result therefore also implies that the interaction is not
dominated by Coulombic forces, since asialo-GM1 has no charge in the headgroup.

Interaction between -Amyloid and Lipid Bilayers

In order to clarify the orientation of lipid bilayers with GM1 ganglioside as
a function of A(1-40) peptide concentration, we have recorded the orientational
behavior of DMPC lipid bilayers using angular-dependent solid-state 31P NMR
experiments. The 31P chemical shift tensors of phospholipids are represented
graphically by an ellipsoid. The three main axes of the ellipsoid represent the
main tensor elements 11,22,33. The chemical shift that is measured at a certain
orientation of the molecule corresponds to the length of a vector parallel to
the magnetic field direction. Thus the chemical shifts indicate the tilt of the
phospholipids with respect to the magnetic field and/or the mobility of the lipid
bilayers. In the case of a completely oriented lipid bilayer, 31P NMR peaks
obtained with aligned samples are narrow and there is only a small powder
resonance under the aligned signal.
In a first set of experiments, we have obtained 31P solid-state NMR spectra of
glass-plate aligned DMPC samples as shown in Figure 4. These results indicate
that the DMPC lipid bilayer is completely oriented on the glass plates. The 31P
chemical shift values of DMPC move to high field on changing the alignment
axis from 0 to 90 relative to the magnetic field B0. Figure 4(c) and (d) show
angular-dependent 31P NMR spectra of a DMPC/GM1 ganglioside system. The
behavior of this system is similar to the result in a DMPC lipid bilayer alone,
implying that the addition of GM1 ganglioside does not significantly perturb the
structure of the DMPC bilayer.
Figure 5 illustrates the 31P NMR spectra of a mechanically oriented DMPC
bilayer as a function of the angle () between the applied magnetic field and bilayer
normal. The left, central and right traces correspond to the 31P NMR spectra of
DMPC/A(1-40) molar ratios of 30:1, 20:1 and 10:1, respectively. Four peaks
can be distinguished whose positions vary with the angle, while an isotropic peak
remains almost at the same position independent of orientation angle (peak 3 in
305
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 5). In particular, the outermost pairs among the four peaks exhibit an
orientation-dependent displacement of peaks (peak 1 and 2 in Figure 5). Peak 4 in
Figure 5 represents the edge of the powder pattern of the lipid bilayers. In addition,
the separation of the outer and inner peaks varies with orientation, corresponding
to the effective chemical shift anisotropies (CSAs) of the phosphodiester moieties
( = // - ), as summarized in Table 1.
The anisotropy values of the outer and inner pairs decrease from 43.4 and
30.9 ppm, respectively, at a low proportion of A (1-40) to 38.0 and 22.9 ppm at
a higher ratio (Table 1). The most anisotropic component (i.e., most ordered) is
attributed to a lamellar bilayer, in agreement with the observed CSA values. The
component with reduced anisotropy is assigned to a different and therefore non-
lamellar structure, most likely the hexagonal H11 form (38), based on its orientation
dependence and the fact that the observed CSA value is approximately one-half the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

value measured for the bilayer structure (39, 40). The isotropic peak is assigned
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

to a disordered component showing rapid isotropic averaging. The most likely

origin of this peak is the micellar components. We note that the chemical shift of
the isotropic signal varies with alignment. This is not inconsistent with micellar
structure because micelle chemical shifts will still show orientation dependence
because of differences in bulk magnetic susceptibility.

Table 1. Chemical shift difference between the parallel and perpendicular

components of the 31P shift of DMPC lipid bilayer in the presence of A
(1-40)

Most significantly, the anisotropies of the ordered components are reduced

with increased proportion of A (1-40) (Figure 5, (a) to (b) to (c)). This observation
is consistent with insertion of A (1-40) into the membrane, because the observed
anisotropies are averages of the large anisotropy arising from free lipids and the
smaller anisotropy arising from lipids in direct contact with the peptide.

306
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 4. 1D 31P 1H-decoupled solid-state NMR spectra of oriented DMPC

samples at 35C. is the angle between the bilayer normal and B0 (see insert).
(a) Oriented DMPC bilayer sample with the order axis parallel to the magnetic
field. (b) Oriented DMPC bilayer sample with the order axis perpendicular to
the magnetic field. (c) and (d) Oriented bilayers of DMPC:GM1-ganglioside in
molar ratio 10:1 in molar ratio, with the order axis respectively parallel and
perpendicular to the magnetic field.

307
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 5. 31P solid-state NMR spectra of macroscopically oriented DMPC

samples incorporating varying ratios of A(1-40) at 35C as a function of the
angle between the bilayer and the magnetic field. DMPC/A(1-40) molar ratios
were (a) 30:1, (b) 20:1, (c)10:1 respectively. is the angle between the bilayer
normal and B0. Asterisk indicates an isotropic peak. From their intensities these
are identified as the parallel and perpendicular components, respectively.

Next, we measure the 31P angular-dependent NMR spectra of a range of

DMPC/ GM1 ganglioside systems in the presence of A (1-40), as shown in
Figure 6. No orientational distribution for DMPC molecules is observed at a
DMPC/GM1 ganglioside ratio of 5:1 (data not shown). Therefore, we decide to
perform an experiment using a ratio of DMPC to GM1 ganglioside of 10:1. In this
case, it is notable that the relative intensity of the isotropic peak is substantially
increased. With increasing concentration of A (1-40) peptide, the intensity of
the isotropic signal gradually increases further.
This result indicates that the lipid bilayer is gradually disrupted with
increasing A (1-40) peptide concentration, again highly suggestive of insertion
of A into the membrane.

308
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 6. 31P solid-state NMR spectra of macroscopically oriented DMPC/GM1

samples incorporating varying ratios of A(1-40) at 35C as a function of the
angle between the bilayer and the magnetic field. DMPC/GM1/A(1-40) molar
ratios were (a) 20:2:1 (b) 10:1:1 (c) 10:1:1.5 respectively. is the angle between
the bilayer normal and B0.

Figure 7. Changes of isotropic signal fraction of DMPC/A and DMPC/GM1/A

depend on lipid/A ratio. Triangle is the DMPC/A sample and square is the
DMPC/GM1/A sample.

309
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The fraction of the lamellar structure of DMPC decreases with respect to
that of DMPC without A(1-40) peptide. A similar decrease in the proportion
of oriented phospholipids on glass plates has also been reported in the presence
of other membrane-bound peptides (41). As expected, the chemical shifts of the
two oriented components are dependent on their angle relative to the magnetic
field. In the case of the DMPC/GM1/A system, the oriented components make
up only about 60% of the observed DMPC signal. It appears that the fraction of
isotropic component in DMPC/ A(1-40) peptide samples depends on the amount
of A(1-40) peptide present. However, the proportion of the isotropic lipids is
almost independent of lipid/peptide ratio when GM1 is present, as shown in Figure
7.
Figure 8 illustrates the 13C CP-MAS ((e) and (f)), DD-MAS ((c) and (d))
and solution NMR (b) spectra of 13C labeled A (1-40), DAEFRHDSGYEVH
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

HQKL[1-13C]V18FF[3-13C]A21EDVGSNK[2-13C]G29AIIGLMVGGVVI, in the
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

presence of DMPC bilayer, recorded both at 38 and 21C, and also the solution
NMR spectrum of sonicated lipid only (a). The 13C NMR spectrum of DMPC in
the sonicated single vesicle has very sharp signals and serves as a reference for the
peaks arising from lipid itself. The CP-MAS technique emphasizes signals from
the more rigidly structured components, whereas DD-MAS tends to emphasise
more mobile components (42, 43). It is noted that the 13C NMR signals of fatty
acyl chains at 32 and 30 ppm can be very conveniently utilized as a probe to
determine the state of the lipid phase as either the gel or liquid crystalline states,
respectively (44). Therefore, it is clear that the 13C CP-MAS NMR peaks of the
fatty acyl chains show the presence of a mixture of gel and liquid crystalline lipids
(6:4) at 38 C but gel phase alone at 21C.
At least three kinds of different forms can be distinguished for A (1-40)
peptide: random coil, -sheet, and transmembrane-type -helix structure; they can
be characterized by liquid state, CP-MAS and DD-MAS NMR, respectively, in the
presence of lipid bilayer as judged from the conformation-dependent displacement
of peaks (45, 46) The 13C CP-MAS NMR spectra recorded at 38C and 21C clearly
exhibit the characteristic peaks of a -sheet form as judged from the peak-positions
of Ala21 C at 20.3 ppm, and Val18 C=O at 170.1 ppm (Figure 8(e) and (f)). It is
therefore concluded that the A peptide is likely to exist in a mixture of monomeric
and oligomeric states in the presence of phospholipids: the A peptides which form
-sheet structure are likely to be oligomeric (i.e., the fibrils or protofibrils observed
in many other studies), whereas the A peptides which take on a small amount of
-helical structure are most likely to be monomeric. Moreover, it is suggested that
these structural differences usually exist in an each domain rather than in single
molecules.
It is noteworthy that very sharp isotropic Ala21 C, Gly29 C and Gly18 C=O
signals (Figure 8(b)) can be ascribed to A incorporated into micelles, the presence
of which is confirmed by an isotropic 31P NMR signal in the presence of this
protein (Figure 5). The Ala21 C signal at 16.5 ppm is characteristic of -helical
conformation and not random coil (16.7 ppm). Surprisingly, this signal, as well
as that from Gly29 C, which are clearly visible on liquid-state NMR, is almost
completely suppressed when the sample is analyzed with both CP-MAS and DD-
MAS NMR. This suppression is ascribed to a well-known phenomenon: the failure

310
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of the necessary high-power proton decoupling to produce high-resolution signals
when the incoherent fluctuation frequency of peptide chains in liquid crystalline
lipids interferes with the coherent frequency of proton decoupling (60 kHz) (45,
47). If so, it is expected that such suppressed peaks can be recovered, at least
partially, either by reducing the fluctuation frequency by lowering the temperature,
or by reducing the decoupling frequency to some extent, from the original 60 kHz
to 37 kHz, for instance. Consistent with this expectation, we found that the peak
intensity of Ala C is more evident when the decoupling frequency is reduced
in the DD-MAS (Figure 7(c)), and the Gly C signal is also more visible at low
temperature in the gel phase lipid, as viewed from the peak-position of fatty acyl
chain (Figure 8(a)).
We have shown in this paper that the bilayer structure is strongly perturbed
by the presence of A (1-40) peptide and also GM1, yielding a fraction of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

anisotropic but non-lamellar structure, suggested to be the hexagonal H11

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

structure. At the same time, the proportion of an isotropic lipid phase such as
micelle or single-vesicle increases, as determined from 31P NMR spectra. Such
perturbation to the bilayer is significantly more prominent when GM1 is present.
The assignment of the non-lamellar phases to H11 and micellar structures is based
partly on the observation that lipid polymorphic structure is strongly related to
the shape of lipids: a cylindrical lipid such as phosphatidylcholine favors bilayer
structure, while cone-like lipids such as (unsaturated) phosphatidylethanolamine,
cardiolipin and phosphatidic acid-Ca2+ prefer to take on a hexagonal H11 forms
(38). Furthermore, an inverted cone such as GM1 favors a micellar structure.
Non-lamellar lipid structures, such as that revealed here by 31P NMR, have
been reported for a number of biological membranes containing proteins such
as erythrocyte (ghost) (47), rat, bovine and rabbit liver microsomal membranes
(48, 49), and cytochrome P-450 membrane (49). Several peptides including
gramicidin, alamethicin, -helical peptides, and antimicrobial peptides have also
been shown to perturb bilayer structure to form non-lamellar hexagonal H11
lipids by peptide-induced changes in lipid phases (5055), in addition to the
effect of lipid molecules themselves as mentioned above. The hydrophobicity
and conformational flexibility of transmembrane peptides in lipid bilayers can
affect their propensity to induce the formation of inverted non-lamellar phases
by mechanisms not dependent on mainly lipid-peptide hydrophobic mismatch
(54), although we note that a theoretical interpretation of a mechanism of the
lamellar-to-inverted hexagonal phase transition has also been proposed, in relation
to the membrane fusion process (56, 57) For this reason, it is likely that the effect
of A (1-40) arises from interaction with the lipid bilayer, following an initial
penetration into the bilayer. The effect of ganglioside, however, on modulation of
lamellar-inverted micelle (H11) phase transition has been examined in relation to
biomembranes and shown to vary depending upon the relative concentration of
ganglioside to lipids (56).
Our 13C NMR study indicates that A (1-40) together with GM1 plays an
important role to disrupt bilayer structure to form non-lamellar lipid and micelle
by its penetration into bilayer and taking on an -helical form. As a result, the
Ala C 13C NMR peak was almost completely suppressed on both CP-MAS and
DD-MAS NMR spectra recorded at 38C but was partially recovered on DD-MAS

311
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 NMR spectra utilizing reduced proton decoupling (Figure 5(c)). This means that
the features of their 13C NMR spectra are strongly influenced by the fluidity of
lipid molecules.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

Figure 8. 13C CP-MAS and DD-MAS NMR spectra of selectively 13C-labeled A

(1-40) peptide incorporated into lipid bilayers. (a) lipid only, (b) liquid state
NMR at 38C, (c) DD-MAS NMR, using reduced high-power decoupling of 34
kHz at 38C, (d) DD-MAS NMR at 38C, (e) CP-MAS NMR at at 38C, and (f)
CP-MAS NMR at 21C. The molar ratio of DMPC : A peptide is 10:1 in each
spectra except of Figure 5(a).

This study is non-physiological in that the lamellar structure is allowed

to reform slowly in the presence of A. It may therefore be argued that our
observation of interaction of A with lipid tails is not relevant. In response,
we note that the current mixed film method is commonly used to incorporate
hydrophobic transmembrane peptides in lipid bilayers. Otherwise, such peptides
remain outside of the bilayer by forming -sheet aggregates. Indeed, Separovic
and co-workers have been shown that addition of A(1-42) to preformed bilayer
312
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 resulted in precipitate on the lipid surface with little effect on the hydrocarbon
region of the lipid bilayer and also found that 31P isotropic peak was formed
only when the peptide and lipid were co-solubilized for A(1-42) peptides. (13).
Therefore, the current mixed film method is very important in gaining insight into
whether or not the peptide is thermodynamically capable of being incorporated
into bilayers. Since the onset of Alzheimers is a very slow process, it is entirely
reasonable to argue that it is the thermodynamically rather than the kinetically
preferred product that will eventually prevail.
It is still premature to relate fibrillation of A (1-40) in vitro with subsequent
neurodegenerative disease. However, it is interesting to note that we observe
-sheet conformation of A, presumably on the lipid surface. GM1 is closely
associated with transmembrane signaling, leading to the interesting possibility that
aggregation and fibrillisation of A on the membrane surface could interfere with
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

signaling at synaptic junctions. Moreover, we have shown that the interaction with
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

ganglioside GM1 micelles is localized to the N-terminal region of the peptide,

particularly residues His13 to Leu17, which become more helical when bound as
referred to above (27). This is consistent with the present data, although here
the importance of interaction with fatty acyl chains is emphasized together with
formation of non-lamellar structure and micelle formation. This study has shown
that GM1 facilitates the membrane disruption caused by A peptide. Indeed, this
view is consistent with insertion of A peptide into lipid bilayer to cause disruption
of bilayer structure (17), although this is in contrast to an alternative view which
emphasizes a cholesterol-dependent cluster of GM1 in membranes (16).

Conclusion
In order to clarify the interaction between -amyloid and lipid bilayers
including ganglioside, we first demonstrate using solution state NMR on
15N-labelled A(140) and A(142) that the interaction with ganglioside GM1

micelles is localized to the N-terminal region of the peptide, particularly residues

His13 to Leu17, which become more helical when bound, leaving the C-terminal
region unstructured (27). Furthermore, the insertion of A peptide into lipid
bilayer to form a micellar structure is proposed on the basis of a monolayer
study (17). Secondly, we obtain 1H decoupled 31P solid-state NMR spectra
of dimyristoylphospatidylcholine (DMPC) and DMPC/GM1 ganglioside in a
mechanically oriented system in the presence of A(1-40) in order to clarify
the behavior of A in the presence of GM1 ganglioside. Finally, we record the
spectra from 13C-labeled A peptide by CP-MAS and DD-MAS NMR, as well as
low-power solution NMR to determine conformational features of the A peptide
in the presence of lipid bilayers.
In the case of solution NMR experiment, we demonstrate A binding in the
low micromolar range to GM1 and asialo-GM1, but no measurable binding to the
isolated GM1 headgroup. Several experiments (both ours and those of others (33))
demonstrate that although there may be a Coulombic element to the binding, the
interactions are not limited to Coulombic ones. Our results are therefore relevant

313
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 to the physiological situation where salt concentration is higher and Coulombic
interactions are weaker.
The chemical shift changes reported here demonstrate binding of A to GM1
micelles, which is localized to the N-terminal half of the peptide. We confirm that
binding to asialo-GM1 is weaker, specifically with differences around residues
1215 which include His residues. This conclusion is radically different from
an earlier NMR study [34]. However, that earlier study was carried out using
SDS(Sodium Dodecyl Sulfate) micelles which might explain the discrepancy.
Above all, we find that A (1-40) strongly perturbs the bilayer structure of
DMPC to form either a non-lamellar structure or micelles. In particular, GM1
accelerates the effect of A (1-40), as viewed from 31P NMR. Such effect may
arise from direct contact with fatty acyl chains of lipids by inserting A (1-40) into
the bilayer. The difference of the isotropic peak intensity between DMPC/A and
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

DMPC/GM1/A suggests the presence of a specific interaction between A and

Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

GM1 ganglioside. We show that in the DMPC/GM1/A system there are three
lipid phases, a completely aligned phase, a hexagonal phase, and non-oriented
lipids(or smaller vesicle structures (13)). The formation of latter two phases are
induced by the presence of the A peptide and facilitated by GM1.

Acknowledgments
This work has been supported by a Grant-in-Aid for Creative Scientific
Research of the Japan Society for the Promotion of Science (JSPS)

References
1. Iversen, L. L.; Mortishire-smith, R. J.; Pollack, S. J.; Shearman, M. S.
Biochem. J. 1995, 311, 116.
2. Glenner, G. G.; Wong, C. W. Biochem. Biophys. Res. Commun. 1984, 122,
11311135.
3. Masters, C. L.; Multhaup, G.; Simms, G.; Pottgiesser, J.; Martins, R. N.;
Beyreuther, K. EMBO J. 1985, 4, 27572763.
4. Kang, J.; Lemaire, H. G.; Unterbeck, A.; Salbaum, J. M.; Masters, C. L.;
Grzeschik, K. H.; Multhaup, G.; Beyreuther, K.; Muller-Hill, B. Nature 1987,
325, 733736.
5. Kirschner, D. A.; Abraham, C.; Selkoe, D. J. Proc. Natl. Acad. Sci. U.S.A.
1986, 83, 503507.
6. Petkova, A. T.; Yau, W. M.; Tycko, R. Biochemistry 2006, 45, 498512.
7. Petkova, A. T.; Ishii, Y.; Balbach, J. J.; Antzutkin, O. N.; Leapman, R.
D.; Delaglio, F.; Tycko, R. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
1674216747.
8. Lambert, M. P.; Barlow, A. K.; Chromy, B. A.; Edwards, C.; Freed, R.;
Liosatos, M.; Morgan, T. E.; Rozovsky, I.; Trommer, B.; Viola, K. L.;
Wals, P.; Zhang, C.; Finch, C. E.; Krafft, G. A.; Klein, W. L. Proc. Natl.
Acad. Sci. U.S.A. 1998, 95, 64486453.
314
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 9. Geula, C.; Wu, C. K.; Saroff, D.; Lorenzo, A.; Yuan, M.; Yankner, B. A. Nat.
Med. 1998, 4, 827831.
10. Muller, W. E.; Koch, S.; Eckert, A.; Hartmann, H.; Scheuer, K. Brain Res.
1995, 674, 133136.
11. Arispe, N.; Rojas, E.; Pollard, H. B. Proc. Natl. Acad. Sci. U.S.A. 1993, 90,
567571.
12. Ambroggio, E. E.; Kim, D. H.; Separovic, F.; Barrow, C. J.; Barnham, K. J.;
Bagatolli, L. A.; Fidelio, G. D. Biophys. J. 2005, 88, 27062713.
13. Lau, T. L.; Ambroggio, E. E.; Tew, D. J.; Cappai, R.; Masters, C. L.;
Fidelio, G. D.; Barnham, K. J.; Separovic, F. J. Mol. Biol. 2005.
14. de Planque, M. R.; Raussens, V.; Contera, S. A.; Rijkers, D. T.; Liskamp, R.
M.; Ruysschaert, J. M.; Ryan, J. F.; Separovic, F.; Watts, A. J. Mol. Biol.
2007, 368, 982997.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

15. Yanagisawa, K.; Odaka, A.; Suzuki, N.; Ihara, Y. Nat. Med. 1995, 1,
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

10621066.
16. Kakio, A.; Nishimoto, S.; Yanagisawa, K.; Kozutsumi, Y.; Matsuzaki, K.
Biochemistry 2002, 41, 73857390.
17. Chi, E. Y.; Frey, S. L.; Lee, K. Y. Biochemistry 2007, 46, 19131924.
18. Brocca, P.; Sonnino, S. Trends Glycosci. Glycotechnol. 1997, 9, 433445.
19. Kasahara, K.; Sanai, Y. Trends Glycosci. Glycotechnol. 2001, 74, 587594.
20. Nagai, Y. Behav. Brain Res. 1995, 66, 99104.
21. McLaurin, J.; Chakrabartty, A. J. Biol. Chem. 1996, 271, 2648226489.
22. McLaurin, J.; Franklin, T.; Fraser, P. E.; Chakrabartty, A. J. Biol. Chem.
1998, 273, 45064515.
23. McLaurin, J.; Chakrabartty, A. J. Biol. Chem. 1996, 271, 2648226489.
24. Choo-Smith, L. P.; Surewicz, W. K. FEBS Lett. 1997, 402, 9598.
25. Yanagisawa, K.; Ihara, Y. Neurobiol. Aging 1998, 19, S6567.
26. Kakio, A.; Nishimoto, S.; Kozutsumi, Y.; Matsuzaki, K. Biochem. Biophys.
Res. Commun. 2003, 303, 514518.
27. Williamson, M. P.; Suzuki, Y.; Bourne, N. T.; Asakura, T. Biochem. J. 2006,
397, 483490.
28. Hou, L.; Shao, H.; Zhang, Y.; Li, H.; Menon, N. K.; Neuhaus, E. B.;
Brewer, J. M.; Byeon, I. J.; Ray, D. G.; Vitek, M. P.; Iwashita, T.; Makula, R.
A.; Przybyla, A. B.; Zagorski, M. G. J. Am. Chem. Soc. 2004, 126,
19922005.
29. Bai, Y.; Milne, J. S.; Mayne, L.; Englander, S. W. Proteins 1993, 17, 7586.
30. Zhang, S.; Iwata, K.; Lachenmann, M. J.; Peng, J. W.; Li, S.; Stimson, E.
R.; Lu, Y.; Felix, A. M.; Maggio, J. E.; Lee, J. P. J. Struct. Biol. 2000, 130,
130141.
31. Basu, A.; Glew, R. H. J. Biol. Chem. 1985, 260, 1306713073.
32. Morrison, J.; Yang, J. C.; Stewart, M.; Neuhaus, D. J. Mol. Biol. 2003, 333,
587603.
33. Laguri, C.; Phillips-Jones, M. K.; Williamson, M. P. Nucleic Acids Res 2003,
31, 67786787.
34. Ma, K.; Clancy, E. L.; Zhang, Y.; Ray, D. G.; Wollenberg, K.; Zagorski, M.
G. J. Am. Chem. Soc. 1999, 121, 86988706.

315
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 35. Choo-Smith, L. P.; Garzon-Rodriguez, W.; Glabe, C. G.; Surewicz, W. K. J.
Biol. Chem. 1997, 272, 2298722990.
36. Bokvist, M.; Lindstrom, F.; Watts, A.; Grobner, G. J. Mol. Biol. 2004, 335,
10391049.
37. Ariga, T.; Kobayashi, K.; Hasegawa, A.; Kiso, M.; Ishida, H.; Miyatake, T.
Arch. Biochem. Biophys. 2001, 388, 225230.
38. Cullis, P. R.; de Kruijff, B. Biochim. Biophys. Acta 1979, 559, 399420.
39. Thayer, A. M.; Kohler, S. J. Biochemistry 1981, 20, 68316834.
40. Smith, I. C. P.; Ekiel, I. H. Phosphorus-31 NMR; Academic Press: New York,
1984.
41. Moll, F., 3rd; Cross, T. A. Biophys. J. 1990, 57, 351362.
42. Tuzi, S.; Naito, A.; Saito, H. Eur. J. Biochem. 1993, 218, 837844.
43. Kamihira, M.; Naito, A.; Tuzi, S.; Nosaka, A. Y.; Saito, H. Protein Sci. 2000,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch018

9, 867877.
Downloaded by DUKE UNIV on June 20, 2012 | http://pubs.acs.org

44. Kimura, S.; Naito, A.; Tuzi, S.; Saito, H. Biopolymers 2001, 58, 7888.
45. Saito, H.; Tuji, S.; Tanio, M.; Naito, A. Annu. Rep. NMR Spectrosc. 2002,
47, 39108.
46. Saito, H. Magn. Reson. Chem. 1986, 24, 835852.
47. Rothwell, W. P.; Waugh, J. S. J. Chem. Phys. 1981, 74, 27212732.
48. Davis, J. H. Biochim. Biophys. Acta 1983, 737, 117171.
49. Stier, A.; Finch, S. A.; Bosterling, B. FEBS Lett. 1978, 91, 109112.
50. Gasset, M.; Killian, J. A.; Tournois, H.; de Kruijff, B. Biochim. Biophys.
Acta 1988, 939, 7988.
51. Cornell, B. A.; Separovic, F. Eur. Biophys. J. 1988, 16, 299306.
52. Keller, S. L.; Gruner, S. M.; Gawrisch, K. Biochim. Biophys. Acta 1996,
1278, 241246.
53. Killian, J. A.; de Kruijff, B. Biochemistry 1985, 24, 78907898.
54. Liu, F.; Lewis, R. N.; Hodges, R. S.; McElhaney, R. N. Biochemistry 2001,
40, 760768.
55. Hallock, K. J.; Lee, D. K.; Ramamoorthy, A. Biophys. J. 2003, 84,
30523060.
56. Siegel, D. P. Biophys. J. 1986, 49, 11551170.
57. Siegel, D. P.; Epand, R. M. Biophys. J. 1997, 73, 30893111.

316
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 19

Characterization of Microbial Poly(-L-lysine)

and Its Derivatives by Solid-State NMR
Shiro Maeda,*,1 Chizuru Sasaki,2 and Ko-Ki Kunimoto3
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

1Division of Applied Chemistry and Biotechnology, Graduate School of

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Engineering, University of Fukui, Japan

2Department of Life System, Institute of Technology and Science,

The University of Tokushima, Japan

3Division of Material Engineering, Graduate School of Natural Science

and Technology, Kanazawa University, Japan

*E-mail: [email protected]

The molecular structure and conformation of microbial

poly(-L-lysine) (-PL) and its derivatives in aqueous solution
and in the solid-state have been studied with liquid-state 1H
and 13C NMR, and solid-state 13C and 15N NMR, together with
FT-IR, FT-Raman, and circular dichroism (CD) techniques.
CD and vibrational spectroscopic results indicate that -PL
assumes a -sheet conformation in aqueous solution and in
the solid state. Solid-state 13C NMR spectra of the crystalline
and the amorphous components were obtained separately by
utilizing difference in 13C spin-lattice relaxation times between
the two components. The degree of crystallinity was estimated
to be 63%. Conformation models of the crystalline component
were also discussed. Chemically modified derivatives of -PL,
-PL/MO and -PL/DC were prepared through reactions with
methyl orange (MO) and dabsyl chloride (DC), respectively.
Characterization of these derivatives was carried out by
solid-state 13C and 15N NMR. In -PL/MO side chain -amino
groups of -PL are involved in ionic bonds with methyl orange
(MO) to form poly-ionic complexes, (-PL)-NH3+ SO3-(MO).
In contrast, the -amino groups react with DC in -PL/DC to
form covalent sulfonamide bonds, (-PL)-NH-SO2-(DC). The
15N NMR data are consistent with these structures.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Introduction
In recent years, there has been considerable interest in biopolymers
because of concern over the environmental impacts arising from the disposal
of petroleum-based plastics (1). Biologically derived polymers have been
extensively investigated as biodegradable polymers as a means of reducing
environmental impact (2). Natural poly(amino acid)s are a class of biodegradable
polymers mainly composed of one type of amino acid. Three poly(amino acid)s
which are known to occur in nature: poly(-L-lysine), Poly(-glutamic acid), and
cyanophycin (3, 4). Microbial production of -PL was first discovered by Shima
and Sakai from a culture filtrate of Streptomyces albulus (5). They studied the
detailed fermentation conditions and reported the physicochemical properties of
microbial -PL (68). The -amino group of L-lysine is linked to the -carboxyl
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

group to form a peptide bond in a -PL molecule. (Scheme 1) Chemical synthesis

of -PL was also reported by Kushwaha et al. (9). Owing to antibacterial activities
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

against a broad spectrum of microorganisms, -PL has found application as a

preservative for various food products (10).

Scheme 1. Repeating units of -PL.

-PL is water-soluble and biodegradable, in addition to exhibiting

antibacterial activities. An understanding of the structure and the conformation
is prerequisite for functionalization of -PL (11). We have studied the molecular
structure and the conformation of -PL in aqueous solution (1214) and those
of dabsylated -PL in aqueous DMSO solution (15). It is conceivable that
the four methylene groups in the backbone of -PL endow the polymer with
considerable conformational flexibility compared to the case of poly(-L-lysine).
The pH dependent IR and CD spectra have indicated that -PL assumes a -sheet
conformation in aqueous solution, the content of which is dependent on the chain
length and the pH (14). We have characterized the structure and the conformation
of -PL in the solid-state for the first time by FT-IR, FT-Raman and 13C solid-state
NMR (16). Since the primary structure of -PL is related to nylon 6, the spectral
features were compared with those of nylon 6. On the basis of these spectroscopic
data, a conformation model was proposed for the crystalline component of -PL.
We also carried out structural investigation of -PLazo dye derivatives using
solid-state 13C and 15N NMR (17). Polymers containing azobenzene moieties in
a side chain have potential application as an optical materials (18, 19). When
methyl orange (MO, 4-dimethylamino-azobenzene-4-sulfonic acid) is added to
an aqueous solution of -PL, the -PL/dye complex precipitates out of solution.
This procedure has been generally utilized for the determination of free amino
groups in peptides and proteins (20). Likewise, -PL also forms a water-insoluble
compound in the reaction with an amino labeling reagent, dabsyl chloride (DC,
4-dimethylaminoazobenzene-4-sulfonyl chloride). 13C and 15N solid-state NMR
318
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 measurements of -PL and its derivatives have been reported, and molecular
interaction between -PL and the dye discussed.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Scheme 2. Preparation of -PL derivatives.

Experimental Section
Materials
Microbial -PL (free form, -PL) was produced from the culture broth of a
variant of Streptomyces albulus (No. 346) by Chisso Corporation, Japan according
to a reported procedure (21). Methyl orange (MO), dabsyl chloride (DC) and
other reagents were purchased from Tokyo Kasei Kogyo, Japan, and used without
further purification. The HCl salt form of -PL, -PL/HCl was prepared with
hydrochloric acid according to a reported procedure (13). -PL/MO and -PL/DC
(Scheme 2) were prepared as follows. -PL (26 mg, 0.2 mmol) and MO (65.4 mg,
0.2 mmol) were mixed and dissolved in 400 ml of 0.1 M phosphate buffer solution
(pH 6.0) and the pH was adjusted to 6.0-7.0 by adding 0.1-1.0 M hydrochloric
acid. The molar ratio of [MO]/[lysyl residue] used here corresponds to 1.0. The
solution was allowed to react for 4 h at room temperature. When the solution
volume was reduced to about 150 ml by evaporation, a precipitate was formed.
After centrifugation for 15 min, the orange products were taken out and dried in
vacuo. -PL/DC was prepared according to the conventional Schotten-Baumann
319
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 reaction. To a 100 ml MeOH solution of dabsyl chloride (162 mg, 0.5 mmol)
containing triethylamine (152 mg, 1.5 mmol), 64 mg of -PL (0.5 mmol) was
added. The mixture was allowed to stand at room temperature for 20 h. The
resulted reddish-orange precipitate was collected by centrifugation. The product
was washed with distilled water several times. -PL/Boc (t-butoxycarbonyl)
(Scheme 2) was also prepared through tert-butoxycarbonylation using Boc
anhydride as a reference compound, which is a completely N-substituted -PL
derivative. The degree of chemical modification was estimated from 1H NMR
spectra of DMSO-d6 solutions, by using integral values obtained from the number
of -CH2- protons of -PL and appropriate protons of the chemical modifier. The
apparent degree of chemical modification was 90 % for -PL/MO and 83 % for
-PL/DC.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Measurements and Instruments

Nuclear Magnetic Resonance (NMR)

High-resolution solid-state 13C and 15N NMR spectra were measured

with Chemagnetics CMX Infinity 300 operating at 75.6 MHz and 30.0 MHz,
respectively, at room temperature. A 62.6 kHz r.f. field strength was used for a
proton 90 pulse, with cross polarization and decoupling. A sample in powder
form was contained in a cylindrical rotor of zirconia ceramic. The rotor diameter
was 5 mm for 13C and 7.5 mm for 15N, and the rotor was spun at 7.0 kHz and
5.0 kHz, respectively. Contact time was 1ms, and repetition time was 2 sec.
The number of scans was about 12,000 for 13C and about 40,000 for 15N. 13C
signal of methyl carbon of hexamethylbenzene was externally referenced to 17.35
ppm. 15N signal of glycine was externally referenced to 32.5 ppm from ammonia
(liq. NH3, 25C). 1H, 13C, 1H-1H COSY and 1H-13C COSY NMR spectra of
-PL were measured either in trifluoroacetic acid-h (TFA-h), TFA-d or D2O
solutions with a JEOL JNM-GSX 400 NMR instrument at room temperature.
Chemical shifts were calibrated through an internal standard, TMS or sodium
2,2-dimethyl-2-silapentane-5-sulfonate-d6 (DSS).

IR, Raman, and Circular Dichroism (CD) Spectroscopies

The IR spectra were recorded on a Perkin Elmer 1650 FT-IR spectrometer by

averaging 64 scans with a resolution of 4 cm-1. The spectra were measured as KBr
pellets and nujol mulls. The FT-Raman spectra were obtained on a Perkin-Elmer
2000R spectrometer equipped with a quartz beam splitter and InGaAs detector.
The 1064 nm line of a Spectron Laser System SL300 Nd: YAG laser was used as
the exciting source with an output power of about 200 mW at the sample position.
All spectra were accumulated for 60 scans with a resolution of 4 cm-1. CD spectra
were measured with a JASCO J-600 spectropolarimeter by courtesy of JASCO
Ltd.

320
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Gel Permeation Chromatography (GPC)

Molecular weights and their distributions were determined using the standard
set-up of GPC-LALLS (Gel Permeation Chromatography combined with a Low
Angle Laser Light Scattering detector, Waters model ALC/GPC 244). Ion-pair
chromatography of -PL was carried out on a reverse-phase ODS column (L-
Column, Chemicals Evaluation and Research Institute, Japan).

Results and Discussion

Physical Properties of -PL
-PL is a white crystalline powder that shows a glass transition temperature
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

(Tg) at 88C and a melting point (Tm) at 172.8C. These values are comparable
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

with the typical values of nylon 6 (Tg 56 C, Tm 221 C). Figure 1 shows a HPLC
ion-pair chromatogram of -PL. The chromatogram is clearly resolved according
to the degree of polymerization (n). -PL exhibits a train of peaks where the major
peak population occurs at retention time around 40 min. Using the pentamer and
the decamer as standards for the n value, it is shown that -PL has the maximum
distribution in the n = 25-32 range. The number-average molecular weight (Mn)
of -PL was determined to be 4,090, which corresponds to n = 32 based on the unit
molecular weight of 128. -PL has a relatively narrow molecular weight

Figure 1. HPLC ion pair chromatogram of -PL. (Reproduced with permission

from reference (16). Copyright 2003 Elsevier.)

Characterization of -PL in Aqueous Solution

Since -PL is water-soluble, we examined the molecular structure and the
conformation of -PL in water (1214). Secondary structure of -PL is expected
to change with the dissociation state of side chain -amino groups and end groups.
We measured the pH (or pD) dependence of NMR spectra of -PL, and a series
of -PL oligomers in order to investigate molecular structure dependence on the
321
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 degree of polymerization. Kushwaha et al.(9) reported the chemical synthesis of
-PL and discussed the conformation of the polymer in aqueous solution, although
Mn and Mw/Mn of the polymer were not determined. Based mainly on the pH
dependence of CD spectra, they concluded that -PL took up an electrostatically
expanded conformation due to repulsion of protonated -amino groups at acidic
pHs, whereas at basic pHs the conformation changed to something similar to an
antiparallel -sheet. The plots of the []215 versus the pH values give sigmoidal
curves for all the oligopeptides and the midpoints of the curves coincide with the
reported pKa values of the -amino groups (pKa=8.5). The spectral similarity
between oligo(-L-lysine)and -PL suggests that the same conformational change
occurs with a pH variation for both cases. As shown in Figure 2, the microbial -PL
exhibited a similar pH dependence in the CD spectra as that reported by Kushwaha
et al. (9).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

The result suggests that a similar conformation change occurred in the

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

microbially produced -PL with pH changes. We synthesized a series of

monodispersed oligo(-L-lysine)mers (2-, 4-, 6-, and 8-mers). The pH variation
study of the CD spectra suggested that the oligopeptides exhibited a -sheet
conformation in aqueous solution at a pH above the pKa of the -amino group.
The CD band intensity at 215 nm for the high pH solution was dependent
on the number of lysine residues in the oligopeptides, indicating an increased
interchromophore interaction in the higher oligopeptides. IR spectra measured in
D2O also supported the pH-induced conformational change for the oligopeptides
and -PL (14).

Figure 2. CD curves of -PL at various pH. (Reproduced with permission from

reference (12). Copyright 1991 Kinki Chemical Society, Japan.)

322
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 3. 1H NMR of microbial and fractionated -PL in D2O solution. (N-),

(non-) and (C-) designate resonances of the N-, non- and C-terminal L-lysine
residues, respectively. (Reproduced with permission from reference (13).
Copyright 1993 Kinki Chemical Society, Japan.)

Another evidence for the conformational change comes from the pH titration
of 1H and 13C NMR spectra. Figure 3 shows the 1H NMR spectra of the oligomers
together with that of microbial -PL. All the 1H and 13C resonances can be readily
assigned using the off-diagonal peaks in the 1H-1H and 1H-13C COSY spectra: 1H
NMR (pD 6.4) =1.40(H), 1.59(H), 1.88 (H), 3.25 (H), 3.93 (H); 13C NMR
(pD 6.4) =24.5 (C), 30.7 (C), 33.4 (C), 41.9 (C), 56.0 (C), 172.6 (C=O).
Figure 4 shows pD dependence of 1H NMR resonances of the 8-mer. Other
oligomers exhibit similar pH dependence. A characteristic feature of the spectra
of -PL and the oligomers are the signals from N-, C- and non-terminal L-lysine
residues. There is only one H(C-) per -PL molecule which is adjacent to the
COOH moiety. This proton appears at slightly upfield position compared with
H(N-, non-) which is adjacent to the amide linkage. Therefore, the ratio I[H
(N-, non-)]/I[H(C-)] plus 1 should correspond to the degree of polymerization.
The ratios of the signal intensities, not only I[H (N-, non-)]/I[H(C-)], but also
I[H(C-, non-)]/I[H(N-)] and I[H (C-, non-)]/I[H(N-)] are correlated with
the degree of polymerization of -PL. The pKa values of the ionizable groups
323
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 can be estimated from these curves; pKa(COOH)<3, pKa(C-NH2)N-,non-term. 8,
pKa(C-NH2)C-term. 10, pKa(C-NH2)N-term. >11. These results suggest that the
dissociation state of the -NH2 groups plays an important role in the conformation
of the polymer.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 4. pD dependence of 1H NMR resonances of 8-mer. (Reproduced with

permission from reference (13). Copyright 1993 Kinki Chemical Society, Japan.)

The 1H resonances of the H, H and H protons neighboring the ionizable -

amino group are sensitive to pH change of solution. They showed high-field shifts
upon increasing the pD from 7.4 to 10.4. Similarly, 13C resonances of the C=O and
the C carbons of -PL showed low-field shifts upon the same pD change. Both
pD titration curves of 1H and 13C NMR spectra showed mid points at pD ca. 9.0.
Furthermore, 13C spin-lattice relaxation times (T1) of all carbons rapidly decreased
upon increasing the pD from 7.4 to 10.4: T1 values (sec) at pD.7.4:C (0.51), C
(0.33), C ( 0.37), C(0.34), C(0.38), C=O (2.93); at pD10.4: C (0.42), C
(0.26), C ( 0.29), C (0.31), C (0.27), C=O (2.67). Saito and Smith studied the
helix-coil transition of poly(-L-lysine) hydrochloride in aqueous solution by 13C
NMR. They found that 13C spin-lattice relaxation times T1 of the carbonyl and
the side-chain carbons decrease sharply at pD 10.2 which is the midpoint of the
transition from the random-coil to the -helix (22). The decreases in T1 values
can be interpreted in terms of a conformational change. The NMR results together
with the CD data support the pH-induced conformational change when pH exceeds
the pKa of -amino groups.

Characterization of -PL in the Solid State

IR and Raman Measurements

-PL samples are highly fluorescent due to impurities, and Raman spectra
obtained with visible excitation contain an enormous fluorescence background.
324
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 FT Raman spectra with near-IR excitation at 1,064 nm were measured in order to
eliminate fluorescence problems (23). Figure 5 shows IR and FT Raman spectra
of -PL. Typically -polyamides exhibit vibrational bands characteristic of the
secondary amide group. It is reasonable to assign the IR and Raman bands of
-PL in comparison with -polyamino acids. In the IR spectrum of -PL, the
NH2 asymmetric stretching is observed at 3382 cm-1. The NH stretching band
of the amide group usually appears as a doublet called amide A and amide B
bands. These bands originate from Fermi resonance between the first overtone of
amide II and the N-H stretching vibration and are observed at 3326 and 3081 cm-1,
respectively. Similar frequencies are obtained for the Raman spectra, although
the amide B is unobserved. The low frequency a(NH2) and the amide A bands
indicate that both the NH2 and the amide NH groups are involved in intra/inter
molecular hydrogen bondings. In general, the amide I, II and III modes for -
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

polypeptides are observed at 1680-1630 cm-1, 1570-1510 cm-1 and 1300-1250

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

cm-1, respectively (24). Since the amide bands are affected by the dipole-dipole
interaction between neighboring amide groups, their frequencies and intensities
are sensitive to the chain conformation (25). The amide I and II bands of -PL are
observed at 1633 and 1534 cm-1, both of which fall within the -sheet conformation
region. If we assume that the spectral region for -peptides also holds for -
peptides, -PL has a -sheet conformation in the solid state. Nylon 6, structurally
related to -PL, is known to assume alpha and gamma crystalline forms where the
amide chains participate in the intermolecular hydrogen bonds in either antiparallel
or parallel -sheet form (26, 27). Frequencies and intensities of the amide bands
in -PL are quite similar to those of nylon 6. This spectral similarity also supports
the -sheet conformation for -PL.

Figure 5. IR (a) and Raman (b) spectra of -PL. (Reproduced with permission
from reference (16). Copyright 2003 Elsevier.)

325
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Solid-State 13C NMR Measurements

13C
CPMAS NMR spectra of -PL are shown in Figure 6(a). Peaks are
well resolved and assigned based on the 1H-1H COSY and the 1H-13C HETCOR
measurements in solution. The chemical shifts and their assignments are
summarized in Table 1. The reported data for nylon 6 are also listed.

Table 1. 13C NMR chemical shifts (ppm) of -PL. (Reproduced with

permission from reference (16). Copyright 2003 Elsevier.)
-PL nylon 6a
carbon type in D2O in the solidb -form -form
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

C=O 180.0 178.6 173.4 173.0

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

177.1
C 57.0 57.9 36.7 37.8
C 36.7 38.0 26.5 30.1
C 25.0 27.9 30.4 30.1
C 30.9 34.0 30.4 34.1
32.9
C 41.6 40.5 43.6 39.9
aTaken from ref. (33). b Peak maxima in the T1CP crystalline component discrimination
experiment.

We have demonstrated through CD and NMR measurements (14) that

the amide groups of -PL are involved in intermolecular hydrogen bonds
with the amide groups of the neighboring molecules, and -PL assumes the
-sheet conformation in D2O solution. Similarity in chemical shifts between
the solution and solid 13C NMR spectra implies that -PL also takes on the
-sheet conformation in the solid state. The C=O and the C signals of the solid
are observed as doublets in contrast with the solution spectrum, which may be
attributed to the co-existence of the two crystal forms or the existence of the two
magnetically inequivalent sites in the asymmetric cell. Hydrogen bonds between
the amide groups in the -sheet structure lead to down field shift of the C=O
carbon signal. There appears to be a small peak at 164 ppm in Figure 6(a). Dos
et al. found that amino groups of poly(-L-lysine) can react with atmospheric
CO2 in aqueous and nonaqueous environments to give the carbamates and they
assigned a peak at 164 ppm in 13C spectrum to C=O carbon of carbamates (28).
Recently, we observed that amino groups of -PL also react with CO2 in
aqueous solution to give the carbamates. Detailed discussion will be published
elsewhere.

326
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Polymer Morphology of -PL

13Cspin-lattice relaxation time in the laboratory frame (T1) can provide

information about polymer morphology. The signal intensities of the respective
carbon are plotted against the relaxation delay time in Figure 7. The decay
curves were fitted to a biexponential function decay, M01 exp(-/T11) + M02
exp(-/T12). Table 2 summarizes the observed T1 values.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 6. 13C CPMAS NMR spectra of -PL. (a) normal spectrum, (b) crystalline
component, and (c) amorphous component. Pulse sequences used are normal
CPMAS for (a), T1CP with7 s delay time for (b), and saturation recovery with
200 ms delay time for (c). A peak marked with an asterisk is a spining side band.
(Reproduced with permission from reference (16). Copyright 2003 Elsevier.)

327
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 7. Plots of peak intensities versus delay times in the 13C T1CP experiment
for -PL. C; C; C; C (34.0 ppm); C (32.9 ppm); C.
(Reproduced with permission from reference (16). Copyright 2003 Elsevier.)

Table 2. 13C spin-lattice relaxation time in the laboratory frame, T1 (s) of

-PL. (Reproduced with permission from reference (16). Copyright 2003
Elsevier.)
carbon type
Component
amorphous 1.81 1.13 1.03 1.07 0.84 1.14
crystalline 25.4 48.8 57.6 57.8 66.1 59.1

Here, shorter relaxation times are those of the amorphous component and
longer ones correspond to the crystalline component. The degree of crystallinity
of -PL was estimated to be 63 % using fit parameters of the decay curve for
C shown in Figure 7. By using the T1CP experiment with the delay time of
7 s and the saturation recovery experiment (29) with the delay time of 200 ms,
signals from the crystalline and the amorphous components can be separately
obtained as shown in Figure 6(b) and 6(c). The amorphous spectrum has much
broader resonances than the crystalline spectrum. This is due to the heterogeneity
of the molecular conformations in the amorphous components. Furthermore,
chemical shifts of the amorphous component are observed upfield compared to
the crystalline one. The upfield shifts and the broadening of the CH2 resonances
328
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 are explained by the conformational heterogeneity and the -gauche effects in the
amorphous components. Recently, Asano et al. have investigated the crystallinity
and the size of crystalline component of -PL in detail (30).

Chain Conformation of -PL

It is useful to compare the chain conformations of -PL and nylon 6. The

stable -form of nylon 6 has trans zigzag conformation, and intermolecular
hydrogen bonds are formed in an anti-parallel sense. The less stable -form
has intermolecular hydrogen bonds in a parallel sense. The principal structural
difference between the - and -forms is that the amide-to-methylene dihedral
angles are nearly trans in the alpha and nearly perpendicular to the peptide plane
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

in the -form. As previously noted, the amide groups of -PL are involved in
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

intermolecular hydrogen bonds with those of the neighboring molecules and

-PL assumes the -sheet conformation in the solid state. As shown in Table
1, chemical shifts are significantly deshielded for the C=O, - and -carbons of
-PL compared with those of nylon 6. These chemical shifts can be interpreted
in terms of the substituent effects of -NH2 group. In contrast, the C carbon of
-PL is shielded due to the through space -gauche effect of the -NH2 group
(31). Hydrogen bonds between the amide groups in the -sheet structure lead to
down field shift of the C=O carbon signal. The large down field displacement of
the C=O of -PL may be due to a combination of the NH2 substituent effect and
the hydrogen bond effect. The chemical shift values of C and C are similar and
close to those of -form nylon 6.
A conformation model for -PL described on the basis of the -form of
nylon 6 is shown in Figure 8. Sasaki et al. have carried out X-ray analysis of
powder sample of -PL using linked-atom Rietveld simulation (32). Based on
the preliminary simulation, they tentatively assigned the -PL chain conformation
to the parallel -sheet similar to the -form of nylon 6. Their assignment is
consistent with the present solid state 13C NMR results.

Figure 8. A conformation model of -PL based on the parallel -sheet model of

the gamma form of nylon 6. (Reproduced with permission from reference (16).
Copyright 2003 Elsevier.)

329
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Characterization of -PL Derivatives with Azo Dyes

Solid-State 13C NMR Measurements of -PL Derivatives

Figure 9 shows the 13C CPMAS NMR spectra of -PL and its azo dye
derivatives. 13C chemical shifts are summarized in Table 3 with those for -PL
and -PL/HCl. The C=O carbon signal of -PL/HCl shifts to upfield about 6 ppm
compared to that of -PL. Signals of other chain carbons, except C, also show
upfield shift. These chemical shift displacements are interpreted mainly as the
substituent effect of the NH3+ group upon protonation.
The spectral pattern of -PL/MO in the aliphatic carbon region is similar
to that of -PL/HCl. The C peak of -PL/MO and the N-methyl carbon signal
of MO molecule overlap. The peaks in the 112-155 ppm region are due to the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

aromatic carbons of -PL/MO. The signals of the -PL/DC in the backbone

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

CH2 region (20-40ppm) are much less resolved. The broad and unresolved CH2
peaks suggest that the conformational inhomogeneity of the CH2 chain is due to
the existence of bulky aromatic side chain groups. Since the aromatic carbon
signals of -PL/MO and -PL/DC overlap one another, it is difficult to distinguish
covalently bonded sulfonamide linkage from poly-ionic complex formation. An
amine-labeling reagent (DC) is expected to be covalently bonded to -amino
group of -PL to make sulfonamide bond, (-PL)-NH-SO2-(DC), whereas MO is
expected to be negatively ionized to form ionic bond, (-PL)-NH3+SO3-(MO).

Table 3. 13C NMR chemical shifts of -PL and its derivatives in the solid
statea). (Reproduced with permission from reference (17). Copyright 2005
Springer.)

330
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 9. 13C CPMAS NMR spectra of (a) -PL, (b) -PL/HCl, (c) -PL/BocC,
(d) -PL/MO and (e) -PL/DC. (Reproduced with permission from reference
(17). Copyright 2005 Springer.)

Solid-State 15N NMR Measurements of -PL Derivatives

Figure 10 shows the 15N CPMAS spectra of -PL and its azo dye derivatives
(17). The chemical shifts and their assignments are summarized in Table 4. -PL
exhibits two distinct 15N peaks at 117 and 26.6 ppm. The former and the latter
peaks are assigned to the main chain NHCO and the -NH2 groups in comparison
with nylon 6 (33), and chitosan and chitin (3436). In -PL/HCl, the NHCO and
the -NH2 peaks are observed at 123 and 44 ppm, respectively. These downfield
shifts are attributed to protonation of the -amino group to form -NH3+Cl- salt in

331
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 -PL/HCl. Similar shift is observed in chitosan blend with poly(acrylic acid) where
2 % formic acid aqueous solution is used as a solvent (34, 35). -PL/MO shows
three 15N peaks at 115, 34.1 and 59.9 ppm. A peak at 59.9 ppm is assigned to
dimethyl amino nitrogen of the dye moiety (37, 38). 15N peaks at 115 and 34.1
are assigned to the main chain CONH and the -NH3+ group, respectively. The
15N chemical shift of the -NH3+ is between those of -PL and -PL/HCl. Since

analogous chemical shift is reported for chitosan/poly(acrylic acid) (34, 35, 39),
the formation of poly-ion complex between -PL and MO, (-PL)--NH3+ SO3-
(MO) is clearly indicated. The azo 15N signals are expected around 510 ppm (37,
38), but no such peaks are observed in this region. -PL/DC shows 15N peaks at
121, 102, 61.6 ppm. A peak at 60.6 ppm is assigned to the dimethyl amino group.
15N peaks at 121 and 102 ppm are assigned to the main chain CONH and the

sulfonamide groups, suggesting the formation of the covalent (-PL)--NHSO2-

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

(DC) bond.
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Figure 10. 15N CPMAS NMR spectra of (a) -PL, (b) -PL/HCl, (c) -PL/Boc,
(d) -PL/MO and (e) -PL/DC. (Reproduced with permission from reference
(17). Copyright 2005 Springer.)

332
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 4. 15N NMR chemical shifts of -PL and its derivatives in the solid
statea. (Reproduced with permission from reference (17). Copyright 2005
Springer.)
-NH3+
-
sample -NHCO -NHSO2 -NHCO ion- -NH2 N(CH3)2
free
complex
Nylon 6 b 116.5
-PL 117 26.6
-PL/HCl 123 44.0
-PL/Boc 120 92.2
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

-PL/MO 115 34.1 59.9

-PL/DC 121, 113 102 32.1 61.6
Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

Chitosan c 21.2
CS/PAA d 31.6
a In ppm with respect to liquid NH3 at 25C. b
-crystalline form (33). c References
(3436). d Chitosan/poly(acrylic acid) (34, 35, 39).

A peak at about 32 ppm which corresponds to the -NH3+ group implies that
there exists a small amount of -PL/MO in -PL/DC. Curve fitting for the -PL/DC
spectrum (Figure 11) reveals a peak at 113 ppm which can be assigned to a signal
of main chain amide group, -NHCO-CH(-NH3+) (40). We conclude that the
-PL/DC sample contains a small amount of ion complexes with MO.

Figure 11. Curve fitting for 15N NMR spectrum of -PL/DC. (Reproduced with
permission from reference (17). Copyright 2005 Springer.)

Conclusion
In this work characterization of -PL was done by using various spectroscopic
measurements, e.g., CD, IR, Raman, solution and solid NMR. The pH dependent
333
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 IR and CD spectra indicated that -PL assumes a -sheet conformation in
aqueous solution, the content of which depends on chain length and pH. 13C
spin-lattice relaxation times for solid -PL were measured to yield two kinds of
T1s corresponding to the crystalline and amorphous components: the degree of
crystallinity was estimated to be 63%. Solid 13C NMR spectra of the crystalline
and the amorphous components were observed separately. The amorphous
components of the backbone give rise to higher field signals due to the -gauche
effect. A conformation model for -PL was presented where the backbone amide
groups participate in inter-chain hydrogen bonds similar to the -form nylon 6.
The -PL derivatives with azo dyes were investigated by using solid 13C and
15N NMR. Side chain -amino group of -PL does not make a covalent bond

with methyl orange (MO) but forms poly-ionic complex, (-PL)-NH3+SO3(MO).

In comparison, dabsyl chloride (DC) makes covalent bond with -PL to form
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

sulfonamide, (-PL)-NH-SO2-(DC). A small amount of DC changes to MO by

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

hydrolysis to form the poly-ion complex, (-PL)-NH3+ SO3 (MO).

Acknowledgments
Microbial poly(-L-lysine) was kindly supplied by Chisso Corporation. The
authors are grateful to Professors Akio Kuwae and Kazuhiko Hanai for their
helpful and valuable cooperation through this work.

References
1. Griffith, L. G. Acta Mater. 2000, 48, 263277.
2. Nair, L. S.; Laurencin, C. T. Prog. Polym. Sci. 2007, 32, 762798.
3. Shih, I-L.; Shen, M-H.; Van, Y-T. Bioresour. Technol. 2006, 97, 11481159.
4. Oppermann-Sanio, F. B.; Steinbuchel, A. Naturwissenschaften 2002, 89,
1122.
5. Shima, S.; Sakai, H. Agric. Biol. Chem. 1977, 41, 18071809.
6. Shima, S.; Sakai, H. Agric. Biol. Chem. 1981, 45, 24972502.
7. Shima, S.; Sakai, H. Agric. Biol. Chem. 1981, 45, 25032508.
8. Shima, S.; Fukuhara, Y.; Sakai, H. Agric. Biol. Chem. 1982, 46, 19171919.
9. Kushwaha, D. R. S.; Mathur, K. B.; Balasubramanian, D. Biopolymers 1980,
19, 219229.
10. Ho, Y. T.; Ishizaki, S.; Tanaka, M. Food Chem. 2000, 68, 449455.
11. Maeda, S.; Mori, T.; Kunimoto, K.-K.; Sasaki, C. Kobunshi Kako 2003, 52,
516522(in Japanese).
12. Lee, H.-H.; Oyama, K.; Hiraki, J.; Hatakeyama, M.; Kurokawa, Y.;
Morita, H. Chem. Express 1991, 6, 683686.
13. Fukushi, H.; Oyama, K.; Hatakeyama, M.; Hiraki, J.; Fujimori, D.; Lee, H.-
H. Chem. Express 1993, 8, 745748.
14. Lee, H.-H.; Yamaguchi, H.; Fujimori, D.; Nishida, A.; Yamamoto, H.
Spectrosc. Lett. 1995, 28, 177190.
15. Sasaki, C.; Hamada, T.; Okumura, H.; Maeda, S.; Muranaka, J.; Kuwae, A.;
Hanai, K.; Kunimoto, K.-K. Polym. Bull. 2006, 57, 747756.
334
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 16. Maeda, S.; Kunimoto, K.-K.; Sasaki, C.; Kuwae, A.; Hanai, K. J. Mol. Struct.
2003, 655, 149155.
17. Maeda, S.; Mori, T.; Sasaki, C.; Kunimoto, K.-K.; Kuwae, A.; Hanai, K.
Polym. Bull. 2005, 53, 259267.
18. Li, N.; Lu, J.; Xu, Q.; Wang, L. Opt. Mater. 2006, 28, 14121416.
19. Smitha, P.; Asha, S. K. J. Phys. Chem. B 2007, 111, 63646373.
20. Chang, J. Y.; Knecht, R.; Braun, D. G. Methods Enzymol. 1983, 91, 4148.
21. Hiraki, J., Morita, H. Jpn Patent No. 61-192157, 1986.
22. Saito, J.; Smith, I. C. P. Arch. Biochem. Biophys. 1973, 158, 154.
23. Hendra, P., Jones, C., Warnes, G., Eds.; Fourier Transform Raman
Spectroscopy: Instrumentation and Chemical Applications; Ellis Horwood:
Chichester, U.K., 1991.
24. Baranska, H.; Labudzinska, A.; Terpinski, J. Laser Raman Spectroscopy,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch019

Analytical Applications; Ellis Horwood: Chichester, U.K., 1987.

Downloaded by DUKE UNIV on June 21, 2012 | http://pubs.acs.org

25. Miyazawa, T.; Blout, E. R. J. Am. Chem. Soc. 1961, 83, 712719.
26. Arimoto, H. J. Polym. Sci., Part A-2 1964, 22832295.
27. Hendra, P. J.; Maddams, W. F.; Royaud, I. A. M.; Wills, H. A.; Zichy, V.
Spectrochim. Acta 1990, 46A, 747756.
28. Dos, A.; Schimming, V.; Tosoni, S.; Limbach, H.-H. J. Phys. Chem. B 2008,
112, 1560415615.
29. Ernst, R. R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear Magnetic
Resonance in One and Two Dimensions; Oxford University Press: Oxford,
U.K., 1987.
30. Asano, A.; Tanaka, C.; Murata, Y. Polymer 2007, 48, 38093816.
31. Tonelli A. E. NMR Spectroscopy and Polymer Microstructure: The
Conformational Connection; VCH: New York, 1989.
32. Sasaki, S.; Hishiyama, T.; Huh, K. M.; Ooya, T.; Yui, N. Polym. Prep. Jpn.
2001, 50, 2003.
33. Hatfield, G. R.; Glans, J. H.; Hammond, W. B. Macromolecules 1990, 23,
16541658.
34. Maeda, S; Ohshima, T.; Sakurai, K. Proc. XVIIIth Int. Conference on Magn.
Reson. Biological Systems 1998, 166.
35. Sakurai, K.; Ohshima, T.; Maeda, S. Adv. Chitin Sci. 1998, 135140.
36. Yu, G.; Morin, F. G.; Nobes, G. A. R.; Marchessault, R. H. Macromolecules
1999, 32, 518520.
37. Levy G. C.; Lichter R. L. Nitrogen-15 Nuclear Magnetic Resonance
Spectroscopy; John Wiley & Sons, Inc.: New York, 1979; Chapter 3.
38. Berger S.; Braun S.; Kalinowski H.-O. NMR Spectroscopy of the
Non-Metallic Elements; Wiley & Sons, Inc.: Chichester, U.K., 1997;
Chapter 4.
39. Nge, T. T.; Yamaguchi, M.; Hori, N.; Takemura, A.; Ono, H. J. Appl. Polym.
Sci. 2002, 83, 10251035.
40. Maeda, S.; Mori, T.; Muto, K; Sasaki, C; Kunimoto, K.-K. The 42th Annual
Meeting Japanese NMR Soc. 2003, 1P34, 188.

335
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 20

NMR, NIR, and Infrared Spectroscopy of

CarbohydrateProtein Interactions and
Glycoproteins
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

I. C. Baianu* and V. Prisecaru

AFC-NMR & NIR Microspectroscopy Facility, College of ACES, FSHN

and NPRE Departments, University of Illinois at Urbana, 350 Burnsides
Research Laboratory, 1208 W. Pennsylvania Ave., Urbana, IL 61801
*E-mail: [email protected]

A review of selected data sets for carbohydrateproteins

interactions and glycoproteins is presented together with
authors results from nuclear magnetic resonance (NMR) and
infrared (NIR and mid-IR) spectroscopy, as well as chemical
imaging (micro-spectroscopy) experiments. Previous reports
by X-ray, neutron and electron diffraction are also considered
in this context, and are combined with high-resolution
NMR spectroscopy and relaxation data for glycoproteins in
solution and in biomembranes. High-resolution NMR and
transverse relaxation results are then presented for wheat gliadin
interactions with sucrose in aqueous solutions, as well as wheat
gluten--glucomannan interactions in hydrated gels that are
relevant to food physical chemistry and novel food formulations
for products with increased stability and shelf-life. NMRI and
FT-NIR microspectroscopy data are reported also for intact,
hydrated seeds that contain proteins and carbohydrates as their
major components.

Introduction
Proteincarbohydrate interactions are important for glycoprotein structure,
dynamics and function. They play key roles in physiology, medicine, biochemistry
and biophysical chemistry. Moreover, protein glycosylation is ubiquitous in living
cells as a post-translational modification. Many proteins that play key structural

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and functional biological roles are glycosylated; such cell-surface or extracellular
glycoproteins are involved in immunity and cell-cell recognition (13). Protein
glycosylation is also of great significance in pharmacology and medicine because
more than one-third of the approved protein therapeutics are glycoproteins (4).
Proteincarbohydrate interactions play important roles in protein stability and
function, and are also involved in complex molecular and cellular processes
such as cell adhesion and aggregation, oncogenesis and lectin biology. Thus,
protein glycosylation is a structural variation of both biological/physiological and
pharmaceutical importance because of its major effects on protein stability, cell
adhesion and aggregation.
On the other hand, carbohydrateprotein interactions are also important
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

in industrial applications and in several, related applied sciences such as food

chemistry/food science, pharmacology and medicinal chemistry. In food
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

chemistry they are important for several reasons, such as their role in determining
food properties containing both proteins and carbohydrates, and the correct
formulation and engineering of foods. In pharmacology and medicinal chemistry
such interactions are important for the design of novel drugs with increased
effectiveness and very high selectivity.
Protein residues, or side chains, can form a variety of non-covalent bonds
with carbohydrates, and this may occur with different sugar interface propensity
for different amino acid residues. Thus, hydrogen bonds may be formed between
the hydroxyl groups of sugars and the charged R-groups of amino acid residues
in a protein such as those of Asp, Glu, Lys, Arg and His. Another case is that of
a hexose in a conformation in which several of its carbon atoms are in a cluster
that allows for energetically favorable CH- interactions to occur with the parallel
aligned aromatic rings of Phe, Trp or Tyr (5, 6), when the sugar and aromatic
rings are within 0.4nm from each other. For example, CH-pi interactions can form
between sugars and the aromatic amino acids Tyr296, Phe241, Phe243 on each
of the two chains of the Fc antibody, whereas Phe243 forms an actual contact
with GlcNAc as found from crystal structures of human IgGFc (7). The N-
glycolsylation of the conserved residue Asn297 in an IgG antibody is an example
of both types of such interactions.
The first step in carrying out detailed physicochemical studies of glycosylated
proteins, and glycoconjugates in general, is their separation with a high degree
of purity for elucidation of structure and function (8). Highly effective methods
for the separation of naturally derived glycoconjugates are various forms of high-
performance liquid chromatography (HPLC) that can also be effectively combined
with their analysis by isoelectric focusing . On the other hand, structurally well-
defined glycoconjugates can also be synthesized and then separated by HPLC.
Conformational studies of highly-purified, both natural and synthetic
glycoconjugates can be then carried out on such highly purified fractions by
high-resolution NMR, including various 2D-FT NMR techniques and/or triple
resonance, 1H 13C 15N, or by FT-IR, including polarized spectroscopy.
Such interesting studies were reported, for example, in the case of the human
glycophorin A, an important structural protein in the erythrocyte membrane (9,
10).

338
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Whereas NMR spectroscopy allows the elucidation of the 3D molecular
structure in solutions, X-ray diffraction (XRD) - also combined sometimes with
neutron diffraction and scattering techniques) - allows the determination of
structure only in crystalline solids such as hydrated protein crystals. On the other
hand, flexibility is essential to the physiological function(s) of many proteins (11),
including glycoproteins. However, XRD cannot distinguish between structural
and dynamic disorder (p.439 of Ch. 12 in ref. (12)), whereas NMR spectra
and relaxation (11, 12) are quite sensitive to both main-chain and side-chain
dynamics (13, 14). NMR spectroscopy has been employed to study the flexibility
of both protein segments and domains; combined X-ray and resolution-enhanced
NMR spectroscopy studies were also reported as in the case of the lac-repressor
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

low-resolution structure (see Figure XII-7 on p. 459 in (12)). Quantitative studies

of internal motions in proteins are also possible by NMR relaxation (p. 465 in
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

(12) ); especially useful are 13C T1 measurements that were reported for resolved
aromatic amino acid residues of proteins in aqueous electrolyte solutions (p.475,
loc. cit. and (14) ).
Both NMR and neutron techniques are also available for the study of isotope
exchanges, such as the 1H 2H isotope exchange of the amide protons in
any protein peptide groups (p. 461 in ref. (12)), or the proton exchange of
Tyrosine-OH, Threonine-OH, SerineOH and Cystine SH groups, or for liquid
water protons in carbohydrate solutions or gels (15, 16); the chemical exchange
of protons between water molecules and proteins also affects the 17O NMR
relaxation of water (ref. (17) and references cited therein).
In the special case of a highly ordered array of bacteriorhodopsin molecules
in the purple membrane of Halobacterium halobium a high-resolution structural
determination was reported from electron diffraction (ED) studies (18, 19).
However, this is not generally possible in either bacterial or mammalian cell
membranes, unless highly-ordered membrane arrays can be formed, and also if
the proteins themselves preserve a definite orientation relative to the membrane
surfaces; this was indeed observed for oriented arrays of both human and porcine
erythrocyte membranes that had a lamellar periodicity of 12.1 nm, as determined
from the XRD low-angle patterns in the fully hydrated state (20). In such oriented
membrane systems, the glycophorin A dimers exhibited discrete, sharp X-ray
reflections (Figure 1) corresponding to an orientation of glycophorin A chains
normal to the erythrocyte membrane surface, as shown in Figure 78c on page 191
for porcine red cell membrane arrays, and in Figure 81b on page 197 for human
erythrocyte membranes in ref. (20).
The transmembrane helix of glycophorin A contains a seven-hydrophobic
residue motif, LIxxGVxxGVxxT, that was proposed to mediate protein
dimerization (9). Then, magic angle spinning (MAS) NMR allowed the
determination of the helix-to-helix contacts in the glycophorin A transmembrane
domain by showing that the methyl groups of Val-80 and Val-84 are close-packed,
respectively, with the Gly-79 and Gly-83 (10). On the other hand, threonines
and serines are the most common polar residues found in the transmembrane
domains of proteins such as glycophorin A because they are able to readily
to form hydrogen bonds back to the carbonyls on the same helix. Polarized
Fourier transform infrared spectroscopy was employed to study the orientation

339
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of glycophorin A in reconstituted membranes and the analysis of the dichroic
ration of the 1655 cm-1 amide I vibration band yielded a value of 35o for the
helix crossing angle (9) for the transmembrane orientation. Rotational-echo
double-resonance (NMR) measurements also showed the close packing of the
Thr-87 -methyl group with the backbone nitrogen of Ile-88 across the dimer
interface, at ~0.4nm distance; such a short inter-helical distance makes it possible
for the -hydroxyl of Thr-87 to hydrogen-bond to the backbone carbonyl of
Val-84 on the opposing glycophorin A transmembrane helix (9, 21).
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

Figure 1. XRD pattern of oriented, human erythrocyte membrane arrays showing

meridional 0.9 nm-1 X-ray reflection of glycophorin A chains oriented normally
to the membrane surface (20), corresponding to a repeat distance of ~1.2 nm
along the transmembrane helix axis of human glycophorin.

In this concise review, several NMR and NIR/mid-IR results are also
presented that are pertinent to practical applications in the formulation of
novel food formulations with proteins and complex polysaccharides as major
components.
Our results are also relevant to several other industrial applications such
as wheat grain conditioning, soybean processing and corn milling, as well as
biomedical and biotechnology applications, such as the early detection/ prevention

340
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of disease, and the development of novel, individualized therapies with highly
effective and selective medicines or vaccines.

Experimental Section
Materials and Measurements

Analytical grade sucrose was a product of ER Industries, Inc, (Addison,

IL 60101). Analytical grade fructose was a gift from General Mills Inc.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

(Minneapolis, MN). Amylopectin purified from corn starch was obtained from
Sigma Chemical Co. (St. Louis, MO). Amaizo Polar Gel 5 and Fro-Dex-55L
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

were gifts from the American Maize Products Co. (Hammomd, IN). Moisture
contents of fructose, glucose and sucrose samples were determined in triplicate
by a vacuum oven method to be less than 0.2%.
1H NMRI microscopy was carried out at 17MHz with a laboratory-built

instrument (Taylor and Baianu, 1978:(23)) .

1H NMR relaxation measurements were carried out in triplicate at 10MHz on

a PC-10, and also on a PC-20 Multispec NMR relaxometer (Bruker- IBM Instr.,
Danbury, CT). The inversion recovery sequence, 180o- - 90o, was employed for
T1 measurements of aqueous solutions, gels and hydrated solids. The Carr-Purcell
multi-pulse sequence was employed with the Meiboom-Gill modification (CPMG)
for transverse relaxation times, T2, measurements in solutions and hydrated gels.
For low moisture solid proteins, small carbohydrates, polysaccharides and gels
the multi-pulse Ostroff-Waugh (O-W) NMR pulse sequence was employed for
recording spin-echo spectra with up to 10,000 pulses in the O-W sequence, as
further detailed in (22, 29).
1H NMR high-resolution spectra for protein and carbohydrate solutions in

D2O were recorded in a 1mm capillary on a Varian UNITY-500 spectrometer

operating at 500MHz as detailed in ref. (23).
13C NMR spectra were recorded at 125MHz on a GE 500 spectrometer with

multi-pulse proton decoupling at 500MHz using WALTZ and WAHUHA pulse

sequences in the decoupling channel.
17O and 2H NMR measurements were carried out with a Bruker multi-nuclear

CXP-200 spectrometer operating with a 98mm bore magnet at 4.7 Tesla (23).
Other NMR techniques employed were as specified in (24).
FT-NIR spectra of glycoproteins, carbohydrates and intact, fully hydrated
whole soybean seeds were recorded with a Perkin-Elmer spectrometer model
SpectrumOne NTS (soybean seeds moisture contents were varied between 6 and
24% by dry weight). FT-NIR chemical/hyperspectral images were recorded with
the same spectrometer attached to the NIR Microspectroscopy AutoImage system
which was optimized for 1 micron spatial resolution and nanogram sensitivity.
High-resolution FT-NIR images of selected sample areas of 4 mm2 were recorded
under PC control in less than 1 hour.

341
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Results and Discussion
NMR Spectroscopy and Relaxation Studies of CarbohydrateProtein
Interactions

Wheat Gliadin Interactions with Sucrose in Acidic Aqueous Solutions

Wheat storage proteins are of prime importance to the US food industry

because of their use in large amounts for making human foods. Such proteins,
not unlike membrane proteins have a high proportion of hydrophobic amino acid
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

residues and are not soluble in water at neutral phi A relatively small fraction
of wheat proteinsthe gliadins, is soluble at pH values lower than about 3.5.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

The 13C NMR spectra of wheat gliadins in solution are compared in Figure 2
with those of corn zeins in a similar solvent mixture, and both were recorded at
room temperature. The observed 13C NMR peaks of wheat gliadins are overall
considerably sharper than those of corn zeins even though the average molecular
weights of such two selected samples of cereal proteins are very close - 35,000
for wheat gliadins vs. 32,000 for the corn zeins. The higher mobility of the
aromatic amino acid residues of wheat gliadins in comparison with those of corn
zeins, which is reflected in the sharper peaks in the spectral range from 100 to 150
ppm, for example, is suggestive of either more compact, folded structures of corn
zeins than those of wheat gliadins or a significant degree of corn zein aggregation
in solution even at low zein concentrations. A similar observation regarding
the sharpness of the 13C NMR peaks holds for the aliphatic and carbonyl and
carboxyl regions in Figure 2. Specific assignments of the resolved 13C peaks were
previously reported (36, 37) for 21 amino acid residues although considerable
overlap did occur between the resonance peaks arising from several amino acid
residue groups.
Their hydration behavior and weak interactions with small carbohydrate
molecules was investigated by 1H NMR transverse relaxation measurements at
low magnetic fields in the presence of a Mn+2 paramagnetic ion probe. Both
atlow (<1.4 T) and high magnetic fields (> 4.69 T) the water proton transverse
magnetization decay of wheat gliadin solutions with/without carbohydrates
was a single exponential up to 5mM gliadin concentrations at pH 3.4. At low
concentrations of the paramagnetic ion probe, e.g. < 0.6 mM as in ref. (22), one
can therefore assume a fast-exchange model with bulk water for the bound water
protons surrounding the Mn+2 ion probe which is attached to the wheat gliadin
negatively charged surface residues, Glu and Asp, in solution at pH 3.2. With
this model, one can derive a slow correlation time s for isotropic tumbling of
water bound to such divalent cations of ~18 ns, consistent with the calculated
value from the tumbling rate of globular wheat gliadins of an average molecular
weight of 37,000 using the Stokes-Einstein equation, as explained on p. 475 in
(12). Moreover, the sucrose concentration dependence of the 10 MHz relaxation
rate of water protons in 3 mM gliadin solutions in the presence of a constant
concentration of 0.5 mm Mn+2 exhibited a linear decrease with sucrose up to 0.4
M sucrose, as shown in Figure 8 on p. 372 in (22). A nonlinear regression analysis
342
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of such data was carried out with a thermodynamic linkage model that yields
the coefficients of non-ideal protein-protein interactions that indicate attraction
rather than repelling between the wheat gliadin and sucrose molecules in solution,
whereas the second order virial coefficient of gliadin-gliadin interactions in these
solutions are characteristic of repulsion between the protein molecules. Such
observations can be also considered as evidence for a preferential hydration
of wheat gliadins in the presence of sucrose (25). In other words, there is a
carbohydrate competition for water with the residues present in the protein
hydration domain, as proposed by Arakawa and Timasheff in ref. (25). Moreover,
using a spin probe, Pearce et al. found that spin label binding to wheat gliadins in
solution decreased significantly upon heating (26). After heating to 50 o C, and
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

then cooling to 20 o C, the 13C NMR peaks of wheat gliadins in solution were
significantly sharper than those before heating suggesting a partial, irreversible
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

denaturation through loss of hydrogen bonds (37).

Figure 2. Multi-pulse, proton decoupled 13C NMR spectra of cereal proteins

recorded at 125 MHz with a GE500 spectrometer: A. Wheat gliadins, and B.
Corn zeins in aqueous solutions at pD 3.4.

343
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Protein Interactions with Polysaccharides in Hydrated Gels

Proton broadband decoupled 13C NMR spectra were previously reported both
for wheat gluten ((23) and references cited therein) and various polysaccharides
from wheat (23), or other plant sources ((2730), and references cited therein).
13C chemical shift assignments were made for wheat gluten CP-MAS spectra

by comparison with those of wheat gliadins in acidic aqueous solutions at pD

3.4 (23). In the interesting, but very complex, case of branched glucomannan
polysaccharides (GP) purified from Amorphophalus konjac K. Kock tuber
powders, partial 13C NMR chemical shift assignments were made by comparison
with glucomannan oligosaccharides derived from GP by partial hydrolysis (27,
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

28). Such GP carbohydrates that are relatively hydrophobic form very readily
gels at GP concentrations in water as low as 0.1%, which become quite firm
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

or rigid at GP concentrations higher than ~ 2%. In spite of the complexity of

both GP and wheat gluten, their mixture in a system of two partially hydrated,
branched hydrophobic materials-- which independently form stable gels-- may
have considerable potential for increasing the shelf-life of wheat bread by the
addition of as little as 0.1% GP to the hydrated gluten. The FT-NIR spectra
of very low moisture GPs were quite similar to those of carbohydrates in the
soybean seed (Figure 3) with the exception of several relatively sharp bands below
5,500 cm-1 that are present only in the GP spectra; such sharp NIR bands may
correspond to a crystalline fraction of GPs that has been also detected previously
by XRD studies of GP films (as described in previous studies cited in (27)). 1H
NMR transverse relaxation data with varying GP concentration in wheat gluten
gels were previously interpreted in terms of the formation of cross-linked GP gels
with the cross-links or junctions being formed with divalent cations such as Ca+2.
Such studies could also be carried with the Mn+2 ion probe in a manner similar to
the experiments described in the previous section with wheat gliadins interacting
with sucrose in aqueous solutions. As in the case of wheat gliadin solutions
it was found from low-field 1H NMR transverse relaxation measurements that
water protons exchange rapidly between the water bound to GP in gels and bulk
water trapped in the GP gel. One can therefore predict that in the hydrated wheat
gluten--GP gel system doped with low concentrations of the paramagnetic Mn+2
ion probe one would observe a marked increase in the transverse relaxation
rates of water protons as the concentration of GP is increased and becomes
cross-linked, thereby resulting in tough, rubber-like, textured gels. Further 13C
NMR observations carried out in parallel with such 1H NMR transverse relaxation
studies may also allow one to identify the cross-linking sites in GPwheat gluten
gels.
On the other hand, in soybean protein gels cross-linking occur more
frequently than in the case of wheat gluten--GP gels and additional ionic
interactions are also present. In spite of the high molecular weights of the soy
proteins one notes in the WALTZ 1H-decoupled 13C NMR spectra of such gels
highly resolved peaks throughout the entire spectrum (Figure 3). Thus, relatively
rapid molecular motions of the amino acid residues are occurring in the highly
hydrated soybean protein gels whose high resolution 13C NMR spectra are shown
in Figure 3.

344
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 In Vivo Microscopy and Microspectroscopy of Proteins and Carbohydrates
in Complex Hydrated Systems

NMR Imaging and Relaxation of Hydrated Wheat Grains

One of the earliest in vivo microscopic images reported for wheat proteins,
polysaccharides, starch, water and oils was obtained with 50 micron resolution
in hydrated wheat grains by 1H NMRI at 17 MHz in 1978 (22, 23); at the time,
the spatial resolution was mostly limited by the available computer FFT/imaging
speed and digitization capabilities. However, T1 --contrast enhancement allowed
the specific visualization of germ oil in the intact, fully hydrated wheat grain (22,
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

23). For higher 1H resonance frequencies in the higher magnetic fields currently
available, it has been hypothesized that the spatial image resolution obtainable by
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

2D-FT NMRI might be on the order of 1 to 5 microns, and does obviously depend
on the size of the imaged object. However, technological NMR advances may
very well push even further this spatial resolution limit of NMRI microscopy, as
it has already happened with FT-NIR hyperspectral imaging (29); an example of
such recent developments is presented next.

Figure 3. WALTZ-16 1H Decoupled 13C Liquid-State NMR of a soy flour gel

sample of 38.7% protein content. Spectrum recorded with 10,000 transients on a
Varian UI600 NMR spectrometer, in a 14.1T external magnetic field. (Source:
ref. (38).)

345
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 FT-NIR Microspectroscopy and Imaging of Soybean Seeds and Embryos

Even though the spectral resolution in FT-NIR spectra of proteins and

carbohydrates is quite limited as a result of overlapping NIR bands from these
components it is possible to quantitate these components in hydrated complex
systems, such as wheat grains or whole soybean seeds and embryos. Moreover,
unlike the case of FT-IR, the presence of thick samples and high levels of either
water or oils does not prevent the investigation of proteins and carbohydrates
in intact hydrated seeds. An example is presented here in Figure 4 for intact
whole soybeans at hydration levels as high as 16% (by dry weight), even though
such studies can be carried out even at higher hydration levels with appropriate
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

multiple scattering and reflectivity corrections. Furthermore, the NIR sensitivity

is sufficiently high to carry out both micro-spectroscopic and microimaging
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

studies with state-of-the-art instrumentation.

FT-IR Chemical Imaging can be carried out with very high sensitivity but
is limited in spatial resolution primarily because of the longer wavelength and
focal-point-array (FPA) detector design limitations. An example was presented in
Figure 2.1.14 of ref (29) for the in vivo detection of protein bodies (Figure 5) in
fully hydrated somatic soybean embryos grown in culture from single cells (29).
On the other hand, FT-NIR hyperspectral imaging has lower sensitivity than the
FT-IR microspectroscopy but is capable of ten times higher spatial resolution in
hyperspectral images and does not have the FPA detector design limitations of
the FT-IR microspectrometers. The latest developments discussed in ref. (29)
indicate that the sensitivity range of FT-NIR microspectroscopy observations
can be extended to the femtogram level, with submicron spatial resolution. Such
FT-NIR/IR microspectroscopy instrumentation developments are potentially very
important for agricultural and food biotechnology-as well as biomedical and
pharmacological applications-- that require rapid and sensitive analyses, such as
the screening of high-content microarrays in Genomics and Proteomics research.
Novel, two-photon NIR excitation fluorescence correlation
microspectroscopy results were obtained with submicron resolution for
concentrated suspensions of plant cells and membranes. With advanced
super-resolution microscopy designs, a further, tenfold resolution increase is
attainable, at least in principle, along the optical (z) axis of the microspectrometer.
Especially promising are current developments that employ multi-photon NIR
excitation which could lead, for example, to novel cancer prevention methodology
and the early detection of cancers using NIR-excited fluorescence. Other
related developments are the applications of Fluorescence Cross-Correlation
Spectroscopy (FCCS) detection to monitoring DNA hybridization kinetics, DNA
binding and ligand-receptor interactions--such as glycoprotein recognition of
individual receptors on cell surfaces (31, 33), as well as HIV, or HBV (viral)
single--particle detection (29, 30) and potential treatments for Alzheimers
disease (32, 33).

346
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

Figure 4. FT-NIR spectroscopy of the soybean major components, including

protein, soluble and insoluble carbohydrates (fiber), water, and oils.

Molecular Interactions and Solute Activity Determination from NMR Relaxation

Measurements on Hydrated Proteins, Carbohydrates, and ProteinCarbohydrate
Systems

In a ternary system such as proteins, carbohydrates and water, one has

to consider proteinwater, proteinprotein, proteincarbohydrate, as well
as carbohydratewater and carbohydratecarbohydrate interactions. For
solutions the problem is readily solved, at least in principle, by an equilibrium
thermodynamic approach in terms of solute and solvent activitities by employing

347
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the well-known Gibbs-Duhem equation. On the one hand, there are several
experimental approaches to determining water activity for both solutions and
crystalline solids. On the other hand, the determination of protein-protein
interactions or protein activity in solutions, as well as proteincarbohydrate
interactions has been traditionally difficult to measure except for solutions
at very low concentrations. The approach presented above in Sections II.a
and II.b by employing NMR relaxation measurements for water nuclei in
concentrated protein solutions has allowed the quantitative determination of the
virial coefficients of protein activity in such aqueous solutions, as well as in
hydrated gels. This was possible for small proteins---such as lysozyme (17),
corn zeins (22) and caseinsand for very large proteins, such as soy glycinins
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

(14) and intact myosin (22). The most reliable protein activity determinations
were made from water 17O NMR measurements in protein solutions with or
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

without added salt; the addition of salt ions to protein solutions has marked
effects on 17O, 2H and 1H NMR transverse relaxation rates of water that vary in
a strongly nonlinear manner with the concentration of salt added to solutions of
the same protein concentration, for salts such as NaCl, NaBr, KCl, KBr, LiCl or
ammonium sulfate. As expected from the thermodynamic theory of solutions, the
protein virial coefficients determined from the protein concentration dependence
of water 17O NMR are also dependent upon the type of both the added cation and
anion, following the lyotropic series (22). A thermodynamic linkage approach
also becomes necessary for the correct treatment of protein solutions in water
with added salt ions (34). This approach has also been tested for amino acids in
aqueous solutions with or without added salt, and in the case of carbohydrates the
prerequisite NMR relaxation studies were carried out for several polysaccharides
of practical importance (15, 16, 22). Although, proteincarbohydrate interactions
can be investigated by NMR relaxation measurements even in complex systems,
such as those discussed above in Section II.b, and in the next section, the question
of the direct determination of protein and carbohydrate activities in such complex
ternary systems remains a difficult one unless coupled with a thermodynamic
linkage approach (34), as well as a careful consideration of irreversible systems
that are far from equilibrium (35). However, it would be possible to estimate
the sum of protein and carbohydrate activities in such ternary systems by
subtracting the measured water activity, in accordance with the Gibbs-Duhem
equation. For solid gels and powders of non-crystalline, or glassy, protein and
proteincarbohydrate systems the equilibrium thermodynamics approach is no
longer valid and alternate, nonequilibrium thermodynamic models are currently
being investigated, as indicated above.

Conclusions
The study of proteincarbohydrate interactions and glycoproteins has
significant potential for both biological and practical applications, such as the
development of foods with improved shelf-life, better pharmaceuticals and
nutraceuticals.

348
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

Figure 5. Transmission FT-IR Chemical Image of a fully hydrated, live soybean

somatic embryo (i.e., cloned): shown in red is the high intensity (~0.65) of the
protein amide II absorption band. The image was recorded in 20 minutes with
a spectral IR resolution of 0.2 cm-1, and a spatial image resolution of ~10 m.
(Image adapted from ref. (29).)

A wide range of NMR techniques, when combined with FT-NIR/IR,

CCS, XRD and neutron scattering offer very substantial advantages for
physico-chemical studies of important proteincarbohydrate interactions both in
vitro and in vivo.
Some of the expected novel therapeutics may include safer biomaterials and
also highly-selective and effective, non-toxic drugs (31) or drug-delivery systems.

Acknowledgments
The first author would like to thank the various researchers who contributed
through fruitful discussions to this work, including: Dr. David G. Taylor of the
University of Surrey, Dr. Eiichi M. Ozu of the University of Osaka, Japan, Dr. M.
Balyuzzi and Prof. Ronald E. Burge, F.Inst.P. of the Cavendish Laboratory at the
University of Cambridge, U.K.
The partial financial support for this research by Renessen Co. is gratefully
acknowledged.
349
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 References
1. Arnold, J. N.; Wormald, M. R.; Sim, R. B.; Rudd, P. M.; Dwek, R. A. The
impact of glycosylation on the biological function and structure of human
immunoglobulins. Annu, Rev. Immunol. 2007, 25, 2150.
2. Kaneko, Y.; Nimmerjahn, F.; Ravetch, J. V. Anti-inflammatory activity
of immunoglobulin G resulting from Fc sialylation. Science 2006, 313,
670673.
3. Bucior, I.; et al. Carbohydrate-carbohydrate interaction provides adhesion
force and specificity for cellular recognition. J. Cell Biol. 2004, 165,
529537.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

4. Lee, J. C. Biochemistry of carbohydrate-protein interactionReview.

FASEB J. 1992, 6, 31933200.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

5. Quiocho, F. A. Protein-Carbohydrate InteractionsBasic Molecular

Features. Pure Appl. Chem. 1989, 61, 12931306.
6. Spiwok, V.; et al. Modelling of carbohydrate-aromatic interactions: ab initio
energetics and force field performance. J. Comput.-Aided Mol. Des. 2005,
19, 887901.
7. Krapp, S.; et al. Structural analysis of human IgG-Fc glycoforms reveals
a correlation between glycosylation and structural integrity. J. Mol. Biol.
2003, 325, 979989.
8. Karnoup, A.; et al. A novel HPLCUVMS method for quantitative analysis
of protein glycosylation. J. Chromatogr., B 2007, 859, 178191.
9. Smith, S. O.; et al. Implications of Threonine Hydrogen Bonding in
the Glycophorin A Transmembrane Helix Dimer. Biophys. J. 2002, 82,
24762486.
10. Smith, S. O.; Bormann, B. J. Determination of helix helix interactions in
membranes by rotational resonance NMR. Proc. Natl. Acad. Sci. U.S.A.
1995, 92, 488491.
11. Baianu, I. C. Structural Order and Partial Disorder in Biological Systems.
Bull. Math. Biol. 1980, 42, 110112.
12. Jardetsky, O.; and Roberts, G. C. K. NMR in Molecular Biology; Academic
Press: New York, 1981.
13. Deslaurier, R.; Smith, I. C. P.; Walter, R. J. Am. Chem. Soc. 1974, 96, 2289.
14. (a) Baianu, I. C.; Kakalis, L. Carbon-13 NMR of Soy Glycinins and
Conglycinins. J. Agric. Food Chem. 1989, 37, 1222-1228. (b) Baianu, I. C.;
Kakalis, L. Carbon-13 NMR of Soy Glycinins and Conglycinins. J. Agric.
Food Chem. 1990, 37, 14791486.
15. Mora-Gutierrez, A.; Baianu, I. C. J. Agric. Food Chem. 1989, 37,
14591465.
16. Mora-Gutierrez, A.; Baianu, I. C. Hydration Studies of Maltodextrins by
Proton, Deuterium and Oxygen-17 Nuclear Magnetic Resonance. J. Food
Sci. 1990, 55, 462467.
17. Kakalis, L. T.; Baianu, I. C. Oxygen-17 and Deuterium Nuclear Magnetic
Relaxation Studies of Lysozyme Hydration in Solution: Field dispersion,
concentration, pH/pD and protein activity dependences. Arch. Biochem.
Biophy. 1988, 267, 829841.

350
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 18. Henderson, R.; Unwin, P. N. T. Nature (London) 1975, 257, 2832.
19. Hayward, S. B.; et al. Molecular orientation of bacteriorhodopsin within the
purple membrane of Halobacterium halobium. Proc. Natl. Acad. Sci. U.S.A.
1978, 75 (9), 43204324.
20. Baianu, I. C. Structural Studies of Erythrocyte and Bacterial Cytoplasmic
Membranes by X-ray Diffraction and Electron Microscopy. Ph.D. Thesis,
University of London, U.K., 1974.
21. McKenzie, K. R.; Prestegard, J. H.; Engelman, D. M. A Trans- membrane
helix dimer structure and implications. Science 1997, 276, 131133.
22. Baianu, I. C. Multinuclear Spin Relaxation and High-resolution NMR
Studies of Food Proteins, Agriculturally Important Materials and Related
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Systems. In NMR Applications in Biopoloymers; Finley, J. W. et al., Eds.;

Plenum Press: New York, 1980; pp 361389.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

23. Baianu, I. C.; Taylor, D. G. NMRI microscopy and T1 experiments on the

hydration of single variety wheat grains at 50 m resolution. Phys. Medicine
Dept. Report; University of Surrey, U.K., 1978; p 4.
24. Ferentz, A. E.; Wagner, G. NMR spectroscopy: a multifaceted approach to
macromolecular structure. Q. Rev. Biophys. 2000, 33 (1), 2965.
25. Arakawa, T.; Timasheff, S. N. Biochemistry 1982, 21, 65436546.
26. Pearce, L. F.; et al. J. Food Microstruct. 1987, 6, 121126.
27. Ozu, M. E.; Baianu, I. C.; Wei, L. S. Physical Chemistry of Food Processes;
Baianu, I. C., Eds.; Intl. Thompson Pubs.: New York, 1992; Vol. 2, Chapter
14.
28. Baianu, I. C.; Ozu, M. E. Gelling Mechanisms of Glucomannans and
Polysaccharide Interactions with Wheat Gluten Proteins. NMR Spectroscopy
of Polymers in Solution and in the Solid State; Cheng, H. N., English, A. D.,
Eds.; ACS Symposium Ser. 834; American Chemical Society: Washington,
DC, 2003; Chapter 23, pp 298305.
29. Baianu, I. C. Near Infrared Microspectroscopy, Fluorescence
Microspectroscopy, Infrared Chemical Imaging and High-resolution NMR
Analysis of Soybean Seeds, Somatic Embryos and Single Cancer Cells. In
Analysis of Oil Seeds; Luthria, D., Ed.; AOCS Pubs.: Champaign, IL, 2004;
pp 241273.
30. Baianu, I. C.; You, T. High-resolution Near-Infrared and Nuclear Magnetic
Resonance Analysis of Food and Grain Composition. In Handbook of Food
Analysis Instruments; Otish, S., Ed.; CRC Press: Boca Raton, FL, 2009;
Chapter 12, pp 247280, ISBN: 978-1-4200- 4566-6.
31. Baianu, I . C. Complex Systems Analysis of Cell Cycling Models in
Carcinogenesis: Cell Genome and Interactome, Neoplastic Non-random
Transformation Models. Quantitative Biology; Urbana, Illinois, 2008;
Update of abstract q-bio.EXT-2004-065.
32. Charles, M.; Baianu, I. C.; Prisecaru, V. Visual Molecular Dynamics (VMD)
and the Possible Impact of Lectins on Alzheimers Disease Treatment.
Internal UIUC Report; 2007.
33. Baianu, I. C. Applications of Novel Techniques to Health Foods, Medical
and Agricultural Biotechnology. Quant. Biol. arXiv, 2005. Available from:
http://arxiv.org/abs/q-bio/0406047.

351
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 34. Kakalis, L. T.; Kumosinki, T. F.; Baianu, I. C. Nonideal Effects of Protein
Aggregation and Protein Activity on the Sorption of Water Vapor by
Proteins: A Thermodynamic Linkage Study. J. Agric. Food Chem. 1992,
40, 20632071.
35. Prigogine, I.; Nicolis, G. Self-Organization in Non-Equilibrium Systems:
From Dissipative Structures to Order through Fluctuations; J. Wiley &
Sons: New York, 1977.
36. Baianu, I. C. Carbon-13 and Proton NMR studies of Wheat Proteins in
Solution. J. Sci. Food Agric. 1981, 32, 309313.
37. Baianu, I. C.; Johnson, L. F.; Waddell, D. K. High-field, high- resolution
Proton, Carbon-13 and Nitrogen-15 NMR of Wheat Proteins in Solution. J.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Sci. Food Agric. 1982, 33, 373383.

38. Baianu, I. C. et al. High-Resolution Nuclear Magnetic Resonance and Near
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch020

Infrared Determination of Soybean Oil, Protein and Amino Acid Residues

in Soybean Seeds. 94th AOCS Symp. Proceed.; Kansas, MO, 2003; AOCS:
Champaign, 2004; Chapter 11.

352
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 21

Advanced Solution 2D-NMR of Fluoropolymers

Peter L. Rinaldi,*,a Silapong Baiagern,a Peter Fox,b Jon L. Howell,b
Linlin Li,a Xiaohong Li,a Donald F. Lyons,b Elizabeth F. McCord,b
Sangrama K. Sahoo,a Eric B. Twum,a and Faith J. Wyzgoskic
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

aDepartment
of Chemistry, The University of Akron,
Akron, Ohio 44325-3601
bE. I. Dupont de Nemours and Co, Experimental Station Laboratory,

Wilmington, DE 19880-0402
cDepartment of Chemistry, The Ohio State University Mansfield,

Mansfield, Ohio 44906

*E-mail: [email protected]

In this paper, new methodology is described for characterizing

the structures of fluoropolymers. The methods are based on
19F/1HH/13C multiple resonance and 2D NMR experiments,

taking advantage of the unique ranges of 19F-19F and 19F-13C

couplings. 19F-13C heteronuclear single quantum correlation
(HSQC) 2D-NMR experiments based on 1JCF and 2JCF
experiments provide short range correlations within monomer
units and in many cases help to distinguish between structure
fragments identified using 3JFF, 4JFF and 5JFF correlations.
COSY and especially selective COSY experiments provide
high resolution 2D-NMR experiments with correlations based
on 3JFF, 4JFF and 5JFF couplings. These permit identification
of long-range structure correlations between monomer
units. Examples using poly(hexafluoropropylene oxide) and
poly(vinylidene fluoride) are shown.

Introduction
In recent years, there has been renewed interest in the synthesis and study
of fluoropolymers. This interest is derived in part from the unique stability
of the C-F bond to chemical, thermal and radiation degradation compared to
hydrocarbon-based structures. The presence of a high level of fluorination in

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 polymers also imparts desireable physical and mechanical properties such as
excellent lubrication characteristics, high boiling properties, and desirable optical
properties such as transparency in regions of the spectrum where hydrocarbons
usually absorb. Because of the strength of the C-F bond (relative to the C-H bond)
and the fact that fluorine on the surface of the molecule can shield the carbon
skeleton from chemical attack, fluoropolymers usually have long lifetimes. They
can be used in applications where replacement or replenishment is not possible
and/or desirable.
The properties of 19F make NMR an interesting and useful method for
studying fluoropolymer structure and composition. 19F has many characterististic
similar to that of 1H including 100% natural abundance, nuclear spin I=1/2, and a
large magnetic moment (such that its signal strength is similar to that of 1H). In
addition, 19F has a large chemical shift range (over 200 ppm) compared to 1H (10
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

ppm); its chemical shift is extremely sensitive to variations in chemical structure;

Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

and it exhibits a large range of J-couplings (couplings over 2 to 6 bonds are

commonly observed which range from 0-300 Hz). Finally, in many polymers, the
presence of 1H, 13C and 19F provides many structure-dependent NMR parameters
to measure.
NMR, especially 19F NMR, has always been a useful tool for characterizing
fluoropolymers, especially PHFPO and PVDF. In a series of papers, Howell
et al. (13) used NMR to characterize PHFPO polymers and short chain
oligomers. Kostjuk et al. (4) used 19F 1D-NMR to study the mechanism of
HFPO polymerization using alkali metal fluorides as initiators. Durand et al. (5)
reported NMR data on PHFPO telomers.
Murasheva et al. (6) used 19F NMR to study PVDF, PVDF/TFE and
PVDF/TFE/HFP copolymers. In their seminal work, Cais and Sloane (7) used
19F 1D-NMR spectra to study inverse addition of monomer units in PVDF, and

made use of additivity relationships to assign resonances from the various triad
sequences formed by normal and inverse addition. Ovenall and Uschold (8) used
19F NMR to detect signals attributed to chain-ends and branches. Pianca et al.

(9) used 1D- and COSY 2D-NMR methods to study the chain-end structures in a
series of fluoropolymers, including PVDF. The branched structures in PVDF were
characterized by Hedhli et al. (10), using 1D- and 2D- NMR methods. Gulot et
al. (11) used 1H and 19F NMR to investigate the microstructure of PVDF; and
Ameduri et al. (12) used NMR and MALDI-TOF mass spectrometry to study
the structures of a range of VDF telomers. Wormald et al. (13) used 1D-NMR
methods, aided by solid state high power decoupling techniques, to get high
resolution spectra of a VDF telomere; and Durand et al. (14) reported NMR data
on PVDF telomers.
Most of the previously published NMR work involved simple 1D-NMR
experiments, however, a few notable papers on 2D-NMR characterization of the
fluoropolymers which are the subject of this work have been published. Katoh
et al. (15) first showed how 1H/19F/13C triple resonance and 2D-NMR methods
could be used to simplify the spectra of fluoropolymers using data from PVDF as
an example. Takasaki et al. (16) used low resolution 19F-19F COSY, and 19F-13C
HSQC- and HMBC-based 2D-NMR experiments to study PHFPO segments in
perfluorinated ionomers. Macheteau et al. (17) have used selective versions of

356
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 1H-19F heteronuclear shift correlation experiments to study low molecular weight
oligomers of PVDF. It is amazing how much useful characterization has been
accomplished using low field instruments and limited instrumental capabilities.
In the past few decades multidimensional NMR techniques have become
routine. Instruments with three or more rf channels have become commonplace.
In addition, the Rf and pulse programming capabilities of modern instruments
(past 5 years) have become tremendously flexible. These factors make it useful
to reconsider the nature of NMR methods for studying fluoropolymer structures.
This paper illustrates some NMR capabilities for studying fluoropolymer
structures. While some of these methods have been around for a decade or more,
their proper execution has required recent improvements in the NMR instruments
rf performance. These methods will be illustrated with applications to the study of
poly(hexafluoropropylene oxide) (PHFPO) and low molecular weight oligomers,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

and poly(vinylidene fluoride) (PVDF).

Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Fluorine NMR
The 19F NMR spectrum of PHFPO in Figure 1 exhibits the NMR
characteristics of 19F that make it both powerful for solving structure problems
and challenging to execute. The resonances are spread over a large spectral
window, in this case over 60 ppm, and their chemical shifts are very sensitive to
variations in the chemical and stereochemical environments.

Figure 1. Peak containing regions from the 470 MHz 19F 1D-NMR spectrum of
PHFPO; s are signals from identified impurities in C6F6 solvent.
Several sets of multiplets are observed near -145 ppm for the CF fluorines
in the B and C units of the structure in the Figure 1. The multiplets are not
resolved from one another because there are four structures for each CF group (4
357
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 diastereomers and their mirror images). Furthermore, the multiplets are broad and
complex due to 3-bond couplings to the CF2 and CF3 groups on the same monomer
unit, and 4-bond couplings to the CF2 groups on the other side of the ether linkage.
The nomenclature adopted in this paper to identify atoms in PHFPO/THFPO
is illustrated using the structure of THFPO in Figure 1, where the monomer units
are labeled A, B, C and Z starting from the initiation end (left) and proceeding
to the termination end of the chain. Subscripts are used to refer to the number of
fluorines on the groups within each monomer; for example, B1, B2 and B3 refer to
the CF, CF2 and CF3 groups, respectively, in monomer unit B.
2D-NMR experiments generally provide the solution to resolving overlapping
signals, and also provide correlations between resonances to identify structure
fragments. While the large spectral window makes it more likely to resolve
many resonances in complex molecules, it also creates problems because short
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

pulse widths are required for uniform excitation of resonances in standard NMR
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

experiments. In 1D-NMR experiments narrow pulses that produce small tip

angles can be used without severe consequences, but it is generally not possible
to use arbitrary flip angles in 2D-NMR experiments
Furthermore, if 19F chemical shift is to be used in the indirectly detected
dimension of a 2D-NMR experiment, a significantly long experiment time
is needed to collect the many t1 increments needed to produce good digitial
resolution in the indirectly detected dimension. In this case, a simple COSY
experiment could provide correlations that prove structure, but excellent digital
resolution is needed to resolve the multiplets in the -145 ppm region and this
same digital resolution must be maintained in the 50 ppm peak-free region in
the middle of the spectral window. A weekend-long experiment was needed to
provide a 2D-COSY spectrum with relatively poor digital resolution in the f1
dimension, even though sufficient signal-to-noise was obtained in only 15 min.
Selective excitation experiments offer a solution to these problems (18). Most
of the needed information exists in relatively small areas of the 2D-NMR spectra.
If experiments are performed to select only these areas, a few short experiments
can be performed in a few hours to provide all the needed structural information.
Selective 1H-1H COSY experiments using shaped pulses were described some time
ago (19). While the large 19F spectral window should facilitate the use of selective
excitation, the large frequency offsets compared to 1H places excessive demands
for fast frequency/phase shifting requirements during the rf pulses and are beyond
the capabilities of older NMR instruments.
The rf circuitry on modern NMR aspectrometers is very flexible and permits
the rapid amplitude, frequency and phase switching to generate arbitrary pulse
shapes. Figure 2 shows the full 19F 1D-NMR spectrum of PHFPO (bottom) along
with several selective spectra. To generate the selective spectra, selective/shaped rf
pulses were sliced into small increments (1-5 us) and the phase was progressively
shifted to track the precession of the magnetization from a group of spins using
shifted laminar pulses (SLP) (20). The selective pulses were designed to excite
(from bottom to top) the CF, CF2 and CF3/OCF2 regions.
Modern hardware is capable not only of selective excitation with large (100
kHz) offsets from the transmitter, but also of arbitrarily combining two or more

358
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 offset pulses to excite/flip the nuclear spins of atoms whose resonances occur in
two or more such regions.

NMR Methods
Typical ranges of couplings in fluoropolymers are shown on structure 1,
which shows a segment from PHFPO. One- and two-bond 19F-13C couplings
in fluorinated materials are large compared to their hydrocarbon counterparts.
This makes them ideal for characterization of fluoropolymers which give broad
lines and suffer from difficulty in detecting 2D-NMR correlations which rely on
weaker couplings. Because 1JCF and 2JCF differ by almost a factor of ten, it is
easy to implement experiments that produce correlations that are selective for one
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

or the other interaction. Two- and four-bond 19F-19F homonuclear couplings are
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

usually large, while other homonuclear couplings are much smaller.

Figure 2. 470 MHz 19F 1D-NMR spectra of PHFPO going from bottom to top:
with broadband excitation of all the resonances in the spectra, and with selective
excitation of the CF, CF2 and CF3/OCF2 regions.

In PHFPO, it is possible to use two different 19F-13C HSQC (21, 22)

(heteronuclear single quantum) 2D-NMR experiments, one tuned for 1JCF and a
second tune for 2JCF, to obtain the atomic connectivities highlighted by the blue
arrows in 1. In this way, all of the monomer fragments can be identified. A COSY
experiment would then primarily contain correlations between fluorines on either
side of the ether linkage, coupled by the large 4JFF.
359
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 3 shows the modified version of the HSQC pulse sequence used to
study the fluoropolymers which are the subject of this work. Details regarding the
implementation of this experiment can be found in reference (23), however, there
are several unique and noteworthy features. 1H decoupling is performed on a third
rf channel throughout the experiment, especially with materials such as PVDF,
which have high hydrogen content. Even the HSQC spectra of perfluorinated
PHFPOs benefit from continuous 1H decoupling as there are sometimes residual
protons present (identified by 1H 1D-NMR spectra and 13C 1D-NMR spectra
obtained with and without 1H decoupling). Several spectra are collected with
different 19F spectral windows (CF, CF2 and CF3 regions). To accomplish this,
all of the 19F pulses, with the exception of the composite inversion pulse in the
middle of t1, are low power region-selective pulses. The composite pulse in
the middle of t1 has high power, and is performed with the transmitter moved
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

to the center of the peak-containing regions of the 19F spectrum for maximum
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

effectiveness at decoupling 19F in the f1 (13C chemical shift) dimension. The

delays (1) during the TANGO (Testing for Adjacent Nuclei with a Gyration
Operator) (24) period are set based on 1JCF; and the phase of the last (135o) pulse
is set based on whether 19F-13C or 19F-C-13C correlations are to be detected. The
delays during the rest of the pulse sequence are set to 1/(4xnJCF), where the
appropriate one-bond or two-bond correlation is to be detected.

Figure 3. Pulse sequence for selective HSQC; see text for details.
Figure 4 shows schematic illustrations of various options used for perfoming
selective COSY experiments. In all cases, the schematic illustrations of the
standard COSY spectra are illustrated by the full boxes with diagonal in Figure
4a-c, while the gray areas represent the detected spectrum obtained with each
variation of the selective COSY. If the pulse scheme in Figure 4d is used with the
blue pulse shaped and selective, and the yellow pulse as a standard high power
pulse, then the strip COSY spectrum illustrated by the gray area in Figure 4a is
obtained. If both blue and yellow pulses are selective for the same region, then
the endo-COSY spectrum in Figure 4b is obtained. If the pulse scheme shown in
360
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 4e is used with blue and yellow pulses selective for the indicated region
of the f1 dimension if Figure 4c and the orange pulse selective for the indicated
region in the f2 dimension, then the exo-COSY (ECOSY) spectrum in Figure
4c is obtained. This spectrum will contain a cross-peak pattern with couplings
in all dimensions. By inserting an inversion pulse (green) in the middle of t1
then selective coupling can be removed in the f1 dimension. On modern NMR
instruments, this green pulse can be the sum of two or more arbitrary pulse shapes,
to remove multiple couplings to fluorines in different regions of the spectrum.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Figure 4. Schematic illustration of selective COSY experiments; see text for

details.

Characterization of PHFPO
Structure

The first material chosen to illustrate the NMR methodology was PHFPO
because it was relatively simple to prepare NMR samples, and had well-defined
19F spectral regions which were well separated from one another. Here we show

results from tetrameric(hexafluoropropylene oxide) (tHFPO) shown in Figure 1.

This material was readily available from the commercially available acid fluoride,
readily dissolves in many solvents, and exhibits many of the NMR characteristics
of the PHFPO backbone and selected chain-end resonances found in the spectrum
of the polymer. As illustrated in Figure 1, four sets of stereoisomers (and their
mirror images) are present due to the three stereogenic centers present in the
structure.

361
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 2D-NMR

Figure 5 shows selected data from the one-bond (top highlighted in red),
and two-bond (bottom highlighted in blue) 19F-13C HSQC spectra; showing the
correlations that prove the CF3 one-bond attachments at the chain-end and two-
bond correlations between the same CF3 fluorine and adjoining CF2 carbon. The
experiments were optimized using the couplings indicated in the figure. Note that
the 19F bound to 13C (top) is shifted by ca. 0.1ppm upfield compared the the 19F
bound to 12C as a result of the well-known isotope shift (25). Many regions such
as this were identified in the HSQC spectra, in order to identify resonances from
each of the unique structure fragments in tHFPO.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Figure 5. Selected regions from the region selective 19F13C HSQC spectra of
tHFPO showing the correlations indicated on the structure; the region from the
one-bond HSQC spectrum is shown in red and the corresponding region from the
two-bond HSQC spectrum is shown in blue.

With the four structure elements circled in blue on the structure in Figure
6 identified by HSQC experiments, the methodology then relies on COSY
correlations to prove how they are attached to one another. An expansion from
the standard COSY spectrum (ca. 1% of the total area from this 2D-NMR
spectrum) showing the correlations indicated on the structure is shown in Figure
6a. Two groups of cross peaks are seen correlating B2-C1 and A2-B1 fragments.
Poor digital resolution hides an enormous amount of detail in this region despite
the fact that thousands of t1 increments were collected over the course of a
weekend-long experiment.

362
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Figure 6. Selected regions from the COSY (a), and ECOSY (b) and decoupled
ECOSY (c) showing the region containing four-bond correlations between the
structure fragments shown on the included structure.

363
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Using selective COSY 2D-NMR experiments the spectrum in Figure 6b was
collected. Even though only 2x256 t1 increments were collected, considerably
better digital resolution is evident in this spectrum. From the data in the figure, it is
clear that there are many overlapping cross-peak patterns. All of the homonuclear
couplings can be removed in the f1 dimension by insertion of shaped inversion
pulses (at the chemical shifts of those 19F atoms to be decoupled) in the middle of
the evolution time. As a result, the spectrum in Figure 6c was obtained. Four sets
of multiplet are observed in each of the regions, from the four sets of diastereomers.
All of the couplings are retained in the f2 dimension where there is good digital
resolution, at little cost of instrument time to produce this digital resolution. At
each slice, AB patterns (2JFF=150-160 Hz) are observed for each of the unique CF2
groups; these are further split by doublet couplings to the CF group on the other
side of the ether linkage. Three-bond couplings are too small, and are not resolved.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

The shaped decoupling pulse during t1 was actually a combination of shapes at

Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

several shifts, for those 19Fs that produce the couplings in f1. An additional delay
was required at the end of the evolution period to allow creation of the antiphase
magnetization from coupling between the nuclei creating the cross-peaks. The
original paper should be consulted for details (23).

Characterization of PVDF
Structure

The primary structure components of the main-chain repeat units in PVDF

are illustrated by those shown in Figure 7. The nomenclature to describe these
structures is based on that first described by Cais et al. (7), where 0 and 2
(number of fluorine in backbone methylene groups) represent CH2 and CF2 groups
along the chain. Normal addition produces -0202- carbon sequences, however,
occasionally an inversion occurs to produce tail-to-tail (-2002-) and head-to-tail
(-0220-) sequences.

Figure 7. Structures and nomenclature used to define PVDF.

364
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 As with tHFPO above, HSQC experiments provide one-bond and two-bond
C-F correlations illustrated by the blue arrows on the structures across the bottom
of Figure 7. These experiments can be used to identify the immediate neighbors (0
or 2) of each 2 unit. COSY spectra exhibit three-, four- and five bond correlations
between 2 carbons on sequential monomer units in all threes types of structures
(-0202-, -2002- and -0220-), however, from these data it is not always possible
to identify which units produce which cross-peaks. Two-bond 19F-13C HSQC
experiments help to distinguish among these: F-C correlation in -0202- will show
correlations to the 13C shifts of the same 0 units; F-C correlations in -2002- units
will not show correlations to the 13C shifts of the same 0 units; and -0220- units
will show correlations to the 13C shifts of each others two resonances.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

2D-NMR

The one-bond and two-bond 19F-13C HSQC 2D-NMR spectra of a PVDF

sample are shown in Figure 8a and b, along with the 19F and 13C 1D-NMR spectra,
which are plotted along the corresponding axes of the 2D-NMR spectra. The 19F
signals of 020 and 022/220 carbon sequences occur in distinct regions (red and
yellow) as indicated by the labels on the figure. These 19F resonances have been
assigned based on empirical shift additivity rules and the substituent effect (7).
Distinction between the 13C resonances of 0 and 2 carbons is straightforward based
on the power of electronegative fluorine to shift the 13C resonances downfield.
These assignments of resonances from three-carbon sequences are confirmed
by the HSQC data. 19F resonances of 020 sequences only show two-bond F-C
correlation to 0 carbons (red circles in Figure 8b). However, 19F resonances of 022/
220 sequences show correlations to carbons in both the 2 and 0 carbon resonance
region (yellow circles in Figure 8b). Note that 222 sequences are not possible in
PVDF.
The double quantum filtered (dq) COSY spectrum of PVDF is shown in
Figure 9. The 19F spectral window for PVDF is relatively narrow compared
to that for PHFPO, so that it is not essential to use the selective versions of
this experiment. The correlations shown on the dqCOSY spectrum permit
identification of attachments between fragments of three carbon sequences: A
(020) to E (022), E (022) to F (220), F (220) to D (020), and D (020) to C (020).
The 19F resonances are correlated with each other via 3-, 4- and 5-bond JFF
couplings through 22, 202 and 2002 attachements, respectively. As mentioned
above, HSQC data permits distinction between these connectvities (i.e. whether
2-2 linkages occur through one-bond, -0-, or -00-).
Through combined use of HSQC and COSY based 2D-NMR experiments,
this 13 carbon sequence has been identified and its resonances assigned. Similar
interpretation of the spectra easily permits identification of 9-carbon sequences
needed to identify pentad sequences in the polymer.

365
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Figure 8. Selected regions from the one-bond (a) and two-bond (b) 19F-13C
HSQC 2D-NMR spectra of PVDF.

Experimental Section
Materials
Solvents were purchased from Aldrich Chemical Co and Cambridge Isotope
Laboratories and used as received; THFPO was purchased as the acid fluoride
from Alfa Aesar and converted to the carboxylic acid as previously described (23).
PVDF was provided by E. I. Dupont de Nemours and Co.

Instrumentation
NMR measurements were made with a Varian direct drive 500 MHz NMR
spectrometer with five rf channels. Channels 1 and 3 were designed for high
frequency (1H/19F) and channel 2 was designed and used for low frequency band
operation (31P-15N). All channels were equipped with waveform generators for
pulse shaping. The probe used was a 5mm four channel (1H/19F/13C/2H) design
with a single high-band input for the 1H and 19F signals, and pulsed field gradient
coils (capable of up to 0.7 T/m gradients). A duplexer was used with internal
filters to combine 1H and 19F rf transmitter signals, direct them to the high band
channel of the probe, take the NMR signal from the probe, isolate the 1H and 19F
signals from the sample, and direct the desired signal to the instrument receiver.
366
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Complete details describing the pulse programs used have been summarized in
reference (23).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

Figure 9. 19F-19F double quantum filtered COSY spectrum of PVDF.

Summary
Recent advances in the capabilities of NMR instrument hardware and the
software used to control that hardware warrants a reevaluation of older NMR
techniques and their utility in solution NMR studies of polymer structures. NMR
studies of fluoropolymers can particularly benefit from application of selective
excitation techniques. The unique characteristics of 19F (compared to 1H)
create difficulities in applications of 2D-NMR method of structure elucidation.
These same characteristics contribute to the successful applications of selective
excitation NMR experiments. In this paper, examples of these applications are
presented involving the characterization of PHFPO and PVDF. Unprecendented
detail is resolved in the selective 2D-COSY and HSQC spectra. These data
permit the assignment of resonances from monomer- and stereo-sequences in the
polymer backbone.
These experiments can and are being used to study polymer chain-end
structures, in order to shed light on the inititiation and termination processes
in polymerization. Similar analyses are also underway to characterize
poly(vinylidene fluoride-co-tetrafluoroethylene) (PVDF/TFE), poly(vinylidene
fluoride-co-hexafluoropropene) (PVDF/HFP) and poly(vinylidene fluoride-co-
tetrafluoroethylene- co-hexafluoropropene) (pCDF/TFE/HFP)
367
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Acknowledgments
The authors wish to thank the National Science Foundation (DMR-0905120)
and E. I. Dupont de Nemours and Co. for financial support during the preparation
of this work. We thank The Ohio Board of Regents and The National Science
Foundation (CHE-0341701 and DMR-0414599) for funds used to purchase the
NMR instrument used in this work. We wish to thank the staff of the Magnetic
Resonance Center at The University of Akron for maintaining the equipment used
in this work.

References
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021

1. Howell, J. L.; Hofmann, M. A.; Waterfeld, A.; Sipyagin, A. M.; Friesen, C.

Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

M.; Thrasher, J. S. J. Fluorinie Chem. 1998, 89, 131135.

2. Howell, J. L.; Lu, N.; Perez, E. W.; Friesen, C. M.; Novak, I.; Waterfeld, A.;
Thrasher, J. S. J. Fluorine Chem. 2004, 125, 15131518.
3. Howell, J. L.; Lu, N.; Friesen, C. M. J. Fluorine Chem. 2005, 126, 281288.
4. Kostjuk, S. V.; Ortega, E.; Ganachaud, F.; Ameduri, B.; Boutevin, B.
Macromolecules 2009, 42, 612619.
5. Durand, N.; Mariot, D.; Ameduri, B.; Boutevin, B.; Ganachaud, F. Langmuir
2011, 27, 40574067.
6. Murasheva, Y. M.; Shashkov, A. S.; Dontsove, A. A. Polym. Sci. U.S.S.R.
1981, 23, 711720.
7. Cais, R. E.; Sloane, N. J. A. Polymer 1983, 24, 179187.
8. Ovenall, D. W.; Uschold, R. E. Macromolecules 1991, 24, 32353237.
9. Pianca, M.; Barchiesi, E.; Esposto, G.; Radice, S. J. Fluorine Chem. 1999,
95, 7184.
10. Hedhli, L.; Mekhilef, N.; Moyses, S.; Lewis, R. H. Macromolecules 2008,
41, 20112021.
11. Guiot, J.; Ameduri, B.; Boutevin, B. Macromolecules 2002, 35, 86948707.
12. Ameduri, B.; Ladeaviere, C.; Delolme, F.; Boutevin, B. Macromolecules
2004, 37, 76027609.
13. Wormald, P.; Ameduri, B.; Harris, R. K.; Hazendonk, P. Polymer 2008, 49,
36293638.
14. Durand, N.; Ameduri, B.; Boutevin, B. J. Polym. Sci., Part A: Polym. Chem.
2011, 49, 8292.
15. Katoh, E.; Ogura, K.; Ando, I. Polym J. 1994, 26, 13521359.
16. Takasaki, M.; Kimura, K.; Kawaguchi, K.; Abe, A.; Katagiri, G.
Macromolecules 2005, 38, 60316037.
17. Macheteau, J.-P.; Oulyadi, H; van Hemelryck, B.; Bourdonneau, M.;
Davoust, D. J. Fluorine Chem. 2000, 104, 149154.
18. Morris, G. A.; Freeman, R. J. Magn. Reson. 1978, 29, 433462.
19. Huth, J.; Kurur, N.; Bodenhausen, G. J. Magn. Reson., Ser. A 1996, 118,
286290.
20. Patt, S. L. J. Magn. Reson. 1992, 95, 94102.
21. Bodenhausen, G.; Ruben, D. J. Chem. Phys. Lett. 1980, 69, 185188.
368
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 22. Vuister, G. W.; Boelens, R.; Kaptein, R.; Hurd, R. E.; John, B.; Zijl, P. C. M.
V. J. Am. Chem. Soc. 1991, 113, 96889690.
23. Li, X.; McCord, E. F.; Baiagern, S.; Fox, P.; Howell, J. L.; Sahoo, S. K.;
Rinaldi, P. L. Magn. Reson. Chem. 2011in press.
24. Wimperis, S.; Freeman, R. J. Magn. Reson. 1984, 58, 348353.
25. Tordeux, M.; Magnier, E.; Guidotti, J.; Diter, P.; Wakselman, C. Magn.
Reson. Chem. 2004, 42, 700703.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch021
Downloaded by CORNELL UNIV on June 21, 2012 | http://pubs.acs.org

369
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 22

Statistical Models and NMR Analysis of

Polymer Microstructure
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

H. N. Cheng*,1 and Massoud J. Miri2

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

1Southern Regional Research Center, USDA Agricultural Research Service,

1100 Robert E. Lee Blvd., New Orleans, LA 70124, U.S.A.
2Department of Chemistry, Rochester Institute of Technology, Rochester,

NY 14623, U.S.A.
*E-mail: [email protected]

Statistical models can be used in conjunction with

NMR spectroscopy to study polymer microstructure and
polymerization mechanisms. Thus, Bernoullian, Markovian,
and enantiomorphic-site models are well known. Many
additional models have been formulated over the years
for additional situations. Typically spectral interpretation
and data treatment can be done through either analytical
or simulation approaches. These can be combined into
integrated approaches for specific situations. An alternative
(and more general) approach considers the kinetics of
the polymerization process and carries out predictions of
polymer microstructures and NMR spectra. These various
methodologies are briefly reviewed here. Also reviewed
is a recent effort in the simulation category involving a
user-friendly Excel program (Polytact) that can simulate the
tacticities of a large number of statistical models, particularly
those that pertain to polyolefins made with single-site catalysts.

Introduction
High-resolution solution NMR is now a routine technique for polymer
analysis (15). It is fairly straightforward to dissolve a polymer in a solvent and
obtain a 1H or 13C spectrum. Analyses of a polymer NMR spectrum can be carried
out at different levels (Figure 1). At the simplest level, one can use the gross

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 NMR spectral features to search the spectral libraries for routine identification
or pattern recognition. At a more involved level, one can interpret the spectrum
(in whole or in part), using a variety of techniques, and assigning the observed
peaks to specific structures. From the assignments, it may then be possible to
derive detailed information on polymer microstructure, such as homopolymer
tacticity, copolymer composition and sequence, branching, defect structures, and
chain ends. With additional work, it may be possible to infer information on
monomer reactivity and polymerization mechanisms. In general, the amount and
the quality of information that can be extracted from the NMR data depend on the
methodologies used and the efforts expended.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

Figure 1. Information content available in the NMR data together with methods
and techniques used to extract the information.

Statistical models (also known as reaction probability models) have been used
extensively in polymer studies (37). They provide fundamental understanding
of polymerization processes and serve as a theoretical framework whereby NMR,
molecular weight, fractionation, and polymerization kinetics data can be analyzed
in a rational manner. In many cases the models can assist in NMR spectral
assignments and help interpret the spectral intensities. Sometimes they provide
information on the mechanisms of initiation and propagation. Suitably applied,
they may permit maximum amount of information to be obtained from each
spectrum.
In this work a brief review is made of common statistical models being used
and selected methodologies being applied in the NMR analysis of polymers, with
a particular emphasis placed on the work done in the authors laboratories.

372
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Statistical Models
Over the years, many statistical models have been used to depict
polymerization. New models have been devised for specific situations. The
interrelationships of some of these models are shown in Figure 2.

One-State Models
The simple one-state models, such as Bernoullian (B), first-order Markovian
(M1), second-order Markovian (M2), and enantiomorphic-site models (E) are
well-known and have been previously described (6, 8, 9). It is important to note
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

that with B, M1 and M2 models, the polymerization are chain-end controlled,

whereas with E model, the polymerization is catalytic-site controlled (8, 1012).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

Models that provide dual control (both chain-end and catalytic-site) are also
known and used (7, 1113). Other useful models include Markovian with
reversible propagation (1416), complex participation (17, 18), and bootstrap
models (19).

Two-State Models
These models are used when two separate states are involved in
polymerization. The two states can be consecutive or concurrent (20). The
consecutive two-state model can be applied to a polymerization where the
catalytic site switches back and forth between two states as it polymerizes; the
result may be a block copolymer consisting of blocks that conform to different
statistics. The consecutive two-state model can also be used for other block
copolymers or for a polymerization where the reaction condition changes during
polymerization (e.g., changes in monomer feed, pressure, or temperature). The
concurrent two-state model is basically a mixture model. It can be used for a
polymerization where two separate active sites are present, each site making
its own polymer according to its propagation statistics; in effect, a blend of
two polymers is formed. The concurrent two-state model can also be used if a
catalytic site switches back and forth between two states but the rate of switching
is low relative to chain propagation and chain termination, and separate chains
are formed that can be attributed to each of the two states (20).
Each of the two states can be Bernoullian (B) or enantiomorphic-site (E),
giving rise to two-state B/B, B/E, or E/E models. The first two-state B/B model
has been formulated by Coleman and Fox (21) and included both consecutive and
concurrent features. The concurrent B/B model has been used to analyze NMR
data of olefin copolymers (2225). Similarly, the concurrent B/E model has been
used to analyze the tacticity of polypropylene (26, 27). Yet another two-state
model has been reported by Ewen, et al (13, 28) and used to analyze the NMR
data of syndiotactic polypropylene made with metallocene catalysts; the model
includes dual catalytic-site and chain-end control. In addition, a general treatment
of the two-state consecutive models has been formulated (20); this is a structural
approach that is different from the Coleman-Fox treatment and is applicable to
two-state B/B, B/E, and E/E models These models have been used for the analysis
373
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of NMR tacticity data of elastomeric polypropylene (29, 30). Variant models have
also been proposed (31).
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

Figure 2. A general scheme for selected statistical models commonly used for
NMR characterization.

Multistate Models

Many polymers are found to have multiple polymeric components due

to polymerization made at different catalytic sites, at different phases, or (in
programmed monomer addition) at different times. The multistate models are
useful for the analysis of these polymers. A general methodology involving
multistate models has been reported (25, 32). The analysis is facilitated
when NMR data of polymer fractions (from fractionation or chromatographic
374
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 separation) are available. The methodology has been applied to a large number
of synthetic polymers, particularly those made from Ziegler-Natta catalysts
which may contain more than two active sites, e.g., for the analysis of copolymer
sequences (25, 3239) and homopolymer tacticity (25, 33, 38). In addition, it has
been used on natural polymers, e.g., pectin (40, 41) and alginate (4143).

Perturbed Models

Many industrial (and some lab-made) polymers exhibit varying degrees of

compositional heterogeneity. Compositional heterogeneity may also influence
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

polymer microstructure, such as tacticity, composition, and sequence distribution.

Perturbed Markovian (44, 45) and enantiomorphic-site (46) models have been
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

designed as convenient tools for the interpretation of compositional and tacticity

data. In general, compositional heterogeneity can arise from many sources
(4750). Depending on the source(s) of heterogeneity, the chemical composition
distribution (CCD) curve may take on a symmetric or non-symmetric appearance
(Table 1) (48).
For the analysis of CCD or NMR data, multistate heterogeneity can usually
be treated with multistate models. Perturbed models can be used for the treatment
of the other three types of heterogeneities. Computations can be done via either
analytical (4448) or simulation approaches (4750).

Table 1. Different types of compositional heterogeneity. (adapted from (45))

type possible source(s) of heterogeneity CCD curve
statistical statistical fluctuations in copolymer symmetric
composition
conver- different comonomer reactivities skewed or tent-shaped
sion
multistate 1. different polymers made at different variable (skewed to
initiator sites or phases multimodal)
2. programmed comonomer feeds
3. polymer blending
4. polymers from biological sources
process fluctuations in polymerization process symmetric or slightly
conditions skewed

Kinetic Models

Instead of reaction probabilities, an alternative treatment (51) uses the actual

reactions taking place during polymerization. The resulting kinetic scheme can
then be used to generate all requisite information corresponding to the observed
NMR data, e.g., composition, sequence, branching, defect structure, and chain
ends. These kinetic models are especially useful for the treatment of complex
NMR data. For example, the kinetic model of a simple binary copolymerization
375
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 has been reported (51). A detailed study of short- and long-chain branching in
low-density polyethylene using kinetic modeling has also appeared (52).

Other Models

Most of the preceding discussion deals with homopolymer tacticity or

sequence distribution of binary copolymers. The first-order Markovian model for
ternary copolymerization (terpolymerization) is also known (53); an excellent
example of its utility has been shown by Carman et al (54) for the analysis of
NMR spectra of ethylene-propylene rubbers. The first-order Markovian model
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

for quaternary copolymerization (tetrapolymerization) has also been formulated

(55) with mathematical expressions for all relevant sequences. It has been used
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

successfully for styrene-butadiene copolymers (56), where 1,2-, 1,4-trans, and

1,4-cis additions of butadiene to the propagating chain are regarded as three
separate comonomers. The same model has been used for stereoirregular and
regioirregular polypropylene (57), where primary and secondary insertions of
propylene in two configurations are considered the four comonomers. The
mathematical expressions for a pentapolymerization (involving 5 comonomers)
would be very complex (55). A simulation approach for a pentapolymerization
has been published for ethylene-propylene rubber (48, 58); in that case, the
five comonomers are ethylene, propylene in the primary insertion (up and
down configuration), and propylene in the secondary insertion (up and down
configuration).
The versatility of metallocene catalysts gives rise to new statistical models.
Some of these have been recently reviewed (59). A new model (E-B gen) will
be described in the last section of this article.

Application Methodologies
Analytical Approaches

Whereas the models are useful for an understanding of polymerization

mechanism and statistics, an appropriate methodology is needed to apply these
models to the treatment of NMR data. Many methods and techniques have been
developed for this purpose (19). Most of the methods can be grouped under the
category of analytical approaches (5, 6, 60):

Thus, the NMR data of a polymer are first acquired and the spectral features
assigned to pertinent polymer microstructures. Spectral intensities are then
obtained, and calculations made to give copolymer composition and sequence
distribution. A more refined treatment takes all assignable intensities in a
spectrum and fits them to an appropriate statistical model, from which reaction
probabilities can be derived (54, 6062). This approach has also been extensively
376
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 used for many statistical models (57, 11, 12, 25, 32, 48), often with some
variations in methodologies (63, 64).

Simulation Approaches

The second category of approaches can be called simulation (or "synthetic")

approaches. The starting point here is an appropriate statistical model, from
which composition and sequence distribution for a set of model parameters can
be calculated through theory or computer simulation. The observed and the
calculated sequence distribution can be compared and adjustments in the model
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

parameters made until a satisfactory fit is obtained. A number of people have

utilized this approach over the years (59, 60, 62, 6568). The advantage of this
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

approach is that it is often easier to calculate the sequence distribution or to

simulate the polymerization than to engage in a direct data fitting procedure for
many polymeric systems.

In one variation of this approach, one can take the reaction probabilities of a
model and use Monte Carlo method to generate an ensemble of polymer chains,
from which composition and sequence (and their heterogeneities) can be gathered
(49, 50, 6668). When good correlations between microstructures and chemical
shifts are known, one can produce a predicted NMR spectrum. As an example,
a general computer program has been written that can be used to calculate the
copolymer sequences and to simulate the 13C NMR spectra of vinyl and vinylidene
copolymers and terpolymers (67, 68). For polymers exhibiting stereoisomerism
and regioisomerism, detailed 13C shift rules incorporating tacticity, head-to-tail,
head-to-head, and tail-to-tail structures are needed to produce simulated shifts or
spectra. For example, such 13C rules for polypropylene and ethylene/propylene
copolymer have been obtained (69). For these two polymer systems, it is
possible to simulate a range of polymer microstructures, including tacticity,
regiosequences, and chain ends (58, 70).

Integrated Approaches

A third category of methods incorporates elements of analytical and

simulation approaches. The use of both approaches in suitable cases provides
more detailed understanding of polymer microstructure and NMR shift/structure
correlations. For example, for copolymers obtained at high conversion, the direct
use of a statistical model can lead to wrong results; an integrated approach permits
more correct results to be obtained (71). Other examples of integrated approaches
have also been published (48, 72).

377
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Kinetic Simulation Approaches
These approaches can be used with kinetic models (above). A schematic of
this type of approaches is shown below.

If we can formulate a realistic kinetic scheme of a polymerization under

consideration, the scheme can then be used for simulation of polymer chain
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

growth and NMR spectral prediction. Of course, rate constants need to be known
either a priori or through model-fitting. Detailed descriptions of this approach
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

have been reported earlier (51, 52).

Program Polytact
Recently a convenient and pragmatic program, called Polytact, was reported
by Miri, et al (59). It was written on the Excel software with the objective
to calculate all relevant characteristics of the polymer tacticity and to present
them in graphical form in a user-friendly manner using macros. Six types of
tacticities can be produced with the program: isotactic, syndiotactic, atactic,
hemiisotactic, stereoblock, and heterotactic (the latter meaning that ideally two
adjacent substituents point in one direction followed by two substituents pointing
in the other direction). Seven one-state statistical models were built into the
program: B, M1, M2, E-B iso, E-B syn, E-B gen, and the E-M1 combination.
The E-B gen model is introduced for the first time in this program (59). It
enables a user to model and simulate four different types of enantiomorphically
controlled tacticities with one single model: isotactic, syndiotactic, atactic and
hemiisotactic polymer. It is based on a simplified scheme, in which it is only
relevant what the probabilities are for each of the two types of enantiofaces of the
monomer to coordinate with each of the catalysts two lateral, enantiotopic sites.
Specific types of catalyst symmetry groups need not be explicitly considered in
this model. Also, the equations for the pentads for the E-M1 model have been
published for the first time in the article cited (59). The E-B iso and E-B syn
models are to be used for isotactic and syndiotactic polyolefins, respectively. The
model simulations have been applied to twenty polymers with different tacticities
found in the literature (59).
A screen shot of the summary sheet of the program is shown in Figure 3. The
programs user can select each of these models on an input window, and enter the
corresponding numerical probabilities, e.g., Pmm and Prr in case of the M1 model.
The program then produces the following output: (1) the numerical dyad, triad,
tetrad and pentad values for the specific input probability(ies), (2) bar diagrams
for the triad and pentad values, (3) 2D or 3D graphs for the triad and pentad
distribution over the entire probability range(s), depending on the model, (4) the
average sequence lengths for the meso and racemic dyads, (5) a sequence length
distribution graph, and (6) a graph for the simulation of 50 units in the polymer.
378
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 In addition, a tab can be selected on the input window to use short-cuts for
the models (with stored ideal probabilities) to produce the six types of tacticities
mentioned above. These are intended to help the beginning user to get familiar
with the concepts more easily. Furthermore, a third tab can be selected to choose
from different metallocene symmetries that cause certain tacticities; for example,
by clicking the icon with the figure representing a metallocene with a C1 chirality,
the program will produce data for a hemiisotactic polymer as output.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

Figure 3. Screenview of the summary sheet of program Polytact with the movable
input window on the right (applying the E-B gen model).

More detailed views of the 2D and 3D plots of the triad and pentad
distributions are available on a separate spreadsheet in Polytact. These can be
used to determine quickly which types of triads or pentads cannot be produced
with certain models or what their maximum or minimum value is. For example,
the [mmmm] pentad reaches a maximum 6.25% for a syndiotactic polymer (with
a = 0.5), which is also the minimum value of [mmmm] for an isotactic polymer
(with b = 0.5) using the E-B syn and E-B iso models, respectively. However, the
[mmmm] pentad can become zero using the E-B gen model (when either both of
its probabilities c and d approach zero or both probabilities approach 1).
A major use of the program Polytact is to compare experimental n-ad
sequences, which can be obtained from NMR spectroscopy, with those calculated
by the program. To test if the experimental data are in agreement with one of
the seven models, one needs to take a set of n-ads and decide on the particular
probabilities associated with a model. These probabilities then need to be
entered into the program, and the resulting n-ad sequences can be compared
to the corresponding experimental sequences. If the differences between the
experimental n-ads and those calculated by the program lie within a reasonable

379
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 margin of error, the actual polymerization is considered compatible with the
particular model within the limit of experimental error.
The sequence distributions obtained can also be used to make predictions
about the degree of order and crystallinity of a polymer or its other physical
properties. If a polymer with a certain tacticity is to be grafted or modified during
post-polymerization, predictions can be made about the polymers reactivity as
well.
The limitation of the program is that it currently includes only the seven
given statistical models. Thus, it does not include tactictities which are based on
two-state or multi-state models.. It may also be noted that the program represents
a simulation approach; thus, optimization of reaction parameters can be done
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

manually, but there is currently no provision for automated iteration.

The program will be made available by the authors upon request (e-mail:
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

[email protected], or [email protected]).

Acknowledgments
Thanks are due to the many collaborators over the years. These include (in
alphabetical order) G. N. Babu, Mark A. Bennett, J. C. W. Chien, John A. Ewen,
Leo J. Kasehagen, R. A. Newmark, Michael T. Roland, and Stanley B. Tam. G.
H. Lee, Thomas G. Neiss, and David A. Smith contributed to the applications of
some of the models to several polymer systems. Moreover, B. P. Pritchard at RIT
contributed to the writing of Program Polytact.
Mention of trade names or commercial products in this publication is
solely for the purpose of providing specific information and does not imply
recommendation or endorsement by the U.S. Department of Agriculture. USDA
is an equal opportunity provider and employer.

References
1. Cheng, H. N., English, A. D., Eds.; NMR spectroscopy of polymers in solution
and in the solid state; ACS Symposium Series No. 834; American Chemical
Society: Washington, DC, 2002.
2. Brandolini, A. J.; Hills, D. D. NMR spectra of polymers and polymer
additives; M. Dekker: New York, 2000.
3. Mirau, P. A. A practical guide to understanding the NMR of polymers; Wiley:
Hoboken, NJ, 2004.
4. Kitayama, T.; Hatada, K. NMR spectroscopy of polymers; Springer: New
York, 2004.
5. Cheng, H. N. Structural studies of polymers by solution NMR; RAPRA
Review Reports, Vol. 11, Number 5; Rapra: Shewsbury, U.K., 2001.
6. Cheng, H. N. In Encyclopedia of NMR; Grant, D. M., Harris, R. K., Eds.;
Wiley: New York, 1996; pp 37133721.
7. Cheng, H. N. In New advances in polyolefins; Chung, T. C., Ed.; Plenum
Press: New York, 1993; pp 1530.
380
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 8. Bovey, F. A. High Resolution NMR of Macromolecules; Academic Press:
New York, 1972.
9. Koenig, J. L. Chemical Microstructure of Polymer Chains; Wiley: New York,
1980.
10. Price, F. P. In Markov Chains and Monte Carlo Calculations in Polymer
Science; Lowry, G. G., Ed.; Marcel Dekker: New York, 1970; Chapter 7.
11. Cheng, H. N. In Transition Metal Catalyzed Polymerizations: Ziegler Natta
and Metathesis Polymerizations; Quirk, R. P., Ed.; Cambridge Univ. Press:
Cambridge, 1988, p 599.
12. Cheng, H. N. J. Appl. Polym. Sci. 1988, 36, 229.
13. Ewen, J. A.; Elder, M. J.; Jones, R. L.; Curtis, S.; Cheng, H. N. In Catalytic
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

Olefin Polymerizations; Keii, T., Soga, K., Eds.; Kodansha: Tokyo, 1990, p.
439.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

14. Izu, M.; ODriscoll, K. F.; Hill, R. J.; Quinn, M. J.; Harwood, H. J.
Macromolecules 1972, 5, 90.
15. Cais, R. E.; Hill, D. J. T.; ODonnell, J. H. J. Macromol. Sci., Chem. A 1982,
17, 1437.
16. Szwarc, M.; Perrin, C. L. Macromolecules 1985, 18, 528.
17. Seiner, J. A.; Litt, M. Macromolecules 1971, 4, 308.
18. Hill, D. J. T.; ODonnell, J. H.; OSullivan, P. W. Prog. Polym. Sci. 1982, 8,
215.
19. Harwood, H. J. Makromol. Chem., Makromol. Symp. 1987, 10/11, 331.
20. Cheng, H. N.; Babu, G. N.; Newmark, R. A.; Chien, J. C. W. Macromolecules
1992, 25, 6980.
21. Coleman, B. D.; Fox, T. G. J. Chem. Phys. 1963, 38, 1065.
22. Ross, J. F. In Transition Metal Catalyzed Polymerizations, Alkenes and
Dienes; Quirk, R. P., Ed.; Harwood Academic: New York, 1983; p 799.
23. Cozewith, C. Macromolecules 1987, 20, 1237.
24. Floyd, S. J. Appl. Polym. Sci. 1987, 34, 2559.
25. Cheng, H. N. J. Appl. Polym. Sci. 1988, 35, 1639.
26. Zambelli, A; Locatelli, P.; Provasoli, A.; Ferro, D. R. Macromolecules 1980,
13, 267.
27. Zhu, S.-N.; Yang, X-Z.; Chujo, R. Polym. J. (Tokyo) 1983, 12, 859.
28. Ewen, J. A. In Catalytic Polymerization of Olefins; Keii, T., Soga, K., Eds.;
Kodansha-Elsevier: Tokyo, 1986; p 271.
29. Babu, G. N.; Newmark, R. A.; Cheng, H. N.; Llinas, G. H.; Chien, J. C. W.
Macromolecules 1992, 25, 7400.
30. Kravchenko, R.; Masood, A.; Waymouth, R. M.; Myers, C. L. J. Am. Chem.
Soc. 1998, 120, 2039.
31. Gauthier, W. J.; Collins, S. Macromolecules 1995, 28, 3779.
32. Cheng, H. N. In New Advances in Polyolefins; Chung, T. C., Ed.; Plenum:
New York, 1993; pp 159174.
33. Cheng, H. N. ACS Symp. Ser. 1989, 404, 174.
34. Cheng, H. N.; Kakugo, M. Macromolecules 1991, 24, 1724.
35. Cheng, H. N. Polym. Bull. 1990, 23, 589.
36. Cheng, H. N. Macromolecules 1991, 24, 4813.
37. Cheng, H. N. Polym. Bull. 1991, 26, 325.

381
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 38. Cheng, H. N. Makromol. Chem., Theor. Simul. 1992, 1, 415.
39. Masson, P.; Llauro-Darricades, M.-F.; Spitz, R.; Cheng, H. N. Int. J. Polym.
Anal. Charact. 1996, 2, 379.
40. Neiss, T. G.; Cheng, H. N.; Daas, P. J. H.; Schols, H. A. Macromol. Symp.
1999, 140, 165.
41. Neiss, T. G.; Cheng, H. N. ACS Polym. Prepr. 2000, 41 (2), 76.
42. Cheng, H. N. Polym. Bull. 1999, 43, 247.
43. Neiss, T. G.; Cheng, H. N. ACS Symp. Ser. 2002, 834, 382.
44. Cheng, H. N. Macromolecules 1992, 25, 2351.
45. Cheng, H. N. Macromolecules 1997, 30, 4117.
46. Cheng, H. N. Makromol. Chem., Theor. Simul. 1993, 1, 561.
Downloaded by UNIV OF DELAWARE MORRIS LIB on June 20, 2012 | http://pubs.acs.org

47. Cheng, H. N. J. Appl. Polym. Sci.: Appl. Polym. Symp. 1992, 11, 21.
48. Cheng, H. N. Polym. News 2000, 25, 114.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch022

49. Cheng, H. N.; Tam, S. B.; Kasehagen, L. J. Macromolecules 1992, 25, 3779.
50. Cheng, H. N.; Kasehagen, L. J. Macromolecules 1993, 26, 4774.
51. Cheng, H. N.; Kasehagen, L. J. ACS Polym. Prepr. 2003, 44 (1), 381.
52. Cheng, H. N.; Kasehagen, L. J. ACS Polym. Prepr. 1997, 38 (1), 863.
53. Ham, G. E. Copolymerization; Interscience Publishers: New York, 1964.
54. Carman, C. J.; Harrington, R. A.; Wilkes, C. E. Macromolecules 1977, 10,
536.
55. Roland, M. T.; Cheng, H. N. Macromolecules 1991, 24, 2015.
56. Cheng, H. N.; Roland, M. T. ACS Polym. Prepr. 1991, 32 (1), 549.
57. Cheng, H. N. Macromol. Theory Simul. 1994, 3, 979.
58. Cheng, H. N.; Bennett, M. A. Makromol. Chem. 1987, 188, 2665.
59. Miri, M. J.; Pritchard, B. P.; Cheng, H. N. J. Molecular Modeling, 2010,
published online: DOI 10.1007/s00894-010-0880-8.
60. Cheng, H. N. J. Appl. Polym. Sci.: Appl. Polym. Symp. 1989, 43, 129.
61. Cheng, H. N. J. Chem. Inf. Computer Sci. 1987, 17, 8.
62. Crowther, M. W.; Begemann, J. H.; Levy, G. C. ACS Symp. Ser. 1989, 404,
161.
63. Cheng, H. N. ACS Polym. Prepr. 2003, 44 (1), 381.
64. Cheng, H. N.; Gillette, P. C Polym. Bull. 1997, 38, 555.
65. Harwood, H. J. J. Polym. Sci.: Part C 1968, 25, 37.
66. Harwood, H. J.; Chen, T.-K.; Lin, F.-T. ACS Symp. Ser. 1984, 247, 197.
67. Cheng, H. N.; Bennett, M. A. Anal. Chem. 1984, 56, 2320.
68. Cheng, H. N. Trends Anal. Chem. 1994, 13, 95 and references therein.
69. Cheng, H. N.; Bennett, M. A. Makromol. Chem. 1987, 188, 135 and
references therein.
70. Cheng, H. N. Macromol. Symp. 1994, 86, 77.
71. Cheng, H. N. Int. J. Polym. Anal. Charact. 1997, 4, 71.
72. Cheng, H. N. Int. J. Polym. Anal. Charact. 1996, 2, 439.

382
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 23

Composition and Phase Morphology of

Hard/Soft Acrylic Latex Polymers: NMR and
AFM Study
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Kebede Beshah* and Robert F. Antrim

The Dow Chemical Company, 727 Norristown Road,

Spring House, PA 19477
*E-mail: [email protected]

Butyl Acrylate (BA) and Methyl Methacryate (MMA)

copolymers with different molar ratios in latex particles lead
to segregated hard and soft nanodomains. The presence
of both monomers in the hard and soft nanodomains is
challenging for traditional techniques for morphology studies
such as Transmission Electron Microscopy (TEM) due to
lack of atomic contrast. We have used NMR experiments
together with statistical analysis of comonomer sequences to
determine the composition of the monomers in each phase
coupled with Atomic Force microscopy (AFM) to probe the
phase morphology and particle sizes. The synergy of the two
techniques provides the opportunity to probe interface mixing
between the continuous soft phase and the segregated hard
phase that has some impact on the final size of the hard and soft
phases after film formation.

Introduction
Emulsion polymer particles for coatings applications (about 100 nm in size)
need to coalesce during film formation in order to form transparent and crack
free films that provide hardness for better scrub and dirt pickup resistance of
paints (13). Such opposing properties of softness for coalescence and hardness
after film formation have been circumvented by applying coalescing agents such
as Texanol (2,2,4-trimethyl-1, 3- pentanediol monoisobutyrate) that help with
plasticization of the polymer during film formation and later evaporates to leave a

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 relatively hard film. With the drive to reduce and even eliminate volatile organic
compounds (VOC) from the environment, formulators are looking for ways to
achieve film formation at ambient temperature while maintaining certain degree
of hardness. One of the approaches involve designing a hard and soft morphology
within the same latex particle of the binder (4). The soft component delivers
the film formation property at ambient temperatures while the hard components
provide the mechanical strength and aids in dirt pick-up resistance. In some
designs the soft phase forms the continuous phase while the hard phases could be
as low as 10% and with nanodomain sizes below 20 nm. Such designs make it
very difficult for the standard analytical methods such as DSC, DMA, and TEM
to study phase segregation (5).
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

The challenge is exacerbated by the design of the hard and soft components
from the same pair of monomers, methyl methacrylate (MMA) and butyl acrylate
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

(BA). MMA rich BA/MMA copolymer provides the hard phase while a BA rich
BA/MMA copolymer forms the soft phase. A lack of atomic mass differentiation
and the similarities in components makes it difficult to image by TEM. In addition,
the lack of specific stains to differentiate MMA from BA is a limitation of TEM.
The widely dispersed nanodomains of less than 20 nm sizes provide challenges to
detect a glass transition temperature of the hard phase by DSC and DMA.
In this study we have used a combination of solution and solid state
NMR together with Scanning Probe Microscopy (SPM) to determine the phase
behaviors and obtain quantitative information on the phase segregated latex
particles. SPM is sensitive to hard and soft components and the technique is
capable of differentiating these in a film. In some instances, the similarities in
chemical composition could also lead to phase mixing that has significant effect
on detecting less that 5 nm domain sizes even by Scanning Probe Microscopy
(SPM). NMR permits the study of such systems as solid films by solid state NMR
or in solution where the comonomer sequence reveals the heterogeneity of the
polymer and the possible phase segregation.

Experimental Section
Materials

The emulsion polymers for this study were made using standard emulsion
polymerization techniques (6). The model ladder statistical emulsion polymers
of varying BA/MMA mole ratios are synthesized using standard procedures with
thermal initiation and a gradual monomer feed time of about 2 hours for the
polymerization. This provides for uniform composition of the copolymers even
for monomers with varying reactivity ratios compared to batch polymerization
that invariably leads to heterogeneous polymer chains as dictated by reactivity
ratios of the monomers than by intentional design. The complex polymers
(Samples 2, 3 and 4) require a staged feed protocol in order to achieve the desired
hard and soft components.

384
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Nuclear Magnetic Resonance Spectroscopy

The solution and solid state NMR experiments were all done on a Bruker
AVANCEIII 400 MHz spectrometer. The solution NMR experiments were done
on latex polymers that were dissolved in tetrahydrofuran (THF-d8) at room
temperature. For the solid state experiments, circular disks of about 6 mm in
diameter were punched out of the films and stacked in a 7 mm sample rotor for
cross polarization and magic angle spinning (CPMAS) experiment. We have also
used 4 mm sample rotors where the film is introduced after a strip had been rolled
to fit the inner diameter of the rotor. The Cross Polarization mix times for the
high temperature experiments were chosen to be 1 ms in order to optimize the
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

discrimination of the hard and soft components.

We have used the well established Bernoullian distribution (7, 8) for a binary
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

radical copolymerization of BA and MMA, the statistical comonomer sequence is

well approximated by:

Where: m = number of sites; for our triads, m = 3.

n = 0,1,2,,m
fBA = the fraction of one of the components; in this case, mole ratio of BA.

Scanning Probe Microscopy ( SPM), a.k.a. AFM

Characterization of these emulsions consisted of evaluating the dried films

and faces cut tangential to the surface of the film. All AFM images have been
obtained using Brukers D 5000 scanning probe microscope, Santa Barbara, CA.
The tapping mode was used to obtain the images and operated under ambient
conditions. A FESP tip (Bruker) was chosen for imaging and operated at a
frequency 5% lower than its harmonic frequency. The engage set point was 0.8,
gains and scan speed were adjusted to allow good tracking of the tip with the
surface. Both phase and height images were collected simultaneously.
The dried films were prepared using 8 mil gauge bar to draw the emulsion
along Mylar substrate. The film was then air dried. Small samples were placed
into the instrument for imaging. All cryo faces were prepared by air drying a
puddle of material. A small piece was cut out and glued onto a bulls eye pin mount
(Electron Microscopy Sciences, Hatfield, Pa) and cryo faced on a Reichert FCS
Cryo Ultramicrotome using a diamond knife (Diatome, Hatfield, Pa) , at -60C.
Image analysis was carried out using either Scanning Probe Image Processor
(SPIP) or image analysis on version 6 software.

385
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Results and Discussion
Solution State NMR: Quantitative Analysis with Standards

The solution state 1H NMR lineshapes of BA/MMA statistical copolymers

have characteristic signals due to the comonomer sequence structures (911).
Model samples with a composition ladder provide a vivid illustration of the effects
of nearest and next nearest neighbors in the chemical shifts of species of interest.
As mentioned above, these copolymers were made with a standard radical
polymerization of emulsion polymers. Hence, they illustrate close to a random
distribution of near neighbors distribution since the reactivity ratio differences are
compensated for by slow feed rates of monomers during polymerization.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Figure 1. Partial 1H NMR spectrum of BA/MMA statistical copolymers and a

block copolymer to illustrate the effect of near neighbors on the spectra.

Figure 1 shows an expanded region of interest of the 1H spectrum from 3.4 to

4.3 ppm of the ladder composition of BA/MMA copolymers. The signals centered
at 3.6 ppm are due to OCH3 (methyl ester) of MMA and the signals centered at
4 ppm are due to OCH2- (methylene ester) of BA as shown by the circles in the
figure. These signals are particularly sensitive to nearest and next nearest neighbor
monomers in the polymer sequence. For example, the OCH2- signal of BA is
sensitive to whether it is in a triad sequence between two other BAs (BBB), or a
BA and MMA unit (BBM), or MBM triad. Similarly, the MMA signal gives rise to
a variety of peaks compared to a poly methyl methacrylate homopolymer, which
has just one peak centered at 3.6 ppm, due to sequencing effects (BMM and BMB).
We are not going to go into details of sequence distribution analysis in this report.
Extensive studies of Bernoullian distribution in copolymers have been reported
by other workers (7, 8). We will focus only on the part of the sequence analysis
relevant to our morphology study.
386
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 As illustrated in the figure, the relative intensities of the fine structures for BA
and MMA change as the molar ratio of BA and MMA changes in the sample. We
have only shown the effect of triad sequences, but one can envision that in some
favorable cases we could also have well resolved signals from tetrad and pentad
sequence effects on the spectra giving us rich information on copolymer sequences
as discussed by many authors in a variety of polymers including heterogeneity
effect studied by perturbed first-order Markovian distribution (720).
Since these lineshapes of random distributions are characteristic of radical
polymerization, we can use them to determine if we have any heterogeneous
composition that would lead to hard and soft morphologies.
The glass transition temperature (Tg) of poly methyl methacrylate
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

homopolymer is greater than 110C depending upon the tacticity while poly butyl
acrylate has Tg about -60C. There is a linear relationship between Tg and MMA
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

or BA fraction in the copolymer as shown in Figure 2 below. Heterogeneous

polymers with segregated BA rich and MMA rich components would lead to a
SOFT (low Tg component from the BA rich composition) and a HARD (high Tg
component from the MMA rich composition). In some instances it is difficult to
register the presence of segregated hard and soft phases due to the total amount
and the particle size of the dispersed phase.

Figure 2. Tg of BA/MMA copolymers as a function of mole fractions.

Heterogeneous Polymer of Known Composition

Figure 3 shows the high resolution 1H NMR spectrum of a heterogeneous

polymer, Sample 2. We have also plotted model polymers of known composition
to assess the composition of the hard and soft phases. As indicated above,
the MMA signals of Sample 2 show fine structures due to triad monomer
sequences (BMM and MMM) similar to copolymer composition in the range
of 20BA/80MMA to 40BA/60MMA. Based on the integration of the BMM
and MMM triad sequences as discussed below, we estimate the composition
387
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of the polymer that is responsible for the MMA signal pattern in Sample 2 is
34BA/66MMA which is essentially the similar to the synthesis ratio of 35%BA
and 65%MMA.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Figure 3. Solution NMR of heterogeneous polymer (Sample 2) and reference

homogeneous statistical copolymers.

On the other hand, the BA signal at 4 ppm for Sample 2 resembles that of 100%
BA. In the model sample with about 80% BA we note that the MMA signals have
strong features at 3.5 ppm due to neighboring BA. These signals are not detected
in the heterogeneous Sample 2 spectrum, which indicates that the BA component
of the SOFT phase of Sample 4 is close to 100%.

Figure 4. 2m AFM image of Sample 2, phase image (left) and height image
(right). A circle is drawn to aid visualization of the individual particles.

Using the Bernoullian distribution equation (experimental section) and the

intensities of the signals of the triad sequences, we estimate the MMA rich
388
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 component to be 24% of the total polymer with composition of 65MMA/35BA
and the BA rich component to be 76% of the total polymer with composition of
greater than 90% BA. These values agree well with the synthesis composition of
25% (64MMA/34BA/2AA) and 75% (92BA/7AN/1AA).
From DSC measurement we could only detect one phase transition at -26C
that is attributed to the BA rich component of the heterogeneous polymer. The
lack of a clear transition for the MMA rich phase in spite of the 25% hard phase
components could indicate the presence of dispersed nanodomains of the MMA
rich phase. We used Scanning Probe Microscopy (SPM), sometimes referred to
as Atomic Force Microscopy (AFM), which excels in the imaging of hard/soft
polymer domains.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

In Figure 4, the 2 m phase image, on the left, shows greater details of

the particles surface. Here we can start to see the individual spherical particles
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

centered at 400 nm in diameter, and the well distributed smaller hard component
(bright spots) on the surface. In addition, bright non-spheroid features are also
seen on top of the particles, which are tentatively assigned to surfactant migration
to the surface. The hard components of the particles are readily observed on the
surface, which are well controlled in particle size and spacing.

Figure 5. 2 m AFM of cryo faced Sample 2; phase image on left and height
image on right.
To determine if the hard component is found only on the outer surface of the
particle or within the particle, a cryo microtomy with perpendicular incision was
used to reveal the interior of the film and particles. The phase image of the cryo
surface in Figure 5 with higher magnification (2m x 2m) shows more clearly
the edges of the particles and the distribution of the harder component. It is seen
that the hard phase is predominately found on the outer edge of the particles. The
emulsion particles after film formation no longer look as spherical as they did
on the surface. The deformation of the particle is commonly seen during film
formation. Deformation of the particles to form the film increases the contact
389
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 surface area between the particles forming a more rigid structure. From image
analysis, we obtained that 20% of the polymer is from the hard component in good
agreement with 25% obtained by NMR.

Heterogeneous Polymer with Unknown Composition

In some instances, the polymerization process could lead to compositional

heterogeneity that is not readily deduced from the synthesis process. These are
probably the best conditions for the application of NMR and SPM techniques
where NMR provides the compositional information and SPM the phase
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

morphology as we saw above for a known system.

In Figure 6, we have solution state 1H NMR spectra of three samples. For
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

better visual comparison, the spectrum of Sample 1 is also overlaid on the spectrum
of Sample 3 and the Sample 3 spectrum on Sample 4 with dashed lines. Note that
the O-CH2- ester signal of BA (at about 4 ppm) is essentially identical for the
three copolymers, while there is a difference in the MMA ester (-O-CH3) signal.

Figure 6. Solution state 1H NMR of Samples 1 (statistical 58BA/42MMA

copolymer) and heterogeneous polymers, Samples 3 and 4.

390
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 From the integrals of MMA and BA signals as well as the triad sequence
structures, we can readily determine that the signal for Sample 1 is from a
homogeneous random polymerization of 58BA/42MMA copolymer as we have
observed in the ladder compositions in Figure 1. On the other hand, the spectrum
for Sample 3 shows similar lineshape as that of Sample 1 except for higher
intensity of one of the multiplicity at 3.6 ppm , the MMM triad; compare this
signal with the spectral patterns in the model ladder compositions in Figure 1. A
random copolymer of BA and MMA in Sample 3 could not explain this higher
intensity at 3.6 ppm. Note that the 3.6 ppm signal is characteristic of MMA
rich copolymer (bottom of Figure 1) or from a blocky poly methyl methacrylate
sequence. So, this additional intensity at MMM has to arise from another
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

heterogeneous sequence distribution of the polymer that is mainly due to MMA

rich phase.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

The signal for Sample 4 shows even higher MMA block (MMM) signal at 3.6
ppm compared to Sample 1 and Sample 3, while the rest of the spectral pattern
fits a similar 58BA/42MMA random copolymer lineshape. Hence, Sample 3 and
4 are heterogeneous polymers with one phase comprised of BA rich copolymer
(58BA/42MMA) and another that is predominantly MMA. We can now assess
the relative composition of the heterogeneous phases based on the precedent
qualitative analysis since we know the actual composition of Sample 1. The BA
signal in Samples 3 and 4 is from a BA rich polymer matrix (phase) similar to
Sample 1, hence the BA ratio in the soft phase has to be 58%, the same as Sample
1. At this stage we can assume that almost all of BA is derived from the soft phase
of the 58% BA copolymer and we can rewrite the relative integration we obtained
from the figure to reflect that we have a 58% BA containing copolymer. Hence,
the proportional amounts of all BA and MMA in Sample 3 copolymer is rewritten
from 53BA/47MMA to 58BA/51MMA, a simple proportional adjustment.
Knowing that we have now a 58%BA/42%MMA component, we further
rewrite the total composition as; 58BA/51MMA = 58BA/42MMA//9MMA =
53.2BA/38.5MMA//8.3MMA, to normalize the total composition to 100%. Given
that we do not observe lineshape difference between Sample 1 and Sample 3 and 4
except for the MMM triad sequence, we conclude that there might be only a small
amount of BA, if any, in the MMA rich component of the heterogeneous polymer.
Similarly, we determine that Sample 4 is 51.8BA/37.5MMA//10.7MMA. Hence,
Sample 3 and 4 are heterogeneous polymers with about 90% BA rich soft phase
and about 10% predominantly poly methyl methacrylate polymers.

Quantitative Analysis without a Standard

In the previous analysis we used a standard copolymer that has essentially
identical BA rich copolymer which helped with a visual comparison of the
heterogeneous polymer. It is not always possible to obtain a copolymer with
exactly the same composition as one of the components of the heterogeneous
polymer. So, we need an approach that would not require a standard sample
to match the compositions until we find the one that fits, which could lead to a
tedious trial and error task in synthesis. Even though the 1H NMR lineshape is
sensitive to comonomer compositions, accurate integration of the triad signals
391
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (BBB, BBM, MBM) is difficult due to peak overlap. So, we turn to 13C NMR
spectroscopy where one of the side chain methylene signals (-O-CH2-C*H2
-CH2-CH3) at about 30.1 30.8 ppm is also sensitive to comonomer sequences.
One of the triad segments (MBM) for -O-CH2-C*H2 -CH2-CH3 is uniquely
resolved from the rest of the signals at 30.28 ppm permitting an integration of
this signal as shown in Figure 7. In the Figure, we have included four random
copolymers with known compositions of random distribution for the validation
of the approach and the two heterogeneous Samples 3 and 4. Since we can only
resolve the MBM triad sequence signal, we will just determine its ratio with
respect to all the triads (MBM, BBB, MBB) intensities.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Figure 7. A partial 13C solution NMR of statistical homogeneous copolymers (top

4) and heterogeneous polymers (bottom 2). Only the highlighted -O-CH2-C*H2
-CH2-CH3 signal is shown.

392
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 We can now calculate the probability of the MBM occurrence for any
composition of BA/MMA using Equation 1. For example, for a 39BA/61MMA
copolymer with mole ratio of 33.3BA/66.7MMA; for triads, m=3 and fBA = 0.333,
the various relevant probabilities (P) are given by:

PBBB = (fBA)3 = (0.333)3= 0.037

PMBB = PBBM= 2/3 (PBBM, MBB, BMB) = 2(fBA)2(1-fBA)1 = 2(0.333)2(0.667)1 =
0.148
PMBM = 1/3 (PMBM, MMB, BMM) = (fBA)1(1-fBA)2 = (0.333)1(0.667)2 = 0.148

The probability, PMBB, is the placement of two BAs and one MMA in a triad
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

sequence; BBM, MBB, BMB. For our BA signal analysis, we are only interested
in those triads with BA at the center. So, we are only interested with 2/3 of the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

PMBB probability. Similarly, only 1/3 of the probability of two MMAs and one
BA give us BA at the center of the triad sequence (MBM, MMB, BMM). Hence,
converting the wt% composition of 0.39BA/0.61MMA to mole%, we obtain
0.333BA/0.667MMA , the sum of these probabilities is (PBBB + PMBB + PMBM)
= 0.037 + 0.148 + 0.148 = 0.333 = fBA, where fBA = mole fraction of BA in
comonomer
Hence the relative amount of MBM from the Bernoullian distribution is given
by: PMBM/(PBBB + PMBB + PMBM) = PMBM /fBA = (0.148/0.333) = 0.444 = IMBM
compared to 0.45 from integration of the experimental data in Figure 3, where
IMBM is the fractional experimental integral of MBM triad.
This approach does not require any standard and the composition of the soft
component is readily determined by comparing the experimental integral and the
calculated probability. The conversion from experimental integral of the MBM
signal to that of the composition is straight forward using the above relations; i.e.,
PMBM/fBA = the MBM fractional integral in Figure 7.
We can rewrite PMBM/fBA = fBA(1-fBA)2/fBA, using Equation 1 to define PMBM
= fBA(1-fBA)2. Hence, the experimental integral, IMBM, and the probability, PMBM,
are correlated as: PMBM/fBA = fBA(1-fBA)2/fBA = IMBM, where IMBM is the fractional,
experimental integral of MBM, and after simplification, (1-fBA)2 = fMMA2 = IMBM,
since (1-fBA) = fMMA. fMMA = (IMBM), remember that fMMA is mole fraction of
MMA in the BA rich component. This relationship enables us to determine the
mole fraction of MMA directly from the integrals of the experimental spectra such
as Figure 5. For example, for Sample 4, this integral is given by 0.24. Using the
relation, fMMA = (IMBM), fMMA is determined from (0.24) = 0.4899, fBA =(1-
fMMA) = 0.5101. We can thus convert it to weight ratio of 57BA/43MMA. Note
that we obtained 58BA/42MMA for this component with a comparison to a known
standard, Sample 1.
Below is a table of experimental and calculated values of the MBM fractional
amount among the triads. For the heterogeneous polymers, the table also provides
the weight ratios calculated from the experimental data using the equation derived
above.

393
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. Experimental and calculated molar ratio of MBM on model
copolymers and soft phase composition of segregated copolymer
SAMPLE MBM_____ P(MBM)_____ Weight ratio
(wt. ratio) (MBM+MBB+BBB) P(MBM+MBB+BBB) calculated
Experimental Calculated fMMA = (IMBM)
79BA/21MMA 0.08 0.07
59BA/41MMA 0.23 0.23
39BA/61MMA 0.45 0.44
19BA/81MMA 0.71 0.71
Sample 3 0.25 56.2BA/43.8MMA
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Sample 4 0.24 57BA/43MMA

Clearly, we have a very good match of the BA rich composition. In this

case, however, we did it without any need for a standard sample. Matching the
spectral features of the unknown with a known standard that happen to have the
same composition is a daunting task. One should remember that in the presence
of crosslinking and interphase diffusion, the Tg would not be a reliable source for
the use of the Fox equation to obtain the composition of each phase. In the case of
Sample 3 and 4, we do not have any indication of the presence of a heterogeneous
polymer by DSC or DMA measurements.
The use of this approach will require high signal to noise and resolution
spectrum in order to calculate the composition accurately. Otherwise, we will get
erroneous results due to poor integrals of the spectra. We are also assuming that
the contribution of the BA from the hard phase is minimal and we need to have
any of the independent checks we have in this report such solid state NMR and
solution state 1H NMR lineshapes to ascertain the presence of phase segregation.
Whenever possible use of integration from other regions of the spectrum, such as
carbonyl region, will enhance the accuracy of the results in Table 1.
In the preceding discussion we have determined that we do have
heterogeneous composition of the copolymers in Sample 3 and 4 with a BA
rich BA/MMA copolymer and another predominantly MMA composition in the
complex. Since all the compositional calculation was done in solution, we do not
have any indication if these heterogeneous polymers segregate to provide a hard
and soft phase in the polymer film. For that we need to study the films by solid
state NMR and AFM.

Solid State NMR

In Figure 8, the solid state CPMAS NMR spectra of Samples 1, 3 and 4, taken
at ambient temperature are shown. The signal of Sample 1, the homogeneous
58BA/42MMA, is also overlaid (dashed lines) on the other two heterogeneous
polymers for better visual comparison. Note that we have excluded the carbonyl

394
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 carbon signal at about 176 ppm from these spectra in order to expand and display
the more informative section.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Figure 8. Room temperature 13C CPMAS spectra of Samples 1, 3, and 4.

The resonance assignments are shown in the spectrum for the various
moieties of the copolymers. Some of the backbone methylene signals of the
nearest neighbor BA/MMA sequences (BBM, MBM, BMM, BMB) are shifted
downfield and upfield respectively from the homopolymer resonances and are
spread between 40 and 50 ppm resulting in the broad signal close to the baseline.
The intensity of the ester OCH2- signal of BA at 64 ppm is normalized for this
comparison. We readily observe that while the BA signals are similar in intensity
except for peaks 3 and 4 of the butyl group that will be addressed later, the MMA
signals at 51, 45, 15-20 ppm for Samples 3 and 4 are higher than those for sample
1. This is consistent with what we have seen from solution spectra where Samples
3 and 4 have similar BA rich component as Sample 1 and, in addition, they both
have a predominantly poly methyl methacryate component in the complex. As
shown from the plot of Tg as a function of MMA composition of the complex, if
there is phase segregation, the predominantly MMA segment of the complex is
expected to have Tg as high as 100oC. So raising the sample temperature during
CPMAS experiment to about 90oC will increase the mobility of the BA rich
segment of the polymer that has Tg close to 0oC, hence reducing the efficiency of
the cross polarization transfer for the BA rich component while retaining the CP
efficiency of the predominantly MMA component of the polymer.
In Figure 9 a set of solid state NMR spectra are shown for Sample 3 at
90C. The top spectrum is a standard direct polarization (DP) 13C spectrum with
quantitative representation of all species irrespective of their mobility (phase).
The middle spectrum is obtained by CP and the intensity of the backbone signals
395
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 at about 40 ppm is normalized to the DP spectrum in the top plot; the backbone
signals being less mobile even at 90C will have comparable intensities with
both DP and CP experiments. The DP spectrum (top) has signals of the butyl
side chains of BA as the most intense peaks due to their highly mobile domain
as they are almost 100 degrees above their Tg. In contrast, and as expected, we
observe that the methyl group signal of BA is the least intense with the cross
polarization experiment. This is because the -CH3 group in the soft, low Tg,
phase of the copolymer is the most mobile with weak dipolar coupling, hence the
least efficient to cross-polarize.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Figure 9. DP and CP experiments of Sample 4 at 90C. The bottom spectrum is a

vertical expansion of the middle spectrum for better visualization.

The CP signals that are most intense at 90C are those of MMA indicating
a higher CP efficiency which is consistent with a strong dipolar coupling
environment for MMA that one would expect from a hard phase. We acknowledge
that we lost a good portion of the MMA signal with CP compared to DP (top
spectrum) even if we detect mainly MMA signals with CP at 90C. This is
because most of MMA of the total heterogeneous polymer is in the BA rich phase
copolymerized with BA, which constitute the soft phase, hence with poor CP
efficiency.
We do indeed observe weak BA signals after cross polarization at 90C and we
attribute these signals partly due to small BA components in the hard phase MMA
rich composition. More likely, it could be from some soft phase polymer segments
that interdiffuse into the isolated hard phases leading to a restricted motion that in
turn leads to CP signal.
Figure 10 shows the 13C CPMAS experiment taken at about 90C for all three
samples. We notice that the signal to noise is poor in general for all samples. This
is an indication that the sample is predominantly of softer material at 90C.
Even though we have significantly reduced the signals by CP experiment at
90C for Sample 1 (the homogeneously low Tg polymer), we do indeed observe
some residual signals from mainly the backbone and the species close to the
backbone, namely the ester methylene (BA) and -methyl (MMA). There is
396
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 one exception to this observation, the relatively restricted backbone species, the
quaternary carbon signal of MMA at 45 ppm, is barely detected in this experiment.
A quaternary carbon with no directly bonded proton and in a soft domain will
have the least efficiency in cross polarization transfer since it is the farthest from
protons; one of the two necessary conditions for efficient cross-polarization as
discussed in the experimental section above rigidity and proximity to 1H are
missing for the 45 ppm signal of MMA (peak b in the figure) in a soft polymer
matrix (Sample 1).
We still observe some signals from the backbone protonated carbons and those
very close to the backbone (-OCH2- for BA at 64 ppm and OCH3 for MMA
at 51 ppm) since these moieties still have some restricted motions even at 90C
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

and owing to the proximity of 1H rich environment. Unlike DSC and DMA,
where we observe the dynamics for the bulk polymer as a whole, NMR, as a
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

powerful molecular probe, discriminates between the dynamics of the backbone

and sidechain groups. As one would expect, the side chains are more mobile
than the backbone groups hence their signals diminish the most with CP at higher
temperatures.
Sample 3 and 4 also show diminished signals for the BA sidechains similar
to Sample 1 indicating a similar dynamics or copolymer environment and the
presence of a soft phase in these films. On the other hand, some of the MMA
signals at 54, 51, 45, 15-23 ppm remain the most intense in the spectra due to their
rigid environment within the polymer. In other words, we have two segregated
phases in Samples 3 and 4; a SOFT phase, consisting of BA/MMA copolymers
that behaved the same as Sample 1 under CP and a hard phase consisting of a
predominantly MMA component with much higher CP efficiency.

Figure 10. CPMAS spectra of Samples 1, 3, and 4 at 90C.

The MMA signals in Sample 3 are slightly stronger than those of Sample
4, especially the 54 ppm broad peak that is due to the backbone CH2- signal
of MMA. The diminished 54 ppm signal in Sample 4 is an indication that the
hard phase MMA in Sample 4 is probably copolymerized with more BA than the
corresponding hard phase in Sample 3. Even a slightly higher level of BA in the
hard phase will have an effect of shifting some of the backbone CH2- signal of
397
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 MMA in Sample 4 from 54 ppm towards 40-50 ppm. A closer look at about 40
ppm region of the spectra indicates that sample 4 has higher intensity consistent
with having more BA in the hard phase compared to sample 3. Hence, Sample 4
is likely to contain more BA in the hard phase that could lower its Tg more than
that of the hard phase of Sample 3. This could lead to a relatively more miscible
hard and soft phases of Sample 4 compared to Sample 3.
Figure 11 shows the 2 micron images obtained from the cryo face of Sample
4. Unlike the images in Figure 4 and 5, the particle morphology is not well defined
and appears to be well distributed throughout the polymer film. This type of
distribution is indicative of a different type of polymerization. In addition, the
contrast in brightness of the phase appears to differ. There is a gradual change of
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

the brightness of the harder component.

From image analysis of the AFM data, we determined that about 26% of the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

polymer constitutes the bright spots that are attributed to the hard phase of the
polymer. This is in contrast to 11% hard composition from the NMR analysis.
Note that the NMR analysis is based only on the heterogeneous composition of the
polymer and could not account for any interdiffusion that might occur resulting
in different amount of hard/soft domains of the film except some qualitative
indication we saw from CP intensities at 90C.

Figure 11. AFM image of Sample 4, 2um image. Phase image on left; height
image on right.

Such interdiffusion between (meth)acrylic heterogeneous polymers has been

studied earlier (21) where an interface structure showed a gradient (gradual)
composition change from BA rich to MMA rich component. Such interface
structures could bring more BA rich component close to the MMA rich phases
leading to an increase in the rigidity of some of the BA rich (soft phase)
components. In other words, we have more material in the hard phase of the
polymer than the 11% we calculated based on the composition from the analysis
of the solution phase spectra. Such gradient mixed phase may also be responsible
for the lack of DSC and DMA phase transitions as we also expect to have a Tg
398
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 gradient which does not permit a clear phase transition by the two methods. We
note that in some favorable cases, for example involving styrene, DSC, DMA and
TEM could also indicate interphase mixing of latex polymers (22).

Summary
We have employed a combined NMR and AFM methods to probe challenging
phase morphology where we have similar types of monomers in both hard and
soft phases. In addition, the MMA rich hard phase is well dispersed in the soft
BA rich phase with particle sizes below 20 nm. The NMR data provided us with
composition in each phase with solid state NMR experiment providing the existent
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

of a segregated hard and soft phase with the chemical composition information.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

Unlike TEM, AFM does not require any atomic contrast between the two
phases. It is based on the modulus of the various phases, hence ideal for hard
and soft phase characterization. AFM data provided us with the distribution of the
various phases including particle sizes of the dispersed phase.
The synergy between the two methodologies goes beyond just determining
chemical composition in each phase by NMR and morphology by AFM. We were
able to obtain the effect of interdiffusion between the two phases where the amount
of hard and soft phases is determined to be different from that of the synthesis
composition.

References
1. Steward, P. A.; Hearn, J.; Wilkinson, M. C. Adv. Colloid Interface Sci. 2000,
86, 195267.
2. Sperry, P. R.; Snyder, B. S.; ODowd, M. L.; Lesko, P. M Langmuir 1994,
10, 26192628.
3. Blandin, H. P.; David, J. C.; Vergnaud, J. M.; Illien, J. P.; Malizewicz, M. J.
Coat. Technol. 1987, 59 (746), 2732.
4. Eckersley, S. T.; Helmer, B. J. J. Coat. Technol., 69 (864), 97107.
5. Stubbs, J. M.; Sundberg, D. C. J. Coat. Technol. Res. 2008, 5 (2), 169180.
6. Athley R. D., Jr. Emulsion Polymer Technology; Marcel Dekker Inc.: New
York, 1991.
7. Bovey, F. A. High Resolution NMR of Macromolecules; Academic Press:
New York, 1972
8. Koenig, J. L. Chemical Microstructure of Polymer Chains; Wiley: New York,
1980.
9. Kim, Y.; Harwood, H. J. Polymer 2002, 43, 3229323.
10. Candau, F.; Zekhnini, Z.; Heatley, F. Macromolecules 1986, 19 (7),
18951902.
11. Anders, M.; Smulders, W.; Acrdts, A. M.; van Herk, A. M. Macromolecules
1997, 30, 322323.
12. Cheng, H. N.; Tam, S. B.; Kasehagen, L. J. Macromolecules1992, 25, 3779.
Cheng, H. N. J. Appl. Polym. Sci.1992, 51, 21.
13. Cheng, H. N. Macromolecules 1992, 25, 2351.
399
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 14. Cheng, H. N. Macromolecules 1997, 30, 4117.
15. Bujak, P.; Henzel, N.; Matlengiewicz, M. J. Polym. Anal. Charact. 2008, 13
(3), 149162.
16. Brar, A. S.; Kaur, S. J. Polym. Chem. 2005, 43, 11001118.
17. Aerdts, A. M.; German, A. L.; van der Velden, G. P. M. Magn. Resonance
Chem. 1994, 32, S80S88.
18. Beshah, K. Macromol. Chem. 1993, 194, 33113321.
19. Beshah, K. Macromol. Symp. 1994, 86, 3546.
20. Randall, J. C. Polymer Sequence Determination, Carbon13 NMR Method;
Academic Press: New York, 1977.
21. Beshah, K.; Molnar, L. K. Macromoleules 2000, 33 (2), 1036.
Downloaded by OHIO STATE UNIV LIBRARIES on June 21, 2012 | http://pubs.acs.org

22. Colombini, D.; Ljungberg, N.; Hassander, H.; Karlson, O. J. Polymer 2005,
46, 1295.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch023

400
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 24

Application of Melt-State NMR Spectroscopy

for Polyolefin Characterization in Industry
Gerhard Hubner, Isa Fonseca, and Matthew Parkinson*
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

Borealis Polyolefine GmbH, St-Peters-Strasse 25, Linz, Austria

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

*E-mail: [email protected]

Melt-state 13C NMR spectroscopy of polyolefins has been

shown to be sensitive, quantitative and allow rapid sample
throughput. The method was implemented in an industrial
research and development environment and has proved to
be both highly sensitive and versatile. For general problem
solving melt-state MAS NMR was found to complement
standard high-temperature solution state NMR. The main
strength of the technique was found to be the possibility of
high throughput, due to the combination of high sensitivity and
rapid sample preparation. This was most clearly demonstrated
for materials of low solubility. A broad scope of application of
melt-state MAS NMR is demonstrated through the study of the
whole family of polyolefins including HDPE homopolymers,
HDPE copolymers, LLDPE, LDPE, LDPE copolymers,
PP homopolymers and PP copolymers. The application of
the techniques to the study of fractionated materials and
filled materials is also demonstrated. Some key advantages
as compared to standard solution-state NMR spectroscopy
are emphasized and recent research on decoupling in the
molten-state is presented.

Introduction
The macroscopic properties of polyolefins are strongly dependant on
their polymer chain microstructure. Through a variety of modern analytical
techniques a clear understanding of the microstructure of polyethylene (PE)
and polypropylene (PP) may be gained. The application of such analytics has

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 allowed a better understanding of the various polymerization processes and
in-turn facilitates greater synthetic control over polyolefin chain microstructure.
Polymerization of olefins is commonly undertaken using either a low pressure
processes using heterogeneous catalysis (PE and PP) or a high-pressure radical
process (PE).
The physical properties of PEs produced by heterogeneous catalysis in a
low-pressure process can be tuned by the presence of branches of various lengths
in the polymer backbone. Through copolymerization of ethylene with -olefins,
short chain branches (SCB) can be introduced in a controlled way (112). Such
branches form structural defects during crystallization and thus strongly affect
crystallization rates, ultimate crystallinity, melting point, glass transition and
other bulk mechanical properties (13). In contrast, macromonomer incorporation
during polymerization leads to long-chain branches (LCB). With these branches
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

typically being longer than the entanglement molecular weight their presence
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

strongly affects the processability of the bulk polymer (14, 15). The nature of
the process allows for the production of high-density polyethylene (HDPE) and
linear low-density polyethylene (LLDPE). In contrast, if produced by radical
polymerization in a high-pressure process, polyethylene typically has a wider
variety of branch types (16, 17). The possible high content of structural defects
results in low-density polyethylene (LDPE). Although a very large number of
reactions are possible, in practice only a limited number of key structures are
observed in significant concentration, due to differences in stability of reactive
intermediates and associated rearrangement reactions (16). The radical nature of
the high-pressure process also allows the copolymerization of polar commoners,
facilitating for further tuning of physical properties. In general SCB content
and distribution as well as comonomer sequence distribution are the defining
properties of PE microstructure.
The physical properties of PPs produced by heterogeneous catalysis in a
low-pressure process are also strongly dictated by their chain microstructure
(1822). For homopolymers stereo- and regio-regularity play a key role in
crystallization, whereas for propene based copolymers chain defects derived
from comonomer incorporation also play an important role. As such tacticity,
regio-regularity, comonomer content, commoner sequence distribution, total
defect content and various sequence lengths are the defining properties of PP
microstructure.
A variety of analytical techniques are sensitive to polyolefin microstructure,
these include infrared (IR) spectroscopy, differential scanning calorimetry (DSC)
(13, 15), triple-detection gel permeation chromatography (GPC) (21, 23) and
rheology (14, 2432). The analytical technique that has provided the most
direct insight into polyolefin microstructure however is NMR spectroscopy,
and in particular quantitative 13C NMR spectroscopy (26, 27). When applied
to polyolefins this method allows direct access to quantitative information
about, tacticity, comonomer content, SCB distribution and monomer sequence
distribution (17, 19).
The typical implementation of NMR spectroscopy undertakes analysis
of the sample in the solution-state, however when applied to polyolefins a
number of problems arise (26, 33, 34). The most important being low solubility

402
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of polyolefins, even at high temperatures (35, 36) and in solvents such as
trichlorobenzene (TCB) or tetrachloroethane (TCE). The low concentration of 13C
nuclei results in extended measurement times, especially for the quantification
of very low levels of branching. Typically between 6-12 hours are needed for
a single high quality spectrum suitable for further spectral analysis. Despite
these limitations considerable effort is still made to optimize both the sample
preparation and acquisition of quantitative 13C solution-state NMR spectra of
polyolefins.
Recently, the alternative implementation of quantitative 13C melt-state
NMR spectroscopy has been shown to address some of the limitation of the
standard solution-state approach, while still providing high quality results
(3742). Through the combination of high spin concentrations and motional
averaging of line broadening interactions (4348) by the application of common
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

solid-state NMR resolution enhancing methodologies, resolution and sensitivity

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

is considerably increased. In practice the use of magic-angle spinning (MAS)

and high-power dipolar decoupling allow high-resolution spectra to be obtained
from polymers in the bulk molten state. When approximations are made with
respect to key acquisition parameters, further considerable timesaving may also
be made with respect to standard quantitative 13C NMR spectroscopy (3842). In
addition to time saving during acquisition further time saving is also made due
to the removal of the solution-state sample preparation stage. The latter also has
important safety advantages, as toxic solvents no longer need to be handled. In
addition to the investigation of chain branching in polyolefins this implementation
has also been used to investigate sparse branching in other polymer systems,
which have proven problematic for quantitative solution-state 13C NMR analysis
(49).

Experimental Section
Materials
All material presented were selected to demonstrate the wide application
possibilities of melt-state NMR spectroscopy. Materials were produced from a
variety of catalysts and polymerization conditions and range from full plant scale
products, through pilot scale lots to materials produced at bench-scale. As such
all material can be considered to be typical commercial polyolefin grades suitable
for a variety of applications with standard microstructures, molecular weights,
molecular weight distributions and rheological behavior.

Hardware Setup
All melt-state NMR spectra were obtained on a Bruker Avance III 500
solid-state NMR spectrometer with a 11.4 T wide-bore magnet. All spectra were
obtained using a custom Bruker high-temperature (WVT) 7 mm MAS probehead
with the double-resonance coil optimized for 13C detection and compensated
for operation under a pure nitrogen atmosphere. Standard 7 mm zirconia rotors
zirconia rotor caps were also used (39). Sample temperature for all PEs were
403
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 150C, whereas for PP based systems typically 180C was used. Temperature
calibration of the MAS probehead was carried out using Pb(NO3)2 in the standard
manner. Samples were spun between 3 and 5 kHz using a standard MAS controller
using a custom spin-up procedure, to ensure no overlap of MAS spinning side
bands with spectral signals. Shimming was undertaken on the 1H line-shape; and
all quantitative 13C spectra were acquired using standard single-pulse excitation
(SPE) with 1H dipolar decoupling (DD). Excitation was achieved using a 10
us 90 pulse. Due to the long spin-spin relaxation times (T2) encountered for
molten polyolefins low-load dipolar decoupling sequences were employed (38,
39). For decoupling 20 us 180 pulses and associated delays were used. To
ensure optimum time efficiency shorter than true quantitative pulse repetitions
delays were used (< 5 T1) e.g. 3 s for PE copolymers. However, close attention
was paid to ensure signals used during spectral analysis were quantitative with
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

respect to each other (3840). For all experiments a dwell time of 15 us was used
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

for a spectral window of 265 ppm (33.3 kHz). An acquisition time of 571 ms
was used based upon a duty cycle of 8 % which resulted in the digitization of
38056 complex data points. The FID was zero-filled to 64k data points prior to
Fourier-transformation and Lorentz-to-Gauss apodization used.

Spectral Analysis Setup

To ensure repeatable analysis of quantitative 13C NMR spectra a set of
fully automated module spectral analysis programs was implemented. Such an
approach was necessary due to the rapid rate at which spectra were produced.
Such programs assess the presence of key substructures, integrate the spectrum,
calculate primary quantitative results, calculate secondary result from the primary
results, fit the results to models and report all quantitative results with the spectra
as PDF and CSV files.

Results and Discussion

Comparison of Melt-State and Solution-State NMR
When standard solution-state and melt-state 13C NMR spectra of a PE are
compared it is clear that high-resolution has been achieved in the molten-state
(Figure 1). All sites needed for further quantification are resolved and limited
significant differences are seen between the two spectra. Upon closer inspection
of the bulk CH2 signals it can be seen that the line widths in the melt-state spectra
are broader. However, as most spectral analysis methods tend to integrate groups
of signals, quantitative analysis is not hindered. The broader line widths are
mostly due to residual dipolar coupling, which is more problematic to remove
in the molten state due to the long spin-spin relaxation times (3740). With
respect to sensitivity, both spectra have the same signal to noise ratio, however
higher quantification accuracy is achieved for melt-state due to the broader lines
observed resulting in larger integrals. Although 200 mg of PE was used for both
implementations and the same signal-to-noise was achieved the solution-state
spectra required 6.5 h of measuring time whereas only a 1 h measurement was
404
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 needed for the melt-state. When combined with sample preparation the total time
for the solution-state setup was 8 h whereas only 1.5 h was needed to obtain a
melt-state spectrum. In practice a compromise between sensitivity and resolution
is made with less sample preparation and measurement time needed for the
melt-state implementation (39).

Scalar Coupling Mediated Method in the Molten State

With melt-state NMR considered to be closer to solid-state NMR than solution

state NMR, the possibility of applying spectral editing with multidimensional
pulse-sequences to molten polyolefins was assessed. The melt-state techniques
were found to be amenable to such methods with only minor modifications to the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

pulse sequences to account for slow MAS. The spectral editing methods of DEPT
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

and INEPT were both applicable with INEPT found to be more robust in the
molten-state (39). As to why DEPT was more susceptible to pulse mis-adjustment
remains unclear, however the influence of residual dipolar couplings or rotor
synchronization issues is suspected. The use of INEPT has been found to
particularly useful for assignment, especially when comparing to chemical shifts
encountered in the solution-state e.g. when solvent effects exchange the location
of CH and CH2 signals. More advanced multi-dimensional techniques were also
found to be possible such as INADEQUATE which benefited from the enhanced
sensitivity and has proved useful for the conformation of assignments (39).

Advanced Low-Load Dipolar Decoupling

With strong residual heteronuclear dipolar couplings present in the

molten-state high-power dipolar decoupling methods commonly applied in
solid-state NMR need to be employed. Whereas in the solid-state the dipolar
network remains fixed during the rotor period in the molten state the dipolar
network is constantly changing. In practice this means that the advances in
heteronuclear dipolar decoupling commonly used in solid-state NMR based on
phase modulation during the decoupling block e.g. TPPM are not applicable
in the molten state. Thus continuous wave (CW) decoupling still provided the
best performance. The alternative approach of composite pulse based low power
decoupling, common in solution state 13C NMR, has previously been shown to
not be beneficial in the molten-state (39). Although CW decoupling provides the
best performance, hardware constraints based on amplifier duty cycle limits its
application to a few tens of milliseconds. For true solids this is not a problem
as T2 are short and FID truncation is limited. In contrast, for molten samples T2
are long and FID truncation is severe. Such FID truncation strongly influences
the quality and repeatability of the spectra following strong line broadening. To
address this issue low-load decoupling is used based on a sequences pulses and
delays of equal length (38, 39). Recently this has also successfully been applied
in solid-state NMR as Hahn-echo pulse train (HEPT) decoupling (50, 51). Such
HEPT decoupling allows acquisition of the whole molten polyolefin FID and thus
improved quantitative results.
405
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 1. Comparison of solution-state and melt-state 125 MHz 13C NMR spectra
of a polyethylene-co-butene-co-hexene terpolymer.

It should be stated that if CW and HEPT decoupling are applied for the
same time HEPT does not provide better resolution i.e. it does not have higher
decoupling efficiency (38). However, as HEPT can be applied for longer times
then less truncation and thus less artificial line broadening is needed. Further
improvements are made if standard duty cycle rules are broken and HEPT is
applied at 6 or 8 % duty cycle.
Recently, methods for heteronuclear decoupling of polyolefins in the solution-
state have been improved by using alternative super cycles of the standard WALTZ
approach (52). Further improvement has also been gained both in efficiently and
decoupling sideband suppression through the use of bi-level composite pulse based
decoupling (52). In such as approach, decoupling is started in a high power CW
mode, most needed at start of the FID, and then as the signal decays, decoupling
is switched to the lower power WALTZ modulation. With clear advantages seen
in solution-state NMR this approach was modified for melt-state NMR with an
initial CW block < 50 ms followed by the low load HEPT decoupling of > 200 ms.
This so called CW-HEPT approach can also be implemented with the initial CW
part being higher in power than the remaining HEPT part i.e. bi-level CW-HEPT
(Figure 2). For such implementation, the duty cycle was further extended to 9 %
with no decernable issues.

Figure 2. The CW-HEPT (left) and bi-CW-HEPT (right) decoupling schemes.

406
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 When line widths are compared, it is clear that the CW-HEPT approach
provides improved resolution as compared to standard HEPT (Figure 3).
Surprisingly, limited influence was seen when raising the power of the CW block
in the bi-level-CW-HEPT manner. Even more surprisingly, similar decoupling
efficiency was achieved with the CW block having a lower power than the HEPT
block. In practice this means that the bi-CW-HEPT approach with lower initial
CW power, has more promise as it places less strain on the hardware, especially
at 9 % duty cycle. Such decoupling, however, strongly heats the sample and the
temperature instabilities need to be compensated for as a part of the temperature
regulation. As such this decoupling method is probably only applicable for
elevated temperature measurements.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

Applications of Melt-State NMR in Industry

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

In general, the melt-state NMR approach has been found to be applicable

to all polyolefins. It has proved most successful for rapid identification and
microstructure quantification. No major differences have been found between the
setup needed for the measurement of ethylene or propylene based systems. The
de-facto standard spectral analysis method used for polyolefins microstructure
quantification have been found to be easily adapted for use on melt-state
spectra with only minor widening or shifting of some key integral regions.
Importantly, comparable results are achieved when comparing results obtained
by high-temperature solution-state NMR spectroscopy with those obtained by
melt-state NMR spectroscopy. With respect to resonance assignments, the
majority of resonances have been found to be comparable to those observed in
solution-state. However some solvent effects have been observed which means
some caution needs to be exercised before results can be used.

Figure 3. Comparison of resolution achieved for different decoupling schemes

illustrating the enhanced decoupling efficiency of CW-HEPT at high duty cycle.

407
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Polyethylene-co--olefin Microstructure

Standard spectral analysis methods for the quantification of comonomer

sequence distributions at the triad level were found to be equally applicable
to spectra recorded in the melt-state. Only minor modifications to the integral
regions were needed. The simultaneous equations used in these approaches
need to be modified to exclude signals know to be non quantitative (e.g. methyl
groups). Triad comonomer sequence distributions were shown to be comparable
between solution-state and melt-state within an acceptable error (0.2%). More
importantly, the properties derived from the triad distribution, such as sequence
lengths, run numbers and cluster indices showed very high agreement between
the two implementations. Due to the higher sensitivity, quantities derived from
melt-state spectra were also found to be highly reproducible and showed less
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

variance as compared to solution-state results. Polyethylene based terpolymers

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

were also studied and again melt-state NMR provided comparable results to
solution-state NMR in considerably less time.

Difficult to Dissolve Polyolefins

The melt-state techniques is particularly useful when considering materials

that are very difficult to dissolve. One such class of material is HDPE
homopolymers. In such materials considerable sample preparation time is needed
to ensure a homogenous solution for standard solution-state NMR analysis.
Even after dissolution, the extremely low content of branches in such materials
often makes the measurement of quantitative spectra for branch quantification
unfeasibly long. In contrast, such materials are easily studied by melt-state NMR,
with limited reduction of spectral resolution as compared to standard LLDPEs.
In such cases the sensitivity of the melt-state technique allows quantification of
branching produced by alternative catalytic reactivity or feed stock impurity.

Rapid LCB Quantification

With respect to the quantification of long-chain branching, extreme measures

have been applied to achieve the desired sensitivity. In such cases levels of the
order of a few branches per 100,000 backbone carbons have been quantified
using weeks of measuring time at high field (14). In contrast, the sensitivity of
the melt-state techniques allows for the quantification of similar levels of sparse
branching in less than one day. In one such example, 4 branches per 100,000
backbone carbons were quantified in a single 13 h overnight measurement (38).
Such quantification opens up new opportunities for understanding LCB in PE.
Similar sensitivity has been seen for measuring LCB in PP, although this is more
difficult and requires assumptions concerning the structure of the branch site.

408
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Application to Isotactic Polypropylene

With respect to PP, the analysis of isotactic PP (iPP) produced by Zeigler-Natta

catalysis is also readily accomplished by melt-state NMR (53). However, for
polypropylenes resolution is limited to measurement of pentad isotacticity and
triad stereo sequence distribution. Resolution is not sufficient to measure the
pentad stereo sequence distribution or higher orders of tacticity (Figure 4). For
such systems, resolution is very important and solution-state NMR is needed
as a complementary technique. Melt-state NMR, however, does have a role to
play with respect to rapid screening of materials before solution-state analysis or
for the quantification of regio-irregularity. Through the application of fitting to
standard polymerization models similar results are obtained to those calculated
from solution-state spectra of the same material.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Application to Polypropylene-co--olefins

As with ethylene-based copolymers the melt-state techniques was found

to be extremely useful for the analysis of propylene-based copolymers with
-olefins. Comonomer incorporation, commoner sequence analysis, tacticity of
propene sequences and degree of regio-irregularity could all be measured with
melt-state 13C NMR spectra (39). In addition to propylene-based copolymers,
propylene-based terpolymers have also been studied using this technique.

Figure 4. Typical melt-state 125 MHz 13C NMR spectrum of a regio-regular

isotactic polypropylene illustrating resolution necessary for both pentad
isotacticy and triad tacticity distribution quantification.

409
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Application to Filled and Pigmented Materials

Unlike IR spectroscopy or solution-state NMR spectroscopy direct analysis

of filled or pigmented material was found to possible by melt-state NMR, without
the time-consuming step of hot filtration. Such a process not only increases
the throughput of such analyses but has been found to play an important role
in general problem solving. With the high amount of information provided by
a single quantitative 13C NMR spectrum often many answers can be provided
after one quick measurement. Through the use of MAS the anisotropic NMR
interactions are address and averaged to their isotropic values, similarly MAS
also averages any sample anisotropy caused by filler or pigment and thus limits
its influence on the final spectrum. Very good agreement has was seen between
the analysis of original and hot filtered materials (Figure 5).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Application to the Analysis of Fractionated HDPE

A common analysis strategy for comonomer containing PE is preparative

fraction followed by commoner quantification. Commonly, such analysis of PE
fractions by NMR is limited to special cases, due to the time consuming nature
of the NMR analysis. Not only is less material available per fraction but also
the commoner content in some of the fraction may be very low. Although time
consuming, the information provided about comonomer distribution with respect
to molecular weight is highly valuable. The higher sensitivity of the melt-state
NMR approach has reduced the time of such analyses from almost a week to
only 15 h. This has allowed us to offer such analysis to a wider community.
With often only 50 mg or less provided for each fraction, the question was raised
what is the smallest quantity of material from which meaningful quantitative
results can be obtained? To study this, a series of spectra were recorded using
different amount of material from 200 mg down to 5 mg (Figure 6). As expected
sensitivity and signal-to-noise ratio decreased will lower sample amounts,
however, the determined quantities remained relatively constant. Although
the actual determined quantity remained constant as the signal-to-noise ratio
decreased, the error, as judged by the reciprocal of the signal-to-noise, increased
(38, 39, 49). The advantage of NMR spectroscopy, however, is that such a loss in
signal can be compensated for through the acquisition of more transients. Thus
for such systems longer measurement can be undertaken to compensate for the
lower than usual sample quantities. Another major advantage of the melt-state
NMR approach is that the sample is not consumed during analysis. It is thus
possible for melt-state NMR spectroscopy to be the first analytical step for a
material. The rotor can then be unpacked and the sample used for further analysis.
This has key advantages when very small amounts of material are produced and a
series of tests are needed.

410
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Figure 5. Comparison of melt-state 125 MHz 13C NMR spectra of two

polyethylene-co-hexene copolymers containing carbon black and blue pigment
illustrating access to comparable hexene contents without hot filtration.

Figure 6. Comparison of melt-state 125 MHz 13C NMR spectra of a

polyethylene-co-hexene. Illustrating constant hexene content with decreasing
signal-to-noise ratios of the CH branch site and associated quantification error.

411
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 7. Typical melt-state 500 MHz 1H NMR spectrum of a
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

polyethylene-co-vinylacetate copolymer.
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Quantitative 1H Melt-State NMR Spectroscopy

As expected 1H melt-state NMR spectra show much lower resolution, as

compared to solution-state, due to the inherent strength of the 1H-1H homonuclear
dipolar interactions. However, unlike solid-state 1H NMR of polyolefins, where
very limited microstructure information can be gained, for some copolymers the
combination of motional averaging and MAS provided sufficient resolution for
basic microstructure quantification. As such for standard HDPEs, LLDPEs and
PP 1H melt-state NMR spectroscopy is of limited use except for 13C NMR setup.
However, for heteroatom containing copolymers, such as ethylene-methylacrylate
(EMA) or ethylene-vinylacetate (EVA), the presence of polar functionality
provides sufficient chemical shift dispersion to allow simple quantification
(Figure 7). Such measurements have been found to be highly reproducible and
experimentally robust due to the shorter 1H spin-lattice relaxation times. The
only major factor that need be considered is if the MAS spinning-sidebands for
all peaks are used in the quantitative analysis.

Conclusion
Quantitative melt-state 13C NMR was successfully integrated into an
industrial research and development laboratory environment. The technique was
found to be applicable for the analysis of a wide range of industrial polyolefins
and their copolymers. Importantly, quantitative results were found to be highly
comparable to those obtained by standard high-temperature solution-state NMR.
Results obtained by melt-state NMR were also found to be considerably more
reproducible due to the high sensitivity of the technique.
Although the melt-state NMR technique does addresses some of the main
issues encountered with the more standard solution-state NMR approach it does
not completely replace the need for access to solution-state NMR. Although
sensitivity is important, some problems will always need to be addressed with the
higher resolution obtained using solution-state NMR. As such melt-state NMR
412
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 does not replace the need for solution-state NMR, however the two experimental
methods strongly complement one another when both are used to emphasize their
particular strengths.
The main advantage found for melt-state NMR has been for rapid
microstructure screening and for the analysis of difficult to dissolve polyolefins
(e.g. highly-isotactic PP, HDPEs with very-low comonomer contents, preparative
fractions of low comonomer content systems and filled systems).
In general, the implementation of the melt-state NMR approach has allowed
much wider access to the wealth of knowledge provided by quantitative 13C NMR
spectroscopy of polyolefins. Consequently, quantitative 13C NMR results may now
be offered at a much earlier stage of problem solving or general research activities.
By the implementation of such an approach we have gained better insight into
polyolefins and facilitated new research and development opportunities.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

Acknowledgments
The authors of the paper would like to thank the numerous researchers who
contributed samples or problems presented in this work. The authors would also
like to acknowledge the continued support of IOS management, and specifically
Jens Reussner, concerning the development and implementation of advanced
polyolefin characterization methods within Borealis.

References
1. Brintzinger, H. H.; Fischer, D.; Mlhaupt, R.; Rieger, B.; Waymouth, R.
Angew. Chem., Int. Ed. Engl. 1995, 34, 1143.
2. Kokko, E.; Malmberg, A.; Lehmus, P.; Lfgren, B.; Seppl, J. J. Polym.
Sci., Part A: Polym. Chem. 2000, 38, 376.
3. Wang, W.; Zhu, S.; Park, S. Macromolecules 2000, 33, 5770.
4. Kaminsky, W. Adv. Catal. 2001, 46, 89.
5. Kokko, E.; Lehmus, P.; Malmberg, A.; Lfgren, B.; Seppl, J. Long-Chain
Branched Polyethene via Metallocene-Catalysis: Comparison of Catalysts.
In Organometallic Catalyst and Olefin Polymerization, Catalysts for a New
Millennium; Springer: New York, 2001; pp 335345.
6. Galland, G. B.; Quijada, R.; Rojas, R.; Bazan, G.; Komon, Z. J. A.
Macromolecules 2002, 35, 339.
7. Capacchione, C.; Proto, A.; Okuda, J. J. Polym. Sci., Part A: Polym. Chem.
2004, 42, 2815.
8. Furlan, L. G.; Kunrath, F. A.; Mauler, R. S.; de Souza, R. F.; Casagrande, O.
L., Jr J. Mol. Catal. A: Chem. 2004, 214, 207.
9. Izzo, L.; Caporaso, L.; Senatore, G.; Oliva, L. Macromolecules 1999, 32,
6913.
10. Lehmus, P.; Kokko, E.; Leino, R.; Luttikhedde, H. J. G.; Rieger, B.;
Seppl, J. V. Macromolecules 2000, 33, 8534.
11. Prosenc, M. H.; Brintzinger, H. H. Organometallics 1997, 16, 3889. 29.
413
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 12. Thorshaug, K.; Stovneng, J. A.; Rytter, E.; Ystenes, M. Macromolecules
1998, 31, 7149.
13. Zhang, M.; Wanke, S. E. Polym. Eng. Sci. 2003, 43, 1878.
14. Wood-Adams, P. M.; Dealy, J. M.; deGroot, A. W.; Redwine, O. D.
Macromolecules 2000, 33, 7489.
15. Shroff, R. N.; Mavridis, H. Macromolecules 2001, 34, 7362.
16. Axelson, D. E.; Levy, G. C.; Mandelkern, L. Macromolecules 1979, 12, 41.
17. Randall, J. C. J. Macromol. Sci., Rev. Macromol. Chem. Phys. 1989, C29,
201.
18. Resconi, L.; Cavallo, L.; Fait, A.; Piemontesi, F. Chem. Rev. 2000, 100,
1253.
19. Busico, V.; Cipullo, R. Prog. Polym. Sci. 2001, 26, 443.
20. Busico, V.; Cipullo, R.; Segre, A. L. Macromol. Chem. Phys. 2002, 203,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

1403.
Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

21. Bovey, F. A.; Schilling, F. C.; McCrackin, F. L.; Wagner, H. L.

Macromolecules 1976, 9, 76.
22. Vestberg, T.; Denifl, P.; Parkinson, M.; Wilen, C-E. J. Polym. Sci., Part A:
Polym. Chem 2010, 48, 351.
23. Bugada, D. C.; Rudin, A. Eur. Polym. J. 1987, 23, 847.
24. Striegel, A. M.; Krejsa, M. R. J. Polym. Sci., Part B: Polym. Phys. 2000,
38, 3120.
25. Vega, J. F.; Santamaria, A.; Munoz-Escalona, A.; Lafuente, P.
Macromolecules 1998, 31, 3639.
26. Shroff, R. N.; Mavridis, H. Macromolecules 1999, 32, 8454.
27. Wood-Adams, P. M. J. Rheol. 2001, 45, 203.
28. Lohse, D. J.; Milner, S. T.; Fetters, L. J.; Xenidou, M.; Hadjichristidis, N.;
Mendelson, R. A.; Garcia-Franco, C. A.; Lyon, M. K. Macromolecules 2002,
35, 3066.
29. Gabriel, C.; Mnstedt, H. Rheol. Acta 2002, 41, 232.
30. Gabriel, C.; Kaschta, J.; Mnstedt, H. Rheol. Acta 1998, 37, 7.
31. Crosby, B. J.; Mangnus, M.; de Groot, W.; Daniels, R.; McLeish, T. C. B. J.
Rheol. 2002, 46, 401. 30.
32. Trinkle, S.; Walter, P.; Friedrich, C. Rheol. Acta 2002, 41, 103.
33. Bovey, F. A.; Mirau, P. A. NMR of Polymers; Academic Press: San Diego,
1996.
34. Morris, G. A. Magn. Reson. Chem. 1986, 24, 371.
35. Liu, W.; Ray, D. G.; Rinaldi, P. Macromolecules 1999, 32, 3817.
36. Seger, M. R.; Maciel, G. E. Anal. Chem. 2004, 76, 5734.
37. Pollard, M.; Klimke, K.; Graf, R.; Spiess, H. W.; Wilhelm, M.; Sperber, O.;
Piel, C.; Kaminsky, W. Macromolecules 2004, 37, 813.
38. Klimke, K.; Parkinson, M.; Piel, C.; Kaminsky, W.; Spiess, H. W.;
Wilhelm, M. Macromol. Chem. Phys. 2006, 207, 382395.
39. Klimke, K. Optimisation of Polyolefin Branch Quantification by 13C
NMR Spectroscopy. Ph.D. Thesis, Universitt Mainz, Germany 2006,
http://ubm.opus.hbz-nrw.de/volltexte/2006/1077/pdf/diss.pdf.
40. Parkinson, M.; Klimke, K.; Spiess, H. W.; Wilhelm, M. Macromol. Chem.
Phys. 2007, 208, 2128.

414
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 41. Stadler, F. J.; Piel, C.; Klimke, K.; Kaschta, J.; Parkinson, M.; Wilhelm, M.;
Kaminsky, W.; Mnstedt, H. Macromolecules 2006, 39, 1474.
42. Vittorias, I.; Parkinson, M.; Klimke, K.; Debbaut, B.; Wilhelm, M. Rheol.
Acta 2007, 46, 321.
43. Dechter, J.; Komoroski, R.; Axelson, D.; Mandelkern, L. J. Polym. Sci., Part
B: Polym. Phys. 1981, 19, 631.
44. Hatfield, G.; Killinger, W.; Zeigler, R. Anal. Chem. 1995, 67, 3082.
45. Litvinov, V. M.; Mathot, V. B. F. Solid State Nucl. Magn. Reson. 2002, 22,
218.
46. Thakur, K. High-Temperature Solid-State NMR of Cross-Linked, Insoluble,
and Unswellable Polymers. In NMR Spectroscopy of Polymers in Solution
and in the Solid State; ACS Symposium Book Series 834; American
Chemical Society: Washington, DC, 2003; p 32.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch024

47. Zeigler, R. C. Macromol. Symp. 1994, 86, 213.

Downloaded by CORNELL UNIV on June 20, 2012 | http://pubs.acs.org

48. Guo, M.; Cheng, S. Z. D.; Quirk, R. P. Variable temperature, solid-state 13C
NMR study of linear low-density polyethylene. In Olefin Polymerization,
Emerging Frontiers; Symposium Book Series 749; American Chemical
Society: Washington, DC, 2000, p 163.
49. Castignolles, P.; Graf, R.; Parkinson, M.; Wilhelm, M.; Gaborieau, M.
Polymer 2009, 50, 2373.
50. Filip, X.; et al. J. Magn. Reson. 2005, 176, 239.
51. Griffin, J.; et al. Magn. Reson. Chem. 2007, 45, S1S198.
52. Zhou, Z.; Kmmerle, R.; Qiu, X.; Redwine, D.; Cong, R.; Taha, A.;
Baugh, D.; Winniford, B. J. Magn. Reson. 2007, 187, 225.
53. Hu, W.; et al. Macromolecules 2007, 40 (9), 3505.

415
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 25

NMR Rescaling Revisited

M. E. Ries,* A. Bansal, and M. G. Brereton
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

School of Physics and Astronomy, University of Leeds,

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

Leeds, LS2 9JT, United Kingdom

*E-mail: [email protected]

The purpose of this work is to examine the approximation

known as NMR rescaling, in which a polymer chain is replaced
by a series of NMR submolecules. This coarse graining of the
atomistic details makes various NMR calculations analytically
possible. To test the accuracy of this approach a computer
simulation of network chains consisting of freely jointed
rigid rods has been carried out. The time dependence of
the orientation of these rods relative to their network chain
end-to-end vector has been recorded. From this the NMR Hahn
spin echo sampled transverse relaxation and double quantum
build-up and subsequent decay have been calculated. The
simulated NMR signals are then compared with theoretical
results derived from a rescaling approach. In this work we
also test the Anderson-Weiss second moment approximation by
using it to derive analytical expressions for the NMR signals
for the simulated chains, which can then be directly compared
with the simulated results. From this work it is shown that
for a polymer melt above its glass transition temperature both
rescaling and the second moment approximations are valid.

Introduction
It is well-known that pulsed nuclear magnetic resonance (NMR) is a useful
tool for the study of dynamics and orientation in polymer melts and networks
(111). For the NMR measurements to be meaningful though there needs to
be a theoretical link connecting the NMR signal to the microscopic properties
of interest, such as entanglement / crosslink density, chain persistence length,

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 dynamic correlation times and order parameters. To make these connections
quantitative has proved very challenging.
The classical starting point for the quantitative analysis of stress optical
anisotropy in strained polymer networks is the Kuhn and Grn theory of 1942
(12). The true molecular network is replaced by an idealized network of identical
chains each of which consists of n freely jointed rods of length l. The motivation
for using a replacement chain comes from the fact that it is generally not possible
to derive the material properties if all the atomistic details are included. For
example, a representative or coarse grained chain such as the Rouse bead and
spring model (13) enables the rheological properties of a polymer melt to be
theoretically calculated.
In the 1970s and 80s J. P. Cohen-Addad introduced and pioneered the idea
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

of the NMR submolecule (1419); a coarse grained unit comprising of several

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

monomers. In this approach the real polymer chains are replaced by a series of
these NMR submolecules connected together. For the Kuhn and Grn model the
parameters (n, l) of the replacement chain are chosen such that both the maximum
end-to-end length and root mean square end-to-end length of the rigid rod chain
have the same values as that of the real chain it is replacing (20). For the case of
the NMR submolecule it must generate an effective P2(cos ) for the part of the
polymer chain it represents, where P2() is the second-order Legendre polynomial
and is the angle between an inter nuclear vector and the applied static magnetic
field. The NMR properties of the polymer chain are then rescaled onto the NMR
submolecule.
Motional averaging of the NMR properties for a polymer chain is considered
to take place in two steps (21). The first involves local, monomer length scale,
fast reorientations. These fluctuations take place within the NMR submolecule
and partially average the NMR properties such that they become dependent on the
end-to-end vector coordinates of their containing submolecule. The second step
is the slower reorientation of the submolecule itself. This submolecule can be, for
example, a Rouse unit, an entanglement vector or a network chain. The key point
being that the submolecule coordinates can be handled analytically, which enables
the calculation of the NMR signal in terms of these variables. The pre-averaging
of the NMR properties so that they depend on the NMR submolecule coordinates
is termed rescaling (20).
The motivation for this piece of work is that, to the best of our knowledge,
this approach has never before been verified. It has of course been used in many
publications, where it has been assumed to be true, but the rescaling approximation
itself has not been directly tested. Rescaling is just an approximation used to make
the NMR problem analytically solvable, but its accuracy and limits of applicability
are not known. For experiments such as the transverse relaxation, as sampled by
a spin-echo experiment, or the double quantum build-up and decay signal to be
quantitatively interpreted in terms of polymeric material microscopic properties,
rescaling must be valid. Otherwise the connection between these properties and
the NMR response becomes broken. It is the purpose of this work to determine if
and when rescaling is valid.

418
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Simulation Procedure
Freely Jointed Rigid Rod Chain

To test rescaling we simulated a pseudo-atomic chain. For this we used the

freely jointed rigid rod chain where each rod was taken to have unit length. The
rods are connected together to form a polymer chain via hinges that allow complete
free rotation, so that consecutive rods can freely take up any orientation with
respect to each other. In Figure 1 a freely jointed chain is shown forming an
end-to-end vector, or network vector, R with one of the rigid rods making an angle
with the external reference direction B.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

Figure 1. Freely jointed chain showing the end-to-end vector R, reference

direction B and the instantaneous angle that a rod makes with the reference
direction. The dashed ellipse and straight line are explained in the text.

A C++ program was written that simulates a chain of n freely jointed rigid
rods. Initially the rods were connected together so that they formed a pre-chosen
network vector R. In each time step a hinge i was chosen, see Figure 1, and the
line that links the previous (i-1) and next hinge (i+1) was determined from the
coordinates of the hinges, which is the dashed line in Figure 1. Then a circle, seen
in perspective in Figure 1 as a dashed ellipse, was determined such that the plane
of the circle is at right angles to the dashed line, the centre of the circle is where
this plane and line meet and the radius of the circle is such that the ith hinge is on
this circle. This circle indicates all the possible positions available to hinge i that
both keeps hinges i-1 and i+1 fixed and all the rod lengths unchanged. A random
location on this circle was chosen as the new position of hinge i. In each time
step each hinge, except for the first and last, were moved in this manner, such that
the chain adopted a new configuration without altering the set up network vector
R. After each time step the orientation for each rod relative to the reference
direction B was recorded. In all this work we have kept B parallel to R. The chain
was always allowed to undergo 1000 time steps to equilibrate before recording
the orientations of the rods. How this data were then used to calculate the NMR
response for our simulated chains will be described in the sections below.
419
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 To confirm that our simulated chain was behaving as a freely jointed rigid rod
chain we determined the time averaged orientation <P2(cos )> for each rod as a
function of network vector length R. Each rod in the chain was found to have the
same time averaged value, as expected, and followed the classical Treloar results
(22). This states that

where Rmax is the fully extended length of the chain, which for our n rods of
unit length is n. Figure 2 shows the simulated values of <P2(cos )> versus R/Rmax,
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

where the solid line is a no parameter fit given by eq 1. In this simulation n was
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

10.

Figure 2. The time average orientation <P2(cos )> of the rigid rods relative to
their end-to-end vector as a function of the ratio of the chains end-to-end length
R to the maximum end-to-end length Rmax. The solid line is given by eq 1.

Background Theory
NMR Signals
The experimental pulse sequences used to measure the Hahn echo sampled
transverse relaxation function (23, 24) and the double quantum (DQ) build-up and
decay curve (25, 26) can be found in the literature. Both these techniques can be
thought of as measuring an intensity I as a function of experimental time t. The
time evolution of these signals depends on the time dependent relative orientation
of an inter-nuclear vector relative to the external magnetic field. In this work we
will consider chains comprising n=11 rigid rods and the central one will be taken
to be this inter-nuclear vector giving rise to the NMR signal.
The Hahn echo sampled transverse relaxation function IH is given by (27)
420
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and the DQ signal IDQ by (27)

where
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

and D is an NMR interaction constant that in this work was set to 0.01 and
the solid lines in eq 2 and eq 3 indicates an averaging over all the chains in the
system. The size of D determines the timescale for the NMR signal and a value of
0.01 was chosen simply to give NMR signals that lasted many time steps (~1000)
in our simulation. In this work the simulations are repeated 1000 times, meaning
that effectively there are 1000 chains to average the NMR signal over. Finally, the
integrals in the above equations are actually discrete summations over each time
step when used to calculate the NMR signals from the simulation data.

Results and Discussions

NMR Rescaling and the NMR Submolecule Approach

It is the aim of this work to test the NMR submolecule approximation

approach, known as rescaling. To do this we need to compare the theoretical
results from applying the method of NMR rescaling with the simulation results
calculated from the simulation data. In the following work the NMR submolecule
is taken to be the network vector R. Next the rescaled NMR signal for the
transverse relaxation function is calculated after which the same is repeated for
the DQ build-up and decay function.

Transverse Relaxation Function

The rigid rod reorientations that take place within the submolecule R result
in an average orientation of these rods and therefore our inter-nuclear vector,
which in this simulation is the central rigid rod in each submolecule. Rescaling
means that just this time averaged value is needed to calculate the NMR signal.
The atomistic submolecule details, such as their fluctuations about this average
orientation are not used. In this way the NMR response becomes dependent
only on the submolecule coordinates, as it is these that determine this average
orientation, recall eq 1. To find an analytic expression for eq 2 we first replace the
integral in eq 4 with the time averaged value. This result is then substituted into
eq 2. In this simulation every chain has the same network vector and therefore all
421
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the rigid rods have the same average orientation. This makes the averaging over
all chains implied by the solid line in eq 2 trivial, giving

In Figure 3 the simulated transverse relaxation function for network chains

with R/Rmax = 0.6 is shown.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

Figure 3. The simulated transverse relaxation function eq 2 for network chains

with R/Rmax = 0.6. The circles are simulation data points and the solid line is
eq 15.
The NMR rescaled result eq 5 predicts an oscillation with a time period
that is dependent on the average orientation of the rigid rods within the NMR
submolecule. In Figure 3 this oscillation, with the correct time period (within
5%), is seen but of key importance is that the signal also decays away. This is
not predicted from the rescaling approximation and therefore rescaling for our
simulation is not sufficiently accurate.
To understand the origin of the decaying part of the signal seen in Figure 3 it
is necessary to look more closely at the time dependence of the orientation of the
rigid rods. To do this we write the instantaneous orientation of a rigid rod in terms
of its time average orientation, as

where all the time dependence is contained within the (t) term that represents
the instantaneous difference between the orientation of the rigid rod and the time
averaged value. We now substitute eq 6 into eq 4 and eq 2 to obtain
422
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The cosine term can be written as the real part of a complex exponential
and recalling that all our network chains have the same time average orientation
enables this to be more clearly written as
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

Eq 8 is the product of two terms, the first is the rescaling result eq 5 and the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

second term is an extra relaxation term that is the origin of the decay seen in Figure
3. The time dependent departures from the average orientation causes the NMR
response from each network chain to go out of phase with each other, resulting in
a signal that decays away.
The first term in eq 8 represents the solid-like contribution to the signal that
gives rise to a coherent, or reversible time evolution, which can be refocused
by appropriate pulse sequences (28). The second term in eq 8 is related to the
fluctuating part or liquid like contribution and corresponds to a homogenous
broadening of the signal (28). Saalwchter et al have pioneered the use of
normalised experimental DQ build up data, which removes the influence of
relaxation processes on the double quantum intensity (26). This allows the
determination of absolute values for the residual dipolar couplings (8).
Next we use the Anderson-Weiss second moment approximation (29) to
quantify the extra relaxation caused by the second term in eq 8. The extra
relaxation is given by

The second moment approximation uses a result that is exact for a variable X
when it has Gaussian statistics,

A comparison of eq 9 with eq 10 gives

From the definition of (t) in eq 6 the average value of X is zero. Therefore

substituting eq 11 into eq 10 gives
423
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 To calculate eq 12 we need to assume a form for the correlations of the
fluctuations of the orientation about their average value. For simplicity we take
this to have the following form
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

where is the correlation time of these fluctuations and represents the

size of these fluctuations.
It should be noted that polymer chains display a spectrum of relaxation times,
such as Rouse modes (13). Our chains consist of only 11 freely jointed rigid rods
and for such short chains and our calculations eq 13 is found to adequately describe
the auto-correlation function of the fluctuations . For an interesting discussion on
the relationship between a polymers spectrum of relaxation times and its NMR
relaxation times the reader is pointed to the work by Gubaidullin et al (30).
The integrals in eq 12 can now be evaluated and in the limit where the
experimental time is longer than the fluctuation correlation time we obtain

Substituting this result into eq 8 gives

The final result is the original rescaling result multiplied by an exponential

decay envelope. We can now turn to our simulation data to obtain the and .
It follows directly from eq 6 and 13 that

This correlation between the orientation at one moment with the orientation
at a later time can be directly calculated from the simulation data. In Figure 4
is shown as a function of time with a fit to eq 16.

424
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

Figure 4. Correlation function of the orientation at one moment with the

orientation at a later time as a function of time. The circles are simulation data
points and the solid line is eq 16.

In Figure 4 the data has been fitted to eq 16 giving a correlation time of 27

time steps, size of fluctuations of 0.18 and an average orientation P2(cos )
equal to 0.21. This is all the information that is needed to predict the NMR signal
in Figure 3 using eq 15. From eq 15 we determine an oscillation with period
3050 time steps and exponential decay envelope with a 1/e time of 2100 time
steps. A free fit of eq 15 to the data in Figure 3 shown as the solid line gives
an oscillation period of 3200 time steps and a 1/e time for the single exponential
relaxation of 2000 time steps. The agreement is therefore excellent, showing that
the second moment approximation is working for this system. Next we develop
similar arguments for the DQ build-up and decay signal.

Double Quantum Build-up and Decay function

The starting point for the DQ build-up and decay function is eq 3. Again
we make the rescaling assumption and replace the integrals in eq 3 with the time
average values and recalling that our network vectors are fixed, we obtain

As before, an oscillating signal with a period that depends on the average

orientation of the rigid rods within the NMR submolecule or network vector is
predicted. In Figure 5 is the DQ signal calculated from the simulation data for
R/Rmax = 0.5.
425
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

Figure 5. Simulated double quantum build-up and decay function eq 3 for

network chains with R/Rmax = 0.5. The circles are the simulation data points and
the solid line is eq 19.

In Figure 5 the oscillation predicted by eq 17 is seen with the correct time

period (within 5%) but similarly to the previously calculated transverse relaxation
function, recall Figure 3, the signal decays away. To understand the origin of this
decaying part we again look at the effect of the time dependence of the orientation
of the rigid rods using eq 6. To carry out this calculation we first rewrite eq 3 using
the mathematical identity sinA.sinB = (cos(A-B) cos (A+B)), to give

Now that the problem is written in terms of cosine functions the derivation
for each term in eq 18 follows exactly the same steps as those for the transverse
relaxation calculation, see eq 7 to eq 15. This gives

which has the form of an oscillation multiplied by a single exponential.

Analyzing data from the simulation as was done before in Figure 4 using eq
16 we can generate a no free parameter fit to the DQ signal. This is the solid
line in Figure 5. The DQ simulation data in Figure 5 has an oscillation period
and decay time both within 5% of the predictions from eq 19. Therefore for
both the transverse relaxation function and DQ build-up and decay signal the
Anderson-Weiss approximation, eq 10, is generating excellent agreement.

426
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Applicability of Rescaling
In the above section it was shown that rescaling for our system of network
vectors, which were our NMR submolecules, does not quite work. The time
dependent fluctuations of the rigid rods about their average orientation causes
an extra relaxation in the generated NMR signal. For rescaling to work only
the average orientation of the rigid rods should be needed to calculate the NMR
signal. Here we find that this approach does not obtain the correct answer to the
integrals in equations 2 and 3. Each NMR submolecule due to its own particular
time dependent path for the rigid rod orientations has a different result for the phase
angle given by eq 4. This loss of phase coherence between the NMR submolecules
is the origin of the decays seen in Figures 3 and 5. Through application of the
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

Anderson-Weiss second moment approximation (29), given by eq 10, it is possible

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

to quantify this extra relaxation. This result connects the atomistic details, such as
the rigid rod orientation correlation time to the NMR response, recall equations
15 and 19. It is therefore now possible to estimate for real polymer melts when
rescaling will be applicable. To do this we will consider the transverse relaxation
signal given by eq 15.
For rescaling to work such that we can replace the real atomistic details with
NMR submolecules and re-write the NMR problem more simply in terms of pre-
averaged orientations eq 5 needs to hold. For this result to be a good approximation
we need the time period of the oscillations in the first term of eq 15 to be much
shorter than the relaxation time of the second term in eq 15, i.e.

If this holds in a real system, one that contains numerous NMR submolecules
each with a different orientation, the NMR signal would relax principally due to
each NMR submolecule having a different oscillation frequency. These different
frequencies would result in the NMR submolecules going out of phase with each
other. The resultant NMR signal would therefore relax and of key importance this
relaxation would not be due to the second term in eq 15. Simply put, when eq 20
holds then eq 15 becomes eq 5 and rescaling is valid. It would then be possible to
correctly calculate the NMR signal using only the rescaling result represented by
eq 5.
For polymer chains in a polymer melt <P2(cos )> ~1/n, for proton and
deuterium NMR D ~ 100 kHz, atomic bond reorientations take place on timescales

shorter than ~ 10-10s and ~1/5, which can be seen from its definition in eq
16 and remembering that the value of averaged over all orientations
is 1/5. Putting these values into eq 20 gives that n << 2000 for rescaling to be
valid. Therefore, as long as the NMR submolecule is not too large in terms of the
number of monomers it replaces, so that it produces a significantly large average
orientation on its constituent atomic units, then the NMR submolecule description
is valid.
427
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Conclusions
A simulation of network chains comprising of freely jointed rigid rods with
fixed end-to-end vectors has been carried out. By monitoring the orientation of
the central rod and taking this to be the NMR relevant inter nuclear vector the
transverse relaxation function and double quantum build-up and decay have been
explicitly calculated. These simulated NMR signals were then compared with
analytical predictions made using the approximation known as rescaling. For
these simulated chains an extra relaxation was observed arising from the time
dependent fluctuations of the orientation of the rods about their average value. To
calculate theoretically this extra relaxation the Anderson-Weiss second moment
approximation (29) was employed. Excellent agreement was found between these
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

theoretical predictions and the simulated data, see the solid lines in Figures 3 and
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

5. Using these results it was then possible to calculate if rescaling in polymer melts
well above their glass transition temperature was valid. It was determined that, as
long as the NMR submolecules used to coarse grain the atomistic details of the
real polymer chains are not too large (n<<2000), rescaling is indeed valid.

Acknowledgments
Many years ago Michael Rubinstein first brought this potential problem with
rescaling to our attention and showed MER the method used in this work of
simulating freely jointed rigid rods. Several undergraduate students have been
involved in different parts of this work. They are Ian Waterman, Rachel Moody,
Adam Royles and Robert Ramsay.

References
1. Ball, R. C.; Callaghan, P. T.; Samulski, E. T. J. Chem. Phys. 1997, 106 (17),
73527361.
2. Brereton, M. G. Macromolecules 1989, 22 (9), 36673674.
3. Brereton, M. G.; Ward, I. M.; Boden, N.; Wright, P. Macromolecules 1991,
24 (8), 20682074.
4. Deloche, B.; Dubault, A.; Durand, D. J. Polym. Sci., Part B: Polym. Phys.
1992, 30 (12), 14191421.
5. Kimmich, R.; Kopf, M.; Callaghan, P. J. Polym. Sci., Part B: Polym. Phys.
1991, 29 (8), 10251030.
6. Klein, P. G.; Ries, M. E. Progr. Nucl. Magn. Reson. Spectrosc. 2003, 42
(1-2), 3152.
7. Saalwachter, K.; Herrero, B.; Lopez-Manchado, M. A. Macromolecules
2005, 38 (23), 96509660.
8. Saalwachter, K.; Heuer, A. Macromolecules 2006, 39 (9), 32913303.
9. Saalwachter, K.; Kleinschmidt, F.; Sommer, J. U. Macromolecules 2004, 37
(23), 85568568.
10. Saalwachter, K.; Sommer, J. U. Macromol. Rapid Commun. 2007, 28 (14),
14551465.
428
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 11. Sotta, P.; Deloche, B. Makromol. Chem., Macromol. Symp. 1989, 23 (Eur.
Symp. Polym. Mater., 1987, Pt. 2), 18390.
12. Kuhn, W.; Grn, F. Kolloid-Z 1942, 101, 248.
13. Rouse, P. E. J. Chem. Phys. 1953, 21, 1272.
14. Cohen Addad, J. P. J. Chem. Phys. 1974, 60 (6), 24402453.
15. Cohen Addad, J. P. J. Chem. Phys. 1976, 64 (8), 34383445.
16. Cohen Addad, J. P. J. Chem. Phys. 1979, 71 (9), 36893696.
17. Cohen Addad, J. P. Polymer 1983, 24 (9), 11281138.
18. Cohen Addad, J. P.; Dupeyre, R. Polymer 1983, 24 (4), 400408.
19. Cohen Addad, J. P.; Feio, G. J. Polym. Sci., Part B: Polym. Phys. 1984, 22
(6), 957978.
20. Ries, M. E.; Brereton, M. G.; Ward, I. M.; Cail, J. I.; Stepto, R. F. T.
Downloaded by UNIV OF GUELPH LIBRARY on June 21, 2012 | http://pubs.acs.org

Macromolecules 2002, 35 (14), 56655669.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch025

21. Cohen Addad, J. P.; Guillermo, A. J. Polym. Sci., Part B: Polym. Phys. 1984,
22 (6), 931955.
22. Treloar, L. R. G. Trans. Faraday Soc. 1954, 50, 881.
23. Carr, H. Y.; Purcell, E. M. Phys. Rev. 1954, 94 (3), 630638.
24. Levitt, M. H.; Freeman, R. J. Magn. Reson. 1981, 43, 65.
25. Baum, J.; Pines, A. J. Am. Chem. Soc. 1986, 108 (24), 74477454.
26. Saalwachter, K. Progr. Nucl. Magn. Reson. Spectrosc. 2007, 51 (1), 135.
27. Ries, M. E.; Brereton, M. G. Phys. Chem. Chem. Phys .2009, 11 (32),
69186924.
28. Sotta, P.; Fulber, C.; Demco, D. E.; Blumich, B.; Spiess, H. W.
Macromolecules 1996, 29 (19), 62226230.
29. Anderson, P. W.; Weiss, P. R. Rev. Mod. Phys. 1953, 25 (1), 269276.
30. Gubaidullin, A.; Shakirov, T.; Fatkullin, N.; Kimmich, R. Solid State Nucl.
Magn. Reson. 2009, 35 (3), 14751.

429
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 26

An NMR Investigation of a Polymer Melt under

Shear
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org

Ute Bhme and Ulrich Scheler*

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

Leibniz-Institut fr Polymerforschung Dresden e.V.,

Hohe Str. 6, 01069 Dresden, Germany
*E-mail: [email protected]

Mechanical shear is part of processing steps of polymers in the

melt. Shear influences polymer dynamics and chain order in the
melt. The effect of mechanical shear on the NMR relaxation
behaviour of polymer melts has been studied. A dedicated
probehead capable for high-temperature rheo NMR has been
developed and applied. Both the transverse and the longitudinal
relaxation time exhibit a strong temperature dependence.
The prolongation of the longitudinal relaxation time T1 with
increasing temperature is attributed to slowed down spin
diffusion as a function of the enhanced motion. The enhanced
mobility of polymer segments between the entanglements is
directly reflected in the prolongation of the transverse relaxation
time T2. Shearing the polymer sample results in an increase of
the transverse relaxation time, indicating enhanced molecular
mobility which implies, that as an effect of the shear a fraction
of entanglements in the polymer have been lost, resulting
in longer chain segments, these in turn exhibit an increased
molecular mobility as manifested in a prolongated T2.

Introduction
The flow behaviour of polymer melts is critical for various steps of the
processing and defines the materials properties in a wide range of applications.
Therefore rheological measurements are an important source of information on
the behaviour of polymer materials (13). For a further understanding molecular
insight into the origin of materials properties is desirable. The combination of

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 rheology with X-Ray scattering (rheo SAXS) has proven to be a valuable tool for
the investigation of molecular order resulting from shear of the material (46).
The combination of rheology with nuclear magnetic resonance (NMR)
as rheo NMR permits to link macroscopic properties as measured in rheology
with molecular parameters and thus yields insight in the rheological behaviour
of materials (7) with all the information content of NMR on structure, order
and mobility. Applications to liquid crystalline materials show shear-induced
orientation and the competition between orientation in the shear field to
magnetically-induced orientation (8, 9). The formation of multilamellar vesicles
has been followed as well (10). Applications to polymer melts sofar have been
limited to the room temperature range. Callaghan and coworkers showed, that
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org

the NMR relaxation of all protons in the polymer is affected by the shear. Results
are complicated by the formation of shear bands (11). Other studies showed, that
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

shearing a polymer melt may result in loosening of entanglements (12).

Especially the transverse relaxation time T2 is sensitive to slow motions (13)
as they are present in the polymer chain segments between chemical or physical
crosslinks (14, 15). Another efficient measure of the polymer dynamics is the
evaluation of the residual dipolar coupling between protons via the generation of
double quantum coherences (1618). A major advantage of the double quantum
experiments is the fact, that a buildup of the double quantum signal is observed as
opposed to the different kinds of decays like in the relaxation time experiments.
The data may directly be converted into a dynamic order parameter. Relaxation
experiments are feasible, even whan the residual dipolar couplings are too weak,
i.e. in liquid-like systems, they can often be performed much faster.

Experimental Section
For the rheo NMR experiments an in-house built high-temperature rheo
NMR probe head with an integrated Couette cell has been used. The probe
head fits in the micro 2.5 gradient system of a Bruker Avance 300 WB NMR
spectrometer operating at a Larmor frequency of 300 MHz for protons. The micro
2.5 gradient system generates magnetic field gradients of up to 1 T/m in three
axis. The Couette cell is composed of a 10 mm NMR tube as the outer cylinder
with an inner diameter of 8.6 mm and exchangeable rotating inner cylinders with
a diameter of 6 mm, generating a gap of 1.3 mm. The rotor is made of PEEK
with a structured surface to minimize slip. Shear is applied by driving the rotor
with an external servo motor located on the top of the magnet and controlled by
the spectrometer. The sample is heated by hot air. The gradient system, the shims
and the superconducting magnet are protected by a cooling gas flow. Between
the sample and the gradient coils a temperature gradient of 10 K/mm has to be
maintained. A dedicated temperature control system is in place to protect the
probehead, the gradient system and the superconducting magnet monitoring the
entire experiment. Initially temperatures have been calibrated in a an off-line
experimente, because temperature sensors would detoriate the sensitivity of the
NMR experiment. Gas flow and gas temperatures have been monitored to adjust
the sample temperature.
432
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The sample has been molten in the outer cylinder outside of the magnet.
Subsequent the components of the rheo cell were assembled in a dedicated holder
centering the rotor in the melt and placed into the probehead.
Imaging and velocity profiles, derived from a combination of NMR imaging
with PFG NMR, have been used to view the melt in the gap and to monitor the
steady shear. Due to the visco-elasticity, increasing rotational velocity causes the
polymer melt to climb up the rotor (Weissenberg effect) and thus disrupting the
shear in the gap. The used rotor frequencies were 0.5, 1, 2 and 5 rps, yielding
shear rates at the wall of the inner cylinder between 12 and 122 s-1.
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

Figure 1. Intensities of the double quantum signal of poly(propylene) as a

function of the sample temperature.

The influence of different temperatures and shear rates of the melt has been
investigated by relaxation and DQ measurements without spatial resolution.
CPMG, spin echo and inversion recovery sequences have been used for T2 and
T1 experiments, for the double quantum experiments a pulse sequence based on
(16, 17) has been applied. The 90 pulse length was 28 s and a repetition delay
of 3 s was employed. The number of points has been restricted to limit the time
duration for which the sample is exposed to high temperatures in order to avoid
sample degradation.
The sample used in this study was isotactic poly(propylene) (i-PP), blended
with 3% PP-g-maleic anhydride (Mw = 250 kg/mol, Mw/Mn = 4.1), which is often
used as an additive in composite materials. The melt temperature of 163.5 C
and the crystallinity of 50 % (based on the melting enthalpy of 207 J/g of 100%
crystalline i-PP (19)), were analyzed by DSC.

433
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Results and Discussion
Homonuclear double quantum spectra have been recorded as a measure of
the residual dipoar coupling during the melting process of the polymers. As is
depicted in Fig. 1 the intensity of the double quantum signal diminishes, when
the polymer melts. The mobility in the polymer becomes sufficiently high, that
the dipolar coupling is averaged to a degree, that no double quantum intensity is
observable. Unexpectedly no double qunatum signals have been observed under
the shear rates applied in the present study, indicating, that only limited order of
the polymer chains has been induced by the shear.
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

Figure 2. Longitudinal relaxation time (T1) experiments by inversion recovery of

poly(propylene) as a function of the sample temperature.

Fig. 2 shows the temperature dependence of the longitudinal relaxation time,

the lines show the fitted curves with the values shown in Table 1. At temperatures
below the nomimal melting point (160 C) biexponential decay is observed. Above
170 C T1 can be described by a single exponential. The increase of T1 with
increasing temperature is attributed to slowing down spin diffusion due to the
enhanced molecular mobility. The small difference of the two components below
the melting point and the fact that above the melting point a single component
is observed, show that spin diffusion is important for T1. The major relaxation
mechanism is the methyl rotation and thus the efficiency of the spin diffusion from
CH and CH2 protons is important for the longitudinal relaxation. The parameters
used to fit the experimental data of all T1 experiments a summarized in
Table 1.
The temperature dependence of the transverse relaxation time T2 is depicted
in Fig. 3 showing a strong increase of T2 upon melting. T2 increases further with
increasing temperature showing the enhanced mobility of polymer chain segments
between entanglements as expected. T2 shows a multiexponential behaviour as for
most polymers, and can sufficiently be described by a biexponential fit. The results
from the fit are shown in Table 2. A short component in the order of 20 to 80 ms
434
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and the longer component raises from 200 ms to 300 ms. The parameters used to
fit the experimental data of all T2 experiments a summarized in Table 2.
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

Figure 3. Decay curves from CPMG experiments as a function of the sample

temperature.

Figure 4. T1 as a function of the shear rate of poly(propylene) at 180 C.

If the inner cylinder is rotated with a torque applied from the servo motor, the
poymer melt in the gap between the two cylinders is sheared. The melt sticks to
both cylinders, thus the polymer is at rest on the outer wall and rotates with the
inner cylinder.

435
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

Figure 5. Decay curves from CPMG experiments at 180 C as a function of

the shear rate.

Shearing the polymer at a constant sample temperature of 180 C does not

result in a significant change of T1 as depicted in Fig. 4. Comparing to the
strong temperature dependence of the longitudinal relaxation seen in Fig. 2 one
can conclude, that the sample temperature has not been increased by shearing the
highly viscous polymer melt. From the viscosity a temperature increase of up to
30 K would have been anticipated. Because the sample is heated by an air flow of
17 l/min, the additional heat is taken away from the sample by convection.

Table 1. Results from the fits for the longitudinal relaxation time T1
T rotation T1 [s] T1 [s]
[C] [rev/s] component 1 component 2
160 0 0.73 0.06
170 0 0.73 0.10
180 0 0.74
190 0 0.83
180 0.5 0.70
180 1 0.63
180 2 0.64
180 5 0.64

436
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 2. Results from the fits for the transverse relaxation time T2
T rotation T2 [ms] T2 [ms]
[C] [rev/s] component 1 component 2
160 0 19 224
170 0 25 227
180 0 34 242
190 0 53 375
180 0.5 47 258
180 1 50 281
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

180 2 65 285
180 5 83 269

The effect of shearing on T2 is shown in Fig. 5, showing an increase in T2 as

a function of the shear. The signal decay curves under shear are well-described by
a double exponential decay. Both components increase from 40 ms to 60 ms and
150 ms to about 300 ms respectively. The fraction of the component with a larger
T2 increases with the shear rate.
The increase in T2 implies an increase in the mobility of chain segments as
a result of the shear. A decrease in T2 would have been indicative for possible
chain ordering, which has not been observed. The increase of T2 and thus of chain
mobility is explained by loosing some of the entanglements of the polymer chains
as a result of the shear-induced motion of the polymer chains.

Conclusions
Rheo NMR of polymer melts at high temperatures has been demonstrated.
Both the longitudinal and the transverse relaxation times exhibit a strong
temperature dependence in the temperature range of the melting point. The
observed temperature dependence of the longitudinal relaxation time has been
utilized to exclude a shear-induced rise in the sample temperature in the present
study.
The transverse relaxation time T2, which is sensitive to slow molecular
motions typical for polymer chain segments between crosslinks or entanglements,
is becoming longer. The polymer network looses entanglements as a result of
the shear. The absence of a significant double quantum signal together with the
increase in T2 indicate, that chain ordering, which would be anticipated, is not the
dominating effect. The effect of orientation of the polymer chains is negligible.
The most reasonable explaination for the prolongation of T2 is the partial loss of
entanglements resulting in longer chain segments between entanglements.

437
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Acknowledgments
This work has been supported by the Deutsche Forschungsgemeinschaft
(DFG) under grant SCHE 524/9 in the Materials World Network. The sample has
been provided by Dr. Harkin-Jones (Queens University Belfast).

References
1. Gogos, C. C.; Tadmor, Z.; Kim, M. H. Advances in Polymer Technology;
John Wiley & Sons, Inc.: New York, 1998; Vol. 17, No. 4, pp 285305.
Downloaded by OHIO STATE UNIV LIBRARIES on June 20, 2012 | http://pubs.acs.org

2. Dealy, J. M.; Wissbrun K. F. Melt Rheology and Its Role in Plastics

Processing; Chapman and Hall: London, 1996.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch026

3. Chenoy, A. V.; Saini, D. R. Thermoplastic Melt Rheology and Processing;

Marcel Dekker: New York, 1996.
4. Phillips, A. W.; Bhatia, A.; Zhu, P.-w.; Edward, G. Macromolecules 2011,
44, 3517.
5. Somani, R. H.; Yang, L.; Hsiao, B. S. Polymer 2006, 47, 5657.
6. Xu, J.-Z.; Chen, C.; Wang, Y.; Tang, H.; Li, Z.-M.; Hsiao, B. S.
Macromolecules 2010, 44, 2808.
7. Callaghan, P. T. In Encyclopedia of NMR; Grant, D. M., Harris, R. K., Ed.;
Wiley: Chichester, 2002; Vol. 9, p 737.
8. Grabowski, D. A.; Schmidt, C. Macromolecules 1994, 27, 2632.
9. Lukaschek, M.; Grabowski, D. A.; Schmidt, C. Langmuir 1995, 11, 3590.
10. Medronho, B.; Schmidt, C.; Olsson, U.; Miguel, M. G. Langmuir 2010, 26,
1477.
11. Callaghan, P. T.; Gil, A. M. Macromolecules 2000, 33, 4116.
12. Badiger, M. V.; Rajamohanan, P. R.; Suryavanshi, P. M.; Ganapathy, S.;
Mashelkar, R. A. Macromolecules 2002, 35, 126.
13. Fedotov, V. D; Schneider, H. NMR Basic Principles and Progress; Springer:
Heidelberg, 1989; Vol. 21.
14. Hiller, W.; Schneider, H. Acta Polym. 1988, 39, 276.
15. Knrgen, M.; Menge, H.; Hempel, G.; Schneider, H.; Reis, M. E. Polymer
2002, 43, 4091.
16. Saalwchter, K. Prog. Nucl. Magn. Reson. Spec. 2007, 51, 135.
17. Graf, R.; Heuer, A.; Spiess, H. W. Phys. Rev. Lett. 1998, 80, 5738.
18. Schneider, M.; Gasper, L.; Demco, D. E.; Blmich, B. J. Chem. Phys. 1999,
111, 402415.
19. Bu, H. S.; Cheng, S. Z. D.; Wunderlich, B. Makromol. Chem., Rapid
Commun. 1988, 9, 75.

438
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 27

NMR Imaging and Its Application in the Study

of Pharmaceutical Tablets
H. Thrien-Aubin, Y. J. Wang, and X. X. Zhu*
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Dpartement de Chimie, Universit de Montral, C.P. 6128, succursale

Centre-ville, Montral, QC H3C 3J7, Canada
*E-mail: [email protected]

NMR imaging is a non-destructive and non-invasive technique

especially well suited to investigate the kinetics of water
uptake, swelling and erosion of pharmaceutical tablets used
for drug delivery. Polysaccharides such as high amylose
starch, cellulose derivatives and chitosan have been used as
excipients in the tablets for the controlled release of drugs.
NMR imaging may be used to quantitatively study the swelling
of the compressed polymer tablets and the penetration of water
in the polymer matrices. These results help to elucidate the
effects of temperature, tablet size and drug loading on the
properties and drug release process of the tablets. The new
physicochemical understanding gained with NMR imaging
techniques may help in the design and development of new oral
tablets for more efficient drug delivery.

Introduction
Nuclear magnetic resonance (NMR) imaging, because of its non-destructive
and non-invasive nature, requires no physical slicing of the sample to observe
the interior of an object and has grown to be a technique of choice not only
for biomedical imaging, but also in a myriad of other applications in material
sciences and engineering (1). NMR imaging has become an interesting technique
in the study of the dynamics of diffusion in polymer systems (28), and has been
adapted to characterize the behavior of pharmaceutical tablets after immersion
(918). A better understanding of the process of controlled release of drugs
from the pharmaceutical tablets is essential to design and prepare more efficient

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 drug delivery systems. Therefore, it is essential to study the diffusion of small
molecules, especially water and drugs, and the swelling and the dissolution of the
tablets. NMR imaging may provide unique information on such phenomena.
The diffusion process in pharmaceutical tablets is usually studied by
gravimetric methods for solvent uptake and by high performance liquid
chromatography (HPLC) or UV spectrophotometry for the kinetics of drug
release. These techniques, and others such as optical methods (1921), allow
an efficient overall general characterization of the system, while NMR imaging
provides both general and spatially-localized information on the distribution of
the compounds of interest inside and outside the polymeric matrix. NMR imaging
is used to provide the spatial mapping of proton signals. The contrast of the
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org

images can be adjusted by the use of a range of physicochemical properties such

as the difference in concentration, relaxation times, self-diffusion coefficients
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

or flow rates. NMR imaging of pharmaceutical tablets has been mainly used to
image the protons of mobile water molecules inside the system during the kinetic
analysis of water uptake (9, 11, 15, 16, 2227) as well as the diffusion of drugs
from the tablets (22).

NMR Imaging Techniques

Basic Principles of NMR Imaging
In 1973, Lauterbur obtained the first image of the interior of an object (23)
using magnetic field gradient to spatially encode information on both frequency
and phase. Lauterbur (23) used frequency-encoding to map his sample, two
tubes of water in a larger D2O tube (Figure 1), by the conversion of spectral
information into spatial information. A linear gradient of magnetic field with
different orientation was applied to obtain multiple profiles of the object. With
different projection acquired at different angle, it was possible to reconstruct the
2D image of the original object using the backprojection technique and algorithms
used in X-ray tomography. Frequency-encoding is based on the fact that the
resonance frequency of a nucleus changes with the strength of the magnetic field.
If a gradient of magnetic field (G) is overlaid to the magnetic field of the NMR
magnet (B0), a variation in frequency is observed. The application of the gradient
during the signal acquisition (Figure 2A) will encode the space in frequency
leading to the recording of the samples profile, the spin density distribution, i.e.,
the projection of the sample along the axe where the gradient is applied.
Subsequently, in 1974, Mansfield introduced a technique to add a third
dimension to the NMR images using selective irradiation (24). As for the
frequency-encoding, the slice selection is based on the fact that the presence of
a gradient of magnetic field changes the Larmor frequency of the nucleus under
observation according to its position in space. To selectively excite a given slice
of the sample, a RF pulse should be applied simultaneously with the gradient
pulse (Figure 2B). A sinc-shaped RF pulse was used to selectively excite a
given Larmor frequency. When a short rectangular RF pulse is generated at a
frequency x, a broad distribution of RF with an average frequency centered at x
is generated, the broadness of the distribution being inversely proportional to the
442
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 length of the pulse. RF pulses of lower intensity and longer duration could be used
to contain only one frequency, or at least a narrower range of frequencies (24).
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 1. Lauterburs experiment. (A) Acquisition of 4 profiles where the gradient

axis is rotated by 45 increments. (B) Image obtained after the backprojection of
the 4 profiles. Adapted by permission from Macmillan Publishers Ltd: Nature
242, 190-191, copyright 1973 (23).

An important development to NMR imaging, allowing faster image

acquisition, was the concept of phase-encoding (25), which permits the
construction of 2D images much faster than with the Lauterbur projection method.
When a gradient of magnetic field is applied to the sample between the excitation
and the signal acquisition, the precession speed of the nucleus spins is modified
and spatially encoded (Figure 2C). The application of a linear gradient (Gx) causes
a dephasing of the spin leading to reduction in the signal intensity. The 2D image
is obtained by the frequency-encoding to provide a profile, and phase-encoding
to relate the position of the nucleus to the spin phase by the application of N
gradients of the magnetic field applied in the orthogonal direction to the slice and
frequency-encoding gradients.

Signal Acquisition

NMR images are usually obtained using pulse sequences based either on a
spin-echo (Figure 3A) or gradient-echo (Figure 3B) technique (26). The gradient-
echo permits faster imaging caused by small flip angles needed for the excitation
and the absence of a second RF impulsion which allows for shorter repetition time
(TR) leading to the faster acquisition of similar S/N ratio in comparison to the
spin-echo sequence. However, since gradient-echo is also more sensitive to small
variation in the magnetic field, artifacts could be observed in region of variable
magnetic susceptibility.
443
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The intensity of the signal measured by NMR imaging varies according to
Equations 1 and 2 for the spin-echo and gradient-echo experiments, respectively.
The equations show that the intensity of the signal (S) varies with sample-specific
factors such as spin density (), longitudinal relaxation time (T1) and transverse
relaxation time (T2), and with experiment-specific factors such as the echo time
(TE) and the repetition time (TR) and the flip angle () in the gradient-echo
experiment.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

In the study of drug release from pharmaceutical tablets, a crucial factor is

the diffusion coefficient or the velocity of the molecule in the presence of flow
at different positions inside the tablet (9, 11, 15, 16, 27). The self-diffusion
coefficients can be easily measured with NMR spectroscopy using the pulsed
gradient spin-echo (28) or stimulated echo sequence (29). A first magnetic field
gradient pulse is used to spatially encode the spin with their phase angle, while a
second gradient pulse of the same amplitude and intensity applied after a given
diffusion time is used to cancel the phase difference created by the first gradient
pulse if the nucleus has not diffused to a new position; otherwise the signal
measured will be attenuated. The NMR signal decreases with the amplitude of
the gradient pulses and the diffusion coefficient following the equation

where S(G) is the intensity of the NMR signal in presence of a gradient of

amplitude G, S0 the signal intensity in the absence of any gradient, the
gyromagnetic ratio of the nucleus, the duration of the two gradients pulses,
the delay between the two gradients pulses and the interval between the 90o
and 180o pulses of the spin-echo sequence. It is possible to obtain a diffusion
coefficient map by merging the pulsed gradient spin-echo with the imaging
spin-echo pulse sequences; one could then acquire a series of spin density images
with different amplitudes of diffusion gradient and build the diffusion coefficient
image (30, 31).

444
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 2. Basic principles of NMR imaging. (A) Frequency encoding of a

cylindrical object, generated by the application of a magnetic field gradient
during signal acquisition. (B) Slice selection with selective excitation pulse in
presence of a magnetic field gradient. (C) Phase encoding generated by the
application of a magnetic field gradient between the initial RF perturbation
and signal recording.

445
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 3. Simplified NMR imaging pulse sequences. (A) Spin-echo imaging;

(B) Gradient-echo imaging.

It is also possible to obtain images where the signal is proportional to the

relaxation times, by the acquisition of multiple images at different echo times,
which could be used to reconstruct a T2 map. A T1 map could be obtained by
the introduction of an inversion recovery segment in the pulse sequence and
the acquisition of a series of images at different inversion times. With these
relaxation time imaging techniques, it is possible to obtain useful results on the
local dynamics of the water molecules in the tablets, which help to gain a better
understanding of the swelling and drug release since the relaxation times vary
with the local molecular environment as shown in Figure 4 (11, 12, 32, 33).
446
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Technical Considerations

Equations 1 and 2 show that if the echo time TE is kept short in comparison
with T2 and if the repetition time TR is long in comparison with T1, the mapping
will be only proportional to the spin density, i.e., the concentration of the nucleus
under observation. If TR is kept short or TE is long, it is possible to obtained T1-
or T2-weighted images, respectively. By choosing the repetition and echo times,
one can change the signal intensity of different species of the image for a better
contrast; this is particularly interesting for whole body imaging since relatively
large variation of relaxation times may be observed for different types of tissues
(34).
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 4. Variation of the relaxation times according to the molecular motion

velocity of the sample.

In the study of diffusion in pharmaceutical tablets, the results obtained by

NMR imaging are usually maps of water concentration inside the tablets. Figure 5
shows examples of the spin density images of the water uptake of a polysaccharide-
based tablet. One can observe a receding dry part of the tablet (purple) and an
increase in the overall dimension of the tablet with the immersion time. At the
water-tablet interface in Figure 5, a red area is observed, giving the impression that
the spin density is higher in this region than in free water (14, 35, 36). This effect
is created to facilitate the identification of the water-tablet interface by setting the
TR shorter than the T1 of free water but longer than the T1 of the water inside the
tablet, leading to the attenuation of the signal of free water because the TR does not
allow the full relaxation of the free water protons, while allowing full relaxation
of the water inside the tablet between two repetitions.
The spin-spin relaxation is so fast in the dry part of the tablet that even with
very short TE it is difficult to observe the signal from this part. When the echo
time used is short enough, only the mobile protons are observed selectively. In
this case the water protons are the only ones contributing to the NMR signal. As
water diffuses in the tablet, the proton signal (or spin density) increases, which is
also partly due to an increase in the T2 of water (Figure 4).
The maximal resolution that could be achieved by NMR imaging varies with
a number of factors, including the concentration and relaxation time of the sample
447
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 and also the spectrometer frequency and gradient strength. In whole body imaging,
a resolution of 1-2 mm is typical, but in microimaging, which operate usually at a
much higher frequency, a resolution of 100 m is standard and resolutions of 10
m are possible.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 5. Radial section of the crosslinked high amylose starch (CHAS) tablet
after immersion in water at 37 C for (a) 45 minutes and (b) 50 hours. Color
gradient scale represents the 1H spin density. (see color insert)

NMR Imaging of Pharmaceutical Tablets

Experimental Setup

NMR imaging is an ideal method to record in situ the swelling behavior of

solid oral dosage forms. It can be used to evaluate the polymer concentration
profile in the tablets (37, 38), quantify dimensional changes (thickness, area and
volume) during the swelling (1416, 35, 36), and define the diffusion front of the
sample (14, 35). Furthermore, the NMR imaging studies can provide the diffusion
coefficient of a liquid component (15, 16), and T1 and T2 values which are related
to the environment of the penetrant and its interaction with the polymer excipients
(12).
The NMR imaging experiments are normally carried out at 37.0 C. We used
a Bruker Avance-400 NMR spectrometer operating at a frequency of 400.27 MHz
for protons and equipped with a microimaging probe with a 20 mm inner diameter.
A standard spin-echo pulse sequence was used to obtain spin density images of
each tablet in a 20 mm o.d. NMR tube containing 20 mL of the media (distilled
water, buffer solutions or simulated physiological fluids). A slice of 0.5 mm in
thickness was selected either perpendicular or parallel to the main magnetic field
using a sinc-shaped pulse. Eight scans were accumulated to obtain 128 128
pixel images for a field of view of 2.0 cm, leading to an in-plane resolution of
156 m. An echo time of 3 ms and a repetition time of 1 s were fixed, leading to
an acquisition time of about 17 min for each image. The tablets were placed on
a home-made Teflon support, which allowed three-dimensional liquid penetration
(Figure 6).

448
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 6. A tablet on a Teflon support placed in a 20 mm o.d. NMR tube.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Diffusion-weighted images were acquired by combining the spin-echo pulse

sequence with the pulsed gradient spin-echo (PGSE) pulse sequence developed by
Stejskal and Tanner (28). The diffusion time () and the length of the gradient
pulse () were set at 10 ms and 2 ms, respectively. The gradient strength varied
from 5 to 100 G/cm.
When the simulated physiological fluids were used, some defects might be
observed due to the high ionic strengths of the fluids, as shown by the bright spots
above the tablets in SGF in Figure 7. 1H tuning and matching strongly depend
on the ionic strength of the sample. A buffer solution may cause de-tuning of the
probe and the appearance of artifacts in the NMR images (39, 40).

Figure 7. The NMR images of the carboxymethylated starch (CMS)-chitosan

complex tablets immersed in water and simulated physiological fluids at 37 C
for 1, 5, and 10 h. (see color insert)

Swelling and Water Diffusion of Polysaccharide Tablets

Cross-linked high amylose starch (CHAS) (41, 42) was used to prepare
selected tablets. The starch of a high content of amylose yields a stronger gel
(43). Covalent cross-links and hydroxypropyl side chains of the final product
allow greater stability by hindering retrogradation over time (43). Retrogradation
449
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of starch is used to define the changes from amorphous state to a more ordered or
crystalline state.
CHAS can provide sustained drug release as an excipient of oral drug dosage
form by the formation of a membrane at the outer layer of the tablets upon
hydration and maintaining the tablet integrity in water for days. The absence of
erosion and limited swelling allow for further explorations of the applications in
drug delivery (44). Review articles on the NMR imaging of starch-based tablets
were published previously (45, 46). We summarize here the NMR imaging results
obtained for selected polysaccharide tablets (1416, 35, 36, 4547) in an effort to
illustrate the usefulness of NMR imaging in such studies.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org

Gel Formation and Water Gradient

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

NMR imaging could be used to spatially map the proton signals, and thus
to follow water uptake by starch tablets. In Figure 5, the tablet is immerged in
water inside a NMR tube and a 2D NMR image of a slice inside the sample could
be obtained using a slice selective pulse sequence. The NMR images show the
formation of a hydrated membrane at the water-tablet interface. Figure 8 shows
that the membrane thickness increases rapidly at first then remains constant at ca.
0.4 mm until 50 h. This membrane regulates the diffusion of water into the tablet
(48). The water concentration in the core of the tablet increases slowly after the
initial swelling of the tablet until a constant water concentration is reached. After
80 h, the membrane thickness increases drastically to reach ca. 2 mm and then the
gelation of the whole tablet is observed (14, 22, 35, 36, 48).
The NMR images provide access to water concentration profile inside the
tablet, which can be used to calculate the water diffusion coefficient (15, 16, 36).
If the diffusion is Fickian, and that the shape of the tablet is assumed to be an
infinite cylinder (to disregard the edge effects). At the initial stage of the diffusion
process in water at 37 C, these assumptions are considered to be fulfilled for the
CHAS tablet, at longer immersion times the edge effects will not be negligible.
The solution of Ficks law of diffusion for an infinite cylinder can be used for the
calculation of the diffusion coefficients (49). The diffusion coefficient measured
decreases by a factor of 10 during the first 10 h following the immersion (Figure 9),
the period during which the outer gel membrane was formed. The formation of a
gel layer with a low porosity resulting from the reorganization of the starch chains
accounts for the reduced diffusivity observed after the immersion. While Figure
9 provides a general trend for the changes in water diffusion coefficient values, it
should be noted, however, that the D values at longer immersion times have larger
errors since the contribution of the edge effects becomes important and, strictly
speaking, the assumptions made are no longer appropriate.

450
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 8. Thickness of the hydrated membrane shell of the CHAS tablet at 37 C

after its immersion in water.

Figure 9. The decrease of the self-diffusion coefficient of water at 37 C in the

CHAS tablet during the formation of the outer gel membrane.

When a CHAS tablet is immersed into water, water penetrates into the
hydrophilic polymer matrix to form a hydrogel, leading to a steep water gradient.
The formation of the membrane at the liquid/tablet interface is essential to the
tablet integrity and to the sustained drug release (15, 50). X-ray tomography
experiments of the membrane showed the significant difference between the
porosity of the outer layer and dry core (50). The formation of the membrane
can be followed with the NMR images adjusted with either proton density or
self-diffusion coefficient contrast (15). The different regions of the tablets are
more obvious on the self-diffusion coefficient profile extracted at the center of
the tablet (Figure 10) than in the water concentration profile (15). The membrane
is mainly composed of a region where the self-diffusion coefficient of liquid is
stable.

451
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 10. Self-diffusion coefficient profile in the center of the tablet for CHAS
tablet () without drug, () loaded with 10% acetaminophen, and () loaded with
10% ciprofloxacin after 15 h of immersion. Adapted from ref. (15).

Anisotropic Swelling

The swelling is anisotropic with the axial swelling being much more
significant than the radial swelling for all the tablets of different sizes and
within the temperature range from 25 to 60 C (Figure 11) (14, 35, 36). The
more pronounced axial swelling is related to the relief of the stresses induced
during compaction of the tablets (51). However, the thicknesses of the gel
layer along both directions are similar (52). The glassy-rubbery transition is
based on the lowering of the glass-transition temperature (Tg) of the polymer,
which is controlled by the water concentration and depends on temperature and
thermodynamic interactions of the polymer-water system. The swelling of the
CHAS tablets approaches its maximum at around 10 h at 37 C but 60% of
swelling can be achieved within the first 4 h (36). The water uptake and the tablet
swelling strongly depend on the size of the tablets (35). The swelling rate is faster
for small tablets along both swelling directions due to their larger surface/volume
ratio.

Effect of Temperature

The water diffusion in the CHAS tablets was studied at four different
temperatures (25, 37, 45, and 60 C) (14, 36). The diffusion process is Fickian
between 25 and 45 C and Case II at 60 C (36). The tablets manifested a faster
swelling with increasing temperature up to 60 C (Figure 11). The difference is
caused by the different degrees of the transformation from V-type single helix
polymorph to B-type double helices during water uptake (35, 36). The double
helical conformation acts as physical cross-links which in turn limit the swelling
and is thus essential for the sustained drug release, which is verified by CP-MAS
13C NMR spectroscopy (44, 53). Water penetrating into a tablet to acts as a

452
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 plasticizing agent. This plasticizing effect increases the intermolecular spaces and
thus allows greater mobility to starch molecules which rapidly reorganize into the
thermodynamically more stable B-type double helix conformation (43).
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 11. Swelling kinetics of the CHAS tablets along (A) radial direction and
(B) axial direction at various temperatures. , 25 C;, 37 C;, 45 C; and ,
60 C. Adapted from ref. (36).

Effect of Drug Loading

The drug loading effect was studied first by comparing tablets loaded with
10 wt% soluble drugs (ciprofloxacin and acetaminophen) and tablets without the
drugs (15). The extent of the swelling of the tablets was not noticeably influenced
by a 10 wt% drug-loading, since the presence of drugs at this content did not
interfere with the formation of double helices which limited the overall swelling of
the CHAS tablets (15). However, the presence of drug molecules caused a faster
water uptake, as reflected by the higher diffusion coefficients of water.
The swelling of the CHAS tablets loaded with different amounts of
acetaminophen (10, 20 and 40 wt%) were studied by NMR imaging (Figure 12)
(16). The presence of drug molecules accelerated water uptake due to the changes
in the chemical potential gradient. The higher drug loading led to a faster water
diffusion into the tablet. The drug solubility and the degree of drug loading
have little influence on the diffusion coefficient of water in the outer membrane
formed. Water diffuses faster for tablets with higher drug loading and increases
gradually with time (Figure 13).

453
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

Figure 12. NMR images of CHAS tablets of different drug loading levels (10 40
wt %) immersed in water at 37 C for 1, 2, 5, 10, 15, and 20 h. The purple part at
the bottom of the images is the Teflon support. Reproduced with permission from
NRC Research Press (16). (see color insert)

Despite the different rates of water uptake and diffusion in the tablets, the
percentage of the drug released remained similar for all the CHAS tablets. The
outer membrane ensured a controlled release of acetaminophen, regardless of the
amount of drug in the tablet (16).
The changes of the water diffusion coefficient and the tablet swelling strongly
depend on the solubility of the drug and the loading level. The drug releasing rate
merely undergoes a minor change with increasing drug loading amount. Compared
to the CHAS tablets without loaded drugs, the tablets loaded with 10 40 wt %
acetaminophen swell substantially faster. The matrix is able to keep the integrity
of all the tablets after their immersion in water for a long time up to 50 hours.

Figure 13. Self-diffusion coefficients of water in the inner core of the tablets
loaded with 10% (), 20% (), and 40% acetaminophen () obtained by
diffusion-weighted imaging. The D values of water in the hydrated gel membrane
of the tablets are shown by closed symbols. Reproduced with permission from
NRC Research Press (16).

454
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The Study of Various Polymer Matrices

The swelling of the CHAS tablets were compared with the tablets made of
polyelectrolyte matrices including chitosan, carboxymethylated starch (CMS) and
CMS-chitosan complex in both acidic and neutral media. Simulated gastric and
intestinal liquids (SGF and SIF) were used for the in vitro study of the swelling
and drug dissolution of the tablets (Figure 7). In the case of SGF, some artifacts
appeared on the NMR images.
The tablets made of chitosan, CMS, and the CMS-chitosan complex showed
pH-sensitive swelling due to the presence of COONa and/or NH2 functional
groups. The CMS-chitosan polyelectrolyte complex tablets have similar swellings
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org

to those of CHAS tablets in neutral media. However, the complex tablets in SGF
showed a greater extent of swelling. The remarkable dimensional change in SGF
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

was mainly contributed by chitosan that was dissociated from its complex with
CMS.
The effect of pH is more pronounced than that of the ionic strength for the
tablets without drug loading. The swelling at different pH values depends on
the pKa of the matrix. In the case of drug-loaded tablets, the drug release rate
is affected by the pKa values of both the matrix and the drug rather than the drug
loading level.

Conclusion
NMR imaging is an effective method that can provide information on
the diffusion and swelling processes of tablets useful in the development and
formulation of new drug delivery vehicles for pharmaceutical applications.
The contrast of the images may be adjusted by the choice of the experiment
parameters such as gradient strength, repetition time, echo time and flip angle.
The self-diffusion coefficients of the diffusing fluid may also be obtained and
correlated with the swelling of the tablets. The formation of the gel membrane,
the disappearance of the dry core and the drug dissolution can be visualized. The
membrane is crucial in obtaining the controlled release of drugs from the tablets.
The effects of tablet size, temperature, drug loading level and different polymer
matrices on the tablet swelling and liquid diffusion in the tablets have been
studied. These studies show that the use of NMR imaging techniques provides
results that facilitate the development and optimization of polymeric excipients
for controlled drug delivery.

Acknowledgments
The authors acknowledge the financial support from the Natural Science
and Engeneering Research Consil of Canada (NSERC), le Fond Qubcois de
Recherche sur la Nature et les Technologie (FQRNT), and the Canada Research
Chair program.

455
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 References
1. Blumich, B. NMR Imaging of Materials; Oxford University Press: Oxford,
2003.
2. Rothwell, W. P.; Holecek, D. R.; Kershaw, J. A. J. Polym. Sci., Part C:
Polym. Lett. 1984, 22, 241247.
3. Blackband, S.; Mansfield, P. J. Phys. C: Solid State Phys. 1986, 19,
L49L52.
4. Weisenberger, L. A.; Koenig, J. L. Appl. Spectrosc. 1989, 43, 11171126.
5. Malveau, C.; Beaume, F.; Germain, Y.; Canet, D. J. Polym. Sci., Part B:
Polym. Phys. 2001, 39, 27812792.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org

6. Mayele, M.; Oellrick, L. R. Appl. Spectrosc. 2004, 58, 338343.

7. Zainuddin; Strounina, E. V.; Hill, D. J. T.; Whittaker, A. K. Polym. Int. 2010,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

59, 15201525.
8. Hill, D. J. T.; Whittaker, A. K. Radiat. Phys. Chem. 2011, 80, 213218.
9. Bowtell, R.; Sharp, J. C.; Peters, A.; Mansfield, P.; Rajabi-Siahboomi, A. R.;
Davies, M. C.; Melia, C. D. Magn. Reson. Imaging 1994, 12, 361364.
10. Rajabi-Siahboomi, A. R.; Bowtell, R. W.; Mansfield, P.; Henderson, A.;
Davies, M. C.; Melia, C. D. J. Controlled Release 1994, 31, 121128.
11. Rajabi-Siahboomi, A. R.; Bowtell, R. W.; Mansfield, P.; Davies, M. C.;
Melia, C. D. Pharm. Res. 1996, 13, 376380.
12. Fyfe, C. A.; Blazek, A. I. Macromolecules 1997, 30, 62306237.
13. Fyfe, C. A.; Blazek-Welsh, A. I. J. Controlled Release 2000, 68, 313333.
14. Baille, W. E.; Malveau, C.; Zhu, X. X.; Marchessault, R. H.
Biomacromolecules 2002, 3, 214218.
15. Thrien-Aubin, H.; Zhu, X. X.; Ravenelle, F.; Marchessault, R. H.
Biomacromolecules 2008, 9, 12481254.
16. Wang, Y. J.; Ravenelle, F.; Zhu, X. X. Can. J. Chem. 2010, 88, 202207.
17. Mikac, U.; Sepe, A.; Kristl, J.; Baumgartner, S. J. Controlled Release 2010,
145, 247256.
18. Tajiri, T.; Morita, S.; Sakamoto, R.; Suzuki, M.; Yamanashi, S.; Ozaki, Y.;
Kitamura, S. Int. J. Pharm. 2010, 395, 147153.
19. Bussemer, T.; Peppas, N. A.; Bodmeier, R. Eur. J. Pharm. Biopharm. 2003,
56, 261270.
20. Gao, P.; Meury, R. H. J. Pharm. Sci. 1996, 85, 725731.
21. Moussa, I. S.; Cartilier, L. H. J. Controlled Release 1996, 42, 4755.
22. Fyfe, C. A.; Grondey, H.; Blazek-Welsh, A. I.; Chopra, S. K.; Fahie, B. J. J.
Controlled Release 2000, 68, 7383.
23. Lauterbur, P. C. Nature 1973, 242, 190191.
24. Mansfield, P.; Grannel, P. K. J. Phys. C: Solid State Phys. 1973, 6,
L422L426.
25. Kumar, A.; Welti, D.; Ernst, R. R. J. Magn. Reson. 1975, 18, 6983.
26. Callaghan, P. T. Principles of Nuclear Magnetic Resonance Microsocopy;
Oxford University Press: Oxford, 1991.
27. Tritt-Goc, J.; Kowalczuk, J.; Pislewski, N. J. Pharm. Pharmacol. 2003, 55,
14871493.
28. Stejskal, E. O.; Tanner, J. E. J. Chem. Phys 1965, 42, 288292.

456
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 29. Tanner, J. E. J. Chem. Phys. 1970, 52, 25232526.
30. Callaghan, P. T.; Eccles, C. D.; Xia, Y. J. Phys. E: Sci. Instrum. 1988, 21,
820822.
31. Callaghan, P. T.; Xia, Y. J. Magn. Reson. 1991, 91, 326352.
32. Kowalczuk, J.; Tritt-Goc, J. Carbohydr. Polym. 2011, 83, 166170.
33. Chowdhury, M. A.; Hill, D. J. T.; Whittaker, A. K.; Braden, M.; Patel, M. P.
Biomacromolecules 2004, 5, 14051411.
34. Rinck, P. A. Magnetic Resonance in Medicine, 5th ed.; Blackwell: Berlin,
2003.
35. Malveau, C.; Baille, W. E.; Zhu, X. X.; Marchessault, R. H.
Biomacromolecules 2002, 3, 12491254.
Downloaded by PENNSYLVANIA STATE UNIV on June 21, 2012 | http://pubs.acs.org

36. Thrien-Aubin, H.; Baille, W. E.; Zhu, X. X.; Marchessault, R. H.

Biomacromolecules 2005, 6, 33673372.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch027

37. Djemai, A.; Sinka, I. C. Int. J. Pharm. 2006, 319, 5562.

38. Baumgartner, S.; Lahajnar, G.; Sepe, A.; Kristl, J. Eur. J. Pharm. Biopharm.
2005, 59, 299306.
39. Bock, C.; Frederich, M.; Wittig, R.-M.; Prtner, H.-O. Magn. Reson.
Imaging 2001, 19, 11131124.
40. Bock, C.; Sartoris, F.-J.; Prtner, H.-O. Magn. Reson. Imaging 2002, 20,
165172.
41. Lenaerts, V.; Beck, R. H. F.; Van Bogaert, E.; Chouinard, F.; Hopcke, R.;
Desevaux, C. U.S. Patent 6,607,748, August 19, 2003.
42. Mateescu, M. A.; Lenaerts, V.; Dumoulin, Y. U. S. Patent 5,456,921, Oct.10,
1995.
43. Ravenelle, F.; Rahmouni, M. In Polysaccharides for Drug Delivery and
Pharmaceutical Applications; Marchessault, R. H., Ravenelle, F., Zhu, X.
X., Eds.; American Chemical Society: Washington, DC, 2006; pp 79104.
44. Lenaerts, V.; Moussa, I.; Dumoulin, Y.; Mebsout, F.; Chouinard, F.; Szabo, P.;
Mateescu, M. A.; Cartilier, L.; Marchessault, R. J. Controlled Release 1998,
53, 225234.
45. Thrien-Aubin, H.; Zhu, X. X. Carbohydr. Polym. 2009, 75, 369379.
46. Thrien-Aubin, H.; Zhu, X. X. In Polysaccharides For Drug Delivery And
Pharmaceutical Applications; Marchessault, R. H., Ravenelle, F., Zhu, X.
X., Eds.; American Chemical Society: Washington, DC, 2006; pp 105120.
47. Wang, Y. J. Masters thesis, Universit de Montral, Montral, Canada, 2011.
48. Kowalczuk, J.; Tritt-Goc, J.; Pislewski, N. Solid State Nucl. Magn. Reson.
2004, 25, 3541.
49. Crank, J. The Mathematics of Diffusion, 2nd ed.; Oxford University Press:
Oxford, 1975.
50. Chauve, G.; Ravenelle, F.; Marchessault, R. H. Carbohydr. Polym. 2007, 70,
6167.
51. Papadimitriou, E.; Buckton, G.; Efentakis, M. Int. J. Pharm. 1993, 98,
5762.
52. Moussa, I. S.; Lenaerts, V.; Cartilier, L. H. J. Controlled Release 1998, 52,
6370.
53. Shiftan, D.; Ravenelle, F.; Mateescu, M. A.; Marchessault, R. H. Starch/
Strke 2000, 52, 186195.

457
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 28

Effect of Molecular Architecture on the

Performance of 19F NMR Imaging Agents
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Hui Peng, Kristofer Thurecht, Steven Hsu, Idriss Blakey,

Oliver Squires, Nyoman Kurniawan, Stephen Rose,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

and Andrew K Whittaker*

Australian Institute for Bioengineering and Nanotechnology and

Centre for Advanced Imaging, University of Queensland,
Brisbane, QLD 4072, Australia
*E-mail: [email protected]

Traditional magnetic resonance medical imaging agents based

on paramagnetic ions or particles have a number of critical
limitations, namely image contrast is often limited at low
concentrations and sensitivity of the agents becomes diminished
at higher magnetic field strengths. In this chapter we describe
the development of a new class of polymeric imaging agent
which relies on direct detection of the NMR signal of fluorine
nuclei and hence circumvents these potential problems. The
imaging performance of the new materials relies on the
maintenance of librational motion of the fluorinated segments,
necessary for partial averaging of the otherwise strong
dipole-dipole interactions and large chemical shift interactions
of the 19F nuclei. The performance of partially-fluorinated block
copolymers assembled into particles is described and related
to the interactions of the polymer chains with the surrounding
solvent medium. A second generation of materials based on a
hyperbranched structure was developed based on the learning
obtained from the earlier experiments and these demonstrated
excellent in vitro and in vivo imaging performance.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Introduction
Over the past several years the field of in vivo 19F NMR imaging has started
to receive renewed attention. This is a result of particular advantages of the 19F
imaging experiment, since unlike in 1H NMR imaging, the body does not contain
a significant concentration of fluorine to provide a background signal. The signal
from injected fluorine-containing molecules will therefore appear bright against a
dark background, and furthermore if the image is superimposed on a proton density
image the location of the probe molecules in the body can be readily discerned. If
probe molecule has included a targeting functionality, the text above describes the
field of 19F molecular imaging.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

The fluorine nucleus possesses several attractive features for NMR imaging.
The NMR-sensitive isotope,19F, is present in high natural abundance, possesses a
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

spin of I=1/2 and has the second-highest gyromagnetic ratio (40.03 MHz/T) of the
commonly-encountered nuclei (1). This last property provides high sensitivity in
the 19F NMR experiment, however in the absence of fast, large amplitude motion
may lead to substantial dipolar couplings to near-neighbor nuclei (1H and 19F). This
is compounded by the large static chemical shift anisotropy of fluorine. Despite
this, the promise of high sensitivity imaging of mobile fluorinated species has lead
to a growing body of work.
Initial studies of the potential of 19F for medical imaging were performed by
Holland and co-workers who reported images of solutions of sodium fluoride and
perfluorotributylamine within phantoms (2). The first in vivo 19F images were
reported by McFarland et al. (3) on rats injected with fluorocarbons. The 19F
longitudinal relaxation time is known to depend on the oxygen tension in the blood
stream (4), and this was exploited by several workers in small animal models (5,
6). These early studies did not progress beyond pre-clinical examination, due
primarily to the lack of clinical scale scanners suitable for 19F observation. This
commercial bottleneck has now been removed.
The field therefore has been reinvigorated, as reflected by the work from
groups in Carnegie Mellon University (79) and Washington University in St.
Louis (1015) who have advanced work on nanoparticles agents based on linear
and cyclic perfluoroethers, and on branched polymers. The former molecules
have the advantage of being able to be labelled covalently in a facile manner with
fluorescent dyes, while the Wickline group incorporates a fluorescently-labelled
phospholipid into an emulsion particle (1013). The principal application
proposed for these agents is cell tracking (16), and these workers and others have
demonstrated in vivo imaging of dendritic cells labelled appropriately with the
fluorinated particles. In more recent work, Higuchi et al. (17) have reported the
imaging of amyloidophilic compounds containing fluorine as potential markers
for the early stages of Alzheimers disease. The potential use of fluorinated
markers for non-invasive studies of physiology is abundantly clear and has been
reviewed by Yu and others (18). In addition Ruiz-Cabello and colleagues have
recently published an excellent review of the application of 19F spectroscopy and
MRI to medicine (19).
Examination of the literature reveals that there remain a number of
fundamental challenges to be overcome for this technology to become widely

460
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 accepted. These are principally the low sensitivity resulting from the low fluorine
content of many potential marker molecules, the large (isotropic) chemical shift
dispersion of the 19F nucleus, and the short T2 relaxation times observed for
potential structures arising from the large shift anisotropy and strong dipole-dipole
couplings to near-neighbor fluorine and proton nuclei. This chapter describes
two generations of polymeric molecules developed to address these issues, firstly
self-assembled particles consisting of partially-fluorinated amphiphilic block,
and secondly hyperbranched polymers consisting of similar building blocks.
Polymeric agents have been chosen because of their potential for high fluorine
content, and the possibility of attachment of multiple targeting agents to the end
groups and side chains. Excretion of the polymers was confirmed as described
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

below. The in vitro and in vivo performance is related in this chapter to the
molecular structure of the polymeric agents.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

Materials and Methods

Materials
Tert-butyl acrylate(tBA), 2,2,2-trifluoroethyl acrylate (TFEA), 2,2,2-
trifluoroethyl methacrylate (TFEMA), hexafluoroisopropyl acrylate (HFPA),
hexafluoroisopropyl methacrylate (HFPMA) and n-butyl acrylate (BA) were
purchased from Aldrich, dried over CaH2 and distilled under reduced pressure
before use. CuBr was obtained from Aldrich and purified by washing with
glacial acetic acid followed by absolute ethanol and ethyl ether, and then dried
under vacuum. Ethyl 2-bromopropionate (EBrP), N,N,N,N,N-pentamethyl
diethylenetriamine (PMDETA) and trifluoroacetic acid (TFA) were purchased
from Aldrich and used as received. Acetone was refluxed over CaH2 and distilled
under nitrogen before use. 2-dodecylsulfanyl thiocarbonyl sulfanyl-2-methyl
propionic acid (20) and the alkyne analogue (21) were synthesized as previously
reported. AIBN was recrystallized from methanol three times before use.
Dried toluene, dichloromethane (DCM) and dimethylformamide (DMF)
were obtained oxygen- and moisture-free using a purification unit under an inert
nitrogen environment (MBraun Solvent Purification System Auto-5).

Characterization
1HNMR spectra of the polymers and intermediates were measured on a
Bruker Avance 500MHz or 300 MHz spectrometer equipped with a TXI probe
and operating at room temperature.
Gel permeation chromatography (GPC) measurements were performed using
a Waters Alliance 2690 Separations Module equipped with an auto-sampler,
column heater, differential refractive index detector, and a Photodiode Array
(PDA) connected in series. HPLC-grade tetrahydrofuran was used as eluent at
a flow rate of 1 mL/min. The columns consisted of three 7.8 300 mm Waters
Styragel GPC columns connected in series, comprising two linear UltraStyragel
and one Styragel HR3 columns. Polystyrene standards ranging from 517 to 2106
g mol-1 were used for calibration.
461
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Differential Scanning Calorimetry (DSC) analyses were performed using a
Mettler Toledo DSC1 Star System calorimeter with a sub-ambient temperature
attachment. The glass transition temperature (Tg) was determined as the centre of
the change in heat capacity measured at a heating rate of 10 C/min.
Dynamic light scattering (DLS) measurement was performed on a Nanoseries
(Malvern, UK) zetasizer. The scattering angle used was 90 and the temperature
was fixed at 298 K.
19F NMR spectroscopy was performed on an AMX300 spectrometer

interfaced to a 7 T vertical super-wide bore magnet. The system is equipped

with a Bruker microimaging gradient set and the probe used was a Bruker 5 mm
19F single-tuned bird-cage resonator probe tuned to 282.404 MHz for fluorine
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

detection. The 90o pulse time was 5.5 s. Samples of the fluorinated block
copolymers were prepared as described below, and either maintained in mixed
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

DMF-water solvent (or DMSO-water) or examined after dialysis in water. The

concentration of the solutions prepared in 5 mm NMR tubes was approximately
2.5 wt. % of polymer. All measurements were performed at 310 K.
19F NMR spectra of the solutions were recorded using a 90o pulse of 5.5 s

and a repetition delay of 10 seconds. The spectrum width was 50 kHz, and 4
k data points were collected. Typically 16 FIDs were co-added to improve the
signal-to-noise ratio.
19F T2 relaxation times were measured using the CPMG pulse sequence, with

from 2 to 256 180o pulses in the echo train. A total of 12 data points were collected.
Spin-lattice relaxation times were measured using the standard inversion-recovery
pulse sequence, with 16 value of inversion time used. A single T1 relaxation time
was observed for all samples.

19F MRI Experiments

19Fimages were collected on the AMX 300 spectrometer using the 3D spin-
echo pulse sequence (22). Typical experimental details are as follows. The field of
view was 2020 mm2, and the slice thickness was 2.5 mm in a 1281288 matrix.
The echo time was 4.8 ms and the repetition time equal to three seconds. The
images presented here were all acquired by co-adding two acquisitions resulting
in a constant imaging time of one hour and 20 minutes.
In vivo 19F MRI images were collected on a Bruker 700 MHz wide-bore
microimaging system. Proton images were taken using a standard gradient echo
sequence. The 19F images were collected using a spin-echo 3D pulse sequence with
a repetition time of 1 s and echo time of 6.4 ms. A total of 32 scans were averaged
to improve the signal-to-noise and 8 slices were generally taken with a field of
view of 3 cm. The total scan time for the in vivo experiments was approximately
8 minutes.
Ethical clearance for testing of the polymers on live mice was obtained from
the University of Queensland (AIBN/076/08). The respiration rate of the mouse
was monitored at all times during the imaging.

462
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Preparation of Block Copolymer Assemblies

Example Synthesis of Amphiphilic Block Copolymer

Monobromo-terminated p(tBA) (Mn=6800, Mw/Mn = 1.09), prepared by

ATRP, and CuBr were added to a 50 mL round-bottom flask which was sealed
with a rubber septum, then degassed with argon. Deoxygenated toluene was
added, after which nBA and the fluoroacrylate (FA, here FA refers to TFEA,
TFEMA, HFPA and HFPMA monomer) were added, both via syringes, then
PMDETA was introduced. The molar ratio of p(tBA) : CuBr : PMDETA : nBA
: FA=1:1:1:30:100. The solution was stirred for approximately 10 minutes until
the Cu complex had formed, as judged by the complete dissolution of the CuBr
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

powder particles and the formation of a deep green solution. After complex
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

formation, the flask was placed in an oil bath at 90 C. After certain times, the
polymer was dissolved in THF and passed through an alumina column to remove
the copper catalyst. The THF was removed by evaporation and the polymer was
then dried under vacuum at 40 C. For further purification, the polymer was
dissolved in THF, precipitated into a 10-fold excess of hexane, and then washed
with hexane twice. The polymer was dried under vacuum at 40 C for two days.
Hydrolysis of the tert-butyl groups was achieved by addition to the solution
of the block copolymer a quantity of trifluoroacetic acid. A clean, 100 mL, round-
bottom flask fitted with a stirrer bar was charged with p(tBA)-b-p(nBA-co-FA),
followed by dried dichloromethane. The mixture was stirred for 10 min to dissolve
the polymer. Trifluoroacetic acid (TFA, 5.0 equiv to the tert-butyl ester) was
then added. After the mixture was allowed to stir at room temperature for 10
h, the dichloromethane and excess TFA were removed at room temperature with
dry nitrogen gently flowing through the flask. The resulting light-brown polymer
solid was dried under vacuum for 3 days under room temperature. For further
purification, the polymer was redissolved in THF, precipitated into a 10-fold excess
of water, and then washed with water twice. The polymer was dried under vacuum
at room temperature for three days to give to a white product which was pAA-b-
p(nBA-co-FA).
Spherical micelles were obtained by the dissolution of the purified
amphiphilic copolymer pAA-b-p(nBA-co-FA) (20 mg/mL) in DMF, a good
solvent for both blocks, followed by the gradual addition (10.0 mL/h) of an
equal volume of non-solvent for the hydrophobic polyfluoroacrylate to induce
micellization. The micelles were allowed to stir for 24 h before being transferred
to pre-soaked and rinsed dialysis bags (molecular weight cut off, MW=3500) and
dialyzed against ultrapure water (18.2 m.cm) for three days for removal of the
remaining organic solvent. The particle sizes of the micelles were determined by
DLS.

Synthesis of Hyperbranched Copolymers

One gram of dimethylaminoethyl methacrylate (DMAEA) (1.06 mL, 6.98 x

10-3 mol), 269 mg TFEA (221 L, 1.74 x 10-3 mol), 157 mg 2-dodecylsulfanyl
thiocarbonyl sulfanyl-2-methyl propionic acid (4.34 x 10-4 mol), 7 mg AIBN
463
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (4.34 x 10-5 mol) and 101 mg EGDMA (96 L, 5.1 x 10-4 mol) were dissolved
in 5 mL toluene and sealed in a round-bottom flask fitted with a septum. The
mixture was purged with argon for 15 minutes then heated to 65 oC for 48
hours. Upon completion of the reaction, the polymer was precipitated in hexane
and the viscous oil dried under vacuum. The polymer was characterized by
gel permeation chromatography couple with multi-angle laser light scattering
(GPC-MALLS) and 1H NMR. The alkyne-terminated PDMAEA-stat-TFEA was
prepared as above using an alkyne-terminated RAFT agent (174 mg, 4.34 x 10-4
mol) in place of the 2-dodecylsulfanyl thiocarbonyl sulfanyl-2-methyl propionic
acid.
The hyperbranched polymer was chain extended with poly(ethylene glycol)
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

methacrylate (PEGMA) in the following manner. 0.5 g of the hyperbranched

polymer (~ 1.5 x 10-4 mol, ~ 6 x 10-4 mol RAFT units) and 2 g PEGMA-475
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

was dissolved in 5 mL THF containing 2.5 mg AIBN. The solution was degassed
with argon and then reacted for 24 hours at 65 oC. The polymer was precipitated in
hexane and dialyzed against water for 5 days to remove unreacted monomer. The
polymer was freeze-dried yielding a slightly yellow, viscous oil.
Folic acid was attached to the hyperbranched polymer via standard DCC
coupling of acid group on folic acid to the amine group on the polymer.
Typically, the alkyne-terminated hyperbranched polymer was chain extended
with 11-azido-3,6,9-trioxaundecane-1-amine via standard azide-alkyne click
chemistry. The terminal amine group was then exploited for DCC coupling. The
concentration of folic acid attached to the polymer was determined by UV-vis
spectrophotometry by measuring absorbance at 363 nm. Typically, 2-3 folic acid
moieties were attached per polymer chain.

Results and Discussion

Synthesis and Evaluation of Diblock Copolymers
A series of amphiphilic block copolymers containing partially-fluorinated
segments were prepared using controlled radical polymerization (ATRP),
subsequent removal of t-butyl groups to produce a hydrophilic polyacid block.
Solutions of assemblies were prepared by addition of water to the polymers in
DMF solution followed by dialysis. The overall synthetic scheme is illustrated
below (Scheme 1) and described in the Experimental section above. Polymers
of low molar mass dispersity and of high structural fidelity were produced and
characterized as described in more detail elsewhere.
The NMR imaging performance of the diblock copolymer assemblies after
dialysis into pure water was poor, however, good images were obtainable for
solutions of the particles in mixed solvent systems, for example a 1:1 mixture
of water and DMF. The 19F NMR images of dilute solutions (2.5 wt%) in the
mixed solvents were measured at 7 T using a standard three-dimensional spin
echo pulse sequence with an echo time of 4.8 ms, and a repetition time of 3
seconds. The imaging time was restricted to 80 minutes, and comparisons were
made of the signal-to-noise ratio of the images under identical conditions. For
a spin-echo imaging sequence the image intensity is a function of the fluorine
464
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 content, the spin-spin and spin-lattice relaxation times, and the imaging echo and
repetition times, as described in equation 1 (23). The fluorine content is known
from the structure of the polymer which was determined by careful chemical
characterization, as described in the Experimental section. The spin-lattice (T1)
relaxation times were measured for all of the samples and were of the order
of 500 ms. On the other hand, the spin-spin (T2) relaxation times did vary
appreciably across the materials prepared (from < 0.5 ms to > 300 ms depending
on the structure) (24), and showed at least two components when measured by
the CPMG sequence. Using these parameters, and assuming that all fluorine
nuclei could be measured under high-resolution NMR condition, it was possible
to predict the relative imaging intensities of all of the polymers prepared in this
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

study. A comparison of the predicted relative intensities and the observed MRI
intensities is provided in Figure 1.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

Scheme 1. The synthetic route and structure of the amphiphilic copolymer

PAA-b-P (nBA-co-TFE (M)A).

Figure 1. Bar plot showing the ratio of the measured NMR image intensity
(arbitrary units) divided by the relative intensity predicted from the observed
molecular parameters and using equation 1.
465
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 As can be seen in Figure 1, the NMR imaging performance of the diblock
copolymer assemblies depends strongly on the chemical structure of the
partially-fluorinated monomer and the identity of the organic solvent in the
mixed solvent system. Several important rules can be obtained from this simple
comparison. Firstly, for an equivalent fluorine content the acrylate polymers
(P1, P3) perform significantly better that the methacrylate polymers (P2, P4).
Secondly, polymers in water:DMF (dark gray bars) consistently have a higher
image intensity compared with the polymers in water:DMSO (light gray bars).
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

Finally the hexafluoroisopropyl-containing acrylate polymer (P3) provided the

best performance relative to the predicted behaviour, however it must be noted
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

that the relatively low reactivity of the HFPA to free radical polymerization lead
to low levels of incorporation of this monomer and hence overall low image
intensities.

Figure 2. Stack plot of the 19F NMR spectra of the diblock copolymer
PAA50-b-P(nBA39-co-TFEA115) in mixtures of DMF-d7 and D2O as a function of
volume fraction of D2O.
A detailed examination of the effect of polymer and solvent on the imaging
performance has been reported elsewhere (24, 25), however an additional
important piece of information can be obtained from the 19F NMR spectra
collected under high-resolution NMR conditions (receiver dead time of 15 s) as a
function of solvent composition. In Figure 2 above we have plotted the 19F spectra
of PAA50-b-P(nBA39-co-TFEA115) in the different solvents. It can be clearly seen
that the linewidth of the observed NMR signal increases as the solvent becomes
more polar. On passing from DMF to D2O the block copolymers are forced to
assemble into particles with most-likely a shell of polar poly(acrylic acid) chains.
The increasing line width is evidence that the mobility of the fluorine-containing
side-chains becomes progressively more restricted on addition of water. Of
more relevance to the imaging performance, the integrated intensities of the 19F
signal of the trifluoroethyl side-chains decreases appreciably as solvent polarity
466
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 increases (data not shown here). We observed an approximately 30% decrease in
intensity for the TFEA-containing block copolymers in D2O. The intensity of the
spectra of the TFEMA-containing materials decreased by 55% in 1:1 DMF:D2O
and approximately 90% in pure D2O. This observation accounts for the lower MRI
imaging intensity observed for the methacrylate polymers. Remarkably there is
a population of nuclear spins with very short T2 relaxation times, of the order of
several microseconds or less, which do not contribute to the high-resolution NMR
experiments (Figure 4) and of course the spin-echo-based MRI experiment. This
population of fluorine atoms is behaving much like a rigid solid polymer and is
due to fluorinated segments associating strongly in the hydrophilic environment.
Based on these observations, we proposed a set of design criteria for
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

successful, i.e. high sensitivity 19F MRI agents. Clearly the agents must have
a high volume/weight fraction of fluorine atoms so as to maximize the potential
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

signal amplitude. The hexafluoroisopropyl acrylate monomer gave the best

imaging performance, however its use is restricted by the low reactivity to
radical addition as discussed above. The comparison between the performance
of the acrylate and the methacrylate polymers demonstrates that the flexibility
of polymeric main chain is of importance. It is well known that methacrylate
polymers are relatively conformationally-restricted compared with acrylate
chains, as reflected in the systematically lower glass-to-rubber transitions in bulk
acrylate polymers. Chain flexibility is important in these application because of
the role chain motion plays in motion of the side chains (26, 27). The importance
of maintaining flexibility of the fluorinated segments was demonstrated by the
effects of the solvent on the NMR properties and the imaging intensities. Thus it
is necessary to design materials in which association of the fluorinated moieties,
driven here by unfavorable polymer-solvent interactions, is frustrated. This
can be achieved for example by including a branched structure. These design
considerations have the overall effect of increasing the amplitude of motion of the
trifluoroethyl groups, which in turn partially averages the dipolar and anisotropic
chemical shift interactions of the fluorine nuclei. Finally we may desire to attach
to the imaging agent a suitable targeting group, to direct the molecules in vivo to
a particular tissue type, for example cancer cells.

Hyperbranched Polymeric 19F Imaging Agents

To overcome the problems encountered with particles self-assembled from
diblock copolymers, we proposed the synthesis of hyperbranched polymers which
demonstrate high segmental molecular mobility while maintaining relatively high
fluorine contents. A degree of control over the structure of the hyperbranched
polymers can be achieved using the controlled radical polymerization method
dubbed reversible addition fragmentation chain transfer (or RAFT). Molecular
mobility is maintained in these molecules by using acrylate (i.e. low-Tg) and
polar repeat units that are well hydrated under aqueous conditions. A difunctional
monomer is incorporated to introduce branching placed in a statistical manner;
this has the effect of frustrating the aggregation of the fluorinated segments
present in the diblock copolymer assemblies, allowing incorporation of up to 20
mol. % of the fluoro-monomer. Hyperbranched polymers possess a degree of
467
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 shape-persistence which means that the disadvantages of polymer assemblies,
such as disassociation under dilute conditions, do not arise. The highly branched
structure and hence high content of chain ends enables the fictionalization of the
polymers with a high concentration of targeting ligands. Finally use of RAFT
polymerization allows one to incorporate chemically-different end groups suitable
for labeling with multiple functional groups via orthogonal chemistries.
The hyperbranched polymers reported here were synthesized using the
method first published by Perrier and his colleagues (28); the reaction scheme is
reproduced in Scheme 2. An alkyne-terminated derivative of 2-dodecylsulfanyl
thiocarbonyl sulfanyl-2-methylpropionic acid was employed as the RAFT agent.
For example, a statistical hyperbranched copolymer of dimethylaminoethyl
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

acrylate (DMAEA - 77 mol. %) and trifluoroethyl acrylate (TFEA 19 mol. %)

was synthesized using ethyleneglycol dimethacrylate (EGDMA 4 mol. %) as
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

branching agent. Chain-extension of the resultant hyperbranched polymer with

polyethyleneglycol monomethylether methacrylate (PEGMA) was performed
resulting in a highly-branched polymers having a polycationic core with pendant
PEGMA arms which because of their pronounced hydrophilicity are likely
to surround the partially-fluorinated core. As reported by us elsewhere (29)
we examined the cytotoxicity of the various generations of hyperbranched
polymers using the MTS cell viability assay (30). These tests demonstrated that
chain-extension with the PEG-containing monomers reduced the toxicity of the
polycationic core to levels comparable to the non-toxic control, PEG(400). This
suggests that the PEGMA chains are able to shield the hyperbranched core.

Scheme 2. Generic scheme for synthesis of hyperbranched polymers using RAFT

polymerization. The intact RAFT end groups are available for extension with
PEGMA.

The 19F NMR imaging performance of this series of hyperbranched polymers

is outstanding, with excellent images being obtained in vitro over short imaging
times. In Figure 3 below we show the cross-sectional images of four plastic tubes
containing solutions of the hyperbranched polymer at various concentrations. The
right hand part of the figure demonstrates the good linear relationship between
polymer concentration and image intensity.

468
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

Figure 3. The left-hand side of this figure shows the 19F image of four solutions of
the hyperbranched polymer (Scheme 2) in water at concentrations of 5, 10, 15, 20
mg/mL, shown as areas of increasing brightness with increasing concentration.
On the right is a plot showing the linear relationship between 19F image intensity
and solution concentration.

Excellent in vivo imaging has also been demonstrated (29). In the absence of
a targetting ligand the polymer was rapidly excreted, being observed in the bladder
two hours after injection into the tail vein of the mouse. Clearly the compact
hyperbranched polymerics are able to be pass through the tight membranes within
the kidney. To demonstrate the ability of these molecules to target specific cell
types the alkyne-functionalised hyperbanched polymers were conjugated via the
Huisgens cycloaddition reaction with azide-functionalised folate. It is well known
that cancer cells over-express the folate receptors. The animal model for this
experiment was a mouse with approximately 106 B16 melanoma cells (a mouse
cell line) injected sub-cutaneaously. One week following implantation of the cells,
20 L of a 20 mg/mL solution of the folate-functionalised hyperbranched polymer
was injected into the tail-vein of the mouse. After waiting one hour the 19F image
of the anaesetised mouse shown in Figure 4 was acquired on a vertical-bore 16.4
T system. The total imaging time was just over eight minutes. In Figure 4 the 19F
image in orange/orange shades has been superimposed onto the 1H density image,
and demonstrates clearly that there is localisation of the targetting agent in the
centre of the tumour. We have also demonstrated that in the absence of the folate
ligand no accumulation of the polymer was observed in the tumour mass. There
is clear evidence that the hyperbranched imaging agent can provide high-quality
images within short time periods, and that the incorporation of folate groups results
in localisation of the imaging agent in the tumour mass. Further experiments are
in progress to extend the number of animal targets, and to for example examine
the limits of sensitivity of the agents especially in non-solid tumours.

469
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

Figure 4. Targeted imaging of B16 melanoma cells. The proton density of the
image of the torso of the mouse shows the kidneys (yellow curve) and the tumor
mass (red curve) highlighted. Superimposed on this image in color is the 19F
image, demonstrating 19F intensity within the mass of the tumors. A threshold
has been applied to the 19F image so that regions of low intensity have no color,
those with a moderate intensity appear red, and high intensity regions appear
yellow. The fluorine imaging time was 8 minutes. (see color insert)

Conclusions
In this chapter we have described the development and testing of two
generations of 19F MRI agents with potential application in so-called molecular
imaging. Extensive learning was gained from the preparation of a large series of
partially-fluorinated amphiphilic diblock copolymers, which when transferred to
aqueous solution formed nanometer-scaled assemblies. The 19F imaging intensity
was found to be profoundly influenced by the details of molecular structure and
the tendency of the fluorinated segments to self-associate. This behaviour of the
diblock copolymers was determined by the interactions of the chain segments with
each other, and with the surrounding medium. Without control the lowest-energy
structures were also those in which solid-like structures were present. Ultimately
the reduction in molecular motion resulted in a decrease in the spin-spin relaxation
times and a pronounced deterioration in the imaging performance.
The learning gained from these self-assembled systems lead to the design and
synthesis of a series of uni-molecular particles of hyperbranched polymers. A
degree of control of the branching could be achieved using RAFT polymerization
methods. The presence of branched resulting from the inclusion into the
polymerizing mixture of a di-functional monomer lead to a shape-persistent
structure, and which in combination with the proximity of hydrophilic monomeric
units lead to a frustration of the association of the fluorinated monomers,
as demonstrated by the extended spin-spin relaxation times, and the excellent
470
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 imaging performance. The molecules were tested both in test tubes, and in
animals, and performed exceptionally well, with only relatively short imaging
times required in the in vivo experiments. In the absence of targeting moieties the
polymers are excreted rapidly, however, when conjugated with the folate ligand,
the molecules were directed to a mass of melanoma cells previously seeded under
the skin of the mouse. The excellent imaging performance demonstrates the
potential of this class of agents, and this potential is being actively explored by
our group.

References
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

1. Ulrich, A. S. Prog. Nucl. Magn. Reson. Spectrosc. 2005, 46, 1.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

2. Holland, G. N.; Bottomley, P. A.; Hinshaw, W. S. J. Magn. Reson. 1977, 28,

133.
3. McFarland, E.; Koutcher, J. A.; Rosen, B. R.; Teicher, B.; Brady, T. J. J.
Comput. Assist. Tomogr. 1985, 9, 8.
4. Parhami, P.; Fung, B. M. J. Phys. Chem. 1983, 87, 1928.
5. Eidelberg, D.; Johnson, G.; Barnes, D.; Tofts, P. S.; Delpy, D.; Plummer, D.;
McDonald, W. I. Magn. Reson. Med. 1988, 6, 344.
6. Longmaid, H. E., 3rd; Adams, D. F.; Neirinckx, R. D.; Harrison, C. G.;
Brunner, P.; Seltzer, S. E.; Davis, M. A.; Neuringer, L.; Geyer, R. P. Invest.
Radiol. 1985, 20, 141.
7. Ahrens, E. T.; Flores, R.; Xu, H.; Morel, P. A. Nat. Biotechnol. 2005, 23,
983.
8. Srinivas, M.; Morel, P. A.; Ernst, L. A.; Laidlaw, D. H.; Ahrens, E. T. Magn.
Reson. Med. 2007, 58, 725.
9. Janjic, J. M.; Srinivas, M.; Kadayakkara, D. K. K.; Ahrens, E. T. J. Am.
Chem. Soc. 2008, 130, 2832.
10. Lanza, G. M.; Winter, P. M.; Neubauer, A. M.; Caruthers, S. D.; Hockett, F.
D.; Wickline, S. A. Curr. Top. Dev. Biol. 2005, 70, 57.
11. Lanza, G. M.; Yu, X.; Winter, P. M.; Abendschein, D. R.; Karukstis, K. K.;
Scott, M. J.; Chinen, L. K.; Fuhrhop, R. W.; Scherrer, D. E.; Wickline, S. A.
Circulation 2002, 106, 2842.
12. Morawski, A. M.; Winter, P. M.; Yu, X.; Fuhrhop, R. W.; Scott, M. J.;
Hockett, F.; Robertson, J. D.; Gaffney, P. J.; Lanza, G. M.; Wickline, S. A.
Magn. Reson. Med. 2004, 52, 1255.
13. Partlow, K. C.; Chen, J.; Brant, J. A.; Neubauer, A. M.; Meyerrose, T. E.;
Creer, M. H.; Nolta, J. A.; Caruthers, S. D.; Lanza, G. M.; Wickline, S. A.
FASEB J. 2007, 21, 1647.
14. Becker, M. L.; Remsen, E. E.; Wooley, K. L. J. Polym. Sci., Part A 2001,
39, 4152.
15. Du, W. J.; Xu, Z. Q.; Nystrom, A. M.; Zhang, K.; Leonard, J. R.; Wooley, K.
L. Bioconjugate Chem. 2008, 19, 2492.
16. Srinivas, M.; Heerschap, A.; Ahrens, E. T.; Figdor, C. G.; de Vries, I. J. M.
Trends Biotech. 2010, 28, 363.

471
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 17. Higuchi, M.; Iwata, N.; Matsuba, Y.; Sato, K.; Sasamoto, K.; Saido, T. C.
Nat. Neurosci. 2005, 8, 527.
18. Yu, J.-x.; Kodibagkar, V. D.; Cui, W.; Mason, R. P. Curr. Med. Chem. 2005,
12, 819.
19. Ruiz-Cabello, J.; Barnett, B. P.; Bottomley, P. A.; Bulte, J. W. M. NMR
Biomed. 2011, 24, 114.
20. Lai, J. T.; Filla, D.; Shea, R. Macromolecules 2002, 35, 6754.
21. Quemener, D.; Davis, T. P.; Barner-Kowollik, C.; Stenzel, M. H. Chem.
Comm. 2006, 5051.
22. Callaghan, P. Principles of Nuclear Magnetic Resonance Microscopy;
Clarendon Press: Oxford, 1994.
Downloaded by STANFORD UNIV GREEN LIBR on June 20, 2012 | http://pubs.acs.org

23. Hendrick, R. E. Magn. Reson. Imaging 1987, 5, 31.

24. Peng, H.; Blakey, I.; Dargaville, B.; Rasoul, F.; Rose, S.; Whittaker, A. K.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch028

Biomacromolecules 2009, 10, 374.

25. Nurmi, L.; Peng, H.; Seppala, J.; Haddleton, D. M.; Blakey, I.; Whittaker, A.
K. Polym. Chem. 2010, 1, 1039.
26. Kuebler, S. C.; Heuer, A.; Spiess, H. W. Macromolecules 1996, 29, 7089.
27. Kuebler, S. C.; Schaefer, D. J.; Boeffel, C.; Pawelzik, U.; Spiess, H. W.
Macromolecules 1997, 30, 6597.
28. Liu, B. L.; Kazlauciunas, A.; Guthrie, J. T.; Perrier, S. Macromolecules 2005,
38, 2131.
29. Thurecht, K. J.; Blakey, I.; Peng, H.; Squires, O.; Hsu, S.; Alexander, C.;
Whittaker, A. K. J. Am. Chem. Soc. 2010, 132, 5336.
30. Cory, A. H.; Owen, T. C.; Barltrop, J. A.; Cory, J. G. Cancer Commun. 1991,
3, 207.

472
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 29

Latex State and Solid-State NMR Spectroscopy

of Elastomers
Seiichi Kawahara*
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Department of Materials Science and Technology, Faculty of Engineering,

Nagaoka University of Technology, Nagaoka, Niigata 940-2188, Japan
*E-mail: [email protected]

Latex state NMR spectroscopy and field gradient fast magic

angle spinning solid-state NMR spectroscopy were utilized for
structural characterization of rubbery polymers and crosslinked
rubbery materials. High resolution NMR spectra for elastomers
were achieved either through Brownian motion of a latex
dispersion or fast magic angle spinning for a solid. Sequence
distribution of cis-1,4-, trans-1,4- and 1,2-butadiene units
was investigated for polybutadiene through latex state NMR
spectroscopy because the polybutadiene used in this work was
prepared by emulsion polymerization. Crosslinking junctions
of vulcanized natural rubber were analyzed by field gradient
fast magic angle spinning solid-state NMR spectroscopy. The
vulcanized natural rubber with quaternary carbons linking to
sulfur as the crosslinking junction was found to be superior in
mechanical properties compared to the rubber samples with
tertiary carbons linking to sulfur.

1. Introduction
Rubbery polymers and polymer gels hold a unique place in polymer science
because they exhibit both liquid and solid behavior in terms of molecular
motion. They are usually crosslinked in practical use to maintain their physical
dimension. This crosslinking makes it difficult to apply solution NMR for
structural characterization. Moreover, the molecular motions present in them also
pose difficulties for solid-state NMR spectroscopy of the crosslinked rubbers;
e.g., two-dimensional NMR measurements cannot be easily performed on them

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 in the solid state. This is explained to be due to very low concentration of the
crosslinking junctions and their unsatisfied molecular motions, compared with
repeating units. It is necessary to develop a special NMR technique for the
rubbery polymers and polymer gels in order to perform two-dimensional NMR
measurements for their structural characterization.
Two-dimensional NMR measurements for the rubbery polymers and polymer
gels may be performed by eliminating dipole-dipole interactions. However, in the
case of the crosslinked rubbers, which consist of the mobile main chain and the
constrained crosslinking junctions, motionary inhomogeneity prevent the efficient
elimination of the dipole-dipole interactions. One way to eliminate the interactions
is a superposition of another motions onto molecular motions. Thus, the rotational
movement of latex particles in a dispersion may be superposed onto molecular
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

motions. An alternative way is to carry out fast magic angle spinning (FMAS) in
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

solid-state NMR (1, 2); in this case, at least 20 kHz spinning is required to obtain
a high resolution NMR spectrum (3), and a large amount of sample is needed for
an analysis of crosslinking junctions. In order to carry out the analysis of the
crosslinking junctions as extremely small amount of building block through two-
dimensional NMR, it is necessary to develop latex state NMR spectroscopy (1, 2)
and use a solid-state NMR spectrometer equipped with a field gradient-fast magic
angle spinning (FG-FMAS) probe (4). Using the FG-FMAS probe, particulary,
we may achieve large increase in sensitivity for the measurement.
Thus far, only several preliminary studies of the latex state NMR spectroscopy
have been reported: an observation of hydrolysis on the surface of PMMA
dispersoid in latex by Tarcha (5) and determination of dried rubber content for
natural rubber latex by Ang (6) and Gambhir (7). In my opinion, latex state
NMR spectroscopy has not been studied systematically in the previous articles,
and the conditions for getting a high resolution spectrum via latex state NMR
spectroscopy have not been determined. Moreover, fast MAS (more than 20
kHz) has not been achieved for solid-state NMR measurement equipped with
FG-FMAS probe, even though some results of structural characterizations of soft
materials have been reported earlier (810).
In a series of earlier studies, we have conducted two-dimensional NMR
measurements of rubbery polymers and polymer gels through latex state NMR
spectroscopy, using a solid-state NMR spectrometer equipped with a FG-FMAS
probe. In this article, I shall attempt to summarize the work we have done so far.

2. Latex State NMR Spectroscopy

2.1. Effect of Crosslinking

Aliphatic carbon signals obtained for polybutadiene (PB) by latex state 13C-
NMR spectroscopy are shown in Figures 1 and 2, and they are compared with those
obtained for solution and solid samples, in which the solid-state measurement was
performed with gated high power decoupling at magic angle spnning of 5 kHz.
Characteristics of the PB latices are tabulated in Table 1.The chemical shifts of
these signals were shown in Table 2 together with the assignment reported thus far
476
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 (1113), where the number of the carbon atoms in 1,4 and 1,2 units of PB belongs
to IUPAC nomenclature.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 1. 13C-NMR spectra for PBD-S.

Figure 2. 13C-NMR spectra for PBD-G.

477
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 1. Glass transition temperature, melting temperature, gel content,
molecular weight, polydispersity, volume mean particle diameter of PBD
latex
Specimen Tg Tm gel content Mw Mw/Mn <D>1
/oC /oC /w/w% /10 6 /m
PBD-S -75.8 -6.7 0 3.2 4.9 0.111
PBD-G -76.0 80 0.106
1 <D> : volume mean particle diameter
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

Table 2. 13C-NMR chemical shifts for solution, latex and solid samples of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

PBD-S and PBD-G

Chemical Shift / ppm
Sato PBD-S PBD-G
et.
al.9
Sequence C Solution Latex Solid Solu- Latex Solid
tion
C-v 4 24.98- 24.93 25.06 24.95 24.93 25.00 24.85
25.10
C-1,4 4 27.42- 27.47 27.64 27.47 27.45 27.58 27.47
1,4-C 1 27.57
T-v 4 30.16 30.10 30.76 30.19 30.14 30.22 30.19
v-v-C (m) 1 31.60- 32.19 32.12
32.13
1,4-v-C 1 32.72 32.72 32.89 32.72 32.73 32.79 32.72
T-1,4 4
1,4-T 1
v-v-C (r) 1 33.35- 33.60 33.21
33.53
1,4-V-1,4 1 33.99- 34.05 34.17 34.06 34.05 34.13 33.98
34.16
1,4-V- 1 34.31
v(m)
1,4-V-v(r) 1 35.63- 35.74 35.90 36.00 35.72 35.84 35.67
36.00
v-v-T(m) 1 37.24- 37.60 37.62 37.58 37.53 37.71 37.40
37.48
1,4-v-T 1 38.18 38.13 38.39 38.26 38.17 38.30 38.26
Continued on next page.

478
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 2. (Continued). 13C-NMR chemical shifts for solution, latex and solid
samples of PBD-S and PBD-G
Chemical Shift / ppm
Sato PBD-S PBD-G
et.
al.9
Sequence C Solution Latex Solid Solu- Latex Solid
tion
v-V-v 2 38.57- 39.05 39.30 39.13 39.06 39.23
39.13
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

v-v-T(r) 1 38.96- 39.58 39.59 39.62 39.44 39.52 39.61

39.13
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

v-V 1 39.43- 40.60 40.75 40.59 40.68 40.75

41.72
1,4-V-v 2 40.55- 41.08 41.22 41.08 41.11 41.21 41.08
41.00
v-V-1,4 2 41.13- 41.57 41.70 41.56 41.59 41.66
41.73
1,4-V-1,4 2 43.47- 43.45 43.60 43.48 43.53 43.51 43.50
43.70
C, cis-1,4 unit; T, trans-1,4 unit; V, 1,2 unit; 1,4, cis- and trans-1,4 units m, meso; r,
racemic

As shown in Figure 1, for PB-S solution sample containing no gel fraction,

all of the expected signals appeared in the spectrum, but several signals were
invisible due to line broadening for latex and solid samples. The invisible signals
are explained by the constraint of molecular movement of the polymer in the solid
phase. The data demonstrate the important role that active molecular movement
of polymer plays in the latex state 13C-NMR spectroscopy. For latex, solution and
solid samples of PB-G, some diad and triad signals were invisible in the spectrum.
Since the gel content of PB-G was more than 80 w/w% while PB-S did not contain
any gels, the invisible signals in PB-G was due to inhomogeneity in the molecular
movement due to the presence of crosslinking junctions.
The values of T1 for most carbon atoms of latex sample were smaller than
those of the solution sample but were actually similar to those of the solid sample.
Thus, the movement of PB molecules in the latex dispersion is similar to that in
the solid sample. Since a value of glass transition temperature, Tg, for PB is about
76 oC, segmental motions are expected to occur at the observed temperature for
the polymers even in the solid state.

2.2. Effect of Dry Rubber Content (DRC)

The signal-to-noise (S/N) ratio and half-width are well-defined parameters,

which express the resolution of NMR spectrum, where high resolution are

479
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 characterized by high S/N ratio or narrow half-width. Thus, the resolution of
latex state 13C-NMR measurement was expressed via S/N ratio and half-width.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 3. Dependence of S/N ratio () and half-width () on dried rubber

content of PBD-G latex.

Figure 4. S/N ratio () and half-width () of PBD-G latex versus surfactant

concentration.

480
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table 3. S/N ratio and half-width of PBD-S and PBD-G
Experimental condition S/N ratio / 102 Half-width / Hz
PBD-S
Complete decoupling
Solution b 6.6 11
Latex c 2.2 10
Gated dcoupling a
Solution b 3.5 11
Latex c 3.0 11
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

Solid 2.1 32
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

PBD-G
Complete decoupling
Solution b 0.7 26
Latex c 1.0 11
Gated dcoupling a
Solution b 0.5 26
Latex c 0.7 12
Solid 0.7 63
aGated decoupling without nuclear Overhauser effect . b 10w/v% CDCl3 solution of
PBD. c Dried rubber content of the latex was 10%.

As shown in Figure 3, the S/N ratio for PB-G latex sample increased linearly
up to 60 w/w% in DRC, while the half-width increased abruptly at about 20 w/w%
in DRC. This implies that an appropriate DRC necessary for high resolution latex
state 13C-NMR measurement hovered around 10 w/w%. Yet, the S/N ratio and
half-width were independent of surfactant concentration as shown in Figure 4.
Thus, the surfactant concentration was kept at 1 w/v% for future experiments.
Table 3 shows the values of S/N ratio and half-width determined for the
signal at 32.7 ppm, for latex, solution and solid samples of PB-S and PB-G,
obtained under the experimental conditions of complete decoupling (COM) and
gated decoupling without nuclear Overhauser effect (NNE). In the absence of gel
fraction, the S/N ratio for PB-S solution sample, measured with a COM pulse
sequence, was higher than that measured with a NNE pulse sequence by a factor
of about 2. This is due to the influence of nuclear Overhauser effect, as reported
in the previous work (11). For PB-S latex sample, the S/N ratio estimated by
COM measurement was similar to that estimated by NNE measurement, being
comparable to the S/N ratio determined by the solid-state 13C-NMR measurement.
This may be due to either the absence of nuclear Overhause effect or other

481
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 unknown effects in the latex state 13C-NMR spectroscopy. The value of half-width
for PB-S latex sample was nearly equal to that of the solution sample, and smaller
than the solid sample by a factor of about three. These results demonstrate that
the resolution of 13C-NMR measurement for PB latex without the gel fraction was
identical to that for the solution sample, even though the latex is heterogeneous,
consisting of a polymer dispersion in water.
In the presence of the gel fraction, as shown in Figures 1 and 2, the S/N ratio
for PB-G swollen with CDCl3 was smaller than that for the latex sample by about
one-half under observation conditions of both COM and NNE, despite the fact
that the polymer content in solution was the same as that in the latex. Here, the
solution measurement of swollen PB-G containing about 80 w/w% gel fraction
was carried out for the sake of comparison with the latex state measurement. The
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

half-width for latex sample was about one-half of that for the swollen sample and
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

one-sixth of that for the solid sample. Several signals for the solution and solid
samples of PB-G were influenced by drifting baseline due to the crosslinks present
in the gel fraction, as shown in Figures 1 and 2. Thus, highly crosslinked rubber
samples may be quantitatively characterized by 13C-NMR spectroscopy, when they
are dispersed in water as small particles.

2.3. Effect of Particle Size

Deproteinized natural rubber (DPNR) latex was fractionated by centrifugation

into four fractions with narrow, Gaussian, and nearly unimodal distributions in
particle size. The volume mean particle diameter distribution of the four isolated
fractions of DPNR is shown in Figure 5. At least three fractions were completely
separated from each other, and two larger fractions were somewhat overlapped.

Figure 5. Volume weighted particle size distribution for DPNR fractionated by

centrifugation: (A) DPNR-1, (B) DPNR-2, (C) DPNR-3 and (D) DPNR-4.

482
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 6. Latex state 13C-NMR spectra for DPNR-3 obtained (A) with spinning
of the sample tube and (B) without spinning.

The latex state 13C-NMR measurements were made under two conditions: one
with the sample tube spinning at a rate of 15 Hz and the other without spinning.
A typical 13C-NMR spectrum for DPNR-3 latex with a volume mean particle
diameter of 0.901 m is shown in Figure 6. The spectrum with spinning was
distinguishable from that without spinning. The S/N ratio decreased to one-half
and half-width increased a little, as the sample tube was spun at 15 Hz. This is
explained to be due to segregation of the dispersed natural rubber particles (=0.91
g/cm3) onto the upper layer of the latex as a supernatant after spinning the NMR
tube. The concentration of natural rubber at the pulse-irradiated region may, thus,
be too low to obtain a sufficient S/N ratio. In Table 6, the values of S/N ratio and
half-width for the other latex fractions are also shown. For DPNR-1 and DPNR-2
latexes with the volume mean particle diameters of 0.12 and 0.36 m, respectively,
the S/N ratio and half-width were independent of sample spinning. For these
two fractions with the small particle diameter, the particles were homogeneously
dispersed in water after NMR measurement. In contrast, as for DPNR-4, no signals
were seen in the NMR spectrum when the sample tube was spun at 15 Hz. This
demonstrates that the latex state 13C-NMR measurement is significantly dependent
upon the homogeneity in the latex particle distribution.
The diffusion coefficient of the Brownian motion (14) was measured by light
scattering technique, since the diffusion coefficient is proportional to the rotational
motion of the particles, which may perhaps be related to the rotational correlation
time. We postulate that the molecular order rotational correlation time is interfered
by the rotational motion of the particles of about 1 m in the average diameter.
483
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The motions of the particles in the latex were expected to play an important role
in eliminating the effect of dipole-dipole interactions.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 7. Dependence of half-width on diffusion coefficient of Brownian

motion for methyl carbon (solid line) and methine carbon (dotted line) :
DPNR-1DPNR-4 latex at 50 oC (), DPNR1 at 3070 oC () and DPNR4 at
3070 oC ().

In order to confirm the relationship between Brownian motion and resolution

of 13C-NMR spectroscopy that is associated with dipole-dipole interactions, the
half width was plotted against diffusion coefficient of the Brownian motion. Figure
7 shows a typical plot of the half width of methyl signal at 24 ppm and methine at
124 ppm versus the diffusion coefficient. The half width for the latex decreased
with increasing diffusion coefficient and was close to that for the corresponding
solution. This is strong evidence showing that the Brownian motion of the latex
particles (or its related motions) eliminates the effect of dipole-dipole interactions
on the latex-state 13C-NMR measurement.

3. Solid-State NMR with FG-FMAS Probe

3.1. Effect of Crosslinking

Solid-state 13C-NMR spectrum for the vulcanized natural rubber, measured

with a FG-FMAS probe, is shown in Figure 8, together with a solution 13C-NMR
spectrum for unvulcanized natural rubber. Five major signals in the spectrum were
assigned to C nuclei of cis-1,4-isoprene units, according to the previous paper
(15). Half width of the signals and signal to noise (S/N) ratio of the solid-state
13C-NMR spectrum for the vulcanized natural rubber were quite similar to those

484
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 in the solution 13C-NMR spectrum for unvulcanized natural rubber. The narrow
half width and the ample S/N ratio of the solid-state 13C-NMR spectrum for the
vulcanized natural rubber may imply that high resolution was maintained for the
solid-state 13C-NMR spectroscopy, even after vulcanization, since heteronuclear
dipole-dipole interactions was eliminated by FMAS. This is distinguished from
low resolution solid-state NMR spectrum for the vulcanized natural rubber
reported by Klppel (16) and Kenig (1722).

3.2. Spectral Assignments

In Figure 8, small signals appeared at 40, 44, 50 and 58 ppm in the solid-state
13C-NMR spectrum for the vulcanized natural rubber, which were not shown in
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

the solution 13C-NMR spectrum for the unvulcanized natural rubber. To assign
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

the signals, we applied DEPT measurement at 45 o (DEPT45), 90 o (DEPT90)

and 135 o (DEPT135) pulses and APT measurement to the solid-state 13C-NMR
spectroscopy. Figure 9 shows DEPT45, DEPT90 and DEPT135 spectra for the
vulcanized natural rubber. The signals at 24, 26 and 32 ppm characteristic of
methyl, methylene and methine carbons of cis-1,4-isoprene unit were shown to be
up, up, and up in the DEPT45 spectrum and almost null in the DEPT90 spectrum.
On the other hand, in the DEPT135 spectrum, the signals were up, down, and
down. Thus, the pulse width determined for DEPT measurements was confirmed
to be correct.
The small signals at 40 and 44 ppm were shown to be up in the DEPT45
spectrum, null in the DEPT90 spectrum, and down in the DEPT135 spectrum;
hence, they were assigned to secondary carbons. The signal at 50.5 ppm was
assigned to quaternary carbon due to null signal in the spectra, while the signal
at 50 ppm was assigned to tertiary carbon due to up signals in the spectra. In
contrast, the signal at 58 ppm was assigned to tertiary and quaternary carbons due
to the very small up signals in the DEPT45, DEPT90 and DEPT135 spectra and
null signals; in fact, almost all signals disappeared. In Figure 9, the APT spectrum
is also shown for the vulcanized natural rubber. The APT spectrum showed up
signals at 40 and 44 ppm, up and down signals at 50 ppm, and up and down signals
at about 58 ppm. Thus, we assigned the signals at 40 and 44 ppm to the secondary
carbons and the signals at 50 and 58 ppm to the tertiary and quaternary carbons.
Figure 10 shows solid-state 1H-NMR spectrum for the vulcanized natural
rubber and solution 1H-NMR spectrum for the unvulcanized natural rubber. Major
signals at 1.7, 2.1 and 5.1 ppm in the spectra were assigned to methyl, methylene
and unsaturated methine protons of cis-1,4-isoprene units, respectively. Values of
half width and signal to noise (S/N) ratio of the signals in the solid-state 1H-NMR
spectrum for the vulcanized natural rubber were a little bit larger and smaller,
respectively, than the values of the half width and the S/N ratio of the signals in
the solution 1H-NMR spectrum for the unvulcanized natural rubber. For instance,
the value of half width of the signals in the solid-state spectrum was about 1.5
times as large as that in the solution spectrum. This may be explained to be due
to a reduced effect of the homonuclear dipole-dipole interactions in the solid-
state 1H-NMR spectrum; i.e., a major portion of the homonuclear dipole-dipole
interactions are eliminated by FMAS. (The homonuclear dipole-dipole interaction
485
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 may be completely eliminated by extremely fast MAS with smaller sample tube,
but this is not the focus of this work.) In Figure 10, small signals at 3.4 and 4.2 ppm
appeared in the solid-state 1H-NMR spectrum for the vulcanized natural rubber,
but not in the solution 1H-NMR spectrum for the unvulcanized natural rubber.
The signal at 3.4 ppm was assigned to aliphatic 1H linking to C-CH-Sx group
and the signal at 4.2 ppm to unsaturated aliphatic 1H linking to =C-CH-Sx group,
according to literatures (23), where x represents the number of S atoms.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 8. Typical 13C-NMR spectra: (A) solid-state 13C-NMR spectrum for the
vulcanized natural rubber, (B) solution-state 13C-NMR spectrum for unvulcanized
natural rubber. The solid-state 13C-NMR measurement was performed with a 4
mm FG-FMAS probe at 18 kHz in spinning rate. The solution state 13C-NMR
measurement was performed with a NM-40TH5AT/FG2SL probe at 12 Hz in
spinning rate.

3.3. 2D NMR Measurement

Figure 11 shows HSQC spectra obtained through solid-state NMR

spectroscopy equipped with the FG-FMAS probe. The signals at 1.7, 2.1 and 5.1
ppm in the 1H-NMR spectrum were correlated with the signals in the 13C-NMR
spectrum; e.g., the signal at 1.7 ppm was correlated with the signal at 23 ppm, the
signal at 2.1 ppm with the signals at 26 and 32 ppm, and the signal at 5.1 ppm
with the signal at 135 ppm. However, no heteronuclear correlation between the
small signals was detected in HSQC spectra due to inadequate accumulation time
for HSQC measurement.

486
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 9. Solid-state 13C-NMR spectra with pulse sequences of distortionless

enhancement by polarization transfer (DEPT) and attached proton test for the
vulcanized natural rubber: (A) DEPT45, (B) DEPT90, (C) DEPT135, (D) APT.

To detect heteronuclear correlation between the small signals, HMQC

measurement was performed in a selective region of chemical shift: 24 ppm
in 1H domain and 3570 ppm in 13C domain. Figure 12 shows HMQC spectra
obtained by the selective experiment. The 13C-signals at 40, 44, 50 and 58 ppm
were well correlated to the 1H-signals at 2.1, 1.7, 2.8 and >3.6 ppm. In the
previous works (15, 24), the signals at 40, 44 and 58 ppm of the vulcanized
liquid cis-1,4-polyisoprene as a model were assigned through solution NMR
spectroscopy with various pulse sequences, viz., DEPT, APT, HETCOR, HSQC
and HMBC. The signal at 40 ppm was assigned to C4 of trans-1,4-isoprene units,
obtained by isomerization of cis-1,4-isoprene units. In contrast, the signals at 44
ppm in the 13C-NMR spectrum were assigned to the secondary carbons adjacent
to carbons linking to S atoms. The signals at 58 ppm in 13C-NMR spectrum were
assigned to the tertiary and quaternary carbons linking to S atoms. In the present
work, the same results were obtained by the solid-state NMR spectroscopy
equipped with the FG-FMAS probe. Furthermore, the 13C-signals at 58 ppm
were correlated to the 1H-signals at 3.4 and 4.2 ppm, which were assigned to the
C-CH-Sx group and =C-CH-Sx group.

487
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 10. Typical 1H-NMR spectra: (A) solid-state 1H-NMR spectrum for
the vulcanized natural rubber, (B) solution-state 1H-NMR spectrum for the
unvulcanized natural rubber. The solid-state 1H-NMR measurement was
performed with a 4 mm FG-FMAS probe at 18 kHz in spinning rate. The solution
state 1H-NMR measurement was performed with a NM-40TH5AT/FG2SL probe
at 12Hz in spinning rate.

Figure 11. HSQC spectra for the vulcanized natural rubber, obtained through
solid-state NMR spectroscopy equipped with a 4 mm FG-FMAS probe at 18 kHz
in spinning rate for 1 h 2 min.

488
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 12. HMQC spectra for the vulcanized natural rubber, obtained through
solid-state NMR spectroscopy equipped with a 4 mm FG-FMAS probe at 18 kHz
in spinning rate for 71.5 h. HMQC measurement was performed in a selective
region of chemical shift: 24 ppm in 1H domain and 3570 ppm in 13C domain.

3.4. Application of FG-FMAS Solid-State NMR Spectroscopy

The crosslinking junctions of three rubbers (CV, EV and SemiEV crosslinked
rubbers) were analyzed by solid-state 1H-NMR and 13C-NMR spectroscopy. Table
4 shows crosslink density of these three rubbers. Values of crosslink density
of the crosslinked rubbers were similar, implying that the crosslink density is
adjustable, as long as we prepare the samples under the optimum condition for
crosslinking. The slight difference in the value of crosslink density may be due to
the comparatively short time for crosslinking and it is difficult to use exactly the
same optimum time for crosslinking.
Figure 13 shows solid-state 1H-NMR spectra for the three rubbers. Major
signals at 1.7, 2.1 and 5.1 ppm in the spectra were assigned to methyl, methylene
and unsaturated methine protons of cis-1,4-isoprene units, respectively. In Figure
13, small signals at 3.4 ppm and 4.2 ppm were found for the three crosslinked
rubbers. The signal at 3.4 ppm was assigned to aliphatic 1H linking to C-CH-Sx
group and the signal at 4.2 ppm to unsaturated aliphatic 1H linking to =C-CH-Sx
group, according to our work (4). The intensity of the aliphatic 1H linking to
C-CH-Sx group (3.4 ppm) and the unsaturated aliphatic 1H linking to =C-CH-Sx
group (4.2 ppm) was estimated from intensity ratio of the signals to the methyl
proton signal at 1.7 ppm. The estimated value of the intensity ratio is tabulated in
Table 5. It was found that the intensity of the aliphatic 1H linking to C-CH-Sx
group at 3.4 ppm was 0.02 % for CV, EV and SemiEV crosslinked rubbers, whereas
that of the unsaturated aliphatic 1H linking to =C-CH-Sx group at 4.2 ppm was
0.03 % for CV and SemiEV crosslinked rubbers and 0.02 % for EV crosslinked
rubber. From these results, it is concluded that the content of the small signals of
the aliphatic 1H linking to C-CH-Sx group and unsaturated aliphatic 1H linking to
489
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 =C-CH-Sx group for CV, EV and SemiEV crosslinked rubbers was almost similar
to each other.

Table 4. Vulcanization characteristics of HANR and DPNR compounds

Crosslink density (x10-4, mol/cm3)
Sample
CV system EV system SemiEV system
HANR 1.86 1.45 1.67
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Table 5. Intensity ratio of the signal at 3.4 ppm (aliphatic 1H linking to

C-CH-Sx group signal) and the signal at 4.2 ppm (unsaturated aliphatic 1H
linking to =C-CH-Sx group signal) to the signal at 1.7 ppm
Intensity of signals (%)
Crosslinked
3.4 ppm (aliphatic 1H
linking 4.2 ppm (unsaturated aliphatic 1H
rubber
to C-CH-Sx group) linking to =C-CH-Sx group)
CV 0.02 0.03
EV 0.02 0.02
SemiEV 0.02 0.03

Table 6. Intensity ratio of the signals at 40, 44 and 58 ppm to the signal at
24 ppm
Intensity of crosslinking junction signals (%)
Crosslinked rubber
40 ppm 44 ppm 58 ppm
CV 0.010 0.005 0.005
EV 0.010 0.000 0.000
SemiEV 0.009 0.001 0.001

490
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

Figure 13. Solid-state 1H-NMR spectra for the crosslinked natural rubbers
prepare by (A) CV, (B) EV and (C) SemiEV.

Figure 14. Solid-state 13C-NMR spectra for the crosslinked natural rubbers
prepared by (A) CV, (B) EV and (C) SemiEV.

491
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Solid-state 13C-NMR spectra for the same samples are shown in Figure 14.
Small signals at 40, 44 and 58 ppm appeared in the solid-state 13C-NMR spectra for
the CV and SemiEV crosslinked rubber, whereas only signal at 40 ppm appeared
in the solid-state 13C-NMR spectrum for the EV crosslinked rubber. The signals
at 40 and 44 ppm were assigned to C4 of trans-1,4-isoprene units (25, 26) and
secondary carbons adjacent to carbons linking to S atom, respectively, and the
signals at 58 ppm were assigned to the tertiary and quaternary carbons linking to S
atoms, according to our work (4). This may imply that the crosslinking junctions
of the CV and SemiEV crosslinked rubbers were not only secondary carbon but
also tertiary and quaternary carbons whereas that of the EV crosslinked rubber may
contain small amount of quaternary carbon. The intensity of the small signals at
40 ppm, 44 ppm, and 58 ppm was estimated from intensity ratio of the signals to
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

the methyl carbon signal at 24 ppm. The estimated values of the intensity ratio of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

the small signals at 40, 44 and 58 ppm to the methyl carbon signal at 24 ppm were
tabulated in Table 6. The intensity of the signal at 40 ppm was similar to each other
(0.010 %) for CV, EV and SemiEV crosslinked rubbers. This indicates that cis-
trans isomerization occurs similarly for CV, EV and SemiEV crosslinked rubbers
in spite of differences in formulation. In contrast, the intensity of the signals at 44
ppm and 58 ppm in the solid-state 13C-NMR spectrum decreased in the order of
CV, SemiEV and EV crosslinked rubbers. The intensity of the secondary carbon
(44 ppm), the tertiary and quaternary carbon (58 ppm) for CV crosslinked rubber
was the highest among the three. They decreased for SemiEV crosslinked rubber
and disappeared for EV crosslinked rubber. This implies that the amount of the
quaternary carbon of the CV crosslinked rubber is larger than that of the SemiEV
crosslinked rubber. The EV crosslinked rubber may contain a small amount of
quaternary carbon as a crosslinking junction.

Figure 15. Stress-strain curves for CV, EV and Semi-EV crosslinked rubbers.

492
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 The stress-strain curves for the CV, EV and SemiEV crosslinked rubbers
are shown in Figure 15. Modulus at 100% strain of the CV crosslinked rubber
is almost similar with that of EV and SemiEV crosslinked rubbers. This is
consistent with the fact that the values of the crosslink density of the CV, EV
and SemiEV crosslinked rubbers were similar to each other, as shown in Table
4. In contrast, tensile strength of the three crosslinked rubbers was different; that
is, the value of stress at break of the CV crosslinked rubber (22.5 MPa) was the
highest among the three crosslinked rubbers and was reduced for the SemiEV
(17.0 MPa) and EV (9.2 MPa), in that order. In the previous work (27), Suchiva
and co-workers reported that better mechanical properties of the CV crosslinked
rubber was due to higher concentrations of the polysulfidic crosslinks compared
to those of the EV crosslinked rubber. Moreover, the superior mechanical
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

properties of CV crosslinking system may be explained to be due to differences

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

in the vulcanization mechanism of the crosslinking systems. It may be noted that

the conventional crosslinking system produces a homogeneous network. Yet,
an efficient crosslinking system results in significant polymerization of double
bonds of adjacent chains leading to a network of unevenly distributed crosslinks
and in impediment to NR crystallization and stress concentration that anticipates
compound failure (28). In the present work, it was found that the excellent
mechanical properties of the CV crosslinked rubber could be attributed to the
quaternary carbon of crosslinking junctions.

Acknowledgments
This work was supported in part by a Grant-in-Aid (21655080) for
Challenging Exploratory Research and Grant-in-Aid (22350100) for Scientific
Research (B) from Japan Society for the Promotion of Science and JST-JICA
SATREPS.

References
1. Kawahara, S.; Bushimata, S.; Sugiyama, T.; Hashimoto, C.; Tanaka, Y.
Rubber Chem. Technol. 1999, 72, 844853.
2. Kawahara, S.; Washino, K.; Morita, T.; Tanaka, Y.; Isono, Y. Rubber Chem.
Technol. 2001, 74, 295302.
3. Duer, M. J. Introduction to Solid-State NMR Spectroscopy; Blackwell
Publishing: Cambridge, MA, 2004.
4. Kawahara, S.; Chaikumpollert, O.; Sakurai, S.; Yamamoto, Y.; Akabori, K.
Polymer 2009, 50, 16261631.
5. Fitch, R. M.; Gajria, C.; Tarcha, P. J. J. Colloid Interface Sci. 1979, 71,
108109.
6. Ang, T. T.; Roberts, J. D. Plast. Rubber: Mater. Appl. 1979, August 4,
138139.
7. Gambhir, P. N.; Joshi, D. K.; Tiwari, P. N.; Mani, J. J. Nat. Rubber Res.
1993, 8, 208212.
493
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 8. Romanova, E. E.; Grinberg, F.; Pampel, A.; Kaerger, J.; Freude, D. J. Magn.
Reson. 2009, 196, 110114.
9. Viel, S.; Ziarelli, F.; Pages, G.; Carrara, C.; Caldarelli, S. J. Magn. Reson.
2008, 190, 113123.
10. Pampel, A.; Fernandez, M.; Freude, D.; Kaerger, J. Chem. Phys. Lett. 2005,
407, 5357.
11. Sato, H.; Takebayashi, K.; Tanaka, Y. Macromolecules 1987, 20, 24182423.
12. Bywater, S. Polym. Commun. 1983, 24, 203205.
13. Tanaka, Y.; Takeuchi, Y.; Kobayashi, M.; Tadokoro, H. J. Polym. Sci. 1971,
A-2 (9), 4357.
14. Einstein, A. Ann. Phys. 1905, 17, 549560.
15. Kawahara, S.; Jinta, U.; Sakai, J.; Yamamoto, Y.; Isono, Y. Rubber Chem.
Downloaded by UNIV OF GUELPH LIBRARY on June 20, 2012 | http://pubs.acs.org

Technol. 2007, 80, 751761.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch029

16. Klppel, M.; Menge, H.; Schmidt, H.; Schneider, H.; Shuster, R. H.
Macromolecules 2001, 34, 81078116.
17. Zaper, A. M.; Kenig, J. L. Rubber Chem. Technol. 1987, 60, 252277.
18. Mori, M.; Kenig, J. L. Rubber Chem. Technol. 1995, 68, 551562.
19. Mori, M.; Kenig, J. L. Rubber Chem. Technol. 1997, 70, 671680.
20. Mori, M. Rubber Chem. Technol. 2003, 76, 12591275.
21. Mori, M.; Kenigh, J. L. J. Appl. Polym. Sci. 1998, 70, 13911399.
22. Patterson, D. J.; Kenigh, J. L.; Shelton, J. R. Rubber Chem. Technol. 1983,
56, 971994.
23. Frust, A.; Pretsch, E. Anal. Chem. Acta. 1990, 229, 1725.
24. Ukawa, J.; Kawahara, S.; Sakai, J. J. Polym. Sci., Part B: Polym. Phys.
2007, 45, 10031009.
25. Tanaka, Y. Rubber Chem. Technol. 1991, 64, 325385.
26. Buzare, J. Y.; Silly, G.; Emery, J.; Boccaccio, G.; Rouault, E. Eur. Polym. J.
2001, 37, 8591.
27. Suchiva, K.; Kowitterawut, T.; Srichantamit, L. J. Appl. Polym. Sci. 2000,
78, 14951504.
28. Gonzalez, L.; Rodriguez, A.; Valentin, J. L.; Marcos-Fernandez, A.;
Posadas, P. Kautsch. Gummi Kunstst. 2005, 67, 638643.

494
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 30

Proton MAS NMR Analysis of Phenyl Silane

Functionalized Silica Nanoparticles
Khalid A. M. Thakur,*,1 Mark McCormick,1b Wendy L. Thompson,2
Chuntao Cao,2 and William J. Schultz2
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

13M Corporate Research Analytical Laboratory, 6801 River Place Blvd,

142-4N-01, Austin, TX 78726
1b3M Corporate Research Analytical Laboratory, 3M Center, 201-BS-08,

St. Paul, MN 55144

23M Corporate Research Materials Laboratory, 3M Center, 201-1W-28,

St. Paul, MN 55144

*E-mail: [email protected]

High temperature magic-angle-spinning (MAS) 1H NMR

was used to identify organic components on the surface of
functionalized silica nanoparticles. The high temperature
data acquisition, by preferential narrowing of resonances,
enabled differentiation between species that were adsorbed or
weakly bound versus those that were covalently bound to the
surface. MAS NMR of a sample as a slurry with appropriate
solvent provided information similar to that obtained at higher
temperatures. With increasing temperature, gradual separation
and up-field shift of silanol proton and water resonances at
different rates was observed, possibly a result of diminished
hydrogen bonding at increasingly higher temperatures.

Introduction
Functionalized or surface modified silica nanoparticles are used in numerous
applications including biomedicine, biotechnology, foam generation and polymer
modification (112). They are prepared by reacting the silanol (-Si-OH)
groups that are on silica nanoparticle surface with an appropriately reactive
organic molecule. The reactive organic molecules most commonly used for
functionalization have the following general formula RSi(-OR)3, where R may

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 be methyl or ethyl group, and R has the desired terminal functional group. The
wish list for characterization of the functionalized nanoparticles includes: (1)
Extent of reaction of the surface silanols; (2) Confirmation of desired functional
group on molecules that are covalently bound to the surface; (3) Identification of
any self-condensed oligomeric silane; (4) Identification of inaccessible cavities in
the silica particles; and (5) Identification of adsorbed water or solvent molecules
that may be present after the drying process. FT-IR is the most common
spectroscopic analytical tool used for the molecular structure characterization of
functionalized or modified silica particles (1317). Solution NMR, though well
known for molecular structure characterization, has found limited applications in
study of functionalized nanoparticles. There may be a number of possible reasons
for it including poor solubility and broad NMR resonance linewidths.
A variety of solid-state NMR techniques have been used for analysis of
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

silica particles including 1H MAS 1D/2D and 29Si CP-MAS NMR (1830). In
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

functionalized nanoparticles, the surface chemistry is of most interest. Because

of the surface area to volume ratio, the NMR sensitivity is inversely proportional
to the silica particle size. 1H MAS NMR spectra are significantly more sensitive
than 13C or 29Si CP-MAS NMR. However, the normal single pulse 1H NMR
spectra of rigid materials acquired at nominal MAS speeds are usually too broad
to be useful. Fast spinning (>15 kHz) 1H MAS NMR has been shown to improve
the resolution of the resonances and enable analysis of the silica particle surfaces
(18, 19).
Here, we describe some of the capabilities of high temperature magic-angle-
spinning (MAS) single pulse 1H NMR for the characterization of functionalized
nanoparticles. MAS NMR as a slurry after addition of a small fraction of solvent
also provided similar information. Interpretations were made based on changes in
proton NMR resonance linewidths as a function of temperature.

Experimental Section
Materials and Measurements

MAS NMR spectra were acquired by spinning the sample in Chemagnetics

probes at speeds of up to 23 kHz in a 3.2 mm rotor. Data were acquired
at temperatures between 22 C and 250 C on a Varian INOVA 400 MHz
NMR spectrometer. The temperature values reported were not corrected for
possible frictional heating. Custom made fluoropolymer (PTFE) end-caps for
the MAS rotors were utilized to enable high temperature data acquisition and
reduce background resonance. Unless specifically mentioned, pulse sequences
employing single pulse Bloch decay measurements were used. Most of the 1H
MAS NMR spectra were acquired with less than a 25 tip angle and 64 scans
were averaged with recycle time of 3 seconds. The 1H NMR chemical shift was
externally calibrated using methyl resonance of cured poly(dimethyl siloxane)
gel at 0.2 ppm. MAS NMR spectra of AD and HMB were collected on a Varian
NMRS 400 MHz NMR Spectrometer equipped with a Varian HXY 3.2 mm probe
at speeds of 22 kHz.
496
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Materials: Hexamethylbenzene (HMB) and Adamantane (AD) were
purchased from Aldrich and Avocado respectively and used without further
purification. Silica nanoparticles with nominal 20 nm particle size distribution
were obtained from Nalco as a 40 weight% aqueous colloidal silica solution (Nalco
2327). Trimethoxy phenyl silane (TPS) and 1,1,1,3,3,3-Hexamethyldisilazane
(HDMS) were purchased from Aldrich.
The procedure for surface functionalization has been described in patent US
5,648,407 (31). All samples were dried for the analysis.
Sample I: 24.8 mmol of TPS in 150g of methoxyl-2-propanol was added
dropwise to 100g of Nalco 2327 while stirring. After stirring for another 30
minutes, the mixture was kept in an oven at 90 C for 24 hours. This was then
air-dried in a heated Aluminum pan.
Sample II: The dried Sample I solid was re-dispersed in 75g of acetone. 5g
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

of HDMS was added, and the mixture was stirred for four hours. The dispersion
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

was dried at 100 C for 30 minutes in an oven.

Sample III: 24.8 mmol of TPS in 150g of Diglyme was added dropwise
to 100g of Nalco 2327 while stirring. After stirring for another 30 minutes,
the mixture was kept in an oven at 90 C for 24 hours. Upon cooling to room
temperature, an additional 100g of Diglyme was added and the mixture was put
on a Buchi rota-vapor to remove water. Subsequently, 10g of HDMS was added,
and the mixture was stirred for four hours. The solution was then dried at 100 C
for 30 minutes in an oven.
In Sample III the two functionalized steps in Sample I and Sample II were
combined in to a pseudo one step process.

Overview of Resonance Linewidth Broadening Mechanisms in NMR

In order to facilitate discussions in subsequent sections, this sub-section
attempts to provide a non-mathematical description of mechanisms that govern
linewidths of NMR resonances and related key concepts and commonly used
acronyms.
Chemical shifts of resonances () obtained in dilute solution NMR
spectra have been routinely used for molecular structure identification and
characterization of organic solutes (3235). The narrow resonances in solution
at their isotropic chemical shifts are a result of (1) molecular motion averaging
orientation effects with respect to the magnetic field direction; and (2) isolation
of the solutes in deuterated (or non-protonated) solvent. Each resonance peak is
often representative of a specific three dimensional chemical structure vibrating
and rotating in isolation. Any change in the molecular structure or its electronic
environment shifts the NMR resonance, the magnitude of which is typically
related to the distance of the change from the observed nuclei.
During NMR measurements of rigid and semi-rigid materials, the rotational
and translational motions are not rapid enough to average out the orientation
dependent effects leading to broader resonances than in dilute solution. Typically,
all possible orientations of molecules are present in a randomly distributed
isotropic sampling. The orientation dependent shielding from electronic
environment results in a range of observed chemical shift that is referred to as the
497
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chemical Shift Anisotropy (CSA). Nuclei spins also have orientation dependent
dipolar interactions that can lead to resonance broadening up to several kHz.
In isotropic samples, solid-state NMR spectrometers use magic-angle-spinning
(MAS) to mechanically average CSA and dipolar broadening, leading to narrower
resonances.
Anisotropic properties, however, are not completely averaged by MAS.
Consider two contrasting examples: (1) Hexamethylbenzene (HMB) that
has significant orientation dependent shielding and (2) Adamantane (AD)
which has a more symmetric environment. In HMB powder, the asymmetric
shielding from each crystallite will also interact with asymmetric shielding from
randomly oriented HMB crystallites in the environment. Even in the first shell
of nearest-neighbor HMB crystallites, there are infinite numbers of possible
combinations of orientations, each with a distinct net shielding effect on the
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

central HMB. Furthermore, as the sample rotates, the shielding effect of the
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

neighboring crystallites changes due to its dependence on relative orientation of

static magnetic field. Such Anisotropic Bulk Magnetic Susceptibility (ABMS)
broadening effect is significantly smaller than the CSA, but is not averaged by
MAS, and manifests as residual broadening of the resonance (3641). This
effect is demonstrated in the Figures 1(a) and 1(b) for 1H decoupled 13C 22
kHz MAS NMR of AD and HMB, respectively. The AD resonances had
Full-Width-at-Half-Maximum (FWHM) linewidth of 8 Hz, while HMB Methyl
resonance had FWHM of 100 Hz. Furthermore, in a physical mixture of HMB and
AD, the AD crystallites experience the ABMS broadening from the asymmetric
environment of HMB crystallites around it. This effect can be observed in Figure
1(c). In this case the HMB Methyl groups still have a FWHM of about 100 Hz,
but the AD peaks broadened to about 80 Hz at FWHM.

Figure 1. 13C Direct Polarization proton decoupled 22 kHz MAS NMR Spectra of
(a) Adamantane powder; (b) Hexamethylbenzene powder (Methyl resonance);
(c) Mixture of AD and HMB.
498
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Spin-diffusion is another source of residual broadening, especially for
abundant nuclei such as 1H and 19F whose dipolar broadening has been averaged
by MAS in an isotropic sampling (18, 4246). In Fourier Transform NMR
(FT-NMR), a Radio-Frequency pulse generates a coherence of nuclei spins that
oscillate in phase at their characteristic Larmor frequency, and are recorded as
Free Induction Decay (FID). Fourier transforms of the FID results in the normally
reported NMR spectrum. Longer phase coherence translates to narrower peaks
in the NMR spectrum. Spin diffusion or adiabatic spin exchange among dipolar
coupled nuclei in close proximity scrambles the encoded phase information
and hence destroys the coherence, leading to broader NMR resonances. This
broadening effect can be reduced by increased motion such that neighboring
interacting/coupled nuclei positions fluctuate more rapidly and to a larger degree.
The higher the frequency of fluctuations, the lower is the efficiency of spin
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

diffusion, and narrower are the observed 1H NMR resonances (4245). Ultra fast
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

MAS speeds also reduce spin-diffusion and the related spectral broadening (18,
4247). However, obtaining 1H MAS spectra of non-spin-isolated molecules with
resonance linewidths comparable to those found in solution NMR is difficult,
except in some favorable cases where ABMS broadening is negligible (4245).

Results and Discussion

1HMAS NMR spectrum of (I) acquired at 22 kHz spinning speed and 200
C temperature is shown in Figure 2. Broad phenyl silane resonances between
6.5-8.2 ppm, narrow residual 2-methoxy-1-propanol solvent resonances at 1.1/
3.3/3.4/3.9 ppm, water resonance at 5 ppm, and silanol resonance at 3 ppm
were observed in this MAS NMR spectrum (vide infra).

Figure 2. Proton 22 kHz MAS 200 C NMR spectrum of (I).

499
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Figure 3 shows a progression of proton NMR spectra at increasing MAS
speeds for (I). At lower MAS speeds, significant fractions of the intensity of the
functionalized phenyl silane and silanol resonances were spread in to its spinning
side bands, primarily due to 1H-1H dipolar broadening. Resonances from these
molecular species covalently bound to the silica surface were better represented
in the center bands of spectrum only at higher spinning speeds. The resonance
intensity of the solvent adsorbed on the silica surface did not appear to spread
into the spinning side bands at nominal speeds (~ 8 kHz), suggesting motional
averaging and mobility of the adsorbed species at room temperature.
Figure 4 shows a progression of spectra of sample (I) acquired at increasing
temperatures and 22 kHz MAS speed. At ambient temperature (24 C),
the fast spinning proton NMR resonance linewidths of all molecules at the
surface whether covalently bound, hydrogen bonded, or adsorbed, were
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

comparable. At increasingly higher temperatures, the resonances of adsorbed

Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

1-methoxy-2-propanol narrowed significantly. In comparison, the line width of

phenyl silane resonances changed marginally with increase in temperature.

Figure 3. 1H MAS 24 C NMR spectra of (I) at MAS speeds of (a) 8 kHz, (b) 13
kHz, and (c) 22 kHz. Note absence of spinning side bands from the narrower
resonances representing weakly bound species. A broad underlying probe
background signal was subjectively removed as part of baseline correction.

At higher temperatures, weakly attached or adsorbed species are expected to

hop sites and reoriented frequently, possibly by being in vapor phase for increasing
durations, and resulting in isotropic averaging of its immediate environment and
reduction in spin diffusion. In comparison, the strongly attached covalently bound
species would have fewer degrees of freedom and minimal change in environment
with increasing temperature since the silica particles would not be expected to be
mobile. The observed marginal variance in phenyl silane resonance linewidth with
500
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 MAS speed and temperature could be due to a combination of ABMS, molecular
structure variations, and spin diffusion. However, no further experiments were
carried out in order to explore the contributions of each to the broadening.
The resonance around 6 ppm in 24 C spectrum was assigned in other reports
to combination of silanol and water (14, 18). In the 100 C spectrum, part of
its intensity appears as a smaller upfield resonance around 4.5 ppm. Hydrogen
bonded silanol has been reported at this proton chemical shift (14). The 150 C
spectrum indicated a more intense resonance shifted upfield to 4 ppm. With
increasing temperature, this resonance shifted further upfield reaching 2.5 ppm
in the 225 C spectrum. The chemical shift of isolated silanol on silica in the
limit of no hydrogen bonding has been reported at 1.8-2.1 ppm (14, 18). Hence,
the upfield shifting resonance can be assigned to unreacted silanol on the silica
nanoparticle surface whose silanol-silanol and silanol-water hydrogen bonding
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

reduces with increasing temperature.

Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

Figure 4. 1H 22 kHz MAS NMR spectra of (I) at increasing temperatures up to

225 C. Note the shift of resonances marked by arrows.
The other part of the 6 ppm resonance in the 24 C spectrum shifted up
field at a slower rate to 4.9 ppm in the 225 C spectrum. This behavior is
consistent with that of water whose hydrogen bonding interactions are reduced.
However, the broad width of the water resonance compared to that of the methoxy
propanol solvent suggests that the water molecules do not experience motional
averaging similar to that done by solvent at increasing temperature. The binding
strength of water hydrogen bonded to surface silanols is expected to be stronger
compared to the weak surface interactions of adsorbed solvent, but if the water
were to hop sites increasingly rapidly or be in vapor phase for longer durations at
temperatures above 100 C, the resulting isotropic environmental averaging would
have resulted in narrower resonances with increasing temperature. As explored
later, this resonance represents water trapped within cavities of the nanoparticles
that is hydrogen bonded with other water molecules and silanols in the cavities.
The identity of the resonance assigned to residual silanols on the
particle surfaces was confirmed by additional functionalization reactions.
501
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 1,1,1,3,3,3-Hexamethyldisilazane (HMDS) reacts aggressively with silanol
groups to form M siloxane groups, viz. O-Si(CH3)3 (14, 15, 48). Silanol
groups on the surface that could not react with the trimethoxy phenyl silane due
to conformational incompatibilities were expected to react with HMDS, even
though at least one report suggests that presence of hydrogen bonding among the
silanol groups can reduce the reactivity (14).
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

Figure 5. (a) 1H 22kHz MAS 150 C NMR of (I). Narrow resonances are residual
solvent; (b) 1H 22 kHz MAS 150 C NMR of (II). Resonance at 2 ppm is
residual acetone; (c) 29Si CP-MAS NMR of (I); (d) 29Si CP-MAS NMR of (II).
The arrows mark silanol resonances that react with HMDS.

In Figure 5, the 1H 22 kHz MAS 150 C NMR and 29Si CP-MAS 24 C NMR
spectra of (I) and (II) are compared. Upon HMDS treatment in (II) shown in
Figure 5(b), compared to (I) shown in Figure 5(a), a reduction in intensity of the
resonance assigned to silanol (1H 3.6 ppm marked with arrow) was detected
along with appearance of a broad M siloxane group resonance. In Figures 5(a)
and 5(b) the narrow resonances represent methoxy propanol in (I) and acetone in
(II), respectively. Comparison of 29Si CP-MAS NMR spectra of (I) and (II) shown
in Figures 5(c) and 5(d), respectively, confirm that the silanols at the silica surface
( -100 ppm) and on the phenyl silane ( -72 ppm) , both marked by arrows in the
figure, had reacted while M groups had formed ( ~ 10 ppm). Due to the long
data acquisition time requirements of the 29Si CP-MAS NMR spectra, no attempt
was made to acquire multiple spectra with varying cross-polarization contact times
that could enable a quantitative comparison of integral intensities from the various
29Si resonances. The water resonance intensity in proton NMR did not appear to

significantly reduce upon HMDS functionalization. The small fraction of silica

502
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 silanols that remained unreacted in (II) are most likely in cavities of the silica
particles along with water, and are probably not accessible to the solvents used in
the functionalization reactions. In Figure 5(b), the width of the methyl siloxane
group resonance is likely a result of ABMS broadening since the methyl groups
are expected to be freely rotating at this temperature (38, 40).
1H MAS NMR spectra of (I) and (II) acquired at lower MAS speeds after

addition of a small fraction of tetrachloroethane-d2 (TCE-d2) are shown in Figures

6(a) and 6(b), respectively. Solvent was added in order to increase the mobility of
adsorbed species. In Figure 6(b), the width and spinning sideband intensity of the
M siloxane resonance was comparable to that of the phenyl resonance, and hence
is consistent with the M groups being covalently attached to the silica surface.
The silanol proton resonances on the other hand (marked by arrow in Figure 6),
though attached to the silica surface did not indicate similar spinning side-band
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

intensity distribution. We speculate that the combination of broad resonance and

Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

lack of spinning side-band intensity, compared to the methyl M groups on the

surface, may indicate rapid proton hopping and exchange among the silanols at
the silica surface.

Figure 6. (a) Proton 9 kHz MAS 100 C NMR of (I) with addition of deuterated
tetrachloro ethane (TCE-d2); (b) Proton 9 kHz MAS 100 C NMR of (II) with
addition of TCE-d2. The arrows mark silanol resonances that do not have
spinning side band intensity.

In order to further probe whether the water resonances observed in the 1H

MAS NMR spectra were from the molecules at the surface or in cavities within
the nanoparticles, NMR spectra were acquired after dispersion in a solvent.
Depending on the size of the particles, type, and coverage of functional groups at
503
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 the surface, the nanoparticles can form non-precipitating dispersions. Presence of
solvable pendant oligomeric chains from the nanoparticles allows the relatively
large (on average) 20 nm particles to form clear dispersed solutions (in absence
of charged species on its surfaces to prevent aggregations). In its solution, the
water on the surface of the nanoparticles was expected to be solvated by the
solvent (THF-d8) and appear as resonance at its characteristic solvated chemical
shift (1H 2.5 ppm) in the NMR spectra. Water within inaccessible cavities of
nanoparticles would not be solvated by the solvent and should appear as resonance
around 1H ppm. Presence of non-solvated ( 5.3 ppm) water resonance in the 1H
NMR of (III) in THF-d8 shown in Figure 7 confirms that the water observed in the
1H MAS NMR spectra as a resonance around 5 ppm is trapped within cavities

of the nano-particles and is not accessible to the solvent. The 1H 1.7 and 3.6
ppm are residual THF-d8 solvent resonances while the 1H 2.6 ppm resonance is
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

water solvated by THF. The resonances between 3.3-3.6 ppm represent diglyme
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

that was present on nanoparticle surface.

Figure 7. 1H NMR of (III) in THF-d8 solution.

In Figure 7, the phenyl silane and methyl M siloxane resonances have

varying linewidths from the two split narrow ones to the broad underlying
resonance. As mentioned earlier, long molecular chain extensions from the
surface, e.g. self-condensed silanes, are solvated and mobile leading to narrower
resonances. These pendant chain extensions can also be identified in their 1H
MAS NMR spectra at high temperature and fast MAS as narrower resonances, as
shown in Figure 8(a). Here again, the narrow resonances between 3.3-3.6 riding
on top of a broad silanol resonance are from residual diglyme. Other researchers
have also used double-quantum 1H MAS NMR techniques to probe the surface
interactions and pendant oligomers (49).
Another method to distinguish the mobile pendant phenyl silane groups from
those closer to the surface is by MAS 1H NMR analysis of the functionalized
nanoparticles as a slurry. The slurry can be formed by adding only a small
fraction of solvent and mixing well. The solvent molecules solvate the pendant
504
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

Figure 8. (a) 1H 17 kHz MAS 150 C NMR spectrum of (III) as dried powder. (b)
1H 8.7 kHz MAS 50 C NMR spectrum of (III) as slurry in THF-d8.

and adsorbed components more than those covalently bound to the surface,
mimicking the influence of high temperature. As a result, similar mobility based
differentiation can be obtained at a lower temperature and lower MAS speeds.
Figure 8(b) shows the 1H 8.7 kHz MAS NMR of the same sample as 8(a), but
in a slurry in THF-d8 at 50 C. Here again, the narrow phenyl silane and methyl
siloxane resonances from more mobile fractions are riding on top of broad
resonances representing rigid groups on the silica surface. The narrow 2.5 ppm
is solvated water in THF while the 5 ppm resonance is water in inaccessible
cavities of the nanoparticles. The rest of the narrow resonances represent THF
and diglyme. 1H MAS spectra in Figure 6 were also acquired as a slurry and
indicated presence of non-solvated water resonances. It may be noted that due
to the lower MAS speeds used with the slurry, the relative intensity of the rigid
components in the central band is often lower than expected.
Figures 7 and 8 demonstrate that depending on the solubility of the
functionalized nanoparticles, a number of methods can be used to obtain
information about binding strengths of the various species at the surface.

Conclusion
Single pulse high temperature 1H MAS NMR was used to characterize
functionalized nanoparticles. The high temperature helped accentuate the
505
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 difference in resonance linewidths of the various species based on mobility and
surface binding strength. In these nanoparticles based on colloidal silica, water
molecules trapped in inaccessible cavities were identified. Gradual reduction
of hydrogen bonding among silanol groups on silica surface with increasing
temperature and rapid exchange of the silanol protons was inferred from the
NMR data. Oligomeric pendant chains attached to the nanoparticles were
readily distinguished from those covalently attached directly to the surface. For
nanoparticles that are soluble in an appropriate solvent, a number of alternative
methods can also be used.

Acknowledgments
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

Strong support from Dr. Dale Perry and Dr. Rebecca Dittmar at 3M enabled
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

a significant portion of this work.

References
1. Giri, S.; Trewyn, B. G.; Lin, V. S. Y. Nanomedicine 2007, 2, 99111.
2. Du, M.; Zheng, Y. Polym. Compos. 2007, 28, 198207.
3. Trewyn, B. G.; Slowing, I. I.; Giri, S.; Chen, H.-T.; Lin, V. S.-Y. Acc. Chem.
Res. 2007, 40, 846853.
4. Schulz, H.; Pratsinis, S. E.; Regger, H.; Zimmermann, J.; Klapdohr, S.;
Salz, U. Colloids Surf., A 2008, 315, 7988.
5. Roy, I.; Ohulchanskyy, T. Y.; Bharali, D. J.; Pudavar, H. E.; Mistretta, R. A.;
Kaur, N.; Prasad, P. N. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 279284.
6. Blute, I.; Pugh, R. J.; van de Pas, J.; Callaghan, I. J. Colloid Interface Sci.
2007, 313, 645655.
7. Ahn, S. H.; Kim, S. H.; Lee, S. G. J. Appl. Poly. Sci. 2004, 94, 812818.
8. Hong, R. Y.; Fu, H. P.; Zhang, Y. J.; Liu, L.; Wang, J.; Li, H. Z.; Zheng, Y. J.
Appl. Poly. Sci. 2007, 105, 21762184.
9. Xu, X.; Li, B.; Lu, H.; Zhang, Z.; Wang, H. J. Appl. Poly. Sci. 2008, 107,
20072014.
10. Xu, X.; Li, B.; Lu, H.; Zhang, Z.; Wang, H. Appl. Surface Sci. 2007, 254,
14561462.
11. Giraldo, L. F.; Echeverri, M.; Lpez, B. L. Macromol. Symp. 2007, 258,
119128.
12. Song, R.; Yang, D.; He, L. J. Mater. Sci. 2007, 42, 84088417.
13. Almeida, R. M.; Pantano, C. G. J. Appl. Phys. 1990, 68, 42254232.
14. Haukka, S.; Root, A. J. Phys. Chem. 1994, 98, 16951703.
15. Kytokivi, A.; Haukka, S. J. Phys. Chem. B 1997, 101, 1036510372.
16. Bruni, S.; Cariati, F.; Casu, M.; Lai, A; Musinu, A.; Piccaluga, G.; Solinas, S.
Nanostruct. Mater. 1999, 11, 573586.
17. Jana, N. R.; Earhart, C.; Ying, J. Y. Chem. Mater. 2007, 19, 50745082.
18. Trebosc, J; Wiench, J. W.; Huh, S; Lin, V. S. Y.; Pruski, M. J. Am. Chem.
Soc. 2005, 127, 30573068.
19. Brus, J.; Dybal, J. Polymer 2000, 41, 52695282.
506
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 20. Hartmeyer, G.; Marichal, C.; Lebeau, B.; Rigolet, S.; Caullet, P.;
Hernandez, J. J. Phys. Chem. C 2007, 111, 90669071.
21. Kumar, R.; Chen, H. T.; Escoto, J. L. V.; Lin, V. S. Y.; Pruski, M. Chem.
Mater. 2006, 18, 43194327.
22. Bronnimann, C. E.; Ziegler, R. C.; Maciel, G. E. J. Am. Chem. Soc. 1988,
110, 20232026.
23. Liu, C. C.; Maciel, G. E. J. Am. Chem. Soc. 1996, 118, 51035119.
24. Alam, T. M.; Fan, H. Macromol. Chem. Phys. 2003, 204, 20232030.
25. dEspinose de la Caillerie, J.-B.; Aimeur, M. R.; El Kortobi, Y.; Legrand, A.
P. J. Colloid Interface Sci. 1997, 194, 434439.
26. Ek, S.; Root, A.; Peussa, M.; Niinisto, L. Thermochim. Acta 2001, 379,
201212.
27. Turov, V. V.; Leboda, R. Adv. Colloid Interface Sci. 1999, 79, 173211.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030

28. Christiansen, S. C.; Zhao, D.; Janicke, M. T.; Landry, C. C.; Stucky, G. D.;
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

Chmelka, B. F. J. Am. Chem. Soc. 2001, 123, 45194529.

29. Bronnimann, C. E.; Chuang, I.-S.; Hawkins, B. L.; Maciel, G. E. J. Am.
Chem. Soc. 1987, 109, 15621564.
30. Shenderovich, I. G.; Buntkowsky, G.; Schreiber, A.; Gedat, E.; Sharif, S.;
Albrecht, J.; Golubev, N. S.; Findenegg, G. H.; Limbach, H.-H. J. Phys.
Chem. B 2003, 107, 1192411939.
31. Goetz, D. P.; Hine, A. M.; Schultz, W. J.; Thompson, W. L. Patent U.S.
5,648,407.
32. Becker, E. D. High Resolution NMR: Theory and Chemical Applications, 2nd
ed.; Academic Press: New York, 1980.
33. Harris, R. K. Nuclear Magnetic Resonance Spectroscopy; Pittman: London,
1983.
34. Sternhell, S.; Kalman, J. R. Organic Structures from Spectra; John Wiley and
Sons: New York, 1986.
35. Sanders, J. K. M.; Hunter, B. K. Modern NMR Spectroscopy: A Guide for
Chemists, 2nd ed.; Oxford University Press: New York, 1993.
36. VanderHart, D. L.; Earl, W. L.; Garroway, A. N. J. Magn. Reson. 1981, 44,
361401.
37. Thakur, K. A. M.; Kean, R. T.; Zupfer, J. M.; Buehler, N. U. Macromolecules
1996, 29, 88448851.
38. Elbayed, K.; Bourdonneau, M.; Furrer, J.; Richert, T.; Raya, J.;
Hirschinger, J.; Piotto, M. J. Magn. Reson. 1999, 136, 127129.
39. Alla, M.; Lippmaa, E. Chem. Phys. Lett. 1982, 87, 3033.
40. Zorin, V. E.; Brown, S. P.; Hodgkinson, P. J. Chem. Phys. 2006, 125, 144508.
41. Schwerk, U.; Michel, D.; Pruski, M. J. Magn. Reson., Ser. A 1996, 119,
157164.
42. Thakur, K. A. M. Polym. News 2002, 27, 237242.
43. Thakur, K. A. NMR Spectroscopy of Polymers in Solution and in the Solid-
State; Cheng, H. N., English, A. D., Eds.; ACS Symposium Series 834;
American Chemical Society: Washington DC, 2002; Chapter 3, 32-42.
44. Thakur, K. A. M.; Newmark, R. A.; Kuehn, N. K.; Gregar, T. Q.
Macromecules 2003, 36, 719723.

507
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 45. Thakur, K. A. Presentation at 221st ACS National Meeting, San Diego, CA,
April 2001; Thakur, K.A. Polym. Prepr. 2001, 42, 5758.
46. Ando, S.; Harris, R. K.; Scheler, U. Fluorine-19 NMR of Solids Containing
Both Fluorine and Hydrogen. In Encyclopedia of Nuclear Magnetic
Resonance; Grant, D. M., Harris, R. K., Ed.; John Wiley & Sons, Ltd:
Chichester, 2002; Volume 9, 531550.
47. Isbester, P. K.; Brandt, J. L.; Kestner, T. A.; Munson, E. J. Macromolecules
1998, 31, 81928200.
48. Slavov, S. V.; Sanger, A. R.; Chuang, K. T. J. Phys. Chem. B 2000, 104,
983989.
49. Saalwachter, K; Krause, M.; Wolfram Gronski, W. Chem. Mater. 2004, 16,
40714079.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch030
Downloaded by CORNELL UNIV on June 19, 2012 | http://pubs.acs.org

508
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Chapter 31

Dependence of the Amount of Xe Sorption

on the 129Xe NMR Chemical Shift in Glassy
Polymers
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

H. Yoshimizu*

Graduate School of Engineering, Nagoya Institute of Technology,

Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan
*E-mail: [email protected]

Microvoids in glassy polymers are considered to be correlated

with unrelaxed volume (so-called excess free volume). In
this article, the microvoids were investigated by Xe sorption
and 129Xe NMR measurements. The Xe sorption isotherms of
glassy polymers were successfully interpreted by the dual-mode
sorption model composed of Henry and Langmuir sorption
sites. 129Xe NMR chemical shifts of 129Xe in glassy polymers
showed a non-linear low-field shift with increasing amount of
Xe sorption because of fast exchange of Xe atoms between
the two sites. From the 129Xe NMR chemical shift, we could
determine the mean size of the microvoids. In several polymers,
this methodology was examined from three perspectives of free
volume, viz., thermal expansion, miscibility with reduced free
volume in a polymer blend, and specific crystalline structure.
From these findings, 129Xe NMR spectroscopy is shown to be a
powerful technique to determine the mean size of microvoids
in glassy polymers.

Introduction
A complete understanding of a solid amorphous polymer at molecular level is
very difficult because of the complex static, dynamic, and steric structures present.
Yet, it is important to understand the detailed structures of glassy polymers, such
as the amount, the size and the shape of inter-spaces between polymer chains, and
connectivity among them. These are not only important from an academic point

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of view but also relevant to many applications. For example, the design and the
development of gas barriers or polymeric separation membranes both need the
information on the qualities and the quantities of the inter-space.
It is often thought that computational chemistry (such as MD simulations)
(1) and positron annihilation lifetime spectroscopy (2) are good approaches for
the detailed characterization of the polymeric inter-space. In this article, another
method is described, which should be complementary to the above approaches.
Xenon is an inert gas and some of its isotopes are NMR active. Through the use
of Xe as a gaseous penetrant and an NMR probe, the micro-interspaces within a
polymeric material can be characterized from the observed NMR signals. Thus,
it is demonstrated here that 129Xe NMR spectroscopy is a powerful technique for
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

the characterization of glassy polymers.

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

Background of 129Xe NMR

129Xe is an isotope of xenon with spin number I = 1/2 and a natural abundance
of 26.4 %. The relative sensitivity for NMR observation of 129Xe is larger than
that of 13C; thus, 129Xe NMR of Xe gas at relatively low pressures can be easily
observed with good sensitivity. The gyromagnetic ratio of 129Xe is about 1.1 times
larger than that of 13C, indicating that the resonant frequency for 129Xe is close
to that for 13C, i.e., easily tunable using standard multinuclear NMR probes in
commercial spectrometers. In addition, Xe atom is slightly larger than methane,
so it is a suitable probe for the sorption environment experienced by typical gases
in polymer membranes. The most attractive aspect of 129Xe NMR is the sensitivity
of the shielding to the sorption environment. The shift range of sorbed xenon
relative to the resonance of the free gas is well over 200 ppm. Since the Xe atom
has a very large polarizability, it is expected that 129Xe NMR signal is sensitive
to its environments. Indeed, the induced 129Xe NMR chemical shifts are strongly
correlated with the size and the nature of micro-pores, because the interactions with
the host system can disturb the Xe electron density. In the case of microporous
materials like zeolites, 129Xe NMR chemical shift ( ) of adsorbed 129Xe obeys the
following equation (37);

where (S) is due to the interactions between Xe and porous inner walls; (Xe)
corresponds to the interactions among Xe atoms; (E), (SAS), and (M) are the
contribution of electric field created by multivalent cations, the interaction of
Xe with the strong adsorption site, and the contribution of the magnetic field
created by the paramagnetic compensation cations, respectively. Figure 1 shows
schematically the relationships between and density of Xe. The (E), (SAS), and
(M) terms can be ignored for glassy polymers because most of them have no
strong charge groups. Therefore, 129Xe NMR chemical shifts of 129Xe sorbed in
glassy polymers can be interpreted by only two terms: (S) and (Xe). As shown in
Figure 1, for common polymers the observed increases linearly with Xe density.
The value of (S) can be experimentally determined through linear extrapolation
of the data obtained at various pressures of Xe. For zeolites, there are many
510
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 reports that (S) values of adsorbed 129Xe are strongly correlated with the sizes
of micro-cage (310). Similarly it is possible to characterize the mean sizes of
micro-interspaces between polymer chains using (S) values.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

Figure 1. Schematic representation of the relationships between 129Xe NMR

chemical shift and density of Xe.
According to Fraissard et.al., when the NMR chemical shift of 129Xe in
zeolites is only determined by the collisions with the walls, where disturbances
due to paramagnetic species or electric fields ((E), (SAS), or (M)) are absent or
negligible, the mean free path is linked to (S) in equation [1]. Hence (S) can be
related to the mean size of the micro cage by the following equation (3, 5).

is a function of both the shape and dimension of micro cage. In the case of
a sphere, the following relationship has been derived,

where DS is the diameter of the sphere and DXe is the diameter of Xe = 4.4 .
When the shape of the cage is cylindrical, the diameter of crosssection, is related
to DC as follows,

Usually, the average shape of micro inter-spaces in the glassy polymer is

assumed to be spherical because of random coil conformations. However, in the
crystalline structures of helical rigid rod chains, the shape of inter space in chain
bundles should be closer to cylindrical.
It is of interest that some researchers have published many articles on
polymeric systems, but they have rarely paid any attention to the density
dependence of 129Xe NMR chemical shift (1116). In order to obtain (S) which
reflects the micropore size in glassy polymer, it is important to evaluate density
511
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 dependence of Xe chemical shift together with the sorptive properties of Xe. In
view of the considerations mentioned above, we have tried to investigate the
relationship between the 129Xe NMR chemical shift and Xe sorptive properties
for glassy polymers in this work.

Gas Sorption Properties of Glassy Polymers

In general, gas sorption of glassy polymers can be rationalized by the dual-
mode sorption model (1720), which is represented by the following equation,
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

where C is the equilibrium sorption amount at pressure p; C was defined as the

molar concentration of a gas per unit volume in the polymer, usually expressed
in unit of cm3 (STP) / cm3polym , which corresponds to the density of the gas
in the polymer. CD is the concentration due to Henrys law contribution, CH
is the concentration due to Langmuir mode contribution, kD is the Henrys law
coefficient, b is the affinity constant of the Langmuir site, and CH is the hole
saturation constant in the Langmuir sorption mode. The value of CH generally
corresponds to the unrelaxed volume (so-called excess free volume) of glassy
polymer as mentioned below. Figure 2 shows a schematic representation of
specific volumetemperature (VT) curve of a typical amorphous polymer
together with its occupied volume. In the temperature region for the glassy state,
the apparent slope of VT curve is almost similar as the curve of occupied volume.
Thus, the volume of non-equilibrium state (i.e., the unrelaxed volume) linearly
increases with decreasing temperature. The polymer free volume has also been
discussed by other researchers from the viewpoints of gas sorption properties
(1923). The mechanisms of gas sorption below and above the glass transition
temperature, Tg, are different, reflecting the differences in physical structures and
thermodynamic states, viz., the non-equilibrium nature of the glassy state and
the equilibrium liquidlike nature of the rubbery state. In general, the sorption
isotherms for gases in glassy polymers are interpreted in terms of a dualmode
sorption model, based on the assumption that the gas is sorbed both by Henry
and Langmuir sorption mechanisms. The former sorption behavior is similar to
the gas sorption in rubbery polymers (20), whereas in the latter mechanism, the
gas sorbs into the microvoids that exist in the glassy polymer and gets saturated
at high pressures.

512
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

Figure 2. Schematic representation of specific volumetemperature (VT) curve

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

of typical amorphous polymer together with its occupied volume.

Therefore, it is possible to understand the behavior of unrelaxed volume by

examining the Langmuir sorption mechanism. The presence of unrelaxed volume
in glassy polymers plays an important role in the gas sorption as microvoids.
As shown in Figure 2, the unrelaxed volume depends on the difference between
measurement temperature and Tg. The aim of utilization of 129Xe NMR for
glassy polymer is to clarify the relationship between the unrelaxed volume and
microvoid size. As another technique, positron annihilation lifetime spectroscopy
(PALS) is also useful in investigating size and distribution of free volume in
polymers (2, 2427). PALS can estimate the micro-space whose size is larger
than that determined by 129Xe NMR, because positron is smaller than Xe atom.
However, PALS technique probably cannot estimate the microvoids in glassy
polymers independently, due to the fact that PALS data rely on the very short
lifetime of o-positronium and are not affected by molecular motions. This means
PALS cannot distinguish between rubbery and glassy states.
Figure 3 shows Xe sorption isotherms of polystyrene (PS), polycarbonate
(PC), tetramethyl polycarbonate (TMPC), and polyphenyleneoxide (PPO) at 25
C (28, 29). All sorption isotherms obtained here are concave toward the pressure
axis, which is commonly observed in glassy polymers. The solid curves show
the results of curve fitting by a non-linear least-square method based on equation
[5]. These isotherms can be explained successfully on the basis of the dual-mode
sorption model. The calculated parameters are summarized in Table I. The value
of CH is followed by the fact that the orders of CH and temperature difference
between Tg and observing temperatures show a similar trend with one another (see
Table I). CH of PS is the minimum of all samples. It indicates that the total amount
of microvoid in PS is the smallest.

513
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

Figure 3. Xe sorption isotherms of polystyrene (PS), polycarbonate (PC),

tetramethyl polycarbonate (TMPC), and polyphenyleneoxide (PPO) at 25 C.

Table I. Dual-mode sorption parameters of Xe and glass transition

temperature of selected glassy polymers
Sample kD 102 CH b 103 Tg*
PS 1.1 4.8 8.8 98
PC 1.8 13.3 6.0 160
TMPC 1.5 17.4 6.0 196
PPO 1.7 20.6 7.8 216
kD : [cm3 (STP)/cm3polym.
cmHg], CH: [cm3 (STP)/cm3polym.], b : [1/cmHg], Tg : [C]. *

Tg was determined by DSC.

Determination of Microvoid Size through 129Xe NMR

129XeNMR spectrum obtained for the PS film in a NMR sample tube with
thick wall at 25 C and 760 cmHg of Xe (with natural abundance of 129Xe) by
single pulse method is shown in Figure 4 as an example. Internal pressure can be
determined from the Xe sorption amount and weight change of the NMR sample
tube, and/or the chemical shift value of gaseous 129Xe signal. All 129Xe NMR
chemical shifts are referenced to an external standard of zero-pressure of Xe gas.
The peak width of the 129Xe in PS is broad compared with that of gaseous xenon,
showing considerably restricted motion of Xe atom in PS. It may be emphasized
that the peak shape of 129Xe in PS is almost Gaussian, completely symmetric,
although two sorption sites in PS are clearly confirmed from Xe sorption isotherm
measurements as mentioned above. Similar results are obtained for the other
samples and pressure conditions. In view of this interpretation, the pressure
dependence of 129Xe NMR chemical shift will be described below.

514
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

Figure 4. 129Xe NMR spectrum obtained for the PS film in a NMR sample tube
with thick wall at 25 C and 760 cmHg of Xe (with natural abundance of 129Xe)
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

by single pulse method at 110 MHz.

Figure 5 shows the plots of 129Xe NMR chemical shift against total Xe sorption
amounts of C for four glassy polymers (28, 29). These downfield shifts with
increasing C are caused by increasing of interactions among Xe atoms, i.e., the
contribution of term (Xe) in equation [1]. The symmetric peak at about 230 ppm
as shown in Figure 4 indicates that the Xe atoms that dissolve in glassy polymer
diffuse quickly, and exchange rapidly between the Henry and Langmuir sites in
the time scale of NMR observation. When contributions of (E), (SAS), and (M)
in equation [1] can be ignored, the 129Xe NMR chemical shift shows linear low-
field shift with Xe density. Actually, the chemical shift of gaseous 129Xe shifts to
low-field linearly with increasing pressure. Thus, it is expected that 129Xe NMR
chemical shifts in the glassy polymers shift linearly with total Xe sorption amounts
(C) because Xe density can correspond to C. As shown in Figure 5, however, the
shift is non-linear. This finding indicates that the density of Xe in glassy polymer
is not proportional to C. Since the value of C for glassy polymer is composed of
CD and CH (see equation [5]), which is different from the case of gaseous Xe, it is
necessary to evaluate the dependence of 129Xe chemical shifts in glassy polymers
on CD and CH.

Figure 5. Plots of 129Xe NMR chemical shift against total Xe sorption amounts
(C) for four glassy polymers.
515
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Assuming fast exchange of Xe atoms between Henry and Langmuir sites,
the pressure-dependence of NMR chemical shifts for each site can be described
via CD and CH at each pressure and calculated using the dual-mode sorption
parameters. Thus, the observed NMR chemical shift, obs. is expressed in the
following equations,
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

where D and H are fractional concentrations of Xe for the Henry and Langmuir
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

sites at each pressure, and D and D are NMR chemical shifts for the Henry
and Langmuir sites, respectively. In subsequent equations, subscripts D and H
correspond to the Henry and Langmuir sites, respectively. From the equation [1],
NMR chemical shift for each site is explained by following equations,

where AD and AH are constants proportional to Xe concentration for each site,

and (S)D,H and (Xe)D,H are NMR chemical shifts due to the interactions among
Xe and porous inner walls, respectively. Using these equations, Xe concentration
dependence of NMR chemical shifts for the Henry and the Langmuir sites can be
calculated. Figure 6 illustrates the result for PS, and it is noted that the unit of the
x axis is the pressure (cm Hg).

Figure 6. The plots of 129Xe NMR chemical shift against pressure of Xe for PS
at 25 C.

516
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Table II. 129Xe NMR parameters of glassy polymers
Sample AD (S)D AH (S)H
PS 0.012 230.8 5.82 205.4
PC 0.18 225.0 1.59 200.7
TMPC 0.18 210.5 1.88 169.7
PPO 0.67 210.1 1.90 166.4
AD,H : [ppm cm3polym./cm3 (STP)], (S)D,H : [ppm].
D is shifted more downfield than obs., and H is more upfield. The parameters
for each polymer are summarized in Table II. It can be seen from Table II that AH is
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

larger than AD by about one order of magnitude for each polymer. It indicates that
Xe density dependence of the 129Xe NMR chemical shift of 129Xe in the Langmuir
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

site is larger than that in the Henry site. In practice, for rubbery polymer that
contains only the Henry site, 129Xe NMR chemical shift of sorbed 129Xe rarely
reveals Xe density dependence. Moreover, the proportion of Xe in the Langmuir
site is larger than that in the Henry site at low-pressure region, while this is opposite
at high-pressure region. From these facts, 129Xe NMR chemical shift of 129Xe
sorbed in glassy polymer apparently shows non-linear low-field shift against total
sorption amount of C. Additionally AH corresponds to total amount of microvoids
in the glassy polymer; it becomes small with increasing amount of microvoids.
For example, the AH value for PS is larger than that for PPO.
It may be said that the NMR chemical shift extrapolated to CH = 0, i.e., (S)H,
reflects the mean size of microvoids in the glassy polymer. The value of (S)H can
be substituted in equation [2] and then the diameter of spherical space, D S can be
calculated. The results are summarized in Table III together with literature data
of PALS (25, 26). It appears that the spherical spaces correlated to microvoids
in glassy polymers are on the order of angstroms. The order of mean size of
microvoid in four glassy polymers, PPO > TMPC > PC > PS, is consistent with
that of Tg and CH for Xe as shown in Table I. In addition, DS value is close to the
corresponding PALS data, suggesting that the 129Xe NMR spectroscopy is a good
method for characterizing microvoids in glassy polymers.

Table III. Mean size of microvoid and/or micro-pore in glassy polymers

determined by 129Xe NMR (DS, NMR ) and PALS (DS, PALS )
Sample D S, NMR / [] D S, PALS / []*
PS 5.15 5.76
PC 5.27 5.88
TMPC 6.17 6.40
PPO 6.29 6.56
* from references (25) and (26).

517
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Reliability of Microvoid Size Determination by 129Xe NMR
This section examines the reliability of microvoid size determination by using
the 129Xe NMR chemical shift from three perspectives of free volume, viz., thermal
expansion (30), miscibility with reduced free volume in polymer blends (31, 32),
and specific crystalline structures (33).
First, we show the Xe sorption properties and 129Xe NMR spectra of
PPO measured at various temperatures. All the sorption isotherms obtained in
temperature range of 60 to +80 C can be analyzed based on the dualmode
sorption model. The Langmuir saturation constant CH, which corresponds to
unrelaxed volume, is then determined. CH increases linearly with decreasing
temperature. When this straight line is extrapolated to CH = 0, the temperature
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

obtained is almost the same as Tg of PPO as shown in Figure 7(a). Yet, the mean
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

diameter of the microvoids (assuming spheres) of PPO which are determined

from the analyses of the pressure-dependence 129Xe NMR chemical shift of the
129Xe in PPO, also increases with decreasing temperature. The straight line drawn

in Figure 7(b) is a rough estimation because only a few data points are available,
but this suggests that the extrapolated value at Tg is close to 4.4 , which is the
diameter of Xe atom. These findings support the interpretation of unrelaxed
volume as shown above in Figure 2.

Figure 7. Langmuir saturation constant of Xe (CH) (a) and the mean diameter of
individual microvoids (b) of PPO plotted against temperature.
The next topic is the relationship between the variations of microvoids and
gas sorption properties for miscible PPO / PS blend in the glassy state (31, 32).
518
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 This has been investigated by Xe sorption and 129Xe NMR measurements. The
composition dependence of the specific volume (Vsp) has been determined from
the density measurement. It has been shown that Vsp values of the blends are lower
than those calculated by the simple additive rule. Thus, a volume contraction has
taken place when blending PPO and PS, and this may be attributed to an attractive
interaction between the methyl groups of PPO and the phenyl rings of PS (34). Xe
sorption isotherms of this blend system can be interpreted successfully on the basis
of the dual-mode sorption model, indicating that both polymers are in the glassy
state. In Figure 8(a), CH is plotted against the volume fraction of PPO in the blend.
It appears that CH is smaller than that expected from a simple additive rule drawn
as dashed line, whereas kD and b follow additive rules. The data indicate that the
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

decrease in the total amount of microvoids in the blends has occurred by blending,
which is consistent with the result of the density measurement. 129Xe NMR spectra
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

of 129Xe in the blends show a non-linear low-field shift with increasing sorption
amount of Xe because of fast exchange of Xe atoms between Henry and Langmuir
sites. From the analysis of 129Xe NMR chemical shifts, it has been found that
the mean volume of individual microvoids (v) varies with a negative deviation
versus volume fraction of PPO in the blend as well as versus Vsp and CH (Figure
8(b)). For PPO / PS blend system, it is confirmed that the contraction of individual
microvoids occurs by blending.

Figure 8. Langmuir saturation constant of Xe (CH) (a) and the mean volume of
individual microvoids (v) (b) plotted against volume fraction of PPO in the PPO
/ PS blend system at 25 C.

Additional data on microvoid size determined by means of 129Xe NMR

chemical shift are summarized here. Figure 9 shows the plots of the mean volume
of individual microvoids (v) against CH (35). A good linear relationship between
v and CH is obtained. This relationship can be seen as a result of an equation (v
= 3.8 x CH + 48.8), indicating that the minimum microvoid volume is 48.8 3,
which is close to that of Xe atom (44.6 3). In other words, if the microvoid size
is smaller than Xe atom, sorption is impossible. These results make it clear that
unrelaxed volume (total amount of microvoids) takes on its value in response
to the changing size of individual microvoids, and CH is a parameter which

519
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 indicates the total amount of microvoids. Also, it is possible that microvoid size
can be predicted from CH.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

Figure 9. Plot of the mean volume of individual microvoids (v) against Langmuir
saturation constant of Xe (CH) for many glassy polymers at various temperatures.

The final topic in this section is the characteristic cavity in a crystalline part of
poly(4-methyl-1-pentene), (PMP) investigated by gas permeation and 129Xe NMR
measurements on PMP membranes with various degrees of crystallinity (33).
PMP is one of the semi-crystalline polymers and is widely used in industry
because of its good transparency, heat stabilty, and solvent-resistance. It is one
of the significant properties of PMP that the density of the crystalline region
is lower than that of amorphous region at around room temperature. Kusanagi
et.al. have investigated the crystalline structure of PMP, which is characterized
by molecular chains in 7/2 helical conformation packed in a tetragonal unit cell
and concluded that the lower density of crystal region is attributed to a coarse
packing of the chains; as a result, the cylindrical cavity with a diameter of about
4 is formed in the crystalline phase (36). This fact means that transport of gas
and vapor molecules occurs not only in amorphous but also in crystalline regions.
Figure 10 presents a schematic model of crystalline PMP as a cross-section of
four 7/2 helical chain bundle. It can be expected that small gases can diffuse in
the center circle of about 4 diameter drawn in the figure. The permeability
coefficient extrapolated to 100 % crystallinity, i.e., the permeability coefficient for
the cylindrical cavity along the helical PMP chains can be determined from the
permeation data of PMP membranes with various degrees of crystallinity. For the
cavity, permeability coefficients of He, N2, O2, CH4, and CO2 have shown finite
values, whereas those of C3H8 and tert-C4H10 are zero. Xe sorption measurements
of PMP membranes lead to the conclusion that Xe with a diameter of 4.4 is
able to penetrate the cylindrical cavity of crystalline PMP. From the analysis of
129Xe NMR chemical shifts of 129Xe in PMP membranes with various degrees of

crystallinity by using equations [2] and [4], it has been able to evaluate the size of
the cavity in PMP crystal as about 4.5 . This value is consistent with the results
of not only crystallographical analysis but also gas permeation and Xe sorption
520
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 measurements mentioned above. From the present study, it is concluded that the
width of the cylindrical cavity in PMP crystal is about 4.5 .
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

Figure 10. Schematic model of PMP crystal as a cross-section of four 7/2 helical
chain bundle: stick (upper) and CPK (lower) models. The circle with a diameter
of ~ 4 has been drawn in the center of the CPK model.

Concluding Remarks
In this article, the potential of 129Xe NMR spectroscopy for the
characterization of glassy polymers is demonstrated through the pressure
dependence of 129Xe NMR chemical shift. Previously it has been reported that the
size of molecular cavity in the crystalline form of syndiotactic polystyrene can be
determined by means of 129Xe NMR spectroscopy (37). In our work, 129Xe NMR
chemical shift of the 129Xe in rubbery polymers and liquids is shown to depend
on Xe concentration, D, and AD, as mentioned above (Table II and Figure 6).
The relationships between 129Xe NMR chemical shift ((S) ) of 129Xe in n-alkane
liquid (38), and its fractional free volume (Vf ) are very useful in predicting the
density of organic and polymeric materials in the liquid and rubbery states. In
addition, we successfully reported the estimation of the density of side chain
regions in low-density liquid-crystalline polyester with n-alkyl side chain (39,
40). We have also demonstrated that the linewidth of 129Xe NMR signal of the
129Xe in this polyester corresponds to its mobility and (therefore) diffusivity. It

is anticipated that 129Xe NMR spectroscopy will be increasingly used in the near
future as a sophisticated technique for the characterization of diffusive properties
521
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 of polymeric materials and for the determination of microvoid size in glassy
polymers.

Acknowledgments
We gratefully acknowledge partial financial support by a Grant-in-Aid for
Scientific Research (C), No. 20550186 (2008) from the Japan Society for the
Promotion of Science.

References
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

1. Koyama, A.; Yamamoto, T.; Fukao, K.; Miyamoto, Y. J. Chem. Phys. 2001,
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

115, 560566.
2. Hill, A. J.; Jones, P. L.; Lind, J. H.; Pearsall, G. W. J. Polym. Sci., Part: A
Polym. Chem. 1988, 26, 15411552.
3. Demarquay, J.; Fraisaard, J. Chem. Phys. Lett. 1987, 136, 314318.
4. Ripmeester, J. A.; Ratcliffe, C. I.; Tse, J. S. J. Chem. Soc., Faraday Trans. 1
1988, 84, 37313745.
5. Fraisaard, J.; Ito, T. Zeolites 1988, 8, 350361.
6. Ito, T.; Springuel-Huet, M. A.; Fraisaard, J. Zeolites 1989, 9, 6873.
7. Ripmeester, J. A.; Ratcliffe, C. I. J. Phys. Chem. 1990, 94, 76527656.
8. Fetter, G.; Tichit, D.; De Mnorval, L. C.; Fraisaard, J. Appl. Catal. 1990,
65, L14.
9. Dybowski, C.; Bansal, N. Annu. Rev. Phys. Chem. 1991, 42, 433464.
10. Barrie, P. J.; Klinowski, J. Prog. Nucl. Magn. Reson. Spectrosc. 1992, 24,
91108.
11. Walton, J. H.; Miller, J. B.; Roland, M. J. Polym. Sci., Part B: Polym. Phys.
1992, 30, 527532.
12. Miller, J. B. Rubber Chem. Technol. 1993, 66, 455.
13. Miyoshi, T.; Takegoshi, K.; Terao, T. Polymer 1997, 38, 54755479.
14. Walton, J. H.; Miller, J. B.; Roland, C. M.; Nagode, J. B. Macromolecules
1993, 26, 40524054.
15. Mansfeld, M.; Flohr, A.; Veeman, W. S. Appl. Magn. Reson. 1995, 8,
573586.
16. Morgan, D. R.; Stejskal, E. O.; Andrady, A. L. Macromolecules 1999, 32,
18971903.
17. Koros, W. J.; Paul, D. R.; Huvard, G. S. Polymer 1979, 20, 956960.
18. Hachisuka, H.; Tsujita, Y.; Takizawa, A.; Kinoshita, T. Polymer 1991, 32,
23822386.
19. Merkel, T. C.; Bonder, V.; Nagai, K.; Freeman, B. D. Macromolecules 1999,
32, 370374.
20. Kesting, R. E.; Fritzsche, A. K. Polymeric Gas Separation Membranes; John
Wiley & Sons, Inc.: New York, 1993; pp 1959.
21. Kamiya, Y.; Mizoguchi, K.; Naito, Y.; Hirose, T. J. Polym. Sci., Part B:
Polym. Phys. 1986, 24, 535547.
22. Fleming, G. K.; Koros, W. J. Macromolecules 1986, 19, 22852291.
522
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 23. Yampolskii, Yu. P.; Kaliuzhnyi, N. E.; Durgarjan, S. G. Macromolecules
1986, 19, 846850.
24. Ruan, M. Y.; Moaddel, H.; Jamieson, A. M.; Simha, R.; McGervey, J. D.
Macromolecules 1992, 25, 24072411.
25. Li, H. L.; Ujihira, Y.; Nanasawa, A.; Jean, Y. C. Polymer 1999, 40, 349355.
26. Liu, J.; Jean, Y. C.; Yang, H. Macromolecules 1995, 28, 57742779.
27. Bohlen, J.; Kirchheim, R. Macromolecules 2001, 34, 42104215.
28. Suzuki, T.; Miyauchi, M.; Takekawa, M.; Yoshimizu, H.; Tsujita, Y.;
Kinoshita, T. Macromolecules 2001, 34, 38053807.
29. Suzuki, T.; Miyauchi, M.; Yoshimizu, H.; Tsujita, Y. Polymer J. 2001, 33,
934938.
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

30. Yoshimizu, H.; Ohta, S.; Asano, T.; Tsujita, Y. Polymer J. 2011, submitted.
31. Suzuki, T.; Yoshimizu, H.; Tsujita, Y. Desalination 2002, 148, 359361.
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ch031

32. Suzuki, T.; Yoshimizu, H.; Tsujita, Y. Polymer 2003, 44, 29752982.
33. Suzuki, T.; Tanaka, T.; Nakajima, M.; Yoshimizu, H.; Tsujita, Y. Polymer J.
2002, 34, 891896.
34. Djordjevic, M. B.; Porter, R. S. Polym. Eng. Sci. 1983, 23, 650657.
35. Yoshimizu, H.; Asano, T.; Saji, J.; Ogawa, Y.; Ohta, S.; Suzuki, T.; Tsujita,
Y. Seni Gakkaishi. 2011, submitted.
36. Kusanagi, H.; Takase, M.; Chatani, Y.; Tadokoro, H. J. Polym. Sci., Polym.
Phys. Ed. 1978, 16, 131142.
37. Sivakumar, M.; Suzuki, T.; Yamamoto, Y.; Mahesh, K. P. O.; Yoshimizu, H.;
Tsujita, Y. J. Membr. Sci. 2004, 238, 7581.
38. Stengle, T. R.; Williamson, K. L. Macromolecules 1987, 20, 14281438.
39. Tsukahara, M.; Tsujita, Y.; Yoshimizu, H.; Kinoshita, T. ACS Symposium
Series 876; Pinnau, I., Freeman, B., Eds.; American Chemical Society,
Washington, DC, 2004; pp 129138.
40. Yoshimizu, H.; Tsukahara, M.; Suzuki, T.; Toida, J.; Ando, A.; Watanabe, J.;
Tsujita, Y. J. Mol. Struct. 2005, 739, 1926.

523
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Epilogue: An Addiction to NMR
What do people think about NMR?
To some its an instrument with a quirk;
To others its boring or just bizarre;
Yet to a chosen few, its our lifes work.
Downloaded by UNIV OF GUELPH LIBRARY on June 19, 2012 | http://pubs.acs.org
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ot001

Liquid NMR is a chemists dream,

That can determine structures yet unknown.
Solid techniques are held in high esteem;
Their elegance has never been outshone.

Whether liquid or solid, we fiddle

With polymers, composites, and resins.
We struggle to unravel the riddle,
Seeking the mysteries of nuclear spins.

We play with pulses, probes, and smart gadgets,

Tuning them like a banjo or a bass.
We use magic angle in large magnets
And relish spinning in a special space.

Thus, spurred by fun and curiosity,

We labor willingly into the night.
We revel in our creativity,
Be it a new technique or an insight.

Why are we so delighted with our lot,

Forever working, not willing to change?
Yes, we have an addiction that, once caught,
We cannot shake off, nor want to shortchange!

H. N. Cheng
July 4, 2011

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Editors Biographies
H. N. Cheng
H. N. Cheng (Ph.D., University of Illinois) is currently a research chemist
at Southern Regional Research Center of the U.S. Department of Agriculture
Downloaded by PENNSYLVANIA STATE UNIV on June 19, 2012 | http://pubs.acs.org

in New Orleans, where he works on projects involving improved utilization

Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ot002

of commodity agricultural materials, green chemistry, and polymer reactions.

Over the years, his research interests have included NMR spectroscopy, polymer
characterization, biocatalysis and enzymatic reactions, functional foods, and pulp
& paper technology. He is an ACS and a POLY Fellow and has authored or
co-authored 148 papers, 24 patent publications, co-edited 7 books, and organized
or co-organized 20 symposia at national meetings since 1997.

Tetsuo Asakura
Tetsuo Asakura (Ph.D., Tokyo Institute of Technology) has been Associate
Professor and Professor in the Department of Biotechnology of Tokyo University
of Agriculture and Technology since 1981. He was Assistant Professor of Nihon
University School of Dentistry at Matsudo in 198081 and Visiting Professor
in the Department of Chemistry at Florida State University in 199091. He
has produced approximately 300 publications. His research specialties include
structural analysis of polymers using NMR, NMR chemical shift calculation,
characterization of silk and transgenic silk, and applications of silk to biomaterials.
He received Awards of Polymer Society in Japan and Fiber Society in Japan

Alan D. English
Alan D. English Ph.D. is a DuPont Fellow in DuPont Central Research
and Development at the Experimental Station in Wilmington, DE. His interests
include solid state NMR spectroscopy of polymers and development of polymer
structure/property/processing relationships. He is both an APS Fellow and an
ACS Fellow and has been the recipient of a number of awards including the ACS
Award in Applied Polymer Science and the International Award of the Society of
Polymer Science Japan. He is the author of nearly 100 papers/patents/books and
has organized or co-organized 15 symposia at national meetings/workshops since
1981. He is also an Associate Editor of Macromolecules.

2011 American Chemical Society

In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Subject Index
A B

A. See Amyloid -peptide (A) BA/MMA copolymers, 386f, 387f, 390f

-CD. See -cyclodextrin (-CD) Bicellar pore size probing, 239
Acidic aqueous solutions, 342 Bicellar self-assemblies, 223f
Acrylic latex polymers Bicelle disk versus perforated lamellae,
emulsion polymerization techniques, 234
384 Bicelle morphology, 231
heterogeneous polymer, 387 bicellar pore size probing, 239
nuclear magnetic resonance DHPC, 234
spectroscopy, 385 membrane crowding, 233
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

overview, 383 PEG diffusion, 237

quantitative analysis, 391 PEGylated lipid diffusion, 232
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

scanning probe microscopy, 385 Bicelles, magnetic alignment, 225f

solid state NMR, 394 Bi-Gaussian RDC distributions, 218f
solution state NMR, 386 Block copolymer assembly preparation,
-cyclodextrin (-CD) 463
AVANCE 500 MHz NMR Spectrometer, 11B MAS spectra
267 borax, 136f
HTC conformations, 268 boric acid/PVA film, 135f
HTC dynamics, 271 PVA film, 140f
overview, 265 B16 melanoma cells, 470f
structure, 267f Borax
Advanced low-load dipolar decoupling, 11B MAS spectra, 136f
405 signal intensities, 138f
(AGSGAG)5, 286f stationary NMR spectra, 137f
Alternative lock channel, 60, 61f 11B 3QMQMAS spectrum
Amorphous macromolecules, advanced boric acid/PVA film, 141f
NMR experiments 11B stationary NMR spectra
molecular weights, 69 boric acid/PVA film, 141f
overview, 67 Bulk polymers
representative structures, 69 anisotropic spin interactions, 19
Amphiphilic block copolymer, 463, 465s, chain microstructure, defects and
466f dynamics, 26
Amyloid -peptide (A) double quantum NMR, 23
angular dependent NMR experiments, empty helical nano-channels, 31
302 hydrogen bonding, 29
chemical shift changes, 304f manipulation of spin interactions, 22
lipid bilayers, 305 overview, 17
overview, 299 packing in polymers, 30
solution NMR experiments, 301 polymeric proton conductor dynamics,
structural change, 302 27
Anisotropic spin interactions, 19 side chain conformation, 29
Anisotropic swelling, 452 two-dimensional NMR spectroscopy, 24
Application-driven polymer designs, 9
Arrhenius plot, iPP, 202f
Atactic polypropylene (aPP), 69s, 75f, 76f,
77f, 80f
C

Canopy
D2O, 155t
NIM diagram, 152f

533
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Carbon T1 versus temperature, 154f Crosslinked EPDM rubbers, chain mobility
Carboxylic acid-carboxylic acid ester 1H NMR, 210
structure, 120s 1H NMR transverse relaxometry, 210
Carboxylic acid dimer structure, 120s multiple-quantum 1H NMR, 211
Carboxylic acid Li+ neutralized dimer overview, 207
structure, 126s systematic T2 relaxometry study, 209
Carboxylic acid Li+ neutralized tetramer Crosslinked high amylose starch (CHAS),
structure, 126s 448f
Carboxymethylated starch (CMS)-chitosan Crystallinity determination methods, 180
complex tablets, 449f 13C spin-lattice relaxation time, 169f, 175f,
13C CPMAS NMR spectra, 283f, 289f, 290f 328t
CH carbon, 91f Curves, 209
-PL, 327f, 331f, 332f CW-HEPT, 406f, 407f
PVIBE, 89f
13C CPMAS spectra, 395f
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

13C CPMAS SSNMR spectra, 98f

CD curves, -PL, 322f
D
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

13C direct polarization proton, 498f

CDT-PTZ copolymer, 31f 2D contour plot, 1H-1H DQ NMR
Centerband-only detection of exchange spectrum, 31f
(CODEX), 70f 2D-COSY spectra, hexafluoropropylene
Chain diffusions, 196f oxide tetramer, 55f
Chain- ends, polymer structure, 38 Decay analysis, transverse magnetization
Chain microstructure, 26 relaxation, 182
Channels alignment, 255 Decay curves, CPMG, 435f, 436f
Chemical shift ranges, organic compounds, 2D exchange NMR, 26f
21f 1D 1H MAS NMR line shape, 129f
13C liquid-state NMR, soy flour gel, 345f 1D 1H MAS NMR spectra, 129f
13C NMR chemical shifts, -PL, 326t, 330t 2D 1H MAS NOESY, 128f
13C NMR observed conformations, 265 DHPC. See Dihexanoylphosphatidyl-
13C NMR spectra, 269f, 270f choline (DHPC)
poly(propylene-co-1-octene), 58f 2D-HSQC NMR spectra poly(ethylene-co-
proteins, 343f octene), 53f
13C NMR spin-lattice relaxation, 274t Diblock copolymer synthesis, 464
13C nuclear shielding, 270f Differential scanning calorimetry,
CH2 resonances, 129f regioregulated P3AT, 165f
Coefficient of thermal expansion, 110t Diffusion anisotropy, 255
Collective motion, precise polyethylene, Diffusion NMR, 52
28f Diffusion NMR, polymers
Compositional heterogeneity, 375t bicelle fundamentals, 222
Confinement effect, 105 bicelle morphology, 231
Conformational transition, 196f bicellar pore size probing, 239
side-chain, 197f bicelle disk versuss perforated
COOH, signal intensity, 129f lamellae, 234
COOH proton chemical shift, 124f DHPC, 234
Core, NIM diagram, 152f membrane crowding, 233
Corona, NIM diagram, 152f PEG diffusion, 237
Correlation spectroscopy (COSY), 46, 46f, PEGylated lipid diffusion, 232
361f, 363f lipid bilayer membranes, 229
Correlation time distribution widths, overview, 221
temperature dependence, 77f Diffusion tensor, membrane frame, 227f
COSY. See Correlation spectroscopy Dihexanoylphosphatidylcholine (DHPC),
(COSY) 222, 234
CPMAS spectra, 397f Dimyristoylphosphatidylcholine (DMPC),
CPMG, decay curves, 435f, 436f 222
CP-OH angle distribution, 29f schematic cross-section, 238f

534
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Disordered structure, -conjugated Elevated temperatures, long term aging,
polymers, 161 186
Distortionless enhancement by polarization Empty helical nano-channels, spectroscopic
transfer, 487f fingerprints, 32f
2D MAS NMR correlation experiments, EPDM. See Ethylenepropylene-diene
126 terpolymers (EPDM)
DMPC. See Dimyristoylphosphatidyl- -PL. See Microbial poly(-L-lysine)
choline (DMPC) (-PL)
DMPC lipid bilayer, A, 306t EPM. See Ethylene-propylene copolymers
2D-NMR methodologies (EPM)
heteronuclear, 48 Estimated repeat lengths, 95t
homonuclear, 46, 46f Ethylene copolymers, 187
Domain size determination, NMR Ethylene-octene copolymer (m-PE), 183
spin-diffusion experiments, 185 Ethylene-propylene copolymers (EPM),
Donor-acceptor group 207
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

local packing, 31f Ethylenepropylene-diene terpolymers

organization, 31f (EPDM), 207
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Double quantum build-up, 425, 426f thermoplastic vulcanizates, 184

curves, 213f Exchange time constants, 80f
intensity, 212f
rates, 215f
Double quantum filtered COSY
(DQCOSY), 46f, 47
F
Double quantum NMR, 23, 24f 19F
1D 31P 1H-decoupled solid-state NMR detected HSQC spectra, 60f
19F 1D-NMR spectrum
spectra, Ab, 307f
DPNR, vulcanization characteristics, 490t hexafluoropropylene oxide tetramer, 56f
DQCOSY. See Double quantum filtered poly(hexafluoropropylene oxide), 44f
COSY (DQCOSY) Field-effect transistors, 30
Drawing rate effect, 188 Filler-matrix interactions, 105
Drawing ratio effect, 188 Fluorinated solvent effects, 191
Drawing temperature effect, 188 Fluorine NMR, 357
Drug loading effect, 453 Fluoropolymers
DSC curves fluorine NMR, 357
PVIBE/e-PL, 92f instrumentation, 366
PVIBE/e-PL/saponite-clay materials, 366
nanocomposites, 91 NMR methods, 359
Dynamic heterogeneity, 128 overview, 355
PHFPO characterization
2D-NMR, 362
structure, 361
E PVDF characterization
2D-NMR, 365
ECOSY, 363f structure, 364
Elastomers 19F NMR imaging agents
latex state NMR spectroscopy block copolymer assembly preparation,
crosslinking effect, 476 463
dry rubber content effect, 479 characterization, 461
particle size effect, 482 diblock copolymer synthesis, 464
overview, 475 hyperbranched copolymers synthesis,
solid state NMR 463
crosslinking effect, 484 hyperbranched polymeric 19F imaging
2D NMR measurement, 486 agents, 467
FG-FMAS, 489 materials, 461
spectral assignments, 485 MRI experiments, 462
Electrophoretic NMR, 261f overview, 459

535
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Fractionated HDPE analysis, 410 1H MAS NMR chemical shifts, 122t
Freely jointed rigid rod chain, 419, 419f 1H MAS NMR spectra, 125f, 195f, 198f
time average orientation, 420f 1H MAS NMR spectroscopy
Frequency encoding, NMR imaging, 445f ab initio calculations, 118
FT-IR chemical image, soybean, 349f 2D MAS NMR correlation experiments,
FT-NIR spectroscopy, soybean, 347f 126
Functionalized silica nanoparticles dynamic heterogeneity, 128
overview, 495 1H MAS NMR, 119
resonance linewidth broadening material preparation, 116
mechanisms, 497 solid-state 1H NMR spectroscopy, 117
FWHM, isotropic resonance, 129f HMBC. See Heteronuclear multiple bond
correlation (HMBC)
HMQC. See Heteronuclear multiple
quantum coherence (HMQC)
G 1H NMR, e-PL, 323f, 324f
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

2H NMR spectroscopy, 255

Ganglioside, 299, 300f 1H NMR transverse relaxation rates,
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Gas sorption properties, glassy polymers, EPDM, 211f

512 1H NMR transverse relaxometry, 210
Gel formation, 450 Homogeneous copolymers, 392f
Gibbs-Thomson relationship, 88 Homonuclear decoupling, 54
Glassy polymers, gas sorption properties, Homonuclear 2D-NMR, 46, 46f
512 HPLC ion pair chromatogram, -PL, 321f
Glycoprotein, 337 1H spin-diffusion rates, 95t
Glycosylation, 337 HSQC. See Heteronuclear single quantum
Graphene oxide, 108, 108f coherence (HSQC)
Graphite, 105, 108f HSQC spectrum, 303f
HSQC-TOCSY, 49f
1H SS NMR fit data, 110t

H 1H SS NMR spectra, 109f

HTC. See Hexatriacontane (HTC)
HANR, vulcanization characteristics, 490t HTC--CD-IC, 268f
H2BC, 49f HTC chains, 272f
1H-11B CP profiles, boric acid/PVA films, HTC conformations, 268
144f HTC dynamics, 271
1H-11B HETCOR spectrum, boric acid Human erythrocyte membrane arrays, 340f
1H wPMLG NMR spectra, PVA films, 143f
doped in PVA, 144f
1H-13C WISE NMR spectra, 195f, 198f Hydrated gels, 344
HDPE, 183 Hydrated membrane shell, CHAS tablet,
Head-to-head polypropylene (hhPP), 69s, 451f
80f Hydrated wheat grains, 345
Helical packing arrangement, macrocyclic Hydrogen bonding, 29
channel, 33f Hydrostatic pressure, 186
Heterogeneous polymer, 388f Hyperbranched copolymers synthesis, 463
Heteronuclear 2D-NMR, 48, 49f Hyperbranched polymeric 19F imaging
Heteronuclear multiple bond correlation agents, 467
(HMBC), 48, 49f, 61f Hyperbranched polymers, 468s, 469f
Heteronuclear multiple quantum coherence
(HMQC), 48, 49f
Heteronuclear single quantum coherence I
(HSQC), 48, 49f, 360f
nested pulse sequences, 60f Information content, NMR data, 372f
Hexafluoropropylene oxide tetramer, 55f, Intensity, 212f
56f Ion associations, 251, 258, 260f
Hexatriacontane (HTC), 265 Ionic liquid, 254, 258
1H-19F intermolecular NOE spectra, 293f

536
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 Ionic polymers Melt-crystallized PE, temperature effect,
channels alignment, 255 182f
diffusion anisotropy, 255 Melt-state NMR applications, 407
2H NMR Spectroscopy, 255 Melt-state NMR spectroscopy, polyolefins
ion associations, 258 advanced low-load dipolar decoupling,
ion motions, 260 405
ionic liquid, 254, 258 hardware setup, 403
mapping local domain information, 259 materials, 403
membrane preparation, 253 melt-state NMR applications, 407
overview, 252 melt-state versus solution-state, 404
pulsed-field-gradient NMR, 254 overview, 401
structures, 253f scalar coupling mediated method, 405
water uptake determination, 254 spectral analysis setup, 404
Ion motions, 260 Melt-state versus solution-state, 404
Ionomers, 115 Membrane crowding, 233
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

IPB1. See Isotactic poly(1-butene) (iPB1) Membrane preparation, 253

IPP. See Isotactic polypropylene (iPP) Methine carbon, 484f
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

I-PP, 183 Methyl carbon, 484f

aging, 186 Microbial poly(e-L-lysine) (e-PL)
thermoplastic vulcanizates, 184 aqueous solution, 321
IR, -PL, 325f azo dyes, 330
Isotactic poly(1-butene) (iPB1), 191, 194 chain conformation, 329
13C CODEX reference, 199f conformation model, 329f
Isotactic-polyolefins derivative preparation, 319s
iPB1, 194 instruments, 320
iPP, 200 materials, 319
measurements, 193 measurements, 320
overview, 191 overview, 318
samples, 192 physical properties, 321
Isotactic polypropylene (iPP), 187, 191, polymer morphology, 327
200, 409 repeating units, 318s
arrhenius plot, 202f solid state, 324
l versus forder, 203f Micro-CT image, rabbit femurs, 292f
polymer chain trajectory, 204f Molecular dynamics, regioregulated P3AT,
unit cell structure, 201f 163
Isotropic signal fraction, A, 309f Molecular mobility, 179
Monomer sequence, polymer structure, 39
MR T2 relaxometry, phase composition
analysis, 181
L Multiblock copolymer, 257f
Multinuclear solid state NMR, 299
Langevin dynamics simulation, 164f Multiple-quantum 1H NMR, 211
Lauterburs experiment, 443f Multiple receiver systems, 59
Lipid bilayer membranes, 229
Lipid bilayers, 299
Local motion, precise polyethylene, 28f
Local packing, donor-acceptor groups, 31f N
Low branch content quantification, 28f
Low-symmetry macrocycles, 31 Nafion 112, 251, 256f
Nanoscale ionic materials (NIM), 150
materials, 150
NMR characterization, 151
M overview, 150
preparation, 150
Macrocycles, helical packing arrangement, 15N CPMAS NMR spectra, PK60PA40, 99f
33f Nested HSQC pulse sequences, 60f
Mapping local domain information, 259

537
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 N-Hexatriacontane, 265 micelles, interfaces, and confined
NIM. See Nanoscale ionic materials (NIM) environments, 8
NIM, D2O, 155t nanoscale ionic materials (NIM), 149
NIM diagram, 152f overview, 3
NIM preparation, 150 polymers in melt, 7
NMR. See Nuclear magnetic resonance polymers in solution, 7
(NMR) polymers with precise control of
NMR characterization, 151 structure, 8
statistical models, 374f solid state 13C NMR, 85
NMR imaging solid-state, boric acid, 133
frequency encoding, 445f solution NMR methods, 37
magnetic field gradient, 445f strategies, 5
overview, 441 structure-dynamics-properties
pharmaceutical tablets relationships, 7
experimental set-up, 448 Nuclei interaction, NMR, 20t
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

swelling, 449
water diffusion, 449
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

phase encoding, 445f

pulse sequence, 446f
O
techniques
principles, 442 Optimum analysis temperature, 181
signal acquisition, 443
technical considerations, 447
NMR rescaling, 421 P
NMR signals, 420
overview, 417 Packing, polymers, 30
simulation, freely jointed rigid rod chain, Paraffinic oil, thermoplastic vulcanizates,
419 184
NMR spin-diffusion experiments, domain P3AT. See Poly(alkylthiophene) (P3AT)
size determination, 185 P3BT. See Poly(3-butylthiophene) (P3BT)
NMR submolecule, 421 P3BT versus P3HT, 174
15N NMR chemical shifts, e-PL, 333t
-conjugated polymers, disordered
15N NMR spectrum, e-PL, 333f
structure, 161
NOESY, 46f PEB66, 72f, 74f, 76f, 77f
Normalized diffusive intensity decays, STE PEG diffusion, 237
PFG 1H NMR spectra, 241f PEG diffusion coefficient versus degree of
Normalized exchange intensities, 72f, 74f, polymerization N, 240f
79f PEGylated lipid diffusion, 232
Normalized quantum build-up curves, PFG. See Pulsed-field gradient (PFG)
214f, 215f diffusion
N1s XPS fit data, 110t PFG diffusion, NaCl, 156f
Nuclear magnetic resonance (NMR) PFG NMR diffusion, 155f
advanced experiments, 67 Phase composition, 179
application-driven polymer designs, 9 Phase composition analysis, NMR T2
biopolymers, biochemistry, and relaxometry, 181
bio-inspired chemistry, 9 Phase encoding, NMR imaging, 445f
blends, composites, and nanostructures, PHFPO, 357f, 359f
8 Phospholipids, reduced transbilayer water
bulk polymers diffusion coefficient, 236f
applications, 25 P3HT. See Poly(3-hexylthiophene) (P3HT)
background, 18 Pigmented materials, 410
overview, 17 PK10PA90, 97f
confinement effect, 105 PK60PA40, 97f
filler-matrix interactions, 105 15N CPMAS NMR spectra, 99f
1H MAS NMR spectroscopy, 115
PK80PA20, 97f
methodologies, 5 PK90PA10, 97f

538
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 PLGA-PEG-folate, 50f Polymer structure
31P NMR spectra, bicelles, 235f chain- ends, 38
Poly(alkylthiophene) (P3AT), 163 defects, 41
Poly(3-butylthiophene) (P3BT), 163 monomer sequence, 39
Polyethylene acrylic acid copolymers, 115, stereosequence, 40
117s Polyolefins, 408
Polyethylene-co--olefin microstructure, annealing effects, 185
408 crystallinity determination methods, 180
Polyethylene-co-butene-co-hexene deformation effects, 186
terpolymer, 406f domain size determination, NMR
Polyethylene-co-butene (PEB), 69s spin-diffusion experiments, 185
Polyethylene-co-hexene, 411f overview, 179
Polyethylene-co-hexene copolymers, 411f phase composition analysis, NMR T2
Polyethylene-co-vinylacetate copolymer, relaxometry, 181
412f Poly(propylene)
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

Poly(hexafluoropropylene oxide), 44f double quantum signal, 433f

Poly(3-hexylthiophene) (P3HT), 163 longitudinal relaxation time experiments,
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Polyimide composites, 107 434f

Polyimides, 106 shear rate, 435f
Polyisobutylene (PIB), 69, 79f, 80f Polypropylene-co--olefin applications,
Polyisoprene (PI), 69s 409
Polyketone/nylon 6 (PK/PA), solid state Poly(propylene-co-1-octene), 58f
13C NMR, 85 Polysaccharides, 344
See also Poly(vinyl isobutyl ether) Polytact, 378, 379f
(PVIBE)/e-PL, solid state 13C NMR Polyvinylethylene (PVE), 69s, 80f
Polymer--CD-IC, 275f Poly(vinylidene fluoride) (PVDF)
Polymer chain crowding, 157f COSY spectrum, 367f
Polymer chain trajectory, iPP, 204f 19F-13C HSQC 2D-NMR spectra, 366f
Polymeric proton conductor dynamics, 27 structure, 364f
Polymer melt, shear, 431 Poly(vinyl isobutyl ether) (PVIBE)/-PL,
Polymer microstructure solid state 13C NMR
application methodologies crystallinity and crystal size, 88
analytical approaches, 376 Gibbs-Thomson relationship, 88
integrated approaches, 377 impact resistence, 96
kinetic simulation approaches, 378 instruments, 320
simulation approaches, 377 measurements, 320
overview, 371 morphology, 96
Polytact, 378 overview, 85
statistical models See also Polyketone/nylon 6 (PK/PA),
kinetic models, 375 solid state 13C NMR
one state models, 373 Precise polyethylene, 28f
perturbed models, 375 Protein-carbohydrate interactions
two state models, 373 carbohydrate-protein interactions, 342
Polymer nanocomposites hydrated systems, 345
characterization, 107 materials, 341
overview, 105 measurements, 341
polyimide composites, 107 overview, 337
Polymers 31P solid-state NMR spectra, A, 308f, 309f
bicelles, 221 Pulsed-field-gradient NMR, 254
diffusion NMR, 221 Pulsed-field gradient (PFG) diffusion, 151
melt, 7 Pulse sequence, 45
nanocomposites, 105 direct polarization, 139f
packing, 30 double quantum NMR, 24f
precise control of structure, with, 8 PVA film
solution, 7 11B MAS spectra, 140f
solution NMR methods, 37 1H wPMLG NMR spectra, 143f

539
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 PVDF. See Poly(vinylidene fluoride) selective excitation, 52
(PVDF) sensitive detection, 57
Rigid-rod copolymers, 29
Rod-coil copolymers, scheme packing, 30f
Rubbers
Q 13C-NMR spectra, 486f
HMQC spectra, 488f, 489f
Quantitative 1H melt-state NMR 1H-NMR spectra, 488f
spectroscopy, 412 latex state 13C-NMR spectra, 483f
Quantitative NMR, 2D-NMR, 51 solid-state 13C-NMR spectra, 491f
Quantitative one-dimensional NMR, 41 solid-state 1H-NMR spectra, 491f
data acquisition, 41 stress-strain curves, 492f
data processing, 43

S
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

R
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

Scalar coupling mediated method, 405

Raman spectra, -PL, 325f Scattering versus NMR, 34t
Random coil structural component, 284f Schematic cross-section through, DMPC,
Random poly(ethylene-co- propylene), 184 238f
Rapid LCB quantification, 408 Scheme packing, rod-coil copolymers, 30f
Rates, 215f Segregated copolymer, 394t
REDOR plots, 284f Selective excitation, 52
Reduced transbilayer diffusion coefficients, Self-diffusion coefficient, 155t, 156t
242f CHAS tablet, 451f, 452f
Reduced transbilayer water diffusion SEM photomicrographs, 291f
coefficient, phospholipids, 236f Sensitive detection, 57
Regio-regular isotactic polypropylene, 409f Sideband patterns, double quantum NMR,
Regioregulated P3AT 24f
differential scanning calorimetry, 165f Side-chain conformation, 29
FTIR spectra, 166f Side-chain conformational transition, 197f
molecular dynamics, 163 Signal intensity, pulse length, 139f
Regioregulated P3BT Silk-based biomaterials
13C CPMAS NMR measurements, 166
13C CPMAS NMR spectra, 167f
applications, 289
13C spin-lattice relaxation time
design, 287
fluorinated solvent effects, 291
measurements, 168 modifications, 287
DSC, 163 overview, 281
FTIR measurements, 165 structural studies, 282
FTIR spectra, 166f Silk-like peptide, 288f
phase diagram, 176f Single-Gaussian RDC distributions, 218f
Regioregulated P3HT Small angle neutron scattering, 34t
13C CPMAS NMR, 170, 172f
13C spin-lattice relaxation time
Small angle X-ray scattering, 34t
Sodium nitride, stationary NMR spectra,
measurements, 173 137f
DSC, 163 Solid-state NMR, boric acid doped in PVA
FTIR measurements, 165 11B MAS NMR, 135
FTIR spectra, 166f interactions between boric acid and
phase diagram, 176f polymer chains, 142
Rescaling applicability, 427 overview, 133
Resonance linewidth broadening possible conformation, 145
mechanisms, 497 quantitative analysis, 139
Rf signal control Solid-state NMR spectra, 22f
alternative lock channel, 60 disordered structure, -conjugated
homonuclear decoupling, 54 polymers, 161
multiple receiver systems, 59

540
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.
 1H, 117, 118f T1H decay curves, 101f
1H MAS, 120s Thermoplastic vulcanizates, i-PP, 184
isotactic-polyolefins, 191 tHFPO, 362f
Solution NMR methods, 37 T1H relaxation curves, PVIBE, 94f
hardware, 52 TOCSY. See Total correlation spectroscopy
rf signal control, 52 (TOCSY)
new experiments Total configurational entropy, 82f
diffusion, 52 Total correlation spectroscopy (TOCSY),
2D-NMR methodologies, 46 46f, 47
quantitative NMR, 51 Transverse magnetization relaxation, decay
polymer structure, 38 analysis, 182
pulse sequences, 45 Transverse relaxation function, 421
quantitative one-dimensional NMR, 41 Two-dimensional NMR spectroscopy, 24
data acquisition, 41
data processing, 43
Publication Date (Web): October 14, 2011 | doi: 10.1021/bk-2011-1077.ix002

Soybean seeds, 346

Spin interaction manipulation, 22
U
Downloaded by 89.163.35.42 on June 19, 2012 | http://pubs.acs.org

SSB spectral region, 129f

Static solid state NMR spectra, 22f Ultrahigh charge carrier mobility, 30
Stationary NMR spectra Unit cell structure, iPP, 201f
borax, 137f
sodium nitride, 137f
Statistical models, NMR characterization, W
374f
STE intensity decays, DMPE-PEG 2000 Water uptake determination, 254
concentration, 233f Wheat gliadin interactions, 342
STE PFG 1H NMR spectra, normalized Wide angle X-ray scattering, 34t
diffusive intensity decays, 241f
STE PFG 1H NMR spectral series, 232f
Stereosequence, polymer structure, 40
Stimulated echo pulsed field gradient NMR X
pulse sequence, 229f
129Xe NMR, 510, 511f
Storage modulus, 110t
Structure-dynamics-properties glassy polymers, 517t
relationships, 7 PS films, 515f
Subject analysis, NMR, 4t Xenon sorption
Swelling kinetics, CHAS tablet, 453f gas sorption properties, glassy polymers,
512
isotherms, 513
langmuir saturation constant, 518f, 519f,
T 520f
microvoid size determination, 514
Temperature dependence overview, 509
correlation time distribution widths, 77f 129Xe NMR, 510
polyolefins, 183
Temperature effect, melt-crystallized PE,
182f

541
In NMR Spectroscopy of Polymers: Innovative Strategies for Complex Macromolecules; Cheng, H., et al.;
ACS Symposium Series; American Chemical Society: Washington, DC, 2011.