Metal and Alloy Bonding:

An Experimental Analysis
R. Saravanan M. Prema Rani
•

Metal and Alloy Bonding:

An Experimental Analysis
Charge Density in Metals and Alloys

123
Dr. R. Saravanan M. Prema Rani
Research Centre and PG Department Research Centre and PG Department
of Physics of Physics
The Madura College The Madura College
Madurai 625 011 Madurai 625 011
Tamil Nadu Tamil Nadu
India India
e-mail: [email protected]; e-mail: [email protected]
[email protected]

ISBN 978-1-4471-2203-6 e-ISBN 978-1-4471-2204-3

DOI 10.1007/978-1-4471-2204-3
Springer London Dordrecht Heidelberg New York
Library of Congress Control Number: 2011936134

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Ó Springer-Verlag London Limited 2012

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as
permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced,
stored or transmitted, in any form or by any means, with the prior permission in writing of the
publishers, or in the case of reprographic reproduction in accordance with the terms of licenses issued
by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be
sent to the publishers.
The use of registered names, trademarks, etc., in this publication does not imply, even in the absence of
a specific statement, that such names are exempt from the relevant laws and regulations and therefore
free for general use.
The publisher makes no representation, express or implied, with regard to the accuracy of the
information contained in this book and cannot accept any legal responsibility or liability for any errors
or omissions that may be made.

Cover design: eStudio Calamar S.L.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Preface

This book has been written based on the experimental results obtained through
several experimental techniques, especially the powder X-ray diffraction method,
on various metals and alloys we encounter frequently. An analysis of the
interactions of electrons in different atoms has been discussed.
Metals have useful properties including strength, ductility, high-melting points,
thermal and electrical conductivity and toughness. The key feature that distin-
guishes metals from non-metals is their bonding. The existence of free electrons in
metals has a number of profound consequences for the properties of metallic
materials. There are a large number of possible combinations of different metals
and each has its own specific set of properties. The physical properties of an alloy,
such as density, reactivity, Young’s modulus and electrical and thermal conduc-
tivity, may not differ greatly from those of its elements, but engineering properties,
such as tensile strength and shear strength, may be substantially different from
those of the constituent materials. Metals and their alloys make today’s manu-
facturing industry, agriculture, construction and communication systems, trans-
portation, defense equipments, etc. possible. Some of the major reasons for the
continuing advancements in alloys are the availability of materials, new manu-
facturing techniques and the ability to test alloys before they are produced. Most
modern alloys are, in fact, preplanned using sophisticated computer simulations,
which help to determine what properties they will display. Semiconductors have
been studied extensively due to their importance in applications. These materials
receive much attention because physical properties such as the band gap, mobility
and lattice parameter can be continuously controlled. Having such continuous
control is of importance in applications such as electronic and optical devices.
Metals and alloys have high-melting temperatures because of the heavy
bonding between the atoms. There are a variety of applications for metals and
alloys. Due to the importance of these materials, a study of their bonding
interactions has been carried out in this monograph using experimentally observed
X-ray diffraction data.
Today’s technological evolution results in developing new and sophisticated
materials of immense use in domestic, technical and industrial applications.

v
vi Preface

Usually, the synthesis of new materials, especially metals and alloys, results in
single-phase materials, but often not in single crystalline form. Hence, a complete
analysis of the structure, local distribution of atoms and electron distribution in
core, valence and bonding region is necessary using powder diffraction methods,
in addition to single crystal diffraction results, since most of the recent materials
will be initially obtained in powder form. Since one can make efforts to grow
single-crystals from powders, a prior analysis is required using powders to proceed
for single crystal growth.
In this context, we have taken some simple metals (Al, Cu, Fe, Mg, Na, Ni, Te,
Ti, Sn, V, Zn) and alloys (AlFe, CoAl, FeNi, NiAl) and collected powder XRD data
sets or used single crystal XRD data sets from the literature, to study the structure in
terms of the local and average structural properties using pair distribution function
(hereafter PDF), electron density distribution between atoms using Maximum
Entropy Method (hereafter MEM) and bonding of core and valence electron dis-
tribution using multipole technique. Particularly, the PDF analysis requires data
sets of very high values of Q (=4pSinh/k) which is achievable only through syn-
chrotron studies, but not accessible for common crystallographer/material scien-
tists. The present work gives reasonable results obtained through single crystal
work or through high Q data sets, using only powder samples. Also, a study on the
electronic structure of metals using the most versatile currently available tech-
niques like MEM and multipole method is worthwhile. If the tools available for
analysis yield highly precise information, then it is appropriate to apply it to precise
data sets available as in this work, and thus the methodology can also be tested. In
order to elucidate the distribution of valence electrons and the contraction/expan-
sion of atomic shells, multipole analysis of the electron densities was also carried
out. Recently, multipole analysis of the charge densities and bonding has been
widely used to study the electronic structure of materials.
Bonding studies in crystalline materials are very important, especially in metals,
because of their extensive use. These studies can reveal the qualitative nature of
bonding as well as the numerical values of mid-bond densities which indicate the
strength of the material under study. With the advent of versatile methods like
MEM and multipole method, bonding studies gained impetus because of the
accuracy of these methods and the fact that the experimental data can be used with
these methods to accurately determine the actual bonding between atoms.
The precise study of bonding in materials is always useful and interesting, yet
no study can reveal the real picture as no two sets of experimental data are
identical. This problem is enhanced when the model used for the evaluation of
electron densities is not entirly suitable. Fourier synthesis of electron densities can
be of use in picturing bonding between two atoms, but it suffers from the major
disadvantages of series termination error and negative electron densities which
prevent the clear understanding of bonding between atoms; the factor intended to
be analysed. The advent of MEM solves many of these problems. MEM electron
densities are always positive and even with limited number of data, one can
determine reliable electron densities resembling true densities. Currently, the
multipole analysis of charge densities has been widely used to study crystalline
Preface vii

materials. This synthesizes the electron density of an atom into core and valence
parts and yields an accurate picture of bonding in a crystalline system.
In this research monograph on metals and alloys, a complete analysis of
bonding has been made on 11 important metals and four alloys. Powder X-ray
diffraction data as well as single crystal data sets have been used for the purpose.
Charge density analysis of materials provides a firm basis for the evaluation of
the properties of the materials. Designing and engineering of new combinations of
metals requires firm knowledge of the intermolecular features. Recent advances in
technology and high-speed computation has put the crystal X-ray diffraction
technique on a firm pedestal as a unique tool for the determination of charge
density distribution in molecular crystals. Methods have been developed to make
experimental probes to unravel the features of charge densities in the intra and
intermolecular regions in crystal structures. In this report the structural details have
been elucidated from the X-ray diffraction technique through Rietveld technique.
The charge density analysis has been carried out with MEM and multipole method,
and the local and average structure analysis by atomic PDF.
This research work reveals the local and average structural properties of some
technologically important materials, which are not studied along these lines. New
understandings of the existing materials have been gained in terms of the local and
average structures of the materials. The electron density, bonding, and charge
transfer studies analysed in this work will give fruitful information to researchers in
the fields of physics, chemistry, materials science, metallurgy, etc. These properties
can be properly utilized for the proper engineering of these technologically
important materials.
Chapter 1 introduces the significance and applications of metals, alloys and
semiconductors studied in this research work. The objectives of this book are
presented. The essential mechanism of ball milling which has evolved to be a
simple and useful method for the formation of nano crystalline materials is
discussed. The current state of art of non-destructive characterisation techniques
such as X-ray diffraction and scanning electron microscope are discussed.
Chapter 2 provides a survey of the current applications of X-ray diffraction
techniques in crystal structure analysis, with focus on the recent advances made in
the scope and potential for carrying out crystal structure determination directly
from diffraction data. The basic concepts of crystal structure analysis, Rietveld
refinement and the concepts used for the estimation and analysis of charge density
in a crystal are discussed. The more reliable models for charge density estimation
like multipole formalism and MEM are discussed in detail. The local structural
analysis technique and atomic PDF is also discussed.
Chapter 3 presents the results and discussions of this research work. A detailed
account of the results of the materials analysed are presented in the subsections.
Section 3.1 (Sodium and Vanadium Metals) describes about the nature of
bonding and the charge distribution in sodium and vanadium metals are analysed
using the reported X-ray data of these metals. MEM and multipole analysis used
for bonding in these metals are elucidated and analysed. The mid-bond densities in
sodium and vanadium are found to be 0.014 and 0.723 e/Å3 respectively, giving an
viii Preface

indication of the strength of the bonds in these materials. From multipole analysis,
the sodium atom is found to contract more than the vanadium atom.
Section 3.2 (Aluminium, Nickel and Copper) describes the average and local
structures of simple metals Al, Ni and Cu are elucidated for the first time using
MEM, multipole and PDF. The bonding between constituent atoms in all the above
systems is found to be well pronounced and clearly seen from the electron density
maps. The MEM maps of all the three systems show the spherical core nature of
atoms. The mid-bond electron density profiles of Al, Ni and Cu reveal the metallic
bonding nature. The local structure using PDF profile of Ni has been compared
with that of the reported results. The R value in this work using low Q XRD data
for the PDF analysis of Ni is close to the value reported using high Q synchrotron
data. The cell parameters and displacement parameters were also studied and
compared with the reported values.
Section 3.3 (Magnesium, Titanium, Iron, Zinc, Tin and Tellurium) describes the
average and local structures of magnesium, titanium, iron, zinc, tin and tellurium
are analysed using the MEM, and PDF. The structural parameters of the metals
were refined with the well-known Rietveld powder profile fitting methodology.
One-, two- and three-dimensional electron density distributions of Mg, Ti, Fe, Zn,
Sn and Te have been mapped using the MEM electron density values obtained
through refinements. The mid-bond density in Ti is the largest value along [110]
direction among the six metal systems. From PDF analysis the first neighbour
distance is observed to decrease as the atomic number increases for all the metals.
Section 3.4 (Cobalt Aluminium and Nickel Aluminium Metal Alloys) describes
the precise electron density distribution and bonding in metal alloys CoAl and NiAl
is characterized using MEM and multipole method. Reported X-ray single-crystal
data used for this purpose. Clear evidence of the metal bonding between the con-
stituent atoms in these two systems is obtained. The mid-bond electron densities in
these systems are found to be 0.358 and 0.251 e/Å3 respectively, for CoAl and NiAl
in the MEM analysis. The two-dimensional maps and one-dimensional electron
density profiles have been constructed and analysed. The thermal vibration of the
individual atoms Co, Ni and Al has also been studied and reported. The contraction
of atoms in CoAl and expansion of Ni and contraction of Al atom in NiAl is found
from multipole analysis, in line with the MEM electron density distribution.
Section 3.5 (Nickel Chromium (Ni80Cr20)) describes the alloy Ni80Cr20 was
annealed and ball milled to study the effect of thermal and mechanical treatments
on the local structure and the electron density distribution. The electron density
between the atoms was studied by MEM and the local structure using PDF. The
electron density is found to be high for ball-milled sample along the bonding
direction. The particle sizes of the differently treated samples were realized by
SEM and through XRD. Clear evidence of the effect of ball milling is observed on
the local structure and electron densities.
Section 3.6 (Silver doped in NaCl (Na1-xAgxCl)) describes the alkali halide
Na1-xAgxCl, with two different compositions (x = 0.03 and 0.10) is studied with
regard to the Ag impurities in terms of bonding and electron density distribution.
X-ray single crystal data sets have been used for this purpose. The analysis focuses
Preface ix

on the electron density distribution and hence the interaction between the atoms is
clearly revealed by MEM and multipole analysis. The bonding in these systems is
studied using two-dimensional MEM electron density maps on the (100) and (110)
planes and one-dimensional electron density profiles along the [100], [110] and
[111] directions. The mid-bond electron densities between atoms in these systems
are found to be 0.175 and 0.183 e/Å3, respectively, for Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl. Multipole analysis of the structure is performed for these two
systems, with respect to the expansion/contraction of the ion involved.
Section 3.7 (Aluminium Doped with Dilute Amounts of Iron Impurities (0.215
and 0.304 wt% Fe)) describes the electronic structure of pure and doped alu-
minium with dilute amounts of iron impurities (0.215 and 0.304 wt % Fe) has been
analysed using reported X-ray data sets and the MEM. Qualitative as well as
quantitative assessment of the electron density distribution in these samples is
made. The mid-bond characterization leads to a conclusion about the nature of
doping of impurities. An expansion of the size of the host aluminium atom was
observed with Fe impurities.
Chapter 4 presents the conclusion of the results of the reported work.
A complete analysis on the electron density of important metals and alloys is
presented in this book. This book will be highly useful for scientists and
researchers working in the areas of metallurgy, materials science, crystallography,
chemistry and physics.
Acknowledgments

The author Dr. R. Saravanan, acknowledges his family for their kind support, help
and for making the atmosphere conducive during the course of the compilation of
this book.
The author Ms. M. Prema Rani, wishes to thank her family, husband and espe-
cially her children for their support and for motivating her in writing this book.
The authors thank the various finding agencies in India, the University Grants
Commission (UGC), Council of Scientific and Industrial Research (CSIR) and
Department of Science and Technology (DST), though they did not fund the
compilation of this book directly. But, the authors believe that the various research
tasks accomplished during the course of the work for the book may involve usage
of the resources arising out of the funds by the above agencies and hence these
agencies are gratefully acknowledged.
The authors wish to render their cordial thank to the authorities of the Madura
College, Madurai, 625 011, India for their generous support in the various research
efforts by the authors which led to the successful compilation of this book.
Research of high quality needs good support from various people including the
authorities in the concerned institutions from where the research efforts originate.
In that respect, the authors thank the principal and the board of management of the
Madura College, Madurai, 625 011, India, particularly the secretary, Mr. M.S.
Meenakshi Sundaram, The Madura College Board, Madurai, 625 011, India for his
support and encouragement in the academic and research efforts of the authors.
Editing a book on a special topic like the present one involves help, support,
and constant motivation by a large number of clause of people, right from clerical
level and up to intellectual level. The authors wish to acknowledge all those people
who could not find a place in this page of this book but who rendered their cordial
help for successfully editing this book.
The authors dedicate this book for real hard working people with real positive
qualities.

Dr. R. Saravanan
M. Prema Rani

xi
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Significance of the Present Work . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.1 Sodium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.2 Vanadium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4.3 Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4.4 Aluminium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4.5 Titanium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4.6 Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4.7 Nickel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4.8 Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.9 Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.10 Tin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.11 Tellurium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Significance of Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.5.1 Alloys in Nuclear Reactors . . . . . . . . . . . . . . . . . . . . 14
1.5.2 Alloy Wheels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6 Significance of the Alloys Dealt With in this Research Work. . . 14
1.6.1 Cobalt Aluminium . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6.2 Nickel Aluminium . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.6.3 Nickel Chromium. . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.6.4 Iron–Nickel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.6.5 Sodium Chloride Doped with Silver . . . . . . . . . . . . . . 17
1.6.6 Aluminium Doped with Iron . . . . . . . . . . . . . . . . . . . 18

xiii
xiv Contents

1.7 Ball Milling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 18

1.7.1 Mechanism for the Formation of Nano Crystalline
Materials by the Ball Milling . . . . . . . . . . . . . . . . . . 19
1.7.2 Effect of Materials of Milling Media . . . . . . . . . . . . . 20
1.7.3 Laboratory Ball Mill . . . . . . . . . . . . . . . . . . . . . . . . 21
1.8 X-Ray Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.8.1 X-Ray Diffraction Methods . . . . . . . . . . . . . . . . . . . . 22
1.8.2 Diffractometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.8.3 Powder X-Ray Diffraction Instrumentation . . . . . . . . . 23
1.8.4 Single-Crystal X-Ray Diffraction Instrumentation . . . . 24
1.9 Grain Size Analysis from X-Ray Diffraction . . . . . . . . . . . . . . 25
1.10 Scanning Electron Microscope. . . . . . . . . . . . . . . . . . . . . . . . 26
1.11 Fundamental Principles of Scanning Electron Microscopy. . . . . 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

2 Charge Density Analysis from X-Ray Diffraction . . . . . . . . . . . . . 31

2.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2 X-Ray Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2.1 Bragg’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2.2 Electron Density . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2.3 Structure Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.3 Crystal Structure Determination from Diffraction Data . . . . . . . 36
2.3.1 Structure Refinement . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3.2 Theoretical Models in Structure Analysis . . . . . . . . . . 38
2.4 Methods in X-Ray Crystallography . . . . . . . . . . . . . . . . . . . . 38
2.4.1 Structure Determination from Single-Crystal
X-Ray Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.4.2 Powder Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.5 The Rietveld Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.5.1 The Rietveld Strategy . . . . . . . . . . . . . . . . . . . . . . . . 42
2.5.2 Rietveld Refinement. . . . . . . . . . . . . . . . . . . . . . . . . 44
2.6 Multipole Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.6.1 Multipole Electron Density Model . . . . . . . . . . . . . . . 46
2.6.2 Mathematical Approach of Multipole Electron
Density Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.6.3 Criteria for Judging Aspherical Atom Refinements . . . 48
2.6.4 Multipole Refinement Strategy . . . . . . . . . . . . . . . . . 50
2.6.5 Significance of Multipole Model . . . . . . . . . . . . . . . . 50
2.7 Maximum Entropy Method . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.7.1 Maximum Entropy Enhancement
of Electron Densities . . . . . . . . . . . . . . . . . . . . . ... 53
2.7.2 MEM Refinement Strategies . . . . . . . . . . . . . . . . ... 56
Contents xv

2.8 Pair Distribution Function . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

2.8.1 Atomic Pair Distribution Function . . . . . . . . . . . . . . . 57
2.8.2 Important Details of the PDF Technique. . . . . . . . . . . 59
2.8.3 Calculation of PDF . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.8.4 Significance of PDF . . . . . . . . . . . . . . . . . . . . . . . . . 61
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3 Results and Discussion on Metals and Alloys. . . . . . . . . . . . . . . . . 65

3.1 Sodium and Vanadium Metals . . . . . . . . . . . . . . . . . . . . . . . . 65
3.1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.1.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 66
3.1.3 Origin of the Data . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.1.4 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.1.5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 68
3.1.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2 Aluminium, Nickel and Copper . . . . . . . . . . . . . . . . . . . . . . . 74
3.2.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.2.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 75
3.2.3 Data Collection and Structural Refinement . . . . . . . . . 75
3.2.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 81
3.2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium . . . . . . . . 85
3.3.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.3.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 86
3.3.3 Maximum Entropy Method . . . . . . . . . . . . . . . . . . . . 86
3.3.4 Pair Distribution Function . . . . . . . . . . . . . . . . . . . . . 87
3.3.5 Data Collection and Structural Refinement . . . . . . . . . 87
3.3.6 MEM Refinements . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.3.7 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 100
3.3.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
3.4 CoAl and NiAl Metal Alloys. . . . . . . . . . . . . . . . . . . . . . . . . 105
3.4.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.4.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 105
3.4.3 Origin of the Data . . . . . . . . . . . . . . . . . . . . . . . . . . 106
3.4.4 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
3.4.5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 115
3.4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3.5 Nickel Chromium (Ni80Cr20) . . . . . . . . . . . . . . . . . . . . . . . . . 118
3.5.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3.5.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 119
3.5.3 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
3.5.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 123
3.5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
xvi Contents

3.6 Silver Doped in NaCl (Na1-xAgxCl) . . . . . . . . . . . . . . . . . . . 130

3.6.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.6.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 131
3.6.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
3.6.4 MEM Refinement . . . . . . . . . . . . . . . . . . . . . . . . . . 132
3.6.5 Multipole Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.6.6 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 137
3.6.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3.7 Aluminium Doped with Dilute Amounts of Iron
Impurities (0.215 and 0.304 wt% Fe) . . . . . . . . . . . . . . . . . . . 139
3.7.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3.7.2 Summary of the Work . . . . . . . . . . . . . . . . . . . . . . . 139
3.7.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.7.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . 144
3.7.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

4 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
4.1 Sodium and Vanadium Metals . . . . . . . . . . . . . . . . . . . . . . . . 148
4.2 Aluminium, Nickel and Copper . . . . . . . . . . . . . . . . . . . . . . . 148
4.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium . . . . . . . . 149
4.4 Cobalt Aluminium and Nickel Aluminium Metal Alloys. . . . . . 149
4.5 Nickel Chromium (Ni80Cr20) . . . . . . . . . . . . . . . . . . . . . . . . . 150
4.6 Silver Doped in NaCl (Na1-xAgxCl) . . . . . . . . . . . . . . . . . . . 150
4.7 Aluminium Doped with Dilute Amounts of Iron
Impurities (0.215 and 0.304 wt% Fe) . . . . . . . . . . . . . . ..... 150
Chapter 1
Introduction

Abstract The properties of a material are a direct result of its internal structure.
The ability to control structures through processing, and to develop new struc-
tures through various techniques, requires qualitative and quantitative analysis of
the atomic and electronic structure. The average and local structure of some
significant metals and important alloys have been analyzed and reported in this
book. This introduction chapter deals with the significance and applications of
metals, alloys and semiconductors. The essential mechanism of ball milling
which has evolved to be a simple and useful method for the formation of nano
crystalline materials is discussed. The current state-of-the-art of non-destructive
characterisation techniques such as X-ray diffraction and scanning electron
microscope are discussed.

1.1 Introduction

The development of improved metallic materials is a vital activity at the leading

edge of science and technology. Metals offer various combinations of properties
and reliability at a cost which is affordable. They are versatile because subtle
changes in their microstructure can cause dramatic variations in their properties.
An understanding of the development of microstructure in metals, rooted in
thermodynamics, crystallography and kinetic phenomena is essential for the
materials scientist.
Alloys can blend the properties of two or more metals to create a hybrid
metal that is more cost-effective, stronger, more durable and overall better suited
to its intended purpose than the pure metals used to create the compound. With
emerging requirement of designing new materials capable of sustaining high-
strain rate and severe operating conditions with reduced wastage of cost, energy
and material, it has become an important issue to develop full understanding of

R. Saravanan and M. Prema Rani, Metal and Alloy Bonding: An Experimental Analysis, 1
DOI: 10.1007/978-1-4471-2204-3_1,  Springer-Verlag London Limited 2012
2 1 Introduction

the nature of enhanced mechanical properties of the materials. New materials

that can be tailored for individual applications are always in a constant demand.
As the range of uses for powder metallurgy, hard metals and electronic materials
expands, customer requirements are causing materials companies to come up
with new products that have the required properties.

1.2 Significance of the Present Work

Metals and semiconductors play an important role in the present world as

evidenced by their variety of applications. Hence, a study on some important
metals, alloys and semiconducting systems is essential in terms of the local
structure and the average structure which are completely different. The usual
methods of analysis using structural refinement of X-ray or neutron data will give
only the average structure of the materials under investigation. The studies on the
local structure of materials seem to be rare because of the complexity of the
problem. There is only limited information available about the investigations of
materials in terms of the local structure. Numerous research papers are being
published every year based on powder as well as single-crystal X-ray diffraction
(XRD) data. The structures reported using those data are only average structures.
Since, the analysis of local structure requires highly precise data up to maximum
possible Bragg angle, accurate refinement of the data is limited. Due to the
complexity of the problem, tasks of acquirement of precise X-ray data from the
samples, and the computational incapabilities, local and average structural analysis
has not been much explored. Atomic ordering is closely related to the materials’
electronic and magnetic properties. Although the physical properties of alloys are
closely related to their electronic structures, studies on the charge transfer and
hybridisation of the electronic states are still insufficient (Lee et al. 2004).
In the present monograph, apart from pure metals, investigations on the local
and average structures of doped metals and alloys are carried out with various
doping concentrations. The average structure has been studied using both single-
crystal and powder XRD data in some cases. The bonding and electron density
distribution of the host as well as dopant atoms have been studied using tools like
maximum entropy method (MEM) (Collins 1982) and multipole analysis (Hansen
and Coppens 1978). For powder analysis, Rietveld refinement technique (Rietveld
1969) (for average structure) and Pair Distribution Function (Proffen and Billinge
1999) (for local structure) have been used. Effects on the electron density distri-
bution by ball milling (El-Eskandarany 2001; Suryanarayana 2004; Ares et al.
2005) of alloy has been analyzed in this work.
The present research work reveals the local and average structural properties of
some technologically important materials, which are not studied in these lines.
New understandings of the existing materials have been gained in terms of the
local structure and average structure of the materials. The electron density,
bonding and charge transfer studies analyzed in this work would give fruitful
1.2 Significance of the Present Work 3

information to researchers in the fields of physics, chemistry, materials science,

metallurgy, etc. These properties can be properly utilised for the proper engi-
neering of these technologically important materials.

1.3 Objectives

Though the materials studied and reported in this research book are all metals and
alloys, the work has been divided into several parts for the sake of convenience.
They have been given as below.
1. The average and electronic structure of the following elemental metals using
Rietveld (Rietveld 1969), multipole (Hansen and Coppens 1978) and MEM
(Collins 1982) by single-crystal XRD data.
• Sodium (Na)
• Vanadium (V)
2. a. The local, average and electronic structure of the following elemental metals
using Rietveld (1969), and MEM (Collins 1982) by powder XRD data.
• Magnesium (Mg)
• Aluminium (Al)
• Titanium (Ti)
• Iron (Fe)
• Nickel (Ni)
• Copper (Cu)
• Zinc (Zn)
• Tin (Sn)
• Tellurium (Te)
b. The local structural information by analyzing the atomic pair distribution
function (PDF) (Proffen and Billinge 1999).
3. The average and electronic structure of the following metal alloys using
multipole (Hansen and Coppens 1978) and MEM (Collins 1982) by single-
crystal XRD data.
• cobalt aluminium (CoAl)
• nickel aluminium (NiAl)
• Iron nickel (FeNi)
4. a. The annealing and ball milling of the alloy nickel chromium (Ni80Cr20)
b. The study of the local, average and electronic structure of the annealed and
ball milled alloy Ni80Cr20 using Rietveld (Rietveld 1969) and MEM (Collins
1982) by powder XRD data.
c. To study the local structure using PDF (Proffen and Billinge 1999).
4 1 Introduction

d. Analysis of the particle sizes of the differently treated samples by scanning

electron microscopy (SEM) and (XRD).
5. The study of the average and electronic structure of the following doped alloys
using Rietveld (1969), multipole (Hansen and Coppens 1978) and MEM
(Collins 1982) by single-crystal XRD data.
• Sodium chloride with iron impurities (Na1-xAgxCl)
• Aluminium, with iron impurities (0.215 wt% Fe and 0.304 wt% Fe)

1.4 Metals

Metals account for about two-thirds of all the elements and about 24% of the mass
of the planet. Metals have useful properties including strength, ductility, high-
melting points, thermal and electrical conductivity and toughness. The key feature
that distinguishes metals from nonmetals is their bonding. (Gallagher and Ingram
2001) Metallic materials have free electrons that are free to move easily from one
atom to the next. The existence of these free electrons has a number of profound
consequences in the properties of metallic materials (Kittel 2007).
The local and average structures of some technologically important metals such
as sodium, magnesium, aluminium, titanium, vanadium, iron, nickel, copper, zinc,
tin and tellurium are analyzed in this work and their typical properties and uses are
presented below.

1.4.1 Sodium

Sodium is a soft, silvery-white, highly reactive metal having only one stable isotope;
23
Na. Sodium ion is soluble in water in nearly all of its compounds. Sodium metal is
so soft that it can be cut with a knife at room temperature (Zumdahl 2007)
Sodium compounds are important for the chemical, glass, metal, paper,
petroleum, soap, and textile industries. A sodium–sulphur battery is a type of
molten metal battery constructed from sodium and sulphur. This type of battery
has a high-energy density, high efficiency of charge/discharge (89–92%) and long
cycle life, and is fabricated from inexpensive materials (Oshima et al. 2004).
NaS batteries are a possible energy storage technology to support renewable
energy generation, specifically in wind farms and solar generation plants. In the
case of a wind farm, the battery would store energy during times of high wind but
low-power demand. This stored energy could then be discharged from the batteries
during peak load periods. In addition to this power shifting, it is likely that sodium
sulfur batteries could be used throughout the day to assist in stabilising the power
output of the wind farm during wind fluctuations (Walawalkar et al. 2007). Due to
its high-energy density, the NaS battery has been proposed for space applications
(Auxer 1986).
1.4 Metals 5

1.4.2 Vanadium

Pure vanadium is a bright white metal, and is soft and ductile. It has good
corrosion resistance to alkalis, sulphuric and hydrochloric acid and salt water. The
metal has good structural strength and a low fission neutron cross section, making
it useful in nuclear applications. (Lynch 1974).
Vanadium is used in producing rust resistant, spring and high-speed tool steels.
It is an important carbide stabiliser in making steels. Vanadium is also used in
producing superconductive magnets with a field of 175,000 gauss (Lide 1999).
The role of vanadium complexes in catalytically conducted redox reactions
(Crans et al. 2004) and potential medicinal applications, such as in the treatment of
diabetes type I and type II (Crans 2000), has stimulated interest in the stereo-
chemistry and reactivity of its coordination compounds (Monfared et al. 2010).
Vanadium oxides and vanadium oxide-related compounds have a wide range of
practical applications such as catalysts, gas sensors and cathode materials for
reversible lithium batteries, electrochemical and optical devices, due to their
structural, novel electronic and optical properties (Zhang et al. 2010). Mixed metal
oxides find applications in a variety of fields due to the wide variation in their
dielectric and electrical properties. The vanadium-based oxide ceramics have high-
dielectric constant, low-dissipation factor and high-quality factor, which favour the
use of these ceramics in many fields (Nithya and Kalaiselvan 2011).
Secure and reliable power is essential in areas such as telecommunications and
information technology to safeguard the vast computer networks that have been
established. Uninterruptible power systems have incorporated battery technology
to allow smooth power feeding switch-over in the case of a power failure. In such
systems lead-acid batteries are commonly being used until generators come online
or for safe computer shutdown. The vanadium redox battery provides many
advantages over conventional batteries for emergency back-up applications. This
system stores all energy in the form of liquid electrolytes which are re-circulated
around the battery system. The electrolytes can be recharged for indefinite number
of times, or the system can be instantly recharged by mechanically exchanging the
discharged solution with recharged solution (Kazacos and Menictas 1997).
Figure 1.1 shows a vanadium redox battery. A vanadium redox battery consists of a
power cell in which two electrolytes are kept separated by an ion exchange membrane.
Both the electrolytes are vanadium based. Vanadium redox batteries are based on the
ability of vanadium to exist in four different oxidation states (V2, V3, V4 and V5), each
of which holds a different electrical charge. The electrolyte in the negative half-cell
has V3+ and V2+ ions, while the electrolyte in the positive half-cell contains V3+ and
V2+ ions. During charging, reduction in the negative half-cell converts the V3+ ions
into V2+ ions. During discharge, the process is reversed, oxidation in the negative half-
cell converts V2+ ions back to V3+ ions. The typical open-circuit voltage created
during discharge is 1.30 V at 25C (Skyllas-Kazacos 2003).
6 1 Introduction

Fig. 1.1 Vanadium redox

battery

Other useful properties of Vanadium flow batteries are their quick response to
changing loads and their extremely large overload capacities. Their extremely
rapid response times also make them perfectly well suited for UPS-type applica-
tions, where they can be used to replace lead-acid batteries and even diesel
generators.

1.4.3 Magnesium

Elemental magnesium is a fairly strong, silvery-white, light-weight metal. The

lightness combined with good strength-to-weight ratio has made magnesium and
its alloys suitable for use in missiles and automotive industry.
Magnesium alloys have low density (1.5–1.8 g/cm3) and high strength in
relation to their weight (Kainer 2000). Magnesium alloys are used for die-casting
due to their good corrosion resistance and low heat of fusion with the mould
material. Most of the magnesium alloy castings are made for the automotive
industry. Lowering car weight by 100 kg makes it possible to save 0.5 l petrol/
100 km. It is anticipated that in the following years the mass of castings from
magnesium alloys in an average car will rise to 40 kg, internal combustion engines
will be made mostly from the magnesium alloys and car weight will decrease from
1,200 to 900 kg (Mordike and Ebert 2001a, b).
A good capability of damping vibrations and low inertia connected with a
relatively low weight of elements have predominantly contributed to the
employment of magnesium alloys for the fast moving elements and in locations
where rapid velocity changes occur; some good examples may be car wheels,
combustion engine pistons, high-speed machine tools and aircraft equipment
elements (Wang et al. 2002).
The concrete examples for the use of castings of magnesium alloys in batch
production in the automotive industry are elements of the suspension of the front
1.4 Metals 7

and rear axes of cars, propeller shaft tunnel, pedals, dashboards, elements of seats,
steering wheels, elements of timer-distributors, air filters, wheel bands, oil sumps,
elements and housings of the gearbox, framing of doors and sunroofs and others
(Dobrzánski et al. 2007).
In recent times, the increased environmental concerns and the rising costs of oil
have again made magnesium and its alloys a material of interest for the automotive
industry. Considering the characteristics of low density of magnesium, its exten-
sive use in structural body parts of vehicles will offer major reductions of weight
and hence reduction in fuel consumption. Such weight reduction provides a sig-
nificant contribution to reducing the carbondioxide emission. It is estimated that an
average new car produces 156 g CO2/km travelled. This could be reduced to
around 70 g CO2/km through the application of magnesium technology (Mehta
et al. 2004).
The advantages of magnesium and magnesium alloys are, lowest density of all
metallic constructional materials, high-specific strength, good castability, suitable
for high-pressure die-casting, can be turned or milled at high speed, good wel-
dability under controlled atmosphere, much improved corrosion resistance, readily
available, better mechanical properties, resistant to ageing, better electrical and
thermal conductivity and recyclability (Mordike and Ebert 2001a, b).
Magnesium alloys have attracted increasing interest in the past few years due to
their potential as implant materials. Magnesium and its alloys are degradable
during their time of service in the human body. Magnesium alloys offer a property
profile that is very close or even similar to that of human bone (Hort et al. 2010).

1.4.4 Aluminium

Aluminium has been the dominant material in the aircraft industry for more than a
half century due to its attractive combination of light weight, strength, ductility,
corrosion resistance, ease of assembly and low cost (Dorward and Pritchett 1988).
Aluminium foam sandwiches (AFS) due to their flexible process ability and
potential of cost reduction find application in space components. Currently, these
light-weight materials find some first applications in particular fields of mechanical
engineering such as race cars, and small series of other land-based vehicles
(Schwingel et al. 2007). The use of high-strength aluminium alloys in automotive
and aircraft industries allows reducing significantly the weight of the engineering
constructions. In these fields, very often the main requirements for the components
include high fatigue and wear-resistance (Lonyuk et al. 2007). Aluminium solar
mirrors are an alternative for solar concentrators. The aluminium reflectors often
offer an initial reflectance of 85–91% for solar irradiance. They have good
mechanical properties and are easy to recycle (Almanza et al. 2009). The high
strength-to-weight advantage of aluminium alloys has made it the material of
choice for building airplanes and sometimes for the construction of land-based
structures. For marine applications, the use of high-strength, weldable and
8 1 Introduction

corrosion-resistant aluminium alloys have made it the material of choice for

weight sensitive applications such as fast ferries, military patrol craft and
luxury yachts and to lighten the top-sides of offshore structures and cruise ships
(Paik et al. 2005).

1.4.5 Titanium

Titanium has many desirable physical properties. The pure metal is relatively soft
and weak, but it becomes much stronger when mixed with other metals to form
alloys. The high-melting point of titanium (1,668C) shows that it is an ideal
material for the construction of high-speed aircraft and space vehicles.
Due to the exceptional strength-to-weight ratios, toughness, high stiffness and
excellent biocompatibility, titanium and its alloys are used extensively in aero-
space, chemical and biomedical applications (Kartal et al. 2010). Titanium alloys
are widely used in the aerospace industry due to their excellent fatigue/crack
propagation behaviour, and corrosion resistance (Markovsky and Semiatin 2010).
The alloys of titanium represent significant advantages over most other
engineering materials used for a variety of industrial applications due to their
resistance to corrosion, oxidation and erosion. Titanium and its alloys have high-
chemical durability as well as high strength. Their use is significant in nuclear
industry, since the mechanical strength, high-heat proof and radiation proof are
desired by many components, such as the steam condenser tubes, the irradiation
targets for transmuting radioactive wastes and the overpacks for geological
disposal of high level radioactive wastes (Setoyama et al. 2004).
Titanium is the preferred choice for surgical instrumentation due to its lighter
weight, bacterial resistance and durability. High strength-to-weight ratio, corrosion
resistance, non-toxic state and non-ferromagnetic property has made titanium ‘‘the
metal of choice’’ within the field of medicine. It is also durable and long-lasting.
When titanium cages, rods, plates and pins are inserted into the body, they can
last for more than 15 years. And dental titanium, such as titanium posts and
implants, can last even longer. Osseo integration is a unique phenomenon where
the body’s natural bone and tissue actually bonds to the artificial implant. This
firmly anchors the titanium dental or medical implant into place. Titanium is the
only metal that allows this integration. Titanium and its alloys are widely used to
replace failed hard tissues, such as artificial hip joints and dental implants
(Li et al. 2008). Mechanical properties such as high strength, ductility and fatigue
resistance, as well as a low modulus make titanium and its alloy suitable for
applications in jet propulsion systems and human body implant (Heinrich et al.
1996). Titanium has long been used as an implant material in different medical
applications, showing excellent performance in forming a close contact to the
surrounding tissues (Petersson et al. 2009).
1.4 Metals 9

1.4.6 Iron

Pure iron is silvery metal with a shining surface. It is a good conductor of heat and
electricity. Iron is used to make bridges, automobiles and support for buildings,
machines and tools. It is mixed with other elements to make alloys, the most
important of which is steel (Sparrow 1999).
Iron-based glassy alloys seem to be one of the most interesting materials due to
their soft magnetic properties including high-saturation magnetisation. They are
suitable materials for many electrical devices such as electronic measuring and
surveillance systems, magnetic wires, sensors, band-pass filters, magnetic
shielding, energy-saving electric power transformers (Nowosielski et al. 2008).

1.4.7 Nickel

Nickel is a silvery-white lustrous metal with a slight golden tinge. It is one of the
four ferromagnetic elements that exist around room temperature, the other
three being iron, cobalt, and gadolinium. Its Curie temperature is 355C. Nickel is
non-magnetic above this temperature (Kittel 1996). Nickel belongs to the transi-
tion metals and is hard and ductile. The isotopes of nickel range from 48Ni to 78Ni.
The isotope of nickel with 28 protons and 20 neutrons 48Ni is ‘‘double magic’’ and
therefore unusually stable (Audi 2003).
The metal is corrosion-resistant, finding many uses in alloys, as plating, in the
manufacture of coins, magnets, common household utensils, rechargeable batter-
ies, electric guitar strings, as a catalyst for hydrogenation, and in a variety of other
applications. Enzymes of certain life-forms contain nickel as an active centre,
which makes the metal an essential nutrient for those life-forms. It is also used for
plating and as a green tint in glass. In the laboratory, nickel is frequently used as a
catalyst for hydrogenation. Nickel is often used in coins, or occasionally as a
substitute for decorative silver.
Rechargeable nickel batteries are one type of alkaline storage cylindrical bat-
tery and classified as secondary batteries.
Nickel battery has a positive electrode made of active material-nickelous
hydroxide. Because of the perfectly, sealed construction and the efficient charge/
discharge characteristics, nickel batteries provide superior features and practical
values in long service life, high-rate discharge and stable performance. As a result,
they are widely used in many fields such as communication and telephone equip-
ment, office equipment, tools, toys and emergency devices and consumer
applications.
A typical jet engine today contains about 1.8 tonnes of nickel alloys and includes
a long list of tailor-made nickel-based-alloys to meet specific needs (Nickel
Magazine 2007). Pure nickel is a strong candidate for protective coating in bio-
diesel storage applications, due to its high resistance to the corrosive nature of
10 1 Introduction

biodiesel and its vapors and minimal catalytic effects on the oxidation of biodiesel
(Boonyongmaneerat et al. 2011).

1.4.8 Copper

Copper is a ductile metal with very high thermal and electrical conductivity. Pure
copper is rather soft and malleable, and a freshly exposed surface has a pinkish or
peachy color. It is used as a thermal conductor, an electrical conductor, a building
material, and a constituent of various metal alloys.
Copper is the most widely used metal because of high conductivity. Copper and
copper-based alloys are unique in their physical and mechanical properties. They
have excellent corrosion resistance, high resistance to fatigue and relative ease of
joining by soldering available in wide variety of forms (Pillai 2007).
Copper is easily worked, being both ductile and malleable. The ease with which
it can be drawn into wires makes it useful for electrical work in addition to its
excellent electrical properties. Copper can be machined, although it is usually
necessary to use an alloy for intricate parts, such as threaded components, to get
really good machinability characteristics. Good thermal conduction makes it
useful for heat sinks and in heat exchangers. Copper has good corrosion resistance;
it has excellent brazing and soldering properties and can also be welded, although
best results are obtained with gas metal arc welding (Sambamurthy 2007). Copper
as both metal and pigmented salt has a significant presence in decorative art.

1.4.9 Zinc

Zinc compounds are actively investigated because of their significant properties.

Zinc oxide, being an n-type semiconductor with a wide direct gap of about 3.2 eV,
has received much attention as a low-cost material for transparent and conductive
films (Futsuhara et al. 1998). Zinc phosphide a II–V compound exists as a p-type
semiconductor with a direct gap of near 1.51 eV, and is a promising low-cost
material for solar cells due to its band structure (Pawlikowski 1981).
Zincprovidesimmunity,fertilityandthecapacityofsensesincludingsight,tasteand
smell, notes the International Zinc Association. Zinc can also be recycled indefinitely,
withoutlosinganyofitsstructuralorfunctionalcharacteristics(http://www.livestrong.
com/article/199141-uses-for-zinc-powder/).
Zinc-air batteries (non-rechargeable) and zinc-air fuel cells (mechanically
rechargeable) are electro-chemical batteries powered by oxidizing zinc with
oxygen from the air. These batteries have high-energy densities and are relatively
inexpensive to produce. Sizes range from very small button cells for hearing aids,
larger batteries used in film cameras that previously used mercury batteries, to very
large batteries used for electric vehicle propulsion.
1.4 Metals 11

In operation, a mass of zinc particles form a porous anode, which is saturated with
an electrolyte. Oxygen from the air reacts at the cathode and forms hydroxyl ions
which migrate into the zinc paste and form zincate, releasing electrons to travel to the
cathode. The zincate decays into zinc oxide and water returns to the electrolyte. The
water and hydroxyls from the anode are recycled at the cathode, so the water is not
consumed. The reactions produce a theoretical 1.65 V, but this is reduced to 1.4–
1.35 V in available cells. Zinc-air batteries have some properties of fuel cells as well
as batteries, with zinc as the fuel, the reaction rate can be controlled by varying the air
flow, and oxidised zinc/electrolyte paste can be replaced with fresh paste. Metallic
zinc could be used as an alternative fuel for vehicles, in a zinc-air battery (Noring
et al. 1993). Zinc-air batteries are considerably more safer in combating situations
and more environmental friendly than lithium batteries (http://www.defense-update.
com/products/z/zinc-air-battery-new.htm).
A Switzerland-based company, ReVolt uses zinc-air battery technology for
hearing aids. ReVolt’s battery claims to store three times more energy than lith-
ium–ion by volume, and could incur just half the costs (http://www.goodcleantech.
com/2009/11/a_powerful_rechargeable_zinc-a.php).

1.4.10 Tin

As the trend towards further miniaturisation of electronic products continues

apace, packaging technology has progressed from the conventional wire and tape
automated bonding to area array flip-chip bonding, which is able to provide
increased input/output (I/O) counts and improved electrical performance (Qin
et al. 2010). The advantages of this technology include high-density bonding,
improved self-alignment, reliability and ease of manufacture (Wolf et al. 2006).
One major step in the flip-chip interconnection process routes involves the
deposition of, normally, solder alloys onto the bond pads of the chips (also known
as solder bumping). With respect to the bumping materials, lead–tin-based alloys
were the most widely used solders for flip-chip applications because of their low
cost, low-melting point and excellent solderability properties. However, with
world-wide legislation for the removal/reduction of lead and other hazardous
materials from electrical and electronic products, development of a large number
of lead-free, mostly tin-rich, alternative solders has been undertaken (Eveloy et al.
2005). Typically containing more than 90 wt% Sn, with a wide range of alloying
elements such as Ag, Cu, In, Bi and Zn, these lead-free alternatives can be binary,
ternary and even quaternary alloys, with variations in compositions. Sn–Ag–Cu
solders can promote enhanced joint strength and creep and thermal fatigue
resistance, and permit increased operating temperatures for advanced electronic
systems and devices (Fabio and Mascaro 2006).
In electronic/optoelectronic packaging, chip bonding serves three major functions,
i.e., mechanical support, heat dissipation and electrical connection (Hunziker
et al. 1996). The choice of solder material for bonding is based on optimisation of a
12 1 Introduction

number of properties, including solderability, melting temperature, Young’s modulus

(or stiffness), coefficient of thermal expansion, Poisson’s ratio, fatigue life, creep rate
and corrosion resistance. In terms of melting temperature, solders are typically clas-
sified as either hard (high-melting temperature) or soft (low-melting temperature). The
Pb/Sn system is an example of a soft solder, which is commonly used for electronic
packaging. Hard solders, e.g., Au/Sn, are used for optoelectronic packaging. Au/Sn
solder, with its combination of good thermal and electrical conductivities, is partic-
ularly attractive for ‘flip-chip’ bonding, where the active area of the device is next to
the submount. Au-20 wt% Sn is the most common composition utilised; it has a
relatively high-melting temperature (280C), good creep behaviour and good corro-
sion resistance (Ivey 1998).
Guide wires, catheters, stents, etc., are being increasingly employed in the
diagnosis and treatment of cancer, diseases of the circulatory system, etc. A guide
wire is used for navigating a catheter, a tube made of plastic, in a blood vessel. The
tip portion of the guide wire must be sufficiently flexible to pass through the
meandering blood vessels. On the other hand, in the body portion of the guide
wire, a high-elastic modulus and strength against bending are also required to
overcome the high resistance to bending and rotation in a blood vessel and to
smoothly transmit the torque from the end to the tip of the guide wire (Sutou et al.
2006). Ti–Mo–Sn alloy is found to be a promising biocompatible material for use
in catheters (Maeshima and Nishida 2004).

1.4.11 Tellurium

Due to the remarkable physical properties of tellurium such as low band gap and
transparency in the infrared region, Te is used extensively in various technological
areas. Te thin films find use in microelectronic devices such as gas sensor (Shashwati
et al. 2004; Tsiulyanu et al. 2004) and optical information storage (Josef et al. 2004).
Tellurium-based thin films, suitable for applications in environmental monitoring
with considerably short-response time and high sensitivity to nitrogen dioxide at
room temperature have been reported (Tsiulyanu et al. 2001).
Tellurium, with a low band gap of 0.32 eV, is one of the most promising
materials for a shield in a passive radiative cooling (Engelhard et al. 2000).
Radiative cooling is the one among today’s challenges in materials science
research. It occurs when a body gets cold by loosing energy through radiative
processes. The phenomenon of radiative cooling uses the fact that the thermal
energy emitted by a clear sky in the ‘‘window region’’ (8–13 mm) is much less
than the thermal energy emitted by a blackbody at ground air temperature in this
wavelength range. Hence, a surface on the earth facing the sky experiences an
imbalance of outgoing and incoming thermal radiation and cools to below the
ambient air temperature. While this concept can work well at night, assuming a
1.4 Metals 13

relatively dry atmosphere, the solar energy input during the day, which is normally
much greater than that radiated out, causes heating of the system. To prevent this,
a shield is required to cover the radiating surface in order to block solar radiation
during the day as well as to prevent convective mixing in the cooled space. An
ideal radiation shield should completely reflect solar radiation, but allow complete
transmission in the ‘‘atmospheric-window’’ region. Solar radiation should prefer-
ably be reflected, as any absorbed radiation will be converted to heat somewhere in
the system (Dobson et al. 2003). The different approaches for the design of shield
are the introduction of optical scattering materials into shield substrates, and
coating of shield substrates with high-solar reflector films. Te thin films with
thickness of 111–133 nm having high-IR transmission across the full 8–13 lm
band region suitable for solar radiation shield devices have been prepared by
chemical vapor deposition method (Tian et al. 2006). Tellurium is used in
photocopiers to enhance picture quality.

1.5 Significance of Alloys

Technology is reshaping our day-to-day life. Metals and their alloys make today’s
manufacturing industry, agriculture, construction and communication systems,
transportation, defense equipments, etc., possible. Our manufacturing industries
are using different metals and alloys as raw materials for their finished goods.
There are a large number of possible combinations of different metals and
each has its own specific set of properties. Alloying one metal with other metal or
non-metal often enhances its properties. For example, steel is stronger than iron, its
primary element. The physical properties, such as density, reactivity, Young’s
modulus and electrical and thermal conductivity, of an alloy may not differ greatly
from those of its elements, but engineering properties, such as tensile strength
(Francis 2008) and shear strength may be substantially different from those of the
constituent materials. This is sometimes due to the size of the atoms in the alloy,
since larger atoms exert a compressive force on neighboring atoms, and smaller
atoms exert a tensile force on their neighbors, helping the alloy resist deformation.
Sometimes alloys may exhibit marked differences in Behaviour even when small
amounts of one element occur (Hogan 1969; Zhang and Suhl 1985). Some of the
major reasons for the continuing advances in alloys are the availability of
materials, new manufacturing techniques, and the ability to test alloys before they
are ever produced. Most modern alloys are, in fact, preplanned using sophisticated
computer simulations, which help determine what properties the alloy will display.
14 1 Introduction

1.5.1 Alloys in Nuclear Reactors

Structural materials employed in reactor systems must possess suitable nuclear and
physical properties and must be compatible with the reactor coolant under the con-
ditions of operation. The most common structural materials employed in reactor
systems are stainless steel and zirconium alloys. Zirconium alloys have favourable
nuclear and physical properties (Kutty et al. 1999), whereas stainless steel has
favourable physical properties. Aluminium is widely used in low-temperature test
and research reactors; zirconium and stainless steel are used in high-temperature
power reactors. Zirconium is relatively expensive, and its use is therefore confined to
applications in the reactor core where neutron absorption is important.

1.5.2 Alloy Wheels

Alloy wheels are automobile (car, motorcycle and truck) wheels which are made
from an alloy of aluminium or magnesium (or sometimes a mixture of both). They
are typically lighter for the same strength and provide better heat conduction and
improved cosmetic appearance. Lighter wheels can improve handling by reducing
vehicle mass which helps to reduce fuel consumption. Better heat conduction can
help dissipate heat from the brakes, which improves braking performance in more
demanding driving conditions and reduces the chance of brake failure due to
overheating (Nunney 2006). Alloy wheels are also purchased for cosmetic pur-
poses as the alloys used are largely corrosion-resistant. This permits the use of
attractive bare-metal finishes, with no need for paint or wheel covers, and the
manufacturing processes allow intricate, bold designs. Magnesium alloy wheels, or
‘‘mag wheels’’ are sometimes used on racing cars, in place of heavier steel or
aluminium wheels, for better performance.

1.6 Significance of the Alloys Dealt With in this Research Work

Some of the commercially and industrially important metals and alloys have been
analyzed in this work and their significance are discussed briefly in the following
lines.

1.6.1 Cobalt Aluminium

Intermetallic alloys, including CoAl and NiAl, are of great importance since these
materials not only have good strength-to-weight ratio but also has excellent
corrosion and oxidation resistance, which make them good candidates for high-
1.6 Significance of the Alloys Dealt in this Research Work 15

temperature and soft magnetic applications (Wan et al. 2010). Among the
intermetallics, Cobalt aluminides are of a considerable technological interest for
high-temperature applications (Lee et al. 2004). Cobalt forms a stable B2 (CsCl)
structure with Al in a wide concentration range (Botton et al. 1996). This structure
can be considered as two interpenetrating primitive cubic sublattices, where each
Co atom has eight Al atoms as nearest neighbors and vice versa. Depending on the
concentration, CoAl alloys exhibit different mechanical and magnetic properties
(Kudryavtsev et al. 1998). CoAl is especially attractive for epitaxial growth due to
its low lattice mismatch with respect to GaAs (Wan et al. 2010). In combination
with semiconductor hetero structures, they are able to be integrated in a number of
devices, e.g., spin injectors and mirrors.
Cobalt-based super alloys consume most of the produced cobalt. The
temperature stability of these alloys makes them suitable for use in turbine
blades for gas turbines and jet aircraft engines. Cobalt-based alloys are also
corrosion and wear-resistant. Cobalt-based alloys are used in total joint
replacement applications where high strength and corrosion resistance are
necessary. Special cobalt–chromium–molybdenum alloys are used for prosthetic
parts such as hip and knee replacements (Michel et al. 1991). Cobalt alloys are
also used for dental prosthetics, where they are useful to avoid allergies to
nickel. Some high-speed steels also use cobalt to increase heat and wear-
resistance. The special alloys of aluminium, nickel, cobalt and iron, known as
Alnico, are used in permanent magnets (Luborsky et al. 1957).

1.6.2 Nickel Aluminium

NiAl, an intermetallic compound, is a promising material for aerospace

applications. It has a combination of important structural properties, such as high
mechanical strength, low density, high-melting point, high thermal conductivity
and excellent oxidation resistances (Choudry et al. 1998; Ng et al. 1997). It has
potential applications such as hot sections of gas turbine engines for aircraft
propulsion systems, coats under thermal barrier coating, electronic metallisation
compounds in advanced semiconductors and surface catalysts (Albiter et al. 2002).
Strong bonding between aluminium and nickel, which persists at elevated tem-
peratures yields excellent high-temperature properties and specific strength that are
competitive with those of super alloys and ceramics. Thus, these alloys offer new
opportunities for applications in gas turbines at temperatures higher than those
currently possible with conventional nickel-based super alloys. The advantages of
NiAl-based intermetallics are a high-melting point above 1,460C and high ther-
mal conductivity. Due to their excellent high-temperature properties NiAl-based
intermetallics are potential candidate materials for combustion chambers heat
shields and first row vanes in industrial gas turbines (Scheppe et al. 2002).
Polycrystalline NiAl exhibits a brittle–ductile transition at temperatures ranging
from 300 to 600C which are significantly lower than those of other intermetallic
16 1 Introduction

compounds. These properties have made NiAl-based alloy a promising candidate

in some high-temperature (HT) structural applications (Gaoa et al. 2005).
Fine grains of a nickel-aluminium alloy, known as Raney nickel are used in
many industrial processes. It is used as a heterogeneous catalyst in a variety of
organic syntheses, most commonly for hydrogenation reactions. Its structural and
thermal stability (i.e., the fact that it does not decompose at high temperatures)
allows its use under a wide range of reaction conditions (Carruthers 1986). Raney
nickel is used in a large number of industrial processes and in organic synthesis
because of its stability and high-catalytic activity at room temperature (Hauptmann
and Walter 1962).

1.6.3 Nickel Chromium

Nichrome wire, an alloy typically made of 80% nickel and 20% chrome, has
specific characteristics that make it ideal for certain uses. It is silvery-grey in color,
is corrosion resistant and has a high-melting point of about 1,400C. Its ability to
conduct electricity, withstand heat, its strength, flexibility and melting point
determine its uses and applications.
Nichrome wire is used to make coils for heating elements. Due to its relatively
high resistivity and resistance to oxidation at high temperatures, it is widely used
in heating elements, such as in hair dryers, electric ovens, toasters and irons.
Industrial uses include a wide variety of heating elements for ovens of all sizes,
and devices that must apply heat directly to surfaces, such as for sealing plastic
packages. Medical laboratories use nichrome wire loops somewhat like small
spoons for handling specimens because they withstand frequent and repeated
sterilisation, then cool rapidly. In industrial use, fine nichrome wire meshes serve
as filters for liquids at high temperatures. Since the wire heats up rapidly with even
a small amount of electricity, it makes safe igniters (sometimes called electric
matches) to light the fuses of fireworks and model rockets. A piece of nichrome
wire about an inch-and-a-half long connected to a length of lead wire allows the
operator to safely detonate the device from a distance with only a battery and a
switch. Fireworks professionals use nichrome igniters with timers at large displays.
NiCr thin films are widely used in several applications in microelectronics such as
thin film resistors, filaments and humidity sensors because of their relatively large
resistivity, more resistant to oxidation and a low-temperature coefficient of resis-
tance (Kwon et al. 2005). Ni80Cr20 matrix offers excellent high-temperature oxi-
dation/corrosion resistance and essential mechanical strength (Ding et al. 2007).
For dental clinical applications, the NiCr-based casting alloys are developed as an
alternative to gold-based alloys due to their corrosion resistance in oral environ-
ment (Huang 2003). Nichrome Ni80Cr20 wt% films are used for strain gauge
applications because of their high resistivity, low-temperature coefficient of
resistance, commercial availability and low-temperature dependence of gauge
factor (Chen et al. 2008).
1.6 Significance of the Alloys Dealt in this Research Work 17

1.6.4 Iron–Nickel

Iron group metals and binary alloys have a number of important industrial
applications (Fabio and Mascaro 2006). Fe–Ni alloys have attracted much atten-
tion due to their interesting mechanical and magnetic properties (Karayannis and
Moutsatsou 2006). Permalloy is a Fe–Ni alloy used in soft magnetic read/write
heads (Cacciamani et al. 2010). Fe–Ni-based alloy powders are interesting in their
applications as soft magnetic materials with low coercivity and high permeability
(Gheisari et al. 2009).
One of the important challenges of electrical motors is to increase the efficiency
to obtain more power with the same electrical energy consumed. The use of iron–
nickel alloys increases the efficiency by reducing the magnetic losses significantly
(Frederic et al. 2007). Fe–Ni thin films and multilayered structures are
materials for high-frequency devices, such as inductors or magneto impedance
effect (MI)-based magnetic field sensors (Kurlyandskaya et al. 2011).
Typical applications that are based on the low coefficient of thermal expansion
of Fe–Ni alloys include thermostatic bimetals, glass sealing, integrated circuit
packaging, cathode ray tube shadow masks, composite molds/tooling and mem-
branes for liquid gas tankers (Cacciamani et al. 2010). Applications based on the
soft magnetic properties include read-write heads for magnetic storage, magnetic
actuators, magnetic shielding and high-performance transformer cores. Due
to their unique low coefficient of thermal expansion and soft magnetic properties,
Fe–Ni alloys are used in several industrial applications (McCrea et al. 2003).

1.6.5 Sodium Chloride Doped with Silver

Silver alkali halides provide interesting model systems for the study of decom-
position processes in ionic solids. This is not only due to the absence of structural
phase transitions but also due to the invariance of the anion sublattice, which is not
involved in the demixing process. The phase separation is entirely confined to the
cationic system and the anions exhibit an almost rigid frame. Along with the strong
polarizability of silver ions, this feature guarantees that even single crystals are not
destroyed during demixing (Elter et al. 2005). The dynamics of Na1-xAgxCl show
that in the homogeneous phase, the doping of NaCl with silver ions leads to a
considerable softening of the lattice (Caspary et al. 2007).
Semiconductors in confined surroundings experience a very significant devel-
opment considering their importance in optoelectronic technologies. The semi-
conductors concerned are those having a direct and wide band gap conferring them
with a radiative character and consequently a rather considerable output of
photoluminescence, contrary to indirect gap semiconductors with a weaker output.
Photosensitivity is observed in ionic semiconductors such as sodium chloride
doped with silver (Madani et al. 2004).
18 1 Introduction

1.6.6 Aluminium Doped with Iron

The addition of iron as a dopant in aluminium for integrated circuit applications

substantially increases resistance to electro migration and creep. The amount of
iron utilised depends to an extent on the electrical requirements of the device, the
geometry of the device, the substrate composition and composition of overlying
layers.
Stress-induced grain boundary movement in aluminium lines used as connec-
tions in integrated circuits are substantially avoided by doping aluminium with
iron. Through this expedient not only is grain boundary movement is avoided but
the electro migration problems are decreased (Ryan et al. 1993).
Aluminium doped with iron is distinguished by high thermal-shock resistance
and, at temperatures of 800C has comparatively good mechanical properties. It
has mechanical properties which permit its use in components which are slightly
stressed mechanically. It has excellent shock resistance and can therefore be used
in those parts of thermal installations which are subject to frequent thermal
cycling, such as in particular as a casing or casing part of gas turbine or of turbo
charger or as a nozzle ring (Nazmy et al. 1995).
To meet the recent trends towards appealing, custom effects, many new effect
pigments have been developed in the last few years in the automotive industry in
recent past. Besides micas, flake-like particles have been doped with ultra thin layers
of metal oxides and launched into coating markets. Aluminium flakes, doped with
thin layers of iron oxide find much demand in this industry (Poth 2008).

1.7 Ball Milling

Most crystalline solids are composed of a collection of many small crystals or

grains, termed polycrystalline. The term nano crystalline materials (Gleiter 1989)
are used to describe those materials that have a majority of grain diameters in the
typical range from 1 to 50 nm (McHenry and Laughlin 2000). The nano crystalline
materials have received much attention as advanced engineering materials with
unique physical and mechanical properties. The mechanical properties of the nano
crystalline materials at room temperature have higher strength and toughness to
those of coarse-grained ones (Gleiter 1981).
Nano crystalline materials can be successfully synthesised by several tech-
niques, including inert gas condensation (Birringer et al. 1984), rapid solidification
(Inoue 1994), sputtering (Li and Smith 1989), crystallisation of amorphous phases
(Lu et al. 1995) and chemical processing (Kear and Strutt 1995). Among the
different options for preparations, the ball milling method has been considered the
most powerful tool for nanostructure materials because of its simplicity relatively
inexpensive equipment and the possibility of producing large quantities that can be
scaled up to several tons (Zhao et al. 2010).
1.7 Ball Milling 19

Fig. 1.2 Balls and powder

sample in a ball mill

A ball mill is a type of grinder and a cylindrical device used in grinding (or
mixing) materials like ores, chemicals and ceramic raw materials. Ball mills rotate
around a horizontal axis, partially filled with the material to be ground and the
grinding medium (Fig. 1.2). Different materials are used as media, including
ceramic balls, alumina balls and stainless steel balls. An internal cascading effect
reduces the material to a fine powder. Industrial ball mills can operate continu-
ously fed at one end and discharged at the other end. Large-to-medium sized ball
mills are mechanically rotated on their axis, but small ones normally consist of a
cylindrical capped container that sits on two drive shafts (pulleys and belts are
used to transmit rotary motion) (Rajagopal 2009).
Ball mills are also used in pyrotechnics and the manufacture of black powder,
but cannot be used in the preparation of some pyrotechnic mixtures such as flash
powder because of their sensitivity to impact. High-quality ball mills are poten-
tially expensive and can grind mixture particles to as small as 5 nm, enormously
increasing surface area and reaction rates. The grinding works on principle of
critical speed. The critical speed can be understood as the speed after which the
balls (which are responsible for the grinding of particles) start rotating along the
direction of the cylindrical device; thus causing no further grinding (Malik and
Singh 2010). Ball mills are used extensively in the Mechanical alloying process in
which they are not only used for grinding but also for cold welding as well, with
the purpose of producing alloys from powders.

1.7.1 Mechanism for the Formation of Nano Crystalline

Materials by the Ball Milling

The mechanism for formation of nano crystalline materials by the ball milling
technique has been summarised as the phenomenology of the grain size reduction
into three stages (Fecht 1995).
First stage: Plastic deformation is produced in the crystal lattices of the ball-
milled powders by slip and twinning. This deformation is localised in shear bands
containing a high-dense network of dislocation. The local shear instability of a
20 1 Introduction

crystal lattice can be triggered by the material heterogeneity and enhance insta-
bilities. These instabilities result from a non-uniform heat transfer during the
mechanically induced deformation of the milled powders. During this stage of
milling, the atomic level strain increases as a result of increasing the dislocation
density.
Second stage: Due to the successive accumulation of the dislocation density,
the crystals are disintegrated into sub-grains that are initially separated by low-
angle grain boundaries. The formation of these sub-grains is attributed to the
decrease of the atomic level strain.
Third stage: Further ball milling time leads to further deformation occurring in
the shear bands located in the unstrained parts of the powders which leads to sub-
grain size reduction so that the orientation of final grains become random in
crystallographic orientations of the numerous grains and hence, the direction of
slip varies from one grain to another.
In ball mills, the useful kinetic energy can be applied to the powder particles of
the reactant materials by
• Collision between the balls and the powders (Fig. 1.2)
• Pressure loading of powders pinned between milling media or between the
milling media and the liner
• Impact of the falling milling media
• Shear and abrasion caused by dragging of particles between moving milling
media
• Shock wave transmitted through crop load by falling milling media.

1.7.2 Effect of Materials of Milling Media

There are many types of milling media suitable for use in a ball mill, each material
having its own specific properties and advantages (Rajagopal 2009). Common in
some applications are stainless steel balls. While usually very effective due to their
high density and low contamination of the material being processed, stainless steel
balls are unsuitable for some applications, such as black powder and other flam-
mable materials require non-sparking lead, antimony, brass or bronze grinding
media. In some applications ceramic or flint grinding media is used. Ceramic
media are also very resistant to corrosive materials. High-density alumina media
are widely used to grind clay bodies, frits, glazes and other ingredients. It is more
expensive than silica media but is more efficient.
1.7 Ball Milling 21

Fig. 1.3 Laboratory ball mill

1.7.3 Laboratory Ball Mill

The laboratory ball mill is a low-energy ball mill which is used in the present
study. It is less expensive and operates with minimum maintenance requirements.
The vial is a cylinder made of stainless steel with a radius of 2.5 cm. Stainless steel
balls of different radii have been used as the milling media. The rotation speed is
about 200 rpm. It produces homogenous and uniform powders. The laboratory ball
mill used in our work is shown in Fig. 1.3.

1.8 X-Ray Diffraction

The modern understanding of metals and alloys, their structures, defects and
various properties would not be possible if their crystal structures had not been
revealed by XRD studies. XRD is a non-destructive technique for analyzing a wide
range of materials, including crystals, metals, minerals, polymers, thin film coating
and ceramics. Crystalline materials are characterised by the orderly, periodic
arrangements of atoms. The unit cell is the basic repeating unit that defines a
crystal. Parallel planes of atoms intersecting the unit cell are used to define
directions and distances in the crystal. These crystallographic planes are identified
by miller indices (Warren 1990). Diffraction from different planes of atoms
produces a diffraction pattern, which contains information about the atomic
arrangement within the crystal.
XRD is based on constructive interference of monochromatic X-rays and a
crystalline sample. These X-rays are generated by a cathode ray tube, filtered to
produce monochromatic radiation, collimated to concentrate and directed towards
the sample. The interaction of the incident rays with the sample produces
constructive interference (and a diffracted ray) when conditions satisfy Bragg’s
Law (nk = 2d sin h) (Bragg 1912). This law relates the wavelength of
22 1 Introduction

Fig. 1.4 Determination of

crystal structure through
X-rays

electromagnetic radiation to the diffraction angle and the lattice spacing in a

crystalline sample. These diffracted X-rays are then detected, processed and
counted.
Figure 1.4 illustrates how crystalline structure may be determined through XRD.
As the crystal and detector rotate, X-rays diffract at specific angles. The detector
reports the intensity (I) of X-ray photons as it moves. Angles of diffraction (where the
Bragg equation is satisfied) are marked by peaks. The peak height is a function of the
interaction of the X-rays with the crystal and the intensity of the source.

1.8.1 X-Ray Diffraction Methods

Classically, the two main ways of studying metals and alloys were metallography
(the examination of polished and etched surfaces) and cooling curves (looking for
discontinuities that indicated some sort of phase change). Both these methods
involved considerable skill and experience, and the results were not always
unambiguous. The introduction of XRD provided a much clearer, simpler and
more objective way of investigation. XRD is now a common technique for the
study of crystal structures and atomic spacing.
In terms of the specimen handled, two methods can be identified,
1. Powder diffraction
2. Single-crystal diffraction
In the former, the specimen is a collection of crystallites. Since these fragments
are completely randomly oriented the incident X-ray beam meets with every
possible lattice plane, oriented in all directions. Whereas in the single-crystal
method, the whole specimen is a single piece, without any discontinuity in the
lattice arrangements (Tareen and Kutty 2001). All diffraction methods are based
on generation of X-rays in an X-ray tube. These X-rays are directed to the sample,
and the diffracted rays are collected. A key component of all diffraction is the
angle between the incident and diffracted rays.
1.8 X-Ray Diffraction 23

1.8.2 Diffractometers

A typical X-ray diffractometer consists of a source of radiation (X-ray tube), a

monochromator to choose the wavelength, slits to adjust the shape of the beam,
sample and a detector. In a more complicated apparatus also a goniometer can be
used for fine adjustment of the sample and the detector positions (Azaroff 1968).

1.8.3 Powder X-Ray Diffraction Instrumentation

The powder method essentially has two different ways of registering the diffracted
X-rays (Tareen and Kutty 2001).
1. The whole diffraction pattern is recorded simultaneously on a photographic film
called the powder photographic method.
2. The diffraction pattern is scanned by a counter device, or a solid-stated semi-
conductor detector. The counter or detector registers the diffracted beam in
successive stages, away from the direct beam.
In X-ray diffractometers, X-rays are generated in a cathode ray tube by
heating a filament to produce electrons, accelerating the electrons towards a
target by applying a voltage, and bombarding the target material with electrons.
When electrons have sufficient energy to dislodge inner shell electrons of the
target material, characteristic X-ray spectra are produced. These spectra consist
of several components, the most common being Ka and Kb. Ka consists, in part,
of Ka1 and Ka2. Ka1 has a slightly shorter wavelength and twice the intensity as
Ka2 (Stout and Jensen 1989). The specific wavelengths are characteristic of the
target material (Cu, Fe, Mo, and Cr). The monochromatic radiation required for
the powder method is usually the Ka1a2 doublet, monochromated by crystal
reflection or by the use of a filter whose K absorption wavelength falls between
the Ka and the Kb wavelengths. The Cu Ka1a2 doublet (k = 1.542 Å ´ ) with a
´
Ni filter kk = 1.488 Å is probably used more than any other source (Warren
1990). These X-rays are collimated and directed onto the sample. The sample is
mounted on a goniometer (Shirane et al. 2002) and gradually rotated while
being bombarded with X-rays, producing a diffraction pattern of regularly
spaced spots known as reflections. As the sample and detector are rotated, the
intensity of the reflected X-rays is recorded. When the geometry of the incident
X-rays impinging the sample satisfies the Bragg Equation, constructive inter-
ference occurs and a peak in intensity is seen. A detector records and processes
this X-ray signal and converts the signal to a count rate which is then output to
a device such as a printer or computer monitor.
24 1 Introduction

1.8.4 Single-Crystal X-Ray Diffraction Instrumentation

The foremost essential criteria for single-crystal XRD are to obtain an adequate
crystal of the material under study. The crystal should be sufficiently large (typ-
ically larger than 0.1 mm in all dimensions), pure in composition and regular in
structure, with no significant internal imperfections such as cracks or twinning.
The crystal is placed in an intense beam of X-rays, usually of a single wave-
length (monochromatic X-rays), producing the regular pattern of reflections.
Molybdenum is the most common target material for single-crystal diffraction,
with MoKa radiation = 0.7107Å. These X-rays are collimated and directed onto
the sample. When the geometry of the incident X-rays impinging the sample
satisfies the Bragg Equation, constructive interference occurs. A detector records
and processes this X-ray signal and converts the signal to a count rate which is
then output to a device such as a printer or computer monitor. Modern single-
crystal diffractometers use CCD (charge-coupled device) technology to transform
the X-ray photons into an electrical signal which are then sent to a computer for
processing.
Single-crystal diffractometers use either three- or four-circle goniometers.
These circles refer to the four angles (2h, v, u, and X) (Shirane et al. 2002) that
define the relationship between the crystal lattice, the incident ray and the detector
as shown in Fig. 1.5. Samples are mounted on thin glass fibers using an epoxy or
cement. The thin glass fiber is attached to brass pins and mounted onto goniometer
heads. The goniometer head and sample are then affixed to the diffractometer.
Samples are centered by viewing the sample under an attached microscope and are
placed under the cross-hairs for all crystal orientations. Once the crystal is centred,
a preliminary rotational image is often collected to screen the sample quality and
to select parameters for later steps. As the crystal is gradually rotated, previous
reflections disappear and new ones appear, the intensity of every spot is recorded
at every orientation of the crystal. Multiple data sets may have to be collected, with
each set covering slightly more than half a full rotation of the crystal and typically
containing tens of thousands of reflections (Massa 2004).
An automatic collection routine can then be used to collect a preliminary set
of frames for determination of the unit cell. Reflections from these frames are auto-
indexed to select the reduced primitive cell and calculate the orientation matrix
(which relates the unit cell to the actual crystal position within the beam). These
data are combined computationally with complementary chemical information to
produce and refine a model of the arrangement of atoms within the crystal and
converted to the appropriate crystal system and Bravais lattice (Giacovazzo 2002).
The final, refined model of the atomic arrangement is the essential crystal
structure.
1.8 X-Ray Diffraction 25

Fig. 1.5 Goniometer

1.8.4.1 Corrections for Background, Absorption

After the data have been collected, corrections for instrumental factors,
polarisation effects of X-ray absorption (Glusker et al. 1994) and (potentially)
crystal decomposition must be applied to the entire data set. This integration
process also reduces the raw frame data to a smaller set of individual integrated
intensities. These correction and processing procedures are typically part of the
software package which controls and runs the data collection.
Single-crystal XRD is most commonly used for precise determination of a unit
cell, including cell dimensions and positions of atoms within the lattice. Bond-
lengths and angles are directly related to the atomic positions. A single, robust,
optically clear sample, generally between 50 and 250 microns in size is essential.
Data collection generally requires between 24 and 72 h.

1.9 Grain Size Analysis from X-Ray Diffraction

The size of the grains in a polycrystalline material has more effect on the prop-
erties of the material, for example, the hardness of a metal or alloy increases with
decrease in the grain size (Cullity and Stock 2001). This dependence of properties
on grain size makes the measurement of grain size important in the control of most
metal forming operations.
In this book the grain size has been reported for some metals which has been
analyzed from XRD. The grain morphology was examined by SEM. The software
GRAIN written by Dr.R.Saravanan was used to estimate approximate grain sizes
from XRD (Saravanan, GRAIN software). The grain size is analyzed using full
26 1 Introduction

width at half maximum of the powder XRD peaks. The Debye–Scherrer formula
given in Eq. 1.1has been used to calculate the particle size.
Kk
s¼ ð1:1Þ
b cos h
where K is the shape factor, k is the X-ray wavelength, typically 1.54 Å, b is the
line broadening at half the maximum intensity (FWHM) in radians and h is the
Bragg angle (Patterson 1939) s is the mean size of the ordered (crystalline)
domains, which may be smaller or equal to the grain size. The dimensionless shape
factor has a typical value of about 0.9, but varies with the actual shape of the
crystallite. The Scherrer equation is limited to nano-scale particles. It is not
applicable to grains larger than about 0.1 lm, which precludes those observed in
most metallographic and ceramographic microstructures.
It is important to realise that the Scherrer formula provides a lower bound on
the particle size. The reason for this is that a variety of factors can contribute to the
width of a diffraction peak; besides particle size, the most important of these are
usually inhomogeneous strain and instrumental effects. If all of these other
contributions to the peak width were zero, then the peak width would be deter-
mined solely by the particle size and the Scherrer formula would apply. If the other
contributions to the width are non-zero, then the particle size can be larger than
that predicted by the Scherrer formula, with the ‘‘extra’’ peak width coming from
the other factors.

1.10 Scanning Electron Microscope

The SEM uses a focused beam of high-energy electrons to generate a variety of

signals at the surface of solid specimens (Malik and Singh 2010). The signals
derived from electron-sample interactions reveal information about the sample
including external morphology (texture), chemical composition and crystalline
structure and orientation of materials making up the sample. In most applications,
data are collected over a selected area of the surface of the sample, and a two-
dimensional image is generated that displays spatial variations in these properties.
Areas ranging from approximately 1 cm to 5 microns in width can be imaged in a
scanning mode using conventional SEM techniques (magnification ranging from
20X to approximately 30,000X, spatial resolution of 50–100 nm). The SEM is also
capable of performing analyses of selected point locations on the sample;
this approach is especially useful in qualitatively or semi-quantitatively deter-
mining chemical compositions, crystalline structure and crystal orientations
(Reimer 1998).
SEM is also widely used to identify phases based on qualitative chemical
analysis and/or crystalline structure. Backscattered electron images (BSE) can be
used for rapid discrimination of phases in multiphase samples. SEMs equipped
1.10 Scanning Electron Microscope 27

with diffracted backscattered electron detectors (EBSD) can be used to examine

microfabric and crystallographic orientation in many materials.

1.11 Fundamental Principles of Scanning Electron Microscopy

Accelerated electrons in SEM carry significant amounts of kinetic energy, and this
energy is dissipated as a variety of signals are produced by electron-sample
interactions when the incident electrons are decelerated in the solid sample. These
signals include secondary electrons (that produce SEM images), backscattered
electrons (BSE), diffracted backscattered electrons (EBSD that are used to
determine crystal structures and orientations of minerals), photons (characteristic
X-rays that are used for elemental analysis and continuum X-rays), visible light
(cathodoluminescence-CL) and heat. Secondary electrons and backscattered
electrons are commonly used for imaging samples, secondary electrons are most
valuable for showing morphology and topography of samples and backscattered
electrons are most valuable for illustrating contrasts in composition in multiphase
samples (i.e., for rapid phase discrimination). X-ray generation is produced by
inelastic collisions of the incident electrons with electrons in discrete orbitals
(shells) of atoms in the sample. As the excited electrons return to lower energy
states, they yield X-rays that are of a fixed wavelength (that is related to the
difference in energy levels of electrons in different shells for a given element).
Thus, characteristic X-rays are produced for each element in a mineral that is
‘‘excited’’ by the electron beam (Goldstein 2003).
SEM analysis is considered to be ‘‘non-destructive’’, i.e., X-rays generated by
electron interactions do not lead to volume loss of the sample, so it is possible to
analyze the same materials repeatedly.

References

Albiter A, Bedolla E, Perez R (2002) Mater Sci Eng A 328:80–86

Almanza R, Hernández P, Martínez I, Mazari M (2009) Sol Energ Mater Sol Cells 93:1647–1651
Ares JR, Guégan P, Cuevas FA (2005) Acta Materialia 53:2157–2167
Audi G (2003) Nucl Phys A (Atomic Mass Data Center) 729:3–128
Auxer W (1986) Electrochemical Society, pp 49–54
Azaroff LV (1968) Elements of X-ray crystallography, Chap. 13. McGraw Hill, New York
Birringer R, Gleiter H, Klein HP, Marquardt P (1984) Phys Lett A 102:356–369
Boonyongmaneerat Y, Sukjamsri C, Sahapatsombut U, Saenapitak S, Sukkasi S (2011) Appl
Energ 88:909–913
Botton GA et al (1996) Phys Rev B 54:1682–1691
Bragg WL (1912) Nature 90:410–410
Cacciamani G, Dinsdale A, Palumbo M, Pasturel A (2010) Intermetallics 18:1148–1162
Carruthers W (ed) (1986) Some modern methods of organic synthesis. Cambridge University
Press, Cambridge, pp 413–414
28 1 Introduction

Caspary D, Eckold G, Elter P, Gibhardt H, Güthoff F, Demmel F, Hoser A, Schmidt W, Ding CH,
Yang ZM, Zhang HT, Guo YF, Zhoua JN (2007) Compost A 38:48–352
Chen R, Zuo D, Sun Y, Li D, Lu W, Wang M (eds) (2008) Key engineering materials, vol 375.
Trans Tech Publications, Switzerland, pp 690–694
Choudry MS, Dollar M, Eastman JA (1998) Mater Sci Eng A256:25–33
Collins DM (1982) Nature 298:49–51
Crans DC (2000) J Inorg Biochem 80:123–131
Crans DC, Smee JJ, Gaidamauskas E, Yang L (2004) Chem Rev 104:849–902
Cullity BD, Stock SR (2001) Elements of X-ray diffraction, Chap. 14, 3rd edn. Prentice Hall,
Upper Saddle
Ding CH, Yang ZM, Zhang HT, Guo YF, Zhoua JN (2007) Compos Part A 38:348–352
Dobrzánski LA, Tánski T, Čížek L, Brytan Z (2007) J Mater Process Technol 192:567–574
Dobson KD, Hodes G, Mastai Y (2003) Sol Energ Mater Sol Cells 80:283–296
Dorward RC, Pritchett T (1988) Mater Des 9(2):63–69
El-Eskandarany MS (2001) Mechanical alloying for fabrication of advanced engineering
materials, Chap. 1. William Andrew, New York
Elter P, Eckold G, Gibhardt H, Schmidt W, Hoser A (2005) J Phys Condens Matter 17:6559–
6573
Engelhard T, Jones ED, Viney I, Mastai Y, Hodes G (2000) Thin Solid Films 370:101–105
Eveloy V, Ganesan S, Fukuda Y, Wu J, Pecht MG (2005) IEEE Trans Compon Packag Technol
28:884–894
Fabio RB, Mascaro LM (2006) Surf Coat Technol 201:1752–1756
Fecht HJ (1995) NanoStruc Mater 6:33–42
Francis J (ed) (2008) Philosophy of mathematics. Global publishing house, New York, p 56
Frederic B, Thierry W, Herve F (2007) Electrical insulation conference and electrical
manufacturing expo, pp 394–401
Futsuhara M,Yoshioka K,Takai O (1998) Thin Solid Films 322:274–281
Gallagher R, Ingram P (eds) (2001) New coordination science, chemistry for higher tier ingram,
3rd edn. Oxford University Press, New York, p 56
Gaoa Q, Guob JT, Huaib KW, Zhanga JS (2005) Mater Lett 59:2859–2862
Gheisari KH, Javadpoura S, Oh JT, Ghaffari M (2009) J Alloys Compd 472:416–420
Giacovazzo C (2002) Fundamentals of crystallography issue 7 of international union of
crystallography texts Oxford science publication, Chap. 5, 2nd edn. Oxford University Press,
Oxford
Gleiter H (1981) Deformation of polycrystals. In: Hansen N, Horsewell A, Leffers T, Lilholt H
(eds) Proceedings second riso intrnational symposium metallurgy and materials science,
Roskilde, p 15
Gleiter H (1989) Prog Mat Sci 33:223–315
Glusker JP, Lewis M, Rossi M (eds) (1994) Crystal structure analysis for chemists and biologists,
volume 16 of methods in stereochemical analysis. Wiley, New York, p 261
Goldstein J (2003) Scanning electron microscopy and X-ray microanalysis, Chap. 6. Kluwer
Adacemic/Plenum Pulbishers, New York
Hansen NK, Coppens P (1978) Acta Cryst A34:909–921
Hauptmann H, Walter W (1962) Chem Rev 62:347–404
Heinrich G, Sgler T, Rosiwal SM, Singer RF, Ley L (1996) Diam Relat Mater 5:304–307
Hogan CM (1969) Phys Rev 188:870–874
Hort N, Huang Y, Fechner D, Störmer M, Blawert C, Witte F, Vogt C, Drücker H, Willumeit R,
Kainer KU, Feyerabend F (2010) Acta Biomater 6:1714–1725
http://itp.nyu.edu/physcomp/uploads/battery_nimh.jpg)
http://en.wikipedia.org/wiki/File:Vanadinite2_sur_goethite_ (Maroc).jpg
http://image.made-in-china.com/2f0j00DvhtLbGPOekg/Grinding-Ceramic-Balls.jpg
http://www.shubhmets.in/full-images/663342.jpg
Huang HH (2003) Biomaterials 24:1575–1582
References 29

Hunziker W, Voyt W, Melchior H (1996) Proceedings of the 1996 IEEE 46th electronic
components and technology conference, Orlando, Florida, pp 8–12
Inoue A (1994) Mater Sci Eng A 179:57–61
Ivey DG (1998) Micron 29(4):281–287
Josef P, Dana P, Jaroslav B et al (2004) Anal Appl Pyrolysis 71:739–746
Kainer KU, Sb (2000) International congress magnesium alloys and their application, Mnichov,
pp 534–608
Karayannis VG, Moutsatsou AK (2006) J Mater Process Technol 171:295–300
Kartal G, Timur S, Urgen M, Erdemir A (2010) Surf Coat Technol 204:3935–3939
Kazacos MS, Menictas C (1997) Telecommunications energy conference, 19th international,
pp 463–471
Kear BH, Strutt PR (1995) Nanostruct Mater 6:227–236
Kittel C (1996) Introduction to solid state physics, 7th edn. Wiley, India
Kittel C (2007) Introduction to solid state physics, Chap. 6, 7th edn. Wiley, India
Kudryavtsev YV et al (1998) J Appl Phys 83:1575–1581
Kurlyandskaya GV, Bhagat SM, Svalov AV, Fernandez E, Arribas AG, Barandiaran JM (2011)
Solid State Phenom 168:257–260
Kutty TRG, Ravi K, Ganguly C (1999) J Nucl Mater 265:91–99
Kwon Y, Kim N, Choi G, Lee W, Seo Y, Park J (2005) Microelectron Eng 82:314–320
Lee YS, Lee CH, Kim KW, Shin HJ, Lee YP (2004) J Magn Magn Mater 272:2151–2153
Li ZG, Smith DJ (1989) Appl Phys Lett 55:919–921
Li SJ, Cui TC, Hao YL, Yang R (2008) A Biomater 4:305–317
Lide DR (ed) (1999) In: CRC handbook of chemistry and physics, 80th edn. CRC Press,
New York, pp 4–32
Lonyuk B, Apachitei I, Duszczyk J (2007) Surf Coat Technol 201:8688–8694
Lu K, Wei WD, Wang JT (1995) Scr Metall Mater 24:2319–2323
Luborsky FE, Mendelsohn LI, Paine TO (1957) J Appl Phys 28:344–351
Lynch CT (1974) CRC handbook of materials science. Vol 1,CRC press
Madani S, Hocine K, Madjid B, Abdelhamid C, Mohamed AB, Sofiane M (2004) Physica E
23:217–220
Maeshima T, Nishida M (2004) Mater Trans 45:1096–1100
Malik HK, Singh AK (2010) Engineering physics, Tata McGraw-Hill, New Delhi, p 6
Markovsky PE, Semiatin SL (2010) J Mater Process Technol 210:518–528
Massa W (2004) Crystal structure determination, Chap. 6, 2nd edn. Springer, Berlin
McCrea JL, Palumbo G, Hibbard GD, Erb U (2003) Rev Adv Mater Sci 5:252–258
McHenry ME, Laughlin DE (2000) Acta Mater 48:223–238
Mehta DS, Masood SH, Song WQ (2004) J Mater Process Technol 156:1526–1531
Michel R, Nolte M, Reich M, Löer F (1991) Systemic effects of implanted prostheses made of
cobalt–chromium alloys, Archives of orthopaedic and trauma surgery, 110:61–74
Monfared HH, Alavi S, Bikas R, Mahedpour M, Mayer P (2010) Polyhedron 29:3355–3362
Mordike BL, Ebert T (2001a) J Mater Process Technol 117:381–385
Mordike BL, Ebert T (2001) Mater Sci Eng A 302:37–45
Nazmy M, Noseda C, Staubli M (1995) Switzerland, United States (19), Patent Number 5411702,
May 2
Newbury DE (1986) Advanced scanning electron microscopy and X-ray microanalysis, Chap. 2.
Springer, Berlin
Ng C, Simkin BA, Crimp MA (1997) Mater Sci Eng A239:150–156
Nickel Magazine (2007), 22(2) March issue
Nithya D, Kalaiselvan R (2011) Physica B 406:24–29
Noring J et al (1993) In: Proceedings of the symposium on batteries and fuel cells for stationary
and electric vehicle applications vols 93–98 of proceedings (electrochemical society), The
Electrochemical Society, pp 235–236
Nowosielski R, Babilas R, Ochin P, Stoktosa Z (2008) Arch Mater Sci Eng 30:13–16
30 1 Introduction

Nunney MJ (ed) (2006) Light and heavy vehicle technology, 4th edn. Butterworth-Heinmann,
London, p 454
Oshima T, Kajitani M, Okuno A (2004) Int J Appl Ceram Technol 1269–1276
Paik J, Veen SVD, Duran A, Collette M (2005) Thin-Walled Str 43:1550
Patterson A (1939) Phys Rev 56:978–982
Pawlikowski M (1981) Infrared Phys 21:181–187
Petersson IU, Löberg JEL, Fredriksson AS, Ahlberg EK (2009) Biomaterials 30:4471–4479
Pillai SO (2007) Solid state physics, New age international (p) limited, New Delhi, Reprint, p 288
Poth U (2008) Automotive coatings formulation: chemistry, physics and practices European
coatings tech files, Vincentz, p 152
Proffen TH, Billinge SJL (1999) J Appl Cryst 32:572–575
Qin Y, Wilcox GD, Liu C (2010) Electrochim Acta 56:183–192
Rajagopal K (2009) Textbook of Engineering Physics, Part II, Publisher-PHI learning Pvt. Ltd,
New Delhi, p 168
Reimer L (1998) Scanning electron microscopy, physics of image formation and micro analysis,
Chap. 1. Springer, Berlin
Rietveld HM (1969) J Appl Crystallogr 2:65–71
Ryan VW, Schutz RJ, Warren (1993) United States Patent (19), Patent No: 5243221 September 7
Sambamurthy K (2007) Pharmaceutical engineering, new age international, p 427
Saravanan R http:\\www.saraxraygroup.net. Grain software
Scheppe F, Sahm PR, Hermann W, Paul U, Preuhs J (2002) Mater Sci Eng A329:596–601
Schwingel D, Seeliger HW, Vecchionacci C, Alwes D, Dittrich J (2007) Acta Astronaut 61:326–330
Setoyama D, Matsunaga J, Muta H, Uno M, Yamanaka S (2004) J Alloys Compds 381:215–220
Shashwati S, Muthe KP, Joshi N (2004) Sens Actuators B 98:154–159
Shirane G, Shapiro SP, Tranquada JM (2002) Neutron scattering with a triple-axis spectrometer:
basic techniques, Cambridge University Press, Cambridge, p 88
Skyllas-Kazacos M (2003) J Power Sour 124:299–302
Sparrow G (1999) Iron—the elements-group 1, the elements series. Marshall Cavendish, p 4
Stout GH, Jensen LH (1989) X-ray structure determination, Chap. 1, 2nd edn. Wiley-
Interscience, New York
Suryanarayana C (2004) Mechanical alloying and milling, Chap. 4, vol 180 of Mechanical
engineering. Marcel Dekker, New York
Sutou Y, Yamauchi K, Takagi T, Maeshima T, Nishida M (2006) Mater Sci Eng A 438:64–69
Tareen JAK, Kutty TRN (2001) A basic course in crystallography, chemistry series, Universities
Press, p 149
Tian MY, Qing GZ, Hong XW, Jian H (2006) Trans Nonferrous Met Soc China 16:693–699
Tsiulyanu D, Mariana S, Miron V, Liess HD (2001) Sens Actuators B 73:35–39
Tsiulyanu D, Marian S, Liess HD, Eisele I (2004) Sens Actuators B 100:380–386
Walawalkar R, Apt J, Mancini R (2007) Energ Policy 35(4):2558–2568
Wan Q, Hey R, Trampert A (2010) 16th International conference on microscopy of
semiconducting materials, IOP Publishing J Phys Conference Series 209, 012023
Wang RM, Eliezer A, Gutman E (2002) Mater Sci Eng A344:279–287
Warren BE (1990) X-ray diffraction, 2nd edn. reprint, Courier Dover Publications, New York,
p 16
Wolf MJ, Engelmann G, Dietrich L, Reichl H (2006) Nucl Instrum Method Phys Res A 565:290–
295
Zhang X, Suhl H (1985) Phys Rev A 32:2530–2533
Zhang Y, Liu X, Xie G, Yu L, Yi S, Hua M, Huang C (2010) Mater Sci Eng B 175:164–171
Zhao KY, Li CJ, Tao JM, Ng DHL, Zhu XK (2010) J Alloys Compds 504S:306–310
Zumdahl SS (2007) Chemical principles,6th edn. Cengage Publications, p 33
Chapter 2
Charge Density Analysis from X-Ray
Diffraction

Abstract This chapter provides a survey of current applications of X-ray dif-

fraction techniques in crystal structure analysis, with some focus on recent
advances that have been made in the scope and potential for carrying out crystal
structure determination directly from diffraction data. The basic concepts of crystal
structure analysis, Rietveld refinement, and the concepts used for the estimation
and analysis of charge density in a crystal are discussed. The more reliable models
for charge density estimation like multipole formalism and maximum entropy
method are discussed in detail. The local structural analysis technique atomic pair
distribution function is also discussed.

2.1 Introduction

Knowledge of the atomic scale geometrical structure of matter is a prerequisite for

understanding and predicting the properties of technologically and scientifically
important materials. The geometrical structure of a material not only consists of
the time and space averaged periodic conformation of atoms in an idealized crystal
lattice but also the microstructure caused by imperfections, dislocations, and all
kinds of disorders that are often responsible for interesting properties of the
material under investigation (Dinnibier 2008).
Establishing the structural properties of metals and alloys is essential for
understanding their function and properties. In order to understand the properties
of materials and to improve them, the atomic structure has to be known. The
advent of X-ray methods to analyze metals was a great boon to metallurgists
(Weiss 1925). The introduction of X-ray diffraction provided a much clearer,
simpler, and more objective way of investigation.
Modern crystallography is not only about knowing the positions of the atoms
and the corresponding bond distances, angles, and related features within a
material, but the most important and interesting aspect of this technique is the

R. Saravanan and M. Prema Rani, Metal and Alloy Bonding: An Experimental Analysis, 31
DOI: 10.1007/978-1-4471-2204-3_2,  Springer-Verlag London Limited 2012
32 2 Charge Density Analysis from X-Ray Diffraction

correlation of structural features with physical properties. This can include optical,
magnetic, and conductive effects as well as the structural behavior. The resulting
knowledge increases our understanding of the complex, underlying processes,
ultimately aiding the design of new materials in which the desired chemical or
physical properties are enhanced.

2.2 X-Ray Diffraction

The field of crystallography developed to great heights after the discovery of X-rays
by Wilhelm Conrad Röntgen in 1895 (Novelline and Squire 2004). The first X-ray
experiment was performed by Max von Laue in Munich in 1912 (Purrington 1997).
At the time of von Laue’s work, there was some experimental evidence that X-rays
were high energy particles; other data indicated that X-rays might be waves. Von
Laue surmised that, if X-rays were waves, they would have rather short wavelengths
(on the order of 1 9 10-10 m) and the dimensions of the objects in crystals would be
the appropriate size to produce the phenomenon of diffraction. He exposed a crystal
of copper sulfate to X-rays and recorded the diffraction pattern on a piece of pho-
tographic film, the experimental setup similar to that shown in Fig. 2.1.
This experiment proved the wave nature of X-rays and began the exploration of
molecular structure by X-ray diffraction methods. Modern X-ray data collection is
substantially the same as von Laue’s experiment though various kinds of electronic
detectors are used rather than photographic film.
An X-ray photo contains two important pieces of information about the crystal
structure, the shape and size of the unit cell and the distribution of electron density
throughout the unit cell. A useful explanation of von Laue’s experiment was
formulated by Bragg (1913) in Cambridge, in 1912. Bragg considered the dif-
fractions to arise from ‘reflections’ of separate X-ray waves from parallel planes of
electron density.

2.2.1 Bragg’s Equation

X-ray diffraction results from the interaction between X-rays and electrons of
atoms. Depending on the atomic arrangement, interfaces between the scattered
rays are constructive when the path difference between two diffracted rays differs
by an integral number of wavelengths. This selective condition is described by
Bragg’s equation also called ‘‘Bragg’s law’’ (Bragg 1912).
Crystals consist of a periodic arrangement of atoms or molecules that form a
crystal lattice. In such an arrangement of atoms numerous planes run in different
directions through the lattice points (atoms, molecules), horizontally, vertically,
and diagonally which are the lattice planes. All the planes parallel to a lattice plane
are also lattice planes and are set at a distance apart from each other which is
2.2 X-Ray Diffraction 33

Fig. 2.1 X-rays passing

through a crystal and falling
on a photographic plate

called the lattice plane distance ‘d’. Bragg diffraction occurs when electromagnetic
radiation or subatomic particle waves with wavelength comparable to atomic
spacings are incident upon a crystalline sample, scattered in a specular fashion by
the atoms in the system, undergo constructive interference. When the scattered
waves interfere constructively, they remain in phase since the path length of each
wave is equal to an integer multiple of the wavelength. The path difference
between two waves undergoing constructive interference is given by 2d Sin h,
where h is the scattering angle. This leads to Bragg’s law which describes the
condition for constructive interference from successive crystallographic planes
(h,k,l) of the crystalline lattice (Azaroff 1968).
In Fig. 2.2, two monochromatic X-ray beams of wavelength ‘k’ are represented
by wave 1 and wave 2. The spacing between the atomic planes occurs over the
distance, d. Wave 2 reflects off at the upper atomic plane at an angle h equal to its
angle of incidence. Similarly, wave 1 reflects off at the lower atomic plane at the
same angle h. although wave 1 is in the crystal, it travels a distance of 2d farther
than wave 2. Hence the path difference between waves 1 and 2 is 2d. If this
distance 2d is equal to an integral number of wavelengths nk, then waves 1 and 2
will be in phase on their exit from the crystal and constructive interference will
occur. If the distance 2d is not an integral number of wavelengths, then destructive
interference will occur and the waves will not be as strong as when they entered
the crystal. Thus, the condition for constructive interference to occur is nk = 2d.
In figure, BC ¼ d sin h; BD ¼ d sin h
The path difference ¼ BC þ BD
ð2:1Þ
BC þ BD ¼ 2d sin h
Thus, nk ¼ 2d Sin h
This is known as Bragg’s Law (Bragg 1913) for X-ray diffraction.
On the basis of Bragg’s Law, by measuring the angle ‘h’, the wavelength ‘k’,
the chemical elements can be determined, if the lattice plane distance ‘d’ is known,
or, if the wavelength ‘k’ is known, the lattice plane distance ‘d’ and thus the
crystalline structure can be determined. This provides the basis for two measuring
34 2 Charge Density Analysis from X-Ray Diffraction

Fig. 2.2 Schematic representation of X-ray reflection from two parallel planes of a crystal

techniques for the quantitative and qualitative determination of chemical elements

and crystalline structures, depending on whether the wavelength ‘k’ or the 2d-
value is identified by measuring the angle ‘h’ (George 2006).

2.2.2 Electron Density

According to the theory of quantum mechanics an electron’s position can only be

described statistically. The probability of finding an electron at one point or another
can be calculated. This calculation produces a quantity called electron density, a
number that tells us the relative probability of finding an electron at a particular point
in space. Quantum mechanics also says that an electron can be viewed as a stationary
wave, or, as a cloud of negative charges. The electron density is a periodic function
of position in the crystal, rising to a maximum at the point where an atom is located
and dropping to a low value in the region between atoms (Cullity and Stock 2001).
An electron’s wave characteristics are described mathematically by its orbital
or wave function, w. A wave function assigns a number to each point in space, and
the numbers oscillate so that they are positive at some locations and negative at
others. The electron density function, q, is equal to W2. This guarantees that q will
always have a positive value, and this value expresses the relative probability of
finding an electron at a particular location.

W $ W2 ¼ q ð2:2Þ
wave function probability
ðprobability amplitudeÞ ðor electron densityÞ
A ‘particle’ electron can also be described by a ‘wave’ function or orbital. The
orbital, whether atomic or molecular, covers a region of space and does not move.
A moving electron looks like a stationary cloud of charges. The physical
2.2 X-Ray Diffraction 35

interpretation of the electron density function q(r) is that q dr is the probability of

finding an electron in a volume element dr, i.e., electron density in this volume.

2.2.3 Structure Factor

The crystal structure, which is determined by the lattice parameters and the atomic
positions within the unit cell, is an essential element for the diffraction intensities.
Each atom scatters according to its electronic distribution following dependence on
sin h/k. When there is more than one atom per unit cell, the interference between the
waves scattered by each atom is considered as the diffraction unit. The addition of
the waves scattered by each of the atoms of the unit cell, considering that there is
coherence in time and space is called the structure factor (Marin and Dieguez 1999).
The structure factor Fhkl is the resultant of j waves scattered in the direction of
the reflection hkl by the j atoms in the unit cell (Stout and Jensen 1968).
The structure factor may be expressed as
X   
Fhkl ¼ Fhkl expðiahkl Þ ¼ fj exp 2pi hxj þ kyj þ lzj ð2:3Þ
j
X    X   
Fhkl ¼ fj cos 2p hxj þ kyj þ lzj þ i fj sin 2p hxj þ kyj þ lzj ð2:4Þ
j j

Fhkl ¼ Ahkl þ iBhkl ð2:5Þ

where the sum is over all atoms in the unit cell, xj, yj, zj are the positional
coordinates of the jth atom, fj is the scattering factor of the jth atom, and ahkl is the
phase of the diffracted beam.
The scattering factor of the atom is,
amplitude of the radiation scattered from the atom
fj ¼ ð2:6Þ
amplitude of the radiation scattered from a single electron
The structure factor describes the way in which an incident beam is scattered by
the atoms of a crystal unit cell, taking into account the different scattering power of
the elements through the term fj. Since the atoms are spatially distributed in the
unit cell, there will be a difference in phase when considering the scattered
amplitude from two atoms. This phase shift is taken into account by the complex
exponential term. The atomic form factor or scattering power, of an element
depends on the type of radiation considered as electrons interact with matter
through different processes.
If the summation over discrete atoms in the following structure factor
expression
36 2 Charge Density Analysis from X-Ray Diffraction

X
Fhkl ¼ fj e2piðhxj þkyj þlzj Þ ð2:7Þ

is replaced by an integration of a continuous, cyclic electron density function, q, an

expression is obtained that is of the form of a Fourier Transform
Z
Fhkl ¼ qð x:y:zÞe2piðhxj þkyj þlzj Þ dV ð2:8Þ
V

This implies that the electron density is the Fourier Transform of the Structure
Factor. Likewise, the Structure Factor is the Fourier Transform of the electron
density.
Z
qð x:y:zÞ ¼ Fhkl e2piðhxþkyþlzÞ dV ð2:9Þ
V

While this equation is a useful depiction of the relationship between electron

density and the structure factor, it is usually used in the form of a summation
1 XXX
qðx; y; zÞ ¼ Fhkl e2piðhxþkyþlzÞ ð2:10Þ
V h k l

According to Eq. 2.10 the electron density, q, can be calculated at any point (x,
y, z) by constructing a Fourier Series which has coefficients that are equal to the
Structure Factors (Warren 1990). This is the basic equation of crystallography.
It enables the calculation of a three-dimensional electron density map
throughout the unit cell. Maxima in the electron density map define the locations
of individual atoms.

2.3 Crystal Structure Determination from Diffraction Data

Structure determination in crystallography refers to the process of elaborating the

three-dimensional positional coordinates (and also, usually, the three-dimensional
anisotropic displacement parameters) of the scattering centers in an ordered crystal
lattice. Crystal structure determination from diffraction data involves unit cell
determination, structure solution, and structure refinement.
The aim of structure solution is to obtain an approximate description of the
structure, using the unit cell and space group determined, but starting from no
knowledge of the arrangement of atoms or molecules within the unit cell. Typical
mineral structures contain several thousand unique reflections, whose spatial
arrangement is referred to as a diffraction pattern. Indices (hkl) may be assigned to
each reflection, indicating its position within the diffraction pattern. This pattern
has a reciprocal Fourier transform relationship with the crystalline lattice and the
unit cell in real space. This step is referred to as the solution of the crystal
2.3 Crystal Structure Determination from Diffraction Data 37

structure. If the structure solution is a sufficiently good approximation to the true

structure, then a good quality structure can be obtained by structure refinement.

2.3.1 Structure Refinement

Once a structure solution has been achieved, there are actually two structure
models, a calculated model based on the approximate co-ordinates obtained from
interpretation of a three-dimensional electron density map,
X
Fhkl ¼ fj e2piðhxj þkyj þlzj Þ ð2:11Þ

and an observed model based on the calculation of structure factors from exper-
imental intensities,
pffiffiffiffiffiffi
jFhkl j ¼ K Ihkl ð2:12Þ
The task now is to adjust the various atomic parameters so that the calculated
structure factors match the observed structure factors as closely as possible. One of
the ways to measure the agreement between the observed and calculated models is
with the Residual Index (more commonly referred to as the R-factor)
P
jFobs  Fcalc j
R¼ P ð2:13Þ
jFobs j
In addition to the approximate atomic x, y, z-coordinates, there is another factor
influencing the magnitudes of the structure factors, the thermal motion. Although
the structure determination is done on the solid phase, the atoms still have some
thermal motion (vibration and rotation). This thermal effect is introduced into the
structure factor equation by a factor that serves to attenuate the atomic scattering
factor,
ðsin2 hÞ
fj ¼ fo eB k2 ð2:14Þ

where the temperature factor (Warren 1990) B, is related to the mean-square

amplitude of vibration,

B ¼ 8p2 u2 ð2:15Þ
As B increases, the scattering power of atom j(fj) decreases. To represent
anisotropic thermal motion, a total of six parameters per atom are necessary. Three
of these parameters provide the orientations of the principal axes of the ellipsoid
produced by anisotropic thermal motion; the other three parameters represent the
magnitudes of displacement along the ellipsoid axes.
With thermal effects included, the complete expression for the calculated
structure factor is
38 2 Charge Density Analysis from X-Ray Diffraction

X ðsin2 hÞ
Fhkl ¼ fj e2piðhxj þkyj þlzj Þ eB k2 ð2:16Þ

In order to refine the structure, the coordinates and temperature factors must be
adjusted so that these calculated Fhkl’s match, as closely as possible, the observed
Fhkl’s derived from the experimentally measured intensities.

2.3.2 Theoretical Models in Structure Analysis

The classical approach to solve the structure of a material is to build a physical

model that is consistent with all the information known about the material. An
obvious way to expedite the model building process would be to exploit the power
of a computer, and several programs have been developed to achieve the accurate
structure. The validity of a structural model is that it accounts for all the data
available. The calculated powder diffraction pattern should not only match the
observed data but should be consistent with spectroscopic data and with the
physical properties of the material. Some of the versatile methods such as Rietveld
method (Rietveld 1969), Maximum Entropy method (Collins 1982), Multipole
method (Hansen and Coppens 1978), and Pair distribution function (Proffen and
Billinge 1999) have been followed in this study to analyze the structural properties
of metals and alloys.

2.4 Methods in X-Ray Crystallography

In terms of the specimen handled, two methods can be identified, viz., the single
crystal method and the powder method.
1. Single-crystal X-ray diffraction
2. Powder X-ray diffraction
The single crystal diffraction technique, using relatively large crystals of the
material, gives a set of separate data from which the structure can be obtained.
However, most materials of technical interest cannot grow large crystals, so one
has to resort to the powder diffraction technique using material in the form of very
small crystallites. In reality, the two techniques are highly complementary, have
their own strengths and weaknesses and domains of applicability, and one will
never supercede the other. However, the domain where powder diffraction is
having an impact is certainly growing and diversifying. In general, the structure
solution process poses more significant challenges, which has prompted much
research in recent years on the development of new strategies and techniques in
this field.
2.4 Methods in X-Ray Crystallography 39

2.4.1 Structure Determination from Single-Crystal X-Ray

Diffraction

Within the realm of crystallography, single-crystal X-ray diffraction is by far the

most commonly used technique available for the determination of the crystal and
molecular structure of crystalline solids (Goeta and Howard 2004). Single-crystal X-
ray Diffraction is a non-destructive analytical technique which provides detailed
information about the internal lattice of crystalline substances, including unit cell
dimensions, bond-lengths, bond-angles, and details of site-ordering. In the single-
crystal method, the whole specimen is a single piece, without any discontinuity in
the lattice arrangements. Typical mineral structures contain several thousand unique
reflections, whose spatial arrangement is referred to as a diffraction pattern. This
pattern has a reciprocal Fourier transform relationship with the crystalline lattice and
the unit cell in real space (Shmueli 2008).
The measured intensities Ihkl (corresponding to scattering from a lattice plane
with Miller indices h, k, l) are reduced to structure amplitudes Fhkl by the appli-
cation of a number of experimental corrections
2
 1
Fhkl ¼ Ihkl KLp A ð2:17Þ

where k is a scale factor, Lp the Lorentz–polarization correction, and A the

transmission factor representing the absorption of X-rays by the crystal.
Once the initial crystal structure is solved, an iterative refinement procedure is
essential to attain the best possible fit between the observed and calculated crystal
structure. The most common approach is to perform a least-squares minimization
between the experimental structure factors and those calculated by varying the
adjustable parameters of the structural model. These normally include atomic
positions, anisotropic displacement parameters, occupancies, chemical bond
lengths and angles, and other geometric characteristics of a molecule.
The results of the structure refinement yield a list of atom X–Y–Z assignments
in the unit cell, shape of the anisotropic intensity center for each atom (thermal
parameters), and the distance of the nearest atomic neighbors. The quality of a
solution is assessed by the values of R, wR, and GooF.
The final structure solution will be presented with the residual R factor, which
gives the percent variation between the calculated and observed structures.
P
jFobs  Fcalc j
R¼ P ð2:18Þ
jFobs j
The R-value is used to indicate improvements or reductions in the quality of fit
between model and observation. In the expression above Fobs and Fcalc are the
observed and calculated structure amplitudes and the deviations are summed over
all experimentally recorded intensities. Ideal solutions would have R-value of 0,
however, due to random errors, this is never achieved. R-values (listed as percents)
of less than 5% are considered good solutions, high quality samples will often
40 2 Charge Density Analysis from X-Ray Diffraction

Fig. 2.3 A typical powder 600

diffraction profile
500

Intensity (arbitrary units)

400

300

200

100

10 20 30 40 50 60 70 80 90 100 110 120

2θ (degrees)

result in R-values lower than 2.5%. wR refers to squared F-values. GooF refers to
the ‘‘goodness of fit’’ of the solution. In addition to the difference in F values, the
GooF also takes into account the number of observed reflections and the param-
eters used. At the end of refinement, the GooF should approach 1.

2.4.2 Powder Diffraction

The possibility of using powder diffraction methods to study materials was rec-
ognized shortly after the discovery of X-ray diffraction by Laue and Von Knipping
in 1910. Although, the powder method was developed as early as 1916 by Debye
and Scherrer, for more than 50 years its use was almost exclusively limited to
qualitative and semi-quantitative phase analysis and macroscopic stress mea-
surements. In 1922, Bain using the powder photograph method of Debye and Hull
carried out some very important pioneering work (Weiss 1925). Within a few
years many others including Bragg and Pauling, had exploited the powder method
to study a wide range of materials, including metals, minerals, and simple organic
solids. Quantitative analysis of the pattern using modern computers and software
yields a wealth of additional information about the sample structure. Figure 2.3
shows a typical powder diffraction profile.
Powder diffraction has played a central role in structural physics, chemistry,
and materials science over the past 20 years. Important advances in structural
studies of materials ranging from high temperature superconductors and fullerenes
to zeolites and high-pressure research have relied heavily on the powder diffraction
technique.
Although most of the structures determined from powder diffraction data have
been solved in the last few years, notable contributions have been from Zacha-
riasen (1948) and Berg and Werner (1977). Zachariasen, in particular, used a
number of ingenious methods to solve crystal structures from powders. His work
on b-plutonium (Zachariasen and Ellinger 1963), for example, utilized differential
2.4 Methods in X-Ray Crystallography 41

thermal expansion to resolve Bragg peak overlap. Many of the early zeolite
structures that were solved from powder diffraction data involved model building
and significant chemical intuition (Breck et al. 1956; Kokatalio et al. 1978), and
these concepts are now implemented in computer algorithms.
Solving a structure from powder diffraction data has developed rapidly over the
last 10 years. Prior to 1990, very few unknown crystal structures had been
determined directly from powder diffraction data. Today, numerous crystal
structures, both organic and inorganic, have been solved from powder data.
Developments in instrumentation, computer technology, and powder diffraction
methodology have all contributed to this increased success rate.

2.4.2.1 Structure Determination from Powder Diffraction

The first choice is the selection of radiation source and instrumentation geometry.
The first step in data analysis is indexing of the diffraction pattern. From the unit
cell and space group the diffraction intensity associated with each reflection should
be determined by applying whole-profile fitting. The intensity extraction can be
performed using either a least-squares method (Pawley 1981) or an iterative
approach (Le Bail et al. 1988).
Rietveld proposed a simple summation approach to the evaluation of an
observed structure factor magnitude for partially and completely overlapped
reflections, if the calculated diffraction pattern is good but the observed and cal-
culated structure factor magnitudes are different from one another.
The peak area is proportional to the square of the structure factor magnitude and
the problem reduces to finding the peak area. For an isolated peak, the observed
peak area is easy to evaluate. For overlapping peaks, the contribution for a given
reflection weighted by the calculated peak contribution for that reflection is
divided by the sum of the calculated peak values for each contributing reflection.
The estimated standard deviations of observed intensities are not normally
included as part of the standard Le Bail approach, they may nevertheless be
evaluated.
Pawley (1981) declared a method for determining Bragg peak intensities from
powder diffraction data in the absence of structural model. The variables associ-
ated with peak positions and widths are the same, the variables associated with
peak areas are peak areas themselves. Although the Pawley method was intro-
duced some 6 years earlier than its counterpart, LeBail method is currently still the
more popular approach.
The final step in the structure determination maze is the completion of the
structure and the refinement of the structural parameters using the Rietveld method
(Rietveld 1969). The structural proposal from the structure determination can be
confirmed only when the refinement has been brought to a successful conclusion.
Chemical information and intuition play important roles in guiding the path in the
maze.
42 2 Charge Density Analysis from X-Ray Diffraction

2.5 The Rietveld Method

The drawback of the conventional powder method is that the data grossly overlap,
thereby preventing proper determination of the structure. The major breakthrough
in the value of the powder method as a quantitative tool was the development of
the Rietveld method (Rietveld 1969), a technique for crystal structure refinement
which, for the first time, made use of the entire powder pattern instead of analyzing
individual, non-overlapped, Bragg reflections separately. This approach minimizes
the impact of overlapped and degenerate peaks by calculating the entire powder
pattern of a crystalline model, including various experimental and sample
dependent peak-broadening effects. Parameters in the model such as atomic
positions, lattice parameters, and experimental factors that affect peak-shape and
background are varied, using a least-squares approach, until the agreement
between the calculated and measured diffraction profiles are optimized and the
model is refined by iterative procedure.
The method was first reported for the diffraction of monochromatic neutrons
where the reflection-position is reported in terms of the Bragg angle 2h. The
method was quickly extended from reactor neutron data, with its nice Gaussian
line-profiles and lack of atomic form-factor, to in-house X-ray powder diffraction,
synchrotron powder diffraction, and time-of-flight neutron data from pulsed
spallation sources, and to refinements of incommensurate and magnetic structures.
The application of the Rietveld method to neutron data in the early 1970s was soon
followed by its extension to laboratory X-ray diffractometer data (Malmros and
Thomas 1977; Young et al. 1977). The technique has been applied to a wide range
of solid-state problems and has been reviewed by several authors during the last
25 years (Cheetham and Taylor 1977; Hewat 1986).
The problem of the more complex peak shape was resolved by employing
alternative peak-shape functions, such as the Lorentzian and the pseudo-Voigt.
Other problems that can plague X-ray studies include preferred orientation and
poor powder averaging (graininess), both of which arise from the fact that X-rays
probe a smaller sample volume than neutrons; these were addressed by paying
closer attention to the data collection strategy. The accuracy and precision of a
structure refinement from X-ray data can normally be optimized by collecting
high-resolution data at a synchrotron source (Cox et al. 1983). The resolution of
the powder diffractometers at second and third generation sources is so good that
sample imperfections now play a major role in determining the shape of the Bragg
peaks.

2.5.1 The Rietveld Strategy

Powder diffraction patterns are collected in a step scan mode. Intensity and two-
theta axes of Bragg diffraction are collected for the crystal structure determination
(Young 1995). The only wavelength and technique independent scale is in
2.5 The Rietveld Method 43

reciprocal space units or momentum transfer Q, which is historically rarely used in

powder diffraction but very common in all other diffraction and optics techniques.
The relation is
4p sinðhÞ
Q¼ ð2:19Þ
k

2.5.1.1 Peak Shape

The shape of a powder diffraction reflection is influenced by the characteristics of

the beam, the experimental arrangement, and the sample size and shape (Rietveld
1969). In the case of monochromatic neutron sources the convolution of the
various effects has been found to result in a reflex almost exactly Gaussian in
shape. If this distribution is assumed then the contribution of a given reflection to
the profile yi at position 2hi is:
 
4 lnð2Þ 2
yi ¼ IK exp ðh i  hk Þ ð2:20Þ
Hk2

where Hk is the full width at half peak height (full-width half-maximum), 2hk is the
center of the reflex, and Ik is the calculated intensity of the reflex (determined from
the structure factor, the Lorentz factor, and multiplicity of the reflection).
At very low diffraction angles the reflections may acquire an asymmetry due to
the vertical divergence of the beam. Rietveld used a semi-empirical correction
factor, As to account for this asymmetry.
" #
sPð2hi  hk Þ2
AS ¼ 1  ð2:21Þ
tan hk

where P is the asymmetry factor and s is +1, 0, -1 depending on the difference

2hi – 2hk being positive, zero, or negative respectively. At a given position more
than one diffraction peak may contribute to the profile. The intensity is simply
the sum of all reflections contributing at the point 2hi.

2.5.1.2 Peak Width

The width of the diffraction peaks are found to broaden at higher Bragg angles.
This angular dependency (Caglioti et al. 1958) was originally represented by

HK2 ¼ U tan2 hk þ V tan2 hk þ W ð2:22Þ

where U, V, and W are the half-width parameters and may be refined during the fit.
44 2 Charge Density Analysis from X-Ray Diffraction

2.5.1.3 Preferred Orientation

In powder samples there is a tendency for plate-like or rod-like crystallites to align

themselves along the axis of a cylindrical sample holder. In solid polycrystalline
samples the production of the material may result in greater volume fraction of
certain crystal orientations (commonly referred to as texture). In such cases the reflex
intensities will vary from that predicted for a completely random distribution. Riet-
veld allowed for moderate cases of the former by introducing a correction factor:
 
Icorr ¼ Iobs exp Ga2 ð2:23Þ

where Iobs is the intensity expected for a random sample, G is the preferred
orientation parameter, and a is the acute angle between the scattering vector and
the normal of the crystallites.

2.5.1.4 Refinement

The principle of the Rietveld Method is to minimise a function M which analyzes

the difference between a calculated profile y(calc) and the observed data y(obs).
Rietveld (1969) defined such an equation as:
X

1 calc 2
obs
M¼ Wi yi  yi ð2:24Þ
i
c

where Wi is the statistical weight and c is an overall scale factor such that
ycalc = cyobs.
Rietveld (1969) method calculates the entire powder pattern using a variety of
refinable parameters and improves the selection of these parameters by minimizing
the weighted sum of the squared differences between the observed and the cal-
culated pattern using least square methods. Thus the systemic and accidental peak
overlap is overcome.
The method has been so successful that nowadays the structure of materials, in
the form of powders, is routinely being determined, nearly as accurately as the
results obtained by single-crystal diffraction techniques. An even more widely
used application of the method is in determining the components of chemical
mixtures. This quantitative phase analysis is now routinely used in industries
ranging from cement factories to the oil industry. The success of the method can be
gauged by the publication of more than a thousand scientific papers yearly using it.

2.5.2 Rietveld Refinement

The Rietveld refinement was performed using the softwares JANA 2000 (Petříček
et. al. 2006) and an improvised version JANA 2006 (Petříček et. al. 2006). JANA
2.5 The Rietveld Method 45

2000 and JANA 2006 are crystallographic programs for structure analysis of
crystals periodic in three or more dimensions from diffraction data. It is focused to
solution, refinement, and interpretation of difficult, especially modulated struc-
tures. It calculates structures having up to three modulation vectors from powder
as well as single-crystal data measured with X-ray or neutron diffraction. The input
diffraction data can be unlimitedly combined, the combination of powder neutron
data with single-crystal X-ray data being a typical example. The structure solution
can be done using the builtin charge flipping algorithm or by calling an external
direct methods program. Jana can handle multiphase structures (for both powder
and single-crystal data), twins with partial overlap of diffraction spots, commen-
surate, and composite structures. It contains powerful transformation tools for
symmetry (groupsubgroup relations), cell parameters, and commensurate—super
cell relations. A wide scale of constrains and restrains is available including a
powerful rigid body approach and the possibility to define a local symmetry
affecting only part of the structure.
In the refinement using the software, the observed profiles are matched with the
profiles constructed similarly by using pseudo-voigt (Wertheim 1974) profile
shape function of Thompson, Cox and Hastings (1987) modified to some extent
that accommodate various Gaussian FWHM parameters and Scherer coefficient for
Gaussian broadening. The profile asymmetry is introduced by employing multi-
term Simpson rule integration devised by Howard (1982) that incorporates sym-
metric profile shape function with different coefficient for weights and peak shift.
Jana 2006 also employs the correction for preferred orientation which is inde-
pendent of diffraction geometry using March—Dollase function (March 1932;
Dollase 1986). The calculated profiles thus evolved are compared with the
observed ones. Finally, the structure factors evolved from the Rietveld refinements
were further utilized for the estimation of charge density in the unit cell.

2.6 Multipole Method

The distribution of positive and negative charges in a crystal fully defines physical
properties like the electrostatic potential and its derivatives, the electric field, and
the gradient of the electric field. The electrostatic potential is of importance in the
study of intermolecular interactions, and has received considerable attention
during the past two decades. It plays a key role in the process of molecular
recognition, including drug-receptor interactions, and is an important function in
the evaluation of the lattice energy. The evaluation of the electrostatic potential
and its derivatives may be achieved directly from the structure factors, or indi-
rectly from the experimental electron density by the multipole formalism.
The ability to measure the experimental charge distribution in crystals from the
intensities of the scattered X-rays was realized almost immediately after the dis-
covery of X-ray diffraction. Notwithstanding this early recognition, the technical
developments of the 1960s and beyond, which occurred in difftractometry,
46 2 Charge Density Analysis from X-Ray Diffraction

automation of data collection, low temperature techniques, computers were needed

to achieve a breakthrough in the method. The accurate crystallographic methods
developed during the past decades led not only to a much better precision in
atomic coordinates, but also to crucial information about the charge distribution in
crystals. This experimentally obtained distribution can be compared directly with
theoretical results, and can be used to derive other physical properties such as
electrostatic moments, the electrostatic potential, and lattice energies, which are
accessible by spectroscopic and thermodynamic measurements. This broad inter-
face with other physical sciences is one of the most appealing aspects of the field.
Recently, developments in theoretical methods have facilitated the calculation
of charge densities of sufficient accuracy so that they can be compared with the
experimental charge distributions. However, these are limited for molecular
crystals consisting of only light atoms. Such comparisons would reveal the defi-
ciencies in both theory and experiment. The state of art now has reached a stage
where it is possible to derive properties like net charges, molecular dipole
moments and electrostatic potentials and to fit atom-centered spherical harmonic
functions (Guru Row 1983).
The charge density effects in Beryllium metal were measured (Larsen et al.
1980), while the absolute scale was established with an Sm source (Hansen et al.
1987). The bonding between Chromium atoms have been studied and analyzed as
dichromium tetracarboxylates have metal–metal bond lengths which vary by as
much as 0.7 Å (Cotton and Stanley 1977). To explain the properties of b’NiAl
(Fox and Tabernor 1991) measured four low angle structure factors which showed
a depletion of density around both Ni and Al atoms, and a buildup of about 0.13
eÅ-3 along the (111) direction halfway between Ni and Al nearest neighbors.

2.6.1 Multipole Electron Density Model

Expansion of the charge density in a crystal in terms of a series of nucleus-

centered spherical harmonic functions was first applied by De Marco and Weiss
(1965) and significantly complemented by Dawson (1967) and collaborators in
their study of bonding in diamond-type structures. In the earliest treatment the
radial function of the valence shell is that of the isolated atom, but the Dawson
formalism allows a modified Gaussian radial function to have the same radial
dependence as the isolated atom. A more flexible radial function consisting of a set
of harmonic oscillator function wave functions was used by Kurki-Suonio (1968).
Least squares adjustment of the parameters in the Dawson model was carried out
by McConnell and Sanger (1970), while Stewart’s functions include a full set of
spherical harmonics truncated at a level chosen by practical considerations such as
adequacy in describing the aspherical deformations and available computing
facilities. A related model of cosn type, which is linear combinations of spherical
harmonics, was developed (Hirshfeld 1971; Harel and Hirshfeld 1975) and applied
to a number of organic molecules.
2.6 Multipole Method 47

Such models have the obvious merit of providing an analytical description of

the charge density and may be used in the calculation of physical properties based
on charge density distribution (Stewart 1972). As their use should increase as more
accurate data sets become available a critical analysis of the results has been
undertaken.
This analysis requires X-ray data sets specifically collected for charge density
studies.

2.6.2 Mathematical Approach of Multipole Electron Density

Model

The density model (Hansen and Coppens 1978) consists of a superposition of

harmonically vibrating aspherical atomic density distributions:
X
atoms
qðr Þ ¼ qðr  rk  uÞ  tk ðuÞ ð2:25Þ
k

where tk(u) is a Gaussian thermal displacement and  indicates a convolution.

Each atomic density is described as a series expansion in real harmonic functions
through fourth–order (ylm).
X
4 X
l !
r
qatomic ðr Þ ¼ Pc qcore þ Pv k 03 qvalence ðk0 r Þ þ 003 Rl ðk00 r Þ Plm ylm
l¼0 m¼1
r
ð2:26Þ
Here Pc, Pv, and Plm are population coefficients. The total number of electrons
associated with one atom is equal to Pc þ Pv þ P00 , since the higher terms with
l 6¼ 0 integrate to zero when integration is performed over all space. The func-
tions qcore and qvalence are chosen as the Hartree–Fock (HF) densities of the free
atoms normalized to one electron, but the valence function is allowed to expand
and contract by adjustment of the variable radial parameter k0 (Coppens et al.
1979). qcore and qvalence are constructed from the canonical Hartree–Fock atomic
orbitals.
The radial functions Rl ðr Þ of the other terms are
nl þ3
Rl ðr Þ ¼ ðnfl þ2Þ! rn lexpðfl rÞ; where in principle nl can take any positive integer
value. Starting values of the orbital exponent f are modified by a variable
parameter k00 ; such that f0 ¼ k00 f:
RThe spherical harmonic functions ylm and the normalization factors based on
jPlm jds ¼ 2:This normalization implies that Plm ¼ 1 one electron has been
moved from the negative to the positive lobes of the deformation functions with
P  1:
48 2 Charge Density Analysis from X-Ray Diffraction

The structure factor corresponding to (2) and (3) becomes:

F ð hÞ ¼
"   X   X  #
atoms X
X symmetry h 4
h l
hp
Pc fcore ðhÞ þ Pv fvalence 00 þ ukl 00 Pklm yklm 
k p
k l¼0
k m¼1
h
exp 2pih:rkp Tk ðhÞ
ð2:27Þ
where fcore and fvalence are the Fourier transforms of qcore and qvalence respectively,
and ukl is the Fourier transform of Rkl defined as:
Z1
l
ukl ðhÞ ¼ 4pi Rkl ðr Þjl ð2phr Þ r2 dr ð2:28Þ
0

The temperature factor Tk is the Fourier transform of tk and hp and rlp are
symmetry transformed scattering and position vectors. We note that the model
reverts to the free-atom model when j0 = 1 and the Pklm0 s are zero. It is in this
respect, similar to Hirshfeld’s deformation model (Harel and Hirshfeld 1975), but
different from the single Slater-type models used by a series of authors (Cromer
et.al. 1976; Price and Maslen 1978; Chen et.al. 1977).

2.6.3 Criteria for Judging Aspherical Atom Refinements

Although an aspherical atom model (multipole) refinement based on Eq. 2.25 may
lead to a lowering of agreement indices, the improvement is not necessarily sig-
nificant because of the large number of parameters involved, and the results are not
necessarily physically meaningful as they may be affected by parameter correla-
tion or systematic errors in the measurements. The following tests of the multipole
refinements have therefore been applied.

2.6.3.1 Significance of Improvement of the Fit Between Calculated

and Observed Structure Factors

For an adequate model the goodness of fit g; defined as

 1
g ¼ wi D2l =v 2 ð2:29Þ

should tend to 1 as refinement proceeds.

For a refinement of F, D equals jFobs j  jkFcalc j; wl represents the weight
assigned to each of the reflections from an estimated of experimental accuracy and
v is the number of degrees of freedom (i.e., the number of independent observa-
tions n, minus the number of independently varied parameters p).
2.6 Multipole Method 49

In comparison of different models we use the error function

X
2¼ vg2 ¼ wi D2i ð2:30Þ

which follows an v2 distribution with v degrees of freedom for a fully refined

model. The error function can be tested by a modification of the R-factor test:
e0  b d
 Fd;np ð2:31Þ
e np
Or
e0 d
¼1þ Fnp ð2:32Þ
e np
where d is the dimensionality of the hypothesis, and the tabulated F distribution is
the ratio of two v2 distributions.

2.6.3.2 Residual Density Maps

 
A residual density map, from a Fourier summation based on Fkobs  jFcalc j as is
commonly used in structure analysis will show to what extent the multipole
expansion has been successful in describing the features of the density distribution.
It does not contain information, however, about the significance of residual fea-
tures. For this purpose the residual density distribution may be compared with a
map representing the position dependence of the estimated standard deviation in
qobs as described by Rees (1976, 1977) and calculated in several previous studies
(Stevens and Coppens 1976; Coppens et.al. 1977).

2.6.3.3 Parameter Bias

A successful refinement should give parameters with minimal bias due to atomic
asphericity. The rationale for this requirement is twofold. First, deformation maps
(i.e., total minus ‘sum of spherical atom’ densities) are quite sensitive to errors in
positional and thermal parameters, and second, physical properties to be calculated
from the refinement results, such as dipole and quadrupole moments, depend on
the nuclear positions as well as charge distribution.

2.6.3.4 Derived Properties

One of the justifications for performing the multipole refinement is the experi-
mental determination of derived physical properties (Stewart 1972). Properties
50 2 Charge Density Analysis from X-Ray Diffraction

such as net molecular charges, dipoles, and quadrupole moments may also be
obtained by direct-space integration of the experimental charge distribution.

2.6.4 Multipole Refinement Strategy

The raw data is refined for their structural parameters with extinction, absorption,
and TDS corrections. Structure factors are refined to yield the smallest possible
reliability indices. The results are compared for the thermal parameters obtained
using MEM. The refined structure factors are further refined for the population and
k-parameters using the formalism proposed by Hansen and Coppens (1978). In this
refinement, the canonical Hartree–Fock atomic orbitals of the free atoms nor-
malized to one electron are used for the construction of qcore and qvalence charge
densities.

2.6.4.1 Static and Dynamic Deformation Density

The multipole charge density is usually analyzed by means of deformation density

maps. The multipole density consists of superposition of harmonically vibrating
aspherical atomic density distribution as
X
atoms
qðr Þ ¼ qk ðr  rk  uÞ  tk ðuÞ ð2:33Þ
k

where tk ðuÞ is a Gaussian thermal-displacement and  indicates a convolution.

When the thermal contribution is de-convoluted, this density becomes the density
of a static atom. The deformation between a static model density and the observed
density reveals the contribution from thermal displacement of charge distribution.
Thus
Dqdeformation ¼ qmultipole  qobserved ð2:34Þ

gives static deformation density when thermal contribution is de-convoluted and

dynamic deformation density when it is convoluted with the radial dependent
multipole charge density function.

2.6.5 Significance of Multipole Model

It is appropriate to take a critical look at the aspherical-atom multipole (pseudo

atom) model, as expressed in a number of algorithms (Hirshfeld 1971, 1977a;
Stewart 1976), including the Hansen and Coppens model (1978). The pseudo atom
2.6 Multipole Method 51

model has significant advantages and its introduction has greatly contributed to the
increasing application of experimental results in charge-density analysis,
1. Experimental noise is generally not fitted by the model functions and therefore
effectively filtered out.
2. Thermal motion is treated separately and de-convoluted from the final result.
3. The resulting static density provides an effective level of comparison with
theoretical results, especially if the latter has been filtered through the model by
refinement of theoretical structure factors.
4. Notwithstanding the development of alternative formulations, including bond-
charge models and orbital-based algorithms, no generally competitive alter-
native has been developed. While the pseudo atom model is widely used in
experimental density analysis, it is important not to lose sight of the implied
assumptions.
5. The results are dependent on the adequacy of the thermal motion formalism
used (Mallinson et al. 1988), which generally is limited to the harmonic
approximation.

2.7 Maximum Entropy Method

Understanding the chemical and physical properties of molecular systems

requires knowledge of their charge distributions (Bader 1991). Experimentally,
electron density distributions (EDD) can be reconstructed from accurate X-ray
diffraction data through a series of elaborate data reduction and data analysis
steps (Iversen et al. 1996). The most widely used method entails least-squares
optimization of models containing atom-centered aspherical density functions
(Stewart 1976; Hirshfeld 1977b; Hansen and Coppens 1978). In the empirical
modeling schemes, estimates of errors of electron density and in derived prop-
erties can be calculated within the framework of the least squares method. Such
estimates rely on several assumptions including the adequacy of the refined
model. Several studies (Figgis et al. 1993; Chandler et al. 1994; Iversen et al.
1997) have shown that even the very sophisticated models currently used in
empirical EDD modeling are inadequate to describe very fine density features
present in data and in general, least-squares estimates of EDDs will therefore
contain systematic bias due to the model. Nevertheless, the least-squares error
estimates allow, to some extent, assessment of the reliability of conclusions
drawn from the model densities.
The Maximum entropy method (MEM) is an information–theory–based tech-
nique that was first developed in the field of radio astronomy to enhance the
information obtained from noisy data (Gull and Daniell 1978). The theory is based
on the same equations that are the foundation of statistical thermodynamics. Both
the statistical entropy and the information entropy deal with the most probable
52 2 Charge Density Analysis from X-Ray Diffraction

distribution. In the case of statistical thermodynamics, this is the distribution of the

particles over position and momentum space, while in the case of information
theory, the distribution of numerical quantities over the ensemble of pixels is
considered.
The maximum entropy method (MEM) (Sakata and Sato 1990; Collins 1982;
Bricogne 1988) has been introduced in charge density reconstruction. MEM can
yield a high-resolution density distribution from a limited number of diffraction data.
The obtained density distribution gives detailed structure information without using
structural model. The ability of MEM in terms of a model-free reconstruction of the
charge densities from measured X-ray diffraction data can be interpreted as
‘‘imaging of diffraction data’’ (Sakata and Takata 1996). From limited numbers of
X-ray diffraction data, EDDs have been reconstructed by MEM in a number of
systems. Maps that qualitatively reveal bonding features have been obtained. The
electron-density map from MEM is one of the accurate methods for structure
analysis. A precise electron-density map can be obtained and the existence of
bonding electrons is clearly visible in the maximum-entropy map. The resolution of
the maximum-entropy map is much higher than the map drawn by conventional
Fourier transformation (Sakata and Sato 1990).
The maximum-entropy method (MEM) of calculating charge density directly
from X-ray and diffractions data is a promising approach for studying the bonding
state, atomic disorder, and ion conduction in detail. In X-ray diffraction, the
observed crystal structure factors applied to the inverse Fourier transform are
known to result in the charge density. However, this relationship between the
observed crystal structure factors and charge density is premised on obtaining all
observed crystal structure factors without any measurement errors. In contrast,
MEM can improve these disadvantages by inferring the unobserved crystal
structure factors from the observed crystal structure factors and maximizing the
information entropy. MEM has been reported in various fields. These reports have
provided a positive impetus for using MEM analysis to study charge density, ion
conduction, and atomic disorder (Itoh et al. 2010).
In the refinement a rigid body model and restraints for intermolecular distances
and appropriate angles are applied. The integrated intensities of each reflection are
evaluated from the observed diffraction patterns using the result of Rietveld
refinement. The final charge density is obtained after several iterative refinements.
The method is thus developed by combining the MEM with the Rietveld
method to create a new sophisticated method of structure refinement in charge
density level, the MEM/Rietveld method (Takata et al. 1995). The MEM/Rietveld
analysis is an iterative way in combination with the MEM and Rietveld analyses.
When the MEM charge density at a certain iteration step can provide a better
structural model for the Rietveld of the next iteration, the iteration process con-
tinues. In this method, the final MEM electron density distribution derived is
compatible with the structural model used in the Rietveld refinement (Takata et al.
2001).
2.7 Maximum Entropy Method 53

2.7.1 Maximum Entropy Enhancement of Electron Densities

The accurate electron density distribution could be obtained if all the structure
factors are known without any ambiguities. It is however impossible to collect
exact values of all the structure factors by X-ray diffraction methods. The number
of observed structure factors by the experiment is always limited and has some
errors. The uncertainties in the results due to the incompleteness of the experi-
mental information must be rectified. Maximum entropy method is one of the
appropriate methods in which the concept of entropy is introduced to handle the
uncertainties properly. The principle of MEM is to obtain electron density dis-
tribution which is consistent with the observed structure factors and to leave the
uncertainties to a minimum. The mathematical formalism of MEM is given in the
following lines.
The probability of a distribution of N identical particles over m boxes, each
populated by ni particles, is given by
N!
P¼ ð2:35Þ
n1 !n2 !n3 !. . .. . .nm !
As in statistical thermodynamics, the entropy is defined as ln P. Since the
numerator is constant, the entropy is, apart from a constant, equal to
X
S¼ ni ln ni ð2:36Þ
i

where Stirlings’ formula (ln N! = N ln N minus; N) has been used.

In case there is a prior probability qi for box i to contain ni particles, then
becomes
N!
P¼  qn11 qn22 . . . . . .qnmm ð2:37Þ
n1 !n2 !n3 !. . .. . .:nm !
which gives, for the entropy expression,
X X X
m
ni
S¼ ni ln ni þ ni ln qi ¼  ni ln ð2:38Þ
i i i¼1
qi

The maximum entropy method was first introduced into crystallography by

Collins (1982), who based on Eq. 2.5, expressed the information entropy of the
electron density distribution as a sum over M grid points in the unit cell, using the
entropy formula (Jaynes 1968).
X  0 
q ðr Þ
S¼ q0 ðr Þ ln 0 ð2:39Þ
s ðr Þ
54 2 Charge Density Analysis from X-Ray Diffraction

where both q0 ðr Þ and prior probability s0 ðr Þ are related to the actual electron
density in a unit cell as
qðr Þ sðr Þ
q 0 ðr Þ ¼ P and s 0 ðr Þ ¼ P ð2:40Þ
r qðr Þ r sðr Þ

where qðr Þ and sðr Þ are the electron density and prior electron density at a fixed r
in a unit cell respectively. In the present theory, the actual densities are treated
hereafter instead of normalized densities, and q0 ðr Þ becomes s0 ðr Þ when there is no
information. The q0 ðr Þ and s0 ðr Þ are normalized as
X X
q0 ðr Þ ¼ 1 and s0 ð r Þ ¼ 1 ð2:41Þ

The entropy is maximized subject to the constraint

1 X jFcal ðH Þ  Fobs ðH Þj2

C¼ ð2:42Þ
N k r2 ð H Þ

where N is the number of reflections used for MEM analysis, rðH Þ; standard
deviation of Fobs ðH Þ; the observed structure factor and Fcal ðH Þ is the calculated
structure factor given by
X
Fcal ðH Þ ¼ V qðr Þ exp ð2piH:r ÞdV ð2:43Þ
r

where V is the volume of the unit cell.

The constraint is sometimes called a weak constraint, in which the calculated
structure factors agree with the observed ones as a whole when C becomes unity.
From Eq. 2.43, it can be seen that the structure factors are given by the Fourier
transform of the electron density distribution in a unit cell. In the MEM analysis,
there is no need to introduce the atomic factors, by which the structure factors are
normally written. It should be emphasized here that it would be an assumption to
use the atomic form factors in the formulation of the structure factors. Eq. 2.43
guarantees that it is possible to allow any kind of deformation of the electron
densities in real space as long as information concerning such a deformation is
included in the observed data.
We use Lagrange’s method of undetermined multiplier (k) in order to constrain
the function C to be unity while maximizing the entropy.
We then have
k
Q¼S C ð2:44Þ
2
X  0 
q ðr Þ k X jFcal ðH Þ  Fobs ðH Þj2
¼ q0 ðr Þ ln 0  ð2:45Þ
s ðr Þ 2N k r2 ð H Þ
dQ
And when dq ¼ 0 and using the approximation,
2.7 Maximum Entropy Method 55

ln x ¼ x  1 we get,
( " # )
kF000 X 1
qðri Þ ¼ sðri Þ exp jFobs ðH Þ  Fcal ðH Þj exp ð2pjH:r Þ
N rðH Þ2
ð2:46Þ
where F000 ¼ Z; the total number of electrons in a unit cell. Eq. 2.46 cannot be
solved as it is, since Fobs ðH Þ is defined on qðr Þ: In order to solve Eq. 2.46 in a
simple manner, we introduce the following approximation which replaces Fcal ðH Þ
as
X
Fcal ðkÞ ¼ V sðr Þ exp ð2piH:r ÞdV ð2:47Þ

This approximation can be called zeroth order single pixel approximation

(ZSPA). By using this approximation the right-hand side of Eq. 2.46 becomes
independent of sðr Þ and Eq. 2.46 can be solved in an iterative way starting from a
given initial density for the prior distribution. As the initial density for the prior
density sðr Þ; a uniform density distribution is employed in this work
Z
sðr Þ  sðr Þ  ð2:48Þ
M
where M is the number of pixels for which the electron density is calculated. The
reason for this choice of prior distribution is that uniform density distribution
corresponds to the maximum entropy state among all possible density distribu-
tions. In the calculation of qðr Þ; all of the symmetry recruitments are satisfied and
the number of electgrons (Z) is always kept constant through an iteration process.
Mathematically, the summation concerning qðr Þ in the above equations should be
written as an integral. Since we must use a very limited number of pixels in the
numerical calculation, the integral is replaced by the summation in the above
equations.
After completion of the MEM enhancement, it becomes possible to evaluate the
reflections missing from the summation. In a Fourier summation, the amplitudes of
the unobserved reflections are assumed to be equal to zero, while the MEM
technique provides the most probable values.
When extinction is present in the data set, it must be corrected before the MEM
procedure is started. The structure factors must similarly be corrected for anom-
alous scattering, if present. Both corrections require a model for their evaluation.
The independent-atom model is usually adequate for this purpose.
The advantage of MEM is a statistical deduction that can yield a high resolution
density distribution from a limited number of diffraction data without using a
structural model. It has been suggested that MEM would be a suitable method for
examining electron densities in the inner atomic region, for example, bonding
region. It gives less biased information about the electron densities as compared to
conventional Fourier synthesis.
56 2 Charge Density Analysis from X-Ray Diffraction

2.7.2 MEM Refinement Strategies

The technological advances in recent years bring demands for integrated 3D

visualization systems to deal with both structural models and volumetric data, such
as electron and nuclear densities. The crystal structures and spatial distribution of
various physical quantities obtained experimentally and by computer simulations
should be understood three dimensionally.
Once the structure factors are refined, they are further utilized for the evaluation
of MEM charge density.
The MEM charge density calculations are done on the same formalism that
Collins (1982) had adopted. In the refinement process, the analysis was performed
for all data sets using Fortran 90 program PRIMA (Izumi and Dilanian 2002), to
get a 3D density file. PRIMA (Practice Iterative MEM Analyses) is a MEM
analysis program to calculate electron densities from X-ray diffraction data. The
input file contains the cell parameters, space group, pixels, total charge, Lagrange
parameter, and structure factors. In the present work, the unit cell was divided into
643 pixels and the initial electron density at each pixel was fixed uniformly as Z/a30,
where Z is the number of electrons in the unit cell. The electron density is eval-
uated by carefully selecting the Lagrange multiplier in each case such that the
convergence criterion C becomes unity after performing minimum number of
iterations. The 3D electron density was plotted using VESTA (Visualization of
Electron/Nuclear densities and Structures) (Momma and Izumi 2006) software
package. VESTA is a 3D visualization program for structural models and 3D grid
data such as electron/nuclear densities. VESTA deals with structural models and
volumetric data at the same window. Virtually unlimited number of objects such as
atoms, bonds polyhedra, and polygons on iso-surfaces are dealt with. Visualization
of inter atomic distances and bond angles are possible. Transparent iso-surfaces
can be overlapped with structural models.

2.8 Pair Distribution Function

The method of crystallographic analysis of atomic structure is so well established

today that we often fail to see the enormity of the feat accomplished by a simple
X-ray diffraction measurement, the positions of as many as 1023 atoms are
determined with an accuracy of 10-4 nm. This is the benefit of lattice periodicity.
But modern materials are often disordered and standard crystallographic methods
lose the aperiodic (disorder) information. The structural analysis of non-periodic
matter is more difficult than crystalline solids. Recent advances in the experimental
methods, have greatly improved our ability to characterize the atomic structure of
non-periodic matter using the atomic pair distribution function (PDF) analysis and
to find local structural deviations from a well-defined average structure (Egami
1990; Billinge et al. 1996).
2.8 Pair Distribution Function 57

By definition, the atomic pair distribution function PDF is the instantaneous

atomic density–density correlation function which describes the atomic arrange-
ment in materials (Billinge et al. 2000). It is the sine Fourier transform of the
experimentally observable structure factor obtained in a powder diffraction
experiment (Warren 1990). Since the total structure function includes both the
Bragg intensities and diffuse scattering its Fourier associate, the PDF, yields both
the local and average atomic structure of materials. By contrast, an analysis of the
Bragg scattering intensities alone yields only the average crystal structure.
Determining the PDF has been the approach of choice for characterizing glasses,
liquids and amorphous solids for a long time (Wagner 1978). However, its
widespread application to crystalline materials, like manganites, where some local
deviation from the average structure is expected to take place, has been relatively
recent (Egami 1998).
We chose to use high-energy X-rays to measure the PDFs because it is possible to
get high-quality data at high-Q values (Q is the magnitude of the wave vector)
allowing accurate high real-space resolution PDFs to be determined (Petkov et al.
1999a, b). It was previously thought that neutrons were superior for high-
Q measurements because, as a result of the Q dependence of the X-ray atomic form
factor the X-ray coherent intensity gets rather weak at high Q; however, the high flux
of X-rays from modern synchrotron sources more than compensates for this and we
have shown that high quality high-resolution PDFs can be obtained using X-rays.
The study of alloys is complicated by the fact that considerable local atomic
strains are present due to the disordering effect of the alloying. This means that
local bond lengths can differ from those inferred from the average crystallographic
structure by as much as 0.1 Å. This clearly has a significant effect on calculations
of electronic and transport properties. To characterize the structure of alloys fully,
it is necessary to augment crystallography with local structural measurements. In
the past the extended X-ray absorption fine structure (XAFS) technique has been
extensively used. More recently, the atomic pair distribution function (PDF)
analysis of powder diffraction data has also been applied to obtain additional local
structural information (Peterson et al. 2001).

2.8.1 Atomic Pair Distribution Function

The elastic and quasi-elastic scattering intensity of X-rays or neutrons by a col-

lection of atoms, after correction for absorption, polarization and multiple scat-
tering and normalizing to the unit of one atom or one electron scattering (Warren
1990) is given by
X    
I ðQ Þ ¼ fi ðQÞfj ðQÞ exp iQ: ri  rj ð2:49Þ
ij
58 2 Charge Density Analysis from X-Ray Diffraction

where fi ðQÞ is the

 scattering
  amplitude of a single atom i; ri is the position of the
ith atom, exp iQ: ri  rj is the quantum and thermal average, and Q is the
scattering vector,
Q ¼ kf  ki ð2:50Þ
     
Q ¼ jQj ¼ 2kfi  sin h if kf  ¼ ki ð2:51Þ

where kf and ki are the momenta of the scattered and incident photons or neutrons,
respectively, and h is the diffraction angle. For X-ray scattering the quasi-elastic
scattering intensity usually includes the inelastic scattering intensity due to phonons,
since unless a special high resolution setup is utilized it cannot be separated from the
elastic scattering, while the single-electron inelastic (Compton) scattering intensity
is excluded. For neutron scattering the phonon is either excluded or included only
approximately. The average, or total, structure factor can be defined as
h  i
I ðQ Þ hf ðQ Þi 2  f ð Q 2 Þ
Sð Q Þ ¼ þ ð2:52Þ
hf ðQÞi2 hf ðQÞi2

where hf ðQÞi2 is the compositional average of structure factor.

If SðQÞ is isotropic, i.e., independent of the orientation of Q, the Fourier
transformation of (2) in three dimensions gives the atomic pair distribution
function (PDF),
Z
1
~ n0 þ 2
nð r Þ ¼ ~ Q½SðQÞ  1 sinðQr ÞdQ; ð2:53Þ
2ð r
where ~n0 is the average atomic number density. When the unit of Q is nm-1, the
unit of ~nðr Þ is nm-3, and ~ nðr Þ corresponds to the atomic number density at a
distance r from an average atom.
Z
2
nð r Þ  ~
4ðr ð~ n0 Þ ¼ Q½SðQÞ  1 sinðQrÞdQ ð2:54Þ
ð
The frequently used atomic PDF, also called G(r) is defined as,
Gðr Þ ¼ 4ðrð~
nðr Þ  ~n0 Þ ð2:55Þ

where G(r) gives the number of atoms in a spherical shell of unit thickness at a
distance r from a reference atom. It peaks at characteristic distances separating
pairs of atoms and thus reflects the atomic structure. G(r) is the Fourier transform
of the experimentally observable total structure function, S(Q) that is

Z
Q¼max
2
Gðr Þ ¼ Q½SðQÞ  1 sinðQr ÞdQ ð2:56Þ
ð
Q¼0

where Q is the magnitude of the wave vector.

2.8 Pair Distribution Function 59

The structure function is related to the coherent part of the total diffracted
intensity of the
  P   2
  ~  ci fi Q
I coh Q ~ 
S Q~ ¼1þ  P  2 ð2:57Þ
 ~ 
 ci f i Q

where I coh ðQÞ is the coherent scattering intensity per atom in electron units and ci
and fi are the atomic concentration and X-ray scattering factor, respectively, for the
atomic species of type i. G(r) is simply another representation of the diffraction
data. However exploring the diffraction data in real space is advantageous, espe-
cially in the case of materials with significant structural disorder.
In order to refine an experimental PDF one needs to calculate a PDF from a
structural model. This can be done using the relation
" #
1 X X bi bj  
Gcalc ðr Þ ¼ a r  rij  4ðr~n0
€ ð2:58Þ
r i j hbi2

where the sum goes over all pairs of atoms i and j within the model separated by
rij : The scattering power of atom i is bi and hbi is the average scattering power of
the sample. In case of neutron scattering, bi is simply the scattering length, in case
of X-rays, it is the atomic form factor evaluated at a user defined value of Q.

2.8.2 Important Details of the PDF Technique

1. The total scattering, including Bragg scattering as well as diffuse scattering,

contributes to the PDF. In this way both the long-range atomic structure,
manifested in the sharp Bragg peaks, and the local structural imperfections,
manifested in the diffuse components of the diffraction pattern, are reflected in
the PDF.
2. G(r) is barely influenced by diffraction optics and experimental factors since
these are accounted for in the step of extracting the coherent intensities from the
raw diffraction data. This renders the PDF a structure-dependent quantity only.
3. By accessing high values of Q, experimental G(r) values with high real-space
resolution can be obtained and, hence, quite fine structural features are revealed
(Petkov et al. 1999a, b). In fact data at high Q values ðQ [ 15Þ are critical to
the success of PDF analysis. Therefore, the PDF can serve as a basis for
structure determination.
Once the PDF is obtained an approach similar to Rietveld refinement is fol-
lowed. A model atomic configuration is constructed and the respective PDF cal-
culated and compared with the experimental one. Structural parameters in the
model such as atomic positions, thermal factors, and occupancies are then varied in
60 2 Charge Density Analysis from X-Ray Diffraction

Fig. 2.4 Result of PDF

3 calculated
refinement observed
2

-0

G(r)(Å) -2
-1

-2

-3

-4
Difference
-5

-6
4.0 10.5 17.0 23.5 30.0
r(Å)

such a way as to improve the agreement between the calculated and experimental
PDFs. This is done with, or without, observing predefined constraints imposed by
the symmetry of the space group of the crystal structure being tested (Proffen and
Billinge 1999). In this way, local distortions away from the average structure or
lower, unresolved, symmetries can be modeled. The PDF is very sensitive to the
coordination environment of atoms over short (\5 Å) and intermediate (5–20 Å)
ranges. The approach has proved to be quite successful in determining the structure
of various crystalline materials exhibiting different degrees of structural disorder
(Gutmann et al. 2000; Proffen et al. 1999; Petkov et al. 1999a, b).
In Fig. 2.4 the result of PDF refinement is plotted against G(r) (the number of
atoms in a spherical shell of unit thickness) and r (the distance from a reference
atom). The solid line is the calculated PDF. The fitted circles are the observed data.
The difference between the calculated and observed PDF is plotted below the data.
Since PDF contains Bragg and diffuse scattering, the information about local
arrangements is preserved. The PDF can be understood as a bond—length dis-
tribution between all pairs of atoms i and j within the crystal (up to a maximum
distance), however each contribution has a weight corresponding to the scattering
power of the two atoms involved.

2.8.3 Calculation of PDF

The study of a measured PDF ranges from a simple peak width analysis revealing
information about correlated motion (Jeong et al. 1998) to the full profile refine-
ment of the PDF based on a structural model either using the reverse Monte Carlo
technique (Toby and Egami 1992) or least-squares regression (Billinge 1998) as
implemented in the program PDFFIT (Proffen and Billinge 1999).
The experimental PDF was obtained by initially extracting the coherently
scattered intensities from X-ray diffraction pattern by applying appropriated cor-
rection for flux, Compton scattering, and sample absorption. The intensity was
2.8 Pair Distribution Function 61

normalized in absolute electron unit, reduced to atomic PDF. All data procession
was done using the program PDFgetX (Jeong et al. 2001).
PDFgetX (Jeong et al. 2001) is a program used to obtain the observed atomic
pair distribution function (PDF) from measured X-ray powder diffraction data. The
observed and calculated PDF has been obtained from the graphical software
PDFgui (Farrow et al. 2007), which is a graphical environment for PDF fitting.
This allows for powerful usability features such as real-time plotting and remote
execution of the fitting program while visualizing the results locally. Modeling of
PDF was done using the software PDFFIT (Proffen and Billinge 1999) to yield
structural parameters.
The program PDFFIT is designed for the full profile structural refinement of the
atomic pair distribution function (PDF). In contrast to conventional structure
refinement based on Bragg intensities, the PDF probes the local structure of the
studied material. The program allows the refinement of atomic positions, aniso-
tropic thermal parameters, and site occupancies as well as lattice parameters and
experimental factors. By selecting individual atom types one can calculate partial
and differential PDFs in addition to the total PDF. Furthermore, one can refine
multiple data sets and/or multiple structural phases. The program is controlled by a
command language, which includes a FORTRAN style interpreter supporting
loops and conditional statements. This command language is also used to define
the relation between refinement parameters and structural or experimental infor-
mation, allowing virtually any constraint to be implemented in the model. PDFFIT
is written in Fortran-77. The basic concept, command language, and some file
formats of PDFFIT are taken from the diffuse scattering and defect structure
simulation program DISCUS (Proffen and Neder 1997).
Using the PDFFIT (Proffen and Billinge 1999) software the PDF refinement
was executed with the refinement of structural parameters like lattice parameters,
phase scale factor, linear atomic correlation factor, quadratic atomic correlation
factor, low r peak sharpening, peak sharpening cut-off and cut-off, for profile setup
functioning to get the absolute phase. The data configuration parameters are PDFfit
range with step size, data scale factor, upper limit for Fourier transform to obtain
data PDF, resolution peak broadening factor, data collection temperature etc.,
which can be refined to get accurate PDF fitting. In the end the observed and
calculated PDFs are visualized and compared.

2.8.4 Significance of PDF

• We would like to be able to sit on an atom and look at our neighborhood. The
PDF method allows us to do that.
• The PDF gives different information about different length-scales. We can see
the structure within a domain at low-r and between domains at high-r.
• PDF gives both local and average structure information.
62 2 Charge Density Analysis from X-Ray Diffraction

References

Azaroff LV (1968) Elements of x-ray crystallography. Mc Graw hill book company, New York,
p 79
Bader RFW (1991) Atoms in molecules, a quantum theory. Oxford University Press, Oxford
Berg JE, Werner PE (1977) Z. Kristallogr 145:310–320
Billinge SJL (1998) Local structure from diffraction. In: Billinge SJL, Thorpe MF. Plenum, New
York, p 137
Billinge SJL, DiFrancesco RG, Kwei GH, Neumeier JJ, Thompson JD (1996) Phys Rev Lett
77:715–718
Billinge SJL, Proffen TH, Petkov V (2000) Phys rev B 1 62(Number 2):1203–1211
Bragg WL (1912) Nature 90:410–410
Bragg WL (1913) The diffraction of short electromagnetic waves by a crystal. Proc Cambridge
Philoso Soc 17:43–57
Breck DW, Everslole WG, Milton RM, Reed TB, Thomas TL (1956) J Am Chem Soc 78:5963–
71
Bricogne G (1988) Acta Cryst A44:517–545
Caglioti G, Paeletti A, Ricci FP (1958) Nucl Instrum 3:223–228
Chandler GS, Figgis BN, Reynolds PA, Wolff SK (1994) Chem Phys Lett 225:421–426
Cheetham AK, Taylor JC (1977) J Solid State Chem 21:253–257
Chen R, Trucano P, Stewart RF (1977) Acta Cryst A33:823–828
Collins DM (1982) Nature 298:49–51
Coppens P, Yang YW, Blessing RH, Cooper WF, Larsen FK (1977) J Am Chem Soc 99:760–766
Coppens P, GuruRow TN, Leung P, Stevens, ED, Becker PJ, Yang YW (1979) Acta Cryst A
35:63–72
Cotton FA, Stanley TT (1977) Inorg Chem 16:2668–2671
Cox DE, Hastings JB, Thomlinson W, Prewill CT (1983) Nucl Instrum Methods 208:573–578
Cromer DT, Larsen AC, Stewart RF (1976) J Chem Phys 65:336–349
Cullity BD, Stock SR (2001) Elements of X-ray diffraction, Pearson education. 3rd edn. Prentice
Hall, Upper Saddle River, p 558
Dawson B (1967) Proc Royal Soc London Ser A 298:255–263
De Marco JJ, Weiss RJ (1965) Phys Rev 137 A1:869–871
Dinnibier RE (2008) Powder diffraction: theory and practice. In: Dinnibier RE, Billinge SJL
(eds.) Royal society of chemistry, p 5
Dollase WAJ (1986) Appl Crystallogr 19:267–272
Egami T (1990) Mater Trans JIM 31:163–176
Egami T, Billinge SJL, Thorpe MF (eds) (1998) Local structure from diffraction. Plenum, New
York
Farrow CL, Juhas P, Liu JW, Bryndin D, Bozin ES, Bloch J, Proffen T, Billinge SJL (2007) J
Phys Condens Matter 19:335219
Figgis BN, Iverson BB, Larsen FK, Reynolds PA (1993) Acta Cryst B49:794–806
Fox AG, Tabernor MA (1991) Acta Metall Mater 39:669–678
George W (2006) Powder diffraction—The Rietveld method and two stage method chapter 1.
Springer, Berlin
Goeta AE, Howard JAK (2004) Chem Soc Rev 33:490–500
Gull SF, Daniell GJ (1978) Nature 272:686–690
Guru Row TN (1983) J Chem Sci 92:4–5
Gutmann M, Billinge SJL, Brosha EL, Kwei GH (2000) Phys Rev B 61:11762–11768
Hansen NK, Coppens P (1978) Acta Cryst A34:909–921
Hansen NK, Schneider JR, Yellon WB (1987) Pearson WH Acta Cryst A43:763–769
Harel M, Hirshfeld FL (1975) Acta Cryst B31:162–172
Hewat AW (1986) Chem Scripta 26A:119–130
Hirshfeld FL (1971) Acta Cryst B 27:769–781
References 63

Hirshfeld FL (1977a) Isr J Chem 16:226–229

Hirshfeld FL (1977b) J Chem 16:226–229
Howard CJ (1982) J Appl Crystallogr 15:615–620
Itoh T, Shirasaki S, Fujie Y, Kitamura N, Idemoto Y, Osaka K, Ofuchi H, Hirayama S, Honma T,
Hirosawa I (2010) J Alloys Compd 491:527–535
Iversen BB, Larsen FK, Figgis BN, Reynolds PA (1996) Acta Cryst B53:923–932
Iversen BB, LarsenFK, Figgis BN, Reynolds PA (1997) J Chem Soc 2227–2240
Izumi F, Dilanian RA (2002) Recent research developments in physics Part II, vol vol 3.
Transworld Research Network, Trivandrum, pp 699–726
Jaynes ET (1968) IEEE Trans Syst Sci Cybern SSC-4:227–241
Jeong IK, Proffen Th, Jacobs F, Billinge SJL (1998) J Phys Chem A 103:921–924
Jeong IK, Thompson J, Proffen T, Perez A, Billinge SJL (2001) PDFGetX a program for
obtaining the atomic pair distribution function from X-ray powder diffraction data. J Appl
Cryst 34:536
Kokatalio GT, Lawton SL, Olson DH, Meier WM (1978) Nature 272:437–438
Kurki-Suonio K (1968) Acta Cryst A 24:379–390
Larsen FK, Lehmann MS, Merisalo M (1980) Acta Cryst B40:159–163
Le Bail A, Duroy H, Fourquet JL (1988) Mater Res Bull 23:447–452
Mallinson PR, Koritsanszky T, Elkaim E, Li N, Coppens P (1988) Acta Cryst A 44:336–342
Malmros G, Thomas JO (1977) J Appl Crystallogr 10:7–11
March A (1932) Z Kristallogr 81:285–297
Marin C, Dieguez E (1999) Orientation of single crystals by back reflection laue pattern
simulation. World Scientific Publishing, Singapore, p 23
Mcconnell JF, Sanger PL (1970) Acta Cryst A 26:83–93
Momma K, Izumi F (2006) Comm Crystallogr Comput IUCr Newslett 7:106–119
Novelline RA, Squire LF (2004) Squire’s fundamentals of radiology. chapter 1, Harvard
University press, Cambridge
Pawley GS (1981) J Appl Crystallogr 14:357–361
Peterson PF, Proffen T, Jeong IK, Billinge SJL, Choi KS, Kanatzidis MG, Radaelli PG (2001)
Phys Rev B 63:165211–165218
Petkov V, Jeong IK, Chung JS, Thorpe MF, Kycia S, Billinge SJL (1999a) Phys Rev Lett
83:4089–4092
Petkov V, Francesko DRG, Billinge SJL, Acharaya M, Foley HC (1999b) Phil Mag B79:1519–
1530
Petříček V, Dušek M, Palatinus L (2000) JANA 2000. The crystallographic computing system
Institute of Physics Academy of sciences of the Czech republic, Praha
Petříček V, Dušek M, Palatinus L, JANA (2006) The crystallographic computing system. Institute
of Physics Academy of sciences of the Czech republic, Praha
Price PF, Malsen EN (1978) Acta Cryst A34:173–183
Proffen T, Neder R (1997) DISCUS a program for diffuse scattering and defect structure
simulations. J Appl Cryst 30:171–175
Proffen Th, Billinge SJL (1999) J Appl Cryst 32:572–575
Proffen Th, Francesco DRG, Billinge SJL, Brosha EL, Kwei GH (1999) Phys Rev B60:9973–
9977
Purrington RD (1997) Physics in the nineteenth century, history of science/physics. Rutgers
university press, New Brunswick, p 164
Rees B (1976) Acta Cryst A32:483–488
Rees B (1977) Isr J Chem 16:180–186
Rietveld HM (1969) J Appl Crystallogr 2:65–71
Ruben AD, Izumi F (2004) Super-fast program PRIMA for the maximum-entropy method,
advanced materials laboratory, National institute for materials science. 1-1 Namiki, Tsukuba,
Ibaraki, Japan 305:0044
Sakata M, Sato M (1990) Acta Cryst A46:263–270
Sakata M, Takata M (1996) High Press Res 14:327–333
64 2 Charge Density Analysis from X-Ray Diffraction

Shmueli U (2008) IUCr series, International tables of crystallography. Volume 2, chapter1, 3rd
edn. Springer, Berlin
Stevens ED, Coppens P (1976) Acta Cryst A32:915–917
Stewart RF (1969) J Chem Phys 51:4569–4577
Stewart RF (1972) J Chem Phys 57:1664–1668
Stewart RF (1976) Acta Cryst A32:545–574
Stout GH, Jensen LH (1968) X-ray structure determination—a practical guide. The Macmillan
Company Collier-Macmillan, London, p 217
Takata M, Umeda B, Nishibori E, Sakata M, Saito Y, Ohno M, Shinohara H (1995) Nature
377:46–49
Takata M, Nishibori E, Shinmura M, Tanaka H, Tanigaki K, Kosaka M, Sakata M (2001) Mater
Sci Eng A312:66–71
Thompson P, Cox DE, Hastings JB (1987) J Appl Crystallogr 20:79–83
Toby BH, Egami T (1992) Acta Cryst A48:336–346
Wagner CNJ (1978) J Non Cryst Solids 31:1–40
Warren BE (1990) X-ray diffraction, chapter 3. Dover publications, New York
Weiss H (1925) Proceedings of the royal society of london series A 108(748):643–654
Wertheim GK, Butler MA, West KW, Buchanan DNE (1974) Rev Sci Instrum 45:1369–1371
Young RA (1995) The rietveld method, Volume 5 of International union of crystallography
momographs on crystallography, Chap 1. Oxford Science publications, Reprint Oxford
University press
Young RA, Mackie PE, Von Dreele RBJ (1977) Appl Crystallogr 10:262–269
Zachiarsen WH (1948) Acta Crstallogr 1:265–8, 277–87
Zachiarsen WH, Ellinger FH (1963) Acta Crystallogr 16:369–75
Chapter 3
Results and Discussion on Metals
and Alloys

Abstract Through centuries of scientific investigation we have come to under-

stand that all materials are composed of atoms and electrons. The number of
electrons surrounding the nucleus defines the characteristics of that atom. Atoms
are held together through electrons in forming a material. Our experience convinces
that there are two types of electron bonding (a) strong bonding and (b) weak
bonding. Strong bonding involves direct electron–electron bonding and can be
differentiated as covalent, ionic and metallic bonding. Weak bonding utilises Van
der Waals’ forces or dipole–dipole interaction in forming bonding. The under-
standing of the bonding may be enhanced by visualisation of the electron bonding
system. This chapter deals with effective analysis of elemental metals, alloys and
dilute doped alloys through versatile techniques such as the maximum entropy
method (Collins 1982), multipole (Hansen and Coppens 1978) and atomic pair
distribution function (Proffen and Billinge 1999). This book deals with the bonding
and electron density distribution of the following metals and alloys through X-ray
single crystal and powder analysis. 1. Sodium and vanadium, 2. Aluminium, nickel
and copper, 3. Magnesium, titanium, iron, zinc, tin and tellurium, 4. Cobalt alu-
minium and nickel aluminium, 5. Nickel chromium (Ni80Cr20), 6. Na1-xAgxCl
(x = 0.03 and 0.10), 7. Aluminium, with iron impurities (0.215 and 0.304 wt% Fe).

3.1 Sodium and Vanadium Metals

3.1.1 Introduction

The study of bonding in metals is interesting because of the various of uses of

metals. The metals sodium and vanadium have found extensive use in the fields of
chemistry, industry, research etc. Sodium and vanadium are useful in their
elemental and compound forms.

R. Saravanan and M. Prema Rani, Metal and Alloy Bonding: An Experimental Analysis, 65
DOI: 10.1007/978-1-4471-2204-3_3, Ó Springer-Verlag London Limited 2012
66 3 Results and Discussion on Metals and Alloys

The precise study of bonding in materials is always useful and interesting yet no
study can reveal the real picture, because no two experimental data are identical.
This problem is enhanced when the model used for the evaluation of electron
densities is not entirely suitable. Fourier synthesis of electron densities can be of
use in picturing the bonding between two atoms, but it suffers from the major
disadvantage of series termination error and negative electron densities which
prevent the clear understanding of the bonding, a factor intended for analysis. The
advent of Maximum Entropy Method (MEM) has solved many of these problems.
MEM electron densities are always positive and even with limited data, one can
determine reliable electron densities resembling the true densities. Currently,
multipole analysis of charge densities has been widely used to study crystalline
materials (Coppens and Volkov 2004; Poulsen et al. 2004; Frils et al. 2004; Pillet
et al. 2004; Marabello et al. 2004). The multipole technique of synthesising the
electron density of an atom into core and valence parts yields an accurate picture
of the bonding in a crystalline system. In this work, such an analysis has been
carried out using the formalism of Hansen and Coppens (1978).
Bonding in materials has been analysed by Israel et al. (2002), Saravanan et al.
(2002a, b), Kajitani et al. (2001) using MEM technique and reported X-ray data
and they obtained precise information about bonding in materials. The nature of
bonding in these materials as analysed are found to be ionic, mixed covalent and
ionic, and ‘oxide’ bonding.

3.1.2 Summary of the Work

The nature of bonding and the charge distribution in sodium and vanadium metals
have been analysed using the reported X-ray data of these metals. MEM and
multipole analysis have been used for elucidation and analysis of bonding in these
metals. The mid-bond densities in sodium and vanadium are found to be 0.014 and
0.723 e/Å3 respectively, giving an indication of the strength of the bonds in these
materials. From the multipole analysis, the sodium atom is found to contract more
than the vanadium atom.

3.1.3 Origin of the Data

In this study on sodium and vanadium, we have attempted to study the metal
bonding in these metals using MEM and multipole techniques. The X-ray structure
factor data of sodium has been taken from Field and Medlin (1974) and that of
vanadium from Linkoaho (1972). Other relevant details can be found from these
papers. Sodium has been chosen in this present study because of its very large
thermal vibration (Field and Medlin 1974) and because it would be interesting to
analyse the highly reactive nature of the metal in terms of strength of bonding and
3.1 Sodium and Vanadium Metals 67

thermal smearing of valence region. In order to make a comparison of the results

with an element having the same structure, vanadium has been chosen, which has a
very low thermal vibration parameter (Linkoaho 1972). Other factors such as the
quality of the data have also been taken into account in selecting these systems.
Hansen and Coppens (1978) proposed a modified electron density model with the
option that allows the refinement of population parameters at various orbital levels
where the atomic density is described as a series expansion in real spherical
harmonic functions through fourth order Ylm. According to this model, the charge
density in a crystal is written as the superposition of harmonically vibrating
aspherical atomic density distribution convolving with the Gaussian thermal dis-
placement distribution as
X
atoms    
qðr Þ ¼ qk !r !
r k!
u  tk !
u ; ð3:1Þ
k

where tk ð~
uÞ is the Gaussian term and  indicates a convolution. The atomic charge
density is then defined as
X4 X l !
! ! 03 0 003 00 r
q r ¼ Pc qcore r þ Pv k qvalence ðk r Þ þ k Rl ðk r Þ Plm Ylm
l¼0 m¼l
r
ð3:2Þ

where Pc, Pv and Plm are population coefficients. Canonical Hartree–Fock atomic
orbitals of the free atoms normalised to one electron can be used for the con-
struction qcore and qvalence but the valence function is allowed to expand and
contract by the adjustment of the variable parameters k0 and k00 : The effect of the
temperature can be distinguished from the convoluted and the deconvoluted form
of thermal contribution to the charge density as dynamic and static multipole
deformation maps. The deformation density in these maps are characterised by

! 1 X ! ! h  ! i
Dqmultipole deformation r ¼ F h F h exp 2 h ; !
r
v multipole spherical atom

ð3:3Þ

where, Fmultipole is the Fourier transform of the multipole charge density with or
without the convolution of thermal contribution, where the Fourier components are
terminated at the experiment resolution.

3.1.4 Data Analysis

Both the reported data sets have been analysed using the GSAS (Larson and
Von Dreele 2004) programme of the Los Alamos National Laboratory, USA.
The reliability index wR for sodium data is found to be 2.2% after successful
68 3 Results and Discussion on Metals and Alloys

Table 3.1 Parameters from Parameters Na V

MEM refinement
Prior electron density (e/Å3) 0.2785 1.6634
Resolution (e/Å) 0.067 0.047
k 0.024 0.024
C 1.0 1.0
R(%) 2.157 0.392
wR(%) 2.356 0.498
Number of cycles 1464 236
k Lagrangean parameter; C Convergence criterion

refinement. The refined Debye–Waller factor is found to be BNa = 7.899 (96) Å2.
This value is very large and comparable with that reported [BNa = 7.86 Å2, Field
and Medlin (1974)], the difference being only 0.039 Å2. The reliability index for
the vanadium refinement is wR = 1.1% and the Debye–Waller factor is found to
be 0.404(24) Å2. This value is also comparable to that reported by Linkoaho
(1972), i.e., BV = 0.37 Å2, the difference being 0.034 Å2.
The reliability indices and the matching of thermal parameters with the
corresponding reported values for both the atoms indicate that the quality of the
data for the electron density analysis is suitable enough. The same number of data
for both the crystal systems has been used for the present analysis in order to make
a comparison also of the bonding (only insignificant changes in the density maps
and numerical results has been observed on including all the reflections for
sodium).
The MEM analysis has been carried out by dividing the cell into 643 pixels. In
the case of sodium data, convergence condition C = 1.0 was obtained after 1464
cycles of refinement; the reliability indices are found to be RMEM = 2.157% and
wRMEM = 2.356%. The Lagrange parameter k used was 0.024 for sodium. For
vanadium, convergence was obtained after 236 cycles with R values as
RMEM = 0.392% and wRMEM = 0.498%. A value of 0.024 was used for k. The
prior densities assumed for sodium and vanadium were 0.2785 and 1.6634 e/Å3,
respectively. The resolution of the map was 0.067 and 0.047 Å/pixel for sodium
and vanadium, respectively. The analysis on the MEM results has been given in
the following section. Some important results from the MEM refinements are
given in Table 3.1.

3.1.5 Results and Discussion

The MEM electron density maps in the density rich regions of sodium and
vanadium on the (100) plane have been given in Fig. 3.1a, b, respectively.
The density distribution of sodium and vanadium on the (110) plane has been
shown in Fig. 3.2a, b, respectively.
3.1 Sodium and Vanadium Metals 69

(a) (b) 3.0

4.0

3.0
2.0
[010]

[010]
2.0

1.0

1.0

Na V

1.0 2.0 3.0 4.0 1.0 2.0 3.0

[001] [001]

Fig. 3.1 a MEM electron density of sodium on (100) plane. Contour lines are between 0.0 and
1.0 e/Å3. Contour interval is 0.05 e/Å3. b MEM electron density of vanadium on (100) plane.
Contour lines are between 0.2 and 4.5 e/Å3. Contour interval is 0.224 e/Å3

(a) 4.0 (b) 3.0

3.0 2.0
[001]

[001]

2.0

1.0
1.0

Na V

0.0
1.0 2.0 3.0 4.0 5.0 6.0 1.0 2.0 3.0 4.0
[110] [110]

Fig. 3.2 a MEM electron density of sodium on (110) plane. Contour lines are between 0.0 and
1.0 e/Å3. Contour interval is 0.05 e/Å3. The plane has been shifted half a unit cell along the z
direction. b MEM electron density of vanadium on (110) plane. Contour lines are between 0.04
and 6.0 e/Å3. Contour interval is 0.298 e/Å3. The plane has been shifted half a unit cell along the
z direction

Figure 3.3a shows the one-dimensional variation of the electron densities of

sodium along the three directions, [100], [110] and [111]. Figure 3.3b shows the
one-dimensional variation of electron density of vanadium along the three crys-
tallographic directions.
The Fig. 3.4a, b represent the error analysis of the data used for the present
analysis.
The density maps of sodium and vanadium show highly resolved electron
density contours suitable for precise analysis. Both the systems belong to the body-
centred cubic structure. Hence, in sodium, only the valence electronic contribution
70 3 Results and Discussion on Metals and Alloys

(a) 1.0 (b) 4.00

0.9 100
111
Electron density (electron/Å3)

Electron density (electron/Å 3)

0.8
110
111 110
3.15
0.7 100
0.6
0.5
2.30
0.4
0.3
0.2 1.45
0.1
0.0

1.275 1.850 2.425 3.000 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Distance (Å) Distance (Å)

Fig. 3.3 a One-dimensional electron density variation of sodium along [100], [110] and [111]
directions. b One-dimensional electron density variation of vanadium along [100], [110] and
[111] directions

Fig. 3.4 a The histogram (a) (b)

of number of reflections 3 6
plotted in the error space
of sodium. b The histogram
Number of reflections

Number of reflections
5
of number of reflections
plotted in the error space of 2 4
vanadium
3

1 2

0 0
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
[Fobs-FMEM ]/σ (K) [Fobs- FMEM ]/σ (K)

of the body-centred atom is seen at the centre of the (100) plane of the cell as seen
from Fig. 3.1a. It also represents the very large thermal vibration of the sodium
atom.
The corners show that the core regions are perfectly symmetric and spherical.
The four voids seen on the (100) plane conform to the crystal system. The conical-
shaped contour lines seem to be attracted towards the centre from the corners. This
may be due to the fact that the core of the centre atom attracts the valence electrons
of the eight corner atoms. Since the atomic number of sodium is only 11, this core–
valence attraction is very flexible and leads to large smearing of the electrons.
Moreover, it should be remembered that the thermal vibration of the sodium atom
being very large (BNa = 7.899(96) Å2), makes this electrostatic interaction pro-
cess highly feasible. The large thermal parameter of sodium and also the smaller
atomic number of sodium makes the body-centred sodium atom possible to reveal
itself at the face centre.
3.1 Sodium and Vanadium Metals 71

Figure 3.1b shows the density distribution of vanadium on the (100) plane. The
centre of the outer contour lines are bent towards the core of each atom, showing
strong attraction of valence electrons by the core, which is due to the large atomic
number of vanadium and its smaller thermal vibration parameter. Two humps are
seen at the outer contour lines of each of the corner atoms. This may be due to the
overlapping of charges between the cores in the bonding direction.
Figure 3.2a, b show the densities of sodium and vanadium respectively, on the
(110) plane (shifted by half the cell distance along z direction). Large thermal
smearing of sodium atom is seen (Fig. 3.2a) as seen from the body-centred atom.
The core of one atom attracting the valence of other atom is also seen. In vanadium
also, this process is taking place as seen from Fig. 3.2b.
From Fig. 3.3a, which shows the one-dimensional variation of the electron
densities of sodium along the three directions, [100], [110] and [111], the mini-
mum density is found at a distance of 1.732 Å at which the electron density is
found to be 0.014 e/Å3. The electron density along the bonding region of the [111]
direction is flat with low level of density similar to the density along the [110]
direction. But along [100] direction, the density is slightly higher than that along
the other two directions. The large thermal vibration of sodium atom and the
smallest inter atomic distance along [100] direction leads to higher density value
along [100]. The inter atomic distances are larger along [110] and [111] directions
and the lone valence electron of sodium atom leads to lower densities along these
directions. As far as the [100] direction is concerned, the amount of electrons
contributing to the bond electron density comes from the corner atom along [100]
direction and the body-centred atom too (due to the large thermal smearing), to
decide the electron density along the shortest [100] direction.
Figure 3.3b shows the one-dimensional variation of electron density of vana-
dium along the three crystallographic directions. There is a slight increasing
undulation of electron density along [111] direction. This starts beyond a distance
of about 0.7 Å from the centre. The electron density at 1.309 Å is found to be
0.723 e/Å3. This is an indication of the bond strength in vanadium compared to
that of sodium, which has a density of 0.014 e/Å3. The small ellipses of charge
density in the [110] plane as seen from Fig. 3.8b and the void in Fig. 3.8a confirm
these magnitudes. The reported (Wilkinson) radii for sodium and vanadium are
1.57 and 1.22 Å. In this work, the radii are estimated from the electron densities
along the [110] direction, because, along [100] direction, the atoms are too close to
decide the inter atomic interactions and along [111] direction, the bonding effects
are predominant. Thus, the estimated radii are 1.517 and 1.136 Å respectively, for
sodium and vanadium, very close to those reported (Wilkinson).
Figure 3.4a, b representing the error analysis of the data used for the present
analysis conform to the quality of good data sets and justify their usage for the
analysis.
The results of the multipole analysis have been given in Table 3.2.
The results of multipole refinement as given in Table 3.2 show the thermal
vibration parameters, the multipole parameters and the reliability indices. The
large j0 parameter of the valence part of the sodium atom (one electron) denotes
72 3 Results and Discussion on Metals and Alloys

Table 3.2 Multipole Parameter Na V

parameters from MEM
refinement Pc 9.941(364) 17.876(403)
Pv 1.132 4.573
k0 1.691(257) 1.108(242)
k00 1.066(187) 1.016(693)
B(Å2) 8.811(713) 0.406(048)
R(%) 2.39 0.95
wR(%) 6.08 1.15
GoF 3.24 2.90
B Debye–Waller factor; Pc population coefficients in the core
region
Pv population coefficients in the valence region; k0 and k00 var-
iable parameters

the contraction needed to restrict the lone valence electron to its nearly ideal value
of one. In this work, the number of core and valence electrons has also been
refined in the multipole formalism, for both sodium and vanadium, which leads to
very close and reasonable values. For sodium, the core and valence electrons are
refined to be 9.94 and 1.13 e- the total being 11.07 e- well within the experi-
mental error limits. Only a large j0 parameter will restrict the large thermal
vibration of sodium atom to regain this amount of electrons in the valence region.
In vanadium, the value of j0 is 1.11, again showing contraction of the atom, but not
as much as in sodium. The number of core and valence electrons in vanadium is
found to be 17.876 and 4.573 e-. There is a deficit of around 0.551 e- in vana-
dium, which actually shows up as a share in the bonding electron and shoots up the
mid-bond density. In sodium there is no deficit in the total number of electrons
leading to meagre share to the bonding and a decrease in the mid-bond density
value. Figure 3.2a, b clearly show these features; clear and density-free regions in
sodium and charge accumulated mid-bond regions in vanadium. Thus MEM and
multipole results are complementary to each other.
The static multipole deformation maps are shown in Fig. 3.5a, b for sodium and
vanadium respectively on the (100) plane. The static multipole deformation maps
are shown in Fig. 3.6a, b for sodium and vanadium respectively on the (‘00)
plane.
Figure 3.5a shows positive electron contours at the centre of the (100) plane,
and at the edge centres and corners, although the values of these densities are very
small. Yet, right from the core, the positive densities are seen showing enhanced
thermal vibration of sodium atom and the contracted spherical core in line with the
j0 parameter of sodium.
The (‘00) section (Fig. 3.6a) also shows the sodium atom with large thermal
vibration of the valence region. Figure 3.5b shows the nullified core and residual
valence region (in terms of thermal vibration) of vanadium atom on (100) plane.
The core is not seen to be as much contracted as in the case of sodium as seen from
Fig. 3.6b.
3.1 Sodium and Vanadium Metals 73

Fig. 3.5 a Static multipole deformation map of sodium on the (100) plane. Dotted lines indicate
negative electron densities. b Static multipole deformation map of vanadium on the (100) plane.
Dotted lines indicate negative electron densities

Fig. 3.6 a Static multipole deformation map of sodium on the (‘00) plane. Dotted lines indicate
negative electron densities. b Static multipole deformation map of vanadium on the (‘00) plane.
Dotted lines indicate negative electron densities

3.1.6 Conclusion

The nature of bonding and the charge distribution in sodium and vanadium metals
has been analysed using the reported X-ray data of these metals. MEM and the
multipole analysis have been used for this purpose. The bonding in these metals
has been elucidated and analysed. The nature of bonding are clearly revealed by
74 3 Results and Discussion on Metals and Alloys

the two-dimensional MEM maps plotted on (100) and (110) planes and the mid-
bond densities are clearly revealed by the one-dimensional electron density along
the [100], [110] and [111] directions. The mid-bond densities in sodium and
vanadium are found to be 0.014 and 0.723 e/Å3 respectively, giving an indication
of the strength of the bonds in these materials. The sodium atom is found to
contract more than the vanadium atom from the multipole analysis.

3.2 Aluminium, Nickel and Copper

3.2.1 Introduction

Technological evolutions result in efforts for developing new and sophisticated

materials of immense use in domestic, technical and industrial applications.
Usually, the synthesis of new materials results in single-phase materials, though
often not in single crystalline form. Hence, a complete analysis of the structure,
local distribution of atoms and electron distribution in core, valence and bonding
regions is necessary using powder diffraction methods, since most of the recent
materials are initially obtained in powder form. Since one can make efforts to grow
single crystals from powders, a prior analysis is required using powders to proceed
for single crystal growth.
In this context, we have taken three simple metals Al, Ni and Cu and collected
powder data sets to study the structure in terms of the local and average structural
properties using pair distribution function (hereafter PDF), the electron density
distribution between atoms using MEM and the bonding of core and valence
electron distribution using multipole technique. Particularly, the PDF analysis
requires data sets of very high values of Q (= 4ksin h/k) which is achievable only
through synchrotron studies and is not always accessible. But this work gives
reasonable results, which could be obtained through single crystal work or through
high Q data sets, using only powder samples with laboratory experiments. Also, a
study on the electronic structure of the metals using the most versatile techniques
like MEM (Collins 1982) and multipole method (Hansen and Coppens 1978)
available today is carried out. If the tools available for the analysis yield highly
precise information, it is appropriate to apply them to the precise available data
sets, as has been done in this work, thereby testing the methodology.
In order to elucidate the distribution of valence electrons, and the contraction/
expansion of atomic shells, multipole analysis of the electron densities is carried
out using the software package JANA 2000 (Petřiček et al. 2000). Recently, the
multipole analysis of the charge densities and bonding has been widely used to
study the electronic structure of materials (Coppens and Volkov 2004; Poulsen
et al. 2004; Frils et al. 2004; Pillet et al. 2004; Marabello et al. 2004). In this work,
the multipole model proposed by Hansen and Coppens (1978) is used for eluci-
dating the electronic structure.
3.2 Aluminium, Nickel and Copper 75

3.2.2 Summary of the Work

The average and local structures of simple metals Al, Ni and Cu have been
elucidated for the first time using MEM, multipole and PDF. The bonding between
constituent atoms in all the above systems is found to be well pronounced and is
clearly seen from the electron density maps. The MEM maps of all the three
systems show the spherical core nature of atoms. The mid-bond electron density
profiles of Al, Ni and Cu reveal the metallic bonding nature. The local structure
using PDF profile of Ni has been compared with that of the reported results. The R
value in this work using low Q XRD data for the PDF analysis of Ni is close to the
value reported using high Q synchrotron data. The cell parameters and displace-
ment parameters were also studied and compared with the reported values.

3.2.3 Data Collection and Structural Refinement

The powder X-ray intensity data was collected at the Regional Research Laboratory
(RRL), Council of Scientific and Industrial Research (CSIR), Thiruvananthapuram,
India, using X-PERT PRO (Philips, Netherlands) X-ray diffractometer with a
monochromatic incident beam, which offers pure Cu Ka1 radiation. The wavelength
used for X-ray intensity data collection is 0.154056 nm with 2h range of data col-
lection from 10 to 1208. The raw intensities were refined using the software pro-
gramme JANA 2000 (Petřiček et al. 2000), considering the fcc unit cell of Al, Ni and
Cu with 4 atoms/unit cell in the space group Fm3m. The fitted profiles of observed
and calculated relative intensities with the Bragg peaks along with the difference are
given in Fig. 3.7a, b, c for Al, Ni and Cu respectively.
The results of the refinements are given in Table 3.3.
The refined structure factors are given in Table 3.4a, b, c respectively, for Al,
Ni and Cu.

3.2.3.1 MEM Refinements

The refined structure factors were used for MEM analysis using the methodologies
discussed in earlier works, e.g., Saravanan et al. (2003), Israel et al. (2003, 2004).
In the present calculation, the unit cell was divided into 643 pixels and the initial
electron density at each pixel was fixed uniformly as Z/a30, where Z is the number
of electrons in the unit cell. The electron density is evaluated by carefully selecting
the Lagrange multiplier in each case such that the convergence criterion C
becomes unity after performing the minimum number of iterations. The resulting
parameters of MEM computations have been given in Table 3.5.
The electron density distribution of Al, Ni and Cu has been mapped using the
MEM electron density values obtained through refinements. Figure 3.8a, b, c show
76 3 Results and Discussion on Metals and Alloys

Fig. 3.7 a The Rietveld refined powder profile for Al along with difference between observed
and calculated relative intensities. b The Rietveld refined powder profile for Ni along with
difference between observed and calculated relative intensities. c The Rietveld refined powder
profile for Cu along with difference between observed and calculated relative intensities

Table 3.3 The Refined structural parameters using JANA 2000

Parameters Al Ni Cu
Cell parameter (Å) 4.05210(4) 3.52660(7) 3.62540(5)
Cell parameter (Å)* 4.0495 3.5238 3.6149
B (Å2) 0.983(80) 0.561(64) 0.432(53)
B (Å2)** 0.746 0.3401 0.526
RP% 9.84 6.21 6.77
wRP% 6.23 7.88 8.63
B Debye–Waller factor

the MEM electron density distribution of Al, Ni and Cu on the (100) plane of the
unit cell.
Figure 3.8d, e, f show the MEM electron density distribution of Al, Ni and Cu
on the (110) plane.
Figure 3.9a, b, c show the one-dimensional variation of the electron densities of
Al, Ni and Cu along the three directions [100], [110] and [111] of the unit cell
respectively.
The electron densities at the saddle points along these three directions have
been given in Table 3.6.
3.2 Aluminium, Nickel and Copper 77

Table 3.4 Observed and calculated structure factors for a Aluminium b Nickel c Copper
hkl Fo Fc Fo - Fc r(Fo)
a Aluminium
111 35.07 35.05 0.03 0.22
0 02 32.87 32.84 0.03 0.16
202 26.63 26.82 -0.19 0.36
113 22.87 23.18 -0.32 0.11
222 22.34 22.17 0.17 0.58
004 18.83 18.64 0.19 0.83
313 15.96 16.47 -0.51 0.22
204 16.23 15.82 0.41 0.37
b Nickel
111 68.16 67.93 0.23 0.42
002 61.40 61.96 -0.57 0.59
202 44.38 45.76 -1.39 0.68
113 38.60 37.59 1.00 0.56
222 34.01 35.38 -1.36 1.31
c Copper
111 78.59 78.49 0.10 0.53
002 72.71 72.65 0.06 0.76
202 55.34 55.46 -0.12 0.87
113 47.00 46.73 0.27 0.75
222 44.06 44.33 -0.27 1.67
004 35.95 36.58 -0.64 2.99

Table 3.5 Parameters of the MEM refinement

System Prior ED Resolution (Å/Pixel) k RMEM (%) wRMEM (%) Number of cycles
(10-3 e/Å3)
Al 0.7806 0.0633 0.010 1.3839 0.7897 1685
Ni 2.5538 0.0551 0.015 1.2011 1.0346 830
Cu 2.4344 0.0567 0.005 1.7610 1.2318 1072
ED Electron density, k Lagrangean multiplier

Table 3.6 Electron density profiles along the three directions of the unit cells
Direction metals [100] [110] [111]
Position (Å) ED (e/Å3) Position (Å) ED (e/Å3) Position (Å) ED (e/Å3)
Al 2.026 0.1467 1.433 0.2636 3.509 0.1467
Ni 1.763 0.2296 1.247 1.1010 3.054 0.2296
Cu 1.813 0.3318 1.282 0.7649 3.139 0.3317
ED Electron density
78 3 Results and Discussion on Metals and Alloys

(a) 1.0 (b)

1.0

0.9 0.9

0.8 0.8

0.7 0.7

0.6 0.6
[001]

[001]
0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[010] [010]
(c) (d)
1.0
1.0
0.9
0.9
0.8 0.8
0.7 0.7
[001]

[001]

0.6 0.6
0.5 0.5

0.4 0.4

0.3 0.3
0.2
0.2
0.1
0.1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[010] [110]

(e) (f)
1.0 1.0
0.9 0.9
0.8 0.8
0.7 0.7
[001]

[001]

0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[110] [110]

Fig. 3.8 a MEM electron density distribution of Al on the (100) plane. Contour range is
0.0–2.0 e/Å3 and contour interval is 0.1 e/Å3. b MEM electron density distribution of Ni on the
(100) plane. Contour range is 0.0–2.0 e/Å3 and contour interval is 0.1 e/Å3. c MEM electron
density distribution of Cu on the (100) plane. Contour range is 0.0–2.0 e/Å3 and contour interval
is 0.1 e/Å3. d MEM electron density distribution of Al on the (110) plane. Contour range is 0.0 to
3.0 e/Å3 and contour interval is 0.15 e/Å3. e MEM electron density distribution of Ni on the (110)
plane. Contour range is 0.0–3.0 e/Å3 and contour interval is 0.15 e/Å3. f MEM electron density
distribution of Cu on the (110) plane. Contour range is 0.0–3.0 e/Å3 and contour interval is
0.15 e/Å3
3.2 Aluminium, Nickel and Copper 79

(a) 1.0 (b) 2.0

0.9 1.8 Al
Cu Cu
0.8 Ni 1.6 Ni
Al
Electron density(e/Å )

Electron density(e/Å )
0.7
3

3
1.4
0.6
1.2
0.5
1.0
0.4
0.8
0.3
0.6
0.2
0.4
0.1
0.0 0.2

1.0 1.5 2.0 2.5 3.0 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3
Distance along [100] direction (Å) Distance along [110] direction (Å)

(c) 1.4
Cu
1.2 Ni
Electron density(e/Å )
3

Al
1.0

0.8

0.6

0.4

0.2

1.5 2.5 3.5 4.5 5.5 6.5

Distance along [111] direction (Å)

Fig. 3.9 a One-dimensional electron density profiles of Al, Ni and Cu along the [100] direction
of the unit cell. b One-dimensional electron density profiles of Al, Ni and Cu along the [110]
direction of the unit cell. c One-dimensional electron density profiles of Al, Ni and Cu along the
[111] direction of the unit cell

A detailed discussion is formulated and given in the following section based on

these two-dimensional MEM maps and one-dimensional profiles.

3.2.3.2 PDF Refinements

For materials whose structure is not reflected in the long-range order of the crystal,
an alternate structural analysis called the Pair Distribution Function (PDF)
approach is used. This method is sometimes called the real-space structure
determination method, because the PDF is modelled in real-space, rather than in
the reciprocal space. The PDF reflects the short-range ordering in a material or the
probability to find an atom at a distance r from another atom. In other words PDF
gives the bond length distribution of the material under consideration (Proffen and
Neder 1997). It well with the inter atomic distances computed from a crystallo-
graphic model, when there are no short-range deviations from the average struc-
ture (Tobi and Egami 1992; Peterson et al. 2003).
This approach has been widely used for studying the structure of glasses and
liquids (Debay and Menki 1930; Bowran and Finney 2003). More recently, it has
been applied to disordered crystalline and partially crystallised materials.
80 3 Results and Discussion on Metals and Alloys

(a) 6.50 Observed

(b)
50 Observed
calculated
Calculated
40
3.25 30
20
G(r) (Å -2)

G(r) (Å -2)
10
0.00
0
-10
-20
-3.25 Difference
Difference
-30
-40

4 6 8 10 12 14 16 18 20 2 4 6 8 10 12 14 16 18 20
r(Å) r(Å)

(c) 0.8500 Observed

Calculated

0.3875
G(r) (Å -2)

-0.0750

-0.5375
Difference

4 6 8 10 12 14 16 18 20
r(Å)

Fig. 3.10 a The observed and calculated PDFs of Al with difference profile. b The observed and
calculated PDFs of Ni with difference profile. c The observed and calculated PDFs of Cu with
difference profile

Quantitative structural information about nanometer length scales can be obtained

by fitting a model directly to the PDF (Proffen and Billinge 1999) based on the
equation,
Z
2 !h ! i !  !
Gðr Þ ¼ 4pr ½qðr Þ  q0  ¼ Q S Q  1 sin Q r d Q ð3:4Þ
p
where G(r) is the atomic pair distribution function and qðr Þ corresponds to
the (atomic) number density at a distance r from the average atom. The atomic
pair distribution function thus obtained from powder diffraction data is a valuable
tool for the study of the local atomic arrangements in a material, since both
Bragg and diffuse scattering information about local arrangements are preserved
in PDF.
In this study, the observed PDFs of Al, Ni and Cu were obtained from the raw
intensity data using a software programme PDFgetX (Jeong et al. 2001) after
performing corrections such as multiple scattering correction, polarisation cor-
rection, absorption correction, normalisation correction and Compton correction.
A comparison between the observed and calculated PDF has been carried out using
the software package PDFFIT (Proffen and Billinge 1999). Figure 3.10a–c give the
3.2 Aluminium, Nickel and Copper 81

Table 3.7 Distances from System First nearest neighbour distance (Å)
the PDF profiles
Observed Calculateda
Al 2.86 2.863
Ni 2.50 2.492
Cu 2.38 2.564
a
Wyckoff 1963

fitted observed and calculated PDF profiles along with the difference between them
for Al, Ni and Cu respectively, and Table 3.7 gives the bond length distribution of
atoms.

3.2.3.3 Multipole Refinements

The multipole model represents an extrapolation from a finite set of experimental

data. An important feature of the multipole model is the possibility to adjust the
radial dependency for each atom type by including the expansion/contraction
controlling j0 parameters in the refinement, and possibly also the j00 parameters, to
change the radial dependency of the valence deformation density.
In our study, we have used a modified electron density model proposed by
Hansen and Coppens (1978) with the option that allows the refinement of popu-
lation parameters at various orbital levels, multipoles up to seventh order and
corresponding radial expansion (j0 ) parameters. The j0 parameters were refined in
separate cycles together only with the scale factor, and convergence was achieved
in an iterative process. Each refinement cycle was considered successful at the
point at which the maximum shift/s.u. was less than 0.001. The parameters and
results obtained from multipole refinements are given in Table 3.8.
The effect of the temperature can be distinguished from the convolution and de-
convolution form of thermal contribution to the charge density as dynamic (DMD)
and static multipole deformation (SMD) maps. The SMD maps offer a chance to
compare the electron densities without the disturbance of the thermal vibration of
atoms. Hence, the SMD maps have been computed and presented in Fig. 3.11a, b,
c for Al, Ni and Cu respectively.

3.2.4 Results and Discussion

The fitted profiles of Rietveld refinements (Fig. 3.7a, b, c for Al, Ni and Cu
respectively) give a clear picture of the quality of the sample as well as the data.
The accuracy of refinements can be judged from the difference curve between the
observed and calculated intensity, which is almost a straight line. The results of the
82 3 Results and Discussion on Metals and Alloys

Table 3.8 Parameters from Parameters Al Ni Cu

multipole refinement
j0 1.0069(0521) 1.1132(0136) 1.0648(3447)
B (Å2) 0.9831(0809) 0.5604(0640) 0.4317(0533)
R (%) 0.84 2.12 0.59
wR (%) 1.2 1.61 0.36
GoF 1.16 1.16 1.09
qmax (SMD) 0.05 0.02 0.02
qmin (SMD) -0.02 -0.08 -0.01
B Debye–Waller factor; j0 variable parameter

Fig. 3.11 a Static Multipole Deformation map of Al on (100) plane. Dotted lines indicate
negative electron densities. b Static Multipole Deformation map of Ni on (100) plane. Dotted
lines indicate negative electron densities. c Static Multipole Deformation map of Cu on (100)
plane. Dotted lines indicate negative electron densities
3.2 Aluminium, Nickel and Copper 83

structural refinement (Petřiček et al. 2000) given in Table 3.3 show reasonable
values of the Debye–Waller factor for Al, Ni and Cu obtainable from a powder
data set, which is comparable to those reported in Peng et al. (1996). The
experimental cell constant of Al, Ni and Cu are highly comparable to the reported
values (Wyckoff 1963) and the differences are 0.0026, 0.0028 and 0.0096 Å for
Al, Ni and Cu respectively, which are very low indicating the accuracy of
refinements and also the precision of the observed data sets. The reliability indices
for all the powder samples are very low indicating the correctness of the refine-
ments (RP = 9.84, 6.21 and 6.77 for Al, Ni and Cu respectively).
The MEM electron density distribution map of Al as given in Fig. 3.8a for
(100) plane reveals the core of the Al atom being spherical and this sphericity
persists even at slightly larger distances away from the centre. The distribution of
charges all over the plane, indicates the distribution of electrons and the charges at
the edge centres on (100) plane show the distribution from the perpendicular face-
centred atoms, indicating the extension of the spread of the charges. The electron
density map of Al on (110) plane shows (Fig. 3.8b) highly concentrated charges at
the core of the atoms, and the distribution of charges on places other than the
atomic positions, due to the valence, free electrons.
The core of Ni and Cu atom as seen from the electron density map on (100)
plane (Fig. 3.8b, c), show the perfect spherical nature of the electronic charge
clouds. The edge centres show the distribution of electron density of the atoms
located at the face centres of the planes perpendicular to the paper. Since the
atomic numbers of nickel and copper are higher than that of aluminium, more
diffuse distribution of charges are seen in the electron density maps of Ni and Cu
on the (100) plane. The electron density map of Ni and Cu on the (110) plane
(Fig. 3.8e, f), show similar trends as observed in (100) plane.
The one-dimensional profiles of electron density constructed along [100], [110]
and [111] directions for Al, Ni and Cu are shown in Fig. 3.9a, b and c respectively.
The positions of minimum electron densities and density values have been given in
Table 3.6. The mid-bond density for Al is found to be 0.263 e/Å3 at a distance of
1.433 Å along the bonding direction [110]. The mid-bond electron density
between the Al atoms along [100] direction is 0.147 e/Å3 at a distance of 2.026 Å.
The first minimum and the mid-bond electron density between the atoms along
[111] direction are 0.119 e/Å3, 0.147 e/Å3 at a distance of 1.755 and 3.509 Å
respectively.
The mid-bond density for Ni is found to be 1.101 e/Å3 at a distance of 1.247 Å
along the bonding direction [110]. Table 3.6 shows that the mid-bond density
along [100] direction of Ni occurs at a distance 1.763 Å with a value 0.229 e/Å3.
The first and second minimum electron densities between the atoms along [111]
directions are 0.841 and 0.229 e/Å3 at a distance of 1.527 and 3.054 Å.
The mid-bond density for Cu is found to be 0.765 e/Å3 at a distance of 1.282 Å
along the bonding direction [110]. The mid-bond density along [100] direction
occurs at a distance 1.813 Å with a value 0.332 e/Å3. The first and second min-
imum electron densities between the atoms along [111] direction are 0.664 and
0.332 e/Å3 at a distance of 1.374 and 3.139 Å respectively. The mid-bond density
84 3 Results and Discussion on Metals and Alloys

in Ni is the largest value (Table 3.6) among the three metal systems we have
studied. But along the [100] and [111] directions (non-bonding), the element with
the higher atomic number has the higher electron density. Hence obviously the
bonding interaction is stronger in Ni as evidenced by the higher electron density
along [110] direction.
Hence, from the qualitative and quantitative analysis of the MEM electron
densities, the bonding in Al, Ni and Cu is predicted to be predominantly metallic
and the interaction of charges along other nonb-onding directions for Ni and Cu
seems to be heavier due to their metallic nature. (Because, if the systems are ionic
one cannot expect higher mid-bond values as in the present cases and if the
systems are covalent, one can expect values higher than the present values due to
overlapping of charges). Also the strength of the electron density is higher in the
case of Ni, which has lower number of electrons in the unit cell compared to Cu.
The electron density of Al in all the three directions [100], [110] and [111] are
relatively low compared to Ni and Cu, which conforms to the loosely packed
structure of Al. (The interaction of atomic charges will be less and hence electron
densities along bonding and directions other than bonding are expected to be
minimum). !
The refined pair distribution function (the Fourier transform of S Q ; the
reduced structure factor) given in Fig. 3.10a, b, c for Al, Ni and Cu show nice
matching of observed and calculated PDFs. The PDF refinement can be considered
as equivalent to matching the observed X-ray powder data with a model with too
little structure parameters. Moreover, high Q data are required for this type of
analysis on short-range order and local structure of materials (Egami 1990). In
spite of these factors, an attempt has been made to use the X-ray powder data for
this analysis, to test how well these powder data sets can be used for purposes like
PDF analysis.
The peaks in the PDF profile correspond to the nearest neighbour distances
given in Table 3.7. In Al, the difference between observed and calculated first
nearest neighbour distances turns out to be 0.003 Å. In Ni, the difference is
0.008 Å and for Cu it turns out to be 0.184 Å. These differences are not too high
considering the fact that we have used only (i) powder data with (ii) limited Q
values. There may also be local undulations in the structure of these three systems
which lead to these differences. The local disorder is expected to affect the repeat
distances, because of the diffuse scattering, which does not have sharp, single
pointed X-ray diffraction phenomenon. Hence, it is reflected in the local structural
analysis using suitable tools like PDF. From these differences (0.003 Å for Al,
0.008 Å for Ni and 0.184 Å for Cu), one can conclude that (i) the local disorder
increases as the atomic number increases due to more concentration of charges and
(ii) the disorder propagates at a longer distance from the origin leading to
enhanced differences in neighbour distances. In the case of Cu, an additional
disorder is possible due to the fact that CuKa radiation has been used for the
powder data collection which gives rise to resonance effects and possible
fluorescence.
3.2 Aluminium, Nickel and Copper 85

The PDF profile of Ni has been compared with that obtained with reported data
(Proffen and Billinge 1999). Very close resemblance is seen between these two
refinement results. The R values are also close to each other [18% in this work
compared to 14% in the work by Proffen and Billinge (1999)].
The multipole refinement shows no expansion/contraction for Al as seen from
the j0 value given in Table 3.8 [j0 = 1.007(0.052)]. In the case of Cu, there is a
slight contraction of the atom, but in Ni there is appreciable contraction compared
to the other two systems. There is a possibility that the contraction is effected to
obey Pauli’s principle and to make the system with the strongest mid-bond
interaction stable (as evidenced by the mid-bond electron density of Ni).
The results of SMD calculations given in Table 3.8 show the positive and
negative (maximum and minimum) electron densities for all the three systems.
Since these calculations are based on difference densities, the lowest electron
densities are seen for all the systems, i.e., the maximum positive difference is
0.05 e/Å3 for Al and the maximum negative density is -0.08 e/Å3 for Ni. These
values indicate the accuracy with which the multipole analysis has been carried
out. Moreover, the positive difference densities are the same for both Ni and Cu
(0.02 e/Å3), thereby justifying the fact that no biassed results are obtained as far as
the comparison of electron densities of Ni and Cu is concerned.

3.2.5 Conclusion

A clear understanding of the structure of the metals A1, Ni and Cu has been gained
from this work, in terms of cell parameters, thermal vibrations and electron density
distribution in the bonding region using MEM and local structural analysis using
PDF. Multipole model analysis reveals the clear qualitative and quantitative pic-
ture of the charge densities both in core and valance regions with the expansion
and contraction of the valence region.

3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium

3.3.1 Introduction

The applications of metallic materials are spread into various areas of science and
technology including electronics, aeronautics and space travel, to name a few.
Hence a study on some technologically important metals is essential in terms of
the local and average structures which are completely different. The usual methods
of analysis using structural refinement of X-ray or neutron data will give only the
average structure of the materials under investigation. There is only limited
information available about the investigations of materials in terms of the local
86 3 Results and Discussion on Metals and Alloys

structure. In the present investigation, powder data sets of six elemental metals
magnesium, titanium, iron, zinc, tin and tellurium were collected to study the
structure in terms of the electron density distribution between atoms using MEM
and the local and average structural properties using PDF. The PDF analysis
requires data sets of very high values of Q (= 4k sin h/k) which is achievable only
through synchrotron studies, which is not always accessible. But this work gives
reasonable results, which could be obtained through high Q data sets.

3.3.2 Summary of the Work

The average and local structures of the metals magnesium, titanium, iron, zinc, tin
and tellurium have been analysed using MEM, and PDF. The bonding between the
constituent atoms in all these systems is found to be well pronounced and is clearly
seen from the electron density maps. The mid-bond electron density profiles reveal
the metallic nature of the bonding. The local structure of Mg, Ti, Fe, Zn, Sn and Te
using the low Q XRD have given results comparable to high Q synchrotron data.
The cell and displacement parameters are also studied.

3.3.3 Maximum Entropy Method

For the accurate charge density study, the MEM (Sakata and Sato 1990;
Collins 1982) has been used. It is a non-linear calculation that has its roots in
information and probability theory, and was originally designed to reconstruct the
most probable and least biassed probability distribution in an underdetermined
situation. The MEM can yield a high-resolution density distribution from a limited
number of diffraction data without using a structural model. The obtained density
distribution gives detailed structural information in the inter atomic region, e.g.
bonding electron densities. The ability of MEM in terms of a model-free recon-
struction of the charge densities from measured X-ray diffraction data can be
interpreted as ‘imaging of diffraction data’. Combining MEM with the Rietveld
method (Rietveld 1969) has been successful in developing a sophisticated method
of structure refinement in charge density level, that is, the MEM/Rietveld (Takata
et al. 1995). The MEM/Rietveld analysis is an iterative method combining the
MEM and Rietveld analyses. The MEM charge density at certain iteration step
normally succeeds in providing a better structural model for the Rietveld of the
next iteration. In this method, the final MEM electron density distribution derived
is compatible with the structural model used in Rietveld refinement. The method
has been successfully applied to the structural studies of metals and alloys (Robert
et al. 2009; Saravanan and Prema Rani 2007), semiconductors (Saravanan et al.
2008), magnetic materials (Syed Ali et al. 2009]) In this study this method has
been used to elucidate the electron density distribution of elemental metals.
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 87

3.3.4 Pair Distribution Function

Conventional structure determination is based on the analysis of the intensities and

positions of Bragg reflections which only allow the determination of the long-
range average structure of the crystal. The key to a deeper understanding of the
properties of the materials is the study of deviations from the average structure or
the study of the local atomic arrangements. Deviations from the average structure
result in the occurrence of diffuse scattering which contain information about two-
body interactions. One method to reveal the local structure of crystals is the
analysis of PDF. PDF can be understood as a bond length distribution between all
pairs of atoms within the crystal (up to a maximum distance). PDF is the
instantaneous atomic number density–density correlation function which describes
the atomic arrangement in materials. It is the sine Fourier transform of the
experimentally observable total structure factor obtained from a powder diffraction
experiment. Since the total structure factor includes both the Bragg scattered
intensities and the diffuse scattering part of the diffraction spectrum its Fourier
associate, the PDF, yields both the local and average atomic structures of mate-
rials. Determining the PDF has been the approach of choice for characterising
glasses, liquids and amorphous materials for a long time. However, its widespread
application to crystalline materials, where some deviation from the average
structure is expected to take place, has been relatively recent. The PDF refinement
can be considered as equivalent to matching the observed X-ray powder data with
a model with too little structure parameters.
Quantitative structural information about nanometer length scales can be
obtained by fitting a model directly to the PDF (Proffen and Billinge 1999) based
on the equation

Gðr Þ ¼ 4pr½qðr Þ  q0 
Z
2 !h ! i !  !
¼ Q S Q  1 sin Q r d Q ;
p

where G(r) is the atomic pair distribution function and qðr Þ corresponds to the
(atomic) number density at a distance r from the average atom. The atomic pair
distribution function, obtained from powder diffraction data is thus a valuable tool
for the study of the local atomic arrangements in a material, since both Bragg and
diffuse scattering information about local arrangements are preserved in PDF.

3.3.5 Data Collection and Structural Refinement

High quality powder samples of Mg, Ti, Fe, Zn, Sn and Te were purchased and
used for this work. The general properties of these metals are given in Table 3.9.
 88

Table 3.9 General properties of the metals

Propertiesa Mg Ti Fe Zn Sn Te
Atomic number 12 22 26 30 50 52
Cell constant(Å) a = 3.2095 a = 2.9512 a = b = c = 2.8665 a = 2.6650 a = 5.8317 a = 4.4568
b = 3.2095 b = 2.9512 b = 2.6650 b = 5.8317 b = 4.4568
3

c = 5.2107 c = 4.6845 c = 4.9470 c = 3.1324 c = 5.9270

Structure Hexagonal Hexagonal Cubic Hexagonal Tetragonal Hexagonal
Space group P63/mmc P63/mmc Im-3 m P63/mmc I41/amd P3121
a
CRC handbook (Ram 1985)
Results and Discussion on Metals and Alloys
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 89

Fig. 3.12 a The Rietveld refined powder profile for Mg along with difference between observed
and calculated relative intensities. b The Rietveld refined powder profile for Ti along with
difference between observed and calculated relative intensities. c The Rietveld refined powder
profile for Fe along with difference between observed and calculated relative intensities. d The
Rietveld refined powder profile for Zn along with difference between observed and calculated
relative intensities. e The Rietveld refined powder profile for Sn along with difference between
observed and calculated relative intensities. f The Rietveld refined powder profile for Te along
with difference between observed and calculated relative intensities
90 3 Results and Discussion on Metals and Alloys

Fig. 3.12 (continued)

3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 91

Table 3.10 Observed and calculated structure factors for a Magnesium b Titanium c Iron d Zinc
e Tin f Tellurium
hkl Fo Fc Fo-Fc r(Fo)
a Magnesium
220 130.19 130.30 -0.11 3.41
101 109.31 109.82 -0.51 0.95
200 158.93 158.44 0.49 0.73
220 130.19 130.30 -0.11 3.41
211 93.60 93.52 0.08 1.34
301 78.01 78.25 -0.24 2.57
112 103.02 103.72 -0.70 2.13
400 95.89 101.90 -6.01 3.17
321 71.97 72.39 -0.42 2.05
420 96.46 93.20 3.26 3.03
411 64.56 62.40 2.17 2.03
312 89.12 85.19 3.93 2.15
501 65.04 54.87 10.17 3.40
431 57.19 48.25 8.94 2.99
103 51.03 47.87 3.16 2.16
332 70.87 66.37 4.50 2.76
440 69.55 64.68 4.87 2.57
521 53.62 47.79 5.83 2.20
213 40.20 43.10 -2.90 2.14
600 56.13 60.17 -4.03 2.99
303 34.93 40.05 -5.12 1.22
512 48.75 55.56 -6.81 1.64
620 47.57 54.20 -6.63 1.60
611 33.15 37.77 4.63 1.12
b Titanium
100 8.85 8.93 -0.09 0.20
002 17.47 17.46 0.01 0.07
101 14.83 14.79 0.04 0.04
102 7.64 7.63 0.01 0.16
2–10 13.60 13.76 -0.16 0.34
103 11.00 11.17 -0.17 0.21
200 5.97 6.14 -0.17 0.61
2–12 12.51 12.10 0.41 0.30
201 10.17 10.31 0.14 0.34
004 11.44 11.56 -0.12 0.55
202 5.53 5.43 0.10 0.10
104 5.28 5.20 0.09 0.24
203 7.76 8.13 -0.36 0.41
3–10 4.52 4.49 0.02 0.85
3–11 7.66 7.57 0.09 0.35
2–14 8.08 8.52 -0.44 0.38
105 6.85 6.99 -0.15 0.28
3–12 3.97 4.03 -0.06 0.16
(continued)
92 3 Results and Discussion on Metals and Alloys

Table 3.10 (continued)

hkl Fo Fc Fo-Fc r(Fo)
204 3.29 3.88 -0.59 0.74
300 8.00 7.45 0.55 1.25
3–13 5.99 6.14 -0.15 0.34
c Iron
101 34.32 34.31 0.01 0.82
002 25.73 26.36 -0.63 3.28
112 21.46 22.31 -0.85 2.62
202 18.57 19.02 -0.44 2.62
103 17.48 16.64 0.84 2.60
d Zinc
002 40.22 40.08 0.14 0.31
100 20.71 20.57 0.14 0.19
101 29.06 29.06 0.00 0.13
102 13.67 13.98 -0.31 0.16
103 16.33 16.52 -0.19 0.16
2-10 20.75 21.14 -0.40 0.19
004 13.09 14.30 -1.21 0.80
2–12 14.89 14.63 0.26 0.25
200 6.34 6.08 0.26 0.59
201 12.98 13.09 -0.11 0.26
104 4.38 1.99 2.39 1.40
202 5.95 6.34 -0.40 0.36
203 2.02 2.40 -0.38 0.58
105 1.44 1.74 -0.30 0.56
2–14 1.35 1.62 -0.28 0.52
e Tin
101 109.31 109.82 -0.51 0.95
200 158.93 158.44 0.49 0.73
220 130.19 130.30 -0.11 3.41
211 93.60 93.52 0.08 1.34
301 78.01 78.25 -0.25 2.57
112 103.02 103.72 -0.70 2.13
400 95.90 101.90 -6.01 3.17
321 71.97 72.39 -0.42 2.05
420 96.46 93.20 3.26 3.03
411 64.56 62.40 2.17 2.03
312 89.12 85.19 3.93 2.15
501 65.04 54.87 10.17 3.40
431 57.19 48.25 8.94 2.99
103 51.03 47.87 3.16 2.16
332 70.87 66.37 4.50 2.76
440 69.55 64.68 4.87 2.57
521 53.62 47.79 5.83 2.20
213 40.20 43.10 -2.90 2.14
600 56.13 60.17 -4.03 2.99
(continued)
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 93

Table 3.10 (continued)

hkl Fo Fc Fo-Fc r(Fo)
303 34.93 40.05 -5.12 1.22
512 48.75 55.56 -6.81 1.64
620 47.57 54.20 -6.63 1.60
611 33.15 37.77 4.63 1.12
f Tellurium
100 36.84 37.00 -0.16 0.82
1–11 27.50 27.52 -0.02 0.34
101 21.66 121.77 -0.11 1.52
102 24.56 24.73 -0.17 0.49
1–12 107.46 108.20 -0.74 2.14
2–10 92.11 92.39 -0.29 1.73
2–11 44.26 44.56 -0.30 1.68
003 108.01 104.82 3.19 5.31
200 35.72 35.56 0.16 2.55
201 56.89 55.73 1.16 1.50
2–21 79.26 77.64 1.61 2.09
2–12 37.87 38.72 -0.85 2.32
1–13 25.72 25.36 0.35 2.84
103 24.46 24.13 0.34 2.70
202 67.88 72.43 -4.55 2.60
2–22 47.88 51.09 -3.21 1.83
2–13 75.057 72.98 2.08 2.58
3–10 54.41 57.02 -2.61 1.80
3–21 37.44 38.82 -1.38 1.34
3–11 49.38 51.20 -1.82 1.77
1–14 15.67 15.20 0.46 0.73
104 79.78 77.42 2.36 3.70
2–23 26.25 25.95 0.30 1.25
203 28.45 28.45 0.33 1.37
3–12 35.00 37.31 -2.31 2.00
3–22 45.53 48.54 -3.01 2.61
300 39.75 38.33 1.43 4.28
301 30.89 27.13 3.78 1.84
3–31 69.52 61.05 8.47 4.15
2–14 25.24 25.24 -2.59 8.76
3–32 23.78 23.64 0.14 1.13
302 55.26 54.92 0.33 2.63
2–24 56.21 56.14 56.14 2.13
204 32.87 32.83 0.04 1.24
3–13 42.36 42.31 0.05 1.60
3–23 44.61 44.55 0.06 1.69
1–15 68.28 63.03 5.25 7.01
105 13.90 12.83 1.07 1.43
4–20 27.79 25.09 2.70 7.39
4–21 47.27 43.46 3.81 3.65
(continued)
94 3 Results and Discussion on Metals and Alloys

Table 3.10 (continued)

hkl Fo Fc Fo-Fc r(Fo)
303 34.53 35.34 -0.81 3.82
3–33 33.62 34.41 -0.79 3.72
4–10 27.16 26.50 0.66 2.36
4–11 50.46 52.94 -2.48 2.48
4–31 19.28 20.23 -0.95 0.95
2–15 17.99 19.28 -1.29 0.76
4–22 36.61 39.04 -2.44 1.62
3–14 38.03 41.12 -3.09 1.95
3–24 23.21 25.10 -1.89 1.19
2–25 27.40 27.03 0.37 1.63
205 49.03 48.37 0.66 2.92
4–12 19.90 20.61 -0.71 1.20
4–32 48.21 49.94 -1.73 2.90
006 52.05 54.31 -2.25 16.61
3-34 40.97 40.88 0.10 3.07
304 22.50 22.45 0.05 1.68
4–23 24.25 23.86 0.38 1.72
400 55.51 53.59 1.92 3.20
106 9.49 9.42 0.07 0.66
1–16 10.89 10.81 0.09 0.76
401 10.10 10.05 0.95 1.35
4–41 12.94 11.82 1.12 1.60
4–13 21.82 24.54 -2.72 4.64
4–33 18.71 21.04 -2.33 3.98
3–15 20.91 23.51 -2.60 2.18
3–25 32.03 36.01 -3.98 3.34
402 12.93 14.41 -1.48 1.32
4–42 9.29 10.35 -1.06 0.95
2–16 41.09 43.38 -2.29 3.13
4–24 29.54 31.93 -2.39 2.00
4–10 27.16 26.50 0.66 2.36
4–11 50.49 52.94 -2.48 2.48
4–31 19.28 20.23 -0.95 0.95
2–15 17.99 19.28 -1.293 0.76
4–22 36.61 39.04 -2.44 1.61
3–14 38.03 41.12 -3.09 1.95
3–24 23.21 25.10 -1.89 1.19
2–25 27.40 27.03 0.37 1.63
205 49.03 48.37 0.66 2.92
4–12 19.90 20.61 -0.71 1.20
4–32 48.21 49.94 -1.73 2.90
006 52.05 54.31 -2.25 16.61
3–34 40.97 40.88 0.10 3.07
304 22.50 22.45 0.05 1.68
Table 3.11 Refined structural parameters from JANA 2006
Parameters Mg Ti Fe Zn Sn Te
B (Å2) 0.87(00) 0.55 (09) 0.55(00) 1.46(00) 2.08(00) 1.53(41)
B (Å2) 1.46 – 0.35 1.26 – 0.74
Robs(%) 1.91 1.65 2.36 1.76 4.96 3.81
wRobs(%) 0.82 1.71 1.18 1.16 3.73 3.33
Rp(%) 10.03 8.92 20.89 9.08 15.84 17.11
wRp(%) 14.23 13.08 28.94 12.72 23.08 23.84
Cell parameters a = 3.2138 a = 2.954 (0.0000) a = b = c = 2.8692 a = 2.6733 a = 5.8022 a = 4.4634
(Å) (0.0011) (0.00) (0.0000) (0.0000) (0.0044)
b = 3.2137 b = 2.954 (0.0000) b = 2.6733 b = 5.8022 b = 4.4634
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium

(0.0000) (0.0000) (0.0000) (0.0000)

c = 5.2150 c = 4.6916 c = 4.9652 c = 3.1654 c = 5.9330
(0.0019) (0.0001) (0.0000) (0.0000) (0.0058)
B reported Debye–Waller factor
95
96 3 Results and Discussion on Metals and Alloys

Table 3.12 Refined parameters from MEM technique

Metal Parameters
Number of cycles Prior density s(ri) Lagrange parameter (k) RMEM(%) wRMEM (%)
(e/Å3)
Mg 159 0.5150 0.0026 2.5822 1.4972
Ti 45 1.2394 0.0052 1.2171 1.1493
Fe 60 2.2034 0.0327 8.5663 4.6957
Zn 205 1.9544 0.0308 2.7252 1.0814
Sn 92 1.8762 0.0423 9.7379 2.1222
Te 222 1.5220 0.0813 3.7133 3.8628

The powder X-ray intensity data were collected at the National Institute of
Interdisciplinary Science and Technology (NIIST), Trivandrum, India, using
X-PERT PRO (Philips, Netherlands) X-ray diffractometer with a monochromatic
incident beam, which offers pure Cu Ka1 radiation. The wavelength used for X-ray
intensity data collection is 0.154056 nm with 2h range of data collection from 10
to 120° and 0.05° step size.
The structural parameters of the metals were refined with the well-known
(Rietveld 1969) powder profile fitting methodology using the software package
JANA 2006 (Petřiček et al. 2006). The initial cell parameters and space group were
fixed as given in Table 3.9. The fitted profiles and the positions of Bragg peaks for
the six metals have been shown in Fig. 3.12a–f. In these figures, the dots represent
the observed powder patterns and the continuous lines represent the calculated
powder profiles of the respective samples. The small vertical lines indicate Bragg
positions. The difference between the observed profile and fitted calculated profile
is shown at the bottom of each figure. The refined profiles show good matching
between the observed and calculated profiles for all the three systems.
The refined structure factors of the samples given in Tables 3.10a–f show very
good agreement of the observed and calculated structure factors with very low
statistical errors.
The refined structural parameters from the software JANA 2006 are given in
Table 3.11. The refinement indices are low which show the accuracy of the
refinement (Rp = 10.03, 8.92, 20.89, 9.08, 15.84, 17.11% respectively, for sam-
ples Mg, Ti, Fe, Zn, Sn and Te).

3.3.6 MEM Refinements

The refined structure factors were used for MEM analysis. In the present calcu-
lation, the unit cell was divided into 643 pixels and the initial electron density at
each pixel was fixed uniformly as Z/a30, where Z is the number of electrons in the
unit cell. The electron density is evaluated by carefully selecting the Lagrange
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 97

Fig. 3.13 a Three-dimensional MEM electron density isosurface of Mg. b Three-dimensional

MEM electron density isosurface of Ti. c Three-dimensional MEM electron density isosurface of
Fe. d Three-dimensional MEM electron density isosurface of Zn. e Three-dimensional MEM
electron density isosurface of Sn. f Three-dimensional MEM electron density isosurface of Te

multiplier in each case such that the convergence criterion C becomes unity after
performing minimum number of iterations. The resulting parameters of MEM
computations are given in Table 3.12.
The electron density distributions of Mg, Ti, Fe, Zn, Sn and Te have been
mapped using the MEM electron density values obtained through refinements.
Three-dimensional MEM electron density isosurface of Mg, Ti, Fe, Zn, Sn and Te
are shown in Fig. 3.13a–f respectively.
Figure 3.14a–f show the MEM electron density distributions of Mg, Ti, Fe, Zn,
Sn and Te on the (100) plane of the unit cell.
Figure 3.15a–f show the MEM electron density distributions of Mg, Ti, Fe, Zn,
Sn and Te on the (110) plane.
Figure 3.16a, b, c show the one-dimensional variation of the electron densities
of Mg, Ti, Fe, Zn, Sn along the three directions [100], [110] and [111] of the unit
cell, respectively.
98 3 Results and Discussion on Metals and Alloys

Fig. 3.14 a MEM electron density of magnesium on (100) plane. Contour lines are between 0.02
and 2.0 e/Å3. Contour interval is 0.33 e/Å3. b MEM electron density of titanium on the (100)
plane. Contour lines are between 0.05 and 5.0 e/Å3. Contour interval is 0.33 e/Å3. c MEM
electron density of iron on (100) plane. Contour lines are between 0.03 and 5.0 e/Å3. Contour
interval is 0.33 e/Å3. d MEM electron density of zinc on the (100) plane. Contour lines are
between 0.05 and 5.0 e/Å3. Contour interval is 0.33 e/Å3. e MEM electron density of tin on (100)
plane. Contour lines are between 0.05 and 5.0 e/Å3. Contour interval is 0.33 e/Å3. f MEM
electron density of tellurium on the (100) plane. Contour lines are between 0.05 and 5.0 e/Å3.
Contour interval is 0.001 e/Å3
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 99

Fig. 3.15 a MEM electron density of magnesium on (110) plane. Contour lines are between 0.0
and 2.0 e/Å3. Contour interval is 0.02 e/Å3. b MEM electron density of titanium on the (110)
plane. Contour lines are between 0.05 and 5.0 e/Å3. Contour interval is 0.001 e/Å3. c MEM
electron density of iron on (110) plane. Contour lines are 0.0 and 5.0 e/Å3. Contour interval is
0.03 e/Å3. d MEM electron density of zinc on (110) plane. Contour lines are between 0.00 and
10.0 e/Å3. Contour interval is 0.05 e/Å3. e MEM electron density of tin on (110) plane. Contour
lines are between 0.00 and 5.0 e/Å3. Contour interval is 0.05 e/Å3. f MEM electron density of
tellurium on the (110) plane. Contour lines are between 0.00 and 5.0 e/Å3. Contour interval is
0.05 e/Å3
100 3 Results and Discussion on Metals and Alloys

(a) 0.4 Fe Sn
(b) 0.3 Sn
Sn Ti
Fe Fe
Te Mg
0.3 Mg Te

Electron density (e/Å 3 )

Electron density (e/Å 3)

Zn Zn
Ti 0.2
Fe Ti
Te
0.2 Ti

Te 0.1
Mg
0.1
Mg
Sn
Zn
Zn
0.0 0.0
0.40 1.05 1.70 2.35 3.00 0 1 2 3

Distance along [100](Å) Distance along [110](Å)

(c) 0.6 Fe Sn
Ti
0.5 Fe
Zn Zn
Te
Electron density (e/Å 3 )

Sn Mg
0.4 Te

0.3

0.2 Ti

Mg
0.1

0.0
0.20 0.85 1.50 2.15 2.80

Distance along [111](Å)

Fig. 3.16 a One-dimensional electron density profiles of Mg, Ti, Fe, Zn, Sn, Te along the [100]
direction of the unit cell. b One-dimensional electron density profiles of Mg, Ti, Fe, Zn, Sn, Te
along the [110] direction of the unit cell. c One-dimensional electron density profiles of Mg, Ti,
Fe, Zn, Sn, Te along the [111] direction of the unit cell

3.3.7 Results and Discussion

3.3.7.1 MEM

The fitted profiles of Rietveld refinements (Fig. 3.12a–e for Mg, Ti, Fe, Zn, Sn and
Te respectively) give a clear picture of the quality of the sample as well as the data.
The accuracy of refinements can be judged from the difference curve between the
observed and calculated intensity, which is almost a straight line. The results of the
structural refinement (Petřiček et al. 2000) given in Table 3.10 shows reasonable
values of the Debye–Waller factor for Mg (1.46 Å2) (Peng et al. 1996), Fe
(0.350 Å2) (Mohanlal 1979), Zn (1.26 Å2) (Bashir et al. 1988) and Te (0.74 Å2)
(Bashir et al. 1988]) obtainable from a powder data set, which is comparable to
those reported. The reliability indices for all the powder samples are very low
indicating the accuracy of the refinements (RP = 10.03, 8.92, 20.89, 9.08, 15.84,
17.11% for Mg, Ti, Fe, Zn, Sn and Te, respectively).
The MEM electron density distribution map of Mg as given in Fig. 3.14a for
(100) plane reveals the core of the Mg atom to be spherical and shows that the
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 101

Table 3.13 Electron density profiles along the three directions of the unit cells
Direction metal [100] [110] [111]
Position (Å) ED (e/Å3) Position (Å) ED (e/Å3) Position (Å) ED (e/Å3)
Mg 0.6527 0.06 0.6751 0.06 0.5741 0.10
Ti 1.4772 0.19 1.4772 0.19 2.7721 0.19
Fe 1.4346 0.16 1.5850 0.13 1.2424 0.52
Zn 0.4593 0.00 0.5011 0.00 3.6121 0.12
Sn 1.6319 0.33 2.1796 0.02 2.1987 0.02
Te 1.1857 0.13 0.9764 0.05 1.1601 0.17
ED Electron density

valence region of Mg is not spherical but smearing of charges is seen. The MEM
electron density distribution map of Ti as given in Fig. 3.14b for (100) plane
reveals the core of the Ti atom to be spherical but the valence region of Ti shows a
concave nature. The MEM electron density distribution map of Fe as given in
Fig. 3.14c for (100) plane reveals the core of the Fe atom being spherical and this
sphericity persists even at slightly larger distances away from the centre. The
distribution of charges all over the plane indicates the distribution of electrons and
the charges at the edge centres on (100) plane showing the distribution from the
perpendicular face-centred atoms and indicating the extension of the spread of the
charges. The MEM electron density distribution map of Zn as given in Fig. 3.14d
for (100) planes reveals the distortion of the core atom which may be due to the
high Debye–Waller factor as seen in Table 3.11. The distortion extends to the
valence region which is also concave. The MEM electron density distribution map
of Sn as given in Fig. 3.14e for (100) plane reveals the distortion in the core region
which is again substantiated by the high Debye–Waller factor as seen in Table 3.11.
The atomic number of Sn is higher and hence, more diffuse distribution of charges
is seen in the electron density map. Due to high-density of charges, there is
smearing of charges in the valence region. The MEM electron density distribution
map of Te as given in Fig. 3.14d for (100) planes reveals the distorted core region
which may again be due to the high Debye–Waller factor as seen in Table 3.11. The
atomic number of Te is high which can be visualised from the dense distribution of
charges in the electron density map. Since the atomic numbers of Sn and Te are
higher than that of Mg, Ti, Fe and Zn, more diffuse distribution of charges are seen
in the electron density maps of Sn and Te on the (100) plane.
The electron density map of Fe on (110) plane shows (Fig. 3.15c) highly
concentrated charges at the core of the atoms, and the distribution of charges on
places other than the atomic positions, due to the valence, free electrons. The
electron density map of Mg, Ti, Zn, Fe on the (110) plane (Fig. 3.15a, d), show
similar trends as observed in (100) plane. The electron density map of Sn and Te
on the (110) plane (Fig. 3.15e, f), show less dense charge distribution compared to
(100) plane.
The one-dimensional profiles of electron density constructed along [100], [110]
and [111] directions for Mg, Ti, Fe, Zn, Sn and Te are shown in Fig. 3.16a–f,
102 3 Results and Discussion on Metals and Alloys

respectively. The positions of minimum electron densities and density values have
been given in Table 3.14.
The electron densities at the saddle points along these three directions have
been given in Table 3.13.
The mid-bond density in Ti is the largest value (Table 3.13) along [110]
direction among the six metal systems studied. The Debye–Waller factor is less for
Ti when compared with Mg, Fe, Zn, Sn and Te as shown in Table 3.11. Since the
thermal vibration is less the bonding interaction is more significant in Ti. Hence,
obviously, the bonding interaction is stronger in Ti as evidenced by the higher
electron density along [110] direction. From Table 3.13, Zn has minimum mid-
bond density of 0.00 10-3 e/Å3 at a distance of 0.4593 Å, among the metals
studied along [110] and this may be due to the high Debye–Waller factor as
observed in Table 3.11. Sn has less mid bond density 0.02 10-3 e/Å3 at a distance
of 2.1796 Å, along [110] due to the high Debye–Waller factor as observed in
Table 3.11. This substantiates the two-dimensional electron density map of tin
(Fig. 3.15e) where sparse distribution of electron density is observed along [110]
direction. From Tables 3.11 and 3.13, it can be observed that the metals with
higher Debye–Waller factors have low mid-bond density along [110] direction
which confirms the higher thermal vibration.
But along the [100] (nonbonding), Sn has higher electron density of 0.3310-3
e/Å3 at a distance of 1.6319 Å whose atomic number is higher than Mg, Ti, Fe, Zn.
This can also be visualized from Fig. 3.14e which gives the MEM electron density
distribution of Sn on the (100) plane.
Along [111] direction Fe has higher electron density of 0.52 e/Å3 at a distance
of 1.2424 Å which is also visualized in Fig. 3.15c which gives the MEM electron
density distribution of Fe on the (110) plane.
Hence, from the qualitative and quantitative analysis of the MEM electron
densities, the bonding in Mg, Ti, Fe, Zn, Sn and Te is predicted to be predomi-
nantly metallic and the interaction of charges along other non-bonding directions
seems to be heavier due to the metallic nature.

3.3.7.2 PDF

In this study, the observed PDFs of Mg, Ti, Fe, Zn, Sn and Te were obtained from
the raw intensity data using a software programme PDFgetX (Jeong et al. 2001)
after performing corrections such as multiple scattering correction, polarisation
correction, absorption correction, normalisation correction and Compton correc-
tion. A comparison between the observed and calculated PDF is carried out using
the software package PDFFIT (Farrow et al. 2007). Figure 3.17a–f give the
observed and calculated PDF profiles along with the difference between them for
Mg, Ti, Fe, Zn, Sn and Te respectively.
The bond lengths deduced from the PDF analysis have been compared with
those calculated using software programme Grenoble Thermal Ellipsoids Plot
Programme (GRETEP) is a Windows interactive programme that displays the list
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 103

Table 3.14 Neighbour distances from PDF

Metal Neighbour distance (from PDF analysis) (Å)
Nearest neighbour Observed (r1) Calculateda (r2) Difference (r1 * r2)
Mg I 3.12 3.21 0.09
II 5.70 5.57 0.13
III 8.24 8.46 0.22
Ti I 2.80 2.95 0.15
II 5.22 5.12 0.10
III 7.36 7.63 0.27
Fe I 2.80 2.85 0.05
II 4.78 4.76 0.02
III 6.44 6.25 0.19
Zn I 2.62 2.67 0.05
II 4.58 4.63 0.05
III 5.44 5.64 0.20
Sn I 2.38 2.19 0.18
II 4.32 4.66 0.34
III 6.38 6.21 0.17
Te I 1.46 1.98 0.52
II 3.32 3.96 0.64
III 4.46 4.46 0.00
a
GRETEP software

of calculated bonds, bond length between two atoms and the distances and angles
around an atom) (GRETEP). The refined pair distribution function (the Fourier
transform of, the reduced structure factor) given in Fig. 3.17a–f for Mg, Ti, Fe, Zn,
Sn and Te show good matching of observed and calculated PDFs. The difference
between the observed and calculated PDFs are also shown in the figures. The peaks
in the PDF profile correspond to the nearest neighbour distances given in
Table 3.14.
The first neighbour distance as observed from Table 3.15 decreases as the
atomic number increases for all the metals. The difference between observed and
calculated nearest neighbour distances are also given in the table. These differ-
ences are not too high considering the fact that we have used only data with limited
Q values. There may be local undulations in the structure of these six systems that
will lead to slightly higher neighbour differences which correspond to lattice repeat
distances. The local disorder would have arisen due to several factors viz., X-ray
diffraction instrumental error or fine impurity in the sample. This is reflected in the
local structural analysis only like PDF.
104 3 Results and Discussion on Metals and Alloys

(a) (b) 6 Cal

5 Obs
Obs
Cal
4
4 Diff
Diff
3 2
2
-0
1

G(r)(Å -2 )
G(r)(Å-2 )

-0 -2
-1
-4
-2
-3 -6
-4
-8
-5
-6 -10

5 10 15 20 25 30 2 6 10 14 18 22 26 30
r(Å) r(Å)

(c) 20 (d) Cal

Cal 8
Obs Obs
6 Diff
Diff
10 4
2

G(r)(Å-2 )
G(r)(Å -2 )

-0
0 -2
-4
-6
-10 -8
-10
-12
-20
2 6 10 14 18 22 26 30 4.0 10.5 17.0 23.5 30.0
r(Å) r(Å)

(e) 8
Obs (f) 3
Cal
Cal Obs
6 Diff 2 Diff
4 1
2 -0
G(r)(Å -2 )
G(r)(Å-2 )

-0
-1
-2
-2
-4
-3
-6
-8 -4
-10 -5
-12 -6

2 6 10 14 18 22 26 30 3 5 7 9 11 13 15 17 19 21 23 25
r(Å) r(Å)

Fig. 3.17 a The observed and calculated PDFs of Mg with difference profile. b The observed
and calculated PDF of Ti with difference profile. c The observed and calculated PDFs of Fe with
difference profile. d The observed and calculated PDFs of Zn with difference profile. e The
observed and calculated PDFs of Sn with difference profile. f The observed and calculated PDFs
of Te with difference profile

3.3.8 Conclusion

A clear understanding of the structure of Mg, Ti, Fe, Zn, Sn and Te has been gained
from this work, in terms of the cell parameters, thermal vibrations and electron
density distribution in the bonding region using MEM and local structural analysis
using PDF. The electron density, bonding and charge transfer studies analysed in
this work will give fruitful information to researchers in the fields of physics,
3.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium 105

chemistry, materials science, metallurgy, etc. These properties can be properly used
for the proper engineering of these technologically important materials.

3.4 CoAl and NiAl Metal Alloys

3.4.1 Introduction

Metal alloys are highly useful because they have specific properties and charac-
teristics that are more attractive than those of pure, elemental metals. For example,
some alloys possess high strength, some have low melting points, some are
refractory with high melting temperatures, some are especially resistant to cor-
rosion and some have desirable magnetic, thermal or electrical properties. These
characteristics arise from both the internal and electronic structure of the alloys.
Hence, a study on the electronic structure of alloys using the most versatile
techniques like MEM (Collins 1982) and multipole method (Hansen and Coppens
1978) available today is worthwhile. If the tool available for the analysis yields
highly precise information, then it is appropriate to apply it to precise data sets
available as has been done in this work, thereby the methodology can also be
tested. Hence, a study on the electronic structure of the metal alloys CoAl and
NiAl has been carried out using accurate single crystal X-ray data sets reported
earlier by Weiss (1966). Electron density distribution and bonding in different
materials using MEM have been analysed for different materials e.g., GaAs
(Saravanan et al. 2003), InP (Israel et al. 2004), Oxides (Israel et al. 2003), ionic
systems (Israel et al. 2003). Research works have also been reported by Sakata and
Sato (1990), Saka and Kato (1986), Yamamura et al. (1998), Ikeda et al. (1998),
Yamamoto et al. (1996). In order to confirm the results and to elucidate the
distribution of valence electrons, and the contraction/expansion of atomic shells,
multipole analysis of the electron densities was carried out using the software
package JANA 2000 (Petřiček et al. 2000). Recently, the multipole analysis of the
charge densities and bonding has been widely utilised to study the electronic
structure of materials (Coppens and Volkov 2004; Poulsen et al. 2004; Frils et al.
2004; Pillet et al. 2004; Marabello et al. 2004). In this work, the multipole model
proposed by Hansen and Coppens (1978) has been used for elucidating the elec-
tronic structure of CoAl and NiAl.

3.4.2 Summary of the Work

The precise electron density distribution and bonding in metal alloys CoAl and
NiAl is characterised using MEM and multipole method. Reported X-ray single
crystal data have been used for this purpose. Clear evidence of the metal bonding
between the constituent atoms in these two systems has been obtained. The mid-
106 3 Results and Discussion on Metals and Alloys

bond electron densities in these systems are found to be 0.358 and 0.251 e/Å3
respectively, for CoAl and NiAl in the MEM analysis. The two-dimensional maps
and one-dimensional electron density profiles have been constructed and analysed.
The thermal vibration of the individual atoms Co, Ni and Al has also been studied
and reported. The contraction of atoms in CoAl and expansion of Ni and con-
traction of Al atom in NiAl is found from the multipole analysis, in line with the
MEM electron density distribution.

3.4.3 Origin of the Data

Weiss (1966) has reported single crystal X-ray structure factor data of CoAl and
NiAl collected using AgKa and CoKa radiations. Among these highly precise data
sets, structure factor sets collected using AgKa were utilised in this work, because,
more number of Bragg reflections is available using this radiation.
The two sets of data were subjected to preliminary data analysis and the
structure of CoAl and NiAl was refined using the software GSAS (Larson and Von
Dreele 2004) (General Structure Analysis System). Further, the same data were
refined using a code written by Dr. R. Saravanan using an algorithm which uses the
standard least-squares refinement. The individual thermal vibration parameters of
atoms were refined both in GSAS (Larson and Von Dreele 2011) and also using
the code by Dr. R. Saravanan (hereafter referred as RS refinements). Exact
matching of the structure factors and structural parameters were realised in both
the refinements. The refined structure factors were used for MEM analysis. Mul-
tipole refinements were carried out using the refinement software package JANA
2000 (Petřiček et al. 2000) to refine the thermal parameters and the j0 parameter
which gives information about the valence electronic status of the atomic shells.

3.4.4 Data Analysis

CoAl and NiAl belong to the CsCl structure in which there are 2 atoms/cell and
the fractional atomic coordinates are 0, 0, 0, and ‘, ‘, ‘. The cell constants are
2.8619 and 2.8864 Å respectively, for CoAl and NiAl, (Weiss 1966). Figure 3.18
shows the structure of NiAl.
It has been reported that no order–disorder transition occurs in these alloys and
there is a possibility of squeezing of charge densities and directionality in the outer
electron charge density. A total of 11 Bragg reflections have been reported (Weiss
1966) for both these systems. The data were found to be precise and the reported r
(Fobs) values are very low indicating the precision of the data. Using these data,
first, the refinements were carried out with GSAS and the results of these refine-
ments are given in Table 3.15 both for CoAl and NiAl. Then RS refinements were
carried out, the results of which are also given in Table 3.15. The quantity
3.4 CoAl and NiAl Metal Alloys 107

Table 3.15 The Debye–Waller factors in NiAl and CoAl in GSAS and RS Refinements
Refinement type System
NiAl CoAl
GSAS BNi (Å2) 0.221(0.070) BCo (Å2) 0.368(0.049)
BAl (Å)2 0.421(0.144) BAl (Å2) 0.591(0.088)
R (%) 1.5 R (%) 1.1
wR (%) 1.9 wR (%) 1.2
RS BNi (Å2) 0.279(0.073) BCo (Å2) 0.353(0.082)
BAl (Å)2 0.389(0.120) BAl (Å2) 0.528(0.130)
Boverall (Å)2 0.288(0.035) Boverall (Å)2 0.382(0.042)
R (%) 1.413 R (%) 1.134
B Debye–Waller factor

Table 3.16 Parameters of Parameters NiAl CoAl

MEM refinement
Prior electron density (e/Å3) 1.705 1.706
Resolution (Å/Pixel) 0.405 0.405
K 0.24 0.24
C 1.0 1.0
R (%) 0.88 0.92
wR (%) 0.80 0.79
Number of cycles 223 188
k Lagrangean parameter; C convergence criterion

Fig. 3.18 Unit cell of NiAl

108 3 Results and Discussion on Metals and Alloys

(a) 1.0 (b) 1.0

0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6

[001]
[001]

0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 Co 0.1 Ni

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[010] [010]
(c) 1.0 (d) 1.0
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
[001]
[001]

0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 Co 0.1 Ni

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[010] [010]
(e) 1.0 (f) 1.0
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
[001]
[001]

Al Al
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 Co 0.1 Ni

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[010] [010]

Fig. 3.19 a MEM high electron density distribution of CoAl on the (100) plane. Contour Range
is between 0.0 and 15.0 e/Å3. Contour interval is 0.75 e/Å3. b MEM high electron density
distribution of NiAl on the (100) plane. Contour range is between 0.0 and 15.0 e/Å3. Contour
interval is 0.75 e/Å3. c MEM low electron density distribution of CoAl on the (100) plane.
Contour range is between 0.0 and 2.5 e/Å3. Contour interval is 0.125 e/Å3. d MEM low electron
density distribution of NiAl on the (100) plane. Contour range is between 0.0 and 2.5 e/Å3.
Contour interval is 0.125 e/Å3. e MEM electron density distribution of CoAl on the (100) plane
showing the extent of electron clouds of Al. Contour range is between 0.0 and 0.2 e/Å3. Contour
interval is 0.008 e/Å3. f MEM electron density distribution of NiAl on the (100) plane showing
the extent of electron clouds of Al. Contour range is between 0.0 and 0.2 e/Å3. Contour interval is
0.008 e/Å3
3.4 CoAl and NiAl Metal Alloys 109

1.0 1.0
(a) Al (b) Al
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
[001]

[001]
Co Ni
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[110] [110]

1.0 1.0
(c) Al (d) Al
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
[001]

[001]
Co Ni
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
[110] [110]

Fig. 3.20 a MEM high electron density distribution of CoAl on the (110) Plane. Contour range:
0.0–6.6 e/Å3. Contour interval is 0.33 e/Å3. The plane has been shifted half the unit cell along the
z direction. b MEM high electron density distribution of NiAl on the (110) Plane. Contour range:
0.0–6.6 e/Å3. Contour interval is 0.33 e/Å3. The plane has been shifted half the unit cell along the
z direction. c MEM low electron density distribution of CoAl on the (110) Plane. Contour range:
0.0–1.2 e/Å3. Contour interval is 0.06 e/Å3. The plane has been shifted half the unit cell along the
z direction. d MEM low electron density distribution of NiAl on the (110) Plane. Contour range:
0.0–1.2 e/Å3. Contour interval is 0.06 e/Å3. The plane has been shifted half the unit cell along the
z direction

P
minimised in RS refinements was D ¼ hkl Whkl ðjFobs j  jkFcal jÞ2 ; where whkl is
the weight assigned on each observation, Fobs is the observed structure factor and
Fcal is the calculated structure
h P factor. Thei estimated standard deviation in any
2
parameter is given by rp ¼ bii wr¼1 ðDF rÞ
mn ; where bii is the diagonal element of
the inverse matrix, wr is the weight of the rth factor, where DFr is the difference
between Fobs and Fcal, m is the number of observations
P and n is the number of
jjFobs jjFcal jj
parameters. The reliability index given by R ¼ P of these refinements is
jFobs j
also given in Table 3.15. The average thermal vibration of the overall system is
studied and reported in Table 3.15. In RS refinements, the analytical coefficients of
the five-parameter model by Waasmaier and Kirfel (1995) were used for the
calculation of the atomic scattering factors.
MEM studies on the electron density distribution of these two systems were
carried out by dividing both the unit cells into 64 9 64 9 64 pixels. The initial
electron densities were fixed as (F000/a3o), where F000 is the total number of
110 3 Results and Discussion on Metals and Alloys

(a) (b)
2.0
400 High density profiles 1.8 Low density profiles
1.6
Electron Density (e/Å 3)

Electron Density (e/Å3)

100
300 110
111 1.4
100
110
1.2 111

200 1.0

0.8

100 0.6

0.4

0 0.2
Co Co
0.0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 0.5500 0.9875 1.4250 1.8625 2.3000
Distance (Å) Distance (Å)
(c) (d)
500 2.0
High density profiles Low density profiles
1.8
400 111
1.6 111
Electron Density (e/Å3)

110
Electron Density (e/Å 3)

110
100 100
1.4
300
1.2

1.0
200
0.8

0.6
100
0.4

0 Ni 0.2
Ni
0.0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 0.5500 0.9875 1.4250 1.8625 2.3000
Distance (Å) Distance (Å)

Fig. 3.21 a One-dimensional high electron density distribution of CoAl along the [100], [110]
and [111] directions of the unit cell. b One-dimensional low electron density distribution of CoAl
along the [100], [110] and [111] directions of the unit cell. c One-dimensional high electron
density distribution of NiAl along the [100], [110] and [111] directions of the unit cell. d One-
dimensional low electron density distribution of NiAl along the [100], [110] and [111] directions
of the unit cell
3.4 CoAl and NiAl Metal Alloys 111

Table 3.17 Profile electron densities along the three directions of the unit cells of CoAl and
NiAl
Direction system [100] [110] [111]
Position (Å) ED (e/Å3) Position (Å) ED (e/Å3) Position (Å) ED (e/Å3)
CoAl 1.431 0.079 2.024 0.083 1.317 0.358
NiAl 1.443 0.050 2.041 0.078 1.328 0.251
ED Electron density

5
[Fobs(K)-FMEM(K)]/ Fobs (K)

(a)
4
3
2
1
0
-1
-2
-3
-4
-5
0.1 0.2 0.3 0.4 0.5 0.6 0.7
sin (Å-1)

5
[Fobs(K)-FMEM(K)]/ Fobs (K)

(b)
4
3
2
1
0
-1
-2
-3
-4
-5
0.1 0.2 0.3 0.4 0.5 0.6 0.7
-1
sin (Å )

Fig. 3.22 a The distribution of errors in the reciprocal space of CoAl. b The distribution of
errors in the reciprocal space of NiAl

electrons in the respective unit cells and ao is the cell dimension. The resolution of
the MEM maps turns out to be 0.045 Å/pixel along the three directions of the unit
cells. The Lagrange parameter k was chosen so that 188 and 223 cycles of
refinement were required for CoAl and NiAl respectively, to achieve the con-
vergence criterion C = 1; C being,
112 3 Results and Discussion on Metals and Alloys

Table 3.18 Parameters from Multipole refinement

Parameter NiAl CoAl
Ni Al Co Al
Pc 18.0 10.0 18.0 10.0
Pv 10.0 3.0 9.0 3.0
j0 0.984(047) 1.258(087) 0.953(039) 0.920(096)
B (Å2) 0.297(076) 0.337(128) 0.265(040) 0.472(065)
R (%) 1.08 0.91
wR (%) 1.10 0.97
GoF 1.71 1.55
B Debye–Waller factor, Pc population coefficients in the core region, Pv population coefficients
in the valence region; j0 and j00 variable parameters

Fig. 3.23 a Static Multipole Deformation map of CoAl on (100) plane. Dotted lines indicate
negative electron densities. b Static Multipole Deformation map of CoAl on (‘00) plane. Dotted
lines indicate negative electron densities. c Dynamic Multipole Deformation map of CoAl on
(100) plane. Dotted lines indicate negative electron densities. d Dynamic Multipole Deformation
of CoAl on (‘00) plane. Dotted lines indicate negative electron densities
3.4 CoAl and NiAl Metal Alloys 113

Fig. 3.24 a Static Multipole Deformation map of NiAl on (100) plane. Dotted lines indicate
negative electron densities. b Static Multipole Deformation map of NiAl on (‘00) plane. Dotted
lines indicate negative electron densities. c Dynamic Multipole Deformation map of NiAl on
(100) plane. Dotted lines indicate negative electron densities. d Dynamic Multipole Deformation
map of NiAl on (‘00) plane. Dotted lines indicate negative electron densities

1 X

Fcal ðkÞ  Fobs ðkÞ2

C¼

ð3:5Þ
N
rðFobs ðkÞÞ

where N is the number of Bragg reflections, r(K) is the standard deviation in

Fobs(K) and V is the volume of the unit cell. Other relevant parameters of MEM
refinements are given in Table 3.16.
The electron density distribution of CoAl and NiAl has been mapped using the
MEM electron density values obtained through refinements. Figure 3.19a, b show
the MEM high electron density distribution of CoAl and NiAl on the (100) plane
of the unit cell.
Figure 3.19c, d show the MEM low electron density distribution of CoAl and
NiAl on the (100) plane.
114 3 Results and Discussion on Metals and Alloys

Figure 3.19e, f show the extent to which the aluminium atom is distributed as
evidenced by the electron clouds on the (100) plane.
Figure 3.20a, b show the MEM high electron density distribution on the (110)
plane of CoAl and NiAl respectively, wherein the shifted atoms (half the unit cell
along the Z direction) have been shown.
Figure 3.20c, d show the MEM low electron density distribution on the (110)
plane of CoAl and NiAl.
Figure 3.21a, b show the one-dimensional variation of the high and low elec-
tron densities along the three directions ([100], [110] and [111]) of the unit cell of
CoAl and Figs. 3.21c, d represent the one-dimensional variation of high and low
electron densities of NiAl.
The electron densities at the saddle points along these three directions are given
in Table 3.17.
Based on these two-dimensional MEM maps and one-dimensional profiles, a
discussion is formulated and given in the following section.
The errors in the observations have also been studied in this work and repre-
sented in Fig. 3.22a, b, which show the distribution of errors in the reciprocal
space CoAl and NiAl respectively.
The results of the multipole refinements done using JANA 2000 are presented
in Table 3.18.
Hansen and Coppens (1978) proposed a modified electron density model with
the option that allows the refinement of population parameters at various orbital
levels where the atomic density is described as a series expansion in real spherical
harmonic functions through fourth order Ylm. According to this model, the charge
density in a crystal is written as the superposition of harmonically vibrating
aspherical atomic density distribution convolving with the Gaussian thermal dis-
placement distribution as
X
atom    
qðr Þ ¼ qk !r !
r k!
u  tk !
u ; ð3:6Þ
k

where tk ð~
uÞ is the Gaussian term and  indicates a convolution. The atomic charge
density is then defined as
X4 Xl !
    r
r ¼ Pc qcore !
q ! r þ Pv k03 qvalence ðk0 r Þ þ k003 Rj ðk00 :Þ P Y
¼l lm lm r
l¼0
ð3:7Þ
where Pc, Pv and Plm are population coefficients. Canonical Hartree–Fock atomic
orbitals of the free atoms normalised to one electron can be used for the con-
struction of qcore and qvalence but the valence function is allowed to expand and
contract by the adjustment of the variable parameters k0 and k00: The effect of the
temperature can be distinguished from the convolution and the deconvolution of
3.4 CoAl and NiAl Metal Alloys 115

thermal contribution to the charge density as dynamic and static multipole

deformation maps. The deformation density in these maps are characterised by

! 1 X !
Dqmultipole deformation r ¼ F h
v multipole deformation
! h ! i
F h exp 2pi h ; !
r ð3:8Þ
spherical atom

where Fmultipole_deformation is the Fourier transform of the multipole charge density

with or without the convolution of thermal contribution, where the Fourier com-
ponents are terminated at the experiment resolution. The SMD and DMD maps are
presented in Fig. 3.23a, b, c, d for CoAl and in Fig. 3.24a, b, c, d for NiAl. That is,
Fig. 3.23a, b represent the SMD map of CoAl on the (100) and (‘00) planes;
Fig. 3.24a, b represent the SMD map of NiAl on the (100) and (‘00) planes. Other
maps are DMD maps of these two systems. The static deformation maps show the
electron distribution without the contribution from thermal vibration of the atoms.

3.4.5 Results and Discussion

The structural refinement gives precise individual Debye–Waller factors as indi-

cated in Table 3.15. The Al atom has a large thermal vibration in both CoAl and
NiAl compared to the constituent atoms Co and Ni. The BAl is larger in CoAl than
in NiAl, indicating slightly larger binding force in NiAl due to more number of
electrons. The reliability indices in these refinements are as low as 1.1% in the case
of CoAl. The RS refinements give similar results for the thermal vibrations. The
reliability indices are also low compared to GSAS refinements in the case of CoAl,
whereas in NiAl, the R values are comparable in both the refinements. The dif-
ference in BNi between these two refinements is 0.058 Å2, larger in the case of RS
refinement. But BCo is 0.015 Å2 smaller in RS refinement compared to that of
GSAS refinement. The difference in BAl between NiAl and CoAl is 0.17 Å2 in
GSAS refinement and 0.139 Å2 in RS refinement. The fact that the atom with
smaller atomic number vibrates more has been realised in both the refinements and
in both CoAl and NiAl (ZAl \ ZCo \ ZNi).
The MEM refinements were carried out with a prior electron density of 1.706
and 1.705 e/Å3 respectively, for CoAl and NiAl. The Lagrange parameter used
was the same for both the cases. The reliability indices from MEM refinements are
given in Table 3.16, which shows very low R values, being less than 1%.
Figure 3.19a, b shows highly spherical core regions of Co and Ni atoms
respectively, on the (100) plane. The contour range and interval for these two maps
were set as the same in order to compare the electron densities. The same criterion
was followed for appropriate low and high-density maps of CoAl and NiAl. In order
to visualise the low electron density/bonding regions, the maps were re-drawn with a
contour range of 0.0–2.5 e/Å3, as shown in Fig. 3.19c, d. Figure 3.19c shows that
116 3 Results and Discussion on Metals and Alloys

the valence region of Co is clearly not spherical and the four corner atoms on (100)
plane show concave outer charges. It appears that the eight Co atoms give way for
the thermal vibration of the Al atom located at the body centre. It is noted from
Table 3.15 that BAl in CoAl is higher than BCo. In the case of NiAl (Fig. 3.19d), no
such concave outer charges are seen, since the thermal vibration of both Ni and Al
atoms are lower compared to Co and Al in CoAl. Figure 3.19e, f show very low
electron densities on (100) plane of CoAl and NiAl. The body-centred Al charges
touching the (100) plane of the unit cell is visible in Fig. 3.19e. The valence charges
of both the atoms (Co and Al) do not repel each other. In fact, charge clouds of both
the atoms touch each other, but with no resemblance of an attraction. This should be
due to the characteristics of metal bonding. Moreover, in multipole analysis, the
expansion/contraction parameter j0 is estimated to be 0.953 and 0.92 for Co and Al
atoms, which means expansion of both the atoms. The expansion is more in Al atom.
Hence, the above said trends in the low-density regions are visible as seen from
Fig. 3.19e. In NiAl (Fig. 3.19f), the charge accumulation is less dense due to
reduced thermal vibration of both the atoms in NiAl. Again from multipole
refinement, the j0 parameter is found to be 0.984 and 1.258 for Ni and Al atoms
respectively. This clearly shows that Ni atom is expanded slightly (not as in Co), but
the Al atom is contracted to a large extent (j0 = 1.258). These facts are clearly
visualized in the MEM maps. Thus, the results from one analysis (MEM) could be
correlated to the other (multipole). Figure 3.20a, b show the electron density dis-
tribution of CoAl and NiAl in the high-density regions of the (110) plane. The core
regions of both the atoms are seen to be spherical both in CoAl and NiAl. Again, no
repulsion of charges is seen in these systems. The outer regions of both Co and Al
atoms are more smeared in CoAl compared to that of Ni and Al atoms in NiAl, a fact
due to the increased thermal vibrations in CoAl. Figure 3.20c, d show the low
electron density distribution of CoAl and NiAl on the (110) plane. Charges are seen
around the periphery of Co and Al atoms in CoAl (Fig. 3.20c) and Ni and Al atoms in
NiAl (Fig. 3.20d). The bonding charges show a metallic nature in both the systems.
Figure 3.21a–d show the one-dimensional electron density distribution along
the [100], [110] and [111] directions of the CoAl and NiAl unit cells respectively.
Both high- and low-density variation of charges is shown in these figures. The
high-density peaks in Fig. 3.21a, c show the peak densities along these three
directions which are larger in NiAl. The low-density maps show that the density
minima do not occur at the same points in both the cases. Moreover, the distri-
bution of charges along the three directions does not overlap. These facts suggest
that the outer regions are not perfectly spherical. The profile along [100] direction
is perfectly symmetrical due to the distribution of charges from like atoms. The
distribution along [111] direction shows the bonding nature between unlike atoms.
The previous discussion clears that there is attraction/repulsion like interaction
between Co and Al and Ni and Al. Yet, the mid-bond charges are considerably
higher as seen from Table 3.17. CoAl has a mid-bond density value of 0.358 e/Å3
3.4 CoAl and NiAl Metal Alloys 117

and NiAl has a value of about 0.251 e/Å3. If the systems are ionic one cannot
expect higher mid-bond values as in the present cases. If the systems are covalent,
one can expect values higher than the present values due to overlapping of charges.
Hence, from the qualitative and quantitative analysis of the MEM electron den-
sities, the bonding in CoAl and NiAl is predicted to be predominantly metallic.
The strength of the electron density is higher in the case of CoAl, which has lower
number of electrons in the unit cell compared to NiAl. This is due to the relatively
larger thermal vibration of aluminium and smearing of charges in CoAl. The
increase in the mid-bond density is about 0.107 e/Å3 consistent with the increased
thermal vibration of atoms, particularly aluminium, in CoAl. Hence, it is important
to study precisely the thermal vibration of atoms in electron density analyses,
particularly in comparing the bonding densities in systems with similar constituent
atoms.
Figure 3.22a, b shows the distribution of errors in the observations in the
reciprocal space of CoAl and NiAl respectively. The error in the form of [Fobs
(K)-FMEM(K)]/r(Fobs(K)), is distributed uniformly around ±2 indicating that
there is no bias in the experimental data or the calculations and justifies the use of
these experimental data for the present analysis.
The B values from the multipole refinement are in close agreement with those
obtained from other refinements. Although all the B values are comparable, those
from the multipole analysis give accurate interpretation of the electron density
distribution in both the systems.
The SMD and DMD maps obtained from multipole refinements show that
(Figs. 3.23, 3.24) more pronounced deformation in density is at the valence
regions than at the core regions for both the systems. Figure 3.23a (SMD—
CoAl—100 plane) shows features similar to those seen in Fig. 3.19e for CoAl. The
expansion of Al atom and the concave nature of charge clouds of Co atom are
visible in this map. Figure 3.24a (NiAl—SMD 100 plane) does not show any
concave nature/repulsive interaction, because of the heavy contraction of Al atom.
The map in Fig. 3.23b (SMD—CoAl—‘ 0 0 plane) shows the expansion of both
the atoms and Fig. 3.24b (SMD—NiAl—‘00 plane) shows nothing but Al atom
because of its contracted nature. Further, the expansion of Ni atom (j = 0.984) is
not large compared to Co (0.953) atom. Hence, only the distribution of Al atom is
seen in this map. The DMD maps show trends in line with the expansion/con-
traction of atoms in the two systems. Comparing Fig. 3.23b (SMD of CoAl) and
Fig. 3.23d (DMD of CoAl), it is seen that the SMD map shows a number of
positive contours at the edge centres and at the centre of (100) plane. These are the
residual densities due to the thermal vibration of aluminium atom. The core of
aluminium atom in the Fig. 3.23b is full of negative charges. The DMD map
shows (Fig. 3.23d) the uncompensated thermal vibration of the valence regions
that show up as the positive contours starting from the core. This suggests slight
expansion of the aluminium atom.
118 3 Results and Discussion on Metals and Alloys

Figure 3.24b (SMD of NiAl) and 3.24d (DMD of NiAl) can be compared in a
similar manner. Figure 3.24b shows positive electron densities not only at the core
but also at far-valence regions, even after the deconvolution process, suggesting
shrinkage of the aluminium atom in NiAl. The DMD (Fig. 3.24d) shows a well
compensated core region (without negative charge densities) but uncompensated
positive islands of valence region confirming the contraction of aluminium atom in
NiAl.

3.4.6 Conclusion

A precise analysis on the thermal vibrations and electron density distribution in

CoAl and NiAl has been carried out using reported X-ray data. The metal
bonding in these two systems is substantiated using qualitative and quantitative
analysis of MEM and multipole electron densities. CoAl is found to have more
mid-bond density than NiAl due to the increased thermal vibration and hence
charge delocalization. The cores of the atoms are found to be perfectly spherical
and slight deviation is observed in the valence electron distribution. More studies
are needed on more alloy systems in terms not only of electron densities but also
in terms of electron densities coupled with thermal vibrations of the atoms
involved.

3.5 Nickel Chromium (Ni80Cr20)

3.5.1 Introduction

Alloying one metal with other metals or non-metals often enhances its properties.
The physical properties of an alloy may not differ greatly from those of its ele-
ments, but engineering properties may be substantially different from those of the
constituent materials.
Whether in the form of element or alloy with other metals, nickel materials
have made significant contributions to present-day society. The favourable phys-
ical and mechanical properties as well as corrosion resistance characteristics of
nickel and its alloys provide aids to the solution of many industrial problems
(Teeple 1954). Alloying with nickel increases the impact strength and ductility of
sintered materials without adversely affecting their strength (Enríquez and Mathew
2003).
Nickel alloys have good corrosion resistance and heat resistance, the standard
alloy being used for electrical-resistance for heaters and electrical appliances is
Nickel–Chromium (Murray 1997). Chromium–nickel steel alloys have good
mechanical properties (Radomysel’’skii and Kholodnyi 1975). Conductive
3.5 Nickel Chromium (Ni80Cr20) 119

nichrome probe tips have been developed as functionalized nanotools for nano-
scale electrical testing and nano-welding, and are advantageous in competition
with other scanning probe tips because of their unique combination of high
resistance to oxidation, high hardness and relatively high resistivity (Peng et al.
2009). Ni80Cr20 is a commonly used material for application in thin-film strain
gauges (Kazi et al. 2003, 2006]. NiCr coatings on stainless steel substrates are used
as a high velocity oxy-fuel technique for corrosion applications (Ak et al. 2003).
NiCr cermet coatings are commonly utilised in the industry as a means to provide
wear resistance in high temperature (Wirojanupatump et al. 2001). NiCr-coated
super alloys show better oxidation resistance due to the formation of a compact
and adhesive thin Cr2O3 scale on the surface of the coating during oxidation
(Kamal et al. 2009).
For practical use in different applications, alloys need to be shaped at elevated
temperatures according to the requirements. It is therefore, essential to study the
changes in the microstructure after heat treatment of materials. One of the popular
heat treatment processes is annealing (Al-Quran and Al-Itawi 2010). Annealing
induces ductility, softens material, relieves internal stresses, refines the structure
by making it homogeneous and improves cold working properties. The resistance
characteristics of NiCr improve in the double annealing process (Jeng et al. 1990).
Annealing processes appear to be responsible for an increase in passivity of Ni/Cr
alloys toward oxidation (Kim et al. 2005).
Ball milling may be referred to as the breaking down of relatively coarse
materials to the ultimate fineness. At present, it has evolved as an important
method for the preparation of either materials with enhanced physical and
mechanical properties or indeed, new phases or new engineering materials. The
morphology of the powders is modified when they are subjected to collisions.
The initial ball-powder-ball collision causes ductile metal powders to flatten
into flake-like structures (Eskandarany 2001). The high-energy ball milling
technique is an effective method for producing powders for metal matrix
composites (Zwick and Lugscheider 2004). There exist methods to control the
microstructure of pellet and among others, mechanical milling of powder is
simple and easy to control. With this method, the characteristics of powder and
the resultant compaction behaviour and sinterability are properly controlled
(Na et al. 2002).

3.5.2 Summary of the Work

The alloy Ni80Cr20 was annealed and ball-milled to study the effect of thermal and
mechanical treatments on the local structure and on the electron density distri-
bution. The electron density between the atoms was studied by MEM and the local
structure using PDF. The electron density is found to be high for ball-milled
sample along the bonding direction. The particle sizes of the differently treated
120 3 Results and Discussion on Metals and Alloys

samples were realised by SEM and through XRD. Clear evidence of the effect of
ball milling is observed on the local structure and electron densities.
In this work, the purchased alloy Ni80Cr20 was subjected to different types of
mechanical and thermal treatments viz., ball milling and annealing. The three
different processes are as follows:
i. annealing the sample (hereafter sample A)
ii. ball milling (hereafter sample BM)
iii. ball milling and annealing (hereafter sample BMA).
The charge density and local structure variations of these three samples have
been analysed in this work. This analysis would reveal the properties of Ni80Cr20
suitable for its wide applications. The electronic structure analysis has been carried
out by the most versatile technique, MEM (Gull and Daniel 1978) and local and
average structure properties have been analysed using PDF (Petkov et al. 2002;
Xiangyun et al. 2005).

3.5.3 Experimental

3.5.3.1 XRD and SEM

Powder sample of Ni80Cr20 (99.9% purity) was purchased from Alfa Aesar,
Johnson’s Mathew Company. To anneal the sample, 20 mg of the sample was
placed in a small silica crucible and was annealed at 1023 K for 3 h in a pro-
grammable furnace (sample A). To prepare the ball-milled sample, the alloy was
milled at room temperature in a low energy laboratory ball mill for 40 h with a
milling speed of 170 rpm, using stainless steel balls (sample BM). Further, a part
of the ball-milled powder was annealed at 1023 K for 3 h (sample BMA). The low
energy laboratory ball mill used to mill the sample and the furnace in which the
samples were annealed are shown in Figs. 3.25a, b, respectively.
To analyse the bonding charge densities (structural behaviour) of Ni80Cr20
under various experimental conditions, powder X-ray diffraction data were col-
lected individually for sample A, sample BM, sample BMA. X-ray powder
diffraction measurements were performed using X-PERT PRO (Philips, Nether-
lands) at the National Institute of Interdisciplinary Science and Technology (NI-
IST), Trivandrum, India using CuKa1 radiation with a 2h range from 10 to 120°
and 0.05° step size.
Particle size is an important parameter that influences the physical properties
of materials. A particle may be made of several different crystallites. A crys-
tallite size is the minimum part of the material that diffracts coherently. Several
techniques could be used for the investigation of the particle size. X-ray
diffraction (XRD) and Scanning electron microscopy (SEM) techniques are
used in this work to obtain the particle size of the samples. From XRD data the
particle size was obtained as the particle size is related to the diffraction peak
3.5 Nickel Chromium (Ni80Cr20) 121

Fig. 3.25 a Laboratory ball mill. b Programmable high temperature furnace

Fig. 3.26 a SEM picture of sample A. b SEM picture of sample BM. c SEM picture of sample
BMA

Table 3.19 Particle size of the three samples

Sample aParticle size (Nx) Particle size (Ns) from Average number of coherent domains
from XRD (nm) (SEM) (nm) in each particle (Ns/Nx)
A 25.65 78.75 3.07
BM 11.45 38.76 3.39
BMA 19.80 59.35 3.00
mean = 3.15
a
coherently diffracting domains

broadening. The software written by Dr.R.Saravanan (GRAIN software) was

used to estimate approximate particle sizes. The particle size is analysed using
full width at half-maximum of the powder XRD peaks. The particle mor-
phology was examined by SEM. The results of SEM analysis are given in
Fig. 3.26a–c.
SEM pictures from Figs. 3.26a–c, show that the particle sizes are in nano-
scale and sample A is elongated with particle size of 212.12 nm (Fig. 3.26a),
sample BM is flattened with particle size of 38.76 nm (Fig. 3.26b) and sample
122 3 Results and Discussion on Metals and Alloys

125.00
250 (a) (111)
A (b) (002) A
BM BM
BMA BMA

Intensity (counts)
Intensity (counts)

200 93.75

150
62.50
100

50 31.25

0
0.00
43.0 43.5 44.0 44.5 45.0 45.5 50.7 50.9 51.1 51.3 51.5 51.7 51.9 52.1 52.3

2 (degrees) 2 (degrees)
75.0000 65.0000

64.2857
(c) (202) A
54.1667
(d) (113)
A
BM
Intensity (counts)

BM

Intensity (counts)
BMA
BMA
53.5714
43.3333
42.8571
32.5000
32.1429
(222)
21.6667
21.4286

10.7143 10.8333

0.0000 0.0000

74 75 76 77 78 89 90 91 92 93 94 95 96 97 98 99 100

2 (degrees) 2 (degrees)

Fig. 3.27 a Comparison of (111) XRD peaks of the samples A, BM, BMA respectively.
b Comparison of (002) XRD peaks of the samples A, BM, BMA respectively. c Comparison of
(202) XRD peaks of the samples A, BM, BMA respectively. d Comparison of (113, 222) XRD
peaks of the samples A, BM, BMA respectively

Table 3.20 Full width at half-maximum of different miller planes

Sample (111) (002) (202) (113) (222)
A 0.265 0.310 0.395 0.533 0.587
BM 0.440 0.707 0.892 1.477 1.912
BMA 0.347 0.396 0.527 0.642 0.801

BMA is with uniform morphology with particle size of 59.35 nm (Fig. 3.26c).
The obtained particle sizes from XRD and SEM are given in Table 3.19.
The sizes of the samples measured by XRD line broadening are about one-third
of the particle sizes calculated from SEM. The individual particles of the samples
may have been composed of coherently diffracting domains rather than each being
a single coherently diffracting crystal.
From Fig. 3.27a–d, it is observed that the intensity of samples decrease in the
order of A, BMA, BM respectively.
The broadening of the peaks is also observed in the case of sample BM. The
full width at half-maximum (FWHM) as observed in Table 3.20 indicates that
peaks are broader for sample BM compared to samples A and BMA. The
reduction in particle size due to ball milling as observed from SEM pictures is
substantiated.
3.5 Nickel Chromium (Ni80Cr20) 123

3.5.4 Results and Discussion

3.5.4.1 Powder XRD Analysis

The structural parameters of the samples (A, BM, BMA) of Ni80Cr20 which have
FCC structure with four atoms/unit cell in the space group Fm-3 m were refined
with the well-known Rietveld (1969) powder profile fitting methodology using the
software package JANA 2006 (Petřiček et al. 2006). The fitted profiles and the
positions of Bragg peaks for the three samples have been shown in Fig. 3.28a–c.
In these figures, the dots represent the observed powder patterns and the con-
tinuous lines represent the calculated powder profiles of the respective samples.
The small vertical lines indicate Bragg positions. The difference between the
observed profile and fitted calculated profile is shown at the bottom of each figure.
The refined profiles show good matching between the observed and calculated
profiles for all the three systems. The refined structure factors of the samples given
in Table 3.21a–c shows very good agreement of the observed and calculated
structure factors with very low statistical errors. The composition of the sample
was also refined and was determined as Ni80.1Cr19.9 for the annealed sample
(sample A).
The refined structural parameters from the software JANA 2006 are given in
Table 3.22. The refinement indices are low which show the accuracy of
the refinement (Rp = 8.17, 8.35, 10.16% respectively for samples A, BM, BMA).
The cell parameters and thermal vibration parameters are observed to reduce in the
order of sample BM, A, BMA respectively. Ball milling and subsequent annealing
appears to have reduced the cell constant.

3.5.4.2 Charge Density of Ni80Cr20 Using MEM

The refined structure factors of the three samples were used for the MEM analysis.
The software package PRIMA (Ruben and Fujio 2004) was used for the numerical
MEM computations. The two- and three-dimensional representation of the electron
densities were picturised using the software VESTA (Momma and Izumi 2006).
The MEM refinements were carried out by dividing the unit cell into
64 9 64 9 64 pixels.
The Lagrange parameter is suitably chosen so that the convergence criterion
C = 1 is reached after minimum number of iterations. The MEM parameters are
given in Table 3.23.
The electron density distributions of samples A, BM, BMA are mapped using
the electron density values obtained through refinements. Figure 3.29a–f shows the
three-dimensional electron densities of samples A, BM, BMA in the unit cell in
two different orientations.
Figure 3.30a–c shows the two-dimensional electron densities on (100) plane in
the unit cell for the three samples respectively.
124 3 Results and Discussion on Metals and Alloys

Fig. 3.28 a The fitted XRD profiles and the positions of Bragg peaks for sample A. b The fitted
XRD profiles and the positions of Bragg peaks for sample BM. c The fitted XRD profiles and the
positions of Bragg peaks for sample BMA
3.5 Nickel Chromium (Ni80Cr20) 125

Table 3.21 Observed and calculated structure factors for a Sample A, b Sample BM, c Sample
BMA
hkl Fo Fc Fo - Fc r (Fo)
a Sample A
111 65.62 65.67 -0.05 0.94
002 59.48 59.67 -0.19 0.93
202 43.77 42.81 0.96 0.78
113 34.36 34.60 -0.24 0.79
22 2 31.55 32.40 -0.85 1.31
b Sample BM
111 65.10 65.15 -0.06 0.93
0 02 59.01 59.09 -0.08 1.11
202 41.87 42.07 -0.20 1.03
113 33.74 33.80 -0.06 0.86
22 2 30.75 31.59 -0.84 1.59
c Sample BMA
111 66.85 67.11 -0.26 0.96
00 2 61.16 61.43 -0.27 1.21
202 45.17 45.37 -0.20 1.11
113 37.33 37.49 -0.16 0.92
222 34.99 35.36 -0.37 1.68

Table 3.22 The refined structural parameters using JANA 2006

Sample Cell parameter (Å) B (Å2) Robs (%) wRobs (%) Rp (%) wRp (%)
A 3.5462 (0000) 0.991(017) 0.98 0.12 8.35 11.48
BM 3.5534 (0012) 1.088(256) 0.54 0.54 8.17 11.00
BMA 3.5432 (0013) 0.681(000) 0.51 0.46 10.16 13.72
B Debye–Waller factor

Table 3.23 Refined parameters from MEM technique

Sample Parameters
Number Prior density s(ri) Lagrange parameter (k) RMEM (%) wRMEM (%)
of cycles (e/Å3)
A 52 2.44 0.021 1.26 1.55
BM 32 2.43 0.026 0.98 1.09
BMA 8 2.45 0.135 1.86 1.90

Figure 3.31a–c shows the two-dimensional electron densities on (110) plane in

the unit cell. The one-dimensional electron density profiles along [100], [110] and
[111] directions are represented in Fig. 3.32a–c. Table 3.24 gives the numerical
values of the electron densities along the different directions.
The MEM electron density distribution map of sample BM given in Fig 3.30b
reveals the core atom to be spherical and shows the enhancement of the electron
126 3 Results and Discussion on Metals and Alloys

Fig. 3.29 a, b Three-dimensional MEM electron density isosurface of sample A in two different
orientations. c, d Three-dimensional MEM electron density isosurface of sample BM in two
different orientations. e, f Three-dimensional MEM electron density isosurface of sample BMA in
two different orientations

density in sample BM. This enhancement reflects the high thermal vibration in this
sample. The two-dimensional MEM electron density distributions on (100) plane
shown in Fig. 3.30a–c exhibit the large thermal vibration in sample BM. This is
observed from the smearing of charges at the corner atoms and valence electrons
in the face-centred atoms. The thermal vibration is comparatively less in sample A.
Expansion of the charge region is observed in sample BMA. Expansion of charge
clouds of the face-centred atom and their extension towards the corner atoms is
observed. This reveals the restriction in the thermal vibration resulting in the low
Debye–Waller factor in this case. The increase in the atomic sizes can also be
visualized. Ball milling appears to have increased the average cell constant, which
is then reduced by subsequent annealing of the ball-milled sample.
The one-dimensional electron density profiles shown in Fig. 3.31a along [100]
direction illustrates that the electron density is higher for sample BM initially but
at the mid-bond region the density for sample A is increased. This can be attributed
to the removal of dislocations and vacancies due to annealing. Along the [110]
direction (Fig. 3.32b), which is the bonding direction, the electron density of
sample BM is observed to be higher. Ball milling has effectively increased the
bond strength in Ni80Cr20 along the bonding direction than in the other directions.
The flattening of the particles as seen from SEM picture (Fig. 3.26b) is in support
of this conclusion. Again, the electron density along [111] direction (Fig. 3.32c),
3.5 Nickel Chromium (Ni80Cr20) 127

Fig. 3.30 a Two-dimensional electron density distributions on the (100) plane for sample A.
Contour lines are drawn between 0.0 and 1.5 e/Å3 with 0.1 e/Å3 interval. b Two-dimensional
electron density distribution on the (100) plane for sample BM. Contour lines are drawn between
0.0 and 1.5 e/Å3 with 0.1 e/Å3 interval. c Two-dimensional electron density distribution on the
(100) plane for sample BM. Contour lines are drawn between 0.0 and 1.5 e/Å3 with 0.1 e/Å3
interval

show that the density is higher for sample A. In general, sample BMA has lower
electron density. Ball milling and subsequent annealing has reduced the charge
density in the alloy Ni80Cr20. The strain produced in the sample BM (as evidenced
from the flat region in electron density in Fig. 3.32b) is relieved along the other
directions, particularly the [111] direction as seen from Fig. 3.32c. This figure
shows undulations in the ball-milled sample which is much higher than the other
two samples.
128 3 Results and Discussion on Metals and Alloys

Fig. 3.31 a Two-dimensional electron density distribution on (110) plane for sample A. Contour
lines are drawn between 0.05 and 1.0 e/Å3 with 0.05 e/Å3 interval. b Two-dimensional electron
density distribution on (110) plane for sample BM. Contour lines are drawn between 0.05 and
1.0 e/Å3 with 0.05 e/Å3 interval. c Two-dimensional electron density distribution on (110) plane
for sample BMA. Contour lines are drawn between 0.05 and 1.0 e/Å3 with 0.05 e/Å3 interval

3.5.4.3 Pair Distribution Function

In this study, the observed PDF was obtained from raw X-ray powder data using
software programme PDFgetX (Jeong et al. 2001). A comparison between
observed and calculated PDFs has been carried out using software package
PDFFIT (Farrow et al. 2007). The results of the PDF studies are presented in
Table 3.25.
The refined pair distribution function given in Fig. 3.33a–c for samples A, BM,
BMA respectively, show good matching of observed and calculated PDFs. The
PDF refinement can be considered as equivalent to matching the observed X-ray
powder data with too little structure parameters. Moreover, high Q data are required
for these types of analyses on short-range order and local structure of materials.
Powder XRD data have been used in this work only for comparison purposes.
The bond lengths deduced from the PDF analysis have been compared with those
calculated using software programme GRETEP is a Windows interactive pro-
gramme that displays the list of calculated bonds, bond length between two atoms
and the distances and angles around an atom) (Laugier and Bochu). On comparing
the observed bond lengths obtained from the PDFs of the three samples, it is found
that maximum deviation in the observed bond lengths is noted in sample BM.
3.5 Nickel Chromium (Ni80Cr20) 129

(a) 1.6 (b) 1.500

1.4 BMA A
BM BM

Electron density (e/Å3)

BMA
Electron density(e/Å3)

1.2 A 1.175

1.0

0.8 0.850

0.6

0.4 0.525

0.2

0.0 0.200
0.5 1.0 1.5 2.0 2.5 3.0 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance(Å)along [100] direction Distance(Å) along [110] direction

(c) 1.500
Electron density (e/Å3)

A
1.175 BM
BMA

0.850

0.525

0.200
0.5 1.5 2.5 3.5 4.5 5.5
Distance along [111] direction(Å)

Fig. 3.32 a One-dimensional MEM electron density profiles of samples A, BM,BMA along
[100] direction. b One-dimensional MEM electron density profiles of samples A, BM, BMA
along [110] direction. c One-dimensional MEM electron density profiles of samples A, BM,
BMA along [111] direction

Table 3.24 Electron density profiles along the three directions of the unit cell
Sample [100] [110] (Bonding direction) [111]
Position (Å) ED (e/Å3) Position (Å) ED (e/Å3) Position (Å) ED (e/Å3)
A 1.7731 0.484 1.2537 0.378 3.0711 0.484
BM 1.7767 0.460 1.2561 0.605 3.0774 0.460
BMA 1.7691 0.337 1.2509 0.479 3.0641 0.337
ED Electron density

Table 3.25 Neighbour distances from PDF Analysis

Neighbour distance (from PDF analysis) (Å)
Sample Observed Calculateda
A 2.480 2.508
BM 2.440 2.513
BMA 2.460 2.502
a
GRETEP software
130 3 Results and Discussion on Metals and Alloys

(a) (b)
3 A cal 3 BM cal
obs obs
2 2
1 1
-0 -0
G(r)(Å)-2

G(r)(Å)-2
-1
-1
-2
-2
-3
-3
-4 Difference
-4
-5 Difference
-5
-6
-6
4.0 10.5 17.0 23.5 30.0 4.0 10.5 17.0 23.5 30.0
r(Å) r(Å)

(c)
2 BMA cal
obs
1

-0

-1
G(r) (Å-2)

-2

-3

-4 Difference
-5

-6

4.0 10.5 17.0 23.5 30.0

r(Å)

Fig. 3.33 a The observed and calculated pair distribution function of sample A. b The observed
and calculated pair distribution function of sample BM. c The observed and calculated pair
distribution function of sample BMA

3.5.5 Conclusion

The effects of ball milling and annealing on the local structure have been studied in
this work. The ball-milled sample shows enhanced electron density along bonding
direction, which decreases on annealing. The cell parameters and the bond lengths
appear to be larger when the sample is ball-milled, but after annealing these
parameter values are reduced due to relieving of strain in the sample.

3.6 Silver Doped in NaCl (Na12xAgxCl)

3.6.1 Introduction

The distribution of electrons in the crystalline system is an interesting and

important subject of research in the field of physics, chemistry, material science
and Engineering. Particularly in doped systems this study attains great importance
to understand the changes in the microscopic and macroscopic properties due to
3.6 Silver Doped in NaCl (Na1-xAgxCl) 131

doping. Many studies have been conducted on alkali halides, for example,
Linkoaho (1969), Banerjee and Mitra (1969), Shahi and Wagner (1981), Valvoda
(1970), Burley (1967). The effect of doping on the physical properties of alkali
halides is a subject of sustained interest. While the changes in electrical and optical
properties accompanying doping of alkali halides crystals have received consid-
erable attention, lattice strains and lattice deformations arising out of the intro-
duction of foreign atoms have not been studied extensively. Metals containing salts
are among the inorganic substances that act as antimicrobial agents. Also, the
metal salts are used in gradient index of optical fibre communication, development
of passive fault current limiter in parallel biassing, mass spectroscopic analysis of
metals etc. The present analysis is carried out on the accurate electron density
distribution using MEM and multipole analysis. This research work is on single
crystal data sets of Na0.97Ag0.03Cl and Na0.90Ag0.10Cl (Mohanlal and Pathinettam
Padiyan 1989). When the system is dilute doped, it is possible to study only the
average electronic structure. But, when suitable tools like MEM and multipole
techniques are available, they can be used effectively in conjunction to elucidate
the fine features of bonding as done in this work. In order to elucidate the valence
electrons distribution, MEM and to analyse the contraction/expansion of atomic
shells, multipole analysis of the electron density was carried out.

3.6.2 Summary of the Work

The alkali halide Na1-xAgxCl, with two different compositions (x = 0.03 and
0.10), was studied with regard to the Ag impurities in terms of the bonding and
electron density distribution. X-ray single crystal data sets have been used for the
purpose. The present analysis focused on the electron density distribution and
hence the interaction between the atoms is clearly revealed by MEM and multipole
analyses. The bonding in these systems has been studied using two-dimensional
MEM electron density maps on the (100) and (110) planes and one-dimensional
electron density profiles along the [100], [110] and [111] directions. The mid-bond
electron densities between atoms in these systems are found to be 0.175 and
0.183 e/Å3, respectively, for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl. Multipole anal-
ysis of the structure has been performed for these two systems, with respect to the
expansion/contraction of the ion involved.

3.6.3 Data Analysis

The two data sets (Na0.97Ag0.03Cl and Na0.90Ag0.10Cl) reported by Mohanlal and
Pathinettam Padiyan (1989) have been analysed using Rietveld (1969) structural
refinement using the software package Jana 2000 (Petřiček et al. 2000). The
refined multipole structure factors were used for the MEM computations of the
132 3 Results and Discussion on Metals and Alloys

Table 3.26 Parameters from Rietveld analysis using JANA 2000

Parameter System
Na0.97Ag0.03Cl Na0.90Ag0.10Cl
BNa (Å2) 2.5208(2296) 2.5284(1483)
BCl (Å2) 1.6912(0959) 1.6251(0774)
R (%) 2.20 2.20
wR (%) 2.42 2.31
B Debye–Waller factor

electron densities. Table 3.26 represents the results of the Rietveld refinements for
Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively.

3.6.4 MEM Refinement

MEM is one of the appropriate methods in which the concept of entropy is

introduced to handle the uncertainty in the electron density distribution properly.
The principle of electron density distribution is to obtain results consistent with the
observed structure factors and to reduce the uncertainties as much as possible. In
this work, MEM is adopted for the refinements of electron densities Na0.97Ag0.03Cl
and Na0.90Ag0.10Cl, in which maximisation of the initially assumed electron
density distribution through refinements and hence the building up of the real
electron density distribution has been done. For numerical computations, the
software package PRIMA (Izumi and Dilanian 2002) was used. For three-, two-
and one-dimensional representation of the electron densities, the programme
VESTA (Momma and Izumi 2006) software package was used.
MEM refinements were carried out by dividing the unit cell into 64 9 64 9 64
pixels. The initial electron density at each pixel is fixed uniformly as F000/
a30 = 0.6455 and 0.7045 e/Å3 for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively,
where F000 is the total number of electrons in the unit cell and a0 is the cell
parameter. The Lagrange parameter is suitably chosen so that the convergence
criterion C = 1 is reached after the minimum number of iterations. The MEM
parameters are given in Table 3.27.
The three-dimensional electron density distribution in the form of isosurface in
the unit cell for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl is represented in Fig. 3.34a, b
for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively.
The two-dimensional electron density distribution on the (100) and (110) planes is
given in Figs. 3.35a, b and 3.36a, b for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively.
The one-dimensional electron density profiles along [100], [110] and [111]
directions are represented in Fig. 3.37a, b for these two systems.
Figure 3.38a gives the peak electron density profiles of Na(Ag) along the [100]
direction for the two compositions, (Na0.97Ag0.03Cl and Na0.90Ag0.10Cl) and Fig.
3.38b gives the mid-bond electron density of Na (Ag) along the [100] direction for
3.6 Silver Doped in NaCl (Na1-xAgxCl) 133

Table 3.27 Parameters from MEM analysis

Parameter System
Na0.97Ag0.03Cl Na0.90Ag0.10Cl
Number of cycles 63 56
Prior density, s(r), (e/Å3) 0.646 0.705
Lagrange parameter (k) 0.279 0.053
RMEM (%) 2.394 2.405
wR MEM (%) 3.001 2.736
Resolution (Å/pixel) 0.0881 0.0880

Fig. 3.34 a Three-dimensional electron density of Na0.97Ag0.03Cl. b Three-dimensional electron

density of Na0.90Ag0.10Cl

Fig. 3.35 a MEM electron density distribution of Na0.97Ag0.03Cl on the (100) plane. b MEM
electron density distribution of Na0.90Ag0.10Cl on the (100) plane
134 3 Results and Discussion on Metals and Alloys

Fig. 3.36 a MEM electron density distribution of Na0.97Ag0.03Cl on the (110) plane. b MEM
electron density distribution of Na0.90Ag0.10Cl on the (110) plane

0.6 0.6
(a) (b)
110 111
0.5 0.5 110
100
100
Electron density (e/Å3)
Electron density (e/Å3)

111
0.4 0.4

0.3 0.3

0.2 0.2

Na Cl
0.1 Na Cl 0.1

1.2 1.6 2.0 2.4 2.8 3.2 1.3 1.8 2.3 2.8 3.3
Distance along direction (Å) Distance along direction (Å)

Fig. 3.37 a One-dimensional variation of electron density along [100], [110] and [111]
directions of the Na0.97Ag0.03Cl unit cell. b One-Dimensional variation of electron density along
[100], [110] and [111] directions of the Na0.90Ag0.10Cl unit cell

(a) (b) 0.25

90
0.24 Na0.97Ag0.03Cl
Na0.97Ag0.03Cl
Electron density (e/Å3)

Na0.90Ag0.10Cl
Electron density (e/Å3)

80 Na0.90Ag0.10Cl

70 0.23

60 0.22
50
0.21
40

30 0.20

20 0.19
10
0.18
0
0.17
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
Distance along [100] direction (Å) Distance along [100] direction (Å)

Fig. 3.38 a The peak electron density profiles of Na (Ag) along the [100] direction for the two
compositions, (Na 0.97Ag0.03Cl and Na0.90Ag0.10Cl). b The mid-bond electron density of Na (Ag)
along the [100] direction for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl
3.6 Silver Doped in NaCl (Na1-xAgxCl) 135

Table 3.28 Parameters from MEM refinement

Direction Na0.97Ag0.03Cl Na0.90Ag0.10Cl
3
Position (Å) Electron density (e/Å ) Position (Å) Electron density (e/Å3)
[001] 1.422 0.175 1.423 0.183
[110] 2.010 0.148 2.012 0.187
[111] 2.462 0.057 1.971 0.084

Fig. 3.39 a Dynamic multipole deformation (DMD) map of Na0.97Ag0.03Cl on the (100) plane.
b Dynamic multipole deformation (DMD) map of Na0.90Ag0.10Cl on the (100) plane

Na0.97Ag0.03Cl and Na0.90Ag0.10Cl. The numerical values of the electron densities

along different directions are given in Table 3.28.

3.6.5 Multipole Analysis

Hansen and Coppens (1978) proposed a modified electron density model with the
option that allows the refinement of population parameters at various orbital levels
where the atomic density is described as a series expansion in real spherical
harmonic functions through fourth order Ylm.
Multipole electron density computation was carried out according to Hansen
and Coppens (1978) in this work. The electron density from multipole analysis is
represented in Fig. 3.39a, b as DMD maps on the (100) plane for Na0.97Ag0.03Cl
and Na0.90Ag0.10Cl respectively.
Figure 3.40a, b represent the SMD maps on (100) plane for Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl respectively.
136 3 Results and Discussion on Metals and Alloys

Fig. 3.40 a Static multipole deformation map of Na0.97Ag0.03Cl on the (100) plane. b Static
multipole deformation map of Na0.90Ag0.10Cl on the (100) plane

Fig. 3.41 a Dynamic multipole deformation map of Na0.97Ag0.03Cl on the (110) plane.
b Dynamic multipole deformation map of Na0.90Ag0.10Cl on the (110) plane

Table 3.29 Multipole electron densities

System Residual densities
Na0.97Ag0.03Cl Na0.90Ag0.10Cl
DMD analysis
Positive max. (e/Å3) 1.96 8.7
Negative max. (e/Å3) -0.04 -0.18
SMD analysis
Positive max. (e/Å3) 0.11 0.11
Negative max.(e/Å3) -0.04 -0.11
3.6 Silver Doped in NaCl (Na1-xAgxCl) 137

Fig. 3.42 a Static multipole deformation map of Na0.97Ag0.03Cl on the (110) plane. b Static
multipole deformation map of Na0.90Ag0.10Cl on the (110) plane

Figure 3.41a, b represent DMD maps on (110) plane for Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl respectively.
Similarly Fig. 3.42a, b represent the SMD maps on (110) plane for
Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively. The results of the multipole
analysis are given in Table 3.29.

3.6.6 Results and Discussion

The effect of impurity doping has been analysed in this work, in terms of the
electron density distribution using the two versatile techniques MEM and multi-
pole analysis. The experimental structure factor data sets of Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl were subjected to Rietveld (1969) refinements to obtain the indi-
vidual thermal vibration parameters of the atoms Na(Ag) and Cl. The refined
thermal vibration parameters as given in Table 3.26 shows that the Debye–Waller
factor of Na(Ag) increases with the concentration of Ag, whereas the Debye–
Waller factor value of Cl decreases. When the NaCl is doped with Ag impurity, the
local distortion increases with impurity concentration. Hence, the Debye–Waller
factor of Na(Ag) increases with Ag. Since the electrostatic interaction between the
Na(Ag) and Cl ions increases with Ag concentration, the Cl atom has less thermal
vibration as reflected in Table 3.26. The Rietveld refinements give very low
reliability indices as given in Table 3.26. The refined results were used for the
MEM computations. The three-Dimensional electron density of Na0.97Ag0.03Cl
and Na0.90Ag0.10Cl in terms of isosurfaces as shown in Fig. 3.34a, b can be
138 3 Results and Discussion on Metals and Alloys

visually compared for the ionic sizes and the interionic distances. It is seen from
these figures that the electronic charge clouds are more diffuse in Na0.90Ag0.10Cl
and the separation between charges decreases as the concentration of Ag increases.
Since the charge clouds of Cl expand due to more electrostatic attraction in
Na0.90Ag0.10Cl, the thermal vibration of Cl is restricted, resulting in smaller
Debye–Waller factor of Cl.
The increase in the ionic sizes can be visualized in the two-Dimentional electron
density contour maps on (100) and (110) plane (Figs. 3.35a, b and 3.36a, b) for
Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively. Particularly, Fig. 3.36b for
Na0.90Ag0.10Cl shows increased ionic sizes compared to those in Fig. 3.36a. Hence,
the effect of impurity addition in the host lattice is substantiated.
From the one-dimensional analysis of electron density it is found that there is an
increase in the mid-bond density in Na0.90Ag0.10Cl compared to that in
Na0.97Ag0.03Cl in all the three directions. The minimum density along [100]
direction occurs at 1.4251 and 1.4227 Å respectively, for Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl indicating an increase in the ionic radius in Na0.90Ag0.10Cl.
Moreover, the electron density at these positions is 0.175 and 0.183 e/Å3 for
Na0.97Ag0.03Cl and Na0.90Ag0.10Cl respectively. This clearly indicates the addition
of Ag ion in the host lattice. A similar trend i.e., increased mid-bond position and
density is observed along [110] direction also, in Na0.90Ag0.10Cl. Along [111]
direction, there is an increase in electron density in Na0.90Ag0.10Cl compared to
Na0.97Ag0.03Cl, but the position of mid-bond density is decreased in
Na0.90Ag0.10Cl. But one can observe horizontal lines along [111] direction and
only the minimum numerical value decides the electron density, but the bonding
characteristics is decided by the electron density distribution along [100] direction
which supports the addition of Ag impurities in the host lattice. The [100] profiles
provide an estimation of the ionic radii of atoms Na(Ag) and Cl.
The DMD maps on (100) plane (Fig. 3.39a, b) for Na0.97Ag0.03Cl and
Na0.90Ag0.10Cl clearly indicate that the residual densities are more concentrated
and contracted in Na0.90Ag0.10Cl than Na0.97Ag0.03Cl. The SMD maps in
Fig. 3.40a, b show positive clusters (contour line) between the atom indicative of
more electrostatic interaction in Na0.90Ag0.10Cl compared to Na0.97Ag0.03Cl.
A similar trend is observed on (110) planes (Fig. 3.41a, b). Figure 3.42b shows
(for Na0.90Ag0.10Cl) charge clouds of larger radius between the atoms compared to
the electron clouds of Na0.90Ag0.10Cl. The multipole analysis gives the multipole
parameters j and j0 which are as follows. The j and j0 values of Na(Ag) in
Na0.97Ag0.03Cl and Na0.90Ag0.10Cl was refined keeping the Cl ion fixed. For
Na0.90Ag0.10Cl the j and j0 parameters are 1.805(0.605) and 1.0703(0.5411)
respectively. These values are 1.776(0.5040) and 1.0009(0.0544) respectively for
Na0.90Ag0.10Cl. Both values for Na0.90Ag0.10Cl reveal an expansion of Na(Ag) ion
in Na0.90Ag0.10Cl. Table 3.29 shows that the residual density is larger in
Na0.90Ag0.10Cl as analysed by DMD method.
3.6 Silver Doped in NaCl (Na1-xAgxCl) 139

3.6.7 Conclusion

Both MEM and multipole analyses give a clear picture of the bonding in the doped
system NaCl. Hence, these two methods can be used in combination to study the
electronic charge distribution in pure and doped systems which will help to
understand the physical and conducting phenomena of these materials.

3.7 Aluminium Doped with Dilute Amounts of Iron

Impurities (0.215 and 0.304 wt% Fe)

3.7.1 Introduction

The analysis of bonding and electronic structure of metals is useful because of the
variety of applications of metals, particularly when they are doped with other
elements. There has been tremendous impetus in the study of electronic structure
of solids, both theoretically and experimentally due to necessity. When the system
is dilute doped, it is possible to study only the average electronic structure. But,
when suitable tools like MEM are available, they can be effectively used to elu-
cidate the fine features of bonding as attempted in this work. Fourier synthesis of
electron densities can be of use in picturing bonding between two atoms, but, it
suffers from the major disadvantage of series termination error and negative
electron densities that prevent the clear understanding of the fine nature of bonding
between atoms, the factor intended to be analysed. The advent of MEM solves
many of these problems. MEM electron densities are always positive and even
with limited data, one can determine reliable electron densities resembling true
densities. In this work, such an analysis has also been carried out using the for-
malism of Collins (1982) using the data sets reported by Mohanlal and Pathinettam
Padiyan (1987).
In our series of research studies on bonding in materials, we have analysed LiF,
NaF, GaAs, CdTe and some oxides (Israel et al. 2002; Israel et al. 2003; Saravanan
et al. 2002a, b; Kajitani et al. 2001) using MEM and reported X-ray data and
obtained precise information about bonding in the above materials. The nature of
bonding in these materials as analysed are found to be ionic, mixed covalent and
ionic and ‘oxide’ bonding.

3.7.2 Summary of the Work

The electronic structure of pure and doped aluminium with dilute amounts of iron
impurities (0.215 and 0.304 wt% Fe) has been analysed using reported X-ray data
sets and MEM. Qualitative as well as quantitative assessment of the electron
140 3 Results and Discussion on Metals and Alloys

density distribution in these samples has been made. The mid-bond character-
ization leads to a conclusion about the nature of doping of impurities. An expan-
sion of size of the host aluminium atom is observed with Fe impurities.
In this study on aluminium with iron impurities, we have attempted to study
metal bonding and bring out the details of doping using MEM. The X-ray structure
factor data sets have been taken from Mohanlal and Pathinettam Padiyan (1987)
who reported precise structure factors collected from a home-built manual X-ray
diffractometer. Data sets for pure aluminium, Al ? 0.189, Al ? 0.215 and
Al ? 0.304 wt% Fe are available from this report Mohanlal and Pathinettam
Padiyan (1987). Other relevant details can be found from this paper. Among these
data sets, the one corresponding to Al ? 0.189 wt% Fe has been excluded from the
present analysis after examining the preliminary results from MEM and multipole
methods (the low doping level does not allow any meaningful conclusions of the
obtained results). The remaining three data sets, i.e., those of pure aluminium,
Al ? 0.215 and Al ? 0.304 wt% Fe were used in this study. It was observed that
the Beff and the expected number of vacant sites in Al ? 0.215 wt% Fe are much
better than those in Al ? 0.189 wt% Fe, although slight variations of these
quantities is observed, compared to pure aluminium and Al ? 0.304 wt% Fe [it
has been reported (Mohanlal and Pathinettam Padiyan (1987) that the number of
vacant sites created in the Al lattice increases with the number of Fe impurities and
in the case of Al ? 0.189 wt% Fe, less vacancies are created than expected].
Hence, a data set for Al ? 0.215 wt% Fe has been included in the present analysis
in order to have a composition of iron between pure aluminium and Al ? 0.304
wt% Fe and to compare the results.

3.7.3 Data Analysis

All the reported data sets have been analysed using JANA 2000 (Petřiček et al.
2000) software package. The refined structure factors were used for MEM
refinements of the electron densities. The characteristics of the structural refine-
ments are given in Table 3.30.
The reliability indices (R-factors) and the matching of thermal parameters (B)
with the corresponding reported values for all three crystal structures indicate that
the quality of the experimental data is suitable enough for the electron density
analysis. The same number of data for all three compositions samples has been
utilised for the present analysis in order to make a comparison of the results.
MEM calculations were done by dividing the cell into 1283 pixels. For
numerical MEM computations, the software package PRIMA (Izumi and Dilanian
2002) was used. For two- and three-dimensional representations of the electron
densities, the programme VESTA (Momma and Izumi 2006) package was used. In
all three cases, the convergence condition C = 1.0, was obtained after several
cycles of MEM refinement. The main characteristics of the refinements are listed
in Table 3.31.
3.7 Aluminium Doped with Dilute Amounts of Iron Impurities 141

Table 3.30 Structure refinement parameters from JANA 2000

Sample Cell constant (Å) Number of reflections B (Å2) R (%) wR (%)
Al 4.0440 18 0.825(069) 2.22 2.82
Al ? 0.215% Fe 4.0501 18 1.038(061) 2.61 2.86
Al ? 0.304% Fe 4.0452 18 1.442(053) 2.61 2.82
B Debye–Waller factor

Table 3.31 Parameters from MEM refinements

Sample Number of Cycles R(%) wR(%) Lagrange Resolution of the
reflections parameter (k) map (Å/pixel)
Al 18 713 0.2276 0.2867 0.010 0.0316
Al ? 0.215% Fe 18 832 0.2939 0.2969 0.012 0.0316
Al ? 0.304% Fe 18 5703 0.2673 0.3189 0.012 0.0316

Table 3.32 Maxima of the MEM electron density revealed in different one-dimensional sections
of a unit cell for different compositions
Composition Direction
[100] [110] [111]
Position (Å) ED (e/Å ) Position (Å)) ED (e/Å ) Position (Å) ED (e/Å3)
3 3

Al 0.000a 128.75 0.000a 128.75 0.000a 128.75

1.485 0.14 1.429 0.22 1.477 0.15
2.022 0.19 1.698 0.20 1.751 0.15
Al ? 0.215% Fe 0.000a 109.21 0.000a 109.21 0.000a 109.21
1.455 0.14 1.298 0.19 1.534 0.15
2.025 0.19 1.432 0.19 1.753 0.15
2.631 0.12
3.508 0.19
4.384 0.12
5.261 0.15
Al ? 0.304% Fe 0.000a 85.79 0.000a 85.79 0.000a 85.79
1.517 0.12 1.430 0.15 1.752 0.11
2.022 0.15 2.080 0.11
2.682 0.09
3.503 0.15
4.324 0.09
5.254 0.11
a
At origin
ED Electron density

The reliability indices are found to be very low. To analyse the MEM results,
one-dimensional profiles along different directions of the unit cell and two-
dimensional electron density contour maps have been constructed. The results of
the one-dimensional analysis have given in Table 3.32.
142 3 Results and Discussion on Metals and Alloys

Fig. 3.43 a 3-D representation of the electron density of pure Al super imposed with the
structural model. Isosurface level is 0.2. b 3-D representation of the electron density of
Al ? 0.215% Fe super imposed with the structural model. Isosurface level is 0.2. c 3-D
representation of the electron density of Al ? 0.304% Fe imposed with the structural super
imposed with the structural model. Isosurface level is 0.2

Fig. 3.44 a MEM electron density distribution of pure Al on the (100) plane. Contour range is
from 0.0 to 5.0 e/Å3. Contour interval is 0.03 e/Å3. b MEM electron density distribution of
Al ? 0.215% on the (100) plane. Contour range is from 0.0 to 5.0 e/Å3. Contour interval is
0.03 e/Å3. c MEM electron density distribution of pure Al ? 0.304% on the (100) plane. Contour
range is from 0.0 to 5.0 e/Å3. Contour interval is 0.03 e/Å3

Fig. 3.45 a MEM electron density distribution of pure Al on the (110) plane. Contour range is
from 0.0 to 5.0 e/Å3. Contour interval is 0.03 e/Å3. b MEM electron density distribution of pure
Al ? 0.215% Fe on the (110) plane. Contour range is from 0.0 to 5.0 e/Å3. Contour interval is
0.03 e/Å3. c MEM electron density distribution of pure Al ? 0.304% Fe on the (110) plane.
Contour range is from 0.0 to 5.0 e/Å3. Contour interval is 0.03 e/Å3
3.7 Aluminium Doped with Dilute Amounts of Iron Impurities 143

(a) (b)
140 0.25 140
0.5
Al Al+0.215 Fe Al
Al+0.215 Fe 0.20 Al+0.215 Fe 0.4
120 Al 120
Al+0.304 Fe Al+0.304 Fe 0.3 Al
Electron Density (e/Å 3)

Electron Density (e/Å )

0.15

3
100 100 0.2 Al+0.215 Fe
Al+0.304 Fe Al+0.304 Fe
1.0 1.3 1.6 1.9 2.2 2.5 2.8 0.7500 1.0875 1.4250 1.7625
80 80

60 60

40 40

20 20

0.00 0.41 0.82 1.23 1.64 2.05 2.46 2.87 3.28 3.69 0.0 0.6 1.2 1.8 2.4 3.0 3.6 4.2 4.8 5.4
Distance along [100] direction (Å) Distance along [110] direction (Å)

(c) 140 0.22

0.20 Al+0.215 Fe
Al 0.18
120 Al+0.215 Fe 0.16 Al
0.14
Al+0.304 Fe
0.12
Electron Density (e/Å3)

100 0.10
Al+0.304 Fe
0.08
0 1 2 3 4 5 6 7
80

60

40

20

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Distance along [111] direction (Å)

Fig. 3.46 a One-dimensional MEM low electron density profiles of pure Al, Al ? 0.215% Fe
and Al ? 0.304% Fe along the [100] direction. The mid-bond variation of the electron density is
shown as insert in which the y-axis represents the electron density in e/Å3 and the x-axis, the
distance along the [100] direction in Å. b One-dimensional MEM low electron density profiles of
pure Al, Al ? 0.215% Fe and Al ? 0.304% Fe along the [110] direction. The mid-bond variation
of the electron density is shown as insert, in which the y-axis represents the electron density in e/
Å3 and the x-axis, the distance along the [110] direction in Å. c One-dimensional MEM low
electron density profiles of pure aluminium, Al ? 0.215% Fe and Al ? 0.304% Fe along the
[111] direction. The mid-bond variation of the electron density is shown as insert, in which the
y-axis represents the electron density in e/Å3 and the x-axis, the distance along the [111] direction
in Å

The three-dimensional electron density superimposed with the structural model

of Al is shown for three compositions in Fig. 3.43a, b, c respectively. The 2-D
electron density contour maps drawn on (100) planes are given in Fig. 3.44a–c.
Figure 3.45a–c represent the 2-D maps on the (110) plane. The one-dimensional
variation of the electron density is shown in Fig. 3.46a–c. Figure 3.46a represents
the variation along the [100] direction for all three compositions, Fig. 3.46b along
the [110] direction and Fig. 3.46c along the [111] direction. The analysis on the
MEM results is given in the following section.
144 3 Results and Discussion on Metals and Alloys

3.7.4 Results and Discussion

As expected from the substitution Al by Fe, the peak density decreases with the
increase of Fe impurity (Fig. 3.46). As more and more Fe atoms occupy the host
aluminium lattice substitutionally, the size of the charge cloud increases due to the
larger atomic radius of iron (1.72 Å—for aluminium, the atomic radius is 1.43 Å).
The electron density between atoms along [110] direction also decrease with
impurities (Table 3.32 and Fig. 3.45b) although it is almost a constant along the
other two directions, [100] and [111] (Table 3.32). The bonding direction shows
(which is [110] direction, because the first nearest neighbours are along this
direction), that the mid-bond electron density also decreases with the addition of
Fe impurities as shown in the inset of Fig. 3.46b. The mid-bond density along the
other two directions, [100] and [111] (insets of Fig. 3.46a, c), do not reveal any
trend with the Fe impurities, because the values of the densities involved are very
low along these non-bonding directions and the humps seen in the insets of
Fig. 3.46a, c are due to the interference of the electron density of the FCC atoms in
the unit cell (NNM—Non Nuclear Maxima, along the non-bonding directions).
The MEM electron density distribution of pure Al, Al ? 0.215% Fe and
Al ? 0.304% Fe on the (100) plane (Fig. 3.44a–c) shows clear densities with very
low noise levels. The large thermal vibration and smearing of charges is seen.
Hence the increased size of the electronic clouds can be attributed to the static and
dynamic thermal vibration parameters and also to the increased host lattice sub-
stitution of the Fe atoms.

3.7.5 Conclusion

The electron density distribution in the unit cell of Al with different iron impurity
has been studied precisely using the currently available most versatile technique,
the MEM analysis of the electron densities. Minute changes in the density details
and the incorporation of impurities in the host lattice have been revealed. The
electron density and the mid-bond electron density decrease with more Fe impurity
doping as evidenced from the 1D electron densities. Hence, the bonding strength in
Al decreases with Fe addition. In addition to the electron density details showing
up themselves as actual distributions of atomic electrons, the thermal parameters
also show a similar trend with respect to impurity addition. Moreover, an
expansion of the host lattice with the addition of more Fe impurities is observed
pictorially as revealed by the two-dimensional electron densities in Fig. 3.44.
Since, the crystallographic system presently studied is highly symmetric, there are
no appreciable changes in the mid-bond positions despite the increase in the
atomic sizes with more addition of Fe impurities.
References 145

References

Ak NF, Tekmen C, Ozedmir I, Soykan HS, Celik E (2003) Surf Coat Technol 174:1070–1073
Al-Quran FMF, Al-Itawi HI (2010) Eur J Sci Res 39:251–256
Banerjee NK, Mitra GB (1969) Indian J Pure Appl Phys 71:13
Bashir J, Butt NM, Khan QH (1988) Acta Cryst A44:638–639
Bowran DT, Finney JL (2003) J Chem Phys 118:8357–8372
Burley G (1967) Acta Cryst 1(23):1–5
Collins DM (1982) Nature 298:49–51
Coppens P, Volkov A (2004) Acta Cryst A60:357–364
Debay P, Menki H (1930) Physika Zeit 31:797–798
Egami T (1990) JIM 31(3):163–176
Enríquez JP, Mathew X (2003) J Cryst Growth 259:215–222
Eskandarany MS (2001) Mechanical alloying for fabrication of advanced engineering materials
(Chapter 1). Noyes publications, New York
Farrow CL, Juhas P, Liu JW, Bryndin D, Bozin ES, Bloch J, Proffen T, Billinge SJL (2007)
PDFfit2 and PDF gui: computer programs for studying nanostructures in crystals. J Phy Cond
Matt Phy 19:335219
Field DW, Medlin EH (1974) An X-ray sturdy of thermal vibrations in sodium. Acta Cryst
A30:234–238
Frils J, Jiang B, Spence J, Marthinsen K, Holmestad R (2004) Acta Cryst A60:402–408
Gull SF, Daniel GJ (1978) Nature 272:686–690
Hansen NK, Coppens P (1978) Acta Crys A34:909–921
Ikeda T, Takata M, Sakata M, Waliszewski J, Dobrzynski L, Porowski S, Jun J (1998) J Phys Soc
Jpn 67:4104–4109
Israel S, Saravanan R, Srinivasan N, Rajaram RK (2002) Phys Chem Solids 64:43–49
Israel S, Saravanan R, Srinivasan N, Rajaram RK (2003) J Phys Chem Solids 64:879–886
Israel S, Saravanan R, Rajaram RK (2004) Physica B 349:390–400
Izumi F, Dilanian RA (2002) Recent research developments in physics, vol 3, Part II. Transworld
Research Network, Trivandrum , pp 699–726
Jeng SP, Holloway H, Douglas A, Hoflund GB (1990) Surf Sci 235:175–185
Jeong IK, Thompson J, Proffen T, Perez A, Billinge SJL (2001) PDFGetX: a program for
obtaining the atomic pair distribution function from X-ray powder diffraction data. J Appl
Cryst 34:536
Kajitani T, Saravanan R, Ono Y, Ohno K, Isshiki M (2001) J Cry Growth 229:130–136
Kamal S, Jayaganthan R, Prakash S (2009) J Alloy Compd 472:378–389
Kazi IH, Wild PM, Moore TN, Sayer M (2003) Thin Solid Films 433:337–343
Kazi IH, Wild PM, Moore TN, Sayer M (2006) Thin Solid Films 515:2602–2606
Kim N, Kwon NH, Park Y, Choi J (2005) Electron Lett 41:982–984
Larson AC, Von Dreele RB (2004) General structure analysis system, Lansce, MS-H805, Los
Alamos national laboratory, Los Alamos, LAUR 86748
Linkoaho MV (1969) Acta Crystallographica A 25 III:450–455
Linkoaho MV (1972) Physica Scripta 5:271–272
Marabello D, Bianchi R, Gervasio G, Cargnoni F (2004) Acta Cryst A60:494–501
Mohanlal SK (1979) J Phys C Solid State Phys 12:651–653
Mohanlal SK, Pathinettam Padiyan D (1987) Cryst Latt Def Amorph Mat 14:651–653
Mohanlal SK, Pathinettam Padiyan D (1989) Radiat Eff Defects Solids 110:385–391
Momma K, Izumi F (2006) Commission on crystallography computing. IUCr Newslett 7:106–119
Murray GT (1997) Handbook of materials selection for engineering applications. Dekker, Inc.
New York
Na SH, Kim SH, Lee YW, Sohn DS (2002) J Korean Nucl Soc 34:60–67
Peng LM, Ren G, Dudarey SL, Whelan M (1996) J Acta Cryst A52:456–470
146 3 Results and Discussion on Metals and Alloys

Peng Y, Cullis T, Möbus G, Xu X, Inkson B (2009) IOP Electron J Nanotechnology 20:

395708–395714
Peterson PF, Bozin ES, Proffen TH, Billinge SJL (2003) J Appl Cryst 36:53–64
Petkov V, Billinge SJL, Larsen PD, Mahanthi T, Vogt KK, Rangen MG, Kanatzidis (2002) Phys
Rev B 65:092105–092108
Petřiček V, Dušek M, Palatinus L (2000) JANA 2000: the crystallographic computing system,
institute of physics. Academy of Sciences, Praha
Petřiček V, Dušek M, Palatinus L (2006) JANA 2000: the crystallographic computing system,
institute of physics. Academy of Sciences, Praha
Pillet S, Souhassou M, Lecomte C (2004) Acta Cryst A60:455–464
Poulsen RD, Bentien A, Graber T, Iversen BB (2004) Acta Cryst A60:382–389
Proffen T, Billinge SJL (1999) PDFFIT: a Program for full profile structural refinement of the
atomic pair distribution function. J Appl Cryst 32:572–575
Proffen T, Neder R (1997) DISCUS: a Program for diffuse scattering and defect structure
simulations. J Appl Cryst 30:171–175
Radomysel’skii ID, Kholodnyi IP (1975) Powder Metall Met Ceram 14(6):478–486
Ram A (1985) CRC handbook of chemistry and physics, 64th edn. CRC press, Florida
Rietveld HM (1969) J Appl Crystallogr 2:65–71
Robert MC, Saravanan R, Saravanakumar K, Prema Rani M (2009) Z Naturforsch 64a:1–9
Ruben AD, Fujio I (2004) Super-fast program PRIMA for the maximum-entropy method,
advanced materials laboratory, National Institute for Materials Science, Ibaraki, Japan,
p 3050044
Saka T, Kato N (1986) Acta Cryst A 42:469–478
Sakata M, Sato M (1990) Acta Cryst A 46:263–270
Saravanan R, Prema Rani M (2007) J All Compd 431:121–129
Saravanan R, Ono Y, Isshiki M, Ohno K, Kajitani T (2002a) J Phys Chem Solids 64/1:51–58
Saravanan R, Israel S, Swaminathan S, Kalidoss R, Muruganantham M (2002b) Cryst Res
Technol 37:1310–1317
Saravanan R, Ono Y, Ohno IM, Ohno K, Kajitani T (2003) J Phys Chem Solids 64:51–58
Saravanan R, Majella MAA, Jainulabdeen S (2007) Physica B 400:16–21
Saravanan R, Syed Ali KS, Israel S, Pramana (2008) J Phys 70(4):679–696
Shahi K, Wagner JB (1981) Phys Rev B 23(12):6417–6421
Syed Ali KS, Saravanan R, Israel S (2009) J Cryst Growth 311:1110–1116
Takata M, Umeda B, Nishibori E, Sakata M, Saito Y, Ohno M, Shinohara H (1995) Nature
377:46–49
Teeple HO (1954) Nickel and high-nickel alloys. Ind Eng Chem 46:2092–2107
Tobi H, Egami T (1992) Acta Cryst A 48:336–346
Valvoda V (1970) Acta Crystallographica A 26:405–406
Waasmaier D, Kirfel A (1995) Acta Cryst A51:416–431
Weiss RJ (1966) X-ray determination of electron densities. North-Holland Publishing,
Amsterdam
Wirojanupatump S, Shipway PH, Cartney DGM (2001) Wear 249:829–837
Wyckoff RWG (1963) Crystal structure, vol 1. Inter-Science publishers, London
Xiangyun Q, Proffen T, Michell JF, Billinge SJL (2005) Phy Rev let 94:177203–177206
Laugier J, Bochu B GRETEP: Domaine universitaire BP 46, 38402 Saint Martin d’Hères. http://
www.inpg.fr/LMGP
Saravanan R GRAIN software. http://www.saraxraygroup.net
Wilkinson K. http://www.chemclub.com/teachers
Yamamoto K, Takahashi Y, Ohshima K, Okamura FP, Yukino K (1996) Acta Cryst
A52:606–613
Yamamura S, Takata M, Sakata M, Suguwara Y (1998) J Phys Soc Jpn 67:4124–4127
Zwick LZJ, Lugscheider E (2004) Surf Coat Technol 82(1):72–77
Chapter 4
Conclusion

The average and local structures of some elemental metals, alloys and dilute doped
alloys have been analysed from X-ray data using versatile techniques like Rietveld
method, MEM (maximum entropy method), multipole and PDF (pair distribution
function). The materials studied in this research work are listed below.

Metals
• Sodium (Na)
• Vanadium (V)
• Magnesium (Mg)
• Aluminium (Al)
• Titanium (Ti)
• Iron (Fe)
• Nickel (Ni)
• Copper (Cu)
• Zinc (Zn)
• Tin (Sn)
• Tellurium (Te)
Alloys
• Cobalt aluminium (CoAl)
• Nickel aluminium (NiAl)
• Nickel chromium (NiCr)
• Iron nickel (FeNi)
Dilute doped alloys
• Sodium chloride with iron impurities (Na1-xAgxCl)
• Aluminium, with iron impurities (0.215 and 0.304 wt% Fe)
The materials studied in this research work have been divided into several parts
for the exhaustive analytical study and the results of this research work are
presented below.

R. Saravanan and M. Prema Rani, Metal and Alloy Bonding: An Experimental Analysis, 147
DOI: 10.1007/978-1-4471-2204-3_4,  Springer-Verlag London Limited 2012
148 4 Conclusion

4.1 Sodium and Vanadium Metals

• The nature of bonding and the charge distribution in sodium and vanadium
metals have been analysed using the reported X-ray data of these metals.
• The bonding in these metals has been elucidated and analysed using MEM
(maximum entropy method) and multipole method.
• The nature of bonding is clearly revealed by the two-dimensional MEM maps
plotted on (100) and (110) planes and the mid-bond densities are clearly
revealed by the one-dimensional electron density along the [100], [110] and
[111] direction.
• The mid-bond densities in sodium and vanadium are found to be 0.014 and
0.723 e/Å3 respectively, giving an indication of the strength of the bonds in
these materials.
• In sodium, the very large thermal vibration of the sodium atom is seen in the
electron density map on the (100) plane.
• From multipole analysis, the large j0 parameter of the valence part of the
sodium atom denotes the contraction of the sodium atom.
• Static and dynamic multipole deformation maps drawn clearly show the thermal
effect on the electron density in the core and valence regions.
• The sodium atom is found to contract more than the vanadium atom from the
multipole analysis.

4.2 Aluminium, Nickel and Copper

• The average and local structure of simple metals Al, Ni and Cu have been
elucidated for the first time using MEM (Maximum Entropy Method), multipole
and PDF (Pair Distribution Function).
• The powder X-ray data sets were refined by Rietveld method using the software
program JANA 2000.
• The refined structure factors were used for the MEM and multipole analysis.
• The electron density distribution of Al, Ni and Cu has been mapped using the
MEM electron density values obtained through refinements.
• The mid-bond density in Ni is found to be larger among the three metal systems
along [110] direction. But along the [100] and [111] directions (non-bonding),
the element with the higher atomic number has a higher electron density.
• The thermal contribution to the charge density has been realized through the
dynamic and static multipole deformation maps.
• From multipole results there is appreciable contraction in Ni compared to the
other two systems.
4.2 Aluminium, Nickel and Copper 149

• The observed PDF of Al, Ni and Cu were obtained from the raw intensity data
using a software program PDFgetX.
• A comparison between the observed and calculated PDF has been carried out
using the software package PDFFIT.
• The PDF results give a clear picture of the local structure in these metals.

4.3 Magnesium, Titanium, Iron, Zinc, Tin and Tellurium

• The average and local structures of magnesium, titanium, iron, zinc, tin and
tellurium have been analysed using the MEM (maximum entropy method), and
PDF (pair distribution function).
• The structural parameters of the metals were refined from the powder X-ray data
with the well-known Rietveld powder profile fitting methodology using the
software package JANA 2006.
• The refined structure factors were used for the MEM analysis
• The one-, two- and three-dimensional electron density distributions of Mg, Ti,
Fe, Zn, Sn and Te have been mapped using the MEM electron density values
obtained through refinements.
• The variation of electron density distribution with respect to the thermal
vibration factor is clearly visualised.
• The electron density has been analysed qualitatively and quantitatively.
• The local structure has been analysed for the accurate nearest neighbour dis-
tances using PDF (pair distribution function).

4.4 Cobalt Aluminium and Nickel Aluminium Metal Alloys

• The precise electron density distribution and bonding in metal alloys CoAl and
NiAl are characterized using MEM (maximum entropy method) and multipole
method from reported single-crystal data sets.
• The mid-bond electron densities in these systems are found to be 0.358 and
0.251 e/Å3 respectively, for CoAl and NiAl in the MEM analysis.
• The two-dimensional maps and one-dimensional electron density profiles have
been constructed and analysed.
• The thermal vibration of the individual atoms Co, Ni and Al has also been
studied and reported.
• The thermal vibration of Al is more pronounced in CoAl than in NiAl.
• The contraction of atoms in CoAl and expansion of Ni and contraction of Al
atom in NiAl is found from the multipole analysis.
150 4 Conclusion

4.5 Nickel Chromium (Ni80Cr20)

• The alloy Ni80Cr20 has been annealed and ball milled to study the effect of
thermal and mechanical treatments on the local structure and the electron
density distribution.
• The electron density between the atoms was studied by MEM (maximum
entropy method) and the local structure using PDF (pair distribution function).
• The electron density is found to be high for ball-milled sample along the
bonding direction.
• The particle sizes of the differently treated samples were realized by SEM and
through XRD.
• Clear evidence of the effect of ball milling is observed on the local structure and
electron densities.

4.6 Silver Doped in NaCl (Na12xAgxCl)

• The alkali halide Na1-xAgxCl, with two different compositions (x = 0.03 and
0.10), was studied with regard to the Ag impurities in terms of the bonding and
electron density distribution using X-ray single-crystal data sets.
• Electron density distribution and hence the interaction between the atoms is
clearly revealed by maximum entropy method (MEM) and multipole analyses.
• The bonding in these systems has been studied using two-dimensional MEM
electron density maps on the (100) and (110) planes and one-dimensional
electron density profiles along the [100], [110] and [111] directions.
• The mid-bond electron densities between atoms in these systems are found to be
0.175 and 0.183 e/Å3, respectively, for Na0.97Ag0.03Cl and Na0.90Ag0.10Cl.
• Multipole analysis of the structure has been performed for these two systems,
with respect to the expansion/contraction of the ion involved.

4.7 Aluminium Doped with Dilute Amounts of Iron Impurities

(0.215 and 0.304 wt% Fe)

• The electronic structure of pure and doped aluminium with dilute amounts of
iron impurities (0.215 and 0.304 wt% Fe) has been analysed using reported
X-ray data sets and the MEM (maximum entropy method) technique.
• Qualitative as well as quantitative assessment of the electron density distribution
in these samples has been made.
• An expansion of size of the host aluminium atom is observed with Fe impurities.
Minute changes in the density details and the incorporation of impurities in the
host lattice have been revealed.
4.7 Aluminium Doped with Dilute Amounts of Iron Impurities 151

• The electron density and the mid-bond electron density decrease with more Fe
impurity doping as evidenced from the one-dimensional electron densities.
• The bonding strength in Al decreases with Fe addition.
The broad applicability of structure determination and refinement with single
crystals and powders make the future of diffraction studies more exciting and
essential. The success of extraction of details mainly depends on three parameters,
choice of measurement device, pattern profile description and structure solving
algorithms. Modern instrumentation, sources yielding data of unprecedented
quality and modern analysis methods continue to increase our ability to harvest
useful information from the data.
In this research work the charge density of metals and alloys analysed using
versatile tools such as maximum entropy method, multipole and pair distribution
function have been done for the first time. These analyses give the fine details of
the electron environment in the unit cell. The changes in the properties of the
alloys due to the inclusion of dopant elements are clearly picturised. These local
and average analyses of the materials facilitate the better understanding of their
properties and enhance the engineering of new materials.