Journal of Chromatography A, 1217 (2010) 2326–2336

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Review

Single drop microextraction—Development, applications

and future trends
Michael A. Jeannot a,∗ , Andrzej Przyjazny b , John M. Kokosa c
a
Department of Chemistry, St. Cloud State University, 366 Wick Science Building, 720 4th Ave. S., St. Cloud, MN 56301-4498, USA
b
Department of Chemistry & Biochemistry, Kettering University, Flint, MI 48504 USA
c
MDRC Consulting/Mott Community College, Flint, MI 48503 USA

a r t i c l e i n f o a b s t r a c t

Article history: Single drop microextraction (SDME) has emerged over the last 10–15 years as one of the simplest and most
Available online 6 November 2009 easily implemented forms of micro-scale sample cleanup and preconcentration. In the most common
arrangement, an ordinary chromatography syringe is used to suspend microliter quantities of extracting
Keywords: solvent either directly immersed in the sample, or in the headspace above the sample. The same syringe
Single drop microextraction is then used to introduce the solvent and extracted analytes into the chromatography system for iden-
Solvent microextraction
tification and/or quantitation. This review article summarizes the historical development and various
Liquid-phase microextraction
modes of the technique, some theoretical and practical aspects, recent trends and selected applications.
© 2009 Elsevier B.V. All rights reserved.

Contents

1. Brief history of development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2327

2. Important theoretical aspects of SDME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2327
2.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2327
2.2. Two-phase SDME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2327
2.3. Dynamic SDME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2328
2.4. SDME with back-extraction (aqueous–organic–aqueous) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2328
2.5. Headspace SDME (aqueous–headspace–organic) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2328
3. A closer look at various SDME modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2328
4. Experimental parameters affecting DI-SDME and HS-SDME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
4.1. Analyte properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
4.2. Solvent properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
4.3. Extracting solvent purity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
4.4. Syringe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2332
4.5. Drop volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333
4.6. Agitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333
4.7. Ionic strength (salting out effect) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333
4.8. Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333
4.9. Sample volume and headspace volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2333
4.10. Automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2334
5. New developments and trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2334
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2335
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2335

∗ Corresponding author. Tel.: +1 320 308 2046; fax: +1 320 308 6041.
E-mail addresses: [email protected] (M.A. Jeannot), [email protected] (A. Przyjazny), [email protected] (J.M. Kokosa).

0021-9673/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2009.10.089
 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336 2327

1. Brief history of development prehensive book detailing the theory and practice of SDME and
other solvent microextraction methods (including hollow fiber and
The use of a single drop as a collector of analytes in analytical dispersive liquid–liquid methods) is also recently available [14].
chemistry can be traced to the work of Dasgupta in the mid- The purpose and scope of this review will be to provide a useful
1990s. Dasgupta’s group first developed methods involving a liquid description of the development and advantages/disadvantages of
droplet as a gas sampling interface to extract substances such as the different modes of SDME, and to summarize new developments,
ammonia and sulfur dioxide from the air [1]. A silica capillary tube future trends and selected current applications. More exhaustive
was used to support a water droplet which was used to collect literature reviews on the subject have been presented elsewhere
gaseous analytes followed by in-line spectrophotometric analysis. [12–14]. Other “liquid-phase” microextraction techniques such
Dasgupta’s group subsequently developed a drop-in-drop minia- as dispersive liquid–liquid microextraction (DLLME), hollow fiber
turized solvent extraction system in which they extracted sodium techniques, and directly suspended droplet/solidification tech-
dodecylsulfate as an ion-pair with methylene blue into a microdrop niques are described in other articles in this issue and will not be
of chloroform [2]. They employed a peristaltic pump flow manifold addressed here.
and an optical fiber-based absorbance detector.
Cantwell’s group was the first to develop single drop microex- 2. Important theoretical aspects of SDME
traction techniques directly compatible with chromatographic
analysis. In their first paper [3], Jeannot and Cantwell used a Teflon 2.1. Overview
rod with a spherical recess to hold an 8-␮L drop of octane immersed
in a stirred aqueous solution. They called this approach “solvent Transport of analyte molecules from aqueous sample solution
microextraction” (SME). After extraction, the rod was removed, and to the microdrop is generally limited by slow diffusion rates of
a gas chromatography (GC) syringe was used to sample and inject the analyte molecules in the condensed (aqueous and/or organic)
a portion of the octane solution into a GC. In their second paper phases. While temperature and solvent viscosity play a role in these
[4], they demonstrated for the first time the direct use of the GC rates of diffusion, the primary mode of rate enhancement is reduc-
syringe needle for both suspension of the extracting solvent and tion of the distance over which the diffusion must occur. Thus,
injection into the GC. Stirring rate and stirring time were primar- samples are normally agitated via magnetic stirring, mechanical
ily investigated to develop equilibrium and kinetic models for the vibration or syringe plunger motion to increase the amount of con-
process. vective mixing or interfacial contact area, and therefore reduce the
He and Lee introduced the notion of using the GC syringe nee- diffusion distance. The time required to reach equilibrium in SDME
dle as a microseparatory funnel in a technique they called dynamic can be anywhere from seconds to hours, depending on the degree of
liquid-phase microextraction [5]. Rather than suspending the drop agitation, phase volumes, interfacial contact area and equilibrium
of organic solvent in the solution from the syringe, they contained distribution constant. Thus, to avoid excessive analysis times, SDME
the solvent within the syringe needle and drew aqueous phase into is often performed under non-equilibrium (kinetically controlled)
and out of the syringe repeatedly. This technique required high conditions.
precision repeated movement of the syringe plunger, but offered Even in cases where distribution equilibrium is attained in
improved organic solvent “drop” stability since the drop was pro- SDME, it is important to note that the extraction is rarely exhaus-
tected within the needle. tive. (A significant fraction of the total analyte normally remains
Ma and Cantwell successfully demonstrated three-phase single in the aqueous (sample) phase at equilibrium.) This is a conse-
drop microextraction with the use of an “unsupported” liquid mem- quence of the small organic (or receiver) to aqueous (sample)
brane and an aqueous microdrop of acceptor solution [6]. A Teflon volume ratio employed in SDME, similar to what is encountered in
ring near the top of the sample vial held the organic membrane solid-phase microextraction (SPME). In some cases, only a negligi-
phase in place, and an aqueous acceptor phase was suspended ble amount of analyte is removed from the sample solution, which
within the organic layer using a high-performance liquid chro- can be advantageous in studying speciation and avoiding perturba-
matography (HPLC) syringe. With this approach, large enrichment tion of sample-phase equilibria. In any case, whether equilibrium
factors could be realized using relative short extraction times for is attained or not, calibration is normally based on aqueous-phase
ionizable analytes. standards which are extracted under identical conditions to the
Liu and Lee developed a continuous flow microextraction unknown sample, with or without the aid of internal standards.
technique compatible with GC analysis [7]. Polyetheretherketone
(PEEK) tubing was used to continuously pass aqueous sample solu- 2.2. Two-phase SDME
tion through an extraction chamber, and a HPLC valve was used
to inject 1–5 ␮L of organic solvent into the flowing stream. After In a study of the uptake of gaseous ammonia by an aqueous
reaching the tubing outlet, the organic solvent drop remained drop [1], Liu and Dasgupta proposed a radial diffusion model in
attached to the PEEK tubing while aqueous phase continued to flow. which the drop was assumed to be spherical and stagnant. Experi-
A separate GC syringe was used to sample and inject a portion of mental data and visual observation suggested that convection was
the solvent phase after extraction. present within the drop, enhancing the rate of extraction. Diffusion
Extension of SDME to headspace (HS) analysis was developed of gaseous NH3 , establishment of equilibrium at the interface, and
in the early 2000s independently by Przyjazny et al. [8,9], Jeannot protonation of NH3 were all assumed to be fast relative to transport
and colleagues [10] and Vickackaite and colleagues [11]. Common of NH4 + within the aqueous drop.
high-boiling organic solvents such 1-octanol or n-hexadecane were Jeannot and Cantwell proposed a general model for equilibrium
found to be suitable for the determination of volatile or semivolatile and mass transfer in a two-phase liquid–liquid microextraction sys-
analytes. HS-SDME allowed for greater drop stability, avoided prob- tem [3,4]. The equilibrium concentration of analyte in the organic
lems of drop contamination or loss from “dirty” sample matrices, phase (Co,eq ) was shown to be
and in some cases provided for faster extraction rates compared to 0
KCw
direct immersion methods. Co,eq = KCw,eq = (1)
The early development of SDME has been reviewed in 2002 1 + KVo /Vw
[12], and a more recent review from 2007 focuses mainly on where K is the equilibrium distribution constant, Cw,eq is the equi-
applications during the first decade of development [13]. A com- 0 is the initial
librium concentration in the aqueous (water) phase, Cw
2328 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336

concentration in the water phase, and Vo and Vw are the organic and tioning into the air phase:
water phase volumes, respectively. (There is negligible depletion of 0
Kow Cw
the sample solution when the second term in the denominator is Co,eq = (3)
1 + (Kaw Va /Vw ) + (Kow Vo /Vw )
much smaller than 1.) A general kinetic model which fit experi-
mental data well showed that concentration (Co ) versus time (t) This equation incorporates both an air–water distribution constant
data follows the following first-order model: (Kaw ) and an overall organic–water distribution constant (Kow ). A
very small air (headspace) volume, Va , or very small Kaw reduces
Co = Co,eq (1 − e−kt ) (2) Eq. (3) to Eq. (1), the two-phase case.
Theis et al. [10] showed that headspace SDME rate data for
The rate constant, k, was shown to increase with increasing inter-
extraction of simple aromatic compounds followed the same first-
facial contact area, enhanced convection (particularly within the
order behavior as seen in two-phase systems (Eq. (2)). Furthermore,
aqueous phase), smaller volumes of both phases, and smaller equi-
it was apparent that enhanced extraction rates could be achieved
librium distribution constant. On the other hand, at equilibrium,
by increasing the degree of convection in both the aqueous solution
larger phase volumes and distribution constant result in more
and the organic drop.
moles in the organic phase, according to Eq. (1).
A kinetic calibration method was developed by Ouyang et al. for
Jeannot and Cantwell also applied SDME to the determination
headspace SDME [16]. They pre-loaded the extracting solvent with
of free (unbound) progesterone in the presence of a binding pro-
a known amount of a standard. During the extraction process, this
tein [15]. A small phase ratio ensured negligible depletion of free
standard was desorbed from the extracting drop while analyte was
progesterone from the sample solution, and the authors were able
absorbed. Matrix effects were automatically corrected since they
to accurately measure the binding constant of the steroid molecule
affected the desorption and adsorption processes similarly.
to bovine serum albumin. Transport of both free and bound forms
Mohammadi and Alizadeh described some equilibrium and
of progesterone to the interface also resulted in an enhanced rate
kinetic aspects of “dynamic headspace organic solvent film
of extraction.
microextraction” in 2006 [17]. In this technique, a portion of the
headspace is repeatedly drawn into and out of the needle which
2.3. Dynamic SDME
contains organic solvent. When the headspace is drawn into the
needle, a film of organic solvent forms along the inside walls of the
In the dynamic technique [5], He and Lee assumed instan-
needle, and analytes can equilibrate between the headspace and
taneous equilibrium between the aspirated circulating aqueous
this solvent film. They showed a linear dependence of the extrac-
sample plug and the film of organic solvent on the inside wall of
tion efficiency on the number of sampling cycles, and also showed
the syringe needle with each aspiration cycle. Extraction into the
the importance of fast plunger motion in maximizing the rate of
organic “plug” within the needle was assumed to be negligible com-
extraction.
pared with the film. With each new aspiration cycle, it was assumed
Fiamegos and Stalikas studied some theoretical aspects of in-
that the aqueous plug and organic film mixed fully with the respec-
drop derivatization and mass transfer in headspace SDME using
tive bulk solutions. A linear relationship between amount of analyte
low-volatility aldehydes [18]. They found that the chemical reac-
extracted and the number of aspiration cycles resulted.
tion rate was either slower than, or comparable to the diffusion rate
into the drop. Also, they determined that diffusion within the drop
2.4. SDME with back-extraction (aqueous–organic–aqueous) was the slow mass transfer step.
A steady-state kinetic model for headspace SDME was proposed
SDME with back-extraction is a two-step process in which an by Schnobrich and Jeannot in 2008 [19]. Benzene, toluene, ethyl-
ionizable solute is first extracted into an organic layer, followed benzene and xylene (BTEX) were used as model analytes, and it was
by extraction and trapping into a second aqueous layer whose pH shown that the headspace moles remained low and steady, par-
results in ionization of the solute. For example, Ma and Cantwell ticularly for the analytes with the largest octanol–water partition
[6] started with amine solutes in pH 13 buffer which ensured they coefficients. Neither the water–air nor the air–organic distribution
were deprotonated (neutral) and extractable into the organic layer. process was at equilibrium during the course of the extraction sug-
The second aqueous (receiver) drop was buffered at pH 2.1 to pro- gesting that neither of these processes alone is rate-limiting for
tonate (ionize) and therefore trap the amine in the receiver drop. these relatively volatile compounds. Thus, it is important to gener-
The enrichment factor was maximized by using as small a receiver ate convection in both the water and organic phases if possible.
drop volume as possible compared to the donor phase volume, and
followed similar first-order kinetics as Eq. (2). In their model, Ma 3. A closer look at various SDME modes
and Cantwell assumed steady-state behavior of the analyte in the
intermediate (organic) phase. At present, there are seven different modes of solvent microex-
traction that fall under the category of single drop microextraction.
2.5. Headspace SDME (aqueous–headspace–organic) They can be classified into either two-phase or three-phase
techniques, depending on the number of phases co-existing at equi-
In headspace SDME (or direct immersion two-phase SDME librium. This classification is depicted in Fig. 1. Two-phase modes
with a headspace above the sample), the headspace is always include direct immersion (DI), continuous flow (CF), drop-to-drop
a “compartment” for analyte molecules. For maximum sensitiv- (DD), and directly suspended droplet (DSD), while three-phase
ity, therefore, the headspace volume should always be minimized modes consist of headspace (HS), liquid–liquid–liquid (LLL), and a
to reduce the amount of analyte present there. At equilibrium, combination of LLL and DSD first introduced just recently [20]. The
it makes no difference how the three phases are “arranged” (i.e. frequency of use of various SDME modes, shown in Fig. 2, is almost
whether the solvent drop is immersed in the aqueous solution or evenly divided between two-phase (DI, CF, DD and DSD at 52%) and
in the headspace), but the kinetics of the extraction process and three-phase (HS and LLL at 48%) modes. By far the most commonly
other processes (e.g. rate of solvent evaporation) are significantly used modes of single drop microextraction are headspace (41% of
affected. all described SDME procedures) and direct immersion (38%), most
The equilibrium concentration in the organic drop for headspace likely due to their simplicity and inexpensive equipment needed
SDME is given by a modified version of Eq. (2) that includes parti- for implementation, but also because they were the first solvent
 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336 2329

Fig. 1. Single drop microextraction (SDME) classification.

microextraction procedures described in the literature. The other was further automated, controlling plunger movement with a vari-
five modes have found limited use, either as a result of additional able speed motor, by Saraji [23] and by Mohammadi and Alizadeh
equipment required, such as a pump (CF), or applicability limited [17], by controlling the motor movement with computer software.
to a small group of analytes (e.g. LLLME is used mostly for ion- Full automation of both the exposed-drop and unexposed-drop
izable compounds), or because they do not offer any significant methods was finally achieved by Ouyang et al. using a commer-
advantages compared to the more common modes. cial computer-interfaced autosampler to control solvent uptake,
In order to improve the rate of mass transfer, headspace and plunger speed, dwell time, and syringe injection [24].
direct immersion SDME can be performed in a dynamic mode, The two most common SDME modes – direct immersion and
in which not only the donor phase (sample), but also the accep- headspace solvent microextraction – have somewhat different gen-
tor phase (extracting solvent) is in motion (see Sections 2.3 and eral fields of applicability, although there are groups of analytes that
4.6 for more detail). Two variants of dynamic SDME are used, have been determined by using both these techniques as a sample
unexposed-drop and an exposed-drop. In the unexposed-drop (or preparation step. Since in direct immersion SDME the microdrop
in-syringe) method, the solvent is withdrawn, along with 1–3 ␮L of an extracting solvent is in direct contact with an aqueous sam-
of sample liquid or headspace, into the syringe, held for a spec- ple (Fig. 3), the solvent must be immiscible with water, which
ified time (dwell time), and the sample expelled. This process is implies the use of nonpolar or very slightly polar solvents. An
repeated for 30–90 cycles. The extract is then analyzed. In the exception to this rule is the use of ionic liquids as extracting sol-
exposed-drop method, the extracting solvent drop is exposed to vents (see Section 5). Consequently, this mode is best suited for
the sample at the needle tip for a specified time and then with- the separation/enrichment of nonpolar or moderately polar volatile
drawn into the needle, held for a specified time, and expelled out to and semivolatile analytes from relatively clean matrices, such as
the needle tip again. The sample is not withdrawn into the syringe, tap water or groundwater. Since volatile compounds are best
however. The unexposed-drop method was first developed by He preconcentrated by headspace SDME, the preferred use of direct
and Lee, using manual manipulation of the syringe plunger move- immersion mode is for semivolatile compounds. Examples include
ment [5,21]. This was followed by the use of a syringe pump to organochlorine pesticides [25–31], phthalates [32–35], or drugs
improve reproducibility of the plunger movement [22]. The method [36–45]. In general, the extracting solvent used in direct immersion
SDME is volatile, e.g. hexane or toluene, which makes this mode
directly compatible with gas chromatography. Consequently, GC
has been the predominant final determination technique used in
conjunction with direct immersion SDME, accounting for over 62%
of analytical procedures described in the literature.
Other final determination methods are also employed. For
example, HPLC (over 21% of DI-SDME analytical procedures) can

Fig. 3. Direct immersion (DI) SDME. Reprinted from: [13], Copyright (2007), with
Fig. 2. Frequency of use of various SDME modes. permission from Elsevier.
2330 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336

there are virtually no restrictions on extracting solvents used other

than low volatility. Thus, examples of analytes often extracted by
headspace SDME include trihalomethanes [50–53], BTEX hydrocar-
bons [9,10,16,17,33,54–56], volatile organic compounds [57–77],
and inorganic and organometallic species [33,78–87], the last
group often being derivatized prior to extraction. Headspace
mode is often applied to extract polar volatile compounds,
such as aldehydes [18,88–93], following or concurrently with
their derivatization. At the same time, HS-SDME has been
used to extract such semivolatile compounds as polycyclic aro-
matic hydrocarbons [33,76,94–96], polychlorinated biphenyls [76],
phenols [33,97–101], and chlorophenols [97,98]. Both nonpolar
and polar extracting solvents are common; the latter include
ionic liquids [33,50,51,54,94,99,101–105] and aqueous solutions
[47,48,69,81–85,96,98,100,106–109] or even pure water [110]. The
use of water-based extracting solvents in headspace SDME is inter-
esting since it totally eliminates the use of organic solvents. In
addition to sample cleanup, analyte enrichment is also possible
with pH control, analogous to the extraction with back-extraction
approach in LLLME.
The most popular extracting solvents in headspace SDME are
1-octanol, hexadecane, dodecane, and decane. Since headspace
SDME is a three-phase system, equilibration times may in some
cases be longer than in direct SDME as a result of two equilib-
ria involved: sample-headspace and headspace-extracting solvent.
However, extraction times in headspace SDME can be reduced sub-
Fig. 4. Headspace (HS) SDME including solvent cooling. Reprinted from: [166], stantially by increasing the headspace capacity, i.e. the amount of
Copyright (2004), with permission from Elsevier.
analyte contained in the headspace. The headspace capacity, which
is equal to the product of headspace (air) volume, Va , and air–water
be used for the analysis of polar semivolatiles, such as phenols. In distribution constant, Kaw , can be maximized by increasing either
this case, however, solvent replacement may be necessary unless Kaw or Va value, or both. If the amount of analyte extracted into the
an ionic liquid was used as the extracting solvent. Solvent replace- organic phase is small compared to headspace capacity (less than
ment involves gentle evaporation of the original extracting solvent, 5%), then analyte extraction takes place almost exclusively from
followed by redissolving the residue in a solvent compatible with the headspace. This results in rapid extraction, taking only several
the HPLC mobile phase or in the mobile phase itself. minutes, since the diffusion coefficients in the gaseous phase are
Another final determination technique that is being increasingly much larger than those in the liquid phase (by about four orders of
used in combination with direct immersion SDME is atmospheric magnitude).
pressure matrix-assisted laser desorption/ionization mass spec- The most common final determination technique used in com-
trometry (AP-MALDI-MS). If direct immersion SDME is employed bination with headspace SDME is by far gas chromatography, which
for the separation/enrichment of inorganic species, such as metal accounts for over 75% of all analytical procedures incorporating
ions, following their derivatization, then atomic absorption spec- HS-SDME. High-performance liquid chromatography is a distant
trometry or inductively coupled plasma-mass spectrometry are second (close to 10%), with atomic absorption spectrometry and
often used for their final determination. The major advantages of capillary electrophoresis accounting for 5% and 3.5%, respectively.
direct immersion SDME are simplicity of equipment used, at least In its simplest implementation, headspace SDME uses the same
in its static version, and low cost. In the simplest implementation, setup as direct immersion mode except that the microdrop of
the equipment used includes an extraction vial with a septum cap, organic solvent hanging from the tip of a microsyringe is not
a stir bar, a magnetic stirrer, a microsyringe, and a small volume of immersed into an aqueous sample, but remains in the headspace
extracting solvent. Disadvantages of DI-SDME are mostly related to above the sample. This setup has been modified in some procedures
the ease of dislodgment of the microdrop hanging from the tip of the by using temperature control of sample and extracting solvent.
microsyringe needle during the extraction process, which limits the In order to accelerate the rate of mass transfer of analytes from
rate of agitation of sample solution and the type of sample matrix to the sample to the headspace and to increase the amount of ana-
relatively clean (no solid particles present). Typical stirring rates in lytes transferred to the headspace, it is desirable to raise sample
direct immersion SDME are no more than 1000 rpm unless specially temperature. At the same time, however, elevated temperatures
modified needle tips are employed [46], which allows higher stir- tend to decrease the organic solvent-headspace distribution con-
ring rates, up to 1700 rpm. The need for vigorous agitation and/or stant, resulting in lower sensitivity of the determination. The loss of
dynamic mode of extraction is a consequence of slow mass transfer sensitivity can be avoided if the extracting solvent is cooled while
in liquid–liquid systems due to small diffusion coefficients in liq- the sample is heated (Fig. 4). However, this approach significantly
uids. This slow mass transfer results in longer extraction times in complicates the experimental setup; therefore, it should be used
DI-SDME compared to other single drop microextraction modes. only for ultra trace analyses or for highly volatile analytes with low
Headspace SDME (Fig. 4) is the sample preparation method of solvent-headspace distribution constants.
choice for volatile and semivolatile compounds, both polar and Drop-to-drop solvent microextraction [111–115] is a miniatur-
nonpolar. Sample matrices that are complex and/or dirty or con- ized version of direct immersion SDME. It was first introduced by
tain solids do not interfere with analyte separation/enrichment. In Wu in 2006 [115]. In this mode, both the sample and organic sol-
addition to liquid samples (typically aqueous matrices), gaseous vent volumes are in the order of microliters (Fig. 5). This approach
[47,48] and solid matrices [49] are also amenable to this mode. is recommended if available sample volumes are small (e.g. blood).
The scope of HS-SDME includes a wide variety of analytes, since Drop-to-drop microextraction has two prominent features. Firstly,
 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336 2331

Fig. 5. Drop-to-drop (DD) SDME. Reprinted with permission from: [115], Copyright
2006 American Chemical Society.

Fig. 7. Liquid–liquid–liquid (LLL) microextraction. Reprinted from: [167], Copyright

as a result of small sample and solvent volumes, equilibrium
(2008), with permission from Elsevier.
between the sample and the solvent is established quickly due to a
large value of the rate constant, k (see Section 2.2). Consequently,
the sample does not have to be stirred, which further simplifies and rapid establishment of equilibrium between the sample and
the experimental setup. Secondly, because the phase ratio, Vo /Vaq , the extracting solvent. Most procedures making use of continu-
is relatively large (see Eq. (1)), the enrichment factor is small, so ous flow microextraction are limited to extraction of nonpolar or
that the main advantage of this mode, other than the small sample slightly polar semivolatiles, such as pesticides [117,118], polycyclic
volume, is selectivity, which is provided by an extensive sample aromatic hydrocarbons [123], or aromatic compounds [7,120–122],
cleanup. Typical applications of drop-to-drop solvent microextrac- owing to the fact that only nonpolar extracting solvents are stable in
tion include extraction of trimeprazine from 8 ␮L of urine and blood the flowing system and the extent of their dissolution in the flowing
of rats using 0.6 ␮L of toluene [111], and extraction of quinine from sample is small. The second shortcoming of this mode is the need
30-␮L samples of urine and plasma with 2 ␮L of m-xylene [114]. for additional equipment, such as a microinfusion pump. Finally, a
Another two-phase SDME mode that deserves mention is con- direct comparison of continuous flow and static direct immersion
tinuous flow microextraction [7,116–123], in which a drop of SDME has proved the latter to yield superior detection limits and
solvent fully and continuously makes contact with a fresh and flow- precision [120,124].
ing sample solution (Fig. 6). The drop can be held at the tip of PEEK Liquid–liquid–liquid microextraction is a three-phase mode
tubing which is immersed in a continuously flowing sample in the which is best suited for the extraction of hydrophilic organic com-
extraction chamber. Alternatively, a microsyringe can be used to pounds, mostly polar semivolatiles, such as phenols, fatty acids or
hold a microdrop of the extracting solvent. The presence of both amines. It is a miniaturized form of extraction with back-extraction.
diffusion and convection results in the high extraction efficiency In this mode, analytes are extracted from an aqueous sample to
an organic solvent and simultaneously back-extracted from the
organic solvent to the acceptor solution, usually a few microliters of
an aqueous solution at the appropriate pH (Fig. 7). The organic sol-
vent is therefore an interface between the two aqueous solutions.
In order to achieve analyte isolation and enrichment, the acid-base
properties of the analytes are used. For acidic analytes, the pH of
the donor solution (sample) is adjusted to a low value so that ion-
ization of the analytes is suppressed and they can be extracted as
neutral species into the organic solvent. At the same time, the pH
of the acceptor solution is maintained at a high value to promote
ionization of the analytes. This way, the analytes are converted into
ionic species which are excluded from the liquid organic membrane
and therefore accumulate in the acceptor solution. In practical
implementation, a Teflon ring [125] or a small volumetric flask are
used for LLLME, which makes the experimental setup very simple.
An organic solvent forms a layer on top of sample and a micro-
drop of the acceptor solution is immersed into the organic solvent
layer. An organic solvent used in liquid–liquid–liquid microextrac-
tion must be immiscible with water and have a density lower
than water. Since the extract in LLLME is an aqueous solution, this
Fig. 6. Continuous flow (CF) SDME. Reprinted from: [123], Copyright (2007), with mode is directly compatible with reverse-phase HPLC and capillary
permission from Elsevier. electrophoresis, and these two techniques of final determination
2332 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336

have been used exclusively in analytical procedures. Most common appropriate for the chromatographic system. It needs to have a high
applications of LLLME include extraction of drugs from physiolog- enough viscosity to cling onto the tip of a syringe needle, but not
ical fluids or water [6,126–130] and aromatic amines or phenols so viscous that the diffusion rate of analyte into the drop affects
from water [125,131–132]. extraction time significantly. The intermolecular attraction char-
A recent modification of liquid–liquid–liquid microextraction acteristics of the solvent must also be compatible with the analyte
avoids the use of a microsyringe as supporting device [20]. Instead, being extracted. The most important interaction types are London
a large aqueous droplet is freely suspended at the top-center posi- dispersion forces (van der Waals forces), permanent dipole–dipole
tion of a layer of immiscible organic solvent, which is placed on top interactions, and hydrogen bonding. Thus, 1-octanol has been a
of a stirred aqueous sample. According to the authors, this configu- popular solvent for SDME, since it has all three interaction capabil-
ration improves mass transfer and results in a reduced equilibration ities. Furthermore, it has a relatively high-boiling point, relatively
time. low water solubility and moderate viscosity. Traditional separatory
funnel extraction solvents, such as diethyl ether, methylene chlo-
4. Experimental parameters affecting DI-SDME and ride, chloroform and ethyl acetate, on the other hand, are not useful
HS-SDME as extracting solvents for SDME because they are too volatile and
too water-soluble. These types of solvents have been used success-
There are to date approximately 600 research and application fully in a limited number of cases, however, as solvent modifiers,
papers dealing with solvent microextraction, and the single drop added to nonpolar, water-immiscible solvents such as toluene [32].
modes account for more than half of these publications [14]. With This is a fruitful area for future research.
this large available database, it is possible to identify the important Compatibility of an extracting solvent with the analytical
parameters that affect the rates and efficiencies of SDME extrac- method is also crucial. As indicated in Section 3, the most widely
tions. Each of the factors listed below will be briefly addressed used instrumentation for SDME extract analysis is GC. In general,
here. volatile analytes are extracted with higher boiling solvents, such as
tetradecane or 1-octanol. The extract must be injected using an inlet
split of 10/1 to 50/1 to obtain sharp, resolved peaks. Semivolatile
1. Analyte properties (volatility, polarity and ionization)
analytes are extracted with lower boiling solvents, such as o-xylene,
2. Extraction solvent properties
using splitless injection.
3. Extraction solvent purity
If reverse-phase HPLC is used, samples extracted with water-
4. Syringe
immiscible solvents such as toluene, must be exchanged or diluted
5. Drop volume
with a solvent compatible with HPLC such as acetonitrile [118,134].
6. Agitation
As an alternative, an ionic liquid [94,103,104,135] or water contain-
7. Ionic strength (salting out effect)
ing a modifier solvent can be used with reverse-phase HPLC [96] or
8. Temperature
capillary electrophoresis [100] directly. A nonpolar solvent could
9. Sample volume and headspace volume
also be used directly for normal phase HPLC.
10. Automation
4.3. Extracting solvent purity
4.1. Analyte properties
Solvent purity is one of the most important factors in SDME,
As discussed in preceding sections, the properties of the analyte especially when analyzing very dilute solutions. Commercial high
and the matrix it is in will determine whether direct immersion (DI- purity solvents may contain impurities that will interfere with
SDME) or headspace (HS-SDME) extraction is appropriate. Thus, analyte analysis. These impurities may include solvent analogs,
one must consider the volatility (boiling point), ionization (for acids such as xylene present in toluene, or oxidation products, such as
and bases) and polarity of the analyte and matrix. These properties aldehydes and alcohols in decane. It may be necessary, for trace
in turn affect two very important parameters, the organic extract- analysis, to doubly vacuum-distill a solvent and then store it in a
ing solvent/water distribution constant (Kow ) and the air/water freezer. Standard high grade solvents may be appropriate for higher
distribution constant (Kaw ). A detailed discussion of these impor- concentration solutions or if impurities do not interfere with the
tant parameters can be found in the literature [133]. HS-SDME is analysis. Trace impurities can be useful, however, since they can be
appropriate for most polar and nonpolar, lower molecular weight, used as internal standards in the solvent. The concentration of the
volatile and semivolatile compounds. Direct immersion (DI-SDME) internal standard does not have to be known precisely, just present
extraction is appropriate for nonpolar or moderately polar higher at a constant level for a series of analyses. The standard, which can
molecular weight, semivolatile chemicals. Highly polar chemicals also be added to the solvent, is used to monitor the integrity of the
may need to be derivatized to ensure recovery, especially when the drop. This is especially important when using an autosampler, to
matrix is aqueous. Examples of typical applications for HS and DI ensure that the drop is not lost during sampling. The standard can
were described in Section 3. also be used to account for small sample-to-sample variations in
drop size and solvent wicking on the needle.
4.2. Solvent properties
4.4. Syringe
A common misconception is that there are only a limited num-
ber of useful solvents that can be used for solvent microextraction. The most effective syringe for SDME is a standard GC microsy-
In fact, more than 2 dozen solvents have been successfully used for ringe. The drop must cling to the tip, without wicking up the
various SME modes, and this does not include solvent combinations exterior of the syringe needle. This requires a maximum needle
and solvents such as water with extraction enhancers (complexing tip surface area. The standard Hamilton #2 curved bevel syringe
agents, derivatizing agents and pH control) [14]. There are some tip provides the greatest surface area, and approximately 90–95%
important restrictions on the selection of a particular extracting of the drop can be withdrawn into the syringe following the
solvent, however. When extracting from an aqueous solution, the extraction. A straight edge bevel GC syringe allows only 80–85%
solvent needs to be water immiscible. The solvent needs to have withdrawal and an HPLC syringe very little withdrawal unless a
a boiling point high enough that it will not evaporate, but also drop size less than 0.5 ␮L is used. If HPLC analysis is used, the drop
 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336 2333

must be diluted with additional solvent in a sample vial and/or 4.7. Ionic strength (salting out effect)
exchanged with an HPLC-compatible solvent and then an HPLC
syringe used for injection. Salting out is a time-tested technique for increasing extrac-
tion efficiency, especially for moderately polar and low molecular
weight volatile chemicals. High ionic strength can also decrease the
4.5. Drop volume
solubility of the extracting solvent. The effect of ionic strength on
analyte water solubility and thus the Kow and the Henry’s constant
As indicated in the theory section, the amount of analyte
(Kaw ) is exponential. It is thus best to accurately weigh added salt
extracted increases with drop volume. Unfortunately, a maximum
and to use a near, but not saturated concentration of salt. Saturated
drop volume for a standard syringe needle is 2-3 ␮L. A drop size
salt solutions may contain undissolved particles which can dislodge
larger than 3 ␮L is unstable and the drop may fall off the needle,
the drop in DI-SDME. If halide exchange is a concern, anhydrous
especially when using direct immersion SDME. (The hollow fiber
sodium sulfate may be used in the same weight/volume concen-
approach, which is outside the scope of this review, is discussed
trations as sodium chloride. Sodium sulfate must be used with
elsewhere in this issue, and is one solution to this problem.) It
caution, however, since at high concentrations it can crystallize as
must be remembered that even high-boiling organic solvents have
the hydrate.
some volatility and many water-immiscible solvents actually have
Adding salt is not beneficial for increased extraction of nonpo-
finite water solubility. As a consequence, a portion of the drop will
lar semivolatile analytes, such as the PAHs, which have Kow values
evaporate and/or dissolve in the sample matrix, especially when
greater than 1000. Addition of salt may, however, be useful for
elevated extraction temperatures, long extraction times and vigor-
minimizing drop loss when DI-SDME is used.
ous agitation of the sample are used. When using direct immersion
There are a few reported cases where addition of salt is detri-
extraction, the sample may contain salts or soluble macromolecules
mental in DI-SDME [28,137,138]. This may be due to changes in the
which would be harmful to the analytical instrumentation used. If
viscosity or surface tension of the water sample.
a 1-␮L drop was used and 1 ␮L of liquid were withdrawn into the
syringe, chances are good that the solvent would be contaminated
4.8. Temperature
with the sample matrix. Even when using headspace SDME, the
volume of the solvent withdrawn into the syringe may be variable,
As previously indicated, temperature control is important in
due to loss of solvent by evaporation/solubility and wicking onto
SDME extractions, especially for headspace extractions. The organic
the surface of the needle. Difficulties with drop size variations and
solvent/water distribution constant (Kow ) is only weakly affected
solvent wicking are minimized if the drop size used is 0.2–0.5 ␮L
by temperature, but the air/water constant (Kaw ) and the organic
larger than the amount of solvent withdrawn into the syringe fol-
solvent/air constant (Koa ) are strongly dependent on temperature.
lowing extraction. At high extraction temperatures (50–80 ◦ C), it
Typically, for nonpolar analytes, increasing the temperature of the
may be necessary to increase this value.
water solution (or solid matrix) increases headspace concentra-
tion. However, some moderately polar analytes become much more
4.6. Agitation water-soluble at elevated temperatures and headspace concentra-
tions therefore decrease with increasing temperature. In addition,
As indicated earlier, sample agitation is important for reduc- solubility in the extracting solvent may decrease with increasing
ing extraction time. Three sample agitation methods are available: temperatures, decreasing the efficiency of extraction. Therefore,
stirring, vibration and vortexing. Stirring, using a magnetic stir a compromise extraction temperature must be found, especially
bar, is effective with stirring rates of 300–600 rpm for DI-SDME when extracting samples containing multiple analytes, and extrac-
and 500–1000 rpm for HS-SDME. The limitations of higher stir- tion time must be minimized to decrease the temperature effect
ring rates are the dislodgement of the drop by the sample solution on the drop. One solution to the decreased solubility of analyte
or splashing when using headspace. Vibration and vortex stirring, in the drop at elevated temperatures is to use a chilled syringe
used with some autosamplers, are also effective, with the limita- needle, which can be accomplished with a laboratory built [52] or
tion that the agitation cannot occur while the drop is exposed at commercial device.
the needle tip. Sample agitation before extraction can lead to repro-
ducible results, especially for HS-SDME, but these techniques are 4.9. Sample volume and headspace volume
most effective when a computer-controlled autosampler is used,
along with dynamic sampling, so the solvent is contained within Many researchers have employed relatively large (5–30 mL)
the syringe needle during the agitation. volumes of aqueous sample and large headspace volumes (rang-
Agitation of the extracting solvent can also lead to decreased ing up to 80% of the vial volume). SDME theory clearly shows
extraction time (though not extraction efficiency), since a rate- that this can be counterproductive, since the maximum amount
limiting step in the extraction process is often transfer of the analyte of analyte extracted is dependent on the Kow and Kaw values
from the surface into the bulk of the drop. This is facilitated by con- and larger samples require longer extraction times. In general,
tinuous renewal of the drop surface. As discussed in Section 2.2, aqueous sample volumes of 1–4 mL are optimal for analytes with
stirring the sample during direct immersion SDME may result in Kow values less than 1000. Analytes with large Kow values, such
convection within the drop, but dynamic extraction has also been as the halogenated pesticides and the PAHs, will yield greater
shown to significantly decrease extraction time [22]. In dynamic extracted amounts with increased sample sizes up to 30–40 mL.
extraction, reproducibility and extraction efficiency depend on Larger volumes, of course, will require much longer extraction
exact repetition of several factors: number of cycles, sample vol- times.
ume drawn into the syringe, extracting solvent volume, plunger Theoretical calculations further indicate that the headspace, for
speed, dwell time at maximum plunger withdrawal, and exposure both DI-SDME and HS-SDME, should be kept to a practical mini-
time (or dwell time) for the exposed drop in contact with the sam- mum to maximize extraction efficiency. For a 2-mL vial, a sample
ple. Plunger movement must be precise and at an optimal rate size of 1–1.5 mL and a headspace of 0.5–1 mL are appropriate. For a
for 30–90 repetitions. While this can be done manually or with 4-mL vial a sample size of 3 mL and a headspace of 1 mL are appro-
a mechanical device, accurate repetitive sampling requires the use priate. The headspace should be no larger than necessary to allow
of a computer-controlled autosampler [136]. the drop to be suspended over the stirred sample for HS-SDME
2334 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336

and only enough to avoid sample contact with the septum cap for
DI-SDME.

4.10. Automation

Very good accuracy and reproducibility can be achieved by

a skilled analyst using manual DI-SDME and HS-SDME extrac-
tion. However, when analyzing large numbers of samples or
using dynamic extraction, a computer-controlled autosampler is
a necessity. The autosampler can perform all steps of DI-SDME
and HS-SDME, including agitation, temperature control, syringe
plunger movement, cleaning and injection, with accuracy and
reproducibility approaching that of a skilled analyst. Several auto-
mated procedures involving direct immersion [24,41,136] and
headspace SDME [16,17,24,136] have been developed.

5. New developments and trends

A new era for solvent microextraction began in 2003 with the

introduction of ionic liquids as extracting solvents [94]. Since then,
a considerable number of published analytical procedures made
use of ionic liquids, including direct immersion [33,135,139–146]
and headspace SDME [33,50,51,54,99,101–105]. Ionic liquids have
many unique properties, such as negligible vapor pressure, excel-
lent thermal stability, and high viscosity, which allow the use of
stable large drops, thus increasing extraction yield. Their polarity is
adjustable through selection of the appropriate cations and anions.
Consequently, their miscibility with water and organic solvents, Fig. 8. Interface used for ionic liquid based SDME combined with gas chromatogra-
viscosity, and extractability of organic analytes are tunable, which phy. Reprinted from: [54], Copyright (2008), with permission from Elsevier.

accounts for their versatility and may result in a much larger use in
the future. Until recently, the use of ionic liquids in analytical proce- added. The heating of the water present in plant cells results in their
dures involving solvent microextraction has been limited to HPLC, disruption and release of essential oils that are evaporated by the
capillary electrophoresis, and spectroscopic techniques as the final water of the plant material and extracted by headspace SDME. A
determination methods as a result of their nonvolatility. However, cooling system outside the microwave oven condenses the distil-
some recent papers describe several approaches which make ionic late. Compared to traditional alternatives, microwave distillation
liquids compatible with gas chromatography. One of them involves offers reduced extraction time and energy savings.
the use of a removable interface enabling direct introduction of The applicability of SDME has been further extended by con-
the ionic liquid extract into a GC–MS system while preventing the verting those analytes, which cannot be directly extracted by
ionic liquid from entering the column [50,51,54,139] (Fig. 8). The SDME, into extractable species through derivatization reactions.
second approach employs a commercially available thermal des- These include, among others, inorganic species (metal ions, anions)
orption system to thermally desorb analytes from ionic liquids and and highly polar volatile organic compounds. Derivatization reac-
introduce them into the GC [102]. In the third approach, analytes are tions employed in solvent microextraction have been reviewed
desorbed from the ionic liquid in the injection port of gas chromato- recently [150,151], and are discussed in a monograph on solvent
graph, and the ionic liquid is then drawn back into the microsyringe microextraction published in October 2009 [14]. The extension of
[101]. These approaches may significantly extend the applicability analytical procedures involving SDME to inorganic species seems
of ionic liquids in single drop microextraction. to be particularly attractive. Examples include derivatization fol-
So far, the two major areas of application of solvent microextrac- lowed by microextraction of heavy metal ions by direct immersion
tion have been environmental (61%) and clinical & forensic (21%) [143–145,152–156], continuous flow [157], or headspace SDME
analysis. The scope of applications has been recently extended to [85]; microextraction of derivatized inorganic anions, such as
include more solid samples, particularly plants and their parts, periodate, iodate, bromate, iodide, bromide, cyanide, and sul-
by the addition of a preliminary step prior to microextraction fide [107,109,158–159], or neutral species, including iodine, nitric
proper. A novel technique, called hydrodistillation–headspace sol- oxide, chlorine, and ammonia [78,160–162].
vent microextraction, used primarily to isolate essential oils from Every new analytical procedure involving solvent microex-
plants and their parts, couples water extraction with solvent traction has to be optimized by adjusting extraction parameters,
microextraction [62,63,67,68,147–149]. In this technique, a small including sample volume, headspace volume (in three-phase
amount of plant material (typically 0.7–4 g) is mixed with water mode), organic solvent type and volume, agitation conditions, tem-
and subjected to hydrodistillation. A microdrop of a high-boiling perature, pH, extraction time, and sample ionic strength, in such a
solvent is suspended in the headspace of a hydrodistilling sample. way as to maximize extraction yield. Most method development
This arrangement results in a short extraction time (10–20 min, procedures described in the literature have used the one-variable-
including refluxing), and consumes a small amount of plant mate- at-a-time approach, where just one parameter is varied and all the
rial. other parameters are kept constant. This approach is inefficient
Microwave distillation is another novel preliminary step suit- and requires a large number of experiments. Recently, how-
able for samples of plant material, combining microwave heating ever, a number of solvent microextraction procedures have been
and dry distillation at atmospheric pressure [60]. Plant material is optimized using experimental design. The predominant designs
placed in a microwave reactor without any water or organic solvent involved simultaneous design making use of response surface
 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336 2335

methodology (RSM). Prior to applying the RSM, screening exper- ment are available. In its simplest implementation, manual direct
iments were carried out to find which experimental variables immersion or headspace mode, no equipment is needed other than
influenced the response significantly, with factorial designs being that already available in any analytical laboratory.
most common. Experimental design has been applied in opti-
mizing direct immersion [25,32,135,163] and headspace SDME
References
[47,83,104,126,149,163–165].
Solvent microextraction is a highly versatile sample prepara- [1] S. Liu, P.K. Dasgupta, Anal. Chem. 67 (1995) 2042.
tion method not only because it can be used for practically all [2] H. Liu, P.K. Dasgupta, Anal. Chem. 68 (1996) 1817.
classes of analytes, but also because it is compatible, directly or [3] M.A. Jeannot, F.F. Cantwell, Anal. Chem. 68 (1996) 2236.
[4] M.A. Jeannot, F.F. Cantwell, Anal. Chem. 69 (1997) 235.
after solvent replacement, with a wide range of final determina-
[5] Y. He, H.K. Lee, Anal. Chem. 69 (1997) 4634.
tion techniques. Gas chromatography is by far the most common [6] M. Ma, F.F. Cantwell, Anal. Chem. 71 (1999) 388.
final determination technique used in combination with single [7] W. Liu, H.K. Lee, Anal. Chem. 72 (2000) 4462.
[8] A. Przyjazny, J.F. Austin, A.T. Essenmacher, Proceedings of the 6th Polish Con-
drop microextraction, followed by high-performance liquid chro-
ference on Analytical Chemistry, vol. 2, Gliwice, Poland, July 9–14, 2000, p.
matography. GC is the preferred technique for the separation and 135.
determination of volatile and semivolatile analytes, while HPLC is [9] A. Przyjazny, J.M. Kokosa, J. Chromatogr. A 977 (2002) 143.
the method of choice in the assays involving nonvolatile analytes, [10] A.L. Theis, A.J. Waldack, S.M. Hansen, M.A. Jeannot, Anal. Chem. 73 (2001)
5651.
such as ionizable analytes. Gas chromatography is compatible with [11] A. Tankeviciute, R. Kazlauskas, V. Vickackaite, Analyst 126 (2001) 1674.
direct immersion SDME, in which nonpolar volatile organic sol- [12] E. Psillakis, N. Kalogerakis, Trends Anal. Chem. 21 (2002) 53.
vents are typically employed. HPLC and capillary electrophoresis [13] L. Xu, C. Basheer, H.K. Lee, J. Chromatogr. A 1152 (2007) 184.
[14] J.M. Kokosa, A. Przyjazny, M.A. Jeannot, Solvent Microextraction—Theory and
are directly compatible with liquid–liquid–liquid microextraction Practice, John Wiley & Sons, Inc., Hoboken, NJ, 2009.
and with headspace SDME using highly polar solvents, ionic liq- [15] M.A. Jeannot, F.F. Cantwell, Anal. Chem. 69 (1997) 2935.
uids or aqueous solutions. HPLC and CE can also be coupled [16] G. Ouyang, W. Zhao, J. Pawliszyn, Anal. Chem. 77 (2005) 8122.
[17] A. Mohammadi, N. Alizadeh, J. Chromatogr. A 1107 (2006) 19.
with direct immersion SDME following solvent replacement. For [18] Y.C. Fiamegos, C.D. Stalikas, Anal. Chim. Acta 599 (2007) 76.
organic analytes, atmospheric pressure matrix-assisted laser des- [19] C.R. Schnobrich, M.A. Jeannot, J. Chromatogr. A 1215 (2008) 30.
orption/ionization mass spectrometry has also been used with [20] A. Sarafraz-Yazdi, F. Mofazzeli, Z. Es’haghi, J. Chromatogr. A 1216 (2009) 5086.
[21] Y. Wang, Y.C. Kwok, Y. He, H.K. Lee, Anal. Chem. 70 (1998) 4610.
SDME, since the small extracting solvent volume is compatible with
[22] L. Hou, H.K. Lee, J. Chromatogr. A 976 (2002) 377.
MALDI-MS. [23] M. Saraji, J. Chromatogr. A 1062 (2005) 15.
Following solvent microextraction, inorganic analytes, such as [24] G. Ouyang, W. Zhao, J. Pawliszyn, J. Chromatogr. A 1138 (2007) 47.
[25] C. Cortada, L. Vidal, S. Tejada, A. Romo, A. Canals, Anal. Chim. Acta 638 (2009)
metal or metalloid ions, are typically detected by one of the
29.
spectroscopic techniques: UV/vis spectrophotometry, spectrofluo- [26] L. de Jager, A.R.J. Andrews, Analyst 125 (2000) 1943.
rimetry, electrothermal or flame atomic absorption spectrometry, [27] L.S. de Jager, A.R.J. Andrews, Chromatographia 50 (1999) 733.
and inductively coupled plasma (ICP)—optical emission spec- [28] M.C. López-Blanco, S. Blanco-Cid, B. Cancho-Grande, J. Simal-Gándara, J. Chro-
matogr. A 984 (2003) 245.
troscopy or ICP-mass spectrometry. Recently, miniaturized modes [29] L.-L. Qian, Y.Z. He, J. Chromatogr. A 1134 (2006) 32.
of two common spectroscopic techniques coupled with SDME have [30] M. Zhang, J. Huang, C. Wei, B. Yu, X. Yang, X. Chen, Talanta 74 (2008) 599.
been reported. The first one involved the determination of acid [31] L. Zhao, H.K. Lee, J. Chromatogr. A 919 (2001) 381.
[32] R. Batlle, C. Nerin, J. Chromatogr. A 1045 (2004) 29.
labile sulfide fraction in water by microvolume turbidimetry fol- [33] J.F. Liu, Y.G. Chi, G.B. Jiang, J. Sep. Sci. 28 (2005) 87.
lowing headspace SDME of hydrogen sulfide [109]. The second [34] J. Xu, P. Liang, T. Zhang, Anal. Chim. Acta 597 (2007) 1.
procedure used fiber optics-based cuvetteless CCD-array micro- [35] R. Zalieckaitė, E. Adomavičiūtė, V. Vičkačkaitė, Chemija (Vilnius) 18 (2007)
25.
spectrophotometry to determine thiols, chlorine, ammonia, and [36] C. Casari, A.R.J. Andrews, Forensic Sci. Int. 120 (2001) 165.
iodine following direct immersion or headspace SDME [78]. [37] M. Cruz-Vera, R. Lucena, S. Cárdenas, M. Valcárcel, J. Chromatogr. A 1202
(2008) 1.
[38] H. Fang, Z. Zeng, L. Liu, D. Pang, Anal. Chem. 78 (2006) 1257.
[39] P.K. Gupta, L. Manral, K. Ganesan, D.K. Dubey, Anal. Bioanal. Chem. 388 (2007)
6. Conclusions 579.
[40] M. Ma, S. Kang, Q. Zhao, B. Chen, S. Yao, J. Pharm. Biomed. Anal. 40 (2006)
128.
The trend toward miniaturization of sample preparation meth-
[41] S.W. Myung, S.H. Yoon, M. Kim, Analyst 128 (2003) 1443.
ods has resulted in the development of several techniques which [42] A. Sarafraz-Yazdi, S. Raouf-Yazdinejad, Z. Es’haghi, Chromatographia 66
may be described as solvent microextraction (SME). Single drop (2007) 613.
microextraction modes constitute a significant part of SME. They [43] R. Sekar, H.F. Wu, Anal. Chem. 78 (2006) 6306.
[44] K. Shrivas, H.F. Wu, Rapid Commun. Mass Spectrom. 21 (2007) 3103.
use much smaller volumes of organic solvents than classical [45] H.F. Wu, C.H. Lin, Rapid Commun. Mass Spectrom. 20 (2006) 2511.
liquid–liquid extraction, permit automation and higher sample [46] F. Ahmadi, Y. Assadi, S.M.R. Millani Hosseini, M. Rezaee, J. Chromatogr. A 1101
throughput, and provide high extraction efficiency. Several areas (2006) 307.
[47] R. Batlle, P. López, C. Nerín, C. Crescenzi, J. Chromatogr. A 1185 (2008) 155.
of growth in the number of SDME applications can be predicted, [48] J.P. Xie, S.H. Sun, H.-Y. Wang, Y.L. Zong, C. Nie, Y.L. Guo, Rapid Commun. Mass
including environmental, clinical and forensic. Further commer- Spectrom. 20 (2006) 2573.
cialization and automation of SDME equipment and procedures [49] E. Hansson, M. Hakkarainen, J. Chromatogr. A 1102 (2006) 91.
[50] E. Aguilera-Herrador, R. Lucena, S. Cárdenas, M. Valcárcel, J. Chromatogr. A
is likely. The advantages of automated SDME procedures include 1209 (2008) 76.
improved precision, increased sample throughput due to reduction [51] E. Aguilera-Herrador, R. Lucena, S. Cárdenas, M. Valcárcel, J. Chromatogr. A
of extraction times when using dynamic SDME, and unattended 1216 (2009) 5580.
[52] Y. Yamini, M.H. Hosseini, M. Hojaty, J. Arab, J. Chromatogr. Sci. 42 (2004) 32.
operation.
[53] R.S. Zhao, W.J. Lao, X.B. Xu, Talanta 62 (2004) 751.
A more expanded use of ionic liquids can be anticipated, owing [54] E. Aguilera-Herrador, R. Lucena, S. Cárdenas, M. Valcárcel, J. Chromatogr. A
to their unique properties and compatibility with a variety of final 1201 (2008) 106.
[55] M. Kaykhaii, M. Moradi, J. Chromatogr. Sci. 46 (2008) 413.
determination techniques. Ionic liquids can prove especially valu-
[56] J.M. Kokosa, A. Przyjazny, J. Chromatogr. A 983 (2003) 205.
able for the extraction of strongly polar analytes. [57] A. Besharati-Seidani, A. Jabbari, Y. Yamini, Anal. Chim. Acta 530 (2005) 155.
Single drop microextraction can find use in most sample prepa- [58] A. Besharati-Seidani, A. Jabbari, Y. Yamini, M.J. Saharkhiz, Flavour Fragr. J. 21
ration procedures, with analytes ranging from volatile organic (2006) 502.
[59] J. Cao, M. Qi, Y. Zhang, S. Zhou, Q. Shao, R. Fu, Anal. Chim. Acta 561 (2006) 88.
compounds, through polar and nonpolar semivolatiles, to ionic [60] C. Deng, Y. Mao, F. Hu, X. Zhang, J. Chromatogr. A 1152 (2007) 193.
compounds and metal ions, as long as suitable solvents and equip- [61] L. Fang, M. Qi, T. Li, Q. Shao, R. Fu, J. Pharm. Biomed. Anal. 41 (2006) 791.
2336 M.A. Jeannot et al. / J. Chromatogr. A 1217 (2010) 2326–2336

[62] P. Hashemi, M.M. Abolghasemi, A.R. Fakhari, S.N. Ebrahimi, S. Ahmadi, Chro- [114] K. Shrivas, H.F. Wu, Anal. Chim. Acta 605 (2007) 153.
matographia 66 (2007) 283. [115] H.F. Wu, J.H. Yen, C.C. Chin, Anal. Chem. 78 (2006) 1707.
[63] M. Jalali-Heravi, H. Sereshti, J. Chromatogr. A 1160 (2007) 81. [116] X. Chen, T. Zhang, P. Liang, Y. Li, Microchim. Acta 155 (2006) 415.
[64] M.J. Jung, Y.J. Shin, S.Y. Oh, N.S. Kim, K. Kim, D.S. Lee, Bull. Korean Chem. Soc. [117] L. Guo, P. Liang, T. Zhang, Y. Liu, S. Liu, Chromatographia 61 (2005) 523.
27 (2006) 231. [118] Y. He, H.K. Lee, J. Chromatogr. A 1122 (2006) 7.
[65] X. Li, X. Xu, X. Wang, L. Ma, Int. J. Environ. Anal. Chem. 84 (2004) 633. [119] Y. Li, T. Zhang, P. Liang, Anal. Chim. Acta 536 (2005) 245.
[66] P. López, C. Sánchez, R. Batlle, C. Nerín, J. Agric. Food. Chem. 55 (2007) 4348. [120] X. Liu, X. Chen, S. Yang, X. Wang, J. Sep. Sci. 30 (2007) 2506.
[67] P. Salehi, B. Asghari, F. Mohammadi, Chromatographia 65 (2007) 119. [121] X. Liu, X. Wang, X. Chen, S. Yang, Chromatographia 65 (2007) 447.
[68] P. Salehi, A.R. Fakhari, S.N. Ebrahimi, R. Heydari, Flavour Fragr. J. 22 (2007) [122] X.J. Liu, X.W. Chen, S. Yang, X.D. Wang, Bull. Environ. Contam. Toxicol. 78
280. (2007) 368.
[69] S. Sun, Z. Cheng, J. Xie, J. Zhang, Y. Liao, H. Wang, Y. Guo, Rapid Commun. Mass [123] Y. Liu, Y. Hashi, J.-M. Lin, Anal. Chim. Acta 585 (2007) 294.
Spectrom. 19 (2005) 1025. [124] Q. Xiao, B. Hu, C. Yu, L. Xia, Z. Jiang, Talanta 69 (2006) 848.
[70] S.H. Sun, J.P. Xie, F.W. Xie, Y.L. Zong, J. Chromatogr. A 1179 (2008) 89. [125] L. Zhu, C.B. Tay, H.K. Lee, J. Chromatogr. A 963 (2002) 231.
[71] G. Wang, C. Dong, Y.A. Sun, K. Xie, H. Zheng, J. Chromatogr. Sci. 46 (2008) 127. [126] H. Bagheri, F. Khalilian, E. Babanezhad, A. Es-haghi, M.R. Rouini, Anal. Chim.
[72] G.-Q. Wang, R.-J. Zhang, Y.-A. Sun, K. Xie, C.-Y. Ma, Chromatographia 65 (2007) Acta 610 (2008) 211.
363. [127] H. Ebrahimzadeh, Y. Yamini, A. Gholizade, A. Sedighi, S. Kasraee, Anal. Chim.
[73] X. Wang, T. Jiang, J. Yuan, C. Cheng, J. Liu, J. Shi, R. Zhao, Anal. Bioanal. Chem. Acta 626 (2008) 193.
385 (2006) 1082. [128] H. Ebrahimzadeh, Y. Yamini, A. Sedighi, M.R. Rouini, J. Chromatogr. B 863
[74] D.C. Wood, J.M. Miller, I. Christ, R.E. Majors, LC–GC Eur. 17 (2004) 573. (2008) 229.
[75] Q. Xiao, C. Yu, J. Xing, B. Hu, J. Chromatogr. A 1125 (2006) 133. [129] Y. He, Y.J. Kang, J. Chromatogr. A 1133 (2006) 35.
[76] X. Yan, C. Yang, C. Ren, D. Li, J. Chromatogr. A 1205 (2008) 182. [130] M. Ma, F.F. Cantwell, Anal. Chem. 70 (1998) 3912.
[77] T. Zhang, X. Chen, Y. Li, P. Liang, Chromatographia 63 (2006) 633. [131] A.S. Yazdi, Z. Es’haghi, Talanta 66 (2005) 664.
[78] N. Sharma, A.K.K.V. Pillai, N. Pathak, A. Jain, K.K. Verma, Anal. Chim. Acta 648 [132] L. Zhao, H.K. Lee, J. Chromatogr. A 931 (2001) 95.
(2009) 183. [133] R.P. Schwarzenbach, P.K. Gschwend, D.M. Imboden, Environmental Organic
[79] M. Chamsaz, M.H. Arbab-Zawar, S. Nazari, J. Anal. Atom. Spectrom. 18 (2003) Chemistry, John Wiley & Sons, 2002.
1279. [134] A. Sarafraz-Yazdi, Z. Es’haghi, Chromatographia 63 (2006) 563.
[80] V. Colombini, C. Bancon-Montigny, L. Yang, P. Maxwell, R.E. Sturgeon, Z. [135] L. Vidal, A. Canals, A. Salvador, J. Chromatogr. A 1174 (2007) 95.
Mester, Talanta 63 (2004) 555. [136] J.M. Kokosa, US Patent 7,178,414 B1 (February 20, 2007).
[81] R. Figueroa, M. García, I. Lavilla, C. Bendicho, Spectrochim. Acta B 60 (2005) [137] E. Psillakis, N. Kalogerakis, J. Chromatogr. A 907 (2001) 211.
1556. [138] T. Zhang, X. Chen, P. Liang, C. Liu, J. Chromatogr. Sci. 44 (2006) 619.
[82] S. Fragueiro, I. Lavilla, C. Bendicho, Spectrochim. Acta B 59 (2004) 851. [139] E. Aguilera-Herrador, R. Lucena, S. Cárdenas, M. Valcárcel, Anal. Chem. 80
[83] S. Fragueiro, I. Lavilla, C. Bendicho, Talanta 68 (2006) 1096. (2008) 793.
[84] S. Gil, S. Fragueiro, I. Lavilla, C. Bendicho, Spectrochim. Acta B 60 (2005) 145. [140] X. Fu, S. Dai, Y. Zhang, Int. J. Environ. Anal. Chem. 86 (2006) 985.
[85] P. Hashemi, A. Rahimi, A.R. Ghiasvand, M.M. Abolghasemi, J. Braz. Chem. Soc. [141] J. Liu, J.F. Peng, Y.G. Chi, G.B. Jiang, Talanta 65 (2005) 705.
18 (2007) 1145. [142] J.F. Liu, Y.G. Chi, G.B. Jiang, C. Tai, J.F. Peng, J.T. Hu, J. Chromatogr. A 1026 (2004)
[86] F. Pena-Pereira, I. Lavilla, C. Bendicho, Anal. Chim. Acta 631 (2009) 223. 143.
[87] F.J.P. Pereira, C. Bendicho, N. Kalogerakis, E. Psillakis, Talanta 74 (2007) 47. [143] J.L. Manzoori, M. Amjadi, J. Abulhassani, Talanta 77 (2009) 1539.
[88] C. Deng, N. Yao, N. Li, X. Zhang, J. Sep. Sci. 28 (2005) 2301. [144] J.L. Manzoori, M. Amjadi, J. Abulhassani, Anal. Chim. Acta 644 (2009) 48.
[89] Y.C. Fiamegos, C.D. Stalikas, Anal. Chim. Acta 609 (2008) 175. [145] F. Pena-Pereira, I. Lavilla, C. Bendicho, L. Vidal, A. Canals, Talanta 78 (2009)
[90] N. Li, C. Deng, N. Yao, X. Shen, X. Zhang, Anal. Chim. Acta 540 (2005) 317. 537.
[91] N. Li, C. Deng, X. Yin, N. Yao, X. Shen, X. Zhang, Anal. Biochem. 342 (2005) 318. [146] C. Yao, W.R. Pitner, J.L. Anderson, Anal. Chem. 81 (2009) 5054.
[92] A.K.K.V. Pillai, K. Gautam, A. Jain, K.K. Verma, Anal. Chim. Acta 632 (2009) 208. [147] A.R. Fakhari, P. Salehi, R. Heydari, S.N. Ebrahimi, P.R. Haddad, J. Chromatogr.
[93] J. Xie, J. Yin, S. Sun, F. Xie, X. Zhang, Y. Guo, Anal. Chim. Acta 638 (2009) 198. A 1098 (2005) 14.
[94] J.F. Liu, G.B. Jiang, Y.G. Chi, Y.-Q. Cai, Q.X. Zhou, J.T. Hu, Anal. Chem. 75 (2003) [148] M.B. Gholivand, M. Rahimi-Nasrabadi, H. Chalabi, Anal. Lett. 42 (2009) 1382.
5870. [149] V. Kiyanpour, A.R. Fakhari, R. Alizadeh, B. Asghari, M. Jalali-Heravi, Talanta 79
[95] S. Shariati-Feizabadi, Y. Yamini, N. Bahramifar, Anal. Chim. Acta 489 (2003) (2009) 695.
21. [150] C.D. Stalikas, Y.C. Fiamegos, Trends Anal. Chem. 27 (2008) 533.
[96] Y. Wu, L. Xia, R. Chen, B. Hu, Talanta 74 (2008) 470. [151] L. Xu, C. Basheer, H.K. Lee, J. Chromatogr. A 1216 (2009) 701.
[97] Y.A. Shi, M.Z. Chen, S. Muniraj, J.F. Jen, J. Chromatogr. A 1207 (2008) 130. [152] N. Goudarzi, J. Agric. Food Chem. 57 (2009) 1099.
[98] H. Xu, Y. Liao, J. Yao, J. Chromatogr. A 1167 (2007) 1. [153] L. Li, B. Hu, L. Xia, Z. Jiang, Talanta 70 (2006) 468.
[99] C. Ye, Q. Zhou, X. Wang, J. Xiao, J. Sep. Sci. 30 (2007) 42. [154] M.Y. Lin, C.W. Whang, J. Chromatogr. A 1160 (2007) 336.
[100] J. Zhang, T. Su, H.K. Lee, Anal. Chem. 77 (2005) 1988. [155] H.F. Maltez, D.L.G. Borges, E. Carasek, B. Welz, A.J. Curtius, Talanta 74 (2008)
[101] F.Q. Zhao, J. Li, B.Z. Zeng, J. Sep. Sci. 31 (2008) 3045. 800.
[102] A. Chisvert, I.P. Román, L. Vidal, A. Canals, J. Chromatogr. A 1216 (2009) 1290. [156] D. Verma, S.K. Verma, M.K. Deb, Talanta 78 (2009) 270.
[103] J.-F. Peng, J.-F. Liu, G.-B. Jiang, C. Tai, M.-J. Huang, J. Chromatogr. A 1072 (2005) [157] A.N. Anthemidis, I.S.I. Adam, Anal. Chim. Acta 632 (2009) 216.
3. [158] M. Gupta, A. Jain, K.K. Verma, Talanta 71 (2007) 1039.
[104] L. Vidal, E. Psillakis, C.E. Domini, N. Grané, F. Marken, A. Canals, Anal. Chim. [159] K. Reddy-Noone, A. Jain, K.K. Verma, J. Chromatogr. A 1148 (2007) 145.
Acta 584 (2007) 189. [160] B. Pranaitytė, S. Jermak, E. Naujalis, A. Padarauskas, Microchem. J. 86 (2007)
[105] C.-L. Ye, Q.-X. Zhou, X.-M. Wang, Anal. Chim. Acta 572 (2006) 165. 48.
[106] Y. He, A. Vargas, Y.-J. Kang, Anal. Chim. Acta 589 (2007) 225. [161] P. Das, M. Gupta, A. Jain, K.K. Verma, J. Chromatogr. A 1023 (2004) 33.
[107] S. Jermak, B. Pranaitytė, A. Padarauskas, Electrophoresis 27 (2006) 4538. [162] K.J. Huang, H. Wang, M. Ma, M.L. Sha, H.S. Zhang, J. Chromatogr. A 1103 (2006)
[108] S. Jermak, B. Pranaitytė, A. Padarauskas, J. Chromatogr. A 1148 (2007) 123. 193.
[109] I. Lavilla, F. Pena-Pereira, S. Gil, M. Costas, C. Bendicho, Anal. Chim. Acta 647 [163] J. Romero, P. López, C. Rubio, R. Batlle, C. Nerín, J. Chromatogr. A 1166 (2007)
(2009) 112. 24.
[110] A.Y. Nazarenko, Am. Lab. 36 (2004) 30. [164] L. Tan, X.P. Zhao, X.Q. Liu, H.X. Ju, J.S. Li, Chromatographia 62 (2005) 305.
[111] K. Agrawal, H.F. Wu, Rapid Commun. Mass Spectrom. 21 (2007) 3352. [165] L. Vidal, C.E. Domini, N. Grané, E. Psillakis, A. Canals, Anal. Chim. Acta 592
[112] N.J. Petersen, H. Jensen, S.H. Hansen, K.E. Rasmussen, S. Pedersen-Bjergaard, (2007) 9.
J. Chromatogr. A 1216 (2009) 1496. [166] Y. Yamini, M. Hojjati, M. Haji-Hosseini, M. Shamsipur, Talanta 62 (2004) 265.
[113] K. Shrivas, H.F. Wu, J. Chromatogr. A 1170 (2007) 9. [167] Z. Fan, X. Liu, J. Chromatogr. A 1180 (2008) 187.