Coacervation of surface-functionalized polymerized vesicles derived from ammonium bromide surfactants. Application to the selective speciation of chromium in environmental samples

Article abstract of DOI:10.1021/ac802018w

The potential of polymerized vesicle coacervates made up of ammonium bromide surfactants for the extraction of metallic ions from natural waters was examined for the first time. Linear linked polymerized vesicles prepared by UV excitation of (4-carboxyben

Full text of DOI:10.1021/ac802018w

Anal. Chem. 2008, 80, 9787–9796
Coacervation of Surface-Functionalized
Polymerized Vesicles Derived from Ammonium
Bromide Surfactants. Application to the Selective
Speciation of Chromium in Environmental Samples
Nikolaos I. Kapakoglou, Dimosthenis L. Giokas,* George Z. Tsogas, and Athanasios G. Vlessidis
Department of Chemistry, University of Ioannina, 45110, Ioannina, Greece
The potential of polymerized vesicle coacervates made up
of ammonium bromide surfactants for the extraction of
metallic ions from natural waters was examined for the
first time. Linear linked polymerized vesicles prepared by
UV excitation of (4-carboxybenzyl)bis[2-(10-undecenoy-
loxy)ethyl]methylammonium bromide monomer were char-
acterized, and several factors affecting their phase behav-
ior were investigated. Evidently, the permeation of metallic
elements through the polymeric membrane was found to
be sensitive to ionic radius excluding ions larger than the
interbilayer space of the vesicle assembly. To this effect,
Cr3+ ions could selectively diffuse through the polymeric
membrane. Optimization of vesicle structure and surface
charge were the regulating parameters in exploiting this
unique feature toward the analytical speciation of Cr
species in natural waters. Detection limits as low as 0.1
µg L-1 were achieved by preconcentrating only 10 mL of
sample volume with recoveries in the range of 97.0-
105.5% and very good reproducibility (RSD ) 1.51%).
ecules predominates the applications of coacervation in analytical
extraction processes. A possible explanation can be sought in the
difficulty of obtaining functional coacervate phases that offer both
distinct phase separation and adequate partitioning of the analytes.
In this context, alkyltrimethylammonium micelle-based coacer-
vates in the presence of high electrolyte concentration (e.g., 400 g
L
-1 NaCl),7 acid-induced coacervation of alkyl sulfates, sulfonates,
and sulfoccinates,8 lamellar phases between anionic surfactants
and alkaline earth metals,9 alkylcarboxylic acids under the addition
of tetrabutylammonium salt,10 coacervates based on alkanoic
(C8-C16) and alkenoic (C18) acid reverse micelles and tetrahy-
drofuran11 have been reported as means for accomplishing solvent-
free extractions based on coacervation. Depending on the system
used, analyte partition is controlled either by hydrophobic interac-
tions in the nonpolar vesicular core or by electrostatic interactions
with the hydrophilic head group or in some cases both.
In the vastness of available literature regarding the coacerva-
tion phenomenon and vesicular media, the polymerized vesicle
chemistry in analytical separations has been overlooked. Several
studies have shown that many surfactants that have unsaturation
in their fatty chains can be polymerized under a variety of
conditions placing the unsaturated bonds of adjacent molecules
in appropriate proximity for orderly polymerization.12 Free radical
initiation causes a chain reaction that results in linking of the
surfactant molecules at the locations of the unsaturation. In most
cases, the macroscopic order of the system is not disrupted by
polymerization, which yields a microscopic “plastic bag”.12 These
molecular sacs have three basic features that render them highly
attractive for analytical separations. The first is that polymerized
membranes are less permeable to ions than those of unpolymer-
ized membranes; therefore, diffusional input or loss of ions may
be minimized. Second, the surface of a polymerized vesicle can
be chemically modified (functionalized) to promote attachment
of various compounds; and third, the surface charge of the vesicles
is easily controlled by varying either the composition of the
reaction mixture or the relative amounts of three types of
Solventless extraction constitutes today one of the most
challenging issues in contemporary chemical analysis with a
growing number of publications dedicated to the development,
modification, and optimization of systems that minimize or
alleviate the use of organic solvents in extraction. Among the most
popular liquid-phase microextraction and single-drop microextrac-
tion,1,2 coacervation has only recently attracted the interest of
researchers as means for obtaining solvent-free extractions.
Up to date, despite the plethora of molecules that can give
rise to coacervate phases (lipids, proteins, polysaccharides,
synthetic polymers, organized assemblies, etc.),3-6 the use of a
dehydrating agent (salt, nonsolvent, pH, or temperature) that
promotes the interactions among neutral or charged macromol-
* To whom correspondence should be addressed. Phone: +30-26510-98400.
Fax: +30-26510-98781. E-mail: dgiokas@cc.uoi.gr.
(1) Lee, J.; Lee, H. K.; Rasmussen, K. E.; Pedersen-Bjergaard, S. Anal. Chim.
Acta 2008, 624, 253–268
(2) Xu, L.; Basheer, C.; Lee, H. K. J. Chromatogr., A 2007, 1152, 184–192
(3) Ishii, F.; Takamura, A.; Ishigami, Y. Langmuir 1995, 11, 483–486
(4) Ringsdorf, H.; Schlarb, B.; Venzmer, J. Angew. Chem., Int. Ed. Engl. 1998,
27, 113–158
(5) Saegusa, K.; Ishii, F. Langmuir 2002, 18, 5984–5988
(6) Gander, B.; Blanco-Prieto, M. J.; Thomasin, C.; Wandrey, Ch.; Hunkeler,
D. Coacervation/Phase Separation. In Encyclopedia of Pharmaceutical
Technology; Marcel Dekker: New York, 2002.
.
(7) Jin, Z.; Zhu, M.; Conte, E. D. Anal. Chem. 1999, 71, 514–517
(8) Casero, I.; Sicilia, D.; Rubio, S.; Pe´rez-Bendito, D. Anal. Chem. 1999, 71,
4519–4526
(9) Giokas, D. L.; Tsogas, G. Z.; Vlessidis, A. G.; Karayannis, M. I. Anal. Chem.
2004, 76, 1302–1309
.
.
.
.
.
.
.
(10) Ruiz, F. J.; Rubio, S.; Per`ez-Bendito, D. Anal. Chem. 2006, 78, 7229–7239
(11) Ruiz, F. J.; Rubio, S.; Per`ez-Bendito, D. Anal. Chem. 2007, 79, 7473–7484
(12) Guo, C. Y.; Shankar, R. R.; Cai, S. H.; Wu, J. Q.; Abe, S.; Thomas, R. N.;
.
.
Kup, J. E. Langmuir 1992, 8, 815–823
.
10.1021/ac802018w CCC: $40.75 2008 American Chemical Society
Published on Web 11/14/2008
Analytical Chemistry, Vol. 80, No. 24, December 15, 2008 9787

Figure 1. Schematic representation of the polymer structure formed from linear linked 4-CBUAB monomers tethered in the aliphatic chain via
CHdCH* reactive groups and potential conformation of head groups.
monofunctional monomers (of negative, neutral, or positive
charge) thus enabling a better control of analyte/polymerized
vesicle interactions.
radius higher than the interbilayer space offers a unique and stable
microenvironment for accomplishing the selective and reagentless
isolation of Cr3+ species. The interaction of the polymeric
membrane with water-soluble and sparingly soluble chelating
agents is also assessed and discussed. From an analytical point
of view, this is the first study reporting on the application of
polymerized vesicles for the coacervate-based extraction of metal
ions from aqueous matrixes.
Despite their high versatility, there is a dearth of information
regarding the analytical potential of polymerized vesicles. Several
studies have demonstrated the ability of metal-sorbing vesicles
(MSVs) that mimic most closely the functional and structural
properties of cell membranes, for metal uptake by chemical
modification of polymerized vesicles. This is accomplished by
carrier-mediated transport of metal ions across the polymerized
vesicle bilayer in the presence of a lipophilic metal carrier, which
serves as the metal ion carrier through the lipid bilayer.13,14
However, MSVs are limited by the fact that they will not take up
metal ions for which the carrier has no affinity and will have no
capacity for metal ions that the chelator does not bind. Neverthe-
less, yet although some systems have been shown to meet both
criteria, no analytical data regarding these applications have been
reported so far.
In view of the lack of published work on the analytical utility
of surface functionalized polymerized vesicles, this work describes
the first analytical application of polymerized vesicular phases.
Parameters related to the formation of the vesicular coacervate
phase and its suitability for the extraction of metal ions from
environmental samples is evaluated. Evidently, the limited perme-
ability of the polymerized membranes to metallic ions with ionic
EXPERIMENTAL SECTION
Reagents. All reagents were of analytical grade and were
employed as purchased, unless otherwise stated. Standard metal
solutions were prepared by dissolving appropriate amounts of
Cr(NO3)3 · 9H2O (Sigma-Aldrich, Athens, Greece, 99.99%) and
K2Cr2O7 (Sigma-Aldrich, min 99.5%) in doubly distilled water. 10-
Undecenoyl chloride (Sigma-Aldrich) N-methyliminobis(ethanol)
(Fluka Chemie AG, Switzerland), dimethylformamide (Merck,
Darmstadt, Germany), ethyl acetate (Merck), diethyl ether
(Merck), NaOH (Merck), CH2Cl2 (Merck), and R-bromo-p-toluic
acid (Fluka) used for the synthesis of (4-carboxybenzyl)bis[2-(10-
undecenoyloxy)ethyl]methylammonium bromide monomer, were
used without any further purification. Ammonium pyrrolidinedithio-
carbamate (APDC) obtained from Sigma (99.0%) was prepared in
doubly distilled water (10%, w/v) while 8-hydroxyquinoline (8-
HQ) was dissolved in 99.8% spectroscopic grade methanol (Carlo
Erba) due to its restricted solubility in water. Humic acid was
obtained from Fluka and used as purchased.
(13) Shamsai, B. M.; Monbouquette, H. G. J. Membr. Sci. 1997, 130, 173–181
(14) Stanish, I.; Monbouquette, H. G. J. Membr. Sci. 2000, 179, 127–136
.
.
9788 Analytical Chemistry, Vol. 80, No. 24, December 15, 2008

Synthesis of the (4-Carboxybenzyl)bis[2-(10-undecenoy-
loxy)ethyl]methylammonium Bromide Monomer (4-CBUAB)
(Figure 1). Polymerized vesicles were synthesized following the
procedure of Guo et al.12 Briefly, 10-undecenoyl chloride (44.7 g,
0.22 mol) was added to a solution of 11.9 g (0.10 mol) of
N-methyliminobis(ethanol) in 80 mL of DMF. The solution was
left to stand for 1 h and the product spontaneously crystallized.
Subsequently, 250 mL of diethyl ether was added and the mixture
was cooled at -10 °C and filtered. The product was recrystallized
with ethyl acetate, affording 43.8 g (90% yield) of the bis[2-(10-
undecenoyloxycarbonylethyl]methylamine hydrochloride (7HCl).
The hydrochloride salt (2.44 g, 5 mmol) was treated with 1 N
sodium hydroxide (30 mL) in methylene chloride (40 mL). After
solvent removal, the liquid amine that was produced was mixed
with R-bromo-p-toluic acid at a ratio of 1.00/1.25 mol. The reaction
system was then refluxed at 130 °C overnight under argon flow.
The product was purified using methylene chloride. The final
product was a yellowish viscous solid with a melting point between
55 and 62 °C, with a yield of ∼55%.
Characterization of the Polymerized Vesicular Phase.
Surfactant polymerization can occur when the unsaturated bonds
present in their fatty chains are linked together at the locations
of the unsaturation. The manner in which molecules link together
in polymer chains largely depends on the number and location of
reactive groups per monomer, as well as the mode of initiation.
In our study, surfactant monomers were photochemically excited
under UV radiation causing the formation of an activated CHdCH*
radical group, which is involved in the polymerization (Figure 1).
Since the examined monomers have their reactive groups at the
hydrophobic terminus of the amphiphilic tails, which renders
insufficient the cross-linking process, we assumed that polymer-
ization proceeds in a linear manner. The linkages among polymers
in the vesicle structure play an important role since linear
polymerizations cause modest changes in bilayer properties,
whereas cross-linking polymerizations can significantly decrease
the membrane bilayer fluidity and permeability.15 The polymer-
ization-induced decrease in membrane permeability is associated
with a parallel increase in the chemical stability of vesicles, which
is observed by resistance to dissolution by either surfactants or
organic solvents.16
Experimental evidence on the manner in which monomers are
linked together in the polymer chains (linear or cross-linked) was
pursued by the addition of Triton X-114 (0-2.5%, w/v) and
methanol (0-50%, v/v) in an aqueous solution of polymerized
vesicles (0.6%, w/v). The absorbance of the solutions was recorded
at 525 nm. Evidently, the presence of <20% (v/v) methanol
produced a rapid decrease in the absorbance signal, which
exceeds 50%. In a like manner, only 0.5% (w/v) Triton X-114 was
adequate to reduce the absorbance of the vesicular solution by
88.7%. These data show that polymerized vesicles are susceptible
to disassembly in the presence of low alcohol and nonionic
surfactant concentrations, suggesting the absence of cross-
linkages among the monomers.13
Coacervation of 4-Carboxybenzyl)bis[2-(10-undecenoy-
loxy)ethyl]methylammonium Bromide Polymerized Vesicles.
Coacervation of the polymerized vesicles occurred spontaneously
when the synthesized polymeric reagent was added to water,
yielding a cloudy (hazy) solution. To investigate the behavior of
the polymerized vesicular coacervates, four parameters were
considered, namely, vesicle concentration, temperature, ionic
strength, and pH, and phase diagrams were constructed. Both
temperature and ionic strength have been recognized for their
importance on the orientation of the polar head groups.17
Figure 2a shows the phase diagram of polymerized vesicle
concentration versus temperature in the absence of any ionic
strength and pH regulators. Three different regions, defined by
the existence of a vesicular suspension, a coacervate, and a
coexisting coacervate and solid phase. As can be inferred in Figure
2a, at vesicle concentrations below 1% (w/v), the coacervate phase
is attained within a relatively wide temperature range (0-60 °C).
At temperatures above 60 °C, vesicles start to disintegrate since
their melting point is below 62 °C.12 These data suggest that the
Formation of Vesicles. The formation of the vesicles was
performed in a 20-mL tube using a heat system ultrasonic
processor in continuous mode. The ultrasonication took 15 min
to transform the heterogeneous monomer into a clear, homog-
enized colloidal solution, colored weak smoky blue. After ultra-
sonication, the vesicle solution was transferred into a quartz tube.
Polymerization was carried out under UV irradiation at 450 W for
8-10 h.
Instrumentation. A Shimandzu AA-6800 graphite furnace
atomic absorption spectrophotometer (GFAAS) with hollow cath-
ode lamp operating at 10 mA was used throughout the measure-
ments, which were made at 357.90 nm. An adjustable-capillary
nebulizer and supplies of argon were used for the generation of
aerosols and atomization. The output signals were collected and
processed in the continuous peak height mode. Infrared (IR)
spectra of the polymerized vesicular coacervates dispersed in KBr
pellets were recorded on a Perkin-Elmer Spectrum GX FT-IR
spectrometer. Absorbance measurements were performed with
matched quartz cells of 1-cm path length in a Jenway 6405 UV/
vis spectrophotometer. A pH meter, WTW 552 model glass
electrode was employed for pH adjustment of the solutions.
Samples. Water samples were collected in glass bottles from
four rivers (Loudias, Aliakmon, Axios, and Edessaios) in central
Macedonia, Greece. The samples were filtered through a What-
man No. 40 (0.45 µm) filter to remove suspended solids and stored
in dark glass containers at 4 °C.
Analytical Procedure. In a typical extraction experiment, 10
mL of aqueous solution containing Cr species (Cr3+ and/or
CrO42-) in the range of 1-50 µg L-1 was spiked with KCl to adjust
the ionic strength of the solution. An appropriate amount of
4-CBUAB polymerized vesicle aqueous solution was added,
followed by the addition of 50 µL of HCl to adjust the pH of the
solution to the value of 4. The mixture was shaken and left to
stand for 15 min at 50 °C in a thermostatic water bath. Separation
of the phases was accomplished by centrifugation for 20 min at
4000 rpm. The bulk aqueous phase was decanted and the vesicular
phase was treated with a methanolic solution of 1 M HNO3 in
order to dissociate the vesicular structure. Twenty microliters of
the resulting solution was injected into the atomizer.
(15) O’Brien, D. F.; Armitage, B.; Benedicto, A.; Bennett, E. D.; Lamparski, H. G.;
Lee, Y. S.; Warunee, S.; Sisson, T. M. Acc. Chem. Res. 1998, 31, 861–868
(16) Sisson, T. M.; Lamparski, H. G.; Ko¨lchens, S.; Elyadi, A.; O’Brien, D. F.
Macromolecules 1996, 29, 8321–8329
(17) Makino, K.; Yamada, T.; Kimura, M.; Oka, T.; Oshima, H.; Kondo, T.
Biophys. Chem. 1991, 41, 175–183
.
.
.
Analytical Chemistry, Vol. 80, No. 24, December 15, 2008 9789

vesicular coacervates was wide enough for analytical application
pinpointing the fact that analyte extraction can be pursued over a
wide range of salinity levels depending on analyte properties.
To assess the influence of hydrogen or hydroxyl concentration
on phase separation, dilute HCl or NaOH was sequentially added
in aqueous solutions of 4-CBUAB polymerized vesicles (Figure
2c). It was observed that addition of HCl induces a gradual
increase in the absorbance of the solution, possibly due to
reduction of the apparent water solubility of the aggregates, thus
enhancing phase separation. On the other hand, the addition of
NaOH caused a decline in the absorbance values (compared to
pure aqueous solution), rendering phase separation insufficient
for extraction. However, the region encompassed by coacervate
phase as a function of HCl and NaOH is wide enough for analytical
application, suggesting that both acidic and basic species can be
extracted provided that adequate partitioning in the coacervate
phase occurs.
Scope and Extraction Mechanism of Polymerized Vesicle
Coacervates. The analytical utility of coacervate phases has
mainly been explored in the analysis of organic compounds and
less frequently for the analysis of metallic ions.8-11 In our previous
works, we have shown the ability of anionic coacervate phases to
extract metal ions from aquatic solutions based on either elec-
trostatic forces or hydrophobic interactions, or even both.9,18,19
To evaluate the ability of polymerized vesicles to extract metal
species, a multitude of metallic cations, namely, Cd2+, Pb2+, Co2+
,
Cu2+, Cr3+, Cr2O4-, Fe3+, Sb3+, and Al3+, were initially evaluated.
For that purpose, aquatic solutions of metallic ions in the range
of 1-20 µg L-1 were prepared and extracted according to the
aforementioned analytical procedure. The condensed vesicular
phase was dissolved in methanol/HNO3 1 M and analyzed by
graphite furnace atomic absorption spectrometry. Evidently, no
signal was recorded for the majority of the metallic species
investigated except for Cr3+and Fe3+, which were recovered at
79 and 12.5%, respectively.
Electrostatic interactions of metallic ions with the negatively
charged carboxylic acid reactive sites were initially suspected to
be the governing mechanism although it could not completely
account for the observed selectivity. To shed more light on the
coacervate-metal ion interactions, the IR spectrum of the polym-
erized vesicular coacervate before and after extraction was
examined. Cross-examination of both spectra reveals that there
are no variations in the carboxylate hydroxyl bands (2582 and
2447 nm) and RCH3N+ bands (1317, 1276, 1227, and 1171 nm).
Furthermore, the characteristic peaks of Cr-O bonds (832.9 nm)
or Cr-N bonds (1247 nm) are totally absent (Figure 3), suggest-
ing that no such interactions occur. Based on these findings,
complexation of Cr3+ ions by the polymerized vesicle was rejected
as a potential mechanism in the proposed extraction pattern.
Two were the leading elements in the interpretation of the
observed phenomenon. The first was the ability of the polymerized
coacervates to extract trivalent metal ions with small ionic radius
and the second the selectivity toward Cr3+ and not CrO42-. It is
generally assumed that bilayers have sufficiently low permeability
so that transport systems in the membrane can maintain suitable
gradients of ions or polar solutes against a substantial leak. The
Figure 2. Phase diagrams of 4-CBUAB aqueous polymerized vesicle
solutions as a function of (a) temperature, (b) ionic strength, and (c)
pH. Dotted lines stand for the phase separation boundaries between
coacervate and coacervate plus solid region.
region encompassed by the polymerized vesicular coacervates was
wide enough for analytical application.
The response of the polymerized vesicular solution to ionic
strength increase at room temperature is depicted in Figure 2b.
It is clear, that the region encompassed by the polymerized
9790 Analytical Chemistry, Vol. 80, No. 24, December 15, 2008

Figure 3. IR spectra of 4-CBUAB polymerized vesicles after the extraction of 5 µg L-1 Cr3+. Other experimental conditions as defined in the
text.
low permeability reflects the Born energy for the dissolution
(partitioning) of an ion or dipole in the low-dielectric hydrocarbon
interior, an energy term so large that it represents a nearly
insurmountable barrier.18 However, several studies have reported
that fluctuations in the bilayer structure induced by various factors
like temperature produce transient defects that allow solutes to
bypass the electrostatic and solvophobic energy barriers.20 More
specifically, thinner bilayers have many transient defects that allow
rapid permeation of small ionic species; whereas in thicker
bilayers, defects become so rare that partitioning mechanisms
dominate the ionic flux.21 We assumed that the gradual hydration
of vesicular bilayers leads to water filling the highly polar
interbilayer spaces.22 Consequently, there is an increase in its
bilayer repeat when fully hydrated at a temperature below that of
its main transition. This condition promoted the incorporation of
trivalent metal ions with small ionic radius (depending on the
radius of their respective hydrated spheres), through the highly
polar interbilayer space in the positions of the transient defects.
At the working conditions, Cr(OH)2+ is probably the predominant
species of Cr (∼0.75 Å),23,24 while iron (∼0.74 Å) is present as
bulky monomeric hydroxo complexes (i.e., Al(OH)4-).26,27 Bulkier
ions like Cu (∼0.87 Å), Zn (0.88 Å), Sb (∼0.90 Å) Cd (1.09 Å), Co
2-
(1.10Å), Pb (1.33 Å), and the tetrahedrically configured CrO4
anion (2.40 Å) can hardly cross the interbilayer space.
In order for this assumption to be valid, the charge of the
vesicle must be minimized since 4-CBUAB polymerized vesicles
are positively charged owing to the CH3N+ group. The require-
ment for minimum surface change is imposed by the effective
charge of the metallic ions. At the same time, this minimum
positive charge acts synergistically to the low permeability of the
2-
bulky CrO4 anions by minimizing electrostatic attraction with
the also bulky CH3N+ group. As will be discussed below in more
detail, our experimental data for optimum extraction are in good
agreement with previous observations concerning the orientation
of bilayer head groups as a function of temperature and ionic
strength.17 Although the reasons why the direction of the head
group changes depending on temperature and ionic strength are
not well understood, it has been proven that certain temperature
and ionic strength conditions cause the head groups to bend, thus
altering the external surface charge (Figure 1). Under certain
experimental conditions, the surface charge can be minimized,
and further discussion will be afforded below in more detail.
Fe(OH)2+ and to a lesser extent as Fe(OH)2+ 25
. In a like manner,
although Al3+ has the smallest ionic radius of the trivalent metal
ions examined (∼0.67 Å), its effective diameter of hydrated ions
is similar to that of chromium and iron and probably exists as its
RESULTS AND DISCUSSION
Effect of Experimental Variables on Chromium Specia-
tion. The effect of experimental variables expected to influence
the extraction efficiency of chromium by polymerized vesicle
coacervates was assessed. Parameters such as salt type and
concentration (0-1 M), temperature (0-70 °C), vesicle concentra-
tion (0-3%, w/v), pH (1-9), and reaction times (0-30 min) were
investigated in the range of values specified in parenthesis and
the results are presented below.
(18) Tsogas, Z. G.; Giokas, L. D.; Vlessidis, G. A.; Evmiridis, P. N. Spectrochim.
Acta, Part B 2004, 59, 957–965
(19) Tsogas, Z. G.; Giokas, L. D.; Paleologos, K. E.; Vlessidis, G. A.; Evmiridis,
P. N. Anal. Chim. Acta 2004, 537, 239–248
(20) Parsegian, A. Nature 1969, 221, 844–847
(21) Volkov, A. G.; S. Paula, S.; Deamer, D. W. Bioelectrochem. Bioenerg. 1997,
42, 153–160
(22) Suwalskya, M.; Castro, R.; Villena, F.; Sotomayor, C. P. J. Indian Biochem.
2008, 102, 842–849
(23) Sperling, M.; Xu, S.; Welz, B. Anal. Chem. 1992, 64, 3101–3108
(24) Rai, D.; Sass, B. M.; Moore, D. A. Inorg. Chem. 1987, 26, 345–349
(25) Flynn, C. M. Chem. Rev. 1984, 84, 31–41
.
.
.
.
(26) IUPAC Technical report. Analytical methodology for the determination of
aluminium fractions in natural fresh waters. Pure Appl. Chem. 1996, 68,
1597-1638.
.
.
.
(27) Tria, J.; Butler, E. C. V.; Haddad, P. R.; Bowie, A. R. Anal. Chim. Acta 2007,
.
558, 153–165.
Analytical Chemistry, Vol. 80, No. 24, December 15, 2008 9791

Figure 4. Effect of polymerized vesicle concentration on the
extraction of 1 µg L-1 Cr3+: pH 5, [KCl] ) 0.25 M, room temperature.
Figure 5. Performance of the method as a function of solution pH:
[Cr3+] ) 1 µg L-1, [4-CBUAB] ) 0.6% (w/v), [KCl] ) 0.25 M, room
temperature.
is slightly lower than this of the p-benzyl carboxylate. However,
the pKa of carboxylic acid functionalities in a monomer inserted
in an aggregate (pKaaggr) is higher than that of carboxylic acid
functionalities in the free monomer (pKamono), and this increase
depends on the structure of the monomer.28 At pH values below
pKaaggr, carboxylates in the hydrocarbon chain are protonated first
thus either increasing the electrostatic repulsion or creating
hydrogen bonds between adjacent molecules. As the pH ap-
proaches the value of pKaaggr, these carboxyls are deprotonated
faster than the benzyl carboxylate (due to their slightly lower pKa
value) and possibly forming an internal salt with the CH3N+ group.
As pH is increased above pKaaggr, carboxylates are deprotonated,
which alters the vesicle charge. Moreover, some interaction
between the COO- and CH3N+ groups of adjacent molecules
cannot be excluded although the hydrocarbon chain (-CH2-)
between the benzyl and CH3N+ groups is rather small. All these
changes alter the surface charge of the bilayers and disrupt vesicle
structure, which influences membrane thickness, and thus alter
ion permeability, which is also affected by the hydrated form of
Cr as a function of pH. As pH increases above 5, Cr(OH)2+ is
transformed to Cr(OH)2+ and Cr(OH)3, Cr(OH)4- at alkaline and
Effect of Polymerized Vesicle Concentration. The extraction
efficiency of Cr3+ ions was significantly influenced by the
concentration of polymerized vesicles. The results of Figure 4
indicate an upward increase in the absorbance signal up to the
concentration of 2% (w/v) with an interim peak at 0.75% (w/v).
However, as concentration increased above 0.75% (w/v), the
coacervate phase started to resist precipitation possibly due to
repulsion among the positively charged polymerized vesicles that
swell resisting precipitation. Similar phenomena have been
reported in our previous work using anionic vesicular media.9 To
accomplish adequate precipitation, multiple or prolonged (above
1.5 h) centrifugations were needed as vesicle concentration
gradually surpassed the value of 1% (w/v). Owing to this
phenomenon, the interim value of 0.6% (w/v) was selected as the
working value for the extraction of chromium. Although some
loss of sensitivity was observed at this concentration, the phase
separation process was easily accomplished with a single cen-
trifugation step, which simplified and accelerated the experimental
procedure.
Effect of pH. The influence of hydrogen or hydroxyl concen-
tration was the next parameter investigated. Although no com-
plexation is involved in the extraction mechanism, pH may effect
the vesicle structure through protonation/deprotonation of the
carboxylates present in the benzyl ring and the hydrocarbon chain.
Furthermore, pH determines to a great extent the hydrated form
of Cr3+; therefore, its value strongly affects the radius of the
hydrated ion sphere.23,24 Therefore, a compromise between vesicle
integrity and ionic radius of hyrdrated Cr3+ must be reached.
With these observations in mind, the effect of pH on the
extraction yield of Cr3+ was assessed and the results are
graphically demonstrated in Figure 5. It is evident that increasing
pH values up to 4 improves the extraction efficiency, while at
higher values the performance of the method is deteriorated. The
reason why protonation or deportonation of COOH deteriorate
the extraction efficiency is a rather intricate phenomenon. The
pH of an aqueous 4-CBUAB vesicle solution (0.6%, w/v) is 3.1 ±
0.1 while the pKa of the benzyl carboxylate group is 4.19 and the
pKa value of the carboxylates present in the hydrocarbon chain
too alkaline solutions, respectively, which are bulkier than
23,24
Cr(OH)2+
.
Effect of Temperature. The reaction temperature is an
important parameter in vesicular solutions since it plays a crucial
role in the formation of coacervates and their structure. In the
case of polymerized vesicles examined in this study, the optimum
temperature for the extraction of chromium ions was 50 °C (Figure
6). This temperature coincides with previous studies who reported
that, when ionic strength is greater than 0.06, the direction of the
polar head groups is positive between 45 and 65 °C and reaches
its minimum at 50 °C.17 This condition favors the incorporation
of Cr3+ ions due to minimization of the electrostatic repulsion
between the highly polar interbilayer space and the large effective
charge of Cr3+ ions. At the same time, the electrostatic attraction
of the bilayer to oppositively charged ions (like CrO42-) is
minimized thus increasing vesicle selectivity.
(28) Monnard, P.-A.; Deamer, D. W. Anat. Rec. 2002, 268, 196–207
.
9792 Analytical Chemistry, Vol. 80, No. 24, December 15, 2008

the regulation of the vesicle surface charge due to the presence
of counterions.
The reasons why the direction of the head group of bilayers
changes depending on temperature and ionic strength are not well
understood. It has been shown, however, that under certain
temperature and ionic strength conditions the head groups bent
altering the external surface charge.17 More specifically, at ionic
strength above 0.06 M and temperature of 50 °C, the surface
charge of the vesicles reach its minimum positive value, thus
reducing both the electrostatic repulsion of Cr3+ and the attraction
2-
of CrO4 ions with the polar bilayer. Unfortunately, no studies
regarding the conformation of 4-CBUAB head groups has been
conducted so far; therefore, any hypothesis buries a certain degree
of uncertainty.
On the other hand, the influence of counterions on vesicle
charge is better understood and is related to the fact that vesicles
derived from substituted ammonium bromide monomers tend to
act like floating ion-exchange resins where the adsorption of small
ions onto the surface of the vesicles is a competitive exchange
process.12 In this context, as counterion concentration around the
charged groups increases with increasing ionic strength, the head
groups of the vesicle monomers can be located closely to each
other, because the electric repulsion between the charged groups
with the same sign is decreased.17 Up to a given concentration,
this may be favorable for Cr3+ permeation through the membrane,
but as ionic strength increases further, the bilayer distance may
be reduced to a critical point where Cr3+ can no longer permeate
through the intermolecular space. At the same time, vesicles
become unstable disrupting the colloidal form of the coacervate
phase, which deteriorates phase separation.12
Figure 6. Absorbance signal of Cr3+ as a function of mixture
temperature: [Cr3+] ) 1 µg L-1, [4-CBUAB] ) 0.6% (w/v), pH 4.0,
[KCl] ) 0.25 M.
Effect of Reaction and Centrifugation Time. The effects of
reaction and centrifugation time were the next experimental
parameters examined for their effect on the performance of the
method. The data suggest that the reaction is kinetically satisfied
within 15 min with only insignificant gains being made at longer
times. On the other hand, the centrifugation time necessary to
ensure complete phase separation was 20 min. However, centrifu-
gation time depends on vesicle concentration, as previously
discussed. At concentrations above 0.6% (w/v), a gradual increase
in centrifugation time with increasing vesicle concentration is
required in order to mechanically force the vesicle coacervates
to precipitate since they resist phase separation due to electrostatic
repulsion among the vesicles.
Influence of Chelating Agents. From an analytical standpoint,
the feasibility for the direct application of the method (as described
thus far) offers a significant simplification, since no additional
reaction or derivatization steps are required. Nevertheless, the
response of the polymerized vesicles in the presence of chelating
agents is an important aspect since it can shed more light on the
involved extraction mechanism and possibly provide the basis for
expanding its analytical use to other metal species that cannot
diffuse through the polymeric membrane.
Figure 7. Influence of ionic strength on the extraction of 1 µg L-1
Cr3+: [4-CBUAB] ) 0.6% (w/v), pH 4.0, temp ) 50 °C.
Following this observation, the incorporation of other metallic
ions was reassessed in order to shed more light on the extraction
mechanism. Evidently, no change in vesicle selectivity was
observed. This supports the notion that permeation of ionic species
is controlled by the interbilayer space, which acts as a barrier to
species with higher ionic radius. It is conceivable that electrostatic
forces play an important role in this process, yet too large ions
cannot pass the membrane irrespective of the vesicle surface
charge.
Effect of Ionic Strength. The influence of various inorganic
electrolytes (KCl, KNO3, KBr, NaCl, NaBr, NaNO3) on the
analytical repose of Cr3+ was examined at concentrations ranging
from 0 to 1 M. Of the inorganic salts examined, KCl provided the
best results, improving Cr3+ extraction with increasing concentra-
tion up to 0.5 M and decreasing again at higher values (Figure
7).The explanation for this behavior is rather intricate and involves
the coexistence of two phenomena. The first is the conformation
of the head groups as a function of ionic strength and the second
To examine the extraction mechanism, specifically designed
experiments were conducted using two chelating agents: 8-hy-
droxyquinoline, which is sparingly soluble is water, and APDC,
which is soluble in water. In the first set of experiments, 1 µg
L-1 Cr3+ ions was extracted under the optimum experimental
conditions established above. The supernatant aqueous phase was
Analytical Chemistry, Vol. 80, No. 24, December 15, 2008 9793

interactions between the metal chelate and the hydrophobic
vesicular core are trivial and support the notion of direct ion
permeation through the vesicular membrane. This notion is
further supported by the transmembrane metal ion transport
mechanism occurring in carrier-mediated uptake of dilute aqueous
heavy metal ions by metal-sorbing vesicles.13,14
Table 1. Analytical Features of the Method
parameter
Cr3+
phase volume ratioa
0.1
preconcetration factorb
enhancement ratioc
∼5
1.7
3.84-0.79
2.0
extraction concentration factord
consumptive index (mL)e
Selectivity of Polymerized Vesicles. As previously discussed,
the coacervate phase formed from polymerized vesicles of
4-CBUAB was indifferent to the present of most metallic species
except those with small ionic radius. Following the optimization
of the experimental conditions that enable the quantitative
f
LOD (µg L-1
LOQ (µg L-1
)
)
0.10
0.33
1.51
g
RSD (%) (n ) 7, 2 µg L-1
regression equation
)
A ) 0.132 ± 3.6 × 10-3 + 0.104
C ± 1.7 × 10-3
correlation coeffieicnt (r)
0.9979 (confidence level 95%)
determination of Cr3+ species, the selectivity of the system in
a Phase volume ratio: the ratio of the final volume of vesicle-rich
phase to that of the aqueous phase. b Preconcentration factor: the ratio
of the slopes of the calibration curves with and without the precon-
centration step. c Enhancement ratio: the ratio of the concentration of
analyte after preconcentration to that without preconcentration giving
the same absorbance peak area. d Extraction concentration factor: the
ratio of concentration in the vesicle-rich phase to that in the original
solution. Extraction concentration factor is decreased as concentration
is increased. e Consumptive index: The sample volume required for
achieving a unit of preconcentration factor. f LOD: limit of detection,
defined as three times the signal-to-noise ratio. g LOQ: limit of
quantitation, defined as 10 times the signal-to-noise ratio.
2-
mixtures of Cr3+-CrO4
was reassessed. Not surprisingly,
CrO42- exhibited no reactivity with the vesicular phase producing
zero absorbance signal.
Interferences. The high selectivity of the polymerized ve-
sicular phase and the lack of complexation reactions that could
be subject to antagonism in the presence of other metallic cations
alleviate most interferences that could reduce Cr recovery. This
fact along with the high selectivity of the atomic absorption
detector offers the possibility of obtaining the interference-free
determination of Cr3+. To examine the validity of the later
statement, various metallic species in a analyte-to-interferant ratio
decanted and the coacervate phase was restored to initial volume
with water containing 1-2 mM 8-HQ (in methanol) or 2-4 g L-1
APDC (in water). The extraction procedure was repeated and the
coacervate and supernatant phases were separated and analyzed
for their content in Cr3+. The data show that no Cr3+ leaching to
the supernatant was observed and all Cr3+ was quantitatively
recovered in the coacervate phase.
ranging from 1:10 to 1:1000 were examined. As expected, Cu2+
Zn2+, Pb2+, Cd2+, Ni2+, Sb3+, Al3+, and As3+ had no effect on
the analytical response of Cr3+. However, in the presence of Fe3+
,
,
which is coextracted with Cr3+, a decrease in the absorbance
signal was observed at Fe-to-Cr ratios above 250, possibly due to
saturation of the vesicles. Although various methods may be
deployed to overcome this interference,31 the addition of EDTA
was the first measure examined. According to previous studies,
EDTA can relieve Cr from competitor metal species through the
formation of strong ionic complexes.18 Increasing EDTA concen-
tration to 25 mg L-1 did not adversely affect the analytical signal
of Cr; therefore, this value was selected as optimum. Too high
values of EDTA though may enhance EDTA-vesicle surface
interaction due to COO- association with the substituted am-
monium bromide group (R-CH3N+). The presence of Cl- coun-
terbalances this interaction, but it is possible that excess of EDTA
may cause some competition for the substituted ammonium
bromide, thus altering Cr3+ ion permeability or even disrupting
the vesicle structure.
There are two possible explanations for this behavior. Either
complexing agents penetrate the vesicle, complex Cr3+, and are
encapsulated in the vesicular core or they cannot penetrate
through the charged interbilayer spaces and thus do not react
with Cr3+. To investigate which mechanism prevails, two experi-
ments were run in parallel. In the first, an aqueous vesicle solution
(0.6% w/v) was spiked with 1-2 mM 8-HQ (in methanol) or 2-4
g L-1 APDC (in water). The sample was centrifuged and the
supernatant was decanted. The coacervate phase was then
redissolved with 10 mL of water and 1 µg L-1 Cr3+ ions was added.
The sample was mixed, centrifuged, and both the coacervate and
the supernatant phases were collected and analyzed for their
content in Cr3+. In the second set, APDC was added to an aqueous
Cr3+ solution. Polymerized vesicles were added and extraction
was followed in a similar way as a typical cloud point extraction
experiment.29 In all experiments, APDC extraction was carried
out at pH 6 while for 8-HQ the sample was adjusted to pH 8. Both
samples were heated at 50 °C for 15 min. Although the extraction
of Cr3+ with 8-HQ at this temperature is semiquantitative,30 no
higher temperatures could be employed due to limitations
imposed by polymerized vesicle response to temperature effects
as previously discussed. The results from the first set of experi-
ments reveal that no Cr3+ was present in the supernatant phase
while most Cr3+ could be found in the coacervate phase (set 1).
In the second series (set 2), no Cr3+ amounts were detected in
the coacervate phase. These results indicate that hydrophobic
With regard to anions likely to be present in natural waters,
SO42-, NO3-, NO2-, F-, and Cl- did not affect the analytical
response of the system at environmentally realistic concentrations.
3-
On the other hand, PO4 were found to decrease Cr extraction
at PO43--to-Cr ratios above 500 possibly due to association of these
ions on the vesicle surface. To overcome this drawback, three
methods were examined involving the use of a Mg(NO3)2 as phase
3-
modifier32 and the removal of PO4 prior to extraction with
the use of Al2SO4 and AlCl3. The fact that Mg(NO3)2 phase
3-
modifier could not improve Cr3+ extraction suggests that PO4
interference is not of spectroscopic origin but most probably alters
the vesicle surface morphology thus deterring Cr3+ diffusion
through the vesicular interbilayer space. On the other hand,
(29) Paleologos, K. E.; Giokas, L. D.; Karayannis, I. M. Trends Anal. Chem. 2005,
(31) Sotelo, M. F.; Andrade, M. J.; Carlosena, A.; Prada, D. Anal. Chem. 2003,
24, 426–436
(30) Paleologos, K. E.; Stalikas, D. C.; Tzouwara-Karayanni, M. S.; Pilidis, A. G.;
Karayannis, I. M. J. Anal. At. Spectrom. 2000, 15, 287–291
.
75, 5254–5261
(32) Thomaidis, S. N.; Piperaki, A. E.; Polydorou, K. C.; Efstathiou, E. C. J. Anal.
At. Spectrom. 1996, 11, 31–36
.
.
.
9794 Analytical Chemistry, Vol. 80, No. 24, December 15, 2008

2-
Table 2. Speciation of Cr3+ and CrO4 in Natural Water Samples and Comparison with GFAASa
2-b
samples
Cr3+
CrO4
total Cr
total Cr (GFAAS)
relative error (%)
t testc
Loudias river (site 1)
Loudias river (site 2)
irrigation cut 66
0.48 ± 0.03
1.06 ± 0.04
0.78 ± 0.03
0.60 ± 0.05
0.86 ± 0.05
0.53 ± 0.03
0.82 ± 0.04
2.19 ± 0.04
1.29 ± 0.05
2.55 ± 0.03
2.73 ± 0.04
3.73 ± 0.09
2.99 ± 0.07
2.22 ± 0.06
2.67 ± 0.04
2.35 ± 0.05
3.33 ± 0.06
3.33 ± 0.05
4.59 ± 0.12
3.52 ± 0.08
3.04 ± 0.06
2.65 ± 0.11
2.26 ± 0.12
3.28 ± 0.08
3.31 ± 0.06
4.34 ± 0.15
3.48 ± 0.10
2.99 ± 0.04
0.86
3.76
1.46
0.12
5.69
1.00
1.87
0.423
1.454
1.142
0.131
6.842
1.079
2.104
Aliakmon river
Axios river
Edessaios river (site 1)
Edessaios river (site 2)
a All concentrations in µg L-1. Average of triplicate measurements ± standard deviation, b [CrO42-] ) total [Cr] - [Cr3+]. c t ) |CThis method
-
CGFAAS|/([(1/nA) + (1/nB)] {[s2A (nA - 1) + sB2 (nB - 1)]/[nA + nB - 2]})1/2, where nA ) nB ) 5. The value of t critical (for P ) 0.95 and number
of degrees of freedom ν ) nA + nB - 2 ) 8) is 2.306.
Figures of Merit. Calibration curves were obtained by pre-
Table 3. Recovery Study of Spiked Natural Water
concentrating 10 mL of standard solutions with 0.6% (w/v)
4-CBUAB under the optimum experimental conditions established
above. Table 1 features the analytical characteristics of the method.
Under the proposed experimental protocol, the calibration curves
were rectilinear in the range 1-50 µg L-1 Cr3+. The limits of
detection were very satisfactory, considering the low phase volume
ratio applied and the fact that vesicle concentration was set to
0.6% instead of 2% (w/v) (Figure 3), but they are fairly adequate
for Cr speciation in natural waters. The preconcentration factor,
obtained as the ratio of the slopes of the calibration curves for
Cr3+ with and without the preconcentration step, was 4.95. Further
improvement is feasible either by preconcentrating larger sample
volumes or by using the optimum quantity of vesicular concentra-
tion (2%, w/v) although, in the latter case, phase separation is
labor intensive due to its resistance in phase separation, as
previously discussed. In the first case, the time of analysis is
reasonable, not exceeding 1 h for multiple samples depending
on the capacity of the centrifuge, while in the second case
prolonged or multiple centrifugations are needed to accomplish
complete separation of the phases and precipitation of the vesicular
phase.
Samples
Loudias
irrigation Aliakmon
cut 66 river
Edessaios
samples
river (site 1)
river (site 1)
Cr3+
measured
spiked
0.48 ± 0.03
2.0
0.78 ± 0.03 0.60 ± 0.05
2.0 2.0
0.53 ± 0.03
2.0
found
2.53 ± 0.14
102.5
2.74 ± 0.08 2.71 ± 0.10
2.55 ± 0.10
101
recovery (%)a
t testb
98.0
0.866
105.5
1.905
0.618
0.346
2-
CrO4
measured
spiked
2.19 ± 0.04
5.0
2.55 ± 0.03 2.99 ± 0.07
5.0 5.0
2.73 ± 0.04
5.0
found
recovery
t test
7.38 ± 0.20
103.8
7.46 ± 0.16 7.84 ± 0.20
7.80 ± 0.18
101.4
98.2
0.820
97.0
1.299
1.645
0.673
Total Cr
measured
spikedc
found
recovery
t test
2.67 ± 0.04
5.0
3.33 ± 0.06 3.33 ± 0.05
5.0 5.0
3.52 ± 0.08
5.0
7.66 ± 0.16
99.8
8.30 ± 0.17 8.36 ± 0.21
8.61 ± 0.20
101.8
99.4
0.305
100.6
0.247
0.091
0.779
a Estimated according to IUPAC recommendations. b t ) (|CTheoretical
-
CFound|)/(s)ꢀn, where s is the absolute standard deviation estimated from
the experimental tests (n ) 3). The tabulated value of Student’s coefficient
for P ) 0.95 and ν ) 2, t(P = 0.95, ν ) 2) ) 4.30, which is the t critical
value.c TotalCrwasspikedatequivalentCr3+ andCrO42- concentrations.
Polymerized Vesicle Coacervate Speciation-Extraction of
Cr in Environmental Water Samples. The ability of polymerized
vesicular coacervates to work under real conditions was assessed
by applying them to the speciation of Cr in natural waters. Samples
from four rivers in central Macedonia (Greece) were collected at
different locations based on the anticipated burden imposed by
anthropogenic activities. The speciation of Cr with the proposed
mechanism is a two-step process. The first involves the determi-
nation of Cr3+ species and the second the determination of total
Cr (as Cr3+) after reduction of CrO42-, the latter being determined
as the difference between total Cr and Cr3+ species. The results
gathered in Table 2 are in good agreement with those obtained
by GFAAS for total Cr with relative error below 5% in most cases.
According to the t-criterion, differences in the measured concen-
trations between the two methods are insignificant at the P )
0.95 probability level. In only one case, the experimental t-value
was higher than t critical possibly due to the high burden of the
sample, which was expected to carry a significant amount of
pollution from human activities; therefore, several interferences
may coexist at elevated levels that may interfere with the analysis.
Spiking of the target analytes served to investigate method’s
recoveries from real water samples (Table 3). The results show
that the method affords recoveries in the range of 97.0-105.5%
while the data of t-criterion provide additional evidence for the
3-
removal of PO4 via coprecipitation as their AlPO4 complexes
alleviated the interference offering considerable improvement in
the extraction yield. Optimum results were obtained at 20 mg L-1
2-
AlCl3, which can be attributed to the fact that high SO4
concentrations probably interfere with the repulsive forces be-
tween amphiphile head groups. On the other hand, Cl- ions were
already present in solution due to regulation of the ionic strength,
thus causing no alterations in the solution mixture.
Organic matter in the form of humic acids was also examined
since it can bind Cr or, even more, can increase the apparent water
solubility of hydrophobic molecules through the formation of
surfactant-like micelles.33 In addition, humic acids could possibly
react with the positively charged polymerized vesicle surface and
thus modify the vesicle structure and interfere with Cr-vesicle
permeability. Organic matter up to 5 mg L-1 did not show any
negative influence on the analytical response of Cr3+. Since the
levels of organic matter in rivers are usually below 6 mg L-1, they
pose as no threat to the speciation of Cr with the proposed
experimental procedure.34
(33) Wershaw, L. R. J. Contam. Hydrol. 1986, 1, 29–45
(34) Michałowski, J.; Hałaburda, P.; Kojło, A. Anal. Chim. Acta 2001, 438, 143–
148
.
.
Analytical Chemistry, Vol. 80, No. 24, December 15, 2008 9795

robustness of the method since no statistically significant differ-
ences were observed between found and expected values at the
P ) 0.95 probability level.
conditions is necessary, but in contrast to conventional extraction
protocols, optimization of the extraction medium (vesicle) struc-
ture and surface charge is the only regulating parameter.
The extraction and phase separation mechanisms involved in
Cr3+ speciation, unraveled within the confines of the present
study, render polymerized vesicle phases a challenging issue that
merits further examination. Most importantly, it pledges on the
development of highly selective extractions based, most probably,
on size exclusion rather than polar or nonpolar interactions. At
present, the interaction of polymerized vesicles with gold nano-
particles in aqueous dispersions seems to produce a unique
microenvironment that transforms spectroscopically inert systems
to highly sensitive photometric reporters. Such a study is
underway in our laboratory, and the results will be available in
due time.
CONCLUSIONS
There are two main reasons for deploying coacervate phases
in analytical extractions. The first is to alleviate the use of organic
solvents and the second is their high versatility, which enables
vesicle/analyte specific interactions thus achieving high selectivity.
The method proposed in this study demonstrates that polymerized
vesicles fulfill both criteria, offering a unique vehicle toward highly
selective and solvent-free extractions.
The polymerized vesicle coacervate phase employed in this
work presented a unique behavior, a nearly insurmountable barrier
for metal ion species with ionic radius larger than the interbilayer
space. In that manner, small ionic species like Cr3+ could be
selectively isolated from the aqueous phase and determined in
the presence of a wide range of coexisting ions. As with all
analytical methods, appropriate regulation of the experimental
Received for review September 23, 2008. Accepted
October 29, 2008.
AC802018W
9796 Analytical Chemistry, Vol. 80, No. 24, December 15, 2008