Electrospun amino-functionalized PDMS as a novel SPME sorbent for the speciation of inorganic and organometallic arsenic species

Article abstract of DOI:10.1039/c3ra43622e

Sol-gel based amine-functionalized SPME fibers (PDMS-weak anion exchanger) were prepared and used for direct mode extraction of dimethylarsinic acid (DMA), monomethylarsonic acid (MMA), and arsenate (As(v)) from aqueous solutions followed by HPLC-ICPMS determination. Two different methods of coating were employed: (i) electrospinning and (ii) dip coating. Electrospinning was used for the first time for preparation of sol-gel based SPME fibers and was found to be superior in terms of extracted amount of arsenicals, coating homogeneity, accessibility of amine groups on the surface, and preparation time for a single fiber. Various parameters such as solution pH, extraction time, agitation speed, extraction temperature, and ionic strength were studied. Optimum extraction conditions were determined as pH of 5.0, extraction time of 30 min, agitation speed of 700 rpm, and extraction temperature of 20 °C. Extraction ability of the novel coating decreased by the addition of NaCl as a consequence of the competition between anionic arsenic species and chloride ions for active sites of the weak anion exchanger. This novel sol-gel coating prepared by electrospinning was found to be promising for SPME applications. Vibrational spectroscopy revealed the alignment of PDMS chains by elongational force under electrospinning process. The chain alignment accordingly orients the pendant amino functional groups perpendicular to the fiber surface, which may develop the free active functional groups available to the medium and lead to the enhancement of the extraction performance. Moreover, the proposed coating strategy through electrospinning might be able to break new ground for various applications in analytical chemistry as well as other disciplines. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ra43622e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Electrospun amino-functionalized PDMS as a novel
SPME sorbent for the speciation of inorganic and
organometallic arsenic species†
Cite this: RSC Adv., 2013, 3, 22261
*
*
˘
˘
Ezel Boyacı,‡ Nesrin Horzum,‡ Ali Çagır, Mustafa M. Demir§ and Ahmet E. Eroglu
Sol–gel based amine-functionalized SPME fibers (PDMS-weak anion exchanger) were prepared and used
for direct mode extraction of dimethylarsinic acid (DMA), monomethylarsonic acid (MMA), and arsenate
(As(V)) from aqueous solutions followed by HPLC-ICPMS determination. Two different methods of
coating were employed: (i) electrospinning and (ii) dip coating. Electrospinning was used for the first
time for preparation of sol–gel based SPME fibers and was found to be superior in terms of extracted
amount of arsenicals, coating homogeneity, accessibility of amine groups on the surface, and
preparation time for a single fiber. Various parameters such as solution pH, extraction time, agitation
speed, extraction temperature, and ionic strength were studied. Optimum extraction conditions were
determined as pH of 5.0, extraction time of 30 min, agitation speed of 700 rpm, and extraction
temperature of 20 ꢀC. Extraction ability of the novel coating decreased by the addition of NaCl as a
consequence of the competition between anionic arsenic species and chloride ions for active sites of the
weak anion exchanger. This novel sol–gel coating prepared by electrospinning was found to be
promising for SPME applications. Vibrational spectroscopy revealed the alignment of PDMS chains by
elongational force under electrospinning process. The chain alignment accordingly orients the pendant
amino functional groups perpendicular to the fiber surface, which may develop the free active
functional groups available to the medium and lead to the enhancement of the extraction
performance. Moreover, the proposed coating strategy through electrospinning might be able to break
new ground for various applications in analytical chemistry as well as other disciplines.
Received 15th July 2013
Accepted 17th September 2013
DOI: 10.1039/c3ra43622e
www.rsc.org/advances
hence it gives place to sorbent-based techniques such as solid
phase extraction (SPE) and solid phase microextraction (SPME).
1 Introduction
Sample preparation is one of the most critical and challenging
steps in a chemical analysis owing to complex nature of the
matrix for most of the environmental and biological samples.
From the analytical point of view, it can help to isolate the
analyte from sample matrix, preconcentrate or reconstitute in a
more appropriate solvent and give a prospect for reliable trace
level determination of analytes. In addition to being time-
consuming and tedious, the entire analytical process can even
fail unless an appropriate sample preparation method is used.
For instance, one of the commonly used sample preparation
method – conventional liquid–liquid extraction (LLE) – requires
high amount of samples and large volume of organic solvents;
Recent approaches in these techniques provide advantages
thanks to the availability of a variety of commercial products
and sorbent materials such as inorganic oxides like silica,
carbon, and polymers. During the last two decades, the devel-
opment of some miniaturized extraction techniques originated
from solid phase extraction (SPE) has allowed simplied sample
preparation and onsite sampling with subsequent instrumental
analysis. However, SPE has limitations due to its difficulty in
method development, not applicable for onsite sampling,
greater complexity as well as cost resulting from being single
use.1 Solid-phase microextraction (SPME) methodology is a
feasible alternative to LLE and SPE. It has been frequently used
owing to its ease of operation, being rapid, portable and solvent-
free. This methodology enables developing various ber coat-
ings with both advanced extraction efficiency and selectivity.
SPME ber consists of a fused silica core coated with an active
layer. The promising commercially available coatings are poly-
dimethylsiloxane (PDMS) for nonpolar, polydimethylsiloxane/
divinylbenzene (PDMS/DVB) for semi-polar, polyacrylate (PA) and
carbowax/divinylbenzene (CW/DVB) for polar compounds.2 The
studies related to production of the new ber coatings are
˙
˙
Department of Chemistry, Izmir Institute of Technology, Urla, 35430, Izmir, Turkey.
E-mail: ahmeteroglu@iyte.edu.tr
† Electronic supporting information (ESI) available: Operation conditions for
HPLC-ICPMS, chromatograms in the presence and absence of the collision gas,
effect of solution pH, agitation time and speed, NaCl concentration, solution
temperature on extraction of arsenic species. See DOI: 10.1039/c3ra43622e
‡ These authors contributed equally to the work.
˙
§ Also at: Department of Materials Science and Engineering, Izmir Institute of
˙
Technology, 35430, Izmir, Urla, Turkey. E-mail: mdemir@iyte.edu.tr.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22261–22268 | 22261

View Article Online
RSC Advances
Paper
extended to develop solvent-resistive and thermally stable active
coatings as well as to enhance mechanical strength of the phase.
Recently reviewed ber coating processes involve sol–gel
methods,3–5 molecular imprinting,6,7 electrochemical methods,8,9
adsorption of polymeric ionic liquids,10–12 and immobilization
of nano and micro sized functional particles. Among these
methods, sol–gel method offers the opportunity for attachment
of various functionalities to the coating which expands the
working area of SPME. Various approaches have been reviewed
in literature for immobilization of the aforementioned active
phases on a SPME ber. These approaches are dipping the ber
into the solution of active phase, spraying, brushing and elec-
trospinning of the solution/slurry onto ber.13 Among these
coating methods, electrospinning has the ability to fabricate
micro/nanobers, to control morphology and for orientation of
nanobrous mat.14–16 The applicability of electrospun bers as
SPME sorbent for the extraction of both nonpolar and polar
compounds was reported by Zewe et al.17 and it was found that
the electrospun polymeric negative photoresist, SU-8 2100, ber
coated wires displayed the best performance with low detection
limits and large linear ranges for phenolic compounds. Bagheri
and Aghakhani evaluated polyurethane and polyaniline-nylon-6
electrospun bers for the extraction of chlorobenzenes.18,19
Similarly, Bagheri et al. suggested polyamide ber coating for the
extraction of chlorophenols20 from aqueous systems. As described
above, the current electrospun SPME ber coatings are based on
electrospinnable polymers. Generally, these polymers suffer
from low thermal stabilities which limit their application areas.
Herein, we present, for the rst time to the best of our
knowledge, a novel method for SPME ber coating by combi-
nation of both sol–gel chemistry and electrospinning. The sol–
gel solution was tailored by blending PDMS and 3-(triethox-
ysilyl) propylamine (APTES). PDMS has unique physical and
chemical properties resulting from the nature of the siloxane
backbone and methyl groups attached to silicon. Due to its non-
toxicity, low chemical reactivity, thermal and mechanical
stability, and high surface activity, PDMS is used in a wide
variety of application areas such as microuidic devices,
biomedical products, personal, household, and automotive care
industry.21–24 PDMS can be functionalized using APTES which
has terminal amine groups providing sorption sites for arsenic
species. The amine-functionalized PDMS electrospun lm
coated SPME bers were utilized for the trace determination of
arsenic species to survey their extraction feasibility. There are
numerous reasons to select the arsenic as a model analyte in
this study. First, arsenic is a ubiquitous element which is
present in the atmosphere, groundwater, soil, rocks, and
organisms. Second, most of the arsenic compounds are known
to be toxic; although their toxicity depends on the oxidation
state and the order of methylation of particular species.
Therefore, it is important to determine individual species in a
sample rather than the total concentration. Finally, most of the
organometallic and inorganic arsenic species are anionic in
2 Experimental
2.1 Reagents and solutions
All the chemicals were of analytical reagent grade. Ultrapure
water (18.2 MU cm) was used throughout the study. Glassware
and stirring bars were cleaned by being soaked in 10% (v/v)
nitric acid for 24 h and rinsed with ultrapure water prior to use.
Stock standard solutions of As(V), 2000.0 mg Lꢁ1, were
prepared by dissolving As2O5 (Merck, product code: 1.09939,
CAS no.: [1303-28-2]) in 1.0% (v/v) HCl. Dimethylarsinic acid
(DMA, 1000.0 mg Lꢁ1) and monomethylarsonic acid (MMA,
553.0 mg Lꢁ1) were prepared by dissolving dimethylarsinic acid
sodium salt trihydrate (Merck, product code: 8.20670, CAS no.:
[6131-99-3]) and disodium methyl arsenate hexahydrate
(Supelco, product no.: PS 281, CAS no.: [144-21-8]), respectively
in ultrapure water. The concentrations of all the stock solutions
described above were given in terms of arsenic. Prepared stock
ꢀ
solutions were stored in refrigerator at 4 C; lower concentra-
tions were prepared daily by diluting appropriate amount of
stocks in ultrapure water. HNO3 (Merck, product code: 1.00456,
CAS no.: [7697-37-2]) and NH3 (Merck, product code: 1.05422)
were used to adjust the pH of the solutions.
For HPLC-ICPMS studies, 10.0 mM and 30.0 mM (NH4)2CO3
solutions (pH 8.50) were freshly prepared by dissolving a proper
amount of HPLC grade (NH4)2CO3 (Fluka, product code: 74415,
CAS no.: [506-87-6]) in ultrapure water, and pHs of the solutions
were adjusted by using various concentrations of HNO3 and NH3.
All solutions including mobile phase and eluting solvent
were ltered through 0.20 mm cellulose acetate lter paper
(Sartorius) and degassed for 15 min in an ultrasonic bath before
being used in the chromatographic column.
Fiber optic cable was kindly supplied by HES Kablo (Kayseri,
Turkey). Acetone (Merck, product code: 1.00014, CAS no.: [67-
64-1]) was used for the removal of protective polyamide coating
of the bers. The introduction of high number of silanol groups
on the surface of the bare ber was achieved by the treatment
with 1.0 M NaOH solution prepared by using NaOH pellets
(Merck, product code: 1.06498, CAS no.: [1310-73-2]). 0.1 M HCl
(Merck, product code: 1.00314, CAS no.: [7647-01-0]) was used
for neutralization of the base.
Polydimethylsiloxane, PDMS (hydroxyl terminated, typical
viscosity: 25 cSt), (Aldrich, product code: 481939, CAS no.:
[70131-67-8]) was used as organic polymer with sol–gel active
terminal. Tetrahydrofuran, THF (Sigma Aldrich, product code:
87368, CAS no.: [109-99-9]) was used as solvent in sol–gel
process. 3-(Triethoxysilyl) propylamine, APTES (Merck, product
code: 8.26619, CAS no.: [14814-09-6]) was used as sol–gel active
silane coupling reagent with amine functionality. Triuoro-
acetic acid, TFA (Merck, product code: 76-05-1, CAS no.:
[808260]) was used as a catalyst in sol–gel process.
2.2 Instrumentation and apparatus
aqueous systems which enable using weak anion exchange The separation of extracted arsenic species was achieved with
capability of novel coating for their determination. The Agilent 1200 Series HPLC system with a 250 mm anion exchange
proposed SPME-HPLC-ICPMS method was found to be prom- column (Hamilton, PRP-X100). Gradients with 10.0 mM and
ising for the individual determination of the arsenicals.
30.0 mM (NH4)2CO3 solutions (pH 8.50) were used as mobile
22262 | RSC Adv., 2013, 3, 22261–22268
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
Table 1 Extraction parameters and ranges used throughout the studya
phase at a ow rate of 1.0 mL minꢁ1. Chromatographic condi-
tions were 10.0 mM (NH4)2CO3 for the rst 4 min, 30.0 mM
(NH4)2CO3 for 4–13 min and nally 10.0 mM (NH4)2CO3 for
2 min for preconditioning of column. pH of the solutions was
adjusted with 0.01–1.0 M HNO3 and 0.01–1.0 M NH3 using Ino
Lab Level 1 pH meter (Weilheim, Germany). Detection of each
species was realized with ICP-MS, Agilent 7500ce (Tokyo, Japan).
Online separation and detection of arsenic species were
accomplished by connecting the HPLC column outlet directly to
ICP-MS concentric nebulizer. Two instruments were connected
by the shortest possible length (67 cm) of peruoroalkoxy (PFA)
tubing (ID: 0.5 mm) to prevent broadening of the peaks. The
operating parameters for HPLC and ICP-MS are provided in
Table S1 in the ESI.† Extraction of arsenic species was per-
formed in multi-position magnetic stirrer RO 10 power
IKAMAG.
Parameter
Range
pH of solution
3.0, 5.0, 7.0, 9.0
5, 15, 30, 60, 120
300, 500, 700, 1000
0, 0.01, 0.10, 1.0
20, 35, 55
Agitation time (min)
Agitation speed (rpm)
Concentration of salt (M)
Solution temperature (ꢀC)
a
Sample volume and arsenic concentrations are kept constant as 15.0
mL and 10.0 mg Lꢁ1, respectively.
on the ber, the plain silica ber was attached on a rotating
drum (300 rpm) which was placed between the metal collector
and syringe (3 cm away from the syringe). During the process
2 cm of the ber was coated. Coated bers were conditioned at
ꢀ
110 C overnight.
The morphology of bare and coated silica bers was exam-
ined using a Philips XL-30SFEG scanning electron microscope
(SEM) (Eindhoven, The Netherlands). In electrospinning
process, the syringe was connected to a high voltage power
supply (Gamma High Voltage Research, Ormond Beach, FL,
USA) and xed on the microsyringe pump (LION WZ-50C6).
Fourier Transform Infrared (FTIR) spectra were registered in a
Perkin Elmer Spectrum 100 spectrometer with a spectral reso-
lution of 4 cmꢁ1. Polarization of the incident IR radiation was
achieved with a ZnSe polarizer. To evaluate the possible orien-
tation of functional groups provided by electrospinning
process, the prepared sol–gel solution was directly electrospun
onto the KBr pellets. In comparison to electrospun coated
sample, the sol–gel solution was drop-casted onto the KBr
pellets.
SPME bers were also prepared by dip coating using the
above mentioned sol–gel solution. 200 mL disposable pipette tip
was blocked and used as a container for sol–gel solution. The
bare activated ber was immersed (2 cm) in the coating solution
and allowed to react for 30 min. Then the ber was removed
with constant speed of 1.0 mm minꢁ1 and conditioned at 110 ꢀC
overnight.
2.4 Extraction of the arsenic species
Extractions were performed in a 20 mL vial containing 15.0 mL
of aqueous arsenic mixture solution (DMA, MMA and As(V))
which was stirred with a magnetic bar. Desorption of the ana-
lytes was carried out in a container which was prepared from
200 mL pipette tip blocked at the base by a gentle application of
heat. Various desorption solutions were tested and KH2PO4 was
chosen as a proper eluent. Optimized desorption conditions
were; desorption time of 20 min, desorption volume of 150 mL,
eluent concentration of 50.0 mM KH2PO4 (results are not
shown). Extraction parameters such as extraction pH, extraction
time, agitation (stirring) speed, ionic strength and extraction
temperature were optimized. The investigated parameters and
ranges are listed in Table 1.
2.3 Preparation of the SPME bers
The protective polyamide layer of the silica ber was removed by
immersing the piece of 7 cm cut into acetone. The bare bers
(diameter of ꢂ125 mm) were washed rst with ultrapure water.
Introduction of maximum amount of silanol groups on bare
silicate surface required for coupling reaction was accom-
plished by dipping segments of the bare silicate bers into 1 M
NaOH solution for 60 min. This step was followed by dipping
the ber into 0.1 M HCl solution for 30 min to neutralize the
excess base. Fibers were washed with ultrapure water and
3 Results and discussion
3.1 Characterization of prepared SPME bers
ꢀ
thermal curing was applied for 24 h at 120 C.
The sol–gel solution used for coating SPME bers was In this study, the main strategy was to prepare a sol–gel based
prepared by end-functionalization of hydroxyl terminated coating for solid phase microextraction of various arsenicals
chains of PDMS with APTES. In a typical procedure, 2.0 mL from water. For this purpose, rst of all, a sol–gel solution
PDMS, 3.0 mL THF and 1.0 mL APTES were mixed in a 20 mL containing PDMS and APTES was prepared. In dip coating,
vial. Subsequently, 1.0 mL of TFA (95 v TFA + 5 v H2O) was added which is the most common method used for sol–gel based
into the solution, mixed thoroughly, and stirred for 48 h at room coating of SPME bers, the bare ber is dipped into a sol–gel
temperature. Electrospinnable sol–gel solution was formed by solution for a pre-determined time and then the ber is
adjusting the viscosity of the solution through evaporation of removed at a constant speed to obtain homogeneous coating. In
ꢀ
the solvent at 80 C for 2 h. The resulting sol–gel solution was addition to dip coating, this study reports, for the rst time, an
loaded into a syringe and electrospinning process was per- alternative method for immobilization which is fundamentally
formed at 25 kV potential difference with the feeding rate of depending on electrospinning of the sol–gel solution on the
5.0 mL hꢁ1. The distance between the syringe and the metal SPME ber surface. An electrospinning time of 12 s was selected
collector was 10 cm. In order to obtain a homogeneous coating to achieve homogeneous coating on the surface of the ber
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22261–22268 | 22263

View Article Online
RSC Advances
Paper
which provided also comparable thicknesses on nal coatings
from both of the aforementioned methods.
SEM micrographs of amino-functionalized PDMS coated
bers by dip-coating and electrospinning are shown in Fig. 1.
The diameters of the bers aer coating processes were
138 ꢃ 0.5 mm and 130 ꢃ 0.5 mm for dip coated and electrospun
coated bers respectively. The subtraction of the diameter of
bare SPME ber (ꢂ125 mm) from the nal diameter gives an
idea about the thickness of the coated lm, which was about
6.5 mm for dip coated and 2.5 mm for electrospun coated bers.
In order to obtain thicker coating by electrospinning method
various electrospinning times were tested. However, by appli-
cation of longer spinning times homogeneity was lost and
mainly bulb like accumulation of the coating material over the
ber was observed owing to the visco-elastic behavior of the
nal sol–gel solution.
Fig. 2 Polarized FTIR spectra of (a) drop-coated and (b) electrospun amino-
functionalized PDMS film coated KBr pellets.
Uniaxially aligned bers were obtained using a special
geometry of counter electrodes. Two metal strips were positioned
parallel to each other with an air gap (nearly 1 mm). The electrical
eld lines are forced to the edge of the strips and the bers are
obtained in between the metal strips. Polarized FTIR spectros-
copy was used to investigate molecular orientation of amino-
functionalized PDMS under electrical eld. Fig. 2 presents the
polarized FTIR spectra of the coating lm prepared by drop-
casting and electrospinning. The spectra of PDMS25 and APTES26
are in line with the literature. Parallel and perpendicularly
polarized spectra of drop-casted lm are found to be analogous,
indicating that there is no preferential alignment of PDMS chains
(Fig. 2a). However, for the electrospun coated lm, a remarkable
difference was observed depending on the polarization of the
incident IR beam (Fig. 2b). Upon polarization of incident IR
beam, the intensities of the vibrational bands undergo change
although they appear at the same wavenumbers. For instance,
the vibrational bands at 1680 cmꢁ1 and 800 cmꢁ1 which corre-
spond to N–H bending and Si–O symmetrical stretching were
taken into consideration. When the incident IR beam is parallel-
polarized to the electrospun amine-functionalized PDMS lm
calculated by the formula R ¼ A /At, where A and At refer to
k
k
parallel and perpendicularly polarized absorbance intensities for
a particular vibration, respectively. The ratio for the bands at 800
and 1680 cmꢁ1 is 0.9 and 1.2 respectively. One can propose that
the high extensional force caused by electrospinning process
provides the functional amine groups to be more aligned normal
to the lm surface.
It has to be noted that PDMS has high molecular elasticity
due to the exible Si–O bond,28 so that PDMS chains have
tendency to recover to their original positions upon electrical
force perturbation. However, chain alignment with higher
dichroic ratio has been shown for electrospun bers with
various chemistries27,29 in lꢁit1erature. On the other hand, the
vibration signal at 1430 cm originates from C–H bonds and
the band centered at 720 cmꢁ1 is attributed to the Si–O–Si bond.
3.2 Optimization of chromatographic parameters for HPLC-
ICPMS
axis, the absorbance of Si–O bond vibration is higher and N–H Most of the buffers used in the chromatographic separation of
bond vibration is lower. The comparison between the absorbance arsenic species result in salt deposition on skimmer and sample
of the backbone and the amine groups which are perpendicular cones of the ICP-MS. To prevent the formation of such deposi-
to the backbone, gives valuable insight into the segmental tions, (NH4)2CO3 solution was chosen as a proper eluent for
orientation of Si–O chains in the elongation direction of the lm. chromatographic speciation. The effect of ow rate of mobile
The degree of uniaxial orientation of polymer chains can be phase and addition of an organic modier such as 1 and 2%
expressed by the dichroic ratio (R).27 The dichroic ratio is (v/v) methanol into the buffer solutions as well as various
concentrations of (NH4)2CO3 were evaluated for optimization of
chromatographic conditions in terms of peak intensity, sepa-
ration and peak symmetry (results are not shown). In the nal
method, 10.0 and 30.0 mM (NH4)2CO3 (pH adjusted to 8.50)
with 1.0 mL minꢁ1 ow rate was used in separation of the
species. The elution order of the species from the strong anion
exchange column is DMA, MMA and As(V). The retention
mechanism of the species was strongly controlled by the pH of
the mobile phase and dissociation of acidic species. The pH
dependent dissociations of arsenic species are given in Table 2.
Depending on the ionic character of the column and the pH of
the eluting buffer (pH 8.50), it can be suggested that each
arsenic species is eluted in accordance with its charge.
Fig. 1 SEM micrographs of amino-functionalized PDMS (a) dip coated, and (b)
electrospun film coated SPME fibers.
22264 | RSC Adv., 2013, 3, 22261–22268
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
Table 2 Arsenic species used throughout the study
As species
Structure
pKa
Reference
30
2.3
6.8
11.6
Arsenate As(V)
3.6
8.2
Monomethylarsonic
acid MMA
31
31
Dimethylarsinic
acid DMA
6.3
The most common interference regarding the determination
of arsenical by ICP-MS is potential polyatomic interference due
to the formation of Ar40Cl35 in the plasma at the same m/z: 75
during the measurement of monoistopic As75 by HPLC-ICPMS.
In this study, the octopole collision cell of the ICP-MS system
was used with He as the collision gas to overcome the potential
interference of chloride ion on arsenic signal. In addition, the
Fig. 3 Effect of mole ratio of APTES/PDMS in the sol–gel solution on extraction
of the As(V) in (a) dip coated fibers, (b) electrospun coated fibers (extraction
conditions; extraction time: 30 min, As concentration: 10.0 mg Lꢁ1, solution pH:
elution peak of Clꢁ ion was determined by injection of 150 mL, 5.0, stirring speed: 700 rpm, solution volume: 15 mL, desorption conditions;
desorption time: 20 min, 150 mL 50.0 mM KH2PO4). Coating volumes are ca. 0.054
mm3 per fiber (dip coated) and 0.020 mm3 per fiber (electrospun coated).
0.010 M NaCl solution into HPLC-ICPMS (Fig. S1†) under opti-
mized chromatographic run conditions in presence and
absence of collision gas. The Clꢁ ion peak was appeared around
500 s in the chromatograms obtained at m/z: 37 (Fig. S1(a) and
(c)†) both in absence and presence of collision gas, respectively.
Hydrolysis of the precursor, the reaction between hydrolyzed
product and PDMS, and schematic representation of electro-
Chromatogram acquired at m/z: 75 (Fig. S1(b)†) shows a small
peak related to Ar40Cl35 which was eliminated by addition of He
spinning set-up for SPME coating are demonstrated in Fig. 5.
The amine groups in the polymeric sol–gel solution are positively
gas into octopole collision cell (Fig. S1(d)†).
charged and it can be speculated that these functional groups
3.3 Optimization of ber coating
The amine-functionalized PDMS coated SPME bers prepared
by both dip-coating and electrospinning were used for the
extraction of arsenicals. The results were evaluated for
comparison of the coating methods in terms of extraction
performances. The variation of amine content in sol–gel solu-
tion was also considered in the same study. Based on the results
(Fig. 3), a signicant difference in extraction of arsenicals was
observed between these methods, and superior extraction
properties were obtained by electrospun lm coated bers.
Under the consideration of the comparable coating thickness
and sol–gel solution, the only difference between these two
methods is the application of the electric eld in the electro-
spinning method. A typical chromatogram obtained aer
application of proposed method (SPME-HPLC-ICPMS) is shown
Fig. 4 A typical chromatogram obtained after application of proposed SPME-
in Fig. 4. The extraction was performed with electrospun coated
HPLC-ICPMS method (conditions are as given in Fig. 3, with electrospun coated
bers from water containing 10.0 mg Lꢁ1 of each analyte.
fiber).
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22261–22268 | 22265

View Article Online
RSC Advances
Paper
adsorption of analytes via electrostatic interaction through
protonated amine groups. Additionally, under neutral and
slightly basic pHs interaction of free amine groups with analytes
should also be considered. Based on experimental ndings and
expected mechanisms of extractions, at neutral and lower pH's
the extracted amounts of the analytes depend strongly on the
interaction between the weak anion exchanger surface of
the SPME coating (protonated amine groups) and charge of
the species. As it is well known, the weak anion exchangers are
positively charged under low pH and are un-ionized under high
pH conditions and this explains the reason for the lower
extraction of As(V) above neutral pHs. Moreover, the correlation
between the peak area and interaction of the species with the
ber shows also the tendency of the extraction. The most and
the least extracted species were As(V) and DMA, correspond-
ingly. This nding is in agreement with the elution order of the
analytes from strong anion exchange column used throughout
the study which also supports that the extraction is mainly
through anion exchange in nature.
3.4.2 Effect of agitation time/speed on extraction of arsenic
species. The effect of interaction time on extraction of the
arsenic species is given in Fig. S3.† For DMA and MMA,
increasing the contact time upto 30 min resulted in an increase
of the extracted amounts. The extraction capability decreased if
longer times were used. For As(V), the extracted amount
increased gradually upto 60 min and decreased aerwards. A
similar trend in sorption of As(V) was also observed with (NH2)
silicate in previous study.32 Results indicate that the contact
time of the ber and analyte solution is one of the most
important optimization parameters. For the subsequent
studies, 30 min was chosen as extraction time.
The effect of stirring speed of the solution during the
extraction of arsenic species was also investigated. The results
are given in Fig. S4.† The most important outcome of the study
was the inconsistency of the extracted amount of the arsenic
species at various stirring speeds. According to the fundamen-
tals of SPME, increasing the agitation speed decreases the
thickness of boundary layer between coating (surface) and bulk
of the solution. As a result, diffusion path decreases and this
improves mass transfer. It is worth mentioning that the agita-
tion only affects the kinetics of the extraction rather than the
thermodynamics associated with the SPME process. This means
that the maximum amount extracted at equilibrium will not be
affected by agitation. The maximum extraction was obtained for
As(V) and MMA at 700 rpm. When the solutions were stirred at
1000 rpm extracted amount of the arsenic species were slightly
decreased due to formation of vortex in the solution. The
Fig. 5 Schematic diagram showing proposed design for SPME coating by
combination of sol–gel chemistry and electrospinning.
which are under the inuence of the electric eld are oriented
normal to the ber surface. They are, therefore, more available
for the extraction of arsenic species than the functional groups
randomly distributed though the matrix where no electrical eld
is applied (dip coating method). The bers coated by electro-
spinning method were used in subsequent studies.
3.4 Optimization of extraction parameters
3.4.1 Effect of pH on extraction of arsenic species. The formed vortex possibly decreased the interaction of the analytes
optimization study was started with the determination of pH of in the solution with the ber. Another important point that
the solution where the maximum extraction of each species was should be considered is the extraction of analytes under static
obtained. Among the tested pH conditions, the highest and the extraction mode. The obtained peak areas for each analyte at
lowest extractions of As(V) were obtained at pH 5.0 and 9.0, static extraction were comparable to each other. However the
respectively. On the other hand, the extractions of MMA and extraction of DMA decreases signicantly by application of
DMA were particularly independent of pH of the solution agitation. In order to explain this unexpected behavior of DMA
(Fig. S2†). The coating, PDMS-weak anion exchanger, is a kind some important aspects of SPME theory should be considered.
of mix-mode sorbent, and two types of interactions were As described in the study, two main criteria are important for
expected; absorption of analytes by PDMS (nonpolar) and the extraction of the analytes; diffusion in the sample and
22266 | RSC Adv., 2013, 3, 22261–22268
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
electrostatic interaction of analytes with the surface of coating. is expected when the coating surface is positively charged and
First of all, the diffusion of the analyte from bulk of the solution analytes are negatively charged. The negative charge on the As(V)
to the surface of coating should be evaluated. In order to explain is higher than that of MMA and the lowest charge is on DMA.
diffusion behavior of the analytes, diffusion coefficient (D) of This fact supports the strongest interaction between As(V) and
each analyte should be considered. It is worth mentioning that coating, and higher extracted amount of this species. It should be
the dissociation of each analyte under various pHs results in reminded that the interaction between the analytes and the
different diffusion coefficients. In the current study all analytes coating is mainly based on adsorption. Considering the limited
under consideration were able to dissociate and different form number of available adsorption sites on the surface, a displace-
of the species can be found in the sample depending on the pH ment can occur.34 In this case As(V), with the strongest affinity to
of solution. For instance, the neutral form of As(V) is H3AsO4, the coating may displace especially DMA, which has the lowest
and it is formed under acidic conditions. As pH of solution affinity to the coating. On the other hand, the displacement effect
increases deprotonation occurs and the other ionic species is not observed when extraction time is signicantly shorter than
ꢁ
become predominant, such as H2AsO43ꢁ. With further increase equilibrium time.35 In the present study with a comparatively
of the pH, HAsO42ꢁ, and nally AsO4 are formed. For MMA short static extraction time (30 min), no displacement effect was
the formed species in aqueous solution are CH3H2AsO3, expected and the results of DMA under static mode and agitation
2ꢁ
CH3HAsO3ꢁ, CH3AsO3 and for DMA are (CH3)2HAsO2 and conditions supported this expectation.
(CH3)2AsO2ꢁ (ordered from low to high pHs). The pH dependent
3.4.3 Effect of salt concentration (ionic strength) on
distribution of species for each arsenical compound can be extraction of arsenic species. The effect of NaCl concentration
determined from pKa values shown in Table 2. As pH of the on extraction of arsenic species was studied in 1.0 M, 0.10 M, and
solution increases the charge on the dissociated molecule 0.010 M NaCl solutions. The results given in Fig. S5† indicate that
increases, which produces stronger electrostatic interaction increasing the ionic strength of the solution results in a decrease
between the dissociated molecule and water molecules and in the extracted amount of each arsenic species. It can be spec-
results in higher hydration level. As hydration level of a ulated that a decrease in the amount of the extracted analytes is
compound increases, diffusion coefficient decreases.33 Regarding related to competitive sorption of Clꢁ ions by protonated amine
the As(V) and MMA, it is possible to predict the diffusion coeffi- groups. The amine groups were thought to be converted into
+
cients for each analyte, such as As(V) and MMA. The extraction surface inactive aminium salt (R–NH3 Clꢁ) which possibly
pH chosen in this study was pH 5.0, where the predominant precluded the extraction of the arsenic species in the presence of
ꢁ
species for As(V) is H2AsO4 (D: 8.12 ꢄ 10ꢁ6 cm2 sꢁ1) and for chloride ion. Since amine groups are known as weak anion
MMA is CH3HAsO3 (D: 7.78 ꢄ 10ꢁ6 cm2 sꢁ1).33 Although a exchangers, any stronger anion than arsenicals will be extracted
ꢁ
diffusion coefficient for DMA is not reported in the aforemen- by the coating in a competitive manner.
tioned study, it can be predicted with respect to MMA. DMA is
3.4.4 Effect of solution temperature on extraction of
almost neutral at pH 5.0 and should have the lowest hydration arsenic species. In the headspace microextraction of volatile
level. Therefore, comparable or slightly higher diffusion coeffi- compounds increasing the solution temperature increases the
cient for DMA can be expected compared to that of MMA. As a amounts of the extracted analytes36 by altering the equilibrium
conclusion, in the absence of agitation (under static extraction concentrations of analyte in the headspace and solution. In case
conditions), the mass transfer rate depends on the diffusion of direct mode microextraction, a prediction of the effect of the
coefficients of analytes; which is the rate determining step. temperature on the extraction of analyte is not easy. It depends
Although there are some differences in diffusion coefficients of on the exothermic or endothermic nature of the extraction as
analytes, still similar extraction amount can be expected. This well as on the volatility of the analytes. In this study, the general
phenomenon can explain the similar extraction behavior of trend in the effect of the solution temperature on the extraction
analytes observed under static extraction mode.
of As(V), DMA and MMA is illustrated in Fig. S6.† Increasing the
On the other hand, in order to explain the observed decrease extraction temperature decreases the amount of extracted ana-
on extraction of DMA by application of agitation, the affinity of lytes. This nding shows the exothermic nature of the extraction
analytes to coating should be taken into account. Under agita- of the analytes by the bers. The effect of the temperature was
tion conditions, the mass transfer is faster for all of the analytes more signicant especially for DMA and MMA. It was observed
and extraction amounts at equilibrium do not depend on that as the temperature was increased, the extraction decreased.
diffusion in the sample. At equilibrium, analytes with high For further studies, 20 ꢀC was chosen as optimum extraction
affinity to coating are extracted more than the analytes with temperature.
lower affinity. Firstly, it should be mentioned that PDMS does
not contribute to the extraction of analytes, and the main
mechanism of extraction, as explained before, is electrostatic
4 Conclusions
interaction between anionic analytes and protonated amine In this study, the sol–gel based mixed mode (PDMS-weak anion
group (cationic). In terms of affinity of analytes to the coating, exchanger) SPME bers were used for direct mode extraction
the higher extraction amount is expected for analytes with of DMA, MMA and As(V) from aqueous solutions followed by
higher distribution constant Kfs (distribution of analytes HPLC-ICPMS determination. Two different approaches for sol–
between ber and sample). In this sense and according to the gel based coating of the bers were used, namely dip coating
extraction mechanism mentioned above, the highest extraction and electrospinning of amino functionalized sol–gel solution.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22261–22268 | 22267

View Article Online
RSC Advances
Paper
The latter one was used for the production of sol–gel based
SPME bers. Coating experiments revealed that the bers
coated by electrospinning process showed a superior perfor-
mance than dip coated bers. The origin of this improvement is
seen the alignment of PDMS chains under electrical force
7 E. Turiel, J. L. Tadeo and A. Martin-Esteban, Anal. Chem.,
2007, 79, 3099.
8 H. Bagheri, A. Mir and E. Babanezhad, Anal. Chim. Acta,
2005, 532, 89.
9 J. C. Wu and J. Pawliszyn, Anal. Chim. Acta, 2004, 520, 257.
during the electrospinning process; thus, the further availability 10 K. P. Huang, G. R. Wang, B. Y. Huang and C. Y. Liu, Anal.
of pendant amino groups inevitably orientated perpendicular to Chim. Acta, 2009, 645, 42.
the surface of the bers. The extracted amounts of analytes were 11 T. D. Ho, A. J. Canestraro and J. L. Anderson, Anal. Chim.
shown to be affected from various parameters: solution pH, Acta, 2011, 695, 18.
ionic strength, and extraction time. Due to weak anion 12 F. Zhao, Y. J. Meng and J. L. Anderson, J. Chromatogr., A,
exchange character of the coating, the optimum extraction pH 2008, 1208, 1.
was found to be 5.0. However, the results are promising for a 13 F. S. Mirnaghi, Y. Chen, L. M. Sidisky and J. Pawliszyn, Anal.
relatively wide range of pH from 3.0 to pH 7.0, which show Chem., 2011, 83, 6018.
potential for various water remediation. Optimum extraction 14 N. Horzum, E. Boyaci, A. E. Eroglu, T. Shahwan and
time and agitation speed were determined as 30 min and M. M. Demir, Biomacromolecules, 2010, 11, 3301.
700 rpm, respectively. Negative effect of the ionic strength on 15 N. Horzum, M. M. Demir, M. Nairat and T. Shahwan, RSC
the extraction was observed by the addition of NaCl due to Adv., 2013, 3, 7828.
formation of aminium salt. Moreover, it was observed that the 16 M. M. Demir, N. Horzum, B. Ozen and S. Ozçelik, J. Phys.
extraction of arsenical compounds is exothermic in nature. The Chem. B, 2013, 117, 10920.
proposed SPME ber coating method can offer a very conve- 17 J. W. Zewe, J. K. Steach and S. V. Olesik, Anal. Chem., 2010,
nient way to develop efficient and selective sorbents for deter- 82, 5341.
mination and removal of wide range of undesirable 18 H. Bagheri and A. Aghakhani, Anal. Methods, 2011, 3, 1284.
¨
¨
a
compounds in environmental eld simply by replacing the 19 H. Bagheri and A. Aghakhani, Anal. Chim. Acta, 2012, 713, 63.
sol–gel active silane coupling reagent with another functional 20 H. Bagheri, A. Aghakhani, M. Baghernejad and
group. The proposed coating method is superior in terms of
availability of free functional groups on surface, achieving very 21 G. Odian, Principles of Polymerization, John Wiley & Sons,
thin coating and homogeneous distribution of active layer on Inc., 2004.
the surface. Beside the present study outlined, this novel 22 F. Abbasi, H. Mirzadeh and A. A. Katbab, Polym. Int., 2001,
approach can be a pioneer for further development of novel 50, 1279.
coatings and surface modications in various elds of chem- 23 Y. Huang, S. S. H. Ho, K. F. Ho, S. C. Lee, Y. Gao and
A. Akbarinejad, Anal. Chim. Acta, 2012, 716, 34.
istry as well as other disciplines.
N. S. Y. Feng, Anal. Methods, 2012, 4, 277.
24 J. C. McDonald, D. C. Duffy, J. R. Anderson, D. T. Chiu,
H. K. Wu, O. J. A. Schueller and G. M. Whitesides,
Electrophoresis, 2000, 21, 27.
Acknowledgements
This work was nancially supported by The Scientic and 25 D. Q. Xiao, H. Zhang and M. Wirth, Langmuir, 2002, 18, 9971.
¨ ˙
Technological Research Council of Turkey (TUBITAK) through 26 A. S. M. Chong and X. S. Zhao, J. Phys. Chem. B, 2003, 107,
¨
¨
the project 108T798. The authors thank Dr Huseyin OZGENER
12650.
for FTIR measurements, Erica A. Souza Silva for useful discus- 27 M. V. Kakade, S. Givens, K. Gardner, K. H. Lee, D. B. Chase
sions regarding SPME results, the Center of Material Research and J. F. Rabolt, J. Am. Chem. Soc., 2007, 129, 2777.
´
for SEM facility, Environmental Research Center for HPLC-ICP- 28 J. E. Mark and B. Erman, Rubberlike Elasticity. A Molecular
˙
MS measurements at Izmir Institute of Technology, and HES
Primer., John Wiley & Sons, New York, 1988.
Kablo for their kind donation of ber optic cables.
29 M. M. Demir, B. Ozen and S. Ozcelik, J. Phys. Chem. B, 2009,
113, 11568.
30 E. H. Larsen and S. H. Hansen, Mikrochim. Acta, 1992, 109,
Notes and references
47.
1 Z. Y. Zhang, M. J. Yang and J. Pawliszyn, Anal. Chem., 1994, 31 M. Calatayud, J. Gimeno, D. Velez, V. Devesa and
66, A844. R. Montoro, Chem. Res. Toxicol., 2010, 23, 547.
2 J. Dugay, C. Miege and M. C. Hennion, J. Chromatogr., A, 32 E. Boyaci, A. Cagir, T. Shahwan and A. E. Eroglu, Talanta,
1998, 795, 27. 2011, 85, 1517.
3 K. Farhadi, R. Tahmasebi and R. Maleki, Talanta, 2009, 77, 33 M. Tanaka, Y. Takahashi, N. Yamaguchi, K. W. Kim,
1285.
G. D. Zheng and M. Sakamitsu, Geochim. Cosmochim. Acta,
2013, 105, 360.
34 Handbook of Solid Phase Microextraction, ed. J. Pawliszyn,
Chemical Industry Press, Beijing, 2009.
4 T. P. Gbatu, K. L. Sutton and J. A. Caruso, Anal. Chim. Acta,
1999, 402, 67.
5 W. M. Liu, Y. A. Hu, J. H. Zhao, Y. A. Xu and Y. F. Guan, J.
Chromatogr., A, 2006, 1102, 37.
35 J. Pawliszyn, J. Chromatogr. Sci., 2000, 38, 270.
6 F. Augusto, E. Carasek, R. G. C. Silva, S. R. Rivellino, A. D. Batista 36 J. Pawliszyn, Application of Solid Phase Microextraction, The
and E. Martendal, J. Chromatogr., A, 2010, 1217, 2533.
Royal Society of Chemistry, Cornwall, UK, 1999.
22268 | RSC Adv., 2013, 3, 22261–22268
This journal is ª The Royal Society of Chemistry 2013