Analytica Chimica Acta 953 (2017) 23e31

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Ionic liquids intercalated in montmorillonite as the sorptive phase for

the extraction of low-polarity organic compounds from water by
rotating-disk sorptive extraction
Jhon A. Fiscal-Ladino a, Mo
nica Obando-Ceballos a, Milton Rosero-Moreano a,
Diego F. Montan ~ o , Wilson Cardona b, Luis F. Giraldo b, Pablo Richter c, *
b

a
n en Cromatografía y T

Grupo de Investigacio ecnicas Afines GICTA, Departamento de Química, Facultad de Ciencias Exactas y Naturales, Universidad de
Caldas, Calle 65 No. 26-10, Manizales, Colombia
b
Química de Plantas Colombianas, Instituto de Química, Escuela de Ciencias Exactas y Naturales, Universidad de Antioquia, Calle 70 No. 52-21, A.A, 1226,
Medellín, Colombia
c
Departamento de Química Inorga
nica y Analítica, Facultad de Ciencias Químicas y Farmac
euticas, Universidad de Chile, Casilla 233, Santiago, Chile

h i g h l i g h t s g r a p h i c a l a b s t r a c t

Clay modified with ILs successfully

extracts PCBs from aqueous
solutions.

A new RDSE sorbent is applied in the
analysis of environmental samples.

A simple and green method is
developed.

MMT-HDMIM-BR phase was more
efficient than other commercial solid
phases.

The novel MMT-HDMIM-BR sorbent
is promising for determination of
other low polarity analytes.

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

Article history: Montmorillonite (MMT) clays were modified by the intercalation into their galleries of ionic liquids (IL)
Received 10 August 2016 based on imidazolium quaternary ammonium salts. This new eco-materials exhibited good features for
Received in revised form use as a sorptive phase in the extraction of low-polarity analytes from aqueous samples. Spectroscopic
26 October 2016
analyses of the modified clays were conducted and revealed an increase in the basal spacing and a
Accepted 24 November 2016
shifting of the reflection plane towards lower values as a consequence of the effective intercalation of
Available online 3 December 2016
organic cations into the MMT structure. The novel sorbent developed herein was assayed as the sorptive
phase in rotating-disk sorptive extraction (RDSE), using polychlorinated biphenyls (PCBs), representative
Keywords:
Montmorillonite
of low-polarity pollutants, as model analytes. The final determination was made by gas chromatography
Ionic liquid with electron capture detection. Among the synthetized sorptive phases, the selected system for
Rotating-disk sorptive extraction analytical purposes consisted of MMT modified with the 1-hexadecyl-3-methylimidazolium bromide
PCBs (HDMIM-Br) IL. Satisfactory analytical features were achieved using a sample volume of 5 mL: the
GC-ECD relative recoveries from a wastewater sample were higher than 80%, the detection limits were between
3 ng L1 and 43 ng L1, the precision (within-run precision) expressed as the relative standard deviation
ranged from 2% to 24%, and the enrichment factors ranged between 18 and 28. Using RDSE, the extraction
efficiency achieved for the selected MMT-HDMIM-Br phase was compared with other commercial solid

* Corresponding author.
E-mail address: prichter@ciq.uchile.cl (P. Richter).

http://dx.doi.org/10.1016/j.aca.2016.11.067
0003-2670/© 2016 Elsevier B.V. All rights reserved.
24 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
phases/supports, such as polypropylene, polypropylene with 1-octanol (as a supported liquid mem-
brane), octadecyl (C18) and octyl (C8), and showed the highest response for all the studied analytes.
Under the optimized extraction conditions, this new device was applied in the analysis of the influent
of a wastewater treatment plant in Santiago (Chile), demonstrating its applicability through the good
recoveries and precision achieved with real samples.
1. Introduction
Sample preparation is an essential part of an analytical process,
particularly in complex samples in which the analyte is very dilute
and mixed with many matrix molecules. Although modern
instrumental technology resolves a large part of this problem, an
appropriate sample preparation procedure allows the necessary
sensitivity and selectivity to be achieved for the measurement.
Together with efficient sample preparation procedures, it is also
important to develop novel analytical technologies that incorporate
the concept of eco-efficiency. Microextraction technologies satisfy
this requirement; attaining high levels of analytical efficiency
through the use of small amounts of dangerous reagents [1].
Since the appearance of solid-phase microextraction [2] in the
early 1990s, a number of other microextraction technologies have
been developed [3]. A critical aspect of extraction technology is the
sorptive phase, considering its affinity for the analyte and its
selectivity [4e7]. It is well known that sorptive phases of different
polarities are commercially available for both solid-phase extrac-
tion and microextraction techniques; however, an important trend
in microextraction is the development of novel sorbents based on
the synthesis of polymers and/or modification of natural materials.
Taking into account that clays play an important role in the
retention/mobility behavior of chemical species in the soil,
different materials based on clays have been proposed for the
removal or analytical extraction of environmental pollutants [8].
Montmorillonite (MMT) layers contain negative charges that are
counterbalanced by inorganic cations (e.g., Naþ and Ca2þ). These
inorganic cations are exchangeable, making MMT an efficient
adsorbent for various cationic contaminants, such as heavy metals,
dyes, surfactants and monomers [9]. However, non-modified clays
do not adsorb low-polarity organic compounds from aqueous so-
lutions [8]. Contrarily, modified clays have been proposed as po-
tential materials for the adsorption of environmental toxicants. In
this context, it has been reported that pillared, delaminated and
hydroxyl-interlayered smectites [8] efficiently adsorb chlor-
ophenols from water in batch equilibrium experiments.
When the modification of clays is performed with organic cat-
ions, the interlayer space of MMT can change from hydrophilic to
hydrophobic, and the resulting materials, also known as organo-
clays, have been used as efficient adsorbents for hydrophobic
organic contaminants [8]. The modification of MMT with cationic
surfactants has been proposed for the removal of bisphenol A from
wastewater [10]. Similarly, a host of conducting polyaniline poly-
mers has been combined with lamellar MMT clay as novel sorbents
for use in the solid-phase microextraction (SPME) of phenolic
compounds [8]. The enhanced sorbent capacity originates from the
properties of the composites, including the large surface area and
homogeneous dispersion. With good linearity and reasonable
repeatability of the response, the extraction device used in that case
exhibited more convenient handling than the other traditional
SPME fibers.
Clays modified with ionic liquids (ILs) are another alternative
and have been recently evaluated as an alternative to other
© 2016 Elsevier B.V. All rights reserved.
materials used to impart organophilic properties for nano-
composite applications [11e13]. The increased basal spacing of
nanoclay improves the attraction of organic analytes for nano-
composite applications. The intercalation of ILs in MMT through
exchange with interlamellar sodium ions has been confirmed by
FTIR, TGA and XRD.
Recently, in the field of microextraction techniques, ILs with
different chain lengths, namely 1-methyl-3-octyl-imidazolium
bromide, 1-methyl-3-undecyl-imidazolium bromide and 1-methyl-
3-octadecyl-imidazolium bromide, were intercalated into the gal-
leries of montmorillonite (MMT) clay and used as solid supports in
the micro-solid-phase extraction of estrogenic hormones for the
first time. The materials developed in that study possess the
advantage of hydrophobic ILs intercalated into the polar structure
of MMT and represent a new organiceinorganic hybrid solid-phase
extraction sorbent system for sample preparation in analytical
applications [14].
Rotating-disk sorptive extraction (RDSE) was developed in 2009
by our research group. This sample preparation technique repre-
sented an alternative to the microextraction/cleanup techniques
available at that time, and a number of advantages and applications
have since been described [6,15e24]. In addition to being a very
simple, relatively rapid and inexpensive approach, other advan-
tages of RDSE are as follows: (1) the design of the extraction device
enables a high surface-area-to-volume ratio; (2) the extractions are
carried out on small sample volumes; (3) the recirculating regime
prevents the collapse of the filter in complex samples, allowing
continuous contact between the solid and liquid phases; (4) the fact
that the extraction phase is only in contact with the liquid sample
permits one to stir at high speeds; and (5) the sorptive phase is
easily replaceable, allowing the use of either commercial, natural or
synthesized sorbents.
In this study, a novel application of RDSE associated with a new
sorptive phase based on an IL intercalated in MMT is presented.
Polychlorinated biphenyls (PCBs) in water samples were selected as
the analytical model system, considering the low polarity of these
analytes (Log Kow values range from approximately 4 to 8). These
organic pollutants were widely used in the past for different pur-
poses and, as a result of their persistence, can still be found in the
environment. The modified material that we prepared in the pre-
sent study was synthesized using MMT and an IL, 1-hexadecyl-3-
methylimidazolium bromide. Considering the hydrophobicity of
the hexadecyl moiety in the IL, efficient extraction of PCBs from
water samples is expected.
2. Experimental
2.1. Reagents
Water from a Millipore Milli-Q Plus water system (Billerica, MA,
USA) was used throughout the experiment. Sonoita was the com-
mercial name of the MMT used in this study, which was obtained
from Bentocol (Bugalagrande, Valle del Cauca, Colombia). Methyl
imidazole (99%), 1-bromobutane (97%), 1-bromooctane (97%), 1-
 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
bromohexadecane (97%), ethyl acetate, sodium hydroxide,
dichloromethane, sodium tetrafluoroborate and ethanol were all of
reagent grade, obtained from Sigma Chemical Co. (St. Louis, MO,
USA) and used for the synthesis of the ionic liquids (ILs): the 1-
hexadecyl-3-methylimidazolium series (1-hexadecyl-3-
methylimidazolium bromide (HDMIM-Br), 1-hexadecyl-3-
methylimidazolium hydroxide (HDMIM-OH) and 1-hexadecyl-3-
methylimidazolium tetrafluoroborate (HDMIM-BF4)), 1-butyl-3-
methylimidazolium series (BMIM-Br, BMIM-OH and BMIM-BF4)
and 1-octyl-3-methylimidazolium series (OMIM-Br, OMIM-OH and
OMIM-BF4). Methanol (Merck, Darmstadt, Germany) and hydro-
chloric acid (Merck) were used for the intercalation of MMT with
the ILs. Methanol was also used in the rotating-disk sorptive
extraction experiments as a matrix modifier and desorption agent.
The hollow-fiber (HF) phase used in comparative studies was made
of polypropylene, Accurel PP 300/1200 acquired from Membrana
GmbH Polypore Cia (Wuppertal, Germany). 1-Octanol from Merck
was used to prepare the liquid membrane. The film thickness was
300 mm with a pore size of 0.2 mm and an internal diameter of
1.2 mm. The specific surface area was 17.3 m2 g1. The C18 and C8
sorbents were obtained from Waters Corp. (Milford, MA, USA).
PCB standards (PCB 81, PCB 105, PCB 114, PCB 123, PCB 126, PCB
156, PCB 157, PCB 167, PCB 169, PCB 189) (99% purity) were pur-
chased from AccuStandard (New Haven, CT, USA). Solutions of each
PCB were prepared in ethyl acetate at 100 mg mL1, and a mixed
standard with each compound at 1 mg mL1 in ethyl acetate was
used as the working standard. Hexachlorobenzene (HCB) was
purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany) and
used as the internal standard.
Nitrogen with a purity of 99.995% was purchased from Linde
(Santiago, Chile) and used in the evaporation of the final extract for
RDSE and as a chromatographic carrier gas.
2.2. Instruments
The ionic liquid HDMIM-Br was characterized by 1H and 13C
NMR, and the spectra were recorded from DMSO-d6 solutions on an
AMX 300 spectrometer (Bruker, Billerica, MA, USA) operating at
300 MHz for 1H and 75 MHz for 13C. The chemical shifts (d) are
expressed in ppm, with the solvent peak as the reference and TMS
as the internal standard; the coupling constants (J) are given in
Hertz (Hz).
X-ray diffraction (XRD) experiments were performed using an
XPert PANalytical Empyrean diffractometer (EA Almelo,
Netherlands). The 2Ө values were obtained from a step size of
0.0394 using CuKa radiation (l of 1.540598 Å). The samples
(100 mg) of raw MMT and modified MMT prepared with different
ILs were placed on a plate with an area of 2 cm2 to minimize the
error in the peak location, and the crystal structure of the clays can
be obtained from these data, also providing the inter-planar space.
The MMT-HDMIM-Br sorptive phase was also characterized by
Fourier transform infrared spectroscopy (FTIR), thermogravimetric
analysis (TGA) and scanning electronic microscopy (SEM). The
infrared spectra were recorded in the range of 4000e400 cm1 on
an FTIR model IRAffinity-1 (Shimadzu, Kyoto, Japan) from KBr
pellets. The TGA spectra were obtained using a Q500 thermogra-
vimetric analyzer from TA Instruments (New Castle, DE, USA), and
the heating rate was 10  C/min from room temperature to 800  C in
air. SEM experiments were performed using an FEI Quanta 250
microscope (OR, USA).
Quantification of PCBs was performed using a HewlettePackard
gas chromatograph, model 5890 series II (Palo Alto, CA, USA),
equipped with an electron-capture detector and fused silica capil-
lary column (Restek RTX-5MS, Bellefonte, PA, USA) (Rtx®-5 ms,
30 m  0.25 mm  0.25 mm film thickness). One microliter of the
25
extract was injected into the gas chromatograph in splitless mode.
The injector and detector temperatures were 225  C and 310  C,
respectively. The initial oven temperature was 100  C, which was
maintained for 2 min, followed by heating to 300  C at a rate of
20  C min1 (7 min). The chromatographic run required a total time
of 19 min using a carrier gas flow rate of 1.5 mL min1.
2.3. Preparation of the sorbent material
2.3.1. Clay material and purification
The MMT used in this work was treated with water to eliminate
the quartz fraction. A portion of 10 g of MMT was treated with
200 mL of water and 150 mL of 5% (w/w) sodium polyphosphate.
The mixture was vigorously stirred for 2 h, and after this time, the
suspended fraction was separated. After this treatment, the clay
was dried in an oven at 60  C for 12 h. The clay used in this study
was a phyllosilicate material with a general chemical formula of
(Si4)IV(Al4)VIO10(OH)8.
2.3.2. Synthesis of the ionic liquid
The ionic liquid HDMIM-Br was synthesized according the
procedure described by Min et al. [25]. Briefly, the procedure
involved mixing equimolar amounts of methyl imidazole (5.0 g,
61 mmol) and 1-bromohexadecane (18.3 g, 61 mmol) in a reaction
flask in ethyl acetate (50 mL). The mixture was vigorously stirred at
reflux for 48 h, and after this time, the solution was evaporated to
remove the solvent and obtain HDMIM-Br as a pale yellow solid.
This compound was characterized by 1H NMR and 13C NMR.
1
H NMR (DMSO-d6, 300 MHz): d 0.86 (3H, t, J ¼ 7.07 Hz),
1.19e1.27 (24H, m), 1.28e1.36 (2H, m) 1.80e1.98 (2H, m), 4.12 (3H,
s, NeCH3), 4.31 (2H, t, J ¼ 7.43 Hz, NeCH2e), 7.45 (1H, d, J ¼ 1.51 Hz),
7.64 (1H, d, J ¼ 1.51 Hz), 10.25 (1H, s).
13
C NMR (DMSO-d6, 75 MHz): d 14.07 (CH3), 22.65 (CH2), 26.24
(CH2) 28.98 (CH2) 29.31 (CH2) 29.36 (CH2) 29.48 (CH2) 29.57 (CH2)
29.61 (3CH2), 29.65 (3CH2) 30.30 (CH2), 31.84 (NeCH3), 50.15
(NeCH2), 121.97 (CH-Im), 123.77 (CH-Im), 137.29 (CH-Im).
The other ILs (1-hexadecyl-3-methylimidazolium hydroxide, 1-
hexadecyl-3-methylimidazolium tetrafluoroborate, 1-butyl-3-
methylimidazolium-based [26,27] and 1-octyl-3-
methylimidazolium-based) were synthetized following a similar
procedure.
2.3.3. Modification of MMT
One gram of MMT was dispersed in a 3.0 g of a solution of
HDMIM-Br in methanol at 13% v/v, and the dispersion was vigor-
ously stirred for 1 h at room temperature. The mixture was filtered,
and the solid material was washed three times with methanol
(20 mL each time) to remove the excess ionic liquid and then three
times with distilled water. The solid was dried in an oven for 24 h at
105  C. MMT samples modified with the other ILs were prepared
following the same procedure.
2.4. Rotating-disk sorptive extraction
The extraction device used included a Teflon disk (1.5 cm
diameter) containing an embedded miniature magnetic stirring bar
(Teflon-coated micro stir bar from VWR International). The disk had
a 0.44 cm3 cavity on one of its surfaces, into which 10 mg of MMT-
HDMIM-Br sorbent was loaded. The cavity was covered with a
fiberglass filter (1.4 cm diameter, 3 mm mean pore size) and sealed
with a Teflon ring (Fig. 1a). Before extraction, the sorptive phase
was conditioned with ethyl acetate, methanol and Milli-Q water for
5 min each.
Comparative studies using other sorptive phases, including
octadecyl (C18), octyl (C8) and MMT without modification (10 mg),
26 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31

were performed using the same disk (Fig. 1a). In addition, a hollow
fiber made of polypropylene (PP) and a hollow fiber embedded
with 1-octanol by occupying the HF pores as a supported liquid
membrane (SLM) (PP with 1-octanol as a SLM) were also immo-
bilized on the disk cavity for comparative studies (Fig. 1b).
The beaker containing the samples (5 mL) and the rotating disk
was placed on an MMS-3000 Boeco magnetic stirrer (Hamburg,
Germany). An analog heat-block evaporator (VWR, USA) was
employed during the extraction process following the RDSE
procedure.
The desorption procedure for all sorbents used was with 5 mL of
methanol, followed by a concentration step with nitrogen to total
dryness and reconstitution in 100 mL of methanol.
2.5. Real samples
A sample was taken from the inflow of a water treatment plant
in Santiago (Chile) in September 2015. A portion of the sample was
spiked with 20 ng mL1 of the target compounds to demonstrate
the extraction capacity of the new device and was then analyzed by
GC-ECD.
3. Results and discussion
the clay was probably modified by the IL, which intercalated into
the polar structure of MMT. These bands are due to the stretching of
the CH bonds present in the methylene long chain in the ionic
liquid. Kim et al. [12] reported similar spectra when MMT was
modified with the following ILs: 1-ethyl 3-methylimidazolium
bromide, 1-hexyl 3-methylimidazolium chloride and N-ethyl-
pyridinium tetrafluoroborate.
The TGA spectra of the non-modified MMT revealed weight
losses in the interval of 25e150  C due to release of the adsorbed
water intercalated in the clay; the release of bonded water occurred
gradually between 250 and 450  C. The dehydroxylation of clay
minerals occurs over a wide range of temperatures (250e900  C),
as was reported by Liptay et al. [28]. The spectrum of MMT-
HDMIM-Br reveals an evident loss of the IL that was intercalated
in the clay at approximately 250  C (Fig. 3).
The micrographs of MMT-HDMIM-Br (Fig. 4) reveal clusters of
particles having a narrow size distribution with diameters of
approximately 25 mm.
As expected, the XRD experiments (Fig. 5) show that the ion
exchange of Naþ in MMT by HDMIM-Br led to an increase in the
spacing of the d001 reflections in comparison with pure MMT and
that such spacing was related to the length of the alkyl chain on the
HDMIM cation. The spacing results are reported in Table 1.

XRD analyses were performed on the raw MMT and modified
MMT prepared with different ILs to establish whether the exchange
of its interlayer sodium cations with the long alkyl chains of qua-
ternary ammonium salts affects the basal spacing of MMT and fa-
cilitates the extraction of low-polarity organic analytes. As shown
in Table 1, the MMT-HDMIM (Br, OH, and BF4) series exhibited
higher d-value increases with respect to the raw MMT, and the
reflection peaks 2Ɵ ( ) of plane (001) were displaced towards lower
values as a consequence of the intercalation process in accordance
with Bragg’s law, with a straight correlation with the materials that
exhibited the best organic compound retention [14]. A longer car-
bon chain intercalated into the clay galleries afforded great
extraction capability that was comparable to commercial C18 sor-
bent. The MMT-HDMIM-Br sorptive phase was selected for further
studies.
3.1. Characterization of the sorbent material
Typical characterization analyses were carried out on the
selected phase, including FT-IR, TGA, SEM and XRD. When the FTIR
spectrum of MMT-HDMIM-Br was compared with that of the non-
modified clay, it can be observed that new signals appeared in the
region of the 3000 cm1 (Fig. 2) in the former case, indicating that
3.2. Study of the extraction variables
Different chemical and preconcentration factors were evaluated
to obtain the highest extraction efficiencies of the selected analytes.
In addition, the variables were studied in deionized water spiked
with a known concentration of the selected PCBs, and conse-
quently, the selectivity of this technique was able to assess the
behavior of each variable.
Matrix modifications such as NaCl (salting-out effect) or meth-
anol addition can modify the extraction efficiency or analyte re-
covery, particularly for low-polarity analytes. It has been observed
that a high concentration of methanol (20%) improves the recovery
of hydrophobic solutes by reducing adsorption onto the glass wall
of the extraction vessel [20]. Considering the high hydrophobicity
of PCBs, the effects of both reagents were investigated using con-
centrations of NaCl and methanol in intervals of 0e10% (w/v) and
0e3% (v/v), respectively, in rotating disks that contained the
sorptive phase. An increase in the concentration of NaCl increased
the response of the analyte within an interval of 0e5%, after which
the signal remains constant. Regarding methanol, the extraction
efficiency increased within an interval 0e2%. Over this range, the
extraction efficiency remained constant. Consequently, concentra-
tions of 5% and 2% for NaCl and methanol, respectively, were
selected for further studies.
The amount of MMT-HDMIM-Br immobilized into the disk was

Table 1
XRD data for MMT and MMT modified with ILs.

Sample XRD analysis

2Ө ( ) d-value (Å)

MMT 8.12 10.86

MMT-BMIM-Br 6.89 12.81
MMT-BMIM-OH 6.67 13.23
MMT-BMIM-BF4 6.78 13.03
MMT-OMIM-Br 6.40 13.80
MMT-OMIM-OH 6.28 14.07
MMT-OMIM-BF4 6.25 14.14
MMT-HDMIM-Br 5.15 17.15
MMT-HDMIM-OH 4.87 18.12
Fig. 1. Device used in RDSE. (A) device used for powder phases; (B) device used for the
MMT-HDMIM-BF4 5.13 17.23
HF phase.
 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
Fig. 2. Overlapped infrared spectra of the raw (MMT) and modified material (MMT-
HDMIM-Br).
Fig. 3. TGA analysis of the raw MMT and MMT-HDMIM-Br.
Fig. 4. SEM micrographs of the (A) raw MMT-Na clay and (B) MMT-HDMIM-Br clay.
27
Fig. 5. XRD analysis of MMT and MMT-HDMIM.
studied between 10 and 50 mg. The extraction efficiency decreased
linearly (R2 ¼ 0.91) when the amount of the sorptive phase
increased by 0.24 relative signal units per mg of sorptive phase. This
effect can be explained by the fact that the cavity tends to be
compressed when a higher amount of sorbent is loaded into it and
that its interaction with the liquid phase containing the analytes
consequently decreases.
A relevant hydrodynamic factor affecting the efficiency and
extraction time in RDSE is the rotation velocity of the disk. Efficient
stirring of the sample in contact with the extraction phase is
necessary to achieve the partition equilibrium as quickly as possible
because the mass transfer of the analyte through the boundary
layer that contacts the surface of the phase in microextraction
techniques is the rate-determining step for extraction. Conse-
quently, in this study, the effect of stirring was studied between 500
and 3000 rpm. The dependence of the extraction efficiency on the
stirring speed was linear between 500 and 2000 rpm, and the
extraction efficiency then sharply decayed at 3000 rpm (Fig. 6). In
this last condition, the vortex produced by the strong rotation of the
disk reached the surface of the device, avoiding an appropriate
mass transfer of the analytes. Hence, a rotation velocity of
28 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
2000 rpm was selected as optimum for PCB transfer.
Under the selected conditions, the extraction time profiles were
obtained between 15 and 75 min for each analyte from 5 mL sample
volumes. Fig. 7 shows the extraction profiles for PCB 81, PCB 123
and PCB 105 as representative analytes. As evident from the figure,
the amount of analyte extracted increased concomitantly with the
extraction time and reached equilibrium at approximately 30 min.
Then, the signals began to decay slowly probably because of the
evaporation of the methanol used as a modifier after long extrac-
tion periods. Leakage of the ionic liquid from the sorptive phase was
not observed after long extraction periods. In spite of proposed
determination resulted faster than other methods based on RDSE
[16,18,20,23,24], its main drawback, compared with the typical
solid phase extraction technology, is the relatively long extraction
time. This is mainly due because RDSE is an equilibrium-based
technology.
The desorption of the analytes from the disk was quantitative
when using 5 mL of methanol in a 10 min extraction at a rotation
velocity of 1000 rpm.
The retention capability of non-modified MMT was also
assessed under the same optimized conditions in order to sepa-
rately establish if the clay contributed to the overall analyte
retention. As expected, considering the polarity of the clay, the
contribution of the clay to the extraction of the determined PCBs
was negligible, reaching approximately 2% of the retention effi-
ciency of the modified clay and confirming that the presence of the
intercalated ionic liquid is mandatory for the extraction of this type
of low-polarity analyte.
3.3. Analytical features and analysis of real samples
By taking into account the selected values for the variables, the
analytical features of the method involving rotating-disk sorptive
extraction were determined using a tap water matrix. Table 2
shows the analytical features for the different analytes under study.
As evident from the results in Table 2, good accuracy and
Fig. 6. Effect of the rotation velocity of the disk on PCB extraction.
precision and a low LOD were obtained for this method. The cor-
relation between the PCB concentrations and the signal obtained
from the GCeECD were almost all near 0.99. The detection limits of
the method were determined by following the 3-s criteria and
were sufficiently low (between 2.5 and 43 ngL1) for the deter-
mination of these compounds in water. The reproducibility and
recovery were determined using different extraction disks (n ¼ 6)
with PCB sample concentrations of 1.0 mg L1 in a water sample. The
absolute recovery values ranged between 35.8 and 55.5%, with
relative standard deviations (RSDs) between 2.1 and 24.8%. The
precision was also determined by the sequential use of the same
extraction disk (containing the same sorbent phase) for various
aliquots of the same sample containing PCBs at 1.0 mg L1 (n ¼ 10).
The RSDs under these conditions were between 10 and 25%, clearly
indicating that the MMT-HDMIM-Br sorbent could be re-used ten
times. A higher number of adsorptioneelution cycles resulted in an
RSD over 30%.
The results obtained using this newly proposed MMT-HDMIM-
Br sorptive phase in RDSE were compared under the same
selected conditions with other non-polar sorbents immobilized in
the same extraction device. Sorptive phases of C18, C8 and HF were
tested.
Fig. 8 shows the comparative results. The most promising results
were obtained using RDSE with MMT-HDMIM-Br. The efficiency of
the liquid HF membrane with 1-octanol as the supported liquid
membrane (SLM) (PP with 1-octanol as the SLM) was also good,
particularly for the most polar PCBs. Similarly, RDSE with MMT-
HDMIM-OH was significant for retention of the most polar PCBs,
associated with the presence of OH anion and possible H-bond
formation with the target analytes. The C16 chain contained in the
structure of the ionic liquid likely favored the retention of the most
hydrophobic analytes with respect to the liquid membrane.
When the analytical performance of the synthetic sorptive
phase MMT-HDMIM-Br was compared with that observed in a
synthetic molecular imprinted polymeric (MIP) sorbent used pre-
viously in RDSE [7], at least two advantages arise: On the one hand,
 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31 29

Fig. 7. Effect of the extraction time on PCB extraction.

Table 2
Analytical features of the method.

Analyte Linearity, R2 Precision RSD, % LOD (ng L1) LOQ (ng L1) Absolute recovery, % Enrichment factor

PCB 081 0.9984 17.5 37.8 97.3 36.8 18

PCB 123 0.9983 4.0 2.5 6.5 36.1 18
PCB 114 0.9980 4.4 43.0 103.8 35.8 18
PCB 105 0.9951 4.9 19.6 46.6 42.2 21
PCB 126 0.9927 2.1 32.8 47.7 48.8 24
PCB 167 0.9983 7.9 6.8 14.6 49.2 25
PCB 156 0.9959 13.2 3.2 7.1 47.6 24
PCB 157 0.9984 24.8 3.2 7.2 47.2 24
PCB 169 0.9593 17.1 5.3 12.5 48.9 24
PCB 189 0.9945 9.9 12.5 23.4 55.5 28

the extraction time was significantly reduced from 120 min (using
MIP) to 30 min (using the proposed new phase). On the other hand,
eco-efficiency of the method was also improved when MMT-
HDMIM-Br is used, taking into account the following points: (a)
The amount of sorptive phase used per analysis was decreased
(from 25 mg in MIP to 10 mg in the present case), (b) reusability of
the phase was improved considering the number of extraction-
elution cycles (from 6 cycles in MIP to 10 cycles in the present
case) and (c) the synthesis of the proposed phase was greener that
the followed for MIP (use of fewer amounts of dangerous reagents).

Fig. 8. Comparative results for different phases used in RDSE for PCB extraction.
30 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
Fig. 9. Chromatogram of wastewater samples consisting of influent and spiked influent.
Under the above extraction conditions, a sample of the inflow of
a wastewater treatment plant in Santiago was analyzed, and
neither of the analytes was detected (Fig. 9). In the sample spiked
with 20 ng mL1 of the PCB standard, the analyte signals clearly
appeared, as observed in Fig. 9. The relative recovery percentages in
this analysis were in the range of 80e86%, demonstrating good
performance, considering the type of studied samples. As can be
seen also in Fig. 9, chromatographic peaks 7 and 8 were not sepa-
rated, which suggest that both analytes (PCB 156 and PCB 157)
could be determined as the sum of both signals, taking into account
that toxicity equivalent factors of both compounds are similar. In
addition, the small signals appearing in the non-spiked sample at
15.69 and 17.25 min did not correspond to PCBs, however to avoid
some interference in the determination by GC with electron cap-
ture detector, it is recommended to use a confirmative mass se-
lective detector.
4. Conclusion
In the present study, a novel eco-material based on montmo-
rillonite clays modified with ionic liquids (MMT-HDMIM-Br) was
proposed for the extraction of low-polarity compounds by RDSE.
This new extraction technique together with GC analysis exhibited
good analytical features for the determination of PCBs in waste-
water samples. The achieved analytical performance indicates that
this technique can be applied to the determination of other low-
polarity analytes.
Using rotating-disk sorptive extraction, the new material pro-
posed as the sorptive phase provided the best efficiency in the
extraction of PCBs from water compared with other non-polar solid
phases.
This newly developed ecosorbent will have great opportunities
in the microextraction field because of the possibilities that fully
comply with the modern trends in sample preparation.
Acknowledgments
The authors would like to thank FONDECYT (Project 1140716)
for financial support. One of the authors (FLJA) would like to thank
COLCIENCIAS and VIP Unicaldas for his “Jo
ven Investigador”
fellowship (call 2014/645).
References
[1] H. Piri-Moghadam, F. Ahmadi, J. Pawliszyn, A critical review of solid phase
microextraction for analysis of water samples, TrAC Trends in Anal. Chem. 85
(2016) 133e143.
[2] C. Arthur, J. Pawlyszin, Solid phase microextraction with thermal desorption
using fused silica optical fibers, Anal. Chem. 62 (1990) 2145e2148.
[3] R. Lucena, Extraction and stirring integrated techniques: examples and recent
advances, Anal. Bioanal. Chem. 403 (2012) 2213e2223.
[4] S.S. Lakade, F. Borrull, K.G. Furton, A. Kabir, N. Fontanals, R.M. Marce
,
Comparative study of different fabric phase sorptive extraction sorbents to
determine emerging contaminants from environmental water using liquid
chromatographyetandem mass spectrometry, Talanta 144 (2015)
1342e1351.
[5] C.E. Nika, E. Yiantzi, E. Psillakis, Plastic pellets sorptive extraction: low-cost,
 J.A. Fiscal-Ladino et al. / Analytica Chimica Acta 953 (2017) 23e31
rapid and efficient extraction of polycyclic aromatic hydrocarbons from
environmental waters, Anal. Chim. Acta 922 (2016) 30e36.
[6] L. Correa, J.A. Fiscal, S. Ceballos, A. de la Ossa, G. Taborda, C. Nerín, M. Rosero-
Moreano, A hollow fiber solvent bar microextraction GC-ECD determination of
disinfection byproducts in water samples, J. Sep. Sci. 38 (2015) 3945e3953.
[7] V. Manzo, K. Ulisse, I. Rodríguez, E. Pereira, P. Richter, A molecularly imprinted
polymer as the sorptive phase immobilized in a rotating disk extraction device
for the determination of diclofenac and mefenamic acid in wastewater, Anal.
Chim. Acta 889 (2015) 130e137.
[8] R.C. Zielke, T.J. Pinnavaia, Modified clays for the adsorption of environmental
toxicants: binding of chlorophenols to pillared, delaminated, and hydrox-
yinterlayered smectites, Clay Clay Min. 36 (1988) 403e408.
[9] M.M. Abolghasemi, S. Parastari, V. Yousefi, Microextraction of phenolic com-
pounds using a fiber coated with a polyaniline-montmorillonite nano-
composite, Mikrochim. Acta 182 (2015) 273e280.
[10] S. Zheng, Z. Sun, Y. Park, G.A. Ayoko, R.L. Frost, Removal of bisphenol A from
wastewater by Ca-montmorillonite modified with selected surfactants, Chem.
Eng. J. 234 (2013) 416e422.
[11] W.H. Awad, J.W. Gilman, M. Nyden, R.H. Harris Jr., T.E. Sutto, J. Callahan,
P.C. Trulove, H.C. DeLong, D.M. Fox, Thermal degradation studies of alkyli-
midazolium salts and their application in nanocomposites, Thermochim. Acta
409 (2004) 3e11.
[12] N.H. Kim, S.V. Malhotra, M. Xanthos, Modification of cationic nanoclays with
ionic liquids, Micropor. Mater 96 (2006) 29e35.
[13] J.U. Ha, M. Xanthos, Sequential modification of cationic and anionic nanoclays
with ionic liquids, Green Chem. Lett. Rev. 4 (2011) 103e107.
[14] C. Aftafa, F.O. Pelit, E.E. Yalcinkaya, H. Turkmen, I. Kapdan, F.N. Ertas, Ionic
liquid intercalated clay sorbents for micro solid phase extraction of steroid
hormones from water samples with analysis by liquid chromatography-
tandem mass spectrometry, J. Chromatogr. A 1361 (2014) 43e52.
[15] P. Richter, C. Leiva, C. Choque, A. Giordano, B. Sepúlveda, Rotating disk
sorptive extraction of nonylphenol from water samples, J. Chromatogr. A 1216
(2009) 8598e8602.
[16] A. Giordano, P. Richter, I. Ahumada, Determination of pesticides in river water
using rotating disk sorptive extraction and gas chromatographyemass spec-
trometry, Talanta 85 (2011) 2425e2429.
[17] Y. Corrotea, K. S
anchez, M.A. Rubio, P. Richter, Extraction of polycyclic aro-
matic hydrocarbons from water samples into a rotating-disk microextractor
31
and the subsequent determination by gas chromatography-mass spectrom-
etry, J. Chil. Chem. Soc. 49 (2014) 2477e2480.
[18] V. Manzo, O. Navarro, L. Honda, K. Sa
nchez, M.I. Toral, P. Richter, Determi-
nation of crystal violet in water by direct solid phase spectrophotometry after
rotating disk sorptive extraction, Talanta 106 (2013) 305e308.
[19] C. Mun ~ oz, M.I. Toral, I. Ahumada, P. Richter, Rotating disk sorptive extraction
of cu-bisdiethyldithiocarbamate complex from water and its application to
solid phase spectrophotometric quantification, Anal. Sci. 30 (2014) 613e617.
[20] A. Can ~ as, P. Richter, Solid-phase microextraction using octadecyl-bonded
silica immobilized on the surface of a rotating disk: determination of hexa-
chlorobenzene in water, Anal. Chim. Acta 743 (2012) 75e79.
[21] V. Manzo, M. Miro
, P. Richter, Programmable flow-based dynamic sorptive
microextraction exploiting an octadecyl chemically modified rotating disk
extraction system for the determination of acidic drugs in urine,
J. Chromatogr. A 1368 (2014) 64e69.
[22] A. Can ~ as, P. Richter, M.G. Escandar, Chemometrics-assisted excita-
tioneemission fluorescence spectroscopy on nylon-attached rotating disks.
Simultaneous determination of polycyclic aromatic hydrocarbons in the
presence of interferences, Anal. Chim. Acta 852 (2014) 105e111.
[23] L. Jachero, B. Sepúlveda, I. Ahumada, E. Fuentes, P. Richter, Rotating disk
sorptive extraction of triclosan and methyl-triclosan from water samples,
Anal. Bioanal. Chem. 405 (2013) 7711e7716.
[24] M. Becerra-Herrera, L. Honda, P. Richter, Ultra-high-performance liquid
chromatographydtime-of-flight high resolution mass spectrometry to
quantify acidic drugs in wastewater, J. Chromatogr. A 1423 (2015) 96e103.
[25] G.H. Min, T. Yim, H.Y. Lee, D.H. Huh, E. Lee, J. Mun, S.M. Oh, Y.G. Kim, Synthesis
and properties of ionic liquids: imidazolium tetrafluoroborates with unsatu-
rated side chains, Bull. Korean Chem. Soc. 27 (2006) 847e852.
[26] J. Dupont, C.S. Consorti, P.A.Z. Suarez, R.F. de Souza, Preparation of 1-butyl-3-
methyl imidazolium-based room temperature ionic liquids, Org. Synth. 79
(2002) 236e243.
[27] B.C. Ranu, S. Banerjee, Ionic liquid as catalyst and reaction medium. The
dramatic influence of a task-specific ionic liquid, [bmIm]OH, in michael
addition of active methylene compounds to conjugated ketones, carboxylic
esters, and nitriles, Org. Lett. 7 (2005) 3049e3052.
[28] G. Liptay (Ed.), Atlas of Thermoanalytical Curves, Heyden and Son Ltd.,
Budapest, 1973.