Polymerized phosphonium-based ionic liquids as stationary phases in gas chromatography: performance improvements by addition of graphene oxide

Article abstract of DOI:10.1039/c5nj01842k

Eight new functionalized, polymerizable phosphonium ionic liquids were prepared and applied as polymeric stationary phases in gas chromatography (GC). These coated GC columns exhibit high thermal stability compared to other ionic liquids (220-380 °C), and column efficiencies between 2700 and 3700 plates per m. The new columns have been characterized using the Abraham model in order to understand the effects of the polymeric cations and anions on the behavior of the system. These stationary phases show unique selectivity for several types of organic compounds such as alcohols, amines, pesticides and polycyclic aromatic hydrocarbons (PAHs), with good peak symmetries in some cases. Moreover, graphene oxide (GO) sheets have been covalently bonded onto the inner wall surface of fused silica capillary columns using 3-aminopropyl-diethoxymethylsilane (3-AMDS) as a cross-linking agent. The use of GO in the preparation of the capillary columns enhances their efficiency, improving peak symmetries because of the reduction of the unspecific absorption.

Full text of DOI:10.1039/c5nj01842k

NJC
View Article Online
View Journal
PAPER
Polymerized phosphonium-based ionic liquids as
stationary phases in gas chromatography:
performance improvements by addition of
graphene oxide†
Cite this:DOI: 10.1039/c5nj01842k
a
a
b
´
Jaime Gonzalez-Alvarez,* Pilar Arias-Abrodo, Marcos Puerto,
´
b
b
a
´
Maialen Espinal Viguri, Julio Perez and Marıa Dolores Gutierrez-Alvarez*
´
´
Eight new functionalized, polymerizable phosphonium ionic liquids were prepared and applied as
polymeric stationary phases in gas chromatography (GC). These coated GC columns exhibit high thermal
stability compared to other ionic liquids (220–380 1C), and column efficiencies between 2700 and 3700
plates per m. The new columns have been characterized using the Abraham model in order to understand
the effects of the polymeric cations and anions on the behavior of the system. These stationary phases
show unique selectivity for several types of organic compounds such as alcohols, amines, pesticides and
polycyclic aromatic hydrocarbons (PAHs), with good peak symmetries in some cases. Moreover, graphene
oxide (GO) sheets have been covalently bonded onto the inner wall surface of fused silica capillary
columns using 3-aminopropyl-diethoxymethylsilane (3-AMDS) as a cross-linking agent. The use of GO
in the preparation of the capillary columns enhances their efficiency, improving peak symmetries
because of the reduction of the unspecific absorption.
Received (in Nottingham, UK)
15th July 2015,
Accepted 18th August 2015
DOI: 10.1039/c5nj01842k
www.rsc.org/njc
access multi-substituent functionalized phosphonium cations
can explain the scarcity of studies focused on phosphonium-
1. Introduction
The term ‘‘ionic liquid’’ (IL) is currently used to describe a broad based ILs. However, some of their properties (i.e. less dense than
class of salts with wide liquid ranges.
water, whereas the density of the nitrogen-based ILs is higher28)
Ionic liquids have been developed for a wide variety of applica- make them more suitable for some applications such as solvents
tions in different areas due to their unique properties.1–11 An in organic and inorganic synthesis. In the field of analytical
emerging application of ILs is as stationary phases in gas- chemistry and, more specifically, in gas chromatography, the
chromatography, which is the subject of this paper. The common phosphonium-based ILs offer great potential due to properties
properties of the ILs, such as high thermal and redox stability, such as higher thermal stability29,30 than many nitrogen-based
high viscosity and negligible vapor pressure, make these types of ILs. However, the paucity of studies carried out on phosphonium
compounds very attractive as stationary phases in gas chromato- ILs is still not sufficient for a wide range of applications. It is
graphy. The vast majority of ILs previously applied in GC are based therefore necessary to perform detailed studies on the behavior
on imidazolium cations.12–22 In contrast, studies focused on novel of phosphonium ILs as stationary phases in gas chromatography.
classes of ILs achieved by a modification of the chemical structure
A major problem in the study of new stationary phases based
of the cations are scarce. In particular, phosphonium, sulfonium on ILs is the difficulty to obtain highly homogenous coatings
and borenium-based ILs23–27 have not received the same literature that would generate good profiles of peaks and more effective
attention as their imidazolium analogs.
separations. For this reason, in recent years, developments
The higher cost and lower air stability of most phosphines in separation sciences have resulted in a new generation of
compared with amines, along with the synthetic difficulties to stationary phases containing carbon nanotubes and GO, which
have been found to provide homogeneous coatings. Graphene
is a monolayer of sp2-hybridized carbon atoms tightly arranged
a
´
Department of Physical and Analytical Chemistry, University of Oviedo, C/Julian
into a two-dimensional honeycomb lattice. GO is an aromatic
macromolecule containing reactive oxygen functional groups
such as epoxide and hydroxyl.31,32
´
Claverıa 8, 33006 Oviedo, Spain. E-mail: gonzalezalvarezjaime@gmail.com,
loly@uniovi.es; Fax: +34 985 103125; Tel: +34 985 103489
b
´
Department of Organic and Inorganic Chemistry, University of Oviedo, C/Julian
Zhao et al.33 demonstrated that singled-walled carbon nano-
´
Claverıa 8, 33006 Oviedo, Spain. Fax: +34 985 103125; Tel: +34 985 103473
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5nj01842k tubes enhanced the enantioseparation on a chiral ionic liquid
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
stationary phase, because the carbon nanotubes increased the stirred for 48 h. Methanol was then evaporated under vacuum and
superficial area in the inner wall, and therefore favored the the resulting solid residue was extracted with dichloromethane
interactions between the stationary phase and the analytes.
(25 mL). The solution was washed with water (3 ꢀ 20 mL) and
Qu et al. have employed graphene and GO as stationary phases dried overnight over magnesium sulfate. Then, the resulting
for capillary electrochromatography, capillary liquid chromato- solution was filtered, and evaporated under reduced pressure to
graphy and gas chromatography, enabling effective separation of yield compound [(nBu)3(allyl)P]NTf2 (IL 2) as an oil (65%).
neutral, basic and protein mixtures,34 and separation of various
types of organic compounds with good separation efficiency.35
[(nBu)3(allyl)P]OTf (IL 3). To a solution of [(nBu)3(allyl)P]Br
(IL 1) (0.100 g, 0.309 mmol) in CH2Cl2 (20 mL), AgOTf (0.095 g,
In order to contribute to the development and study of new 0.371 mmol) was added and the mixture was stirred for 48 h.
stationary phases with improved selectivities, in this paper, we Afterwards, the solution was filtered through a short plug of
present a complete study of the solvation properties of eight diatomaceous earth, dried under reduced pressure, and after
monocationic phosphonium-based polymeric ILs (phosphonium 6 h, IL 3 was obtained as an oil (78%).
PILs).30 Moreover, with the aim of improving the column effi-
[(nBu)3(allyl)P]BAr04 (IL 4). Compound IL 4 was prepared as
ciency, to eliminate non-specific interactions due to silanol described above for IL 3, starting from IL 1 (0.100 g, 0.309 mmol)
groups, sheets of graphene oxide covalently bonded to the inner and NaBAr04 (0.329 g, 0.371 mmol). IL 4 was obtained as a white
wall of the column have been incorporated. For the first time, solid, with a melting point of 75 1C (yield 70%).
phosphonium ionic liquids have been polymerized onto this
[(nBu)3(crotyl)P]Cl (IL 5). Crotyl chloride (0.94 mL,
modified inner wall, giving good results. The thermal stability 9.60 mmol) was added to a solution of tri-n-butylphosphine
of the new phosphonium PIL stationary phases has been evalu- (2 mL, 8.00 mmol) in CH2Cl2 at 0 1C. The mixture was allowed
ated, and the Abraham model has been used to study the variation to warm to room temperature, stirred for 1 hour and then the
of the interactions between the new chromatographic phases and solvent was evaporated to dryness under reduced pressure. The
the analytes as the cations and anions are changed.
residue was dissolved in methanol (20 mL), to which lithium
bis(trifluoromethyl)sulfonamide (2.756 g, 9.60 mmol) was
added and the reaction mixture was stirred for 48 h. The solvent
was then evaporated under vacuum, the residue was extracted
with CH2Cl2 (20 mL) and the organic layer was washed with
distilled water (3 ꢀ 20 mL). Dichloromethane was evaporated
and IL 5 was obtained as a colorless oil (63%).
2. Material and methods
Most chemical reagents were purchased from Sigma-Aldrich
and used without further purification. The PAH mixture was
purchased from Restek (Bellefonte, United States). NaBAr04
(Ar0 = 3,5-bis(trifluoromethyl)phenyl) was prepared as previously
reported.36 Methanol, dichloromethane and n-hexane were
obtained from Merck. Untreated fused silica capillaries (0.25 and
0.10 mm i.d.) were purchased from Supelco (Madrid, Spain).
Dichloromethane was distilled from CaH2 and stored under nitro-
[(nBu)3(crotyl)P]NTf2 (IL 6). Crotyl chloride (0.94 mL,
9.60 mmol) was added to a solution of n-tributylphosphine
(2 mL, 8.00 mmol) in CH2Cl2 at 0 1C. The mixture was allowed
to warm to room temperature, stirred for 1 h and then the
solvent was evaporated to dryness under reduced pressure. The
residue was dissolved in methanol (20 mL), to which lithium
bis(trifluoromethane)sulfonamide (2.756 g, 9.60 mmol) was
added and the reaction mixture was stirred for 48 h. The solvent
was then evaporated under vacuum, the residue was extracted
with CH2Cl2 (20 mL) and the organic layer was washed with
distilled water (3 ꢀ 20 mL). Dichloromethane was evaporated
and IL 6 was obtained as a colorless oil (2.709 g, 63%).
gen in a Young tube and used without further purification. 1H, 31
P
and 13C NMR spectra were recorded on a Bruker Advance 300, DPX-
300, or Advance 400 spectrometer. NMR spectra are referred to the
1
internal residual solvent peak for H and 13C {1H}, whereas phos-
phoric acid (H3PO4, 85%) is used as the external reference for 31P
NMR. The NMR samples were prepared under nitrogen using Konte
manifolds purchased from Aldrich. All spectra are given in the ESI.†
[(nBu)3(methallyl)P]NTf2 (IL 7). IL 7 was prepared as described
above for IL 6, starting from n-tributylphosphine (2 mL, 8.00 mmol),
methallyl chloride (0.94 mL, 9.60 mmol) and lithium bis(trifluoro-
2.1. Synthesis of polymerizable phosphonium ionic liquids
Some manipulations were performed under an inert atmo- methane)sulfonamide (2.756 g, 9.60 mmol). IL 7 was obtained as a
sphere of dinitrogen using standard Schlenk techniques due colorless oil (57%).
to the air sensitivity of n-tributylphosphine.
[(Ph)2(allyl)2P]NTf2 (IL 8). IL 8 was prepared as described
[(nBu)3(allyl)P]Br (IL 1). Allyl bromide (0.83 mL, 9.60 mmol) above for IL 6, starting from allyldiphenylphosphine (1.90 mL,
was added to a solution of n-tributylphosphine (2 mL, 8.00 mmol), allyl bromide (0.83 mL, 9.60 mmol) and LiNTf2
8.00 mmol) in dichloromethane (20 mL), previously cooled to (2.756 g, 9.60 mmol). IL 8 was obtained as a colorless oil (66%).
0 1C. The reaction mixture was allowed to reach room tempera-
Schemes S1 and S2 show the structures of all the polymeriz-
ture and stirred for 1 h. The resulting solution was evaporated able phosphonium ILs used in this study (see ESI†).
to dryness, and after aprox. 6 h under vacuum IL 1 was obtained
as a white solid (2.379 g, 92%).23
2.2. Methods
[(nBu)3(allyl)P]NTf2 (IL 2). LiN(CF3SO2)2 (0.106 g, 0.371 mmol)
2.2.1. Preparation of polymeric phosphonium-based IL
was added to a solution of [(nBu)3(allyl)P]Br (IL 1) (0.100 g, capillary columns. All capillary columns were coated using
0.309 mmol) in methanol (20 mL), and the resulting slurry was the static method on a 10 m capillary column (0.25 mm i.d.)
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
at 40 1C using 0.25% (w/v) of ILs dissolved in dichloromethane. and phosphonium PILs. As it can be seen, a homogeneous film
Prior to adding the solvent to the IL monomer, 3 mg of AIBN is formed.
[2,20-azobis(2-methylpropionitrile)] (B10 wt%) were added as
To evaluate the capillary columns, each individual probe
an initiator of the radical polymerization. Capillaries were filled molecule (dissolved in dichloromethane) was injected into the
with the solution of the initiator and ionic liquid. The AIBN columns at three different temperatures: 50 1C, 80 1C and
decomposition constant is lower than the coating rate at this 110 1C. At 50 1C some of the used probe molecules, (particularly
temperature, so polymerization was almost insignificant dur- the H-bond donors) showed peak asymmetries (As) higher than
ing the filling of the column.37,38 After coating, the ends of the 1.5 and were therefore rejected from the data set. At 110 1C
capillaries were sealed and the capillaries placed in a GC oven, some probe solutes coeluted with the solvent; therefore, these
where the column was heated from 40 to 80 1C at 1 1C minꢁ1 results were also discarded.
and then kept at 80 1C for 5 h in order to ensure complete
For the determination of the system constants, 43 probe
polymerization. Eqn (1) can be used to approximate the sta- molecules were used, ensuring that a minimum of four test
tionary phase film thickness (df) of capillary columns.22,38
compounds were used to assess each of the parameters under
study. Selected compounds differ in size, dipolarity/polarizability,
and hydrogen bond donor or acceptor properties. Moreover, the
peak asymmetries were from 0.9 to 2.3, and the retention times
ranged from 2 to 15 min for the probe molecules at 110 1C. The
solute descriptors for the 43 probe molecules are listed in Table 1.
df = dcc/400
(1)
where dc is the diameter of the capillary tubing (in micro-
meters), and c is the percentage by weight (%) concentration of
the stationary phase dissolved in an appropriate solvent.
Helium carrier gas was then flushed through the capillaries at
a rate of 1 mL minꢁ1, and the capillaries were then conditioned
from 30 to 120 1C at 3 1C minꢁ1 and held at 120 1C for 2 h.
A triplicate of each column was prepared and the efficiencies of
the IL columns were determined to be higher than 3200 plates
per m by using naphthalene at 100 1C.
2.2.2. Preparation of polymeric phosphonium-based IL capillary
columns incorporating sheets of graphene oxide bonded to the inner
wall columns. In order to avoid the nonspecific adsorptions due to
the low-stability of the graphene dispersion in dichloromethane,30
GO sheets were covalently bonded onto the inner surface of fused
silica capillary columns using 3-aminopropyl-diethoxymethyl-silane
(3-AMDS) as a cross-linking agent. First, the capillaries were pre-
treated with 1 mL of 3-AMDS toluene solution (1 vol%) in order to
modify the inner surface of the capillaries through a covalent
interaction between the silanol groups and 3-AMDS.35 Then, GO
sheets were linked to the amino functionalized surface of fused silica
capillary columns by rinsing the capillaries with the GO suspension.
To obtain a stable homogeneous colloidal suspension of GO sheets,
a long time of sonication is required (over 6 h) in an ultrasound bath,
followed by centrifugation at 3000 rpm for 5 min. After coating,
the ends of the capillaries were sealed and the capillaries were
introduced in a GC oven, heating the columns from 50 to
120 1C at 3 1C minꢁ1. After this, the capillaries were filled with
the solution of initiator and ionic liquid.
Table 1 Probe molecules and values of solute descriptors used in the
characterization of phosphonium PIL stationary phases
Probe molecules
E
S
A
B
L
1-Hexanol
Anisole
0.21
0.44
0.344
0
0
0
0
0
0
0
0.09
0.049
0.344
0
0
0
0
0
0.52
0.311
0.53
0.054
0.139
0.414
0.438
0.135
0.117
0.27
3.643
3.808
3.759
4.516
4.734
4.838
3.409
4.548
2.544
2.878
3.128
3.006
4.04
4.378
4.539
5.154
3.856
3.781
4.424
4.005
4.648
6.175
3.956
6.921
4.972
4.244
5.974
5.475
3.844
4.533
2.963
2.578
5.032
4.741
5.589
3.017
3.814
3.409
6.034
3.86
0.712
0.403
0.872
0.595
0.121
0.079
1.182
0.167
0.243
0.219
0.635
0.742
0.648
0.846
1.24
0.768
0.895
0.771
0.499
0.636
0.57
0.784
0.274
0.925
0.44
0.843
1.135
1.055
1.138
0.906
0.642
0.662
0.819
1.025
0.44
Cyclohexanone
1,2-Dichlorobenzene
Butylbenzene
Nonanal
Butyl acetate
Iodobenzene
1-Hexyne
1-Nitropropane
1-Pentanol
Pyridine
Benzonitrile
Phenyl acetate
Nitrobenzene
Naphthalene
Heptanal
2-Heptanone
1-Phenylethanol
Benzaldehyde
1-Octanol
0.52
0.532
0.331
0.521
0.269
0.193
0.441
0.496
0.648
0.394
0.52
0.135
0.425
0.252
0.309
0.557
0.393
0.606
0.319
0.503
0.501
0.52
0.14
0
0
0.123
0.823
0.813
0.199
1.419
0.955
1.664
1.017
0.803
0.668
1.119
0.769
0.806
0.198
0.224
0.694
0.787
0.191
0.101
0.613
0.079
1.269
0.894
0.105
0.084
1.0217
0.351
0
0.344
0
0.249
0
0.366
0.4
1-Chloronaphthalene
Aniline
Fluorene
0.951
1.003
1.12
4-Chloroaniline
Benzyl alcohol
Butyl benzoate
Cinnamic alcohol
Phenol
Acetophenone
3-Methyl-1-butanol
1-Butanol
Ethyl benzoate
2-Phenylethanol
1-Decanol
Triethylamine
Ethylbenzene
1,3-Dinitrobenzene
Coumarin
1.128
0.882
0.851
0.971
0.759
1.026
0.423
0.44
0.886
0.797
0.44
0.14
0.499
0.57
0
Fig. 1 shows the SEM images of capillary columns with a
mixture of sheets of graphene oxide bonded to the inner wall
0.451
0.716
0
0.351
0.344
0
0.39
0.344
0
0
0
0
0
0
0
0.444
0.636
0.52
0.78
0.139
0.438
0.524
0.177
0.496
0.456
0.309
1.61
Styrene
0.671
0.662
0.564
1.128
Fig. 1 SEM images of (a) the inner wall of the capillary with GO sheets
covalently bonded and phosphonium PILs (b) an amplified zone of the
image shown in (a).
Undecan-2-one
Methyl hexanoate
4-Chloroaniline
5.732
3.970
4.972
0.366
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
The probe molecules were injected and retention times were
NJC
measured in triplicate. Multiple linear regressions were per-
formed using Statgraphics Centurion XV for Windows version
15.2.06. The linearity regression (R2) for all evaluations was
Z0.97. The values of all the system constants are listed in
Table 2 for phosphonium PILs.
All separations were performed using a Shimadzu GC-2010
gas chromatograph (Shimadzu, Kyoto, Japan) equipped with a
flame ionization detector. Analyses of alcohols/amines, PAHs
and pesticides were performed using helium as carrier gas at a
flow of 1 mL minꢁ1 and a split ratio of 100/1. The injector and
detector temperatures were 250 1C and 300 1C, respectively.
Methane was used to measure the column hold-up time.
Fig. 2 Representation of efficiency variation of phosphonium PIL stationary
phases.
3. Results and discussion
PIL film thickness of about 0.16 microns. Therefore, capillary
columns were prepared with a thickness of approximately
0.16 microns film (0.25%, w/v) for all subsequent studies.
Phosphonium-based ionic liquids (IL 1–8) have been synthe-
sized by quaternization of phosphines, and in several cases by a
subsequent methatesis reaction to allow the anion exchange.
With the new ILs in hand, we have determined the influence of
both cations and anions on the properties of the ionic liquids,
as well as in their performance as stationary phases in GC.
3.2. Thermal stability
Thermal stability is highly dependent on both anions and the
cations of the ILs, and directly determines the operating
temperature range and lifetime of the GC columns.
3.1. Optimization of film thickness capillary columns
Fig. 3 shows the loss of the stationary phase due to the IL
The variation of the column efficiency as a function of the film decomposition or volatilization. As seen in Fig. 3, PIL 8, which
thickness was evaluated using naphthalene as a probe solute. has two allyl groups, shows the highest thermal stability,
As can be seen in Fig. 2, a percentage of 0.25% w/v PIL provides presumably due to crosslinking during polymerization. PILs
greater separation efficiency. This percentage corresponds to a 1, 4, 5 and 7 show lower thermal stability. It is well known that
Table 2 Regression parameter coefficients of phosphonium PILS
IL
T (1C)
c
e
s
a
b
l
R2
n
SE
F
IL 1
50
80
110
ꢁ4.68 (0.12)
ꢁ4.32 (0.16)
ꢁ4.61 (0.15)
ꢁ0.20 (0.09)
ꢁ0.23 (0.14)
ꢁ0.28 (0.12)
1.83 (0.12)
1.74 (0.16)
1.22 (0.17)
5.68 (0.19)
4.83 (0.20)
3.60 (0.26)
0.09 (0.14)
0.72 (0.03)
0.59 (0.02)
0.48 (0.03)
0.99
0.99
0.98
28
32
24
0.05
0.07
0.09
338
212
188
ꢁ0.64 (0.22)
ꢁ0.66 (0.29)
IL 2
IL 3
IL 4
IL 5
IL 6
IL 7
IL 8
50
80
110
ꢁ3.39 (0.19)
ꢁ4.32 (0.14)
ꢁ4.35 (0.23)
ꢁ0.12 (0.14)
ꢁ0.23 (0.18)
ꢁ0.80 (0.13)
2.22 (0.19)
1.27 (0.13)
1.15 (0.23)
1.85 (0.32)
1.69 (0.16)
1.32 (0.28)
1.28 (0.29)
0.70 (0.16)
ꢁ1.07 (0.22)
0.52 (0.05)
0.58 (0.03)
0.55 (0.04)
0.98
0.98
0.97
24
27
32
0.12
0.08
0.1
123
175
66
50
80
110
ꢁ4.73 (0.19)
ꢁ3.99 (0.08)
ꢁ5.48 (0.10)
ꢁ0.32 (0.14)
ꢁ0.15 (0.07)
ꢁ0.04 (0.07)
1.80 (0.20)
1.74 (0.09)
1.31 (0.09)
4.21 (0.32)
3.14 (0.16)
2.15 (0.11)
ꢁ0.37 (0.24)
ꢁ0.64 (0.15)
ꢁ0.79 (0.12)
0.77 (0.05)
0.61 (0.02)
0.40 (0.02)
0.98
0.99
0.99
28
34
33
0.08
0.04
0.04
124
461
802
50
80
110
ꢁ3.98 (0.11)
ꢁ3.38 (0.13)
ꢁ5.33 (0.22)
ꢁ0.14 (0.07)
ꢁ0.28 (0.08)
ꢁ0.35 (0.16)
1.57 (0.08)
0.96 (0.12)
0.94 (0.19)
4.22 (0.14)
3.75 (0.18)
2.93 (0.22)
1.27 (0.10)
0.90 (0.15)
0.29 (0.23)
0.62 (0.03)
0.55 (0.02)
0.41 (0.02)
0.99
0.98
0.98
25
26
21
0.07
0.06
0.11
216
120
174
50
80
110
ꢁ4.70 (0.25)
ꢁ4.36 (0.09)
ꢁ4.55 (0.13)
0.09 (0.18)
ꢁ0.28 (0.07)
ꢁ0.36 (0.11)
1.87 (0.25)
1.80 (0.10)
1.78 (0.15)
7.23 (0.46)
5.56 (0.14)
4.79 (0.23)
ꢁ0.21 (0.28)
ꢁ0.33 (0.15)
ꢁ0.39 (0.28)
0.67 (0.06)
0.56 (0.02)
0.49 (0.02)
0.97
0.99
0.99
23
31
25
0.15
0.04
0.07
77
544
239
50
80
110
ꢁ3.52 (0.13)
ꢁ3.85 (0.11)
ꢁ3.99 (0.12)
ꢁ0.20 (0.09)
ꢁ0.21 (0.07)
ꢁ0.24 (0.08)
1.28 (0.12)
1.05 (0.08)
0.90 (0.10)
0.97 (0.17)
0.91 (0.11)
0.88 (0.11)
1.72 (0.11)
0.97 (0.10)
0.55 (0.14)
0.49 (0.03)
0.42 (0.02)
0.40 (0.03)
0.98
0.98
0.98
32
35
30
0.08
0.07
0.07
137
171
116
50
80
110
ꢁ3.86 (0.14)
ꢁ3.94 (0.10)
ꢁ3.83 (0.16)
ꢁ0.12 (0.09)
ꢁ0.13 (0.03)
ꢁ0.24 (0.10)
1.57 (0.14)
1.56 (0.08)
1.00 (0.16)
1.62 (0.21)
1.42 (0.11)
0.66 (0.19)
0.81 (0.16)
0.51 (0.07)
0.04 (0.21)
0.67 (0.03)
0.58 (0.02)
0.45 (0.03)
0.99
0.99
0.98
31
36
33
0.07
0.03
0.07
214
450
106
50
80
110
ꢁ3.90 (0.20)
ꢁ4.26 (0.13)
ꢁ3.74 (0.10)
ꢁ0.06 (0.15)
ꢁ0.10 (0.09)
ꢁ0.43 (0.08)
1.39 (0.21)
1.33 (0.13)
1.27 (0.12)
1.42 (0.32)
1.07 (0.16)
0.08 (0.14)
1.40 (0.26)
1.09 (0.17)
0.14 (0.16)
0.63 (0.04)
0.51 (0.03)
0.39 (0.02)
0.97
0.98
0.98
29
32
30
0.12
0.1
0.06
91
193
221
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
parameter model developed by Abraham43 has been used to
characterize the interactions between molecules of solute and
stationary phases.
The Abraham solvation parameter model equation is as follows-
log k = c + eE + sS + aA + bB + lL
(2)
According to eqn (2), the retention factor (k) of a given solute
at a specific temperature is determined chromatographically.
E, S, A, B, and L are solute descriptors that represent excess
molar refraction, dipolarity, H-bond acidity, H-bond basicity
and the gas-hexadecane partition coefficient at 298 K, respec-
tively. Poole and co-workers44 optimized these descriptors by
gas chromatography, reversed-phase liquid chromatography
and micellar electrokinetic chromatography. The system con-
stants from eqn (2) are defined as follows: e indicates the ability
of the stationary phase to interact with the p- and n-electrons of
the solvent; s is the dipolarity/polarizability of the solvent; a and
b define the solvent hydrogen bond basicity and acidity, respec-
tively. Finally, l is the phase ability to distinguish between, or to
separate the members of any homologous series.
The system constants for the tested stationary phases were
obtained by the use of the inverse GC method and are listed
in Table 2. Two of the five interaction parameter coefficients,
namely s (dipole-type interactions) and a (H-bond basic inter-
actions) have larger magnitudes for all ILs studied. This fact
has been already observed for imidazolium-based ILs19 and
phosphonium PILs.29 It is well known that hydrogen bond
basicity and dipolarity of ILs are determined by the nature of
the counter-anions. The cross-correlation between descriptors
were checked and low correlations between descriptors (o0.5 in
absolute value) were observed.
Fig. 3 Thermal stability of the phosphonium PIL stationary phases.
the presence of halides in the ILs lowers the thermal stability of the
resulting stationary phases due to their nucleophilic nature, which
opens the way to decomposition via SN1 and SN2 processes.39 PIL 4
shows lower stability than PIL 1 because of the instability of the
tetrakis[3,5-bis(trifluoromethyl)phenyl]borate anion. On the other
hand, PILs with NTf2 and OTfꢁ anions show higher thermal
ꢁ
stabilities, except in the case of PIL 7, a fact that could indicate
instability due to the methallyl group. For PIL 6, the slight increase
in the length of the aliphatic chain compared to PIL 2 makes the
thermal stability somewhat higher.
These results provided a rough idea of the thermal stability
of the ILs studied. To evaluate quantitatively the thermal
stability of PIL stationary phases a study was conducted using
an isothermal method. Thus the GC columns were held at a
series of constant temperatures for at least 12 h, and the
retention of probe compounds was measured at a lower tem-
perature before and after the isothermal conditioning.
Table 3 shows the isothermal measurements for naphtha-
lene at different temperatures with the phosphonium PILs. As it
can be seen, the thermal stabilities of these PILs ranged from
220 to 380 1C, and are higher than those of other stationary
phases based on ILs.18,40,41
Hydrogen bond basicity values ranged from 0.66 to 5.56 at
80 1C. The values of the a-term of PILs 1 and 5 were higher in
comparison with the other PILs, since the high electron density
of chloride and bromide makes them strong hydrogen bond
acceptors. The other studied PILs showed lower hydrogen bond
basicities due to the large delocalization of the negative charge
in NTf2ꢁ, OTfꢁ and tetrakis-(3,5-bis(trifluoromethyl)phenyl)-
borate anions. This observation has already been made for
imidazolium cations.16,19,23,45–47
3.3. IL solvation parameters: the effect of anions and cations
on the system constants
A number of methods have been developed to characterize the
solvation behaviour of liquids.20,42 In this study, the solvation
It was observed that the values of the coefficients decrease as
the temperature is increased, showing that the solute retention
decreases as a function of temperature, as usually occurs in gas
chromatography analysis.
Table 3 Variation of the naphthalene retention factor with phosphonium
PIL columns using the isothermal method
T (1C) IL 1
IL 2
IL 3
IL 4
IL 5
IL 6
IL 7
IL 8
100
150
180
200
220
250
270
300
320
330
350
380
400
0.64
0.52
0.47
0.43
0.38
0.31
0.70
0.65
0.57
0.46
0.37
0.31
0.97
0.92
0.84
0.80
0.74
0.69
0.61
0.83
0.76
0.69
0.62
0.55
0.35
0.28
0.23
0.17
0.28
0.24
0.20
0.17
0.59
0.54
0.50
0.45
0.38
0.31
0.61
0.56
0.52
0.49
0.45
0.39
The values obtained for the b-term were negative for some of
the PILs tested. Negative values of the b-terms have been also
reported for some of the imidazolium ILs, despite the fact that
the imidazolium cation possesses the ability to act as a hydrogen
bond donor. In this case, it seems that these unrealistic values
cannot be explained by the structure of the PILs because none
of them have acidic hydrogens that could act as strong hydro-
gen bond donors. Therefore, b-term was forced to be zero and
the regression was performed using e, s, a and l terms. The
results obtained for the correlation coefficient and system
0.20a 0.15
b
0.59a
0.10
b
0.40a 0.27
0.08
0.39a 0.33
b
b
0.36a 0.52
0.22a
0.26
b
b
0.46
0.21
0.18
0.11
0.51a
b
0.20a
b
a
Peaks tailing were observed. b No retention was observed for naphthalene. constants remained unchanged, implying that the b-term does
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
not affect the model. It is important to note that this phenomenon
has been reported previously for other phosphonium ILs.29,30
On the other hand, all the PILs that we have studied have
shown low values for non-bonding and p-electron interactions
(e-term). The negative values of this system constant are related
to the way in which the solute descriptor E is determined.42 The
comparison of phosphonium and imidazolium ILs shows that
the e-term values are lower for phosphonium ILs tested here,
due to the p-electron-rich imidazolium cations, implying that
the e-term is dominated by the cations. Other phosphonium
PILs studied show similar values for the e-term (from ꢁ0.28 to
0.12 at 70 1C). The positive values are due to the presence of
lone pair electrons on the oxygen atoms of the ether chain.29
In the case of PIL 8, which contains two phenyl groups in its
structure, negative values of the e-term can be explained by the
presence of p-stacking interactions between the phenyl groups.
The values obtained for the s-term were similar for all PILs
studied. For PILs 1–4, the anions change, whilst the cations
remain constant. There are no major changes in the s-term and
this implies that the dipolar interactions did not appear to be
significantly influenced by the anion type. When the cations are
changed (PILs 6–8), no changes were observed and it can be
concluded that all phosphonium PILs interact through dipole-
type interactions.
The l coefficient is a combination of exoergic and endoergic
processes. The magnitude of this term indicates how well the IL
separates compounds in a homologous series. As the studied
PILs had positive l coefficients and the dispersion interactions
seem to dominate. The term values fall within the range 0.61–
0.42 at 80 1C. These values are similar to those of most of the
studied phosphonium PILs. Therefore, dispersion interactions
must not be significantly influenced by the cations, and the
dipolarities (s-term) of these PILs are nearly identical.
Fig. 4 Separation of a mixture containing alcohols and amines using IL 1
and IL 6: (1) hexane; (2) 1-butanol; (3) 1-pentanol; (4) 2-methylpentanol; (5)
1-hexanol; (6) 1-octanol; (7) aniline; (8) diisopropylamine; (9) 1-decanol; (10)
2-phenylethanol; (11) cinnamic alcohol. Program temperature: initial oven
column temperature 100 1C for 2 min, increased at a rate of 25 1C minꢁ1 to
200 1C and then held at this temperature for 2 min.
IL film on the capillary wall. Moreover, PIL 5 shows the highest
retention times for polar compounds due to their increased
ability to interact through hydrogen bonds with this stationary
3.4. Separation of complex mixtures
In order to study the chromatographic behavior of these IL columns, phase (higher value of the a-term).
different standard mixtures were prepared and injected on
phosphonium PILs stationary phases.
Fig. 5 shows the chromatograms for the separation of a PAH
mixture employing PILs 1 and PIL 6. It has been demon-
As an example, the chromatograms of a mixture of amines strated48 that some PAHs display high carcinogenic and muta-
and alcohols using PIL 1 and PIL 6 coated columns are displayed genic activity, and have been listed as contaminants in waste
in Fig. 4. Good symmetries for PILs 1 (0.8–1.9) and 6 (0.8–1.5), water, solid waste and sediments.
compared to those obtained with other stationary phases based
As can be seen in Fig. 5, the PIL 1 column allowed to
on imidazolium and phosphonium ILs17,18,21,35 are diagnostic of separate indeno(1,2,3-cd)pyrene, dibenz(a,h)anthracene and
a homogeneous distribution of the polymerized ILs on the benzo(g,h,i)perylene in less than 30 min, with a resolution of
capillary inner wall. The higher tailing peaks for PIL 1 compared 3.4 for indeno(1,2,3-cd)pyrene and dibenz(a,h)anthracene; and
to those for PIL 6 are in agreement with the higher values of a resolution of 1.2 for dibenz(a,h)anthracene and benzo(g,h,i)-
a- and b-terms observed for PIL 1, due to the presence of the perylene. It is important to stress that this column is able to
bromide anion, which make this PIL a relatively strong hydrogen give good results for the separation of PAHs using only 10 m of
bond acceptor. Moreover, variations of the retention factor are the column.
observed for some analytes. Alcohols exhibited an increase in
The presence of the fronting peaks 12, 15, 16 and 17 with PIL 6
the retention time on the IL 6 stationary phase compared to IL may be due to a strong interaction between these compounds and
1, due to the presence of hydrogen bond-donating moieties.
the column, saturating the column even at low concentrations.
The chromatograms obtained for other phosphonium PILs
The chromatograms obtained for the other phosphonium
PILs studied are in the ESI.† As it is shown there, some under study herein are showed in the ESI.† All stationary
stationary phases (PILs 4 and 8) show high peak asymmetries phases show the same elution order, except for the stationary
(As 4 2.0) for alcohols, which is indicative of a non-homogeneous phase based on PIL 4. In this case, dibenz(a,h)anthracene
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
PIL 4 column than for the less bulky dibenz(a,h)anthracene, the
former being therefore more retained than the latter.
3.5. Incorporation of graphene oxide on phosphonium-based
polymeric ionic liquids
The addition of a GO dispersion to the phosphonium PILs prior to
use for separating a pesticide mixture by GC analysis improves the
chromatographic profile notably.30 The elution order remains the
same for all pesticides, indicating that the employment of GO
does not change the nature of the stationary phase, although it
increases the surface area, affording a more homogeneous film of
polymer ionic liquid along the capillary wall.
Despite these improvements, the GO dispersion in dichloro-
methane exhibited a low stability against agglomeration and
sedimentation, and this implies that the GO film along the
capillary wall was far from homogeneous. Considering that a
more homogeneous GO coating would further improve the column
performance, GO sheets were covalently bonded onto the inner
surface of fused silica capillary columns using 3-aminopropyl-
diethoxymethylsilane as a cross-linking agent. The column
Fig. 5 Separation of a polycyclic aromatic hydrocarbon mixture using IL 1 efficiencies of these new stationary phases have been evaluated,
and IL 6. (1) Hexane; (2) naphthalene; (3) acenaphthylene; (4) acenaphthene;
(5) fluorene; (6) phenanthrene; (7) anthracene; (8) fluoranthene; (9) pyrene;
(10) benz[a]anthracene; (11) chrysene; (12) benzo[b]fluoranthene; (13)
and in all cases an improvement of the efficiencies has been
observed, so that impressive 4200 plates per m were found for
some PILs. When naphthalene was injected at 100 1C, the
benzo[k]fluoranthene; (14) benzo[a]pyrene; (15) indeno(1,2,3-cd)pyrene;
retention time was measured after every 50 runs, and the results
(16) dibenz(a,h)anthracene; (17) benzo(g,h,i)perylene. Program temperature:
initial oven column temperature 90 1C for 2 min, increased at a rate of indicated that the columns coated with the new phases endured
15 1C minꢁ1 to 255 1C and held at this temperature for 20 min.
a series of 450 runs without loss of activity.
The thermal stabilities of these new stationary phases were
evaluated. As shown in Fig. 6, when GO sheets are covalently
eluted before the indeno(1,2,3-cd)pyrene. This behaviour could
be explained due to the nature of the anion (BAr04ꢁ) having four
aromatic rings in its structure. Steric effects seem to be the
responsible of this inverted elution order, as for the non-linear
indene compound it would be more difficult to go through the
bonded to the inner wall of capillary columns, a slight incre-
ment of the thermal stability is observed, compared to that of
the phosphonium PILs capillary columns. This phenomenon
could be explained because GO displays hydroxyl (–OH) and
carboxylic groups (–COOH) that allow p–p, n–p and halogen–p
interactions with phosphonium PILs. This fact implies that the
adsorption of the PILs onto the GO is stronger than onto silica.
GO–PIL 8 was chosen to test the effect of the covalently
bonded GO on the ability of the PIL columns to separate
different types of compounds by GC. The choice of this PIL 8
in particular was based on it, displaying the worst symmetry
peaks in the separation of the mixtures described above, so that
clearly an improvement was needed.
Table 4 shows the resulting system constant values for the
GO–PIL 8 stationary phase. The a and b terms are lower than
those of PIL 8 (with no GO). This can be attributed to the
generation of a more homogeneous film along the capillary as a
consequence of the covalently bonded GO sheets to the inner
wall, and to the existence of a lower number of free silanol
groups thus reducing the non-specific absorption.
Fig. 6 Thermal stability for the mixtures of phosphonium PILs and GO
sheets covalently bonded to inner wall stationary phases.
Table 4 Regression parameter coefficients of a mixture of PIL 8 and GO sheets covalently bonded to the inner wall stationary phase
IL
T (1C)
c
e
s
A
b
l
R2
n
SE
F
IL 8
50
80
110
ꢁ3.90 (0.18)
ꢁ4.30 (0.12)
ꢁ4.68 (0.10)
ꢁ0.10 (0.12)
ꢁ0.16 (0.19)
ꢁ0.39 (0.10)
1.45 (0.18)
1.21 (0.17)
1.09 (0.20)
1.05 (0.12)
0.82 (0.08)
0.11 (0.14)
1.11 (0.23)
0.86 (0.10)
0.19 (0.17)
0.79 (0.03)
0.68 (0.03)
0.61 (0.02)
0.99
0.98
0.97
30
34
36
0.11
0.08
0.07
561
227
105
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Paper
NJC
The chromatograms obtained when alcohol/amine and pesticide
mixtures were injected in the column coated with covalently linked
GO sheets and phosphonium PIL 8, are depicted in Fig. 7 and 8,
respectively. As shown in Fig. 7, the good peak symmetries obtained
(As o 1.1) compared to those observed by the use of a dispersion of
graphene for the alcohol/amine mixtures constitute a clear improve-
ment, and strongly suggest that the phosphonium PIL film is
homogeneously distributed on the capillary inner wall.
In agreement with the formation of a very homogeneous film
along the capillary wall, Fig. 8 shows also a good peak tailing
compared to the stationary phase based on GO dispersion and the
PIL. As a result, the new stationary phase allows the separation of
some pesticides that previously, using the columns discussed
above, co-eluted and could not be separated.
4. Conclusions
In this study, eight polymeric phosphonium ionic liquid stationary
phases were prepared and their chromatographic properties
used for gas chromatography analysis have been studied. The
thermal stabilities observed for all the stationary phases based
on the phosphonium ionic liquids studied are higher than those
of stationary phases based on traditional imidazolium mono-
cationic ionic liquids, therefore allowing separations at higher
temperatures than imidazolium ionic liquids.
Moreover, the incorporation of graphene oxide sheets through
covalent bonds to phosphonium PILs clearly improved their
properties and behavior as gas chromatography stationary phases,
due to the formation of a more homogeneous film along the
capillary. This fact implies an enhancement of the interactions
with the analytes and a decrease in non-specific interactions. As a
consequence, the new phases allowed separation of several types
of organic compounds with good separation efficiencies.
Fig. 7 Separation of a mixture containing alcohols and amines on (a)
polymeric IL 8 and the graphene dispersion column; (b) polymeric IL 8 and
the graphene oxide covalently bonded column: (1) hexane; (2) 1-butanol;
(3) 1-pentanol; (4) 2-methylpentanol; (5) 1-hexanol; (6) 1-octanol; (7)
aniline; (8) diisopropylamine; (9) 1-decanol; (10) 2-phenylethanol; (11)
cinnamic alcohol. Program temperature: initial oven column temperature
100 1C for 1 min, increased at a rate of 30 1C minꢁ1 to 200 1C and then held
at this temperature for 2 min.
This research should lead to the development of a wide range
of chromatographic columns displaying improved selectivities.
Acknowledgements
´
We thank Dr Lucıa Riera (CSIC-University of Oviedo) for helpful
comments.
References
1 D. R. MacFarlane, M. Forsyth, P. C. Howlett, J. M. Pringle,
J. Sun, G. Annat, W. Neil and E. I. Izgorodina, Acc. Chem.
Res., 2007, 40, 1165–1173.
2 C. F. Poole and S. K. Poole, J. Chromatogr. A, 2010, 1217,
2268–2286.
3 E. Stanisz, J. Werner and A. Zgola-Grzeskowiak, Trends Anal.
Chem., 2014, 61, 54–66.
4 X. Han and D. W. Armstrong, Acc. Chem. Res., 2007, 40,
1079–1086.
5 D. W. Armstrong, L. Zhang, L. He and M. L. Gross, Anal.
Chem., 2001, 73, 3679–3686.
Fig. 8 Separation of a pesticide mixture on: (a) a mixture of graphene
oxide dispersion and the PIL 8 column; (b) a mixture of covalently bonded
graphene oxide and the polymerized IL 8 column: (1) hexane; (2) hexa-
chlorobenzene; (3) aldrin; (4) bromocyclen; (5) HCH; (6) heptachlor; (7)
endrin; (8) isodrine; (9) chlorophenothan; (10) endosulfan; (11) 4,4-DDE;
(12) dieldrin; (13) DDD; (14) metoxychlor; (15) bromacil; (16) coumaphos;
(17) dialifos; (18) ruelene; (19) carbaril; (20) pyraclofos. Program tempera-
ture: initial oven column temperature 130 1C for 2 min and a heating rate
of 10 1C minꢁ1 to 250 1C and then held at this temperature for 15 min.
6 J. P. Hallet and T. Welton, Chem. Rev., 2011, 111, 3508–3576.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Paper
7 Q. Zhang, S. Zhang and Y. Deng, Green Chem., 2011, 13, 28 R. E. Del Sesto, C. Corley, A. Robertson and J. S. Wilkes,
2619–2637. J. Organomet. Chem., 2005, 690, 2536–2542.
8 S. Xiaojie, X. Jun, Z. Yuxiu and L. Zhaoxin, Prog. Chem., 29 Z. S. Breitbach and D. W. Armstrong, Anal. Bioanal. Chem.,
2014, 26, 647. 2008, 390, 1605–1617.
9 A. X. Zeng, S.-T. Chin, Y. Nolvachai, C. Kulsing, L. M. Sidisky 30 J. Gonzalez-Alvarez, P. A. Abrodo, M. Puerto, M. E. Viguri,
´
´
´
´
and P. J. Marriott, Anal. Chim. Acta, 2013, 803, 167–173.
10 K. Dettmer, Anal. Bioanal. Chem., 2014, 406, 4931–4939.
11 P. Delmonte, A. R. Fardin-Kia, J. K. G. Kramer, M. M. Mossoba,
J. Perez and M. D. G. Alvarez, RSC Adv., 2013, 3, 21377–21380.
31 G. I. Titelman, V. Gelman, S. Bron, R. L. Khalfin, Y. Cohen
and H. Bianco-Peled, Carbon, 2005, 43, 641–649.
L. Sidisky, C. Tybur-czy and J. I. Rader, J. Chromatogr. A, 2012, 32 O. C. Compton and S. T. Nguyen, Small, 2010, 6, 711–723.
1233, 137–146.
12 C. F. Poole, J. Chromatogr. A, 2004, 1037, 49–82.
33 L. Zhao, P. Ai, A.-H. Duan and L.-M. Yuan, Anal. Bioanal.
Chem., 2011, 399, 143–147.
13 P. Sun and D. W. Armstrong, Anal. Chim. Acta, 2010, 661, 34 Q. Qu, C. Gu and X. Hu, Anal. Chem., 2012, 84, 8880–8890.
1–16.
35 Q. Qu, Y. Shen, C. Gu, Z. Gu, Q. Gu, C. Wang and X. Hu,
Anal. Chim. Acta, 2012, 757, 83–87.
36 M. Brookhart, B. Grant and A. F. Volpe, Organometallics,
1992, 11, 3920–3922.
37 J. P. Van Hook and A. V. Tobolsky, J. Am. Chem. Soc., 1958,
80, 779.
38 J. Bouche and M. Verzele, J. Gas Chromatogr., 1968, 6,
501–505.
39 W. H. Awad, J. W. Gilman, M. Nyden, R. H. Harris,
T. E. Sutto, J. H. Callahan, P. C. Trulove, H. C. Delong and
D. M. Fox, Thermochim. Acta, 2004, 409, 3–11.
40 Z. Hai-Yan, L. Xian-Bo, T. Yu-Zeng and C. Ji-Ping, Chin.
J. Anal. Chem., 2010, 38(7), 1003–1006.
14 G. A. Baker, S. N. Baker, S. Pandev and F. V. Bright, Analyst,
2005, 130(6), 800–808.
15 D. W. Armstrong, L. He and Y.-S. Liu, Anal. Chem., 1999, 71,
3873–3876.
16 Q. Zhao, J. Eichhorn, W. R. Pitner and J. L. Anderson, Anal.
Bioanal. Chem., 2009, 359, 225–234.
17 K. Huang, X. Han, X. Zhang and D. W. Armstrong, Anal.
Bioanal. Chem., 2007, 389, 2265–2275.
18 T. Payagala, Y. Zhang, E. Wanigasekara, K. Huang, Z. S.
Breitbach, P. S. Sharma, L. M. Sidisky and D. W. Armstrong,
Anal. Chem., 2009, 81, 160–173.
´
´
19 J. Gonzalez-Alvarez, D. Blanco-Gomis, P. Arias-Abrodo,
´
´
´
D. Dıaz-Llorente, E. Busto, N. Rıos-Lombardıa, V. Gotor- 41 Q. Wei, M. Qi, H. Yang and R. Fu, Chromatographia, 2011,
´
´
Fernandez and M. D. Gutierrez-Alvarez, Anal. Bioanal.
Chem., 2011, 400, 1209–1216.
74, 717–724.
42 M. Vitha and P. W. Carr, J. Chromatogr. A, 2006, 1126,
143–194.
43 M. H. Abraham, C. F. Poole and S. K. Poole, J. Chromatogr. A,
1999, 842, 79–114.
44 C. F. Poole, S. N. Atapattu, S. K. Poole and A. K. Bell, Anal.
Chim. Acta, 2009, 652, 32–53.
45 C. F. Poole and S. K. Poole, J. Sep. Sci., 2011, 34, 888–900.
20 Y. Cong and J. L. Anderson, J. Chromatogr. A, 2009, 1216,
1658–1712.
21 Q. Q. Baltazar, S. K. Leininger and J. L. Anderson,
J. Chromatogr. A, 2008, 1182, 119–127.
22 J. L. Anderson and D. W. Armstrong, Anal. Chem., 2005, 77,
6453–6462.
´
´
´
23 J. McNulty and P. Das, Tetrahedron Lett., 2009, 50, 46 J. Gonzalez-Alvarez, D. Blanco-Gomis, P. Arias-Abrodo, D. Dıaz-
´
´
´
5737–5740.
Llorente, E. Busto, N. Rıos-Lombardıa, V. Gotor-Fernandez and
´
´
24 B. Ines, M. Patil, J. Carreras, R. Goddard, W. Thiel and
M. D. Gutierrez-Alvarez, J. Sep. Sci., 2012, 35, 273–279.
´
´ ´
47 J. Gonzalez-Alvarez, D. Blanco-Gomis, P. Arias-Abrodo, D. Dıaz-
M. Alcarazo, Angew. Chem., Int. Ed., 2011, 50, 8400–8403.
25 B. D. Powell, G. L. Powell and P. C. Reeves, Lett. Org. Chem.,
2005, 2, 550–553.
´
´
´
Llorente, E. Busto, N. Rıos-Lombardıa, V. Gotor-Fernandez
´
and M. D. Gutierrez-Alvarez, Anal. Chim. Acta, 2012, 721,
26 E. Frackowiak, G. Lota and J. Pernak, Appl. Phys. Lett., 2005,
86, 164104.
173–181.
48 Guidance for Reporting Toxic Chemicals: Polycyclic Aromatic
Compounds Category, United States Environmental Protection
Agency, EPA 260-B-01-03, 2003.
´
27 A. Farzad, M. T. Garcıa, R. D. Singer and P. J. Scammells,
Green Chem., 2009, 11, 1595–1604.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.