See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/256609509

Fast comprehensive two-dimensional gas

chromatography method for fatty acid methyl
ester separation and...

Article in Journal of Chromatography A · September 2013

DOI: 10.1016/j.chroma.2013.08.099 · Source: PubMed

CITATIONS READS

21 221

4 authors, including:

Asia Nosheen Blagoj Mitrevski

COMSATS Institute of Information Technology Monash University (Australia)
28 PUBLICATIONS 144 CITATIONS 22 PUBLICATIONS 296 CITATIONS

SEE PROFILE SEE PROFILE

Philip Marriott
Monash University (Australia)
342 PUBLICATIONS 9,147 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Biofuel, Phytoremediation, Stress Physiology and Plant Microbe Interaction View project

Biofuel, Bioremediation/Phytoremediation, Plant Microbe Interaction, Stress Physiology View project

All content following this page was uploaded by Asia Nosheen on 09 July 2017.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
 Author's personal copy

Journal of Chromatography A, 1312 (2013) 118–123

Contents lists available at ScienceDirect

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

Fast comprehensive two-dimensional gas chromatography method

for fatty acid methyl ester separation and quantification using dual
ionic liquid columnsq
Asia Nosheen a , Blagoj Mitrevski b , Asghari Bano a , Philip J. Marriott b,∗
a
Department of Plant Sciences, Quaid-i-Azam University, P.O. BOX 1090, Islamabad 44000, Pakistan
b
Australian Centre for Research on Separation Science, School of Chemistry, Monash University, Wellington Rd, Clayton, VIC 3800, Australia

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

Article history: Safflower oil is a complex mixture of C18 saturated and unsaturated fatty acids amongst other fatty
Received 12 June 2013 acids, and achieving separation between these similar structure components using one dimensional gas
Received in revised form 26 August 2013 chromatography (GC) may be difficult. This investigation aims to obtain improved separation of fatty
Accepted 27 August 2013
acid methyl esters in safflower oil, and their quantification using comprehensive two-dimensional GC
Available online 3 September 2013
(GC × GC). Here, GC × GC separation is accomplished by the coupling of two ionic liquid (IL) column
phases: the combination of SLB-IL111 with IL59 column phases was finally selected since it provided
Keywords:
excellent separation of a FAME standard mixture, as well as fatty acids in safflower and linseed oil, com-
Comprehensive two-dimensional gas
chromatography
pared to other tested column sets. Safflower oil FAME were well separated in a short run of 16 min. FAME
Safflower validation was demonstrated by method reproducibility, linearity over a range up to 500 mg L−1 , and
Fatty acids limits of detection which ranged from 1.9 mg L−1 to 5.2 mg L−1 at a split ratio of 20:1. Quantification was
Ionic liquid column carried out using two dilution levels of 200-fold for major components and 20-fold for trace components.
The fatty acids C15:0 and C17:0 were not reported previously in safflower oil. The SLB-IL111/IL59 column
set proved to be an effective and novel configuration for separation and quantification of vegetable and
animal oil fatty acids
© 2013 Elsevier B.V. All rights reserved.

1. Introduction different abundances make FA profiling a challenging task. Gas

Vegetable oils contain fatty acids (FA) that are implicated in var-
ious biological functions [1,2], and their FA profile is an important
commercial parameter [3]. Safflower is an oil seed crop which has
unique oil properties because of its fatty composition. It represents
a rich source of polyunsaturated fatty acids (PUFA) which play a
fundamental role in growth, coronary heart disease prevention,
and reduction in the risk of high blood pressure and atherosclerosis
[4]. Another important application of safflower oil is as a potential
source of biodiesel; this is a renewable resource, and large scale
production ensures adequate supply as a biodiesel feedstock [5]; it
is composed of 75–80% linoleic acid, which improves the low tem-
perature properties of biodiesel fuel. Rashid and Anwar [6] reported
that safflower biodiesel can be utilised effectively in ignition com-
pression engines due to its good quality fatty acid composition.
Structural similarities of fatty acids, their molecular heterogene-
ity in a wide variety of food samples and feedstocks, and their
q Presented at the 37th International Symposium on Capillary Chromatography
and 10th GC × GC Symposium, Palm Springs, CA, 12–16 May 2013.
∗ Corresponding author. Tel.: +61 3 99059630; fax: +61 3 9905850.
E-mail address: philip.marriott@monash.edu (P.J. Marriott).
chromatography (GC) is by far the most commonly used analyt-
ical procedure for FA analysis, usually after conversion into their
corresponding methyl esters (FAME). Non-polar (NP) and mid-
polarity columns give good FAME separation according to their
carbon number but they lack separation of positional and geomet-
ric isomers of unsaturated FA [7,8]. Polar columns (P) give better
separation of the aforementioned isomers within a given carbon
number, and are now the most common column type employed
for this purpose [8,9]. Recently, several applications have appeared
in the literature on using ionic liquid columns [7–9] as the most
polar stationary phases commercially available, even though some
of them were applied for C18 [10,11] or C22 isomers separation
[12]. The IL columns have shown equal or better selectivity towards
FAME positional and geometric isomers of unsaturated FA than
the most widely used cyanopropyl columns [9], reportedly with
column bleed reduction [7]. However, as phase polarity increases,
FAME differing in their number of carbons and number, geometric
configuration and/or position of double bonds increasingly over-
lap, so again co-elution is obtained. Longer columns can address
this problem, but the initial column cost and long run time are
drawbacks.
Multidimensional GC (MDGC) is an alternative method to over-
come co-elutions and potentially delivers improved separation for

0021-9673/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.chroma.2013.08.099
 Author's personal copy
A. Nosheen et al. / J. Chromatogr. A 1312 (2013) 118–123
selected groups of analytes [13]. However, the limited number
and frequency of targeted regions analysed in a single run lim-
its the applicability of MDGC due to the time requirement for
the completion of each heart-cut on the 2 D column before the
subsequent heart-cut. Comprehensive two-dimensional gas chro-
matography (GC × GC) achieves bi-dimensional separation across
the whole sample, with good peak capacity, and an enhanced sensi-
tivity arising from cryogenic or flow zone compression. The unique
feature of structurally related elution of compounds within the 2D
chromatograms is an oft-stated benefit [14,15]. Now a modulation
process introduces focussed analyte bands to a second dimension
(2 D) column for fast separation. The separation on the 2 D column
must finish before the next 1 D portion is introduced, otherwise
wrap-around occurs which leads to polar and non-polar analyte
overlap in the 2D space. Both NP/P [7,11,16,17] and P/NP column
combinations have been applied [11,17]. While a NP/P column set
gives excellent separation of multicomponent standard mixtures
[16], or for moderately abundant major components, it may fail at
separation of complex samples containing several geometric and
positional isomers of unsaturated FAME, as is the case for safflower
oil and linseed oil. IL columns, either as a 1 D or as a 2 D column, have
already been applied to GC × GC of fatty acids [7,11], essential oils
[18,19], and other applications [20–22]. An excellent review cov-
ering the application of ionic liquid columns in MDGC and GC × GC
has been recently published [23].
Here a fast GC × GC method is proposed based for the first time
to the best of our knowledge on two ionic liquid columns for FAME
profiling in different samples. The method has been applied to
selected samples where the separation of very abundant isomers
is required. Method parameters (linearity, limit of detection and
reproducibility) were validated.
2. Experimental
2.1. Chemicals and reagents
Dichloromethane (DCM), methanol, n-hexane (all chromatog-
raphy grade), boron trifluoride-methanol reagent (12.5%, for
GC derivatisation), nonadecanoic acid methyl ester and sodium
chloride (99.0%) were purchased from Supelco (Sigma–Aldrich,
Bellefonte, PA) while sodium hydroxide pellets were purchased
from Merck (Darmstadt, Germany). A 37-component FAME mixture
(Supelco, p/n 47885-U) was used for method optimisation and vali-
dation, where the individual FAME concentrations were between
2% and 6% of the total FAME concentration. The original 10 mg mL−1
(total FAME) concentration was progressively diluted with DCM to
six concentrations in the range 30–1000 mg L−1 . Note that the con-
centrations were recorded in terms of total FAME, rather than for
the individual FAME, and so each individual FAME will be 0.6–20.0;
1.2–40.0, or 1.8–60 mg L−1 depending on their composition in the
mixture.
2.2. Safflower oil extraction and sample preparation
Safflower seeds were extracted by Soxhlet apparatus using hex-
ane as a solvent, according to the AOCS Ag 1–65 method [24].
Safflower and linseed FAME were prepared according to the
AOCS standard method Ce 2-66 [25]. A stock solution of non-
adecanoic acid methyl ester (C19:0) was prepared at 1000 mg L−1
in DCM, and further diluted to 100 mg L−1 for use as an internal
standard (IS). Equal volumes of IS and FAME at various concentra-
tions (20× and 200× dilutions) were mixed just before GC injection.
Fish oil and butter samples were prepared according to Chin et al.
[26]. The upper hexane layer was collected and diluted by 200-fold
for GC analysis.
119
2.3. Instrumentation
A range of different ionic liquid columns (SLB-IL59, SLB-IL61,
SLB-IL76, SLB-IL82, SLB-IL100, SLB-IL111, all from Supelco) were
used to investigate the IL/IL arrangement. IL111 was used as the 1 D
column in all cases. For the final analytical method, the most polar
commercially available ionic liquid column IL111 (30 m × 0.25 mm;
0.20 mm) was used as 1 D, and the least polar IL59 (1 m × 0.1 mm;
0.08 mm) was used as the 2 D column in the IL/IL column set. FAME
were eluted under a moderately high temperature ramp: an ini-
tial temperature of 100 ◦ C was held for 1 min and then increased to
230 ◦ C at a rate of 8 ◦ C min−1 . The injector and detector tempera-
tures were 250 ◦ C and 260 ◦ C, respectively.
Different modulation periods (4 s, 5 s, 6 s, and 8 s) and modula-
tion temperatures (30 ◦ C, 60 ◦ C and 90 ◦ C isothermal) were applied
to determine best separation during the optimisation stage. FID
acquisition rate was set at 50 Hz. One microlitre aliquots of sam-
ples or standard mixtures were injected at 20:1 split ratio. Agilent
Chemstation was used for data acquisition and processing, an
in-house application for data conversion, and Fortner Transform
(Fortner, Inc., Savoy, IL) for data visualisation (2D plots). Accurate
mass analysis of FAME in safflower oil samples was performed on
an Agilent 7200 Q-TOF mass spectrometer.
3. Results and discussion
3.1. Separation of FAME on IL/IL column sets
Although IL columns have been reported in combination with
other stationary phase columns, including for FAME separation [9],
the present study employs combinations where an IL phase is cou-
pled with another IL phase. This is possible due to the relatively
large polarity differences between the IL phases, even though all
IL phases themselves are classified as polar. IL phase polarities
are available on the Sigma-Aldrich website [27]. For IL phases the
following McReynold’s constants (Supelco data) have been deter-
mined IL59 (2624), IL61 (2710), IL76 (3379), IL82 (3638), IL100
(4437) and IL111 (4938). By comparison, a 5% phenyl column is
characterised by a McReynold’s constant of about 252 and a Wax
phase about 2324 (Supelco phase data).
As expected from the polarity scale, an IL111/IL100 column
set (Fig. 1A) provided very little spread of the 37 FAME over the
2D space for the standard mixture, most likely due to the simi-
lar separation mechanisms for these components on this column
set. However, when the polarity of the 2 D column was decreased,
the components in the mixture increasingly spread over the 2D
space as evidenced by increased 2 D retentions between saturated
and polyunsaturated FAME (Fig. 1B–E). Not surprisingly, IL111/IL59
(Fig. 1E) gave the best spread of components; wrap-around on a
1 m × 0.1 mm ID 2 D column was not excessive. Interestingly, IL59
functions as a lower polarity column phase when coupled to the
IL111, giving a similar FAME profile as that on the SolGel-WAX/VF-5
column set (data not shown), even though IL59 has similar polarity
to SolGel-WAX. Certainly, phase selectivity also influences compo-
nent spread, but this is out of scope of this communication.
In all cases, the modulation process in these examples was per-
formed at the end of the 1 D IL111 column, rather than modulation at
the beginning of the 2 D column. Data demonstrate that the IL111
phase appears to provide effective modulation of FAME, since an
average peak width at half height of about 150 ms was obtained –
which is a very efficient value for solutes that wrap-around. Based
on the best spread, with minimum wrap-around, and good modula-
tion (peak shape) across the whole 2D plot, the IL111/IL59 column
set was selected as column set of choice for all analytical sample
analyses. All 37 FAME in the Supelco mixture were well separated
 Author's personal copy
120 A. Nosheen et al. / J. Chromatogr. A 1312 (2013) 118–123
Fig. 1. Contour plots of Supelco FAME mixture obtained on IL111 column as 1 D col-
umn combined with (A) IL100, (B) IL82, (C) IL76, (D) IL61, and (E) IL59 as 2 D columns.
Increased spread of the FAME, thus better use of separation space (increased degree
of orthogonality) is obtained when the 2 D phase polarity decreased. Each panel has
a range of 0–6 s for 2 D retention.
in less than 16 min, and their location in the 2D separation space
on this column set is presented in Fig. 2.
The study was extended to another 5 non-IL 2 D column phases,
all combined with IL111 as 1 D column. Note that the retention
of FAME on polar IL phase columns leads to significant reduction
in elution temperature (Te ) compared to non-polar or moderately
polar phases, in similar manner to polar cyanopropyl phases [28].
This suggests that the very high polarity IL111 phase may exhibit
Fig. 2. Contour plot of Supelco FAME mixture on IL111/IL59 column set. All 37
FAME were baseline separated in just over 15 min. C4:0, C6:0 and C8:0 do not fit the
saturated FAME trend because they are not modulated well under these conditions.
Fig. 3. Contour plots of Supelco FAME mixture obtained on IL111 column as 1 D
column combined with (A) BPX70, (B) INNOWax, (C) BPX50, (D) BPX35, and (E) BPX5
as 2 D columns. The column sets in D and E appear to exhibit too much separation
with excessive wrap-around. Each panel has a range of 0–6 s for 2 D retention.
excessive polarity difference when coupled with very low polarity
2 D phases (e.g. a 5% phenyl phase), owing to the low T of FAME on
e
the IL111 phase and consequent high retention factor for non-polar
solutes.
As the 2 D phase polarity decreased (Fig. 3A–E), the retention on
the 2 D column (2 tR ) increased, and may produce multiple wrap-
arounds for later eluting FAME (i.e. lower polarity FAME) on the
least polar columns (BPX35 and BPX5). As a consequence, the aver-
age peak width at half height (wh ) increased proportionally. The
only way to counter this is to have either very thin phase coatings
(i.e. a very high phase ratio), or an excessively short 2 D column. On
the other hand, the polar cyanopropyl BPX70 phase produced only a
small spread of the FAME, with a profile similar to the one obtained
for IL111/IL76. Wax and BPX50 2 D columns produced relatively
good separation, though still with appreciable wrap-around.
An interesting “exponential” trend in the 2 D retention values of
saturated FA on IL111/BPX5 was observed (Fig. 4), when the orig-
inal 2D plot in Fig. 3E was reconstructed by taking into account
the known mixture composition and the possible number of wrap-
arounds. This is mimicked by the Cx:1 FAME series. FAME located
in the panel 0–6 s were already wrapped around once. FAME which
showed more than one wrap-around were placed in the corre-
sponding 6–12 s, 12–18 s or 18–24 s panels, and removed from
the original 0–6 s panel. As the retention of higher carbon FAME
increased, the peak width at half-height (wh ) increased propor-
tionally, being 150 ms for C9:0, 250 ms for C15:0, and 600 ms for
C24:0. Polyunsaturated FAME showed peak widths similar to those
of saturated FAME with similar 2 tR values. Thus both C14:1 and
 Author's personal copy

A. Nosheen et al. / J. Chromatogr. A 1312 (2013) 118–123 121

Fig. 4. Reconstructed 2 D time corresponding to the IL111/BPX5 column set data

shown in Fig. 3E. These data are simply based on the 2 D positions of the peaks in the
panel shown in Fig. 3E, translated to respective panels from 6–12, 12–18 or 18–24 s,
and removed from the original 0–6 s panel. Amount injected in the column was
approximately 1 ng per component.

C20:5 had wh of about 150 ms; C23:1 had wh ∼ 450 ms. From these
data, it can be estimated that C24:0 FAME had a theoretical plate
number of up to 11,000, assuming that its 2 tR is about 27 s. This
is an excellent efficiency for the 1 m column. On this column set,
FAME C15:0 and C14:1 both with 1 tR ∼ 8.5 min have a difference
in 2 tR of about 2.15 s, and Rs ∼ 6.6, corresponding to the vertical
displacement between the Cx:0 and Cx:1 lines. The approximate
exponential increase in 2 tR for the saturated FAME under the mod-
erately high temperature ramp of 8 ◦ C min−1 is unusual in GC × GC
and it is worth further study. It may be partly contributed by a
Fig. 5. Contour plots of (A) safflower oil, (B) linseed oil, (C) butter, and (D) fish oil
reduced flow velocity as T increases. Use of a second oven for inde-
sample on the IL111/IL59 column set. Panels A and B have a range of 0–6 s, and
pendent 2 D column oven temperature programme incrementing panels C and D a range of 0–4 s for 2 D retention. C19:0 is added to the safflower and
could possibly give better spread and peak shape, and permit eas- linseed oil samples as an internal standard.
ier optimisation, but this has not been explored since this option
was not available for this study. Given the excessive 2 tR for FAME
FAME (C15:0 and C17:0) were detected and reported for the first
with the IL111/BPX5 set, this suggests that these phases exhibit
time in safflower oil. The identification was confirmed using accu-
too disparate polarity, although one might believe that they ‘max-
rate mass time-of-flight mass spectrometry. Fig. 5A and B clearly
imise orthogonality’. In this case, using BPX5 without wrap-around
demonstrates that C20:0 is completely separated from the highly
would require a much shorter 2 D column, a much thinner 2 df film
abundant C18:2 component, which co-elutes on the 1 D column,
[29], or a longer or more retentive (or thicker film) 1 D column.
owing to the excellent modulation and narrow peaks of the major
Based on the vertical separation between C14:1 and C15:0
C18 isomers. It is common to use polar column phases in the 1D
(these two solutes co-elute on the 1 D IL111 column phase at about
separation of FAME, but on these columns co-elution of polyunsatu-
8.5 min, refer to Figs. 1 and 3) the following separation magnitudes
rated fatty acids (PUFA) of lower carbon number often overlap with
were observed on the 2 D columns when coupled with IL111:
saturated and monounsaturated FAME of higher carbon number,
IL100 < IL82 < IL76 < BPX70 < IL61 < IL59 < Wax < BPX50 which can lead to confounded elution overlaps. GC × GC overcomes
such problems by locating FAME with different degrees of unsatu-
< BPX35 < BPX5. ration at different 2 tR positions: the spread in 2D space is based on
two complementary separation mechanisms.
The IL111/IL59 column set was additionally applied for sep-
aration of FAME in safflower oil, linseed oil, fish oil and butter 3.2. Method performance
(Fig. 5A–D). Excellent separation of major C18 FAME was obtained
for safflower and linseed oils, and considerably less peak broad- Repeatibility of peak area for the most abundant FAME in ana-
ening for these abundant FAME was obtained compared with a lysed samples was calculated from 5 replicates at two different
NP/P column set observed (data not shown). Also a considerably concentrations (25 mg L−1 and 250 mg L−1 ). The results are given in
greater number of other FAME were also detected at the injected Table 1. Generally, an RSD of <10% has been obtained for saturated
concentration; it is possible to inject higher concentrations without and unsaturated fatty acids.
excessive broadening of major components, and so this enhances Limit of detection (LOD) was calculated as the FAME con-
minor component delivery to the column, and their detection. Two centration which gives signal-to-noise ratio of 3, and limit of
 Author's personal copy

122 A. Nosheen et al. / J. Chromatogr. A 1312 (2013) 118–123

Table 1
Method performance parameters (linearity, r2 , LOD) obtained from 5 replicates at two different concentrations (RSD), or from 5 different concentration levels.

FAME Linearity equation* r2 RSD (%)# LOD (mg L−1 )

2× LOD 20× LOD

C16:0; palmitic y = 0.000471x + 0.012 0.999 9.0 3.4 0.16

C18:0; stearic y = 0.000467x − 0.019 0.998 9.8 9.5 0.16
C18:1; oleic y = 0.000585x − 0.012 0.999 7.4 7.6 0.16
C18:2; linoleic y = 0.000166x + 0.000 0.999 7.8 4.6 0.21
C18:3; linolenic y = 0.000162x + 0.002 0.999 9.0 8.4 0.21
#
RSD are given for solution concentrations of approximately 2× and 20× LOD.
*
y = relative response area of the modulated peaks for each solute vs. that for internal standard; x = concentration of total FAME level, not accounting for each individual
FAME. The small slope arises from the relatively high internal standard amount (100 mg L−1 ).

Table 2 mixtures. The greatest spread of components in 2D space applies

List of FAME detected or quantified in selected safflower oil samples, with their
to the case where the highest and the lowest polarity IL columns
relative amounts. Concentrations less than 0.2% are not reported. Fatty acids in bold
have not been reported previously in this type of oil. were coupled, in this case as an IL111/IL59 column set. The spread
of components on the 2 D column (IL59) was excellent even at a
Fatty acid Concentration (%)
moderately high temperature ramp of 8 ◦ C min−1 , which resulted
1 Myristic acid (C14:0) in a fast overall runtime of less than 16 min. In addition, detection
2 Pentadecanoic acid (C15:0) of many, less abundant fatty acids, can be readily determined, and
3 Palmitic acid (C16:0) 4–6.39
two new FAME were reported in safflower oil for the first time.
4 Palmitoleic acid (C16:1)
5 Heptadecanoic acid (C17:0) Method performance was validated on a 37-component FAME
6 Ginkgolic acid (C17:1) mixture. Excellent linearity, peak area repeatability and LOD were
7 Stearic acid (C18:0) 2–3.6 obtained. The proposed method was successfully applied for anal-
8 Oleic acid (C18:1) 10–15.83
ysis of FAME derivatives in several food samples (safflower oil,
9 Linoleic acid (C18:2) 73–79
10 Linolenic acid (C18:3) 0.4–0.7 linseed oil, fish oil, butter). Good agreement with prior results
11 Arachidic acid (C20:0) 0.2–0.3 was obtained for abundances of the major FA in safflower oil. The
12 Eicosadienoic acid (C20:2) method is especially attractive for samples where some compo-
13 Eicosatrienoic acid (C20:3) nents are of large abundance (as in safflower oil and linseed oil),
14 Behenic acid (C22:0)
and with low abundant FAME present.
15 Docosahexaenoic acid (C22:6n3)
16 Lingoceric acid (C24:0)
17 Nervonic acid (C24:1)
Acknowledgements

quantification (LOQ) as the concentration which gives S/N of 10.
Similar LOD and LOQ were obtained for most of the FAME due to
their structural similarities. The results are given in Table 1. Lower
LOD/LOQ could be obtained if splitless instead of split mode of
injection was applied.
Linearity was checked across the concentration range
1.25–60 mg L−1 for different individual FAME, at 5 different
concentration levels. Excellent linear correlation was obtained
for all FAME, with coefficients ranging from 0.998 to 0.999 (see
Table 1).
Representative samples from safflower oil and linseed oil were
quantitatively analysed using the proposed method. Amongst the
prior reported FAME in safflower oil, two other unreported FAME
were detected in almost all samples and unambiguously identified
using high resolution MS. This includes C15:0 (0.3 ppm mass accu-
racy) and C17:0 (−0.4 ppm mass accuracy) for their corresponding
molecular ions. Even though quantitative data were not determined
for FAME in other samples, it is demonstrated here that the pro-
posed method offers a fast profiling approach for FAME in a variety
of food and related samples, which includes safflower oil, linseed
oil, fish oil and butter. Data for safflower oil generally agreed with
previous data [30]. The combination of IL111/IL59, with fast elution
on the very polar first column, and suitable elution performance on
the rather polar IL59 (without too much wrap-around) contributes
to the fast overall analysis time of just above 15 min (Table 2).
4. Conclusion
Ionic liquid columns, even though all exhibited a polar nature,
showed interesting trends in their degree of ‘orthogonality’ when
coupled in the GC × GC experiment, for separation of FAME
A.N. is grateful to the Higher Education Commission of Pakistan
for providing an IRSIP scholarship under the International Research
Support Initiative Program, and Agilent Technologies for facility
support. Sigma–Aldrich is acknowledged for their generous pro-
vision of ionic liquid phase columns.
References
[1] O.O. Fasina, H. Hallman, M. Craig-Schmidt, C. Clements, J. Am. Oil Chem. Soc.
83 (2006) 899.
[2] M. Vosoughkia, M. Ghavamib, M. Gharachorloo, M. Sharrifmoghaddasi, A.H.
Omidi, Adv. Environ. Biol. 5 (2011) 897.
[3] U. Gecgel, M. Demirci, E. Esendal, M. Tasan, J. Am. Oil Chem. Soc. 84 (2007) 47.
[4] X.J. Han, L.M. Cheng, R. Zhang, J.C. Bi, J. Food Eng. 92 (2009) 370.
[5] P.D. Patil, S.G. Deng, Fuel 88 (2009) 1302.
[6] U. Rashid, F. Anwar, Energy Fuels 22 (2008) 1306.
[7] Q. Gu, F. David, F. Lynen, P. Vanormelingen, W. Vyverman, K. Rumpel, G.W. Xu,
P. Sandra, J. Chromatogr. A 1218 (2011) 3056.
[8] P. Delmonte, A.R. Fardin-Kia, J.K.G. Kramer, M.M. Mossoba, L. Sidisky, C. Tybur-
czy, J.I. Rader, J. Chromatogr. A 1233 (2012) 137.
[9] P. Delmonte, A.R.F. Kia, J.K.G. Kramer, M.M. Mossoba, L. Sidisky, J.I. Rader, J.
Chromatogr. A 1218 (2011) 545.
[10] C. Ragonese, P.Q. Tranchida, P. Dugo, G. Dugo, L.M. Sidisky, M.V. Robillard, L.
Mondello, Anal. Chem. 81 (2009) 5561.
[11] C. Villegas, Y. Zhao, J.M. Curtis, J. Chromatogr. A 1217 (2013) 775.
[12] K. Shimizu, Y. Ando, J. Oleo Sci. 61 (2012) 421.
[13] A.X. Zeng, S.-T. Chin, P.J. Marriott, J. Sep. Sci. 36 (2013) 878.
[14] P.J. Marriott, T. Massil, H. Hugel, J. Sep. Sci. 27 (2004) 1273.
[15] B. Vlaeminck, J. Harynuk, V. Fievez, P. Marriott, Eur. J. Lipid Sci. Technol. 109
(2007) 757.
[16] W. Tiyapongpattana, P. Wilairat, P.J. Marriott, J. Sep. Sci. 31 (2008) 2640.
[17] P. Manzano, E. Arnaiz, J.C. Diego, L. Toribio, C. Garcia-Viguera, J.L. Bernal, J.
Bernal, J. Chromatogr. A 1218 (2011) 4952.
[18] C. Cagliero, C. Bicchi, C. Cordero, E. Liberto, B. Sgorbini, P. Rubiolo, J. Chromatogr.
A 1268 (2012) 130.
[19] C. Ragonese, D. Sciarrone, P.Q. Tranchida, P. Dugo, G. Dugo, L. Mondello, Anal.
Chem. 83 (2011) 7947.
 Author's personal copy
A. Nosheen et al. / J. Chromatogr. A 1312 (2013) 118–123
[20] M. Zapadlo, J. Krupcik, P. Majek, D.W. Armstrong, P. Sandra, J. Chromatogr. A
1217 (2010) 5859.
[21] V.R. Reid, J.A. Crank, D.W. Armstrong, R.E. Synovec, J. Sep. Sci. 31 (2008) 3429.
[22] J.V. Seeley, S.K. Seeley, E.K. Libby, Z.S. Breitbach, D.W. Armstrong, Anal. Bioanal.
Chem. 390 (2008) 323.
[23] C. Ragonese, D. Sciarrone, P.Q. Tranchida, P. Dugo, L. Mondello, J. Chromatogr.
A 1255 (2012) 130.
[24] AOCS, Official Methods and Recommended Practices of the American Oil
Chemists’ Society, 4th ed., American Oil Chemists’ Society Press, Champaign,
IL, USA, 1993.
View publication stats
123
[25] AOCS, Official Methods and Recommended Practices of the American Oil
Chemists’ Society, 4th ed., American Oil Chemists’ Society Press, Champaign,
IL, USA, 1990.
[26] S.-T. Chin, Y.B.C. Man, C.P. Tan, D.M. Hashim, J. Am. Oil Chem. Soc. 86 (2009)
949.
[27] http://www.sigmaaldrich.com/content/dam/sigma-aldrich/countries/czech/
gc kolony.pdf (accessed 07.05.13).
[28] J. Harynuk, P.M. Wynne, P.J. Marriott, Chromatographia 63 (2006) S61.
[29] B.S. Mitrevski, P. Wilairat, P.J. Marriott, J. Chromatogr. A 1217 (2010) 127.
[30] B. Bozan, F. Temelli, Bioresour. Technol. 99 (2008) 6354.