Food Control 44 (2014) 191e197

Contents lists available at ScienceDirect

Food Control
journal homepage: www.elsevier.com/locate/foodcont

The analysis of trans fatty acid profiles in deep frying palm oil and
chicken fillets with an improved gas chromatography method
Yi Chen 1, Ying Yang 1, Shaoping Nie, Xi Yang, Yuting Wang, Meiyan Yang, Chang Li,
Mingyong Xie*
State Key Laboratory of Food Science and Technology, Nanchang University, Nanchang 330047, PR China

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

Article history: An improved gas chromatography method for the simultaneous separation of 52 fatty acids (FAs) has
Received 21 December 2013 been developed. For both oleic acid and linoleic acid, a good resolution was achieved for their positional
Received in revised form and geometrical (cis/trans) isomers. This method was validated to be precise, accurate and sensitive. With
2 April 2014
this method, the FAs profiles in palm oil and chicken fillets were analyzed. In general, small changes were
Accepted 3 April 2014
Available online 13 April 2014
observed in the composition of FAs and formation of trans isomers after 8 h frying at the temperature
lower than 200
C. However, with extreme deep-frying process, the thermal degradation and TFAs
formation in frying oil increased in direct proportion to frying temperature and time. Moreover, the FAs
Keywords:
trans Fatty acids
composition of fried chicken fillets was found to be mostly in correspondence with that of the frying oil,
Gas chromatography which might be due to the oil absorption or interaction between the frying oils and frying materials.
Palm oil Ó 2014 Elsevier Ltd. All rights reserved.
Fried chicken fillets

1. Introduction
trans Fatty acids (TFA) refer to a group of unsaturated fatty acids
that contain one or more isolated, non-conjugated, double bonds in
a trans geometric configuration. It has been reported that a higher
intake of TFA may increase the risk of cerebrovascular disease
(Mozaffarian, Katan, Ascherio, Stampfer, & Willett, 2006), coronary
heart disease (Oomen et al., 2001), diabetes (Kim, Chunawala, &
Linde, Reaven, 2006) and even breast cancer (Kohlmeier, 1997). In
light of this, Food and Drug Administration (FDA) and Canadian
Food Inspection Agency (CFIA) have issued the rule requiring
manufacturers to include TFA contents in food products label (CFIA,
2008; FDA, 2003). This rule prompts the urgent need to develop a
simple and accurate method for the quantification of TFAs in food
products.
TFAs in foods are mainly derived from the partially chemical
hydrogenation of vegetable and fish oils. Deep frying has been
considered a source for the production of TFAs. As one of the most
common methods of food cooking, frying is a process of
* Corresponding author. State Key Laboratory of Food Science and Technology,
Nanchang University, 235 Nanjing East Road, Nanchang, Jiangxi Province 330047,
PR China. Tel.: þ86 791 83969009; fax: þ86 791 83969069.
E-mail address: myxie@ncu.edu.cn (M. Xie).
1
These authors contributed equally to this work and should be considered co-
first authors.
immersing food in hot oil with a contact among oil, air and food at
high temperatures from 150
C to 200
C. The simultaneous heat
and mass transfer of oil, food and air during deep-fat frying pro-
duces a food product with desired sensory characteristics,
including fried food flavor, golden brown color and a crisp texture,
which is very popular to consumers. However, this process will
also initiate the thermal oxidative deterioration of unsaturated
FAs, which is considered to be linked with TFAs accumulation in
edible oils by heating or frying. The formation of TFAs during
frying has been investigated in several hydrogenated vegetable
oils. Aladedunye and Przybylski (2009) reported that the final
concentrations of total TFAs after 7 days frying of French fries in
canola oil increased from its initial value of 2.4e3.3% at 185  5
C,
to 5.9% at 215  5
C. Similar results have been reported by Yang,
Yang, Nie, Xie, and Chen (2012) for corn oil. However, Romero,
Cuesta, and Sánchez-Muniz (1998) reported a very minimal pro-
duction of elaidic acid (less than 0.5 g/100 g) after 20 frying cycles
of extra virgin olive oil, high oleic sunflower oil and regular sun-
flower oil. Tsuzuki, Nagata, Yunoki, Nakajima, and Nagata (2008)
also reported that statistically significant differences were
observed for the amounts of trans 18:1 FAs in rice bran oil, saf-
flower oil and sesame oil heated in a glass tube at 180
C for 4 h.
Thus these previous studies indicated that several factors such as
the frying conditions, the FA composition and coexisted antioxi-
dants in frying materials and even the methods of TFAs mea-
surements would contribute to the variation in the thermal TFAs

http://dx.doi.org/10.1016/j.foodcont.2014.04.010
0956-7135/Ó 2014 Elsevier Ltd. All rights reserved.
192 Y. Chen et al. / Food Control 44 (2014) 191e197
accumulation profiles. Furthermore, when the frying materials
containing TFAs were introduced to frying process, the trans level
in frying materials and the frying oils can affect each other due to
fast exchange of fats during frying (Zhou, Tan, Neo, & Lo, 2009).
Thus respective monitoring of TFAs accumulation in various frying
materials and frying oil will be necessary in order to investigate
TFAs formation in edible oils during frying.
The analysis of trans fatty acid is extremely challenging and
complex due to the various positional isomers of unsaturated fatty
acids. The reported analytical approaches mainly include silver-ion
thin-layer chromatography (Agþ-TLC), infrared spectroscopy (IR),
reversed phase high performance liquid chromatography (RP-
HPLC), gas chromatography (GC), GCeMS, and comprehensive
two-dimensional gas chromatography. The IR method is consid-
ered to be inaccurate and unreliable for quantifying isolated trans
fats due to the interference of absorption from other functional
groups other than the trans bond, especially in samples containing
below 5% trans fats (Albuquerque, Costa, Castilho, & Sanches,
2011). Agþ-TLC is able to separate the trans isomer from the cis,
but for quantitative analysis of trans fat, the isolated trans fraction
still needs to be analyzed with subsequent capillary GC method
(Golay, Dionisi, Hug, Giuffrida, & Destaillats, 2006). RP-HPLC also
has several disadvantages, such as complexity of derivatization
method for fatty acids, limited availability of derivatives standards
and lower separation efficiency compared with capillary GC. The
drawbacks outlined above can be partly avoided by GC with mass
spectrometry (MS) detectors (Manzano, Diego, Nozal, Bernal, &
Bernal, 2012; Schröder & Vetter, 2013) and the use of multidi-
mensional gas chromatography (Manzano, Arnáiz, et al., 2012;
Manzano, Diego, et al., 2012; Tranchida, Franchina, Dugo, &
Mondello, 2012), which can be a good alternative to determine
minority fatty acids and quantify the different isomers. However,
these particular instruments are expensive and bulky and not
typically used in most small enterprises or laboratories. In com-
parison, the most acceptable and conventional method is GC
coupled with a flame ionization detector (GC-FID) with appro-
priate accuracy, convenience and cost (Antolín, Delange, &
Canavaciolo, 2008). However, due to the complexity of the fatty
acid composition in different food, GC-FID methods for the anal-
ysis of fatty acids described so far suffer from analysis times longer
than 80 min, no separation or bad resolution of geometric and
positional isomers. For instance, Berdeaux, Dutta, Dobarganes, and
Sébédio (2009) reported in their review that numerous peaks
were poorly resolved, or eluted together in one peak, or co-eluted
with cis-18:1 isomers even using a 100-m BPX70 capillary column
when mono-trans-18:2 or di-trans-18:2 positional isomers are
present. Similarly, Sébédio and Ratnayake (2008) developed
several improved GC methods on highly polar columns such as
100 m CPSil 88 or equivalent, however, a complete separation of
the seven geometrical isomers from linolenic acid still cannot be
achieved. Actually, individual separation of each isomer is essen-
tial for an optimal quantitation of trans fatty acids in various foods,
preventing any risk of under- or over-estimation due to the
method. Therefore, future developments to simplify and improve
the current GC procedure will be in great need to meet the di-
versity of food products.
The aim of this study was to optimize a simple and reliable GC
method based on the AOCS Official Method (AOCS, 2005) for the
determination of FAMEs, and to prove its applicability for deter-
mining the fatty acid profile especially the trans fat in frying food
samples. Palm oil is commonly used by consumers as a medium
to fry food. However, the effects of deep-frying on the TFA for-
mation of unhydrogenated palm oils during the frying process,
have received little attention. Therefore, TFAs formation during
frying process was investigated using chicken fillets and the
commercially available palm oil, as the frying materials and the
frying oil.
2. Materials and methods
2.1. Chemicals and reagents
Reference standards GLC-463, internal standards mono-
heneicosanoin (C21:0, TAG) and methyl heneicosanoate (C21:0,
FAME) were purchased from Nu-Chek Prep, Inc. (Elysian, MN,
USA). FAMEs of C18:2 isomer mixture (No. 47791) and C18:3 iso-
mer mixture (No. 47792) were bought from Supelco Inc. (Belle-
fonte, PA, USA). Standard FAMEs (C12:0, C14:0, C16:0, 9c C16:1,
C17:0, C18:0, 9t C18:1, 9c 12t C18:2, 9t 12c C18:2, 9c 12c C18:2,
C20:0, 9c 12c 15c C18:3) were purchased from Sigma (St. Louis,
MO, USA). HPLC grade n-heptane was obtained from Fisher Sci-
entific (Fair Lawn, NJ, USA). Palm oil was bought from South Sea
Oils & Fats Industrial Ltd. (Shenzhen, China) and chicken fillet was
bought from local supermarket (Shengyuan Food Co., Ltd., China).
All solvents and chemicals were of analytical grade and used
without further purification unless otherwise specified. Stock
standard solutions were prepared in n-hexane and stored at 4
C
until use. Potassium hydroxide, ethyl alcohol, ether, phenol-
phthalein, hydrochloric acid, peroxide, chloroform, methanol,
ferrous chloride, potassium thiocyanate and reduced iron powder
were of analytical grade.
2.2. Frying procedure
Palm oil (3 L) was poured into a commercial electrical fryer
(Laierjia Co., Shanghai, China), which could control the oil tem-
perature within 5
C from the set temperature. The oil was
heated to 150, 200 and 250
C, and then a batch of 40  5 g
chicken fillets was added and fried for 10 min as one frying cycle
at each temperature. In each day, the oil was continuously heated
for 8 h with 48 frying cycles performed. Frying oil (10 mL) was
collected every 2 h with 12 frying cycles. The first, 24th and the
last batches of fried chicken fillets were also collected. All the
samples were stored at 20
C for further chemical analysis.
2.3. Lipid extraction from chicken fillets
The total lipids contained in fried chicken fillets were
extracted according to the method by Folch, Lees, and Sloane-
Stanley (1957) with minor modification. Briefly, 15 g of crushed
chicken fillets sample was poured into a round bottom flask,
followed by addition of 60 mL chloroform-methanol (1:1, v/v)
and homogenization for 1.5e2.0 min. Successively, the mixture
was reflux-heated in a water bath at 60
C for 1 h and then
cooled to room temperature. After being filtered through a fluted
filter paper, the resultant filtrate was collected and dried under
reduced pressure using a vacuum evaporator. Then the fat ob-
tained was resolved in 25 mL petroleum ether, and mixed with
5 g sodium sulfate anhydrous to further remove the residual
solvent. Subsequently, the mixture was separated by centrifu-
gation at 3000 g for 5 min and the supernatant was collected,
evaporated and dried at 105
C in succession. The residue was
then weighed and stored at 20
C.
2.4. Preparation of fatty acid methyl esters (FAMEs)
Corresponding FAMEs of the oil samples and the lipids extracted
from fat samples were prepared by esterification with potassium
hydroxide following the AOCS Official Method Ce 2-66 (AOCS, 1997)
with minor modification. Approximately 10 mg of sample was
 Y. Chen et al. / Food Control 44 (2014) 191e197
dissolved in 2 mL of n-heptane in a tube with tight sealing cap, and
0.1 mL of methanolic potassium hydroxide solution (2 mol/L) was
added. The tube was capped tightly and shaken vigorously for
about 30 s. The solution was then centrifuged at 3000 g for 5 min,
and 1 mL of n-hexane upper phase was transferred into 2 mL auto-
sampler vial for subsequent GC analysis. Monoheneicosanoin
methyl ester (C21:0, 5 mg/mL) was used as the internal standard for
the GC analysis.
2.5. Gas chromatography
Analysis of FAMEs was performed on Agilent 6890N gas chro-
matograph equipped with a flame ionization detector (FID) and a
split injector. A highly polar capillary column, CP-Sil 88
(100 m  0.25 mm i.d.  0.39 mm o.d., 0.20 mm film thickness;
Varian Inc., USA) coated with a 100% cyanopropyl polysiloxane
stationary phase was selected for the analysis. The column tem-
perature was programmed as follows: initial temperature 60
C
(held for 5 min); increased to 170
C at 11.5
C/min (held for
25 min); further increased to 200
C at 5
C/min (held for 5 min);
finally ramped to 215
C at 2
C/min with a final holding time of
20 min. The detector and injector temperatures were maintained at
250
C. Hydrogen was used as carrier gas at a velocity of 26 cm/sec.
A split ratio of 10:1 was used, and samples were injected into the
GC for analysis.
The peaks of FAMEs in the samples were identified by
comparing their retention time to that of the corresponding refer-
ence samples. Quantification of individual FAs was performed ac-
cording to AOCS Official Method Ce 1h-05, and theoretical flame
ionization detector correction factor (TCF) was used to calculate the
amount of individual fatty acids.
2.6. Methodology validation
Intra- and inter-day variability was chosen to evaluate the pre-
cision of the developed method. For the intraday variability test, the
mixed standards solution was analyzed for six replicates within 1
day, while for the inter-day variability test, the solution was
examined once per day for consecutive 6 days. Variations of the
peak area were taken as the measures of precision, and expressed
as percentage of relative standard deviations (RSD).
Sample stability was tested by periodically analyzing the sample
solution kept at room temperature for various periods (2, 4, 6, 8, 10
and 12 h). The RSDs were used as the indicator of stability.
For recovery studies, known volumes of standard solutions
(C12:0, C14:0, C16:0, 9c C16:1, C17:0, C18:0, 9t C18:1, 9c 12t C18:2, 9t
12c C18:2, 9c 12c C18:2, C20:0, 9c 12c 15c C18:3) were added to a
test oil sample at three levels. The spiked samples were analyzed
using the method described above.
The limits of detection (LOD) and quantification (LOQ) under the
present chromatographic conditions were determined at a signal-
to-noise ratio (S/N) of about 3 and 10, respectively.
2.7. Measurements of peroxide value (POV) and acid value (AV)
The peroxide value and acid value were determined following
the national standard of the People’s Republic of China (Code of
China). For determination of POV, 5 g oil samples were weighed
into a conical flask and 30 mL of solvent mixture of acetic acide
isooctance in the ratio of 3:2, respectively, were added to the oil
samples. 50 mL of acetic acid/isooctane solution was added to a
conical flask and the stopper was replaced. The flask was swirled
until the sample was dissolved. Then 0.5 mL of saturated potas-
sium iodide (KI) solution was added to the solution and allowed to
stand for 1 min, thereafter 30 mL of distilled water were added
193
and titrated with 0.01 N sodium thiosulfate solution using starch
indicator until the yellow color was discharged. A blank was
prepared alongside the oil samples. Peroxide value was calculated
as Eqn. (1):
10  ðV1  V2 Þ
Peroxide value ¼ (1)
m
where V1 is the volume of Na2S2O3 for determination of test sample
in mL, V2 is the volume of Na2S2O3 for determination of blank so-
lution in mL, and m is mass of test portion in g (5 g).
For determination of AV, each oil sample (1.0 g) was weighed and
dissolved with 50 mL of natural ethereethanol mixture in a conical
flask. Two drops of phenolphthalein indicator were added and titrated
to pink end point (which persisted for 15 min) with 0.1 N potassium
hydroxide solution. Acid value was calculated according to Eqn. (2):
56:1  V  C
Acid value ¼ (2)
m
where 56.1 is equivalent weight of KOH, V is the volume in mL of
standard volumetric KOH solution used, C is the exact concentra-
tion in KOH solution used (0.1 N); m is the mass in grams of the test
portion (1 g).
2.8. Statistical analysis
All experiments were performed in triplicates. The data are
given as means  SD. The results were analyzed by t-test to identify
significant differences among groups. P values <0.05 were
considered significant. All analyses were performed using SPSS
(SPSS Statistics 17.0).
3. Results and discussion
3.1. Optimization of GC operating conditions
In real samples it is necessary to separate positional and geo-
metric isomers of fatty acids to avoid mis-interpretation (Antolín
et al., 2008). Since it is reported that polar columns are well
suited for isomer separation, a high polar column coated with
cyanopropyl polysiloxane was chosen for this method. On this
column, starting temperature of the column oven and the rate of
the temperature program were optimized. Previous studies have
shown that the most difficult regions of separation were the 18:1,
18:2 and 18:3 (Aro et al., 1998). Therefore, the resolution of these
regions could be used as an index for optimizing the temperature
condition. Compared with isothermal temperature condition which
was specified by AOCS Official Method Ce 1h-05 (at the oven
temperature of 180
C with the flow rate of carrier gas at 1 mL/min)
and other previously reported GC methods (Hou, Wang, Wang, Xu,
& Zhang, 2012), the temperature program that we optimized
showed a much better separation in trans 18:1, trans 18:2 and trans
18:3 regions. Fig. 1A shows a typical GC chromatogram of a FAME
standard mix containing 52 fatty acids with chain lengths between
C4 and C22 under the optimized conditions. The following char-
acteristics of retention time concerning position and geometry of
FAs could be found: Rt Cn < Rt Cn þ 1,2,3 .; Rt trans-Cn < Rt cis-Cn.
For instance, 9t C18:1 was eluted before 9c C18:1, 9c 12t C18:2; and
9t 12c C18:2 were eluted prior to 9c 12c C18:2, 9c 12c 15c C18:3.
Under the GC analysis conditions, all FAMEs were adequately
resolved except a partial overlap from C20:0 to 9c 12c 15c C18:3
(Fig. 1A, peak 1e9), which were also reported in the previous
studies (AOCS, 2005; Liu, Inbaraj, & Chen, 2007; Tsuzuki et al.,
2008). However, this partial overlap will not affect the
194 Y. Chen et al. / Food Control 44 (2014) 191e197

Fig. 1. The separate chromatography of different samples: (A) mixed fatty acid methyl esters standards (Region 2: peak 1-9t 12t 15t C18:3, peak 2-C20:0, peak 3-9t 12t 15c/9t 12c 15t
C18:3, peak 4-6c 9c 12c C18:3, peak 5-9c 12t 15t/9c 12c 15t C18:3, peak 6-5c C20:1, peak 7-8c C20:1, peak 8-11c C20:1, peak 9-9c 12c 15c C18:3), (B) fried palm oil, (C) fresh chicken
fillets and (D) fried chicken fillets.

quantitation of the FAs in real samples since the long chain FAs
(C  20) are usually not present in substantial quantities. Thus, this
GC condition was adequate in terms of retention time and sepa-
ration for TFAs measurement in edible oils.
3.2. Method valuation
As shown in Table 1, the different ratios of LOQ and LOD may be
derived from the difference of GC response to the analytes, and the
LODs were between 0.42 and 2.50 mg/mL, and the LOQs were be-
tween 1.39 and 9.09 mg/mL, respectively. These results showed that
this method was sensitive enough for this purpose.
The RSD of the intra-day variability ranged from 0.38 to 1.82%
for 22 fatty acid methyl esters standards, while the RSD of the inter-
day variability ranged from 2.19% to 5.67% (Table 1). Koning, Meer,
Alkema, Janssen, and Brinkman (2001) reported a GC method for
the separation of fatty acids in fish oil, in which the RSD of
variability were in the range of 1.5%e11%. These values confirmed
the precision of our new GC method.
The stability test data were shown in Table 2. The storage sta-
bility (RSD) for fatty acid methyl esters in palm oil was in the range
of 1.82% (9c 12c 15c C18:3)e2.62% (9c 12t C18:2). These results
suggest that the sample after derivatization was stable for at least
12 h at room temperature.
The recoveries of 13 fatty acid standards were found in the range
of 89.84%e115.72% (Table 2). According to the literature reported by
Cajak, Hajslova, Lacina, Mostovska, and Lehotay(2008), the
acceptable range for percent recovery was 70%e120% with
a RSD < 20%. Owing to some inevitable errors either from the
preparing of samples or chromatographic integration, some of the
recoveries in our manuscript were above 110%, however, they were
still within the acceptable limits. Moreover, these values were
higher than that reported previously (Indarti, Majid, Hashim, &
Chong, 2005), in which the recovery was shown to be as low as

Table 1
Stability data, detection and quantitation limits of 22 fatty acid methyl esters standards analyzed by GC.

FAME Intra-daya Inter-dayb LOD (mg/mL) LOQ (mg/mL) FAME Intra-day Inter-day LOD (mg/mL) LOQ (mg/mL)

Variability Variability Variability Variability

RSD (%) RSD (%) RSD (%) RSD (%)

C12:0 0.88 4.59 0.42 1.39 11c C18:1 1.41 5.67 1.50 4.67
C14:0 0.96 4.51 0.48 1.61 9t 12t C18:2 0.60 3.88 1.07 3.75
C16:0 0.68 4.38 0.92 2.60 9c 12t C18:2 0.69 4.18 0.78 2.59
9t C16:1 0.56 4.32 0.75 2.28 9t 12c C18:2 0.38 5.05 0.75 2.50
9c C16:1 0.71 4.37 1.03 3.00 9c 12c C18:2 0.68 4.18 0.75 2.03
C17:0 0.76 3.92 1.30 3.06 9t 12t 15t C18:3 1.03 4.02 0.80 2.78
C18:0 0.74 3.50 1.58 4.62 C20:0 0.93 2.19 0.69 2.50
9t C18:1 0.55 3.47 1.32 5.00 6c 9c 12c C18:3 1.53 3.64 0.68 2.25
11t C18:1 0.59 4.42 1.32 4.67 9c 12c 15c C18:3 0.89 4.77 2.14 7.14
6c C18:1 1.01 4.25 1.32 5.38 C20:4 0.64 4.52 1.18 3.55
9c C18:1 0.74 4.48 1.43 4.76 C22:6 1.82 2.62 2.50 9.09
 Y. Chen et al. / Food Control 44 (2014) 191e197 195

Table 2 progressive decrease in the content of both (cis) linoleic and (cis)
Repeatability and recovery data of FAME of sample analyzed by GC. linolenic acids was observed throughout the frying period
FAME RSD Recovery (%) FAME RSD Recovery (%) (Table 3). The content of (cis) linoleic acid was reduced by 10.99%
(%)a (%)a and 16.21% after frying for 8 h at 150 and 250
C, respectively. The
C12:0 2.20 107.04e112.82 9t C18:1 2.18 92.46e102.22 deterioration of (cis) linolenic acid was more pronounced, with its
C14:0 2.08 101.65e109.48 9c 12t C18:2 2.62 91.88e115.72 contents being decreased by 20.86% and 62.58%, respectively.
C16:0 2.20 94.44e104.58 9t 12c C18:2 2.54 94.00e113.55 Aladedunye and Przybylski (2009) reported decreases of 13.3% in
9c C16:1 2.48 91.09e91.25 9c 12c C18:2 2.06 89.84e104.98
linoleic acid and 47.1% in linolenic acid when fresh canola oil was
C17:0 2.52 105.30e110.39 C20:0 2.49 97.63e102.32
C18:0 2.51 97.84e103.23 9c 12c 15c C18:3 1.82 98.30e104.45 heated at 215
C for 7 days.
a
In the case of trans isomers, small changes were observed in
Mean relative standard deviation of six experiments in parallel with palm oil
sample.
their formation at lower frying temperatures (200
C, 8 h)
(Table 3). However, when comparing the relative percentage of
80%. These results indicated that the performance of the estab- TFAs during the frying sessions at a higher temperature of 250
C, a
lished method was satisfactory. clear increase in TFAs with 18 carbon atoms was observed (Fig. 2).
The total contents of trans C18:2 isomers in frying oils were
3.3. Fatty acid composition of fresh oils and non-fried chicken fillets significantly increased with an increase in the length of deep frying
time (P < 0.05). The final concentrations of total trans C18:2 in fried
The FA composition in fresh palm oil and non-fried chicken palm oil was 3.64 mg/g at 250
C, which was increased by 53.02%. A
fillets was shown in Supplementary Tables 1 and 2. The major fatty linear relationship between the amount of total trans-linoleic acids
acids in fresh palm oil were oleic acid, palmitic acid and linoleic and the frying time was also observed ( R2 ¼ 0.984). Whereas the
acid, whereas oleic acid and linoleic acid were the major compo- trans C18:1 isomer appeared only after 4 h frying at 250
C in palm
nents in non-fried chicken fillets. The main TFA isomers found in oil, the trans C18:3 isomers were found after 2 h frying at 250
C.
fresh palm oil before the start of frying were 9c 12t C18:2, 9t 12c Once produced, the amount of ctt/cct C18:3 was found to be very
C18:2 isomers, among which 9c 12t C18:2 constituted the largest stable afterward throughout the repeated frying process. In the
portion (0.92  0.07 mg/g), followed by 9t 12c C18:2 case of 9t 12c 15c C18:3, its accumulation increased slightly with the
(0.79  0.07 mg/g) (Supplementary Table 1). Interestingly, 9t C18:1 increasing frying cycles. The formation of trans isomers at elevated
was not detected. These results were quite different from the fatty temperatures indicated that a specific amount of energy was
acid profile of other vegetable oils such as soybean oil, rapeseed oil, required to transfer double bonds from cis to trans configuration,
sunflower oil and corn oil, which were reported to be rich in 9t and when the numbers of cis double bonds increased, the activa-
C18:1 with a high level of 0.03e0.08 g/100 g (Hou et al., 2012). The tion energy for isomerization decreased. This result was consistent
initial amounts of total trans fatty acids in our oil samples were with a previous study (Tsuzuki et al., 2008), which found a sub-
1.71 mg/g (0.18%) (Supplementary Table 1). stantial increase in trans 18:2 FAs and in trans 18:3 FAs in safflower
Fats extracted from the non-fried chicken fillets contained oil, when the oils were heated at 250e350
C.
147.83  10.56 mg/g of palmitic acid (C16:0), 25.49  0.88 mg/g of It has been suggested that the ratio of linoleic acid to palmitic
stearic acid (C18:0), 231.46  12.76 mg/g of oleic acid (C18:1) and acid (C18:2/C16:0) can be used for the evaluation of oil deteriora-
193.31  5.67 mg/g of linoleic acid (C18:2), and trans isomers were tion (Normand, Eskin, & Przybylski, 2001). A progressive decrease
not found in the fats. These results were different from a previous in this ratio with the frying time was observed during frying, and
study in which trans C18:1 of 0.13 mg/g was found in chicken breast this reduction was more apparent at higher frying temperatures.
fillets (Lee, Tweed, Kim, & Scollan, 2012). The value decreased about 6.3, 9.72 and 26.52% during frying at
150, 200 and 250
C, respectively (Table 3). This implies that frying
3.4. Changes of fatty acid profiles in palm oil during frying at higher temperatures could induce more PUFA deterioration. The
decrease in percentage in the ratio of linolenic acid to palmitic acid
After frying at three different temperatures of 150, 200 and (C18:3/C16:0) were more pronounced, which decreased about 6.67,
250
C, TFAs in frying oils were analyzed and compared. A 12.90 and 57.94% during frying at 150, 200 and 250
C, respectively

Table 3
Unsaturated fatty acid composition change during frying process of palm oil at 150, 200, 250
C for varied length of time.

Temperature Fatty acid (mg/g)f

(
C) and time
9c C16:1 cis C18:1 9c12t C18:2 9t12c C18:2 9c12c C18:2 ctt/cct C18:3h 9t12c 15c 11c C20:1 9c12c15c C18:2/ C18:3/
(h)
C18:3 C18:3 C16:0 C16:0

150 2 1.07  0.04 326.23  17.42 0.88  0.07 0.71  0.04 65.67  3.76 NDg ND 0.31  0.02 1.35  0.07 0.1450 0.0032
4 1.10  0.05 336.62  11.59 0.92  0.07 0.80  0.07 66.39  2.42 ND ND 0.32  0.04 1.43  0.08 0.1424 0.0031
6 1.13  0.02 345.66  10.47 0.92  0.00 0.78  0.03 67.83  2.85 ND ND 0.34  0.01 1.56  0.08 0.1407 0.0030
8 1.07  0.01 321.12  4.74 0.87  0.02 0.76  0.02 61.28  0.59 ND ND 0.30  0.01 1.29  0.04 0.1358 0.0029
200 2 1.09  0.02 323.64  3.06 0.90  0.03 0.76  0.05 65.51  0.28 ND ND 0.34  0.00 1.36  0.01 0.1501 0.0031
4 1.09  0.08 326.18  20.74 0.94  0.05 0.83  0.05 63.80  4.16 ND ND 0.37  0.04 1.36  0.10 0.1440 0.0031
6 1.16  0.09 340.66  22.27 0.97  0.03 0.81  0.00 65.42  4.42 ND ND 0.36  0.01 1.33  0.04 0.1408 0.0029
8 1.10  0.03 320.63  6.20 0.89  0.01 0.76  0.04 59.67  1.06 ND ND 0.38  0.01 1.19  0.01 0.1355 0.0027
250 2 1.19  0.02 331.55  2.77 1.19  0.02b 1.08  0.01b 70.75  7.18ab 0.31  0.00b NDa 0.42  0.02 1.78  0.72a 0.1561 0.0252
4 1.16  0.02 319.69  4.45 1.42  0.00c 1.29  0.01c 58.90  0.77a 0.31  0.02bc 0.16  0.00b 0.44  0.00 0.83  0.01ab 0.1320 0.0141
6 1.16  0.01 325.24  1.07 1.56  0.00d 1.45  0.01d 56.53  0.29a 0.27  0.00c 0.19  0.00c 0.46  0.01 0.70  0.00ab 0.1241 0.0124
8 1.22  0.04 353.65  2.38 1.8  0.00e 1.76  0.02e 57.69  0.20a 0.31  0.02bc 0.25  0.01d 0.48  0.00 0.61  0.00b 0.1147 0.0106
aee
Symbols bearing different letters in the same row are significantly different (P < 0.05).
f
Means of duplicate analyses  standard deviation.
g
ND: not detected.
h
Total content of 9c 12t 15t C18:3 and 9c 12c 15t C18:3.
196 Y. Chen et al. / Food Control 44 (2014) 191e197

0.02
0.01
0.03
0.02
trans C18:3





0.75
0.57
0.78
0.79
ND
ND
ND
ND
ND
0.09
0.03
0.07
0.04
0.04
0.01
0.04
0.05
0.12
trans C18:2










1.46
1.40
1.37
1.61
1.49
1.56
1.57
2.02
2.62
1.65  0.10
trans C18:1

ND
ND
ND
ND
ND
ND
ND
ND
0.04
0.02
0.12
0.09
0.08
0.03
0.06
0.03
0.16
Fig. 2. Contents of trans C18:1, trans C18:2, trans C18:3 in frying oil during the frying
sessions at 250
C.










C20:0

1.34
1.29
1.27
1.39
1.30
1.26
1.28
1.18
1.27
(Table 3). These results were consistent with previous observa-
tions by Aladedunye and Przybylski (2009). These authors also

2.93
3.98
2.10
3.16
2.38
4.24
4.94
4.55
4.32
9c12c C18:2
found a significant decrease in the ratio during frying, and the










decrease in this ratio was 1.2 times (for C18:2/C16:0) and 1.9 times

80.82
75.97
77.17
86.61
86.32
95.67
82.49
90.54
84.91
(for C18:3/C16:0) greater in oil heated at 215
C as compared to
185
C.

10.32
To investigate the relationship between TFAs formation and

7.28
8.28
8.29
9.22
3.98
7.92
9.34
8.23
lipid oxidative deterioration, the frying palm oils were also eval-










cis C18:1
uated for their thermal deterioration by acid value (AV) and

297.62
280.50
275.14
298.85
276.66
257.30
285.53
253.30
279.76
peroxide value (POV). As shown in Supplementary Fig. 1A, the AV

Unsaturated fatty acid composition change during frying process of chicken fillets at 150, 200 and 250
C for varied length of time.
of frying oils increased gradually and statistical differences were
found to depend on the frying temperature. This was supported by

1.05
1.77
2.76
1.87
1.11
1.98
2.03
1.73
2.02
a report that the hydrolysis reaction by water from frying mate-










rials would relate to a rapid deterioration of the frying oil (Hara,
C18:0

36.77
34.50
34.72
38.61
37.25
33.31
35.37
34.28
36.36
Ogawa, & Totani, 2006). In contrast, POV was not well correlated
with frying temperature (Supplementary Fig. 1B). The highest
value was found at 150
C, followed by a significant decrease at
0.04
0.02
0.01
0.03
0.06
0.05
0.02
0.02
0.01
200
C and then slightly increase at 250
C. This observation could









C17:0

be explained by the thermal decomposition of hydroperoxides to

0.70
0.68
0.65
0.71
0.67
0.65
0.67
0.62
0.70
carbonyl and aldehyde compounds during deep-fat frying at
higher temperatures, which results in a lower accumulation in the
14.00
19.26
19.22
10.10
16.97
21.42
12.00
18.42
19.33
oil at the higher frying temperatures. These findings suggested
that the frying at low temperatures (for example at 150
C) might









395.07
378.42
367.23
381.25
349.50
319.02
363.87
324.09
373.59
be preferable not only for less TFAs intakes but also for the con-
C16:0

sumption of less deteriorate oils.

0.00
0.01
0.01
0.02
0.02
0.01
0.02
0.01
0.03

3.5. Changes of fatty acid profiles in fried chicken fillets










C15:0

0.39
0.36
0.36
0.40
0.35
0.35
0.35
0.33
0.39

The amount of trans isomers in oil can affect the trans level in
fried product due to fast exchange of fats during frying. We
0.26
0.12
0.31
0.56
0.12
0.23
0.43
0.55
0.23

observed that the FA composition pattern of the fried chicken

fillets was similar to that of the frying oil rather than to that of the









C14:0

8.30
7.91
7.68
8.03
7.32
6.78
7.57
6.73
7.93

original raw chicken fillets (Fig. 1B, C, and D). For instance, C12:0
Fatty acid (mg/g)

and C15:0 did not exist in the raw chicken fillets but appeared
after frying (Table 4). And higher amounts of C14:0, C16:0, C17:0,
0.06
0.03
0.07
0.02
0.04
0.03
0.01
0.05
0.09

C18:0 and C20:0 in fried samples were found as compared to










C12:0

unfried samples. This observation suggested that several lipids in

1.69
1.60
1.55
1.75
1.60
1.50
1.55
1.43
1.70

the fried chicken fillets would be transferred from the frying oil.
The TFAs amount in the chicken fillets fried at 150, 200 and
250
C was monitored after the first, 24th and 48th frying process
Temperature (
C) and cycle

(Table 4 and Supplementary Fig. 2). When frying was conducted at

150
C and 200
C, the trans FAs formed in the chicken fillets
1
24
48
1
24
48
1
24
48

mainly consisted of C18:2 isomers. And the amount of trans C18:2

FAs kept relatively stable regardless of the prolong of frying time.
These results suggested that the number of frying operation did
not have a significant effect on the TFAs level in the fried chicken
Table 4

fillets at these two temperatures. In contrast, when frying at

150

200

250

250
C, most of the trans 18:3 FAs and trans 18:1 FAs appeared
 Y. Chen et al. / Food Control 44 (2014) 191e197
besides the trans 18:2 FAs. These results are similar to the obser-
vation in frying oils.
The results in Table 3 showed that the raw chicken fillets used in
the study contained a high level of cis 18:3 FAs, and its trans isomers
formed earlier than that in palm oils during frying. We speculated
that during frying part of the trans 18:3 FAs may be released from
the chicken fillets into the frying palm oils. The amount of the
frying oil absorbed in fried chicken fillets and the TFAs concentra-
tion of the fried chicken fillets itself would contribute to TFA intakes
from such fried food.
4. Conclusions
An improved GC method for the rapid separation of positional
and geometrical isomers of fatty acid methyl esters has been
developed. The resolution of cis and trans isomers was better in
comparison with previous reports, especially for elaidic and oleic
acid. Its application to the analysis of fatty acids in frying palm oil and
chicken fillets has been shown to be robust and reliable. The results
showed that the rate of isomerization and thermal degradation of
PUFA was drastically higher at the elevated temperatures tested,
forming larger amounts of components with potential harmful
health effects. This suggested that an increase in frying temperature
above 200
C can cause intensive isomerization of PUFA. Besides, by
respective monitoring of TFAs accumulation in frying materials and
frying oil, we found that trans 18:3 FAs could be released from the
chicken fillets to the oil, and similarly the elaidic acid could be
transferred from frying oils to chicken fillets. These results indicated
that the trans level in frying materials and the frying oils can affect
each other due to fast exchange of fats during frying. In addition, this
developed method may be used for the labeling and quality control
of food which contains specific fatty acid profiles.
Acknowledgments
The financial support from the National Basic Research Program
of China (973 program) (No: 2012CB720805), the National Key
Technology R&D Program (No: 2012BAK17B02), and the Program of
Jiangxi Provincial Department of Science and Technology (No:
20122BAB214002) is gratefully acknowledged.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.foodcont.2014.04.010.
References
Aladedunye, F. A., & Przybylski, R. (2009). Degradation and nutritional quality
changes of oil during frying. Journal of the American Oil Chemists’ Society, 86,
149e156.
Albuquerque, T. G., Costa, H. S., Castilho, M. C., & Sanches, S. A. (2011). Trends in the
analytical methods for the determination of trans fatty acids content in foods.
Trends in Food Science and Technology, 22, 543e560.
Antolín, E. M., Delange, D. M., & Canavaciolo, V. G. (2008). Evaluation of five
methods for derivatization and GC determination of a mixture of very long
chain fatty acids. Journal of Pharmaceutical and Biomedical Analysis, 46, 194e199.
AOCS. (1997). Preparation of methyl ester of fatty acids. Champaign US: AOCS Press.
AOCS. (2005). Determination of cis-, trans-, saturated, monounsaturated and poly-
unsaturated fatty acids in vegetable or non-ruminant animal oils and fats by
capillary GLC. Champaign US: AOCS Press.
Aro, A., Kosmeijer-Schuil, T., Bovenkamp, P., Hulshof, P., Zock, P., & Katan, M. B.
(1998). Analysis of C18:1 cis and trans fatty acid isomers by the combination of
gaseliquid chromatography of 4,4-dimethyloxazoline derivatives and methyl
esters. Journal of the American Oil Chemists’ Society, 75, 977e985.
Berdeaux, O., Dutta, P. C., Dobarganes, M. C., & Sébédio, J. L. (2009). Analytical
methods for quantification of modified fatty acids and sterols formed as a result
of processing. Food Analytical Methods, 2, 30e40.
197
Cajka, T., Hajslova, J., Lacina, O., Mostovska, K., & Lehotay, S. J. (2008). Rapid analysis
of multiple pesticide residues in fruit-based baby food using programmed
temperature vaporizer injection-low pressure gas chromatography-high reso-
lution time-of-flight mass spectrometry. Journal of Chromatography A, 1186,
281e294.
CFIA (Canadian Food Inspection Agency). (2008). Labelling of trans fatty acids. URL
http://www.inspection.gc.ca/english/fssa/labeti/inform/transe.shtml Accessed
16.03.13.
Code of China. GB/T 5009.37-2003 method for analysis of hygienic standard of edible
oils (English). URL http://www.codeofchina.com/gb/medical/24993.html
Accessed 16.03.13.
FDA (U.S. Food and Drug Administration). (2003). Guidance for industry: Trans fatty
acids in nutrition labeling, nutrient content claims, health claims; small entity
compliance guide. URL http://www.fda.gov/Food/GuidanceRegulation/
GuidanceDocumentsReguatoryInformation/LabelingNutrition/ucm053479.htm
Accessed 16.03.13.
Folch, J., Lees, M., & Sloane-Stanley, G. H. (1957). A simple method for the isolation
and purification of total lipids from animal tissues. Journal of Biological Chem-
istry, 226, 497e509.
Golay, P.-A., Dionisi, F., Hug, B., Giuffrida, F., & Destaillats, F. (2006). Direct quanti-
fication of fatty acids in dairy powders with special emphasis on trans fatty acid
content. Food Chemistry, 101, 1115e1120.
Hara, E., Ogawa, Y., & Totani, Y. (2006). Evaluation of heat-deteriorated oils (part I):
TLC-FID method for determining polar compounds content. Journal of Oleo
Science, 55, 167e172.
Hou, J. C., Wang, F., Wang, Y. T., Xu, J., & Zhang, C. W. (2012). Assessment of trans
fatty acids in edible oils in China. Food Control, 25, 211e215.
Indarti, E., Majid, M. I. A., Hashim, R., & Chong, A. (2005). Direct FAME synthesis for
rapid total lipid analysis from fish oil and codliver oil. Journal of Food Compo-
sition and Analysis, 18, 161e170.
Kim, S. H., Chunawala, L., Linde, R., & Reaven, G. M. (2006). Comparison of the 1997
and 2003 American Diabetes Association classification of impaired fasting
glucose: impact on prevalence of impaired fasting glucose, coronary heart
disease risk factors, and coronary heart disease in a community-based medical
practice. The Journal of the American College of Cardiology, 48, 293e297.
Kohlmeier, L. (1997). Trans-fatty acids and breast cancer. Food and Chemical Toxi-
cology, 35, 1228.
Koning, S. D., Meer, B. V. D., Alkema, G., Janssen, H. G., & Brinkman, U. A. T. (2001).
Automated determination of fatty acid methylester and cis/trans methyl ester
composition of fats and oils. Journal of Chromatography, 9, 391e397.
Lee, M. R. F., Tweed, J. K. S., Kim, E. J., & Scollan, N. D. (2012). Beef, chicken and lamb
fatty acid analysis a simplified direct bimethylation procedure using freeze-
dried material. Meat Science, 92, 863e866.
Liu, W. H., Inbaraj, B. S., & Chen, B. H. (2007). Analysis and formation of trans fatty
acids in hydrogenated soybean oil during heating. Food Chemistry, 104, 1740e
1749.
Manzano, P., Arnáiz, E., Diego, J. C., Toribio, L., Nozal, M. J., Bernal, J. L., et al. (2012).
Improved separation of fame isomers using comprehensive two-dimensional
gas chromatography. Application to broccoli samples. Food Analytical Methods,
5, 920e927.
Manzano, P., Diego, J. C., Nozal, M. J., Bernal, J. L., & Bernal, J. (2012). Gas
chromatography-mass spectrometry approach to study fatty acid profiles in
fried potato crisps. Journal of Food Composition and Analysis, 28, 31e39.
Mozaffarian, D., Katan, M. B., Ascherio, A., Stampfer, M. J., & Willett, W. C. (2006).
trans Fatty acids and cardiovascular disease. New England Journal of Medicine,
354, 1601e1613.
Normand, L., Eskin, N. A. M., & Przybylski, R. (2001). Effect of tocopherols on the
frying stability of regular and modified canola oils. Journal of the American Oil
Chemists’ Society, 78, 369e373.
Oomen, C. M., Ocke, M. C., Feskens, E. J. M., Erp-Baart, M. J., Kok, F. J., & Kromhout, D.
(2001). Association between trans fatty acid intake and 10-year risk of coronary
heart disease in the Zutphen Elderly Study: a prospective population-based
study. Lancet, 357, 746e751.
Romero, A., Cuesta, C., & Sánchez-Muniz, F. J. (1998). Effect of oil replenishment
during deep-fat frying of frozen foods in sunflower oil and high-oleic acid
sunflower oil. Journal of the American Oil Chemists’ Society, 75, 161e167.
Schröder, M., & Vetter, W. (2013). Detection of 430 fatty acid methyl esters from a
transesterified butter sample. Journal of the American Oil Chemists’ Society, 90,
771e790.
Sébédio, J. L., & Ratnayake, W. M. N. (2008). In A. J. Dijkstra, R. J. Hamilton, &
W. Hamm (Eds.), Trans fatty acids (pp. 102e131). Oxford: Blackwell.
Tranchida, P. Q., Franchina, F. A., Dugo, P., & Mondello, L. (2012). A flow-modulated
comprehensive gas chromatographyemass spectrometry method for the
analysis of fatty acid profiles in marine and biological samples. Journal of
Chromatography A, 1255, 171e176.
Tsuzuki, W., Nagata, R., Yunoki, R., Nakajima, M., & Nagata, T. (2008). cis/trans-
Isomerisation of triolein, trilinolein and trilinolenin induced by heat treatment.
Food Chemistry, 108, 75e80.
Yang, M. Y., Yang, Y., Nie, S. P., Xie, M. Y., & Chen, F. (2012). Analysis and formation of
trans fatty acids in corn oil during the heating process. Journal of the American
Oil Chemists’ Society, 89, 859e867.
Zhou, W. B., Tan, T. W., Neo, F. L., & Lo, H. L. (2009). Analysis of trans fatty acids in
deep frying oils by three different approaches. Food Chemistry, 116, 535e541.