Application of ethyl esters and d3-methyl esters as internal standards for the gas chromatographic quantification of transesterified fatty acid methyl esters in food

Article abstract of DOI:10.1021/jf053022j

Ethyl esters (FAEE) and trideuterium-labeled methyl esters (d ₃-FAME) of fatty acids were prepared and investigated regarding their suitability as internal standards (IS) for the determination of fatty acids as methyl esters (FAME). On CP-Sil 88, ethyl esters of odd-numbered fatty acids eluted ～0.5 min after the respective FAME, and only coelutions with minor FAME were observed. Depending on the problem, one or even many FAEE can be added as IS for the quantification of FAME by both GC-FID and GC-MS. By contrast, d₃-FAME coeluted with FAME on the polar GC column, and the use of the former as IS requires application of GC-MS. In the SIM mode, m/z 77 and 90 are suggested for d₃-methyl esters of saturated fatty acids, whereas m/z 88 and 101 are recommended for ethyl esters of saturated fatty acids. These m/z values give either no or very low response for FAME and can thus be used for the analysis of FAME in food by GC-MS in the SIM mode. Fatty acids in sunflower oil and mozzarella cheese were quantified using five saturated FAEE as IS. Gravimetric studies showed that the transesterification procedure could be carried out without of loss of fatty acids. GC-EI/MS full scan analysis was suitable for the quantitative determination of all unsaturated fatty acids in both food samples, whereas GC-EI/MS in the SIM mode was particularly valuable for quantifying minor fatty acids. The novel GC-EI/MS/SIM method using fatty acid ethyl esters as internal standards can be used to quantify individual fatty acids only, that is, without determination of all fatty acids (the common 100% method), although this is present. This was demonstrated by the exclusive quantification of selected fatty acids including methyl-branched fatty acids, erucic acid (18:1n-9trans), and polyunsaturated fatty acids in cod liver oil and goat's milk fat.

Full text of DOI:10.1021/jf053022j

J. Agric. Food Chem. 2006, 54, 3209−3214 3209
Application of Ethyl Esters and d -Methyl Esters as Internal
3
Standards for the Gas Chromatographic Quantification of
Transesterified Fatty Acid Methyl Esters in Food
SASKIA THURNHOFER AND WALTER VETTER*
University of Hohenheim, Institute of Food Chemistry (170b), Garbenstrasse 28,
D-70599 Stuttgart, Germany
3
Ethyl esters (FAEE) and trideuterium-labeled methyl esters (d -FAME) of fatty acids were prepared
and investigated regarding their suitability as internal standards (IS) for the determination of fatty
acids as methyl esters (FAME). On CP-Sil 88, ethyl esters of odd-numbered fatty acids eluted ∼0.5
min after the respective FAME, and only coelutions with minor FAME were observed. Depending on
the problem, one or even many FAEE can be added as IS for the quantification of FAME by both
GC-FID and GC-MS. By contrast, d
use of the former as IS requires application of GC-MS. In the SIM mode, m/z 77 and 90 are suggested
for d -methyl esters of saturated fatty acids, whereas m/z 88 and 101 are recommended for ethyl
3
-FAME coeluted with FAME on the polar GC column, and the
3
esters of saturated fatty acids. These m/z values give either no or very low response for FAME and
can thus be used for the analysis of FAME in food by GC-MS in the SIM mode. Fatty acids in sunflower
oil and mozzarella cheese were quantified using five saturated FAEE as IS. Gravimetric studies
showed that the transesterification procedure could be carried out without of loss of fatty acids. GC-
EI/MS full scan analysis was suitable for the quantitative determination of all unsaturated fatty acids
in both food samples, whereas GC-EI/MS in the SIM mode was particularly valuable for quantifying
minor fatty acids. The novel GC-EI/MS/SIM method using fatty acid ethyl esters as internal standards
can be used to quantify individual fatty acids only, that is, without determination of all fatty acids (the
common 100% method), although this is present. This was demonstrated by the exclusive
quantification of selected fatty acids including methyl-branched fatty acids, erucic acid (18:1n-9trans),
and polyunsaturated fatty acids in cod liver oil and goat’s milk fat.
KEYWORDS: Fatty acid methyl esters; fatty acid ethyl esters; sunflower oil; mozzarella cheese; erucic
acid; cod liver oil; goat’s milk fat; GC-EI/MS-SIM
INTRODUCTION
examples for IS are methyl esters of 17:0 (17:0-ME) or other
fatty acids with odd carbon numbers such as 7:0-ME, 9:0-ME,
Determination of fatty acids is a routine method in food
chemistry and food control. Typically, the lipid phase of food,
which mostly consists of triacylglycerides, is extracted with
organic solvents. The fatty acid profile is then determined after
transesterification into fatty acid methyl esters (FAME) followed
by GC-FID determination (1). GC-FID is the method of choice
for the determination of the relative contributions of individual
fatty acids to the fatty acid pattern because of relative constant
response factors irrespective of the chain length and number of
double bonds. However, the sensitivity and selectivity of GC-
FID are relatively low, and some applications require the use
of GC-MS for confirmatory measurements. For quantitative
determination of individual fatty acids in food samples, methyl
esters of fatty acids not found at remarkable concentrations in
food have been applied as internal standards (IS) (2). Typical
1
3:0-ME, 19:0-ME, 21:0-ME, and 23:0-ME as well as 26:0-
ME and 21:1-ME (2-7). However, even these rare fatty acids
are often present in food samples. Furthermore, coelutions with
other fatty acids can lead to inaccurate results. On the other
hand, using several odd-chain fatty acids as IS for the deter-
mination of even-chain fatty acids proved to be superior to the
use of a single IS (2). There are two strategies for the use of
IS: (i) spiking a certain fatty acid into the lipid matrix before
extraction (5, 8) or (ii) addition of one or more esterified fatty
acids after conversions of lipid fatty acids into FAME (3, 9,
1
0). In the latter case, the IS must not necessarily be a FAME.
In the present study we have explored under which conditions
fatty acid ethyl esters (FAEE) or d3-methyl esters of fatty acids
(d3-FAME) can be useful IS for the determination of FAME.
Both GC-FID and GC-MS in the full scan mode require that
the IS do not elute at the retention time of relevant FAME in
food or other samples. Recently, we have presented a GC-EI/
*
Corresponding author (telephone 0049 711 459 4016; fax 0049 711
4
594377; e-mail w-vetter@uni-hohenheim.de).
1
0.1021/jf053022j CCC: $33.50
© 2006 American Chemical Society
Published on Web 04/05/2006

3210 J. Agric. Food Chem., Vol. 54, No. 9, 2006
Thurnhofer and Vetter
MS method in the selected ion monitoring (SIM) mode that
enables the determination of food-relevant fatty acids as methyl
esters (11). This novel GC-EI/MS/SIM method is based on four
low mass fragment ions (m/z 74, 79, 81, and 87) and combines
high sensitivity with high selectivity (11). We thus tested
whether the respective SIM masses enable the use of internal
standards even if the IS would coelute with the native FAME
derivatives as derived from food samples. The resulting method
was used for the quantification of fatty acids in food samples.
Between 21.6 and 27.3 µg/mL of the five FAEE were 1:1 (v/v)
combined with a 1:10 diluted 37c-FAME mix solution to determine
relative response factors of FAME and FAEE (see Results and
Discussion). A similar mix was prepared from methyl-branched fatty
acids for the determination of their responses.
For the quantification of fatty acids, 50-100 µL of transesterified
oil or fat and 20 µL of the IS mix were added in a calibrated vial and
filled to 500 µL with n-hexane. In the GC-EI/MS scan mode, individual
areas of the selected ion traces of FAME were determined and corrected
by the known concentration and response factors of the five IS (see
above). FAME without respective FAEE as IS were quantified using
the response of 16:0-EE. In the GC-EI/MS SIM mode, methyl esters
of saturated fatty acids were determined with m/z 87, those of
monoenoic fatty acids with m/z 74, and those of PUFA with the sum
of m/z 79 and 81 (11). The FAEE used as IS were determined with
m/z 101. Ratios of FAME (m/z 88) to FAEE (m/z 101), n ) 4, were
0.64 ( 0.03 for 12:0, 0.67 ( 0.03 for 14:0, 0.70 ( 0.03 for 16:0, 0.72
( 0.03 for 17:0, and 0.75 ( 0.03 for 18:0. All further fatty acids in
the 37c-FAME mix were determined using the response of the
quantification ion relative to 16:0-ME, whereas other fatty acids were
determined with the average response factor of the class of fatty acid
(saturated, monoenoic, or polyenoic fatty acid).
MATERIALS AND METHODS
Chemicals and Samples. Cyclohexane (purest, VWR, Darmstadt,
Germany) and ethyl acetate (purest, Acros Organics, Geel, Belgium)
were combined (1:1, v/v) and distilled to obtain the azeotropic mixture
(
54:46, v/v). n-Hexane (HPLC gradient grade) and methanol (HPLC
gradient grade) were from Fluka (Taufkirchen, Germany). Isooctane
analytical reagent grade) was from Fisher Scientific (Ulm, Germany),
and isolute-HM-N was from Separtis (Grenzlach-Wyhlen, Germany).
Boron trifluoride-methanol complex solution (13-15% BF in metha-
nol) was from Riedel-de-Ha e¨ n (Taufkirchen, Germany). BF ethyl
(∼10%, ∼1.3 M, purris) were
(
3
3
etherate (purum, dist.) and ethanolic BF
from Fluka.
3
RESULTS AND DISCUSSION
A Supelco 37 component FAME mix (37c-FAME mix, Sigma-
Aldrich, Taufkirchen, Germany) as well as additional standards of free
fatty acids and FAME (Larodan, Malm o¨ , Sweden) were used. Addition
retention times were derived from fatty acids in the following food
samples: sunflower oil (Heess, Stuttgart, Germany), goat’s milk
Gas Chromatographic Feature of Methyl, Ethyl, and d3-
Methyl Esters of Fatty Acids. The alternative esters could be
successfully prepared as was found for FAME (see also below).
Moreover, the proposed reaction scheme for derivatization (see
Materials and Methods) resulted in virtually identical GC peak
patterns for all three classes of esters.
Use of GC-MS clarified that none of the investigated food
samples (goat’s milk, sunflower oil, mozzarella cheese, fish oil,
and suet) contained FAEE or d3-FAME after conversion of the
lipid fraction into FAME; likewise, the alternatively trans-
esterified samples were free of FAME.
(
(
Andechser Creamery, Andechs, Germany), buffalo mozzarella cheese
Padania Alimenti, Casalmaggiore, Italy), and cod liver oil (R u¨ gen
Fisch, Sassnitz, Germany).
Sample Preparation. Food samples except oils were lyophilized
prior to extraction. Lipids were gained by accelerated solvent extraction
(ASE, Dionex, Idstein, Germany) with ethyl acetate/cyclohexane (54:
46, v/v, see above) as the solvent (11, 12). After removal of the solvent,
the lipid phase (namely, the fatty acid glycerides) was transesterified.
For this purpose 2-10 mg of fat or oil and 0.5 mL of alcoholic KOH
As anticipated, FAEE eluted after the respective FAME from
the GC column. The difference in the (netto) retention times
(0.5 M) were heated for 5 min at 80 °C. After cooling, 1 mL of BF
3
solution (see below) was added and heated for an additional 5 min at
0 °C. Two milliliters of saturated sodium chloride solution and 2 mL
(
∆t′R) was ∼0.5 min for early eluting fatty acids (Table 1).
8
Owing to the longer retention times, the separation factors of
the corresponding alkyl esters subsequently decreased with
increasing chain length (Table 1). FAEE of saturated fatty acids
abundant in food (16:0 and 18:0) eluted about one-third between
the respective FAME and the FAME with one more carbon
(Figure 1). In this time window, only low abundant monoenoic
FAME isomers were observed. However, no coelution of any
saturated FAEE with relevant FAME was observed (Table 1).
In the case of particular interests in the determination of certain
FAME, suitable FAEE can be selected accordingly.
of n-hexane were added to the cooled solution (ice bath). The esters
were extracted and analyzed by GC-MS in full scan and SIM mode
(
11). For preparation of FAME we used 0.5 mL of methanolic KOH
and 1 mL of methanolic BF , for d -FAME we used KOH in
-methanol and BF ethyl etherate, and for FAEE we used ethanolic
KOH and ethanolic BF . Esters from standards of free fatty acids were
prepared by treatment with BF (in diethyl ether) and either ethanol or
the NMR solvent d -methanol (CD OD).
3
3
d
4
3
3
3
4
3
Gas Chromatography Coupled to Electron Ionization Mass
Spectrometry (GC-EI/MS). Analyses were performed with a Hewlett-
Packard 5890 series II gas chromatograph interfaced to a 5971A mass
selective detector. One microliter of sample dissolved in n-hexane was
injected with a 7673A autosampler (splitless mode, split opened after
Deuterium-labeled FAME were prepared using the readily
available NMR solvent d4-methanol. Three of the heavy
hydrogen isotopes are found on the resulting esters, whereas
the deuterium originally attached to the oxygen was lost due to
the reaction scheme. Smith and Schewe used a similar technique
for the preparation of five individual d3-FAME (13). FAME
and d3-FAME were only partly resolved, with the d3-FAME
2
min). The injector and transfer line temperatures were kept at 250
and 280 °C. The temperature of the ion source was 165 °C. Helium
purity 5.0) was used as the carrier gas at a constant flow rate of 1
fused-silica capillary
(
mL/min. A 50 m × 0.25 mm i.d., 0.20 µm d
f
column coated with CP-Sil 88 (Chrompack, Middelburg, The Nether-
lands) was installed in the GC oven. The GC oven program was the
following: after 5 min at 45 °C, the oven was heated at 7 °C/min to
eluting slightly prior to the native FAME (∆t ∼ 2 s, Table 1)
R
due to the higher volatility of the labeled compounds (14). GC
1
80 °C, at 3 °C/min to 200 °C (hold time 1 min), and finally at 3
separation of native and deuterium-labeled compounds is mainly
due to the slightly higher vapor pressures of the latter, and this
effect is most pronounced on nonpolar stationary GC phases
°
C/min to 220 °C (hold time of 10 min). The total run time was 51.62
min. In the full scan mode m/z 50-450 were recorded after a solvent
delay of 8 min. In the SIM mode, the six to eight fragment ions were
determined including m/z 74 and 87 for FAME, m/z 88 and 101 for
(15). Thus, peak resolution of unlabeled and d3-labeled FAME
on the polar CP-Sil 88 column was very unlikely and, indeed,
was not observed at any condition applied. When we changed
to a nonpolar CP-Sil 8 (equivalent to DB-5) column, a partial
resolution of unlabeled and labeled was obtained (see Supporting
Information). Note that the full resolution of d3-labeled and
3
FAEE, and m/z 77 and 90 for d -FAME as well as m/z 81 and 79 for
all esters of PUFA.
Quantification of Fatty Acids. Free fatty acids of 12:0, 14:0, 16:0,
7:0, and 18:0 were accurately weighed (7-15 mg) in triplicates and
1
ethylated as shown above. The yields of the FAEE were 96.6 ( 2.0.

Quantification of Fatty Acids as Methyl Esters in Food
J. Agric. Food Chem., Vol. 54, No. 9, 2006 3211
Table 1. Gas Chromatographic Retention Times (in Rising Order) of
3
the Methyl, d -Methyl, and Ethyl Esters of 29 Fatty Acids Determined
on CP-Sil 88a
fatty acid
FAME (t
′
r
)b
3
d -FAME (t
′
r
)
FAEE (t
′
r
)
Rc
4
6
8
1
1
1
1
:0
:0
:0
0:0
1:0
2:0
3:0
6.29
10.75
14.03
17.01
18.47
19.63
20.89
21.44
21.98
22.61
22.84
22.91
23.20
23.83
24.47
25.15
25.30
25.41
25.79
27.17
27.78
27.97
28.07
28.98
29.28
30.90
32.54
36.53
41.29
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.01
0.00
0.00
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.01
0.01
0.00
0.00
6.25
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.01
0.00
0.00
0.02
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.00
0.01
0.01
0.00
0.01
0.01
0.00
0.03
0.01
0.01
0.00
0.00
0.01
0.01
0.01
0.00
0.00
7.05
11.24
14.67
17.53
18.92
20.08
21.29
21.85
22.41
23.04
23.28
23.32
23.64
24.29
24.96
25.60
25.72
25.86
26.27
27.70
28.28
28.44
28.56
29.41
29.69
31.28
33.01
36.77
41.48
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.01
0.01
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.01
0.01
0.01
0.01
0.00
0.01
0.00
0.00
0.00
0.01
0.00
0.01
0.00
0.01
0.00
0.01
0.01
0.00
0.01
0.00
1.080
1.036
1.037
1.026
1.021
1.020
1.017
1.017
1.017
1.017
1.017
1.016
1.017
1.017
1.017
1.016
1.015
1.016
1.016
1.017
1.016
1.015
1.016
1.013
1.013
1.013
1.014
1.006
1.004
10.73
14.01
17.00
18.46
19.61
20.87
21.42
21.96
22.59
22.83
22.90
23.18
23.82
24.44
25.13
25.28
25.39
25.77
27.15
27.78
27.93
28.04
28.95
29.24
30.85
32.54
36.48
41.22
d
i14:0
4:0
i15:0
1
d
a15:0
1
1
4:1n-5
5:0
i16:0
6:0
i17:0
1
16:1n-7
a17:0
Figure 2. Mechanism of the formation of (a) the McLafferty ion and (b)
m/z 87 as well as the respective masses for the three ester derivatives
1
1
1
1
1
1
1
1
2
2
2
7:0
8:0
8:1n-9trans
8:1n-9
8:1n-7
9:0
the McLafferty ion at m/z 74 and the fragment ion at m/z 87,
also dominated in the mass spectra of 17:0-EE and 17:0-d3ME,
but they were shifted ∆14u or ∆3u toward higher mass. The
proposed mechanism (Figure 2) illustrates that both fragment
ions contain the ester group, which explains the different masses
observed. For instance, the McLafferty ion of FAME (m/z 74)
is m/z 77 for d3-FAME and m/z 88 for FAEE (Figure 2a).
Likewise, m/z 90 (d3-FAME) and m/z 101 (FAEE) correspond
to m/z 87 in FAME (Figure 2b). The same mass shifts were
also observed for the respective molecular ions. Likewise, the
fragment ions at m/z 241 and 143 in the GC-EI/MS of 17:0-
ME were shifted in the same way, but m/z 253 was present in
the mass spectra of all alkyl esters of 17:0 (see Supporting
Information). These examples illustrate that the diverse esters
may be used for a simple classification of fragment ions in the
mass spectra of FAME. Those fragment ions that still bear the
ester group will have different m/z values in the case of FAME,
d3-FAME, and FAEE, whereas fragment ions formed after
elimination of the ester group will have identical masses. Thus,
8:2n−6
8:3n-3
0:2n-6
0:5n-3
2:6n-3
a
Retention times were established from standard compounds as well as fatty
b
acids in food samples (suet, cod liver oil, sunflower oil, and goat’s milk fat). Netto
retention time (dead time
standard deviation (n
a indicate iso-fatty acids (fatty acids with a methyl branch on the second carbon
counted from the tail) and anteiso fatty acids (fatty acids with a methyl branch on
the third carbon counted from the tail).
)
3.17 min, based on n-hexane); mean value and
d
)
3). c Separation factor of FAEE to FAME. Letters i and
-
the fragment ion at m/z 253 ([284 u - 31 u] ) in 17:0-ME must
arise from an R-cleavage next to the carbonyl group ([M -
+
OR] ), which corresponds to [287 u - 34 u] and [298 u - 45
u] in the cases of d3-FAME and FAEE. Owing to the different
molecular masses, this fragment ion will vary with the carbon
+
chain length. However, [M - OR] along with m/z 55
+
(
[C4H7] ), 97, and 111 were the only important fragment ions
Figure 1. GC-MS/SIM determination (CP-Sil 88 column) of the fatty acid
pattern of buffalo mozzarella cheese derived after transesterification into
methyl esters with five ethyl esters added as internal standards. Arrows
identify added FAEE of 12:0, 14:0, 16:0, 17:0, and 18:0. (Inset) Enlarged
part of the late eluting fatty acids
with identical m/z values in the GC-EI/MS spectra of the three
classes of alkyl esters.
The situation was similar for monoenoic fatty acids but
completely different for PUFA. Most fragment ions from esters
of PUFAs are formed after elimination of the (uncharged)
headgroup, so that the respective ions are found in the mass
spectra of all three classes of esters of PUFAs. Thus, the mass
spectra of the three alkyl esters of EPA (20:5n-3) and other
PUFA looked virtually identical in the mass range below m/z
150 (Figure 3). In other words, none of these fragment ions
contained the ester group. This is in agreement with the
observation that the major ions in saturated and monoenoic fatty
acids (m/z 74 and 87 in the case of FAME) do not play any
role in the mass spectra of PUFA (11). Because the molecular
ion was barely detectable in the case of the simple esters of
PUFA, GC-MS was not well-suited for the identification of
different simple esters of PUFAs.
native FAME would have required about five hydrogens
substituted with deuterium. Whereas d3-FAME have no advan-
tage over native FAME when GC-FID is used for the determi-
nation, they might be suitable for GC-MS determination in the
case of different mass fragment ions.
Mass Spectrometric Feature of Methyl, Ethyl, and d3-
Methyl Esters of Fatty Acids. The pattern, that is, the relative
intensity of key fragment ions of FAME, d3-FAME, and FAEE,
was virtually identical for all saturated fatty acids (see Sup-
porting Information for an example). The most abundant
fragment ions in the mass spectra of saturated FAME, that is,

3212 J. Agric. Food Chem., Vol. 54, No. 9, 2006
Thurnhofer and Vetter
Table 2. Amounts of Fatty Acids in Sunflower Oil (n
) 3) and Mozzarella Cheese (n ) 3)
sunflower oil (n
SIM amount
)
3)
buffalo mozzarella (n
SIM amount
(g/100 g)
) 3)
responsea
scan amount
(g/100 g)
scan amount
(g/100 g)
FAME
(n
)
4)
(g/100 g)
4
6
8
1
1
1
:0
:0
:0
0:0
1:0
2:0
0.94
0.94
0.94
0.94
0.89
0.98
0.75
3.43
0.75
0.84
0.75
0.92
0.75
0.75
3.43
0.82
0.75
0.70
3.80
0.75
3.80
3.80
0.75
1.09
3.80
0.75
0.90
3.84
4.39
4.39
4.39
4.39
0.92
0.06
0.94
0.05
4.04
0.86
0.92
0.05
0.91
0.97
0.97
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.08
0.08
0.08
0.08
0.00
0.01
0.02
0.05
0.02
0.01
0.02
0.00
0.02
0.02
0.05
0.01
0.02
0.04
0.08
0.02
0.08
0.08
0.02
0.01
0.09
0.02
0.00
0.04
0.07
0.07
0.07
0.07
0.02
0.00
0.02
0.00
0.04
0.00
0.02
0.00
0.00
0.00
0.00
nd
nd
nd
±
nd
nd
nd
nd
nd
nd
0.21
0.33
0.26
0.71
0.03
2.95
0.05
0.07
0.03
0.15
0.22
9.43
0.40
0.51
1.01
1.56
0.49
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.02
0.02
0.03
0.06
0.00
0.22
0.00
0.01
0.00
0.01
0.02
0.43
0.06
0.04
0.03
0.12
0.05
0.26
0.30
0.24
0.69
±
±
±
±
nd
±
nd
nd
nd
±
±
±
±
±
±
±
±
±
±
±
0.04
0.03
0.05
0.01
0.003
0.0027c
d
0.11
±
nd
nd
nd
nd
nd
±
nd
nd
nd
±
nd
±
nd
nd
±
±
nd
±
nd
nd
±
±
±
±
0.01
0.09
0.10
±
0.01
0.01
2.65
0.03
e
i13:0
3:1
nd
nd
nd
nd
nd
±
nd
nd
nd
nd
nd
±
nd
nd
nd
±
nd
nd
nd
nd
±
1
e
a13:0
3:0
i14:0
1
0.12
0.22
9.88
0.51
0.65
0.96
1.58
0.52
24.9
0.11
0.38
0.41
2.10
0.75
0.94
0.38
0.03
12.1
0.00
0.00
0.19
0.01
0.02
0.01
0.02
0.00
0.12
0.02
0.01
0.04
0.03
0.02
0.08
0.01
0.00
0.24
d
14:0
0.14
0.00
i15:0
a15:0
1
1
4:1n-5
5:0
0.02
8.21
0.00
0.07
i16:0
d
f
1
1
6:0
6:1
8.29
0.08
0.05
0.01
23.6
±
0.31
0.24
0.34
0.54
1.99
0.52
0.99
0.39
0.05
±
±
±
±
±
±
±
±
0.08
0.03
0.13
0.09
0.04
0.13
0.07
0.02
i17:0
1
1
6:1n-9
6:1n-7
0.06
0.13
0.00
0.00
±
±
±
±
±
±
±
a17:0
d
17:0
0.05
0.00
17:1n-7
i18:0
d
1
1
1
1
1
1
1
1
2
1
2
2
2
2
2
2
2
8:0
8:1n-9trans
8:1n-9
8:1n-7
8:1
5.69
0.32
28.7
1.04
0.05
0.02
0.25
0.09
5.13
0.02
11.7
1.04
18.2
±
0.65
±
nd
±
±
nd
nd
nd
±
0.14
21.3
±
0.38g
26.7
0.72
0.16
0.02
± 0.69
0.81
0.58
0.68
0.11
1.82
0.35
0.36
0.12
0.08
0.13
0.07
0.08
0.08
0.03
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.10
0.17
0.09
0.08
0.03
0.05
0.05
0.04
0.03
0.02
0.01
0.04
0.02
0.01
0.62
0.45
0.47
0.12
2.21
0.33
0.44
0.11
0.08
0.13
0.09
0.06
0.07
0.05
±
0.02
0.02
0.03
0.06
0.14
0.03
0.03
0.08
0.01
0.01
0.01
0.00
0.01
0.01
f
nd
nd
nd
±
±
±
±
±
±
±
±
±
±
±
±
±
±
f
8:1
9:0
8:2n-6
0:0
8:3n-3
0:1n-9
1:0
2:0
0:3n-6
3:0
53.9
0.40
0.24
0.01
57.3
0.31
0.90
0.01
±
±
nd
nd
±
±
nd
±
nd
nd
nd
±
nd
nd
±
0.01
0.93
0.00
0.01
0.78
0.23
0.00
0.01
0.03
0.28
0.00
0.00
4:0
6:0
±
nd
nd
sum
99.8
99.7
83.2
87.1
a
b
Response factor relative to 16:0-EE; multiplication factor of the area of the quantification ion compared to 16:0-EE equal amounts. Not detected (detection limit,
c
d
<
0.003 g/100 g based on S/N < 10). Mean value and standard deviation. Calculated with the response of the respective FAEE. All other fatty acids were determined
with the relative response of 16:0-EE. Letters i and a indicate iso-fatty acids (fatty acids with a methyl branch on the second carbon counted from the tail) and anteiso
fatty acids (fatty acids with a methyl branch on the third carbon counted from the tail). Position of double bond not determined, owing to the elution order <n-9. Coelution
of 18:1n-9trans and 18:1n-9cis.
e
f
g
SIM Masses of Simple Esters of Fatty Acids. Our recent
GC-MS/SIM method is based on the determination of m/z 87
and 74 for saturated and monoenoic FAME (both fragment ions
contain the ester group), whereas di- to hexaenoic acids were
determined by the sum of m/z 79 and 81, which arise from
C6H7 and C6H9 radical cations (11). The two fragment ions
suggested for PUFA were thus identical for all three classes of
esters (see Figure 3). Due to the coelution of FAME and d3-
FAME, d3-FAME of naturally occurring PUFA are not suitable
IS for the SIM determination of FAME, whereas FAEE may
be used in the case of noninterfered elution of FAME using
m/z 79 and 81 (see below). By contrast, saturated FAEE and
d3-FAME appeared to be well-suited IS for the determination
of FAME, particularly when GC-MS is applied. FAME showed
virtually no response for the suggested SIM masses m/z 90
of m/z 74) of d3-FAME. Moreover, m/z 74 and 87 were not
present in the mass spectra of saturated d3-FAME.
FAEE did not produce any significant amount of m/z 74 and
87 so that FAEE could even be used in case of coelutions when
GC-EI/MS is applied in the SIM mode. This was verified by
the analysis of mixes of FAME and FAEE (data not shown).
Vice versa, a small response of the SIM masses for saturated
FAEE, m/z 88 and 101, was observed in the mass spectra of
FAME. Because no coelution of FAEE with saturated or
monoenoic FAME was observed, this did not play a role in
CP-Sil 88. However, partial coelution of 18:0-EE and 18:1n-
9-ME was observed on the SP-2331 column recently used (11).
Therefore, this issue should be checked when other GC capillary
columns are applied. In our case, no interference was observed
and the ethyl ester of any saturated or monoenoic fatty acid
can be used as IS for the GC-MS/SIM determination of FAME.
•
+
•+
(<0.02% of m/z 87) and very little response for m/z 77 (∼0.25%

Quantification of Fatty Acids as Methyl Esters in Food
J. Agric. Food Chem., Vol. 54, No. 9, 2006 3213
Table 3. Quantitative Determination of Selected Fatty Acids in Cod
Liver Oil and Goat’s Milk Fat
cod liver oil (n
(g/100 g)
)
3)
goat’s milk fat (n
(g/100 g)
) 3)
FAME
i15:0
a15:0
0.34
0.05
0.10
0.56
0.10
12.3
0.87
4.93
0.53
0.15
18.3
4.29
3.43
12.4
17.5
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.04
0.21
0.28
0.28
0.95
0.14
20.4
0.65
0.73
0.87
2.03
13.6
0.49
1.27
±
±
±
±
±
±
±
±
±
±
±
±
±
0.01
0.03
0.00
0.01
0.01
0.28
0.08
0.08
0.03
0.13
0.05
0.06
0.04
0.01
0.01
0.06
0.08
0.26
0.02
0.31
0.02
0.01
0.63
0.21
0.09
0.24
0.72
1
1
4:1n-5
5:0
i16:0
1
1
1
1
1
1
1
1
2
2
6:0
6:1n-9
6:1n-7
7:0
8:1n-9trans
8:1n-9
8:1n-7
8:2n-6
0:5n-3
2:6n-3
a
nd
nd
a
nd, not detected (limit of detection 0.003 g/100 g of lipids).
fatty acids to be quantified in the range of 0.003-54 g per 100
g of lipids in the sunflower oil (Table 2). Furthermore, these
1
8 FAME include 4 of the 8 FAME currently used as IS (see
Introduction). However, some variations occurred for unsatur-
ated fatty acids between the SIM and scan modes. The higher
value for 18:1n-9 and the lower value for 18:2n-6 in the SIM
mode (Table 2) are in agreement with previous reports (11).
Although still in an acceptable range, we recommend that
quantification of unsaturated fatty acids in the SIM mode should
be carried out with unsaturated FAEE as IS. On the other hand,
the good precision in repetitive analysis of minor fatty acids
such as 15:0 which was not detected in the scan mode makes
the SIM method particularly valuable for low abundant fatty
acids in food. Whereas branched-chain fatty acids were not
present in the sunflower oil, nine branched-chain fatty acids
could be quantified in buffalo mozzarella (Table 2). The lower
amount quantified (∼85%) must be due to the presence of
phospholipids, which contain a lower amount of fatty acids and
nonfatty acid lipid components. Thus, the yield after trans-
esterification was only 91.0 ( 3.1% (n ) 4) of the sample
weight after extraction. The results of scan and SIM analyses
agreed well except for the coelution of oleic and elaidic acids
in the scan mode, which was the reason for the difference in
the fatty acid composition (Table 2). However, the better
resolution in the SIM mode allowed for the correct determination
of this trans-fatty acid. Because our method is not based on
the conventional 100% method, it is not necessary to analyze
all fatty acids to determine the concentration of specific fatty
acids in a sample. For instance, the concentration of ∼1 g of
elaidic acid per 100 g of lipids of mozzarella cheese (Table 2)
can directly be determined without determination of the 40
additional fatty acids found in this mozzarella sample. This
technique was applied to cod liver oil and goat’s milk fat (Table
3). Note that only the fatty acids shown in Table 3 have been
quantified. No determination of the relative contribution to all
fatty acids (100% method) has been carried out, which is
currently the standard technique of FAME determination by GC-
FID. As can be seen from the low standard deviations, the
method provides good reproducibility even for complex lipids.
Only for low abundant fatty acids were the variations between
the replicates slightly higher but still acceptable. Hence, the
method allows the determination of not only minor fatty acids
such as odd-chain fatty acids, methyl-branched fatty acids, and
trans-fatty acids but also major fatty acids such as palmitic acid
or eicosapentaenoic acid (Table 3).
Figure 3. GC-EI/MS mass spectra (excerpt m/z 50
−150) of the (top)
methyl ester, (center) -methyl ester, and (bottom) ethyl ester of
d
3
eicosapentaenoic acid (EPA, 20:5n-3).
Quantification of Fatty Acids in Food Samples as FAME
Using a Series of FAEE as IS. The method of quantification
of FAME by GC-MS using five FAEE as IS was carried out
with sunflower oil and mozzarella cheese (Figure 1). Because
the IS can be added only after conversion of the fatty acids
into FAME, we had to verify that the transesterification
procedure could be carried out quantitatively. For this purpose,
the extracted lipids were accurately weighed and transesterified.
Sunflower oil contains fatty acids as triacylglycerides, and the
sample weight of the transesterified FAME differs by only 4
Da from that of triacylglycerides because glycerol is not found
in the FAME fraction. Transferred to three fatty acids with a
mean chain length of 16 carbons, the error would be ∼0.5%,
and this was considered to be negligible, so no molar correction
was performed. The recovery rate of the transesterified and
purified sunflower oil was gravimetrically determined to be 97.4
(
2.5 (n ) 6). Therefore, the transesterification procedure could
be carried out without significant loss of fatty acids. This was
confirmed by the conversion of free fatty acids into FAEE (see
Materials and Methods).
The methylated fatty acids of sunflower oil were combined
with five internal FAEE standards and determined by GC-MS
in the scan and SIM modes (Table 2). In the scan mode, 11
fatty acids were detected ranging from 0.08 to 57 g/100 g of
lipids. More than 99% of the sample weight could be quantified
in any of the determinations (Table 2). This is in very good
agreement with the gravimetric determination shown above.
Because of the higher sensitivity, GC-EI/MS/SIM allowed 18

3214 J. Agric. Food Chem., Vol. 54, No. 9, 2006
Thurnhofer and Vetter
Although not exclusively shown, d3-FAME will also be
(7) Noti, A.; Biedermann-Brem, S.; Biedermann, M.; Grob, K.
Determination of central nervous and organ tissue in meat
products through GC-MS analysis of marker fatty acids from
sphingolipids and phospholipids. Mitt. Lebensmittelunters. Hyg.
suitable IS for quantification purposes so that both classes of
alternative esters are real substitutes for saturated odd-numbered
FAME, which are currently most widely used as IS. When GC-
MS is used, an almost unlimited number of IS (FAEE and d3-
FAME) can be applied for the determination of FAME. The
alternative esters (d3-FAME and FAEE) are readily prepared
from standard chemicals so that the standards can be produced
at low costs in any laboratory. Note, however, that our concept
is applicable only when the internal standard is added after
transesterification of food lipids (6, 9, 10). Nevertheless, we
are convinced that the proposed method will be helpful for both
quality assessment and quantitative determination of FAME as
was illustrated in this study.
2
002, 93, 387-401.
(
8) Ackman, R. G. Remarks on official methods employing boron
trifluoride in the preparation of methyl esters of the fatty acids
of fish oils. J. Am. Oil Chem. Soc. 1998, 75, 541-545.
9) Yurawecz, M. P.; Molina, A. A.; Mossoba, M.; Ku, Y. Estimation
of conjugated octadecatrienes in edible fats and oils. J. Am. Oil
Chem. Soc. 1993, 70, 1093-1099.
(
(
(
(
10) Dodds, E. D.; McCoy, M. R.; Rea, L. D.; Kennish, J. M. Gas
chromatographic quantification of fatty acid methyl esters: flame
ionization detection vs. electron impact mass spectrometry. Lipids
2
005, 40, 419-428.
11) Thurnhofer, S.; Vetter, W. A gas chromatography/electron
ionization-mass spectrometry-selected ion monitoring method for
determining the fatty acid pattern in food after formation of fatty
acid methyl esters. J. Agric. Food Chem. 2005, 53, 8896-8903.
12) Weichbrodt, M.; Vetter, W.; Luckas, B. Microwave-assisted
extraction and accelerated solvent extraction with the solvent
mixture ethyl acetate/cyclohexane (1:1, v:v) in view of quantita-
tive determination of organochlorines in fish tissue. J. AOAC
Int. 2000, 83, 1334-1343.
13) Smith, P. K.; Schewe, L. R. Preparation of deuterated esters of
carboxylic acids. Br. Patent 1976; GB 74-48109 19741107.
14) G a¨ umann, T.; Bonzo, R. Gas-chromatographic retention indexes
of deuterated compounds. HelV. Chim. Acta 1973, 56, 1165-
Supporting Information Available: Separation of FAME and
d3-FAME on a Factor Four CP-Sil8MS column and GC-EI/
MS full scan spectra of alkyl esters of 17:0. This material is
available free of charge via the Internet at http://pubs.acs.org.
LITERATURE CITED
(
1) Christie, W. W. Preparation of ester derivatives of fatty acids
for chromatographic analysis. In AdVances in Lipid Methodologys
Two; Christie, W. W., Ed.; Oily Press: Dundee, Scotland, 1993;
pp 69-111.
(
(
(
2) Shantha, N. C.; Napolitano, G. E. Gas chromatography of fatty
acids. J. Chromatogr. 1992, 624, 37-51.
1
176.
(
3) Roz e` s, N.; Garbay, S.; Denayrolles, M.; Lonvaud-Funel, A. A
rapid method for the determination of bacterial fatty acid
composition. Lett. Appl. Microbiol. 1993, 17, 126-131.
4) Lepage, G.; Roy, C. C. Improved recovery of fatty acid through
direct transesterification without prior extraction or purification.
J. Lipid Res. 1984, 25, 1391-1396.
(
15) Falconer, W. E.; Cvetanovic, R. J. Separation of isotopically
substituted hydrocarbons by partition chromatography. Anal.
Chem. 1962, 34, 1064-1066.
(
Received for review December 2, 2005. Revised manuscript received
March 8, 2006. Accepted March 11, 2006. We are grateful to the
Deutsche Forschungsgemeinschaft for financial support.
(5) Eder, K. Gas chromatographic analysis of fatty acid methyl esters.
J. Chromatogr. B 1995, 671, 113-131.
(
6) Mansour, M. P.; Volkman, J. K.; Jackson, A. E.; Blackburn, S.
I. The fatty acid and sterol composition of five marine di-
noflagellates. J. Phycol. 1999, 35, 710-720.
JF053022J