food& flavor
Applications
note
RestekCorporation • (800)356-1688 • (814) 353-1300 •
#59585A
food& flavor
UsingComputerModeling toOptimize FAMEAnalysis
Background
Gas chromatography (GC) is an effectivemeans of characteriz-
ing fatty acids in food aswell as othermatrices. TheAssociation
ofOfficial Chemists (AOAC) and theAmericanOil Chemists
Society (AOCS) provide severalmethods forGC analysis of
fatty acidmethyl esters (FAMEs). AOACMethod963.22
provides general guidelines and conditions for analyzing awide
range of saturated andunsaturatedFAMEs, fromC8:0 toC24:0,
usingpacked columnGC
1
. If capillaryGC is used as an alterna-
tivemethod, specifications canbemet or exceeded easily.
AOACOfficialMethod991.39
1
andAOCSOfficialMethodCe
1b-89
2
describe the analysis of polyunsaturatedFAMEs in fish
oils using capillaryGC
2
. Thesemethods list conditions for
separating all of theFAMEs in complex fishoils onpolyethylene
glycol (PEG) and cyanopropyl stationary phases. By properly
optimizing analytical conditions, you can improve resolution and
decrease analysis time.
Unfortunately, the typical optimization process can be extremely
time consuming and frustrating. Furthermore,many analysts
identify unsaturated fatty acids, including polyunsaturated fatty
acids (PUFAs) by equivalent chain length (ECL) values rather
thanby retention time
3
. Changing conditionsmay alter the
elutionorder, and require peak re-identification. Trial and error
withdifferent column configurations andparameters, alongwith
the necessary re-identifications, canwaste additional analyst and
instrument time.Amore efficient approach tooptimizing
analytical parameters is touse computermodeling software.
Programs such as
Pro ezGC
™
use thermodynamic retention
indices (TRIs) tomodelGC analyses andprovide optimized
conditions inminutes. Additionally, this program can recalculate
theECLvalues for the new set of run conditions.
Pro ezGC
™
can
provide optimizedFAME analyses thatmeet or exceed the
resolution andECL specifications stated in the officialmethods.
HowWellDoesComputerModelingWork?
To illustrate the accuracy and efficiencyof themodeling
software, amixture of 21 saturatedFAMEswere analyzedon a
60m, 0.25mm ID, 0.25µmRtx
®
-Wax column (cat.#12426). The
FAMEs ranged frommethyl butanoate (C4:0) tomethyl
tetracosanoate (C24:0). TRIswere generatedby analyzing the
mixturewith twodifferent temperature programs. The first
programwas a relatively slow ramp,while the secondwas a fast
ramp. The resulting retention times for each componentwere
entered into the program, which automatically calculated the
TRIs. The software thenwas able to evaluate awide range of run
conditions and predict the resulting retention times under each
set of conditions. Once the predictions aremade, the software
selects the set of conditions that provides the best separation in
the fastest analysis time. Todemonstrate the accuracyof this
FAMEsMSDData
60m, 0.25mm ID, 0.25µm
Rtx
®
-Wax column (cat.# 12426)
Oven temp.:
45°C@ 6°C/min. to 265°C
(hold 7min.)
Carrier gas:
hydrogen (constant pressure)
Linear velocity:
53.5cm/sec.@ 45°C
Dead time:
1.980min.@ 45°C
Component
Exp. tR
Calc. tR
Exp. -Calc.
(Exp. -Calc.)/
(min.)
(min.)
Error (min.)
Exp. % Error
1. me butanoate
3.918
3.924
-0.006
-0.1
2. me pentanoate
5.315
5.327
-0.012
-0.2
3. me hexanoate
7.158
7.171
-0.013
-0.2
4. me heptanoate
9.293
9.299
-0.006
-0.1
5. me octanoate
11.573
11.577
-0.004
-0.0
6. me nonanoate
13.855
13.854
0.001
0.0
7. me decanoate
16.083
16.075
0.008
0.0
8. me undecanoate
18.228
18.205
0.023
0.1
9. me dodecanoate
20.282
20.253
0.029
0.1
10. me tridecanoate
22.247
22.191
0.056
0.3
11. me tetradecanoate
24.127
24.078
0.049
0.2
12. me pentadecanoate
25.927
25.855
0.072
0.3
13. me hexadecanoate
27.653
27.567
0.086
0.3
14. me heptadecanoate
29.310
29.238
0.072
0.2
15. me octadecanoate
30.902
30.856
0.046
0.1
16. me nonadecanoate
32.432
32.382
0.050
0.2
17. me eicosanoate
33.907
33.824
0.083
0.2
18. me heneicosanoate
35.330
35.205
0.125
0.4
19. me docosanoate
36.702
36.542
0.160
0.4
20. me tricosanoate
38.120
37.872
0.248
0.6
21. me tetracosanoate
39.710
39.396
0.314
0.8
Avg. error:
0.070
0.2
Table I
Experimental retention times versus predicted retention times.
