Chromatography Forum

Hello all,
I need some help in calculating the empirical correction factor (ECF). The AOCS method notes that the analyst can confirm the detector correction factors with pure reference standards. If I have a EPA methyl ester and DHA methyl ester standard and determine the ECF, what ECF would I use to correct all other fatty acid methyl ester derivatives (FAME) in the sample ? I have read that some labs use no individual response factors when analyzing FAME's. Here is the AOCS method Ce 1b-89 calculation:


EPA or DHA mg/g= ((Ax)(Wis)(CFx)/(Ais)(Ws)(1.04)) x 1000


where Ax: area counts for EPA or DHA
Ais: area counts for internal standard
CFx: theorectical correction factor for EPA or DHA
Wis: weight of ISTD added to the sample, in mg
Ws: test portion weight, in mg
*1.04: is needed to convert methyl esters into fatty acids

This equation is used to calculate all other FAME's.


The previous chemist, as part of the system suit, compared the known amount of EPA and DHA (ug) in the standard sample to the amount (ug) calculated by the system with use of an internal standard. These factors would then be applied to correct by weight all other FAME's. The chemist also called for adding known amounts of the ISTD to all samples, blanks, and standards. The ISTD was added to the sample after sample preparation. Since we would determine instrumental precision by the standard injections which bracket the sample, there is no need to add the same amount of ISTD to all vials, correct? System suitablity is verified by the standard peak areas producing an RSD of < 5% . Even though RSD % should be calculated using a minimum of 3 injections, is it okay to report the RSD % of two standard injections, before and after sample injections? This of course would just be the difference between the two injections. I think that the ISTD should be added before sample preparation so that any sample loss from extraction is accounted for and to increase the intra assay precision. Can someone please advise or direct me to a good source for more information. Thanks !!

From the "Scope" section of AOCS Ce 1b-89 (emphasis mine):

This method is designed to determine the fatty acid composition of marine oils and marine oil esters in relative (area %) values, and eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in absolute (mg/g) values using a bonded polyglycol liquid phase in a flexible fused silica capillary column.

Ce 1b-89 is not intended to give you absolute fatty acid/ester content for anything other than EPA and DHA; other esters are quantified on an area percent basis only. For quantifying absolute amounts of other fatty esters in marine oils you should look at AOCS Ce 1i-07, which contains a table of FID TCFs (theoretical correction factors) and a very complete section on how to calculate ECFs and how much they are allowed to deviate from the TCF.

Thanks so much for your guidance ! I will check that method out

Hello Brewer or anyone else who can answer,

Why is it that only the "relative area percent" of the other FAME's can be calculated, while the absolute values of EPA and DHA are determined from AOCS Method Ce 1b-89 ? I need to quantitate total n-3's ,n-6's, and n-9's.

Shouldn't I be able to report the absolute weights of all other FAME's in relation to the weight of the ISTD ? I will be using the GLC-714 (standard FAME mix) according to AOCS Method Ce 1i-07 to calculate my empirical correction factors for FAME's containing 20 or more carbons, using the TCF for FAME's containing less than 20. Thank you in advance for your help

To answer your questions: Yes, you should be able to report the absolute concentrations of all other FAMEs by relating them to the ISTD concentration and correcting them for detector response ... but only if you're using Ce 1i-07. And, the reason you can't calculate absolute concentrations of other FAMEs using Ce 1b-89 is that Ce 1b-89 contains correction factors only for EPA and DHA. If you want total omega-3s, 6s, and 9s, you will have to calculate your ECFs ... which means using Ce 1i-07, not Ce 1b-89.

Does that make sense? The two methods aren't that different other than the approach to quantitation.

Great ! Thanks for your help Brewer

The ECF has been working out good with the GLC standards in comparison to TCF. However, the most current GLC standard solution I made the response to the saturated esters ( C23:0, C24:0, etc) gave smaller peak areas compared with the unsaturated esters. I changed out the liner, and did a bake out on the detector. I know the ECF is supposed to correct for inlet discrimination, column loss, and detector response but how do we know there is not something wrong with the standard itself? For instance, the certificate of analysis states that all the esters each are 4 % by weight. If the internal standard C23:0 is actually 3 % by weight then this affects the ECF. I am getting around 0.81 for ECF of DHA when the TCF is 0.96. I typically get 0.95 for the ECF for DHA. I am just going to make a new standard mixture but if anyone can explain the ECF more clearly that would be great

Hello All:

The best two references I've read concerning measurements of DHA/EPA and other unsaturated fatty acids in Fish Oils is the following:

J Chrom. A. 2004 Aug 6;1045(1-2):197-202. Determination of the carbon deficiency in the flame ionization detector response of long-chain fatty acid methyl esters and dicarboxylic acid dimethyl esters. Authors: Schreiner, Matthias, Hulan HW.

Abstract

Carbon deficiencies (CDs) of long-chain fatty acid methyl esters (FAMEs) and dicarboxylic acid dimethyl esters (DDMEs), which lead to decreased response in a flame ionization detection (FID) system, were determined by using full responding hydrocarbons (heptadecane, eicosane and alpha-cholestane) as references. For saturated FAMEs ranging from C12 to C22 and for DDMEs ranging from C4 to C10, CDs between 1.3+/-0.12 and 1.7+/-0.36 per ester group were recorded. All values were significantly (P < 0.05) greater than 1. Generally, response factors for gas-chromatographic analysis using FID have been calculated on the theory that the CD of FAMEs is 1 per ester group. However, this theory could not be confirmed experimentally for short-chain FAMEs of less than 8 carbons as CDs of around 1.5 were reported for C4 and C6 FAMEs. The study presented here contributes an approach to this problem by confirming the validity of response factors calculated from a CD of 1.5 per ester group as well as for long-chain FAMEs and DDMEs.


J Chrom. A. 2005 Nov 18;1095(1-2):126-30. Quantification of long chain polyunsaturated fatty acids by gas chromatography. Evaluation of factors affecting accuracy. Author: Schreiner, Matthias.

Abstract

The accurate and reproducible analysis of long-chain polyunsaturated fatty acids (PUFA) is of growing importance. Especially for labeling purposes, clear guidelines are needed in order to achieve optimum accuracy. Since calibration standards cannot be used for method validation due to the instability of PUFAs, there is no direct way to check for the absence of systematic errors. In this study the sources of error that weaken the accuracy were evaluated using theoretical considerations and calibration standards with corrected composition. It was demonstrated that the key role for optimum accuracy lies in the optimization of the split injection system. Even when following the instructions outlined in the official methods of the American Oil Chemist's Society (AOCS), systematic errors of more than 7% can arise. Clear guidelines regarding system calibration and selection of appropriate internal standards (IS) can improve precision and accuracy significantly.

These articles are "must-read" in my opinion. At my old job, we also decided that the best way to perform DHA/EPA measurements was to use cold-on column injection.

Cyberlipid and lipidlibrary.aocs.org/ are two very useful websites as well...if you folks already know of these, I apologize for "preaching to the choir."