fatty acid methyl ester split peak

[skunked_once]

My fatty acid methy ester analysis sometimes results in a split peak for palmitic acid methyl ester. Normal retention time is 2.32 minutes but there is often a second peak at 2.34 minutes which can be up to 50% of the total of the two peaks. GC conditions: DB-23, 30m x 0.25 mm (0.25u) column at 29 psi He and 190 C. Split flow anywhere from 50 to 150 mL/min. Straight glass liner with a 1/4" glass wool plug. Samples are prepared by crushing or grinding the seed and mixing a portion with hexane/chloroform/sodium methoxide solution. Has anyone else seen this or does anyone have an explanation?

[chromatographer1]

artifact from injection technique or needle tip not uniformly eluting sample

deposits on front end of column or in glass wool or in injection port liner

an isomeric form of palimitic acid me nearly coeluting with the Palmitic acid ME.

These are common causes for split peaks in capillary chromatography.

[skunked_once]

chromatographer 1,

Thanks for the reply and suggestions. I don't think it is from the injection process. Injection is by autosampler but the split peak problem appears to be random. A dirty liner, glass wool, column inlet end may be the cause. I will monitor the palmitic peak after I change the liner and the septum. Question for other experts; Is it possible to have isomers of palmitic acid or palmitic acid methyl ester? If so, are they naturally occuring or produced during the derivatization or injection (high temperature) process?

[Tony]

Palmitic acid cannot have any isomers, as it has a straight alkyl chain with no double bonds. Are you partitioning your sample prep into hexane? Normal procedure would be to transmethylate with sodium methoxide in methanol, but then partion into hexane against a dilute salt solution (e.g. 0.9% KCl). If you don't do this, you will be injecting glycerol and salts into your liner which will invariably lead to problems.

[chromatographer1]

I have to disagree with Tony, but one can have isomers of a C16 fatty acid, in the same manner you can have isomers of butyric acid, these are branched chain fatty acids having 16 carbons. These are not plentiful in nature but can be found. They are not formed under most derivitization reactions. However, they would not be separated well using a DB-23 polar column and can not be ruled out as a possibility.

Nevertheless, it is most likely that active sites in the sample flow path are causing the splitting of your peak. These are probably from leftover seed components that are not removed thoroughly from the derivitization solution.

I have personally seen deformed, nearly split peaks from a damaged needle tip of a syringe not 'spraying' the sample uniformly into the injection liner. Again not a likely cause but one I would suggest you investigate.

Also be certain you are not injecting too large a sample which can overload the expansion space of your injection liner, a problem which can also express itself as a split analyte peak.

Good luck,

Rod

[skunked_once]

Tony and Rod,

Thanks for your replies. I'm guessing that it is not the needle since the standard runs never have split peaks. Injection volume is only 1 uL which should be small enough to prevent flashback into the injector lines. That narrows it down to sample matrix and dirty liner. I will monitor both more closely next time I run samples.

[Consumer Products Guy]

I've assayed a gazillion FAME samples from soaps, fatty acids, triglycerides over a quarter-century, using both packed and capillary columns, and have never seen this. I agree that it sounds more like a chromatography, and agree that one should separate the FAME from other components using a salt solution. We make FAME using either boron trifluoride-methanol or sulfuric acid methanol, extract into neutral organic using salt solution. For fatty acid esters such as triglycerides, we just saponify first, in same vessel. We observe only a very trace of branched C16 FAME.

[wsayers]

If your instrument is a 6890, try running a fairly concentrated sample in split mode to see if the split peak goes away.

6890s have a habit of splitting early eluting peaks in splitless mode. This can be overcome by:

1) running in split mode at a low split ratio ~10:1 for example

2) pulsed mode may help


Good luck

[DR]

Are you cold trapping (if not, maybe you should try it)? Are you injecting a small enough sample (or injecting very slowly) to prevent any chloroform flashback?

[skunked_once]

I think I have it figured out. I was having problems with the syringe gumming up so I added some ethanol to the acetone in the needle wash vial to improve the needle washing. I should have realized that there is a possibility of formation of ethyl esters from traces of ethanol. This seemed to happen after several injections when the injection liner started getting dirty. After eliminating the ethanol (and replacing it with methanol), the split peaks have disappeared. Thanks for all of the suggestions which helped me eliminate other possibilities.