Chromatography Forum

hiya,
new to this forum but thought i quickly try to pick your brains. i'm trying to run some FAMEs on a GC-FID to do oil profiling for a practical. I've done the derivatisation using 100uL of oil in 10 mL pentane, vortex then adding 0.5mL 2M KOH in methanol. vortex for 15 s, let layers separate, take top layer, filter and inject.
I'm using a PE clarus 500 GC-FID with a DB-5 (15m x 0.25 mm x 0.25 um) column with a semi slow ramp from 60oC to 260oC working with a 20psi helium flow and a split of 25:1.
i get 3 triangular peaks which show huge fronting and i can't seem to get them any better shaped no matter how much i turn the split up or tweak the program. same peaks on a 30 m DB-1 and a 30 m DB-23.
i've injected a FAMEs standard mix from supleco - no change. blanks are clear btw.
any ideas how to separate my peaks?
thanks a lot.

A DB-5 cannot handle acidic based material; you will get those triangular peaks no matter what you do. I would suggest trying a different column. You might try a different column

Work out how much FAMES you are putting onto the column, then dilute the solution enough to give you less than 100 ng per component.

Peter

You seem to be saponifying the oil to a free acid, rather, a K salt of the acid. How are you preparing the Methyl esters?

Rod

thanks for your replies. I'm pretty sure the DB-5 can handle the FAMEs as I have a paper comparing exactly that and they even get good peaks using a DB-1. anyways, the much more polar Db-23 didn't come up with anything better so i start to wonder whether it is some kind of system error.

my method of preparation is:
100uL oil
10mL pentane
vortexing
add 0.5mL of 2M KOH in methanol
vortex for 15 s
after layer separation take top layer and filter through 0.45um filter
inject

i've now also done a oleic acid standard which was a 250 ng/mL hexane solution and i've taken 100uL and poceeded like above. i got no peaks whatsoever.

i understand that maybe the esterification of the oil is going wrong but that doesn't explain why my bought-in standard mix of c16:0, c18:0, c18:1, C18:2 and c18:3 as FAMEs isn't giving me any good peaks. the concentration i'm injecting is 4mg/mL in pentane.
thanks

You may be injecting splitless or with too little split. Your vent should have a flow of 50cc/min or more.

What does it have? Your peaks are indicative like Peter said, to too large an amount of FAME entering the column for the film thickness and column ID you are using. It is a classical example of column overload.

chromatogram please, and show everyone what we are talking about.

Rod

We assay FAME routinely, for decades, as FAME dissolved in hexane, and still 0.1 ul injection. We use polar packed column, polar capillary, or non-polar capillary (split injection), and get nice peak shapes, no issues.

With polar columns, the unsaturated components elute after the saturates of same carbon number (e.g. 18-0, 18-1, 18-2).

With nonpolar columns, the unsaturated components elute before the saturates of same carbon number (e.g. 18-2 18-1, 18-0).

i'm trying to run some FAMEs on a GC-FID to do oil profiling for a practical.

So, a practical is like a project or test? And basically you have little practical experience in chromatography. Good luck.

So, a practical is like a project or test? And basically you have little practical experience in chromatography. Good luck.

no, i'm setting up a practical for others, i'll be the one who is running it. i have lots of experience in GC (various detectors) that's why it bugs me so much that i can't make it work. i've done FAMEs at uni once but can't remember what the settings where as it was quite a few years ago. was just hoping to pick your brains in case i'm completely overlooking something.
will open the split even more, lower the injection volume and maybe increase sensitivity as peaks are quite small, and then see what i get today. try to post a chromatogram later.
i usually don't give up but these ugly peaks are bugging me

So, a practical is like a project or test? And basically you have little practical experience in chromatography. Good luck.

no, i'm setting up a practical for others, i'll be the one who is running it. i have lots of experience in GC (various detectors) that's why it bugs me so much that i can't make it work. i've done FAMEs at uni once but can't remember what the settings where as it was quite a few years ago. was just hoping to pick your brains in case i'm completely overlooking something.
will open the split even more, lower the injection volume and maybe increase sensitivity as peaks are quite small, and then see what i get today. try to post a chromatogram later.
i usually don't give up but these ugly peaks are bugging me

If you have small peaks that are front tailed from overloading it suggests a polarity mismatch between phase and analytes - which brings us back to the possibility that what you are seeing are free acids. The problem then is how are free acids forming during a base catalysed transesterification ? Nonetheless, small peaks can be generated from large quantities of analyte for any number of reasons, and you will then see small overloaded peaks from analytes that are a good match to the stationary phase. You still need to know what quantities you are loading onto the column.

Simple FAMES are among the most tractable of compounds for GC, and do not need any very fancy settings beyond the generic optimums for flow rate and column programme rate, and inlet and detector temperatures around the 250 - 300C to get a decent looking chromatogram that can be the basis for further optimization as required.

Peter

Thanks Peter,
this is what i thought that nothing overly fancy is required. I have just diluted my FAMEs standard mix to 40ug/mL pentane to get around 40ng onto the column when injecting 1 mL. silly me, i also set the split to 100:1 so no peaks at all... i think i'm trying to change too many things at the same time at the moment so will go back a few steps.

i understand your comments about producing free acids but at the moment i'm testing everything with a standard FAMEs mix bought in from supelco. that comes as a neat mix which i dilute down in pentane. but my first injected concentration was probably too high hence ugly peaks.

Use hexane instead of pentane and inject 0,1 µL to 1 µL with 1 µL solvent wash, fast injection. Injection around 250 °C, splitt about 20.

Use hexane instead of pentane and inject 0,1 µL to 1 µL with 1 µL solvent wash, fast injection. Injection around 250 °C, splitt about 20.

As far as anyone can tell, the peaks are overloaded. How would these conditions help with that ?

Peter

These are the settings I am using for FAME analysis and it will give you no severe overload when you inject 0.1 µL of 10 mg/mL FAME mixture (>10 components).

i have sorted it, it was a mixture of poor derivatisation and therefore column contaminated and then column overload.
i've changed the procedure from the paper doing the following now:
100uL of oil in 10 mL pentane
take 1 mL of this mix plus 1 mL of 2M NaOH in methanol, vortex for about 30 s
after layers have separated, take top layer and add a spatula tip of sodium sulfate anhydrous to dry in case there is any water left.
inject 1uL with a split of 20:1 and voila, small but separated peaks
like peter said, no need for fancy settings.
thanks everyone.

Thanks for the feedback, it's good to hear what works in the end.

Peter