Chromatography Forum

Hello,

I'm trying to use FAME's in Olive Oil using an Ultra Trace GC with FID. I'm forced to use nitrogen as carrier gas and TR-225 60 m x 0,25 mm x 0,25 um column. May somebody help me with the setup of the rest of the GC method parameters?

Many thanks in advance.

You can go with the typical methods with methanolic NaOH, and BF3/methanol I believe Consumer Products Guy has a simple procedure with that.

I do some Fames work only very sparsely and most of the reagents, especially BF3/methanol have limited shelf life as they degrade and form water poisoning the reaction. I mainly just need a relative amount C16 C18 C18:1 C18:2 to identify an oil so I wound up using a standard wax column and some prepared methanol NaOH (not fully dry) then, aqueous HCl to acidify, then isooctane to extract and methyl chloroformate for esterfication.

Another method that looks promising is mild methanolysis. This link has the method as well as several standard ones.
http://www.jlr.org/content/51/3/635.full.pdf+html

As for column ramps and flow you can just pick a method and use Agilent's method translator for your column dimmensions and carrier gas.

Thanks for your reply MSCHemist. At this moment i'm just interested at the GC method parameters, so i can run some standards. The Agilent's method translator for your column dimmensions and carrier gas is not very useful since almost all the methods I've seen use He or H2 as carrier gas and 30 m columns, so when i try to translate them using N2 as carrier gas and the 60 m column, the translated run time results extremely long.

That is because the optimal linear velocity as seen on the van deempter curve is 1/2 of that for He and resolution is critical for FAME's so you are stuck with run times twice as long with nitrogen. Is hydrogen an option. Since you are running an FID you must have hydrogen available why not run that as a carrier instead?

You are exactly right as far as concerning the Van Deemper curve estimation. Hydrogen is not an option as carrier gas due to illogical safety issues. That is way i'm looking for some HELP here.

You don't have many options with nitrogen I'm afraid. You can exceed the van deempter recommendation and shorten your run and see how much resolution you give up (which is kind of pointless to spend the money on a 60m column to improve separation then waste it by a suboptimal flow parameter).

I'd see if I can push back on the hydrogen as you already have hydrogen coming out your FID at ~40 ml/min what difference does a 10 or so out the split vent and septa purge cause especially if you can plumb the exhaust ports to the hood vent or something.

You mean shortening the run by keeping the N2 flow rate above the recommended value for capillary columns of 0,25 mm ID?

Yes. You can run the flow rate higher than van deempter optimal and experiment and see how much resolution you loose but as I said you are going to the expense of getting a 60m column to do Fame and then you might be blowing it by not maximizing its separation potential.

There was one guy who claimed there might not be substantial loss of resultion but I never tried it as I only use helium for GC.
http://www.slideshare.net/jvercamm/the- ... n-12741949

I have already tried what that guy claims, 'leaving everything unchanged' and running with Nitrogen carrier gas, but i got just the solvent peak and nothing else.

http://www.scielo.br/scielo.php?script= ... 2000400023

Here is one article. They run fames with nitrogen at 20 cm/min that isn't too bad.

Yeah, I've read it already but the max temp of my column is 230 C, while in the method the max temp goes to 240 C. I do not prefer to go above it. Plus they mention 'operated under programmed temperature conditions: 140-240 ºC at 5 ºC min-1 in 30 min', while 140-240 C with a ramp of 5C/min makes 20 min. I guess they hold the 240 C for 10 minutes.

just take your column up to the the limit 230 and hold a bit longer to compensate. I got fames out to C24 at 230 on my wax column. 10 deg C at 230 deg won't make a huge difference.

i'm really having difficulties. I'm not getting any peaks except the solvent peak, whatever method i use. The solvent peak appears strange too, since it gets lower and lower from injection to injection. On my forth injection the solvent peak almost splits in two. Should i setup a post run method? Is anything wrong with the solvent? Should i condition the column?

Before worrying about the FAMES method you need to do basic GC maintenance and set-up. Change the inlet liner, and put in a new septum, make sure that column is properly installed at both ends. Check that you have the correct and expected gas flow rate by injecting gas from a cigarette lighter and measuring how long it takes to elute - it should take about 3.5 min. Check for leaks.

Check that your injection volume ans split ratio is what it is supposed to be, and that the screen display settings would allow you so see FAME peaks if they are there.

Check gas flows and signal processing settings of the detector.

If you can see that the solvent peak is getting lower than you must be seeing the top of the peak, which either means that you have seriously low sensitivity, or your screen settings are wrong.

Peter

I'd start by injecting a test mix to check the GC-FID. I have a kovats mix of alkanes as a test mix but any simple inactive compounds should suffice.