Chromatography Forum

We are encountering difficulty in GC / FID analysis and quantitation of Methanol content in alcoholic beberages due to a very distorted methanol peak, perhaps due to the complex methanol-water.
Is there any way of improving the methanol shape either by sample preparation or via a proper selection of chromatography parameters?

I think you need to give more info for a useful answer:
Direct or headspace injection?
Column?
Temperatures?
etc
WK

What kind of column are you using?, sometime you see distorted methanol peak because you have a quite small cuantity of it, but as I say you What kind of column are you using if you have a big peak well ,I could think that you don't have a good separation and you could have other compound inside, try to be sure that you have an adecuate split ratio, some times that is the first thing that you hava to check, if not, you could prove with HS, or SPME as well, and check all the other parameter.

We are using on-Column injection on a DB-624 60m x 0.25mm di x 1.4 um.
Injection volumne : 0.1 ul
Injection Port T : 200ºC
Detector T : 300ºC
Oven: Tini: 40ºC for 5 min
Ramp A : 5ºC to 75ºC , 0 min
Ramp B : 10ºC to 140ºC, 2.5 min
Column Flow : 3.3. ml/min
Carrier Gas : N2.

GC: Agilent 6890N with EPC and ChemStation.

You are shooting on-column 100L of beverage onto a 250micron ID column !

This does not seem wise to me.

How about shooting onto a retention gap of 0.53mm ID, say about 1-2 meters worth.

Your deposits of non-volatile material will give you lots of reasons to have a poor methanol peak shape.

The excessive injector temperature seems uncalled for IMHO.

I would try a 0.32 or a 0.53mm column with a retention gap and a faster flow rate to achieve volatilization of your sample.

I would recommend to use headspace, much more accurate than direct injection. Lots of breweries use that technique. I know, I used to live in St. Louis !

The beberages are normally standardized to 40ºGL.
The injection volume is 1/10 of a microliter.
We do use a guard-column 1m x 0.53 micron id
The methanol concentration in the samples is normally around 10-20 ppm.
thanks
Bill

Sorry Bill, I meant to type 100nL, not 100L, which is 0.1µL

I suspect your column needs rinsing, your retention gap needs replacing, and you really should lower your injection temperature a BIT.

You can do an excellent job of using static headspace at a temperature of 80°C for methanol from a lot of matrices.

I would cut 3m from the inlet of your column and start with a new clean retention gap.

Good luck.

Why are you using nitrogen as carrier gas-is helium unavailable to you?

The optimum flow of carrier gas is much lower with nitrogen than helium. I suspect that your flow rate is much too high.

However, this may not solve your problem because it seems that you are getting at least some reasonable peaks from your system. Injecting water is always a problem due to the large volume of vapour it generates. The impact of the large sample volume will be greatest on the early peaks. The retention gap idea is a good one. Otherwise you might try a larger diameter column which will accommodate the sample volume better.

I used to on-column injection some times ago. In optimising on-column injection, I install a guard column of 5m in front of the analytical column for solvent focusing such that the peaks did not distort in shape. For your case, as your sample is water base, you may need to install a polar guard column (SGE provides guard columns of different polarity) for focusing your analyte before entering the analytical column for separation. Also, as you may know, on-column injection is only applicable for very clean sample. Thus, peak distortion is usually observed after sample injections, in particular to dirty samples, as dirt may deposit inside the guard column. At that moment, you have to re-install a new guard column. Hope this help.

With an injection port temp of 200C there may not be time for the methanol to focus on the column. Try reducing the injection temp to 40C and get it to follow the oven temp profile.
GCguy