Could anybody advise me the way to analyse Ethyl alcohol in excess of water by GCMS? Maybe I could use a head space analisis or it is possible to exctract it from water?
Thanks,

You can easily do by Head space GC-MS analysis.You can keep your sample(5ml) in a head space vial with 1 gm of anhydrous sodium sulphate & thermostat at around 60-80 deg.c for 30 minutes with shaker on. Similarly keep your aqeous std with 1 gm of anhyd.sodium sulphate in head space vial.In PerkinElmer headspace analyzer you can control the injection volumes of head space vapour very precisely & accurately.You can use either a wide bore with a restrictor column or a narrow bore column.The MS filament can be kept off intially to avoid any disturbances from water.

Thanks,
Sunil

In my humble opinion:

This gross sample size is not necessary and will be counter productive.

Use at MOST, 100µL of sample and use the 6mL vial for the PE unit or the smallest vial for other manufacturers's units.

Likewise salt is not necessary or advisable. It is only needed with gross samples.

Heating for 15 min is overkill at 80°C. 10 min is adequate for linearity of response.

Sunil's other comments I agree with, but I prefer to use a wide bore with restrictor. A 0.25 mL sample loop (for non-PE units) will be adequate.

See AC Vol 69 Number 11, p2221-2223 for details.

Rod

Thank you very much for your help!
I have got Varian CP-3800 GCMS, MS - Saturn 2000 with a CPSIL8MS capillary coloumn ( 30 m, D 0.25 mm ). Is my coloumn OK for the headspace analyses you advised?

Could you please explain what AC Vol 69 meanes?
Thanks.

I'd say it means Analytical Chemistry journal, volume 69.

Quite right. 1 June 1997 issue of Analytical Chemistry

Sorry if I was unclear in my reference.

FYI

This article shows linearity of Head Space response of 18 common solvents including the following alcohols from water:

methyl R2 = 0.9996
ethyl R2 = 0.9998
isopropyl R2 = 0.9998
propyl R2 = 0.9998
and isobutyl R2 = 0.9998

spiked 0.50 to 1000.25 ng levels per sample of 25µL (25 mg) water with 6 minutes of heating at 85°C and mixed for 30 seconds and equilibrated for 30 seconds. recovery of a 50ng sample spike for ethanol was 98%, or 49ng as measured from the regression of the stds.

That level is 1 ppm wt/v in pure water or 50 ppm for a 1 mg sample dissolved in the water (a 4% solution).

I used the Varian Genesis HS (Tekmar 7000)analyzer but the method also works on the PE HS40 analyzer.

BTW I performed the baseline separation of these 18 solvents in 6 minutes.

Rod

My apology:

Due to typing error in my previous post I said (incorrectly) "1 ppm wt/v in pure water". It is of course, "2 ppm wt/v in pure water" at a 50 ng spike of ethanol into 25 mg of water.

The lowest std (0.5ng) is 20 ppb in water. Also I was using a FID not a MSD or quad MS

Rod

Rod,
Thanks a lot for your help!
I am still waiting for the article you advised but I have tried to do analyses.
I need to work with the concentration from 0.1% to 1% ethanol in water v/v.
I used 1 ml of 0.1% in 10 ml vial ( this is what I have got )by using the conditions you described. 100 microl of sample volume for the injection. As I have no info about the GC temperature programme I tried at 40C for 10 min.Injector temperature 250C ( I can not understand the relationship between the oven temperatures and the injector), split ratio 20 .I had few peaks at about 2-3 min without a proper separation. As it was nearly "on the way home" I didn't have a chance to look at the peaks spectra. What do I need in order to get a better separation?
Thanks in advance.

0.1% ethanol by volume in water is about 1mg per mL.

That is pretty concentrated for headspace! and pretty concentrated for mass spec as well.

Don't you have a TCD or FID detector on a GC?

You are not using the proper equipment for your analysis requirements.

I believe I understand your level of experience with GC more clearly now.

I would not recommend that you perform this analysis with a mass spec as water might elute closely to your ethanol peak (you never mentioned the column you are using), even at headspace levels it will cause you great trouble with your MS.

I wish you well.

Rod

Rod,
I am using Varian CP-3800 GCMS, MS - Saturn 2000 with a CPSIL8MS capillary coloumn ( 30 m, D 0.25 mm ) which I have mentioned before.
Unfortunately, I have not got TCD or FID detector on a GC. Do I need to dilute samples??

The problem is that you are going to inject water into your GCMS and it will elute very close to the ethanol.

I do not think it will be feasible to perform the analysis with the equipment you have.

You could perform a lot of method development and do extractions of the ethanol in water at different concentrations and generate recovery curves but I do not think it would be worth all the labor involved when all you need is a simple isothermal GC with a FID or TCD ($12000?).

Send the samples out for testing at a contract lab and save yourself and your company money and time.

Best wishes, Rod

Well, I changed my mind.

Go ahead and try the analysis. Be sure to scan above 18 for the ethanol ion. It may be feasible but I would feel better about the peak shape and broadening if you had a different column, perhaps a 1301 with a thick film.

Split the sample or dilute the sample a great deal in a non aqueous matrix so your water will not puddle on the column as much. (It goes against my principles to shoot water on a cap column at temperatures that allow the water vapor to liquify.)

Do what you have to do. Good luck Rod

If you have the latest Varian autosampler, I believe that it allows you to do headspace (unheated) from 2 mL vial. I would try this, but I would definitely use a different column. Either a wax column or a PLOT column, probably a PLOT-Q, will separate the alcohol from the water and allow decent results. I would not scan down to 18 and see the water, I would start at 30.

Good luck. The analysis is possible with the right column, but this is really an FID application.

I have a chromatogram of a sample containing 10% ethanol and isopropanol in water on a GCMS with the conditions used.

Please email me for the conditions and for a copy of the chromatogram.

An Equity 5 30m 0.25mm 0.25µm - a bonded 5% Phenyl Methyl silicone was used, similar to the CPSIL8MS column you are using.

rgeorge@sial.com

Rod

As i do in my studies, you can determine the amount of ethanol easily in water by FID detection and a internal standard with a column who accept water sample. I try this with unalcohlic water

for the last mail: unalcohlic water is a mistake and will be: unalcohlic beer