IPA and MEK detection in polymers

[mush]

I am assigned to detect IPA (Partition Coefficent = 825 @ 40C) and MEK (Partition Coefficent = 139 @ 40C) in polymer product using headspace GC.
I wonder if I should:
1. Prepare aqueous or non-aqueous sample prep.
2. Keep temp. at around 80C - 100C or more than 100C. BP for IPA is 82C and BP for MEK is 79C.
3. Keep phase ratio of 50 or more or less.
4. Use capillary or packed column.

[Peter Apps]

Question 4 is the easiest - capillary, but with the 1 ml loops that are standard in headspace autosamplers you will have to do a split injection to avoid volume overloading.

I cannot think of a reason to add water to the samples.

80 degrees C is probably hot enough - the hotter you go the more chance there is that the polymer will decompose into volatiles that interfere with the analytes. The only reason to go hotter would be to melt the polymer (assuming that it is solid) in order to speed up equilibrium.

The main problem is that solid samples take several hours to equilibrate with the headspace. Your samples need to be very finely divided, or dissolved in a high MW solvent that will not interfere with the headspace analysis.

If by phase ratio you mean how much of the vial is filled with sample (and not the phase ratio of the column), you have to compromise; the more sample that you put in the longer it takes to equilibriate, the less you put in the lower the headspace concentration of analyte. Try about a 5 mm depth in the bottom of a vial to start with. Unfortunately the shakers that are built into headspace smaples do not help with solid samples.

Peter

[dcav]

Hi! i have been involved in HS-GC analisys of volatile in complex matrix, so i understand perfectly your doubts!

Have you done some progress?
What polymer is it if you can tell me? In my case ( it was static headspace of hexan in cocoa powder) the best solution, was to get a solubilization of the matrix. You colud solubilize the polymer, if you keep the solution well diluted. Then mixing would enhance diffusion of analytes off from the polymer coil to the solvent and partition between it and headspace wouild be easyer and quicker.
Hope to help you some way...

[chromatographer1]

I pretty much agree with Peter but capillary can be used if you use a 0.53mm ID column and a flow restrictor. Spitter use will increase your RSD values by about 1% but if that is no problem, great.

80°C is adequate. Try to find a solvent to dissolve the polymer if possible. Your results will be more accurate and reproducible.

Peter's remarks about solid samples are excellent. Also note that larger samples will also give larger RSD values but this is not due to equilibrium issues in many cases.

Find out if you require 3% RSD or 20% RSD results and if true values are more important than just getting an answer. Dissolve your solid if you want true answers. Although that might not be necessary it often is.

best wishes,

Rod