GC Using High Boiling Solvents

[mbicking]

I have been asked to "optimize/fix" a method for some residual solvents (IPA, methylene chloride, toluene) in a pharmaceutical. The solvent is DMAC (dimetylacetamide).

DB-5 30M X 0.53 um X 5 um, 5 mL/min He, Inj Vol. 1 uL, Inj Port 150, Splitless for 0.3 min., Temp ramp starting at 36, FID, etc. The main problem is peak distortion/splitting for IPA (first out) and tailing/shoulder for MC. Toluene is OK, except for an impurity in the solvent.

For the record, this is not my method. It's a poor column choice, the inj port is below the BP of the solvent, and who does splitless on a megabore column?

Since this is a GMP environment, I have limited ability to change settings, but I can justify some. Any ideas? The high boiling solvent really makes this a more difficult issue (at least for me).

[Peter Apps]

Hi Merlin

I'n not sure that you can get where you want to be from where you are.

You are pretty certainly getting some poorly controlled solvent effects due to the combination of a high boiling solvent with low boiling analytes (the column temp is presumably constrained by separating the target analytes ?).

Sneak a retention gap in - the peak distortions are probably due to having a condensed solvent film on top of stationary phase.

Changing almost anything would probably be an improvement - what CAN you alter ?

Peter

[Bruce Hamilton]

I wonder if this is a method that's been transferred from headspace sometime in the dark, distant past..

I agree with Peter, it's hard to make easy substantive improvements. If you have reasonable sensitivity, you could try to reduce injection volume. I suspect changing solvent composition, and many other good things, would be prohibited.

Good luck,

Bruce Hamilton

[chromatographer1]

Bruce and Peter have made good suggestions.

Another possibility is to reduce the speed of the injection if your autosampler will allow that change.

An increase in carrier flow is usually allowed by USP regulations, but I don't remember by what degree. That is another adjustment you could explore.

best wishes,

Rodney George
consultant

[mbicking]

Thanks for the suggestions/comments! I have a few more runs in now and have some possble solutions.

Split injection is always an option, and eliminates the tailing on the first peak. Of course there are the senstivity issues. Oddly enough, raising the inlet to 200 (a more reasonable value) did not make any difference.

Pulsed splitless made the problem worse. However, reducing the purge time from 0.3 to 0.1 min. also eliminated the tailing for the first peak. Increasing the delay made it worse, as expected.

We will now have to discuss whether or not these changes are allowed or not during this method transfer phase.

Other issues are still around. What I thought was splitting of the first (IPA) peak from injection effects turns out to be an impurity somewhere in the lab/solvent/glassware. We are having some trouble finding the source, but will have to look at it more next week. We also see a significant amount of eluting material at 250 C - broad bands that seem to appear and disappear, and are not directly related to the current sample. Anyone know of polymerization or reactivity issues with DMAC?

Thanks again. Appreciate the comments.

[krickos]

Other issues are still around. What I thought was splitting of the first (IPA) peak from injection effects turns out to be an impurity somewhere in the lab/solvent/glassware. We are having some trouble finding the source, but will have to look at it more next week. We also see a significant amount of eluting material at 250 C - broad bands that seem to appear and disappear, and are not directly related to the current sample. Anyone know of polymerization or reactivity issues with DMAC?

Thanks again. Appreciate the comments.

Hi

"Ugly" method you got there. Anyways, issues with impurities in DMAC is not unheard of especially if you run splitless or on low split, we had some issues with impurities coeluting with IPA and other solvents when our supplier suddenly changed something in the DMAC production so we had to switch DMAC supplier.
In some quite old methods we also had a comment about DMAC that you may have to heat it at like 60°C for some time to drive off some more volatile impurities.

How does the blank runs look like?

Reactivity issues: Normally rare in normal headspace injection for residual solvents but with increased temperature the risk is of course higher. I have seen it twice, the case I recall directly was DMAC with an "omeprazole" like API at a headspace oven temp above 50°C, caused "false" methanol peaks.

DMAC may also hydrolyse in precence of acids and halogenated compunds should be avoided if remebering right.

Another idea: Have you discussed with your client/QA the "file" status of this method? USP was recently udpated with regard to residual solvents, you might be able to switch to a better/more gentle general method depeding of regulatory filing.

[mbicking]

Kirckos:

This is a direct injection method, not headspace, which makes all of the issues worse, of course. This is "HPLC" grade solvent, but we all know that doesn't mean the same thing in GC. I have checked and we are not seeing residual acetic acid from hydroysis; did not have dimethyl amine available.

The blanks are OK, except for the one interference with IPA. I suspect it is from an environmental source. Will have to watch things more closely.

I agree this is an "ugly" one. I will spend this week trying to sell the changes as "adjustments for instrument differences."