DMSO for GC-HS

[GT]

Hi All,
We have a problem with DMSO used as diluent for GC-HS analysis of residual solvents in Active Pharmaceutical Ingredients.
We have already tested two different DMSO suppliers but in all case we have a lot of interfering peaks.
We use GC-HS Agilent 6890/G1888 with a DB-Wax column with FID. To reach the desired sensibility we have to set split flow between 3:1 to 10:1 and in some cases we have to add water to sample solution to increase the vaporizzation process.
Could you suggest us a good DMSO supplier for this kind of application or a procedure to purified in lab commercial DMSO?
I thank you in advance

[chromatographer1]

Distillation under a nitrogen blanket or simply heating an amount with an inert gas blanket will remove any volatiles from you DMSO. Methyl and dimethyl amine are often found in DMSO especially if water is a contamination.

Purging the liquid with helium when warmed will also remove volatiles.

Use HS to determine the effectiveness of any procedure. Store the cleaned DMSO in small containers with an inert gas headspace and you should be fine.

Best wishes,

Rod

[chromatographer1]

Be aware also that if water is present in the sample being dissolved or in the DMSO you are using, adding heat to the DMSO will produce small artifactual peaks (DMS, DMDS, etc) one of which will closely elute to methanol on many columns.

best wishes,

Rod

[kalidassa]

Hi we have done lot of HSGC Method by using Spectrochem DMSO once try with Spectrochem.

[Charles A. Burger]

Be aware also that if water is present in the sample being dissolved or in the DMSO you are using, adding heat to the DMSO will produce small artifactual peaks (DMS, DMDS, etc) one of which will closely elute to methanol on many columns.

best wishes,

Rod

Hi Rod been looking at some older posts about this because I have run into a similar issue. I went way back and found this post http://www.lcresources.com/discus/messa ... /4161.html from 2003.

Now I'm using the direct injection technique when shooting my 0.5µL plug. Methanol is one of the solvents I'm looking for and when trying to determine my LOQ I noticed that the Methanol peak was larger than it should have been.

So I injected my diluent numerous times with the same outcome. A peak that appears to be methanol. Then I shot DMSO neat and nothing appeared. Jumped back to the diluent, and there it was again. Since I normally don't see many injector parameters at 180°C, I started to suspect that something had to be happening with the mix of water and DMSO.

So after reading what I have, my next question is, with my water so high in the diluent, would distilling the DMSO solve my problem or am I just wasting my time? I've been trying to investigate how the solvent Mfg's are producing the grades they sell, but most are just redistributing what somebody else made for them and they either can't or won't share if it's distilled.

Also, Kalidassa, I tried to find the Spectrochem DMSO you mentioned, but was not sure about what I had found. Is there a link to the company?

Chuck

[krickos]

Hi

We have had a similar issue with DMAA with a hydrate drug substance quite low on residual moist and no water in diluent. Apart from water content and temp/time in HS oven we also noted a pH dependence. With drug substance samples with pH at or just below 10 the issue was less or none but at high 10s larger peaks showed up in sample, luckely in our case the peaks did not interfer at least.

So pH/pKa of your sample might be an additional factor.

My gut feeling is that chasing the purest DMSO will not help as you add water to diluent.

[chromatographer1]

If your levels are quite low then yes, your methanol peak will appear larger than it really is.

The peak may be a thiol or mercaptan, which coelutes with methanol.

Even the best DMSO will show the artifact if your sample contains water or your lab air is very humid.

Yes, pH is also a factor in the reaction.

Extremely low levels of methanol are usually not required to be accurate. If a 10 ppm peak becomes a 25 ppm peak of methanol, usually, no one cares about those levels.

Good luck,

Rod