Tircky HS analysis of residual solvents

[Rolandas Plausinaitis]

Hello,

We have compound (can not mention here) which is not soluble in most of the solvents. So during production "cocktails" of different solvents are used. Which means we have to check 8 different solvents of all kinds (aromatics, THF, chlorinated, DMSO). Which leaves basicaly no solvent left to dissolve the compound. Water could be used, but DMSO can not be measured this way....

So now to the questions:
1) I can melt the compound in the vial and measure the solvents- is that acceptable from HS theory point? What about calibration? Anybody done this kind of analysis?

2) We use CombiPAL sampler for static headspace with the syringe injection. HS heater can go up to 200 degC, syringe heater up to 150 C which suggests we can inject "colder" vapor phase. Is this really the case as usually it is recommended to have syringe hoter compared to the sample. Our sample melts at around 185 C.

Many thanks!

[Peter Apps]

As long as you allow long equilibrium periods (probably several hours) you do not need to have the sample in the liquid phase (i.e. dissolved or molten).

You should never have the sample hotter than the syringe - if you do the sample headspace can condense in the syringe.

Peter

[chromatographer1]

I disagree with Peter concerning residual solvent analysis by headspace with the sample remaining in a solid form, not dissolved or in a molten state.

While it is possible with some materials to achieve reproducible results the results may not be accurate. A huge amount of research may be performed to give data which support the results but still PROOF of the results may not ever be attained. (Consider multiple crystaline forms, affects of other variable solvents in the partition into vapor state, limit of detection, and linearity of recovery) In addition, heating the samples for hours increases factors which may cause inaccurate results. This opinion is based on my experience and the requirement for reproducible and ACCURATE determination of residual solvents for regulatory auditors.

It is more appropriate to analyze the sample by headspace using two or more different dissolution solvents that do give accurate results for the other solvents of interest using Henry's Law.

Peter's other comments are well taken and I am in complete agreement.

Of course if your results are not so critical to your product development or government approval then approximations of solvent content may be suitable.

best wishes,

Rodney George
consultant

[Peter Apps]

Hi Rod

You are right, I should have added "and as long as you can validate recovery and calibration"

I got the impression that the propblem was both a multitude of different residues, and that the sample would not dissolve in anything except water, so that it would not be possible to use a suite of different solutions.

Peter

[rbardsley]

Sir

I am an applications chemist at Teledne Tekmar in Mason Ohio USA (East Coast USA Time Zone). I have been looking for a headspace application the requires temperatures above 230C. Yours sounds like one. Please contact me at rbardsley@teledyne.com with more information about your request.

[Ron]

Since you have a CombiPal you could dissolve the substance in water, put one portion in a headspace vial and a second portion in a 2 mL vial, then do two injections. Use headspace for the majority of the compounds, then change to a liquid syringe for the DMSO. If the concentrations are high enough you could do all by liquid injection, but I prefer headspace for this application if possible.

[krickos]

Hi

Considering the vast mix of solvents including aromatics, I think the best way is first to try some classic headspace sample solvents other than water first.
One also have to consider standard preparations, aromatics like toluene and such usually becomes very tricky and hard to reproduce at times if water is used.

One usually is suggested to use DMI if one looks for residuals of DMSO or DMF. So please consider DMI or DMF as sample solvent. Another possible sample solvent could be DMAA.

Another way to go if sample solvations becomes and issue and water is the only option is to dissolve the sample in water, transfer for example 4,0ml of sample to HS vial then add 1,0ml DMF/DMAA/DMI and for standard vials add 4,0ml water then 1,0ml of standard prepared in DMF/DMAA/DMI. This has several benefits, less interferance from DMF/DMAA/DMI with regard to higher boiling solvents, standards will last longer and easier to reproduce than if prepared in pure water.

And as mentioned if liquid injection works go for that.

[sdegrace]

Do you absolutely have to have just one method? I understand it's twice as much validation work to validate two methods, but on the other hand, it sounds like there might be so much trouble finding a single method for all your solvents that validating two methods sounds like an escape. The extra validation and analysis overhead might be worth avoiding having to do anything whacky with your headspace.

Would it be possible to use say DMSO as the diluent and analyze for the others, then develop a parallel method to analyze for the diluent? Maybe not even using GC.

Stephen