Benzene peak shape issues - residual solvent method help

[jlsackett]

Good day all,

I am working to develop/validate a direct injection method for residual benzene content within pharmaceutical intermediates. The test article is dissolved in DMAC, run on RTX-624 (30m x 0.32mm x 1.8µm). In short, I have found that my peak shape suffers severely as split is stepped down to a split-less configuration or larger injection volumes are made.

Initially I believed there was a second compound co-eluting with the benzene, as the peak did not present typical tailing shape. Rather, as the peak dropped off towards baseline, approximately half way down it would 'bludge' out and tail off.

A split of 1:10 does not demonstrate the peak shape (approx USP Tailing 1.3), but moving down towards split-less it appears.

My attempts at resolving this issue to this point have not provided an acceptable outcome. Too many parameter adjustments (pulsed splitless, temperature settings, injection volume) have been evaluated to include all here, initially. I believe that the larger injection volume or split-less mode are not producing a narrow enough band on the column.

Does anyone have insight on trace benzene analysis WRT peak shape?

I very much realize that without my parameters the input may be limited... please let me know if I can provide more details.

Thanks in advance for any replies!!

Cheers,
Jeremy

[chromatographer1]

Perhaps you should get an injection liner that permits a connection to a megabore capillary column DIRECTLY.

Then you will find that with frequent liner changes you will be able achieve very low levels of benzene analysis.

With your present dissolution solvent you are seeing the problems that attempting to perform low level analysis with a capillary column can produce. You are flooding the column with DMAc which acts like a phase on the column surface and distorts the 'plug' of benzene which never forms a symmetrical plug.

You can perform headspace, or you can get a packed column or a directly connected megabore column and avoid these problems by injecting the sample directly. onto the column.

Headspace can measure 100 ppb levels and SPME can reach ppt levels for benzene. A hundred dollars worth of SPME syringe and fibers may perform vial headspace for benzene at levels that will meet your needs.

best wishes,

Rod

[krickos]

Agree with Rod.



You could try CS2 as solvent for direct injection in splitless mode, bp of 46°C and "no" response in FID. If done right you should get a solvent focusing effect and better peak shape.

[chromatographer1]

That is a good suggestion to try.

Rod

[jlsackett]

Thank you both for your input on this matter!! I apologize for not getting back to the board sooner to express my appreciation and give a status update... I was able to implement HS methodology very successfully in this case. Achieving both excellent sensitivity and peak shape with essentially little change to the procedure (except of course headspace sampling )

Anyhow, thanks again for your input on my first post to this forum. This is an excellent resource that I anticipate continued use. I look forward to reading through other posts as time allows and hopefully I will be able to give back some information to others in the future!

Cheers,
Jeremy