EP Cholesterol GC Assay

[shaun78]

Does anyone around here have experience with the GC procedure for cholesterol in the EP?

I have gotten word that our raw material qualification group simply can not get this assay to work. So, I took a look at it. It seems no matter what I do, I either consistently fail peak area ratio %RSD, peak tailing, or a combination of both. The peak area of the ISTD is consistent from injection to injection, the peak area of the cholesterol is not.

The method is being run exactly as stated in the EP and we are using a Restex Rtx-1 column (though I have also tried with a DB-1 and what ever Supelco's version of the XX-1 columns are. All produced similar results). I have tried the assay using a regular split injection linter, a focusing injection liner, siltek coated split injection liners; liners packed with silanized glass wool as well as quartz wool; modification of injection port temperature by 10% to increase reproducibility.

As you can probably guess I am operating under the assumption that, due to the fact that only the cholesterol peak area is changing, the cholesterol must be in some way reacting with the liner.

Any ideas as to how to fix the problem would be greatly appreciated!

[Consumer Products Guy]

I'm not familiar with the EP cholestrol procedure, but when we evaluated official methods years ago, we decided that there was room for improvement. Yes, we retained use of cholestane internal standard, added it early, and used it for quantitation, unlike an official method that added it after extractions. We also made trimethylsilyl derivatives of the cholesterol (cholestane remains the same) which improved its chromatography. We did not try to do industry round-robin or publish because of that not being pursued.

[zokitano]

Dear Shaun,

I don't have the EP at this moment, but I searched the EDQM Knowledge Database and I found a chromatogram available on-line for the Section (2.4.32)(Total cholesterol in oils rich in omega-3 acids):

http://extranet.edqm.eu/4DLink1/4DCGI/W ... mono/20432

and

http://extranet.edqm.eu/4DLink1/pdfs/ch ... /20432.pdf

Hope this helps

Regards

[shaun78]

TMS derivatives is an interesting approach. I am unsure if I could get the quality folks here to buy off on taking that approach .... but at the same time the monograph does not say that I can't do that either.

Unfortunately, as far as the EDQM is concerned the chromatograms that they have listed for various methods were not run using the same instrumental parameters or column as the cholesterol monograph. Though those chromatograms appear much better than what I am getting.

I guess I should have also stated in my original post that I must follow the EP monograph explicitly due to the fact that we are performing raw material testing on a component that will be used in the manufacturing of a product that is to be sold in Europe. Aside from having a little bit of room where the monograph does not give explicit instructions/detail, I am pretty much stuck.

[Bruce Hamilton]

The EP method relies on an inert injector/column system. and uses Pregnenolone isobutyrate as the internal standard, unless the method has changed.

Nearly two decades ago I used to analyse Cholesterol and TAGS together underivatised using cholesteryl acetate as the Internal standard, so the issues are probably similar. The ester forms are much less sensitive to injector and column activity.

You may not be giving the system time to stabilise before you try the next change. Condition the column at recommended max. temperature ( say 300-320C ) for several hours, preferably overnight, with the injector at recommended temp (285C ). Set oven at the initial temp ( 275 ) and start injections of standards and blanks.

Take 1 mg/ml standard solution ( with Int std ) and keep injecting it. You should find the cholesterol area and retention time will slowly stabilise after increasing peak area for several injections, whereas the Int std will be more stable in area.

I've always attributed the slow stability to removal of activity via the multiple injections. Once it's stable I'd keep injecting until all analyses were performed, not letting anything cool.

If the peak areas don't increase and stabilisie, you have too much activity in your injector/column. Check quality of carrier gas, must be very low in Oxygen. Try heating injector for several hours then injecting larger volumes of solvent to rinse injector.

Please keep having fun,

Bruce Hamilton

[shaun78]

No the method has not changed much since you ran it.

Thank you very much for the insight. I'll get working on that today and will post back to let everyone know how things turn out for me.

Thanks again.

Shaun

[shaun78]

The raw materials group was able to get the assay to pass after following the instructions Bruce Hamilton posted.

The column was conditioned over night at 315C. The next morning after about 10 standard injections peak area had normalized. However, after those 10 injections the peak summetry also had fallen from about 1.0 to 0.8, which still does meet the requirments of the assay. Peak symmetry seems to have held at 0.8 throughout the run. Is this something that we should also expect to see as well?

Thanks again for the method pointers!

[Bruce Hamilton]

Glad you were able to get a usable result without any staff suicides.

I'd normally expect the peak shape to improve with more injections, so I'd suspect there was still some activity being introduced or residual in the instrument/column.

Peak shape can introduce a whole compost heap of worms, so my comments are just a guide. I'd always start with the injector, especially the split liner. I used a packed one in the 5890, and ensured the surfaced were thoroughly cleaned and deactivated.

Then I'd condition the column, but if that didn't work after an overnight run, I'd consider injecting larger volumes of solvent through at low temp and try a sequence of stds again.

I tended to favour using narrow bore 0.20mm ID and thin film columns where possible, just to keep peaks sharp and columns at lower temperatures.

I hope this helps,

Please keep having fun,

Bruce Hamilton