Quantitation on Flame ionization detector

[LabProARW]

My company runs GC using an FID detector to quantitate weight percent assay of samples. I find that we calibrate our method based upon the assay given by a standard material supplier who gives their certificate of analysis to us as determined by their HPLC area percent method. This doesn’t seem proper in my thinking since the Flame ionization detector differs greatly to a UV HPLC detector. What am I missing that it is accurate to use an assay from a different category detector in order to calculate a weight percent analysis on a different detector?

Thanks.

[Peter Apps]

The result of an analysis must be independent of the method that produced it. Is the composition of the standard expressed as a mass fraction i.e. g/100 g would be a mass percent, or do they report only the area % for HPLC peaks ?

Peter

[LabProARW]

Sigma Aldrich COA reports Area Percent by HPLC analysis. Knowing (or believing that I know) that the response factors for HPLC-UV can vary considerably... the fundamental question is: how much confidence to place in an analysis where I weigh out grams and make an assay correction on that gram weight for the assay based upon an HPLC certificate of analysis. Then proceed to run my analysis on a GC with FID detector.

[Consumer Products Guy]

A decade or two ago, some rocket scientist in the QA department had the idea that the analytical chemistry department should assay EVERY chemical that was to be used in the department and in all of R&D, even assaying stuff like sodium chloride, acetonitrile, etc.

Of course, that would've meant that something, somewhere would need to be trusted to use as a reference. Plus that such analyses would dwarf the real purpose of the department, be technically out of reach, and have unbelievably exhorbitant cost. So that "idea" was quickly put to rest.

[Peter Apps]

Sigma Aldrich COA reports Area Percent by HPLC analysis. Knowing (or believing that I know) that the response factors for HPLC-UV can vary considerably... the fundamental question is: how much confidence to place in an analysis where I weigh out grams and make an assay correction on that gram weight for the assay based upon an HPLC certificate of analysis. Then proceed to run my analysis on a GC with FID detector.

Unless the two substances have the same response factor by HPLC, which is not impossible but should not be assumed, the peak area percent is not the same as the mass fraction percent.

Rod's law applies; the more you don't tell us, the more we can't help you. What substances we are talking about would be a useful start.

[James_Ball]

I see it often in industrial labs that the FID and UV detectors are considered "Universal" detectors.

If you have a phenolic compound and shoot it on FID and just look at the total peak areas versus the area of the known compound and take that as %purity, it is what is accepted but not always accurate. If you had Phenol, and the impurity was Carbon Disulfide, then just looking at the relative peak areas would grossly underestimate the amount of CS2 present since it barely responds to the FID while Phenol responds quite well.

If using UV then the impurities would need to be UV absorbing in the wavelength being used, if not, then those would be underestimated also.

It is not easy to define an all encompassing method to determine all impurities in something.