Internal Standard FID response not stable

[BrianDozd]

I am using an FID to detect the hydrocarbons in a few reference sample to verify that the internal standard I want to use with some unknowns is working. The internal standard I am looking to use is 2-Heptene, because I can clearly distinquish both cis- and trans- peaks, and they do not co-elude with any expected compounds.

However, there is a problem. The FID is producing vastly different areas when I add the same amount of standard to the references, depending upon the reference sample used. One reference sample produces about 900 pA s of area, another reference about 800 pA s, and a third 1100 pA s. For each of those, about 1 g of sample and 10 mg of the internal standard was used, and the cis-trans ratio came out as expected, as did the proportions of each compound in relation to the internal standard, which is how the problem wasn't noticed for so long.

Does anyone have any idea why the FID would be responding differently to the same amount of the standard?

[Peter Apps]

Since you give no clue at all what it is that you are analysing, or how, I can only guess that the FID is not in fact seeing the same quantity of IS each time because there is variability in the sample introduction.

Peter

[AICMM]

BrianDozd,

You are shooting 10K ppm and only getting about 800 pa of signal. As a reference point, Agilent check out standard is 400 pA for ~ 20 ppm standard splitless/split. So, either you have a huge split or you have an FID that is not working as expected. IF the former, try reducing your split ratio, if the latter, you have other issues.

Best regards,

AICMM

[karlssa2]

The problem should not be int the FID. I think it is more likley in the inlet, and some problem with transfering the analytes to the column. Try split or splitless, and check if you get more reproduceable responses. You can also try to change type of liner. A split liner with glass wool can sometimes give much better results in splitless mode. An increase in temperature of inlet is also worth checking.