FID vs. TCD

[Labgurl]

Another lab is using an FID, Split/Spliless, capilary column, using an ISTD method to test the exact same sample as we are and obtain a total of about 2.5%volume light end hydrocarbon content.

We use a 7890A, TCD, packed column, straight sample injection (external standard for calibration purposes.) On the same sample we get about 5.5% volume light end hyrdrocarbon content. The instrument is calibrated using a volume% certified standard.

Our 5.5%volume is matching closer to the other labs Mol% results. Is there any way to confirm on Chemstation that our results are indeed in volume%?

[Bruce Hamilton]

If you have correctly calibrated your standards and samples, you have already confirmed your results in whatever units you chose.

% volume suggests your samples may be pressurised liquids. If so, I would consider valve sampling, and would be very wary of discrimination in a split splitless system, and would focus on careful sample handling to avoid loss of volatiles.

If you're using headspace, I'd be trying to ensure standards and samples are closely matched in composition.

[Labgurl]

Thank you for the reply. Are you aware of any liquid standards that I may be able to use for the testing of hydrocarbons? (methane, ethane, propane, butane, & pentane?) Right now the standard we use is a pressurized liquid in a pitson cylinder.

[Labgurl]

Another question...what causes concern with the split/splitless? I'm not too familiar with it.

[Bruce Hamilton]

You only really have the choice of gaseous or compressed liquid sample if you have methane present.

I certainly would be using pressuried valve sampling if you are trying to quantify them at % concentrations, and probably even at ppm concentrations.

The concern with any sampling that doesn't load a complete and representative sample onto the column is that there is a selection process occurring in the injector or sampling system.

In the case of split/splitless the more volatile components ( eg methane or permanent gases ) may be favoured over the less volatile ( eg pentane ).

It's possible to use split/splitless injectors, but you have to watch for discrimination, usually by analysing standards of differing composition at the extremes of your product range.

Any calibration is dependent on the split system behaviour remaining constant for all samples and standards - a big ask, especially if sample volatility is different to standards.

The piston is designed to ensure there is no headspace, because headspace of such a pressurised liquid will have a different composition to the sample liquid.

For pressurised liquid sampling, high pressure, low volume, liquid sampling valves are commonly used to ensure the sample pressure and composition is maintained into the injector.

Bruce Hamilton

[Labgurl]

Thank you for the excellent explanation Bruce! I understand