I am analyzing for gaseous halocarbons in a tedlar bag by FID on a 100m x .53 x 3.00 DB624. For simplification I will only discuss ethyl chloride (EC). I am calculating the EC result based on area count using a 1:1 ratio of sample vs. standard. I determined my response for EC two ways. First method: I shot (manual direct injection) EC @99% neat in 5uL and 50uL injections while using the instrument to range the result to be on scale. Second method: I diluted 99% EC (same as above) to 150ppm and 50ppm working standards and also introduced them by manual direct injection.
When comparing the area count of the first method with the 5uL and the 50uL injections they where within 5% of each other (numerically speaking), both yielding good chromatography (no deformities in the peak shape). When comparing the 50ppm and 150ppm standards they also were within 5% of each other with good chromatography.
When calculating a sample by the area count determined by the second method the yielded result was about a 33% lower than when calculated by the first method.
I obtained two independent EC standards at 5ppm and at 20ppm both from a different source. Shot them both and compared them with the same sample as above. The result agreed with my second method results.
I have determined that both methods gave a linear response with an R squared value of .995 or higher when the injection volume is changed (5, 10, 20, 50, 100uL) all with good chromatography.
My question: I believe that the second method to be the correct value for EC, but why is this correct? I don’t understand why there is such a difference between both methods even though both methods gave no indications (chromatography speaking) of error.

Thanks for taking the time to read this and for your input!!!!


Jeff

The assumption you have made is that the integration will be accurate for the concentrated samples (neat) and the diluted samples (ppm).

While you may have the integration parameters set for accurate measurement given an useable dynamic range for each it is possible that with the broader and more concentrated peak the numbers are skewed toward another slope of calibration.

To the original poster,
You don't say what mass range you think you are injecting in the diluted samples and what type/volume syringes are used. Even if linearity was maintained over many decades (and it may not be as pointed out by anon above), some gas syringes blow back more than others during manual injection, which is why an in-line gas sample loop is better if you have one.

To Anon 1:37

The type of syringe is a gastight, but I tried numerous capacity syringes and types to if something would change to no avail. I see your point, but I would expect different responses for different syringes (non gastight) that I did not see. A sample loop is a better technique, but I think it would yield similar results.

Thanks for the replies,
Jeff