peak area irreproducibility (sample by loop and valve)

[abcdshi2001]

GC HP6890

Note: sampling by loop and valve, not the case of manual injection.

For a given composition of a gas mixture (e.g. cyclohexane in hydrogen, flow rate precisely controlled), peak area for cyclohexane did not reproduce during sequential runs.
To be specific, the area remained kinda stable within an hour or so, but if runs of longer duration were executed, gradual increase in peak area were observed (from 3700 to 6400 after 12 h!!).

In addition, retention time, peak width did not change.

pressure (ambient), temperature (set in the Aux. option) in the sampling loop should remain stable, therefore resulting in a constant amount of hydrocarbon in the sampling loop.

What's wrong with my GC?

[chromatographer1]

Assuming you are not splitting the sample and you allow adequate time to fill the loop and enough time to empty the loop there is nothing wrong with your GC. There is something wrong in how you are bringing the sample to your valve.

If your sample has condensed (liquid) cyclehexane in its cylinder and the pressure of the cylinder is reduced with the withdrawl of gas sample the molar amount of cyclohexane per gas volume will increase as the pressure decreases in the cylinder.

But I have made assumptions from the limited amount of information given.

Good luck,

ROdney George

[AICMM]

Perm tube or cylinder? I would also suspect the transfer lines bringing the sample to the valve injector if it is cyclohexane, heavy enough for it to condense in the lines if not careful and, thus, cause creeping baselines.

Best regards.