Need help with CO2 analysis using a TCD

[gurjotbmc]

I'm using an old, local make of Thermo Scientific GC (Chemito 8610) with a SS Porapack Q column (3m) and a TCD to measure CO2 from lake water samples using headspace equilibration technique. The expected range of CO2 values is 100-5000 ppm. I tried using methanator/FID but was getting erratic values at higher concentrations. Was asked to switch to a TCD. Have been trying to optimize the TCD with different combinations of oven, detector, injector (packed) temperatures and carrier (ultrapure helium) flow rates, but all I'm getting is a beautiful and huge peak for nitrogen (my calibration standards are CO2-N2 mix) followed by a tiny peak (that too only in some cases) which I'm not sure belongs to CO2 or not. Also, the area of the tiny peak does show any linearity with increasing concentrations of CO2. I'm using four standards - 100 ppm, 1000 ppm, 2500 ppm, 5000 ppm. So I badly need help to find out what I'm doing wrong and how best to optimize this method. I'm new to this (as is probably evident from my description) so kindly be gentle. Thanks.

[thohry]

What gas is used for carrier gas ? You can change to N2 as carrier and the huge peak will disappear.

[AICMM]

gurjotbmc,

You are probably near the detection limit for your TCD (depending on your loop size) at the 100 ppm level but the other levels should start to be linear. Next question I would ask is how you are sampling and what column you are using.

I would not advise switching to nitrogen as a carrier because you will loose a ton of sensitivity, to the point where you might not see even the 1000 standard.

Finally, have you considered HID?

Best regards,

AICMM