I am having an ongoing fight with baseline problems on my Varian 3800 GC. We use ECD's and do PCB analysis in waste fuels. The problem is that almost all of the baselines take a dive below the zero line immediately after the solvent peak at the beginning of the run. Resolution is not a real problem, as the individual peaks of the Aroclors come out fine, but the baselines are all over the place. I have checked for oxygen leaks and carrier gas leaks and have found none. Any suggestions would be appreciated. I am fairly new to GC (less than 2 years) and this problem is really baffling me.
By John Hinshaw on Tuesday, December 7, 1999 - 07:53 am:
Your description points to a contamination problem.
What is the base frequency of the detector? It should be about 100-200 Hz (see the users' manual to get their exact recommendation and to learn how to measure this). A high base frequency implies that the detector is contaminated.
Even though you have checked for leaks, how good is the carrier gas going in? Do you have oxygen and hydrocarbon filters in line? Are you using a high-purity dual stage regulator at the tank? Is the gas in the tank rated for ECD use? And, how did you leak-check? Soap bubbles won't cut it here: you need an electonic device.
Is the baseline excessively noisy and does it wander when the instrument is standing by between runs?
How hot is the ECD running? It should be at 300 °C or higher to avoid trapping high-boiling eluants and column bleed. You might benefit from running the ECD with makeup gas only (see next paragraph) overnight at a higher temperature.
A good test for an ECD is to disconnect the column and close the column--detector connection with a blank ferrule. Keep the makeup gas on, and see if the baseline improves. Make a blank run without the column and compare the baseline with one with the column. This may point you in the right direction.
Finally, recheck your sample prep. I see from your other posting here that you are using a cleanup procedure to remove the hydrocarbons from your sample. If you have an FID GC system available, you could evaluate the hydrocarbon content of your injected sample by running the same GC method with an FID instead of the ECD.
Posting is currently disabled in this topic. Contact your discussion moderator for more information.