GC/Micro ECD ghost peaks

[ToxChem]

I run a 6890 with dual micro ECD detectors. I am having problems with 3 ghost peaks showing up. They come and go meaning I can inject the same sample and they will be in one run but not the next. They are also at the same retention times and occur on both the front and back. They are occasionaly in a blank run directly after a sample has been run. I can run mulitple blank runs and they will all come out normal without extra peaks. They are not in solvent blanks. They are not in all samples or in multiple injections of the same sample. Any ideas?

[jdezeeuw]

The only thing I can advise: Systematically check the system by changing one thing at the time.
ECD is very sensitive, but it has to be electronegative. Ghost peaks can come from many sources : septum, O-Ring, Vial septum, syringe contamination, solvent, gas lines, split lines;

First without injections, run several programs; also increase the "cold" time between the programs and see what happens;
Do te same with a Cold-injection port;

Then inject only an empty needle,(no solvent) see if something is happening. It may be septum; also here do the same with a cold injection port;

do the same with solvent only.

Check the split lines, they may be contaminated. If you use splitless, components can be extracted from several areas.

[thohry]

Probably, you ve got the backflush problem.