490 Micro GC - Peaks getting smaller over series of injections


We've recently been having an issue with our 490 Micro GC where peaks are getting smaller of the course of any series of injections.  The GC has 3 channels with TCD detectors.  Shown below is an example of an experiment where the issue first cropped up.  During the experiment, we are flowing a constant flowrate of N2 gas through an airtight three-electrode cell with ~5mM phosphate buffer solution (pH ~ 9) and then into the GC.  So the only gases expected at the GC are N2, H2O vapor, and trace CO2.  Channel 1 has a peak associated with N2 that remained constant over the course of the experiment.  On channel 2 + channel 3, N2 comes out with several other gases, so I have labeled these peaks as "Other" on the graph below.  This peak suddenly grew smaller at ~20hrs until it eventually reached zero intensity, despite the fact that channel 1 continued reading a constant N2 signal (as expected).  

This issue has persisted even when keeping the GC inlet open directly to air.  Shown below are the chromatograms (Channel 2 only) for a series of injections when the GC was open directly to air several weeks after the problem was first noticed.  Similarly, the peak associated with most gases grows smaller with each injection.

Here are some technical details related to the channels for this GC and the method used during the experiment:

Any idea of what could be causing these issues or how to troubleshoot the problem?  Could this be an issue with the pump, the reference gases, or some other GC internals?

Any help is much appreciated.  Thanks,
Mat R.

Was this helpful?