I have 6 GC's operating with the same method/samples - and 4 of them are having problems with quantitation of peak trapping methods (ie. trap oxygen/nitrogen - send to MSieve) while rest of sample goes to Plot Q column for analysis. Two TCD's -
Any suggestions as to where to look for this problem? Even with two gases only though my gas sampling valve I get terrible peak ratio reproducibilities for a very easy separation.
By Anonymous on Friday, April 4, 2003 - 02:40 pm:
Could you possibly have "bad" carrier gas?
I once spent 8 hours of maintenance work on my ion trap MS, only to find the problem was a bad tank of He.
Just a thought.
By Les on Friday, April 4, 2003 - 06:24 pm:
My only thought would be pressuring up the sampling loop, injecting varying volumes of sample to the system. Does your He have traps? Could they be spent? Splitter flow problems?
By Rodney George on Saturday, April 5, 2003 - 05:46 pm:
There are several possible problems. Eliminate them one at a time.
Valve leak: inject carrier gas only watch for peaks from leaks
Valve timing: are other peaks in sample consistent? like CO2 and ethane methane? If they are consistent then the sampling or std is not the problem. However the timing of the diversion of the nitrogen-oxygen-methane peak may be inaccurate.
Valve sticking: does the valve actually react to the switch command consistently
Integration of peaks: how much separation do you really have. Can you monitor the peaks visually? Can you see differences in peaks when plotted which vary in accordance with the integration results?
I hope this helps.
Senior Research and Development Scientist
Supelco Gas Separations Research
By Athar Wasim on Thursday, May 29, 2003 - 02:08 am:
MSieve are badly effected by moisture, in one application we found H2 (carrier) bottle contained 40 PPM water contents (measured with portable) was a reason for contineous N2/O2 peak shift. We put moisture trap in series of H2 supply and it solved the problem.