Peak became weird when using GC-MS

[Joaneee313]

Hi everyone,

We are having problems with our peaks recently when we ran our samples. The instrument we are using is Agilent 6890 connected with a 5973 network mass spectrometer detector. Before we were getting good results. Each peak was smooth bell shape but after a power outrage, the peaks right now are weird looking, with the straight line on the top of the peak.


Does anyone have the same problem before? We thought it might be the dwell time problem, however after we tried different dwell time combination, it still didn't work out right. Any idea about this issue? Thanks a lot.

Qiong

[nvalentine]

Are you overloading the detector? Your picture isn't loading, but it sounds like you are flat-topping your peaks. Maybe check that your split ratio is the same as it was before the power outage. Your voltage settings could also affect the signal. You might want to run a new tune and compare the voltage offset with an older tune.

[tlahren]

I agree with nvalentine on the new tune. I have experienced before where power outs will cause the MS to default to factory tune values (temp zones as well as voltages). Even if you try to restore your previous tune ("File" - "Restore previous tune") it doesn't always work. You need to do a couple new auto tunes. The newer 5975 systems seem to be better about restoring previous tunes. It doesn't happen all the time (during ever power out) but it has happened to me before. An Agilent Tech told me that in such event the "Restore Previous Tune" option doesn't always work out best. And like nvalentine said, check the voltages and compare with the new tune especially the multiplier voltage.

[Joaneee313]

https://docs.google.com/leaf?id=0B1z4sS ... y&hl=en_US
Thanks everyone for your reply. So above is the picture and I tried to upload again. Please click the url and it should bring up the image. Hope this works.

Actually we didn't overload samples and I'm pretty sure about it. We even tuned several times and tuning peak also looked terrible as all of them had splits on top. I might now go and check the multiplier voltage.

[tlahren]

Actually it looks "OK" for a chromatographic peak. I can't really think of anything other than increasing your total number of scans (i.e., reducing your dwell time) which you already covered. Ideally usually run with 2 scans/s or more. It also depends on how wide your peak is (which is hard to tell in the picture provided).

Another possibility: Did you possibly have "Peak Smoothing" enabled in Chemstation prior to the power out and it's now somehow "Disabled?" I don't every really use it but it uses an algorithm to make the peaks more aesthetically appealing. I'm not sure if it affects the integration data at all. If this were the case however, the older "good" data files would now look like this. I guess i'm just thinking out loud now.

Sorry I can't be more help.