Air leak problem

[jkloepping]

I have an odd problem tracking down an air leak in my 5890/5972. During a run, about every 15 minutes I get a brief spike that looks like an air leak. I've tried isolating the MS first by capping the transfer line. No more spikes. After reconnecting with a new ferrule, I checked for leaks with dust-off by scanning ion 51, and everything inside the oven seemed fine, so I did another run, and still had peaks. So, I disassembled the inlet completely and replaced everything from septum, liner, seal & etc. On my next run, the peaks are back.

I did notice something strange. This problem started when I switched to a new SGE column, and looking at it carefully, I see that the polyimide coating on the outside of the column seems to be in poor shave with pitting. Could a column flaw cause this problem?

[Rndirk]

I have an odd problem tracking down an air leak in my 5890/5972. During a run, about every 15 minutes I get a brief spike that looks like an air leak.

What are you detecting? Are you referring to a peak in column bleed in full MS scan?

Could a column flaw cause this problem?

Yes. Do you have an electronic leak detector?

[jkloepping]

I'm detecting mostly ion 18 but also 28 and 22. Background peaks have counts around 1000 or so (a couple of hours after pumping down). Then I will get peaks with counts of 50k or higher with 18 as the dominant ion.

Yes I have an electronic leak detector, and I know it works. I haven't found any leaks with it.

[Rndirk]

So you get peak at m/z 18 when you do an injection measuring in full scan from m/z 10.

If your air/water check is OK before/after a run, i wouldn't see this as a big deal. To be honest, I never record from m/z 10 in full scan. Do you have indications that it affects the performance of your instrument?

Maybe some water is stuck somewhere around your inlet or a column connection. If your system is oxygen free, you could do a sort of bakeout overnight: inlet and oven at 250°C for example, with a good flow of carrier gas.

[jkloepping]

It doesn't affect the analysis, but I like to control all the variables and not knowing the source of this problem is very irritating. I'm going to switch columns back and see if it is a defect in the new column.

[James_Ball]

I sometimes have a similar problem with a 5975 and a 5973 both running purge and trap. Scanning from 35-350m/z there will be random spikes on the chromatogram that show 40 and 44 masses like there is a momentary air leak. These appear at random retention times maybe one to three times during a 20minute run. I have never found the source. No signs of leaks. The random retention time is what leads me to think that something before the column is outgassing air at random times.

I first thought it was related to the rotation of the valve in the purge and trap, but the moving retention time doesn't correspond to any of the valve switches. After a few days it will disappear, maybe for a few weeks then return again. Sounds like we have similar problems, I hope one of us can figure it out.

[Bigbear]

Running 524.2 I see 44 for the first 4 min ( desorb time), I believe it is from disolved CO2 in the sample. From this system's autotunes there is <2% air and water.