7000A MSD: Grass or zero abundance

[cb23]

Dear Chromforum community,

I have been having trouble with our Agilent MSD 7000A. During runs I get "okay-ish" chromatogram, but zero abundance between the peaks. At times, we get grass, i.e. every m/z at around 200 counts in the entire scan range. The peaks consist of the expected main m/z, i.e. 74, 87 and 143 when running a FAME standard mix.

Autotune completes, but not very nice. Tune report.pdf
The full auto tune incl all ramps is here for download

Good example: http://de.tinypic.com/r/fc49kp/9
Grass example:

To tackle the problem, I performed a full GC maintenance (to rule out GC issues before). Regarding the MS I cleaned ion source (twice), installed two brand new filaments, replaced both repeller insulators and the quad cap. Wiped the outside of the EM horn with MeOH, dusted off the outside of HED, EM horn and ion source with helium. EM horn was replaced about 6 months ago.

Actually, all of the above have helped a lot as the chromatogrammes were really hideous before and are okay-ish now, but I still get grass and zero abundance. Occasionally the tune fails to adjust the gain on filament 2. Also, the repeller voltage and EMV is not constant between tunes.

I suspect that I have not removed all dirt from the HED or EM horn.

* Can one actually solvent clean the triple axis HED of the 7000A? So far, I have only seen the tutorial on the HED of the 59xx series.

* If the HED was broken or dirty, why does the autotune complete?

* Somebody suggested to sonificate the EM horn in iso-propanol. This sounds very counterintuitive. Can I do it?

* Are the ion sources of the 7000A and the 5977B compatible? Since our single quad’s been running fine, I could rule out the 7000’s ion source by swapping them (also EM horn will be swapped for testing)

* What does the dark current reading actually tell me?
Could some lint in the HED be causing the high dark current reading?


Thank you and all the best,
cb23

[James_Ball]

I had a similar problem with our 7000C when we first got it. Agilent sent me a new HED/EM assembly and the problem went away. It returned again and a third assembly fixed it. I believe we were having some bad bleed from a column that was bad, but can't be certain. I have never cleaned the HED on one of these but it should be possible I believe the one on these is more of a bar shape than the button of the 5973/75.

Running hydrogen carrier for a while also seemed to clean it up quite a bit. I am not certain, but I think the source is the same between the 7000 and the 5977, if you take a quick look at each you should be able to tell if they are. If they are compatible then swap them and see if the problem remains. If it does, then it is probably a dirty HED.

If the collision cell has picked up some contamination, then running hydrogen carrier for a week or so could also help clean that out too.

[cb23]

Thank you James, I think tomorrow I am going call Agilent.

At this point, I have no longer any idea how to fix this rather erratic issue. Yesterday I vented the system and used helium to de-dust the EM horn, HED, second quad another time... now I am again getting hideous chromatograms and weird tunes like four weeks ago (e.g. repeller voltage sometimes at 9V, sometimes at 36V, also the same for EMV). The brand new filament 2 does not seem to autotune anymore (it worked a few days ago) - even though I did not touch the ion source....

Unfortunately, H2 is not an option since we don't have a H2 detection system in that room.....

[James_Ball]

One thing I did while waiting for a replacement HED was to raise the threshold just enough to keep the system from reading the noise. If that doesn't make it tune properly then it could be something electronic causing the problem.

[cb23]

The threshold did not do the trick, unfortunately...

Anyway, I called Agilent. They suggested to clean the dynode.

So to answer my own question, yes you can clean the dynode not just on the 59xx series, but also on the 7000 series. Just undo the two Torx screws (the two neighbouring ones) and the connection to the analyser door. Then you can slide it out and clean it. Basically, it looks like a tiny mirror. I was told to use 2-propanol and Kimwipes.

Also make sure that the O-rings are super clean, apparently is possible that the smallest leak (even not visible in the air water check) causes noise.

Let's see if that helps...

If not, Agilent will replace the HED and its power supply...

[James_Ball]

Once you put it all back together and it pumps down overnight, take a can of Dustoff computer duster and spray around every seal while scanning from 10-150m/z in the Manual Tune window. If you see a sudden spike in masses around 55-90m/z depending on the brand of duster, you will know you just passed by a tiny leak.

Even a tiny piece of lint on the O-ring around the analyzer can cause a leak that will be a problem.

[cb23]

Thanks James. One question though, does that also work with the rear analyser door (no ionisation)?

For other users with that noise issue, I am going to post some more updates...

Turns out that cleaning the dynode and carefully checking the O-ring for lint has made the noise go away. Make sure to check the dark current as it may indicate a dirty dynode. My dark current reading is now:

Detector Dark Current Check: Baseline 47, Threshold 34, HED On Pulse Count 9, HED Off Pulse Count 0

before it was

Detector Dark Current Check: Baseline 47, Threshold 34, HED On Pulse Count 11030, HED Off Pulse Count 0.

Really nice, indeed. Now during a run, the noise looks "normal" again and I have a baseline again.

However, the chromatograms still look a bit messy, i.e. split peaks, etc.
The base peak of the most recent tune was 219 and all ratios looked off. I will investigate and tune again, after checking for leaks (AWC looked fine, though).

[James_Ball]

The analyzer is open from one end to the other inside so anything leaking in at the rear will make its way to the source.

Glad you got it to work by cleaning the dynode. Seems these units are very sensitive to any contamination there.

[Rndirk]

Thanks for giving updates cb23.

We've had a 7000C for 3 years now and no similar issues (yet!). We're buying another unit soon so i'll keep a link to this topic for future reference.