"Grass"

[Darber99]

Upon changing the column in my 7890/5975 GC/MS, I've been seeing alot of "grass" in my scans and my chromatograms. I installed a Phenomenex ZB-1 (60m x 0.32mm x 1.0um) to do method development for TO-15 air testing and conditioned it at it's max temperature (340'C) for 2 hours into the atmosphere. Upon injection of BFB and a Volatiles check standard, I've been noticing alot of detector noise at different times in my run. My vacuum pump (Pfeiffer Duo 2.5) is also running very high (1.4x10-4) and there are bubbles forming in the oil through the watch glass. I changed the oil in the pump and inspected all the connections and O-rings and it still appears to have a leak. Would a vacuum problem cause the detector noise that I'm seeing? Where else on the rough pump could there be a leak if all my connections and O-rings seem fine? My instrument conditions are that of a regular Volatiles run. Split injection, 3.0mL/min flow, starting pressure of ~18.0psi, Injector 225'C, transfer line temp 280'C.

[rb6banjo]

That pressure is too high. I think you have a substantial air leak. Is 28 your base peak in your tune?

[James_Ball]

That pressure is too high. I think you have a substantial air leak. Is 28 your base peak in your tune?

Do you have the performance turbo pump on the 5975? Even those are rated at optimal flow of 2ml/min with a max of 4ml/min in CI mode.

The grass in the spectra is from the mass filter being overloaded probably by nitrogen or water. You should keep your analyzer pressure below 5x10^-5torr for good results.

Is this done direct on column with a cryofocusing module or through the split interface like regular purge and trap is done now? If not cryofocusing by dropping to -50c initial temp then I would look as switching to a .25 or .18 column with a flow into the analyzer less than 1ml/min for better sensitivity and resolution.

[Darber99]

@rb6banjo-No, when I have PFTBA open, 69 is my base peak (~600,000 abundance) water is next @ about 50% followed by about 20% N2. I let it pump down overnite and it reached 1.09x10-4, nowhere near where I'd like it to be.

@James-I'm pretty sure there's a turbo pump attached since it sounds like a jet engine taking off during pump down. I only increased my flow and pressure b/c that's what was recommended for a 60m column with a 0.32mm ID. Anything less than that and my peaks turn into blobs and I lose resolution. We really haven't gotten to injecting air samples yet. I'm doing basic manual injections of BFB and a VOC check just to make sure the GC/MS is in good working order. We'll be adding the liquid nitrogen and pre-concentrator as soon as we get the bugs worked out.

So am I looking at a leaky pump b/c of the air bubbles forming through the sight glass or could those be coming from anywhere in my system?

[Steve Reimer]

The simplest test would be to switch the foreline pump out for a spare. If you had one, of course.
My 2.5 year old 7890/5975 is on its 2nd pump (Pfeiffer) so failure is possible. The symptom I saw first was high enough pressure in the MS to shut off the turbopump, but I am not trying to put 3 mL/min into it so your symptoms may vary.

[Darber99]

Unfortunately, we don't have a spare pump. Would've been one of the 1st things I would've tried. Seems that my flow was a bit too much for the pump to handle. I changed my flow from 3.0mL/min to 1.2mL/min and the pressure dropped from 1.2x10-4 to 4.6x10-5. However my "grass" has "grown" exponentially and I can't even get spectra from the instrument now.

[carras]

Hello Darber99,

We had the same symptoms with a 5973 some months ago. The first thing Agilent recommended was blowing with He the zone further from the EI source (i.e. were the EM assembly is located). I must say that corrected the situation but only for a while before the “grass” came back. In the end Agilent changed the high energy dynode (part no. G1099-80001) which fixed the failure. In hindsight, though, I guess that blowing was of no use at all and that the effect observed was due to that the MSD was switched off for blowing and back on afterwards. Perhaps you can try to power cycle the MSD and see what happens. Hope it helps

[rb6banjo]

That was my next thought - the HED has failed. I had that happen twice (different instruments). It was when they were each about 7 years old.

[Darber99]

Would an HED failure cause my filaments to not turn on during tuning? I just looked through the sight glass of the MS and noticed that neither filament will turn on when I run an autotune. Checked all my connections and everything looks fine.

[rb6banjo]

I don't know about that. Sorry. Can you go into "Edit Tune Parameters" and force the filaments on?

[James_Ball]

I think the filaments will fail to turn on if the analyzer pressure becomes too high to protect them.

The bubbling in the oil could be from an air leak anywhere from the analyzer interface onward.

If m/z 18 has not dropped to less than 10% after a couple hours of having the vacuum on and source heated up then you definitely have an air leak somewhere. Can be as simple as lint on the o-ring seal around the side plate. Does this unit have the CI gas fitting on the interface? If so could be the aluminum sealing ring at that point. Also check the screws that hold the front glass and the rear plate of the analyzer on, I have had those be loose before also.

From the looks of things air is getting in somewhere, need to find that and get the analyzer pressure down to around 1.0*10^-5 at least before proceeding. Use some Dustoff(Difluoroethane) and spray around the analyzer at all joints and see if you get a response somewhere around m/z 66 I think. That is if you can get the pressure low enough to get the filaments to come on, if not you should still see a fluctuation in the analyzer pressure on the ion gauge if any is leaking in.

Apezion grease on the o-ring, very lightly then wipe off with chemwipe can also help with sealing.

One thing I liked better about the 5971 and 5972 design was gravity helped seal the analyzer with the top opening instead of hindering the sealing process as with the side opening analyzers.

[Darber99]

I think the filament issue is seperate from the pressure issue. I was able to get the filaments to turn on when the vacuum was at 1.2x10-4 b/c I was putting too much flow into the MS for the vacuum to evacuate. Pressure is now @ 3.0x10-5 and I can't get either filament to turn on even though neither of them is burnt out. I can't even get spectra when I scan so there's no way to tell how much H20 or N2 is in the system. Is there anyway to check if there is voltage/current to the filaments using a multimeter?

[James_Ball]

I think the filament issue is seperate from the pressure issue. I was able to get the filaments to turn on when the vacuum was at 1.2x10-4 b/c I was putting too much flow into the MS for the vacuum to evacuate. Pressure is now @ 3.0x10-5 and I can't get either filament to turn on even though neither of them is burnt out. I can't even get spectra when I scan so there's no way to tell how much H20 or N2 is in the system. Is there anyway to check if there is voltage/current to the filaments using a multimeter?

If you remove the black cover at the front of the side plate there should be pins there labeled FIL where you can check the voltage. You should get about 70V from that pin to ground. If you can find it there should be two for each filament and you should be able to check for continuity also.

[MSCHemist]

That vacuum pressure is to high check. My 6890-5975b standard turbo runs ~4.5E-5 torr with a flow of about 1.3 ml/min of He into it. Try servicing the side plate o-ring. Get a little apiezon l vacuum grease on that and the vent valve o-ring. Wipe it off with MeOH first. Smear some grease on a kim whipe rub it over the seals, then wipe off the excess with another kim wipe. Also I had a crushed o-ring where the transferline weldment meets the analyzer chamber (no grease on that one).

I don't like the Pfieffer duo rough pumps. They lasted on average 2 years for me before dying. I replaced mine with an Edwards E2M1.5 with no noticible difference in performance (takes the same amount of time to pump down and same operating pressure I had with the duo).

BTW I use the compressed air and look for m/z 52.

[Darber99]

We ended up switching out the Pfeiffer Duo 2.5 and installing an Edwards E2M2 and inspected the turbo pump (which had a loose bolt). The pressure is now at 3.5x10-5 with a 1.2mL/min flow, the filaments are lighting up and we can get a proper tune. H20<2% and N2 < 0.5%, however I'm still getting over 600 peaks in a scan between m/z 10-700. Am I pretty much looking at an HED failure now or could something else be causing this?