Mass spec tuning problem

[laytox]

New to the forum, hope I'm posting this in the right place. I'm working in a lab as a grad student and have been tasked with bringing an Agilent 6890GC with 5975MSD back from the dead and am having issues getting it to tune. I have experience running GC-MS but this is the first time I've been tasked with assembling and getting them operational. The MSD has been sitting collecting dust for the past 7-8 years or so, so the internals have been exposed to air for at least that long. I got everything put together, got some fresh high vacuum grease on the o-ring and vent valve, and the system pumps down just fine to around 3e-5 torr with appropriate temperatures, etc. However, while trying to tune I keep getting a message about the peaks being too wide. The ions it is looking for are PFTBA at m/z of 502, 219, and 69 (I'm assuming this is standard). The 219 and 69m/z peaks look fine (symmetrical, good shape, very little noise) but the 509 peak is a mess:



I've cleaned out the ion source with methanol and did an MSD bake. Filaments looked a little rough so we ordered a couple new ones which should be in today. Other than that I can't figure out what's wrong with it... is this a detector issue or an ion source issue? Don't know if the higher m/z noise is from improper ionization or if the detector needs some sort of maintenance.

Thanks in advance for any input, this has been a bit of a headache.

[JMB]

The shape of the m/z 502 signal (spiky, low signal to noise) is characteristic of low intensity.
That relatively low intensity at high mass is itself characteristic of (in a quadrupole mass analyzer) dirt from ion burn on the first half-inch or so of the rods.
The theory is that at identical kinetic energy (1/2 mv2) the slower, heavier ions are deflected more by the fringing fields at the rods than the lighter ions. Dirt on rods also disturbs the ion-focusing fields.
That being said, what is the mass analyzer in the Mass Selective Detector ?
Also, I would have expected to see a vacuum of ca.x 10 -6 torr range. Check your pumps (run on ballast to purge volatiles and or change the oil) and check for air leaks.
When system is up and running, do not EVER power down!! Except for routine maintenance.
PS Since you do not know history of maintenance, change all pump oils (turbo & rotary)right away, much better than lost turbo pump bearings.

[laytox]

The analyzer is a single quad, so that would make sense. Foreline pump oil has been changed, was considering changing the diffusion pump oil as well so I suppose I'll go ahead and do that. The system did get taken down by a power surge, but it was having the same issues before that as well. In the service manual it says to never ever remove the quad, so if it is in fact fouled how would you go about remedying the issue? or is it just time to replace the detector unit as a whole. We're about to get some UPS units to plug our instruments in to so I figure I'll go ahead and vent and power down and try to get all the maintenance done at once.

Don't know if it makes a difference but we are running in EI

[James_Ball]

Newer Agilent quads (5973,5975) are gold foil encased in fused silica. They rarely get ion burns on the quads themselves. They can get crud like diffusion pump oil in them.

You may want to clean the HED. That is the white ceramic part attached near the multiplier. It has a single hard wire connecting to the plate, just slide it out and on the inside is a silver button. Clean the button with isopropanol then reassemble. Sometimes if that is dirty it can cause similar problems.

Also if it has been sitting exposed to atmosphere for several years, the ceramics will be full of moisture and air and will need several days to bake out. Use a blank no hole ferrule to seal the inlet to the MSD and let it pump for a few days at temperature then check the abundances again. You will probably want to replace the electron multiplier horn also after so much time open to air.

[laytox]

Awesome, I appreciate the input. Just noticed something though I dont know if it will help with troubleshooting - went to replace the septum in the GC inlet and as soon as I removed the old one the HiVac pressure dropped to 10^-6 torr and frontline pressure dropped as well... would've guessed it would have the opposite effect, allowing ambient air through the column? Might be nothing, just seemed odd that it would have issues achieving the correct vacuum with a septum in but not when it's removed.

[LALman]

I have an earlier model 6890N/5973inert system with std turbopump. That vacuum of 3x10^-5 is just fine if that includes a flow of 1mL/min of carrier gas. It will drop a little lower if carrier gas is shut off.

It looks to me like you have very dirty lenses. Its not enough to sonicate with methanol. Get either the agilent green abrasive paper or some 15 micron alumina grit and some no linty cotton cloth and give the repeller, the ring, the draw out lens, and the other two lenses a light polishing. If you use the alumina, I like to moisten it with methanol and use the cloth saturated with alumina grit as my grinding medium. But the green abrasive paper is faster and better. There should be a kit floating around somewhere in your lab (that came with the instrument) with rods for holding the paper or cloth so you can run the ring and lenses back and forth along the paper on the rod.

Then rinse in a sonicator with DIW, then rinse/sonicate with MeOH a couple times and dry. Don't get any of the insulators wet. reassemble and get that vacuum back down. After a few hours you should be able to see peaks with the EV turned way down. Wait at least 24 hours and you should be able to get a good tune.

When you put the vacuum gasket back in... Draw it through your fingers between some clean lint free material (I use kimwipes) and remove most of the grease so its just a very light coating and wipe out the groove so there is no visible grease. You will get a better vacuum if the grease is barely coating the rubber.

P.S. Don't even think about cleaning the quads. But otherwise I agree with all previous posts.

[LALman]

Awesome, I appreciate the input. Just noticed something though I dont know if it will help with troubleshooting - went to replace the septum in the GC inlet and as soon as I removed the old one the HiVac pressure dropped to 10^-6 torr and frontline pressure dropped as well... would've guessed it would have the opposite effect, allowing ambient air through the column? Might be nothing, just seemed odd that it would have issues achieving the correct vacuum with a septum in but not when it's removed.

It depends on the GC program; but if a pressure pulse or high flow is programmed or if perhaps the column in the oven is not the one in the method, then pressures might be wrong (high) removing the septum would drop pressure on the column and vacuum would get better.

[laytox]

So after changing the septum, I let the MS pump back down over night and today I'm seeing 28 and 32m/z at a ratio of about 1:1, both of which are less than 1/50 the abundance of 69m/z, is that safe to say the air leak is contained? I also went around with a can of keyboard duster to everywhere I could think a leak would occur and saw no new peaks pop up while monitoring. Still wont tune, but it won't be until probably next week some time before I can vent it and replace/clean things (planning to replace filaments, diffusion pump oil, clean HED). If the 502 abundance is still relatively low would this point to an electron multiplier issue?

[James_Ball]

What are the counts on m/z 69 and at what EM voltage are the readings taken?

If it is about 300,000 and you still don't get a peak at m/z 502 then it is probably not the multiplier, but the lenses. The tuning program can't set the 502 peak width if it has nothing there to work with, that is why it will just be a pile of peaks.

Also what is the abundance and peak width at m/z 219?

I have had times before that I had to reset to default settings and let it work from there after everything is good and clean before it will make a good tune.

[gb_gcms]

We had this issue when a tune file was saved while the system had a big leak, so the parameters were all out of whack and the program couldn't correct itself. If you happen to have a backup atune file, try tuning with that one. Alternatively, if you can dig up a record of the last successful tune on that machine, try plugging in those values in the Manual Tune window, save, then run atune again. A screenshot from the Manual Tune window will help give us an idea of where the parameters are currently set. Best of luck! One of the GC/MS systems I manage is the same setup as yours, so hopefully my experience can be of some help to you.

[laytox]

Alrighty, so I spent yesterday taking Jethro apart (thought he might run better with an identity). Changed diffusion pump oil, refilled PFTBA, gave him new filaments, re-cleaned the source (4000 grit abrasive paper, 15min in DI, 15min in methanol, 10min bake), wiped down the sideplate o-ring, changed ferrule in GC... I think that's about it. Let him pump down overnight, now he's down to around 3e-6 (torr? mtorr?) hi-vac and 8.5 (torr? mtorr?) foreline pressures. Purged the calibration valve and he successfully tuned, but there are secondary peaks that I'm not sure are supposed to be there



There's still some stuff under 69m/z but I'm guessing it'll just take another day or two for that to go away.

[LALman]

That looks very nice, good job! Those side peaks to the right of each peak are always there and in fact the smoothness of the side peak on the 502 peak gives you an idea of how clean the lenses remain. When it starts looking raggity no matter what you do; you can expect that the clock is ticking before the next source cleaning.

Look up PTFBA in your NIST library and you will likely see those other peaks belong to PTFBA. But here is a spec sheet showing the peaks you may find.
Note that there are those side peaks at 70, 220, and 503 which is what you see.
https://tools.thermofisher.com/content/ ... -amine.pdf

Here is a handy way to reach a good BFB tune if you have trouble getting the Agilent BFB tuning method to tune successfully...
BFB Tuning for Environmental Analysis: Three Ways to Succeed
https://www.agilent.com/cs/library/Supp ... f05011.pdf

[laytox]

Ah, those are probably the M+1 peaks eh. Well I appreciate y'alls help! I think ol Jethro is ready to roll.

[James_Ball]

Those are called the Iso peaks for Isomer and come from the small percentage of C13 that can be in the molecules. 1%, 4% and 8% for the 69/70, 219/220 and 502/503 are considered good values. The ratio % will show up in the tune printout.

It really does look great for an instrument that sat for so long, good job!