Aux EPC or PCM for Collision Gas 7890/7200

No.  There is a QTOF specific collision cell EPC module. The software needs to recognize it.

G3850-60039 | Agilent   EPC module

G3470-60550 | Agilent,  Fitting and 1/16" tubing to go from the EPC module to the QTOF CC 1/16" Swagelok connection.

Is the 7890 setup for the MS transferline heater/sensor connection?  Has that GC ever been run with a mass spec?

Is it a 7200A or 7200B ? What's the serial number?

It's a 7890A - it hasn't been run with a mass spec before but it has the correct aux heater connection and we removed the rebated hole for the transfer line and got everything set up and configured so it recognizes the transfer line and heats it. The 7200 was purchased from a reputable biotech liquidation auction company by itself as working and come to find out it was missing the entire transfer line module, the TOF gauge, the foreline gauge, the #3 turbo, half the foreline connections, and the #2 turbo had a lunched bearing. It's taken quite a bit of work but we've managed to get it running smoothly for uunder $10,000 in parts from Agilent and resellers, finally giving good signal, holding excellent vacuum, turbos are running extremely smooth/quiet/low current consumption, no errors etc. 

For now just to test everything out we've been able to run samples by feeding N2 to the collision cell directly via a manual Omega 0-6ml/min flow meter and everything seems to work fine, and we're getting good results so we were hoping to avoid having to spend another $3k on parts if at all possible but if that's the only way, I guess that's what we have to do. Thanks!

Yeah, that's what I thought as well but the AUX EPC flow controllers only are capable of pressure regulation, not flow correct? The problem we were having was since the outlet is connected to the collision cell under vacuum, and since the n2 flow is so tiny, the EPC measures -14 psi. And since there is no way to set the EPC for flow rate only pressure 0-100, it throws an error and gives up. I experimented with using it in a reverse pressure regulator configuration with output weldment as the N2 input and the normal input going out to the flow cell, which fixes the negative pressure issue but since the PID loop is set to open when the pressure is low and close when it's high, it does the reverse of what's needed and the firmware update program only gives a few predefined options for PID - I don't believe there's any way to modify the PID settings directly without directly editing of the PID setting XML files, and they consist of eeprom memory locations not human editable variables from the looks of it)

That's why I was asking if an AUX PCM might work since I believe they have flow sensors as well as pressure, correct? Mass hunter methods seem pretty flexible in configuration of different modules so even if it didn't directly recognize it was collision cell gas, it could still be set up manually. 

As for the 7200, the serial is US1247H001, model G3850-64201, and I believe it was at atmospheric for at LEAST a year (I can look Monday but I believe the last service sticker was from 2021 and it was purchased mid January 2024, so it's somewhere between 7 months and 2.5 years at atmospheric (with no TOF turbo or blanking plate, just a wide open hole, and a Kimwipe blocking the opening for the missing transferline. It was probably stored in a lab from whenever it was vented (sometime post 2021- late 2023) then moved to an auction warehouse in RI, then to the lab here in Farmington CT).

We haven't run a real tune or calibration yet (just manual calibration checking bbackground and signal from the EI calibrant, and quick autotune reports and water/air reports) but it's pretty low noise and descent signal so far. The source is definitely very dirty and needs a good clean which we're doing Monday, and the only issue we're having at the moment is o2 and n2 spikes.

We initially had an extremely small air leak somewhere - small enough that the system got to 10-5/10-6 qused/TOF from atmosphere in less than 2 hours (with a DS-602 foreline pump FWIW) and 10-7/10-8 overnight, but we were getting peaks at 28 & 32, so I cleaned and regressed all the o rings and tightened all the clamps, bolts and screws which significantly dropped the signal but somehow there are still peaks at 28 & 32 but instead of 4 to 1 they are EXACTLY equal - and I mean to the decimal: precisely 3,497,070.0 counts each with collision gas off and the transferline blocked. If you introduce some n2 to the transferline, the oxygen peak drops off and the nitrogen picks up but then when you stop they both go right back to 3,497,070.0 within maybe 15-20 seconds. I leak tested every seal, joint and fitting with nitrogen to look for leaks, but nothing, even when bagging the entire 7200 and purging with n2. So I'm hoping it's just source contamination but if you have any other ideas, please let me know. 

Aux EPC are electronic pressure control, not forward flow controllers. Controlling forward flow at 1.5 ml/min into vacuum is difficult.

That 7200A was first installed about March 2014. It passed the End of Guaranteed Support 11-30-2021.

Yes, the peaks are exactly equal -- they are maxed out !  That means that the nitrogen is more than 4x that number of counts.  In manual tune, set the Low mass to 25 amu. the quad cutoff to 38 amu....never lower or too many ions hit the detector. Use the autotune emission current - whatever was last, I guess?  Or try really low, like 5µA or lower to start. Use an emission current that does NOT max out the detector.  If the tune was working and the emission current was right, 69 abundance about 1M,  the m//z 32 should be about to less than 1x10^5. 

Either the emission current is too high, the detector setpoints too high, the quad cutoff too low, or you have leaks to find or some combination.

Interesting - the turbos that were installed have 2013 date codes on them so I figured it was some time around then but I was a little off for the PM - the last sticker on it shows a 20 APR 2018 date with all the boxes checked: IQ, PM, INSTALL, OQ, CAL, and OTHER (with nothing filled in) all performed. I'm not sure if that was actually the original install date or if it had been sold by the original owner to another company and was installed by Agilent but was out of warranty so they didn't bother to get the PM for 2019 (or it broke before it was due and they just used it for parts)

And AHHH! Okay, that makes sense, I was confused because I've seen counts way higher than that but does that just mean the detector is saturated for the current emission current setting? I had set the low mass cutoff to 20 and high to 100 with 20 for the quad cut just to get a good look at the o2 and n2 peaks while testing for vacuum leaks. We have been manual tuning at between 3 and 25ua, with some tests at 50ish. 

I just removed the source and gave it a good cleaning and the o2 and N2 counts fell to ~500 & 1500 before bakeout. I'm baking out now and will let you know what happens. 

One thing that confuses me still is where those peaks are coming from - I obviously suspected a leak but after bagging every joint and pumping in n2, it doesn't change at all - even if I bag the entire machine and pump in n2, it stays exactly the same, and not at 4 to 1 either - the n2 peak is LESS than o2 (28m/z is approximately ~80% of the 32m/z peak) and the machine has been at vacuum for well over a week now. If I introduce air into the transferline, they instantly jump to being at the expected 1:4 ratio of o2 to n2, and if I just introduce n2, it completely swamps the o2, but then it goes back to n2 above o2. I turned off the n2 bottle with the regulator off to see if any flow was leaking in but the gauge stayed bang on what it was on Friday.

And one other strange thing is when the collision gas is turned on, the o2 peak stays the same but the n2 peak actually disappears! Or at least falls significantly. For instance with the n2 off, the emission current at 5 and the mass range from 20-100, they are both registering around the 50,000 counts range. As soon as the n2 is turned on, the o2 stays the same and the n2 falls to a few hundred counts. It's so counterintuitive I have to keep double checking that I didn't somehow reverse which peak was which. 

There will always be measurable oxygen - or at least I've never seen a QTOF with none.  On a 7200A, one week is barely pumped down. We recommended two weeks between installation and checkout.  Getting all of the molecules off that are adhered to the walls of the flight tube takes time.  For every molecule flying around there are ~500 adhered to all of the surfaces.

The collision cell is part of the magic. Gas on and gas off totally changes the way it functions and totally changes the way the QTOF system calibrates. You cannot necessarily troubleshoot this like a single quad system.

The entrance lens is not part of the Removable Ion Source, it is inside on the analyzer. To clean it means venting, removing the source radiator, removing the entrance lens, cleaning it, and going backwards to reassemble. 

Wow, I wish I had posted earlier in the process because you have been a tremendous help - I probably could have saved a week if I knew all this beforehand. 

That all makes a lot of sense. I just finished the one hour bakeout and here's how the EI calibrant spectrum looks:

Pretty happy with that signal for now anyway. The transferline is just capped off from when I split the system to try and track down if/where it was leaking so I still need to reinstall the column and 7890 but do you recommend letting it pump down for a few more days before we test/tune/calibrate everything?

Yes, wait longer, especially since you just cleaned the source.   You might get a no-column-flow tune out of it maybe Monday?  The spectral tilt, the relative abundances of 69/131/219/502 won't be ...right... with no column flow, but if they are there and the resolution is even sort of good, I'm impressed it got this far. 

Yes, wait longer, especially since you just cleaned the source.   You might get a no-column-flow tune out of it maybe Monday?  The spectral tilt, the relative abundances of 69/131/219/502 won't be ...right... with no column flow, but if they are there and the resolution is even sort of good, I'm impressed it got this far.