DFTPP not passing, could use some help.

[GimmeYoLunchBox]

Hi there,
I’ve been desperately trying to keep a very old Agilent GC-MS (5890 Series II/5972 MSD) alive. I’ve replaced the main PCB board and have brought it back on-line after 3 years of collecting dust. For the past two years it was running flawlessly, however recently I’ve had issues passing the DFTPP standard. Mass 51 and 442 ratios are too high and too low respectively, I’ve run a bunch of tests to eliminate possible causes however I can’t seem to pinpoint it down. The very last thing I did before the problem occurred was install a new column. In hopes of getting some advice from the community I figured I would list all the things I’ve done up to this point to provide better details.

-created a new DFTPP standard to make sure it was not failing from evaporation of the solvent.
-tested injection port and transfer line for any leaks, none were found.
-cleaned the ion source, re-tuned the MS but still failed.
-Noticed in my chromatogram I had fairly high 44m/z abundance (around 10000), so I figured I could potentially have a leak somewhere.
----Used to leak detector to check all the connections on the GC side, no leaks were present.
----Used Argon to spray MS and observe MS profile for 40m/z and notice very small spikes (tripled in abundance) when sprayed around manifold o-ring. Vented the MS and checked the o-ring for any damage, no damage noticed so reinstalled and pumped down again. Currently I’m waiting for the MS to reach equilibrium so I can test the O-ring again for leaks. However not sure if that small of a leak would prevent DFTPP from passing.
----Here are my last relative abundances for air/water check before I opened up the MS
18/69=6.78 Water%
28/69=0.46 Nitrogen %
32/69=0.26 Oxygen%
44/69=0.22 CO2%
28/18=6.80 Nitrogen/Water %

Any suggestions or tips would be greatly appreciated as I’m out of ideas. Thank you.

Luke

[cleh]

It would be helpful if you could post your tune results. I have found that as these instruments get older, you have to be careful when pumping down and pushing the MS against the GC. I move the MS so the transfer line nut is just visble through the hole when looking at it in the oven. I then press against the top of the analyzer and turn on the power. Once the mech pump kicks in and stops gurgling, I move the MS the rest of the way so it's flush with the GC. If you find the analyzer O-ring still seems to be leaking, wipe it, the groove it sits in, and the edges of the analyzer sitting against the O-ring down with a Kimwipe and MEOH. If you have some Apiezon grease, apply a small amount to the O-ring. Did you change the liner, septum, seal, and O-ring on the GC? Also, if you have the blunt ended transfer line tip, make sure you have moved the column back a little so it isn't up against the tip. If you have the tip that is open at the end, make sure the column is not protruding out more than 1mm. I would check the DFTPP ratios in the tune file. Sometimes they get corrupted. I use 131/45% 219/40%, and 502/1-2%. Also try tuning on the opposite filament. Hope this helps.

[Bigbear]

Wonder where the water is not pumping out. All heated zones stable.

I do EPA 525 and find using a manual tune works best for me. I start with an autotune then cut the ion foccus enough to get the 198/442 ratio in spec. I use the method tune evaluation where 442 is the major peak.

[Steve Reimer]

If the water won't go away, is the source at the right temperature? It has been a long time since I looked at a 5972 so I don't remember the details of how they are set up.
But if the actual source temp is lower than what it reads, water will hang around and the tune won't look good.

[X]

One thing you can try: go to Tune/Tune Wizard and change the Tune target ration. In this case you should decrease the % of mass 50, and increase that of 414, so that you can get lower 51 and higher 442 response.

[GimmeYoLunchBox]

Thank you for all the feedback, I haven’t had any luck but ill keep listing some of my approaches.

-I’ve replace all the injection port part multiple times. I’m thinking I might completely remove the parts and actually clean the injection port tube with a brush followed by a rinse of DCM, Acetone, and methanol.

-The leak near the manifold o-ring was sealed but still showing high levels of m/z 44 although it’s only been about 20 hours since the pump down.

-All the heating zones appear to be stable. The Detector has two heating sensors, one at the GC-interface/transfer line (that is at 280C), and the other is in the radiator (reading at 167C). In this model the radiator is conductively heated by the GC interface.

-I’ve tuned on both filaments and have tried both manual and dynamic tuning.
Here are my tune results:

EMVolts-1812
Xray-38.5
Emission-35.0
MS Temp-169
Vacuum -40
Repeller -27.72
IonFocus-90.0
EntLens-7.03
EntOffs-19.58

I noticed on the chromatogram if I were to shift my fragmentation pattern slightly right from the center of the peak (shifting mass axis) the DFTPP passes, making me think I should play around with narrowing the peak. This would be uncharted territory for me as I’ve never messed around with altering the repeller, ion focus, or entrance lens. Any advice? Suggestions? Don’t really have any experience with manually adjusting these parameters. Thanks again.

Luke

[Yama001]

Perhaps this is related to the quad contact springs? I recall those caused problems, although that was more related to mass assignment.

If the software has a Target tune option, that can be pretty useful getting pesky ion ratios to behave. It will allow you to change the entrance lens setting from a linear ramp to more of a discrete function that can 'push' ion abundances up and down over the mass range. I believe this was call 'Dynamic Lens Ramping'.

[GimmeYoLunchBox]

Thank you for all the help. After reading a bunch of previous forum topics on this issue I think I zoned in on the problem, BigBears input helped out a lot. My tune file was corrupt (I recall getting an error initially that gave a message in regards to amu gain being 0 and peak width being too narrow), after resetting them to the fresh one (Through renaming and copying the original ones), I was able to get more reliable tuning results. However, under a standard manual tune I still manage to fail on 442, it’s still too high (442/198 is roughly 103%). I know it’s not the source, I know there aren’t any leaks, and I know its not the DFTPP solution. At this point I’m not sure if I need to play around with adjusting the ion focus or keep trying to retune it under different options (Max sensitivity manual tune or Dynamic Target Tune). Here are my latest Tune results:
Standard Spectra AutonTune
EMVolts- 1847
Xray- 42.8
Emission-35.0
MS Temp- 173
Vacuum-37
AmuGain- 539
AmuOffs- 56
Repeller-25.03
IonFocus-90.0
EntLens-40.16
EntOffs-4.02

[Peter Apps]

I noticed on the chromatogram if I were to shift my fragmentation pattern slightly right from the center of the peak (shifting mass axis) the DFTPP passes, making me think I should play around with narrowing the peak. This would be uncharted territory for me as I’ve never messed around with altering the repeller, ion focus, or entrance lens. Any advice? Suggestions? Don’t really have any experience with manually adjusting these parameters. Thanks again.

Luke

If you really mean chromatogram (instead of an expanded spectral scan peak) then you might have a bizarre effect of capillary peaks being sharp enough to bias mass intensities as the mass analyzer scans. Do your results get worse instead of better on the other side of the peak ?

Peter

[Yama001]

Peter has it right here. I believe most Agilent instruments scan from high mass down to low these days. Unless you have a very fast scan time, a spectrum taken from the up slope of a peak will have greater intensity for low masses, and the reverse on the down slope of the peak.

The US EPA's Contract Lab Program requires an average of the 3 scans around the apex of the peak. This does not perfectly account for scan to scan variations, but it does provide a systematic way to evalute the tune. Personally, I think this means the tune criteria is not very important (isotope ratios perhaps make sense), but I don't get to make the rules.

A faster scan time will give you less variation between scans, as well as better chromatographic peak definition, but will also result in a lower signal to noise ratio and noisier mass spectra. It is always a trade off.

[mckrause]

You need to slow down your chromatography - you are almost certainly not getting enough scans across your DFTPP. For current column technology (say, 20 m x 0.18 mm) you can easily get DFTPP out in 2-3 seconds, leading to the situation of having only 3-4 scans across the peak. This is not enough to get reproducible DFTPP. We are running 5972s as well. We have set our scan rate to 1 (the fastest) and slowed down the chromatographic run for DFTPP such that we're getting DFTPP out over a 6-8 second timeframe. This gives me 10-12 scans across the peak, which allows me to average the apex scans.

Another thing: as the 5972 quad ages, for whatever reason you start to get more 219 relative to 69. If your 131 and 219 will not come down to around 45% you are going to have a very difficult time tuning to DFTPP criteria. If this is the case, metal quads will fix that issue. CSS Scientific and Cal Tech Scientific both offer metal quad upgrades for the 5972, and it does make a huge difference.

[James_Ball]

You need to slow down your chromatography - you are almost certainly not getting enough scans across your DFTPP. For current column technology (say, 20 m x 0.18 mm) you can easily get DFTPP out in 2-3 seconds, leading to the situation of having only 3-4 scans across the peak. This is not enough to get reproducible DFTPP. We are running 5972s as well. We have set our scan rate to 1 (the fastest) and slowed down the chromatographic run for DFTPP such that we're getting DFTPP out over a 6-8 second timeframe. This gives me 10-12 scans across the peak, which allows me to average the apex scans.

Another thing: as the 5972 quad ages, for whatever reason you start to get more 219 relative to 69. If your 131 and 219 will not come down to around 45% you are going to have a very difficult time tuning to DFTPP criteria. If this is the case, metal quads will fix that issue. CSS Scientific and Cal Tech Scientific both offer metal quad upgrades for the 5972, and it does make a huge difference.

The 5972s were also the first to have variable Emission current settings and if you are getting too much low mass decrease the emission current, too little increase it. Also can play around with the Variable settings on the Entrance Lens Offset to knock down high abundances if needed.