MSD 5975 Tune Problems

[Absolutdude]

Hello,
last week I did my first maintenance on a GC/MS. I cleaned the source, changed both filaments (trimmed Column at the GC, changed Liner and so on ).
After pumping down the MSD and starting it all I did my first Tune after maintenance yesterday.

EM volts were about 1600 (before that the maintenance, they raised over half a year from 1400 to 1700) Repeller was 15,05.
The Values for water (10%) nitrogen (17%) oxygen(5%) an Co2(0,7%) were quite high but felt down with every tune. So I thought, this is maybe from venting the MSD.
When I tried to start a run I didn’t get any signal. So I thought I might have overtighten the column Nut to the MSD.
I loosed the nut a little bit and got signal in the run I did after.

When I did a Tune after this first run I got huge values for Water/Nirogen/Oxigen (>1000%) so I obviously produced a leek with loosing the nut. After tighten the nut again the values for Water/Nirogen/Oxigen looked way better (4%/7%/2%) but the EMVolts and the Repeller went up a lot.
The EMVolts is now 2275 and the Repeller 23,25.(on both Filaments)

What does this rise mean?
I read that you should change the Electron Multiplier when reaching the about 2200 Volt. Is this true?
How can I reduce the optimum for the EMVolts. Is it just by cleaning the source (which I have just done) or are there other procedures?

What a normal values for water, nitrogen and oxygen? Less than 3 %? or is less than 5% okay (which I had in The past)?

Thank you!

[tlahren]

I don't believe your EM is "bad" yet. The 2200 volts is most likely a false setting due to an improper/poor tune. You have an air leak still at your MSD transfer line. The multiplier is offsetting for poor ionization of PFTBA. It increases the voltage to obtain a certain abundance of each ion. This is also hard on your multiplier. What type of ferrules are you using. 100% Graphite are no good. Use Vespel/Graphite ferrules and make sure they are the right size for the nut you are using. Also make sure the ferrule is pointed in the proper direction (The flat side-bottom of cone faces the MS, the top of cone to the GC oven). I usually tighten the nut until snug with one finger on the wrench. Then I go another 1/4 turn. Once you get the MSD pumped down again (you will have to vent again to replace the ferrule) then turn the transfer line heat on and GC etc. and let it sit for a couple minutes heated. Now open the oven and give that transfer line ferrule another "snug" (usually less than a 1/4 turn). Now go into "Manual Tune" and view the "Scan" option on the bottom. Select to "Open" the PFTBA valve. You should see NO peaks below m/z 69 (maybe a little water 18 m/z but). There should be no water or N2 or (<1% of 69).

If you see peaks during this Scan below 69 then tighten the TXFR line nut another 1/4 turn. It should not take much force to this. If you have a leak there you should see the peaks drop almost instantly (about 3-5 second delay maybe).

Another way to test for MINOR leaks in the GC/MSD system is to use a can of Air/Duster. This usually contains difluoroethane or some other light weight fluoroethane. You spray a gentle stream on a fitting and then watch in the same Tune-Scan window to see the respective ion/fragments increase in the spectrum. If you see new peaks, you know you have a leak there. NOTE: If doing this at the inlet column connection expect a delay time similar to your hold-up time or solvent delay so be patient when watching this.

Graphite/Vespel ferrules are very forgiving of over-tightening. However, it is still very possible to do. If you can see part of the ferrule through the hole in the nut around the column then it is way beyond over-tightened.

Hope this helps a little. Post back if you still have issues or questions.

[cleh]

You may have a leak on the o-ring at the analyzer door. Use the dust off as suggested by tlahren and spray it along the o-ring on the analyzer door. If you see it that it is leaking, vent the instrument. After it's vented, open the door and make sure nothing is on the o-ring or the inside of the analyzer door. You can wipe it down with a Kimwipe and MEOH. When you pump down, hold the door against the manifold with your hand (don't screw the two screws into place), turn on the power switch, wait for the air to clear the manifold, and then use the data system to pump down. You can then screw down the front screw on the analyzer (only finger tight) and leave the back screw unscrewed (it's only there for shipping purposes). Hope this helps.

[MSCHemist]

Keep in mind the o-ring on the analyzer door and the one on the vent valve need a light coating of apiezion (not sure the spelling) vacuum grease.

There is also an o-ring where the transferline weldment meets the main MSD that can wear out. No grease on that one though.

[tlahren]

Keep in mind the o-ring on the analyzer door and the one on the vent valve need a light coating of apiezion (not sure the spelling) vacuum grease.

There is also an o-ring where the transferline weldment meets the main MSD that can wear out. No grease on that one though.

Absolutdude,

If you are weary of using vacuum grease (I usually am) then you can also use MeOH to "wet" the O-ring while seating it. However, last time I talked to an Agilent Service Engineer they told me that the 5975 was designed a little different to prevent problems with that O-ring on the analyzer door. The 5973 was notorious for having problems sealing though. I once pumped down and gave the door a little tug and it seemed sealed. I came back the next morning and the analyzer door was wide open. Now I "wet" the o-ring on the '73s as MSChemist said. I think the O-ring in the transferline is a good place to check too.

[MSCHemist]

I just applied some last week on my 5975b. My side door seal was so dry (5 years old) that it was starting to deteriorate in spots and I was having occasional pump down failures due to it having difficulty forming a good seal. I wiped the o-ring a few times with methanol and a kim wipe. Then I smeared a little bit of Apiezon L on a kim wipe and ran it over the entire o-ring a few times. Finally I wiped the excess off with a clean kim wipe. The instrument has no further issues making a seal and pumps right down. There is no contamination of any sort evidient in the spectra and I assure that is why Agilent specificaly says to use Apiezon L.

[tlahren]

I just applied some last week on my 5975b. My side door seal was so dry (5 years old) that it was starting to deteriorate in spots and I was having occasional pump down failures due to it having difficulty forming a good seal. I wiped the o-ring a few times with methanol and a kim wipe. Then I smeared a little bit of Apiezon L on a kim wipe and ran it over the entire o-ring a few times. Finally I wiped the excess off with a clean kim wipe. The instrument has no further issues making a seal and pumps right down. There is no contamination of any sort evidient in the spectra and I assure that is why Agilent specificaly says to use Apiezon L.

Thanks for the information. I was unaware that Agilent was recommending this. Nice to know.

[MSCHemist]

yep It is in the manual. Also the nice Agilent Service technician that replaced my magnet gave me a smidge that I store in a 2ml autosampler tube. He said that should be enough for 50 instruments. You definately don't want to glob it on. Just get a smear on a kim wipe and run it over the o-ring then wipe off any excess with a fresh Kim wipe.

[Steve Reimer]

A tube of Apiezon L will last a lifetime, it was $24 back in 2007.

[tlahren]

yep It is in the manual.

Why would they put that info in the manual. Don't they know people don't read those things I suppose I should.

Sometimes I take for granted how easy the Agilent instruments are to work on and just push forward without reading.

[MSCHemist]

A tube of Apiezon L will last a lifetime, it was $24 back in 2007.

I checked Agilent's site a few weeks back I believe it is going for fourty something now.

[cleh]

I love Apiezon! Had my tube since 1979. Use it often and still plenty left. Scientific Instrument Services has Apiezon L, 25 gram tube for $95.00. You definitely need it for the 5971 analyzer seal (even when new). If you have to use it on the 72's and beyond, you may want to purchase the o-ring you're using it on and replace it during the next PM.

[MSCHemist]

you can replace the o-ring but my guess is the Apiezon will help prevent the o-rings from dry rotting. Mine was dry as a bone and showing signs of light flaking in one spot but after I cleaned and applied some it looks like new and seals like a dream.

[Absolutdude]

Thank you all for your response

@ tlahren: What do you mean by improper tune settings? I do an Autotune, so I don’t change the settings, or should I?
My Values for Water/N2/O2CO2 are (3,89% /5,57% / 2,16%/ 0,14%).
They should be less than 1%, right?

I ask because they never was since I use the MSD (half a year) and all the time before(about one more year). So there was maybe a leak all the time the GC/MSD was working…

I use Vespel/Graphite ferrules and I did not open the connection to the MSD (I just loosed it a little bit and tighten it again) so I wonder if this is really the problem?

I trimmed the column at the frontend, so there could be the leak too. Unfortunaly I don’t have a can of Air/Duster or Argon to test for leaks. But I think I will order one soon.

[tlahren]

@ tlahren: What do you mean by improper tune settings? I do an Autotune, so I don’t change the settings, or should I?
My Values for Water/N2/O2CO2 are (3,89% /5,57% / 2,16%/ 0,14%).
They should be less than 1%, right?

I ask because they never was since I use the MSD (half a year) and all the time before(about one more year). So there was maybe a leak all the time the GC/MSD was working…

What I meant by "improper tune" is that if you have an air leak it can cause poor ionization of PFTBA which will cause the autotune to adjust the MSD to some "way off" values (the 2200 volts EM for example) to try and compensate for poor ionization. Once the Autotune values are "out-of-whack" it can be hard to get a successful autotune after the fact even if you fix the air leak issue. Once you're fairly sure you have no leaks then restore a previous tune and try an Autotune from this point. You can restore to factory tune if you would like.

I don't remember if it was mentioned in this topic but another source of a leak can be the PFTBA valve or bottle. If the bottle isn't screwed in tight enough or there is a crack in the gasket there will be a continuous air leak. If this is the case though, I think you would see the air peaks (N2 and O2 etc) drop significantly in abundance if you "Close" the PFTBA valve. If you do manual tune and set the scan parameters from 10 to 50 m/z or anything below 69 m/z you should see the air peaks only and you can watch the total abundance of each ion increase and decrease as you make changes.

Yes the air water values should be close to 1% or below. Water is expected to be higher especially after venting (H2O can take a long time to pump out of the aluminum MS body interior).

I hope this helps answer your questions.