Help me! Tune

[Thuypham]

Hello, evrybody

I work at a GC 7890B/MSD 5977A Agilent
I have problem like this:
After change colum -> Run Atune, but result not good.
Before
http://i.imgur.com/yRUVLIT.jpg
After
http://i.imgur.com/YhiBsY5.jpg
Why are EMV and Gain Factor so high? What can the problembe?
I hope to get help. Thank you.
(Sorry, my english is very bad. )

[soccerdad007]

Hi Thuypham,
In both of your tunes, both the before and after, the base peak ion is 18, indicating a huge problem with moisture in the system. I would eliminate the moisture problem before trying to tune the instrument. The moisture could come from a variety of places. Agilents website should give you specific instructions how to find the source of the moisture and how to remove it from the system. Removing the moisture from the system should improve the tune.

Kevin

[Rndirk]

Hello,

The calibration gas hardly gets ionized/reaches the detector because there's a huge source of water, as the poster above me said. This autotune is worthless and the first thing I would do now is cool down the source before more damage occurs to your system.

My guess is that you have injected water in your system (accidently?).

You will probably have to vent the MS and clean the source. Remember that depending on your inlet type, injection of water can lead to backflash very easily and you might have to clean your gas supply lines.

[James_Ball]

Also when you changed the column, how did you gauge the amount of column inserted into the interface?

Best was is to open the instrument and push back the ceramic interface and see that about 1-2mm of column protrudes past the end of the transfer tubing, any more and it will be inside the source and the electron beam will be eating away at the column itself. If you use the measurement tool, be sure that the column does not move further in as you tighten the interface nut.

As above, the water is the biggest problem right now if the column is inserted properly. If you cleaned the source, the metal parts can be sonicated in DI water to remove any grit, but they must be sonicated in methanol or acetone after that to remove any water then should be sonicated in methylene chloride or hexane to remove any residual oils, and dryed before reassembly. Never use water or solvents to clean the ceramic or vespel insulators, replace them if they are dirty.

[Thuypham]

Hello,

Sorry for replying late

Thanks for replying me

The previous day, I replaced ferrule for GC-MS interface, and better result . However, m/z 18, m/z 28 and m/z 32 still high.
http://i.imgur.com/Zy1GbTd.jpg

[Rndirk]

How much time was there between replacing the interface ferrule (= venting the MS) and this autotune? Did you follow the instrument guidelines of doing this operation?

When pumping down, I think you should make sure the vacuum is at an appropriate level with a much lower concentration of air & water before starting to heat the MS parts (source, quads) above 100°C, let alone perform an autotune. Your autotune indicates a leak.

With this concentration of oxygen, chances are you will burn trough your filament quickly, and parts of the MS will get damaged.

[James_Ball]

Does this MS have the ion gauge for the high vacuum? I notice the Hi vacuum reading on the tune report has N/C on it. If it does not have the gauge then it is more difficult to diagnose a leak for certain, if it does have the gauge and it is not reading then the pressure in the analyzer is way too high.

Also new graphite/vespel ferrules will shrink when heated, so after about a day of being installed on the heated interface you will need to retighten the nut about 1/8-1/4 turn to remain sealed. One trick you can use to prevent the ferrule from leaking is to keep the ferrule in the GC oven in a beaker and the temperature cycles will pre-shrink the ferrules so they are ready to go when you need to change one.

[BMU_VMW]

A can of 'dust off' will help for sure in finding the leak(s).
I can hardly imagine it will only be one leak and considering the size the first thing I would check are transferline ferrule and the seals of the MS itself.

a bit of dust, a hair, .... on the seal of the analyser chamber can cause a huge leak. this is easy to check.

If it turns out to be leaking at the seal:
on page 78 and 79 of the manual you can see how to open it.
( http://www.agilent.com/cs/library/userm ... -90003.pdf )

Check the seals for any 'dirt' and check if they aren't damaged. If necessary you can put A LITTLE BIT !!! of grease on them to make them seal better.
also make sure that the thumbscrews (see page 78) are loose when the MS is in operation. The are only ment to keep the analyser chamber closed during transport. When in use it is the vacuum that will keep the chamber closed. If the screws are tightend when the vacuum is applied, it can cause bending of the chamber-wall and create leaks (at least that is what our Agilent tech told me).

good luck.

[James_Ball]

Also if the rear screw is engaged and you try to open the side plate you can bend the plate at the rear of the unit and make it not seal. My engineer always backs it out and puts a ziptie on it to prevent it from being tightened up just in case.

A hair will definitely cause a leak. I once had one leaking(5971) and when I opened it up there was a hair that went all the way across the analyzer and seal on both sides. The girl running the instrument had really long hair so I knew where it came from.

I also use grease on the o-ring seal, Apezion L grease, rub plenty on then clean it off with a Kemwipe. When you are applying it you will be able to feel any nicks or flaws in the o-ring. When putting the o-ring back into the analyzer groove, it will be stretched and will take a little work getting it back in good and level.

[Bigbear]

I like to have the front screw just about touching the seat, this way if we loose power the door will not pop open. I wait until the "pump down" window goes away and the system says to wait 2 hours before I tighten the screw.
I got a hair across the O ring of a 5970 and it sounded like a jet taking off!
I don't use grease, all I do before pump down is to moisten a kimwipe with MeOH and wipe the O ring and the mating part of the door.

[Thuypham]

Hi everybody,
Thank you very much for your help.

How much time was there between replacing the interface ferrule (= venting the MS) and this autotune? Did you follow the instrument guidelines of doing this operation?

I've been to about 6 hours and then proceed tune
And I haven't do this step:
“… Set a minimum velocity of 30 cm/s, or as recommended by the column manufacturer. Allow the carrier gas to flow through the column at room temperature for 15 to 30 minutes to remove air. “

Also new graphite/vespel ferrules will shrink when heated, so after about a day of being installed on the heated interface you will need to retighten the nut about 1/8-1/4 turn to remain sealed. One trick you can use to prevent the ferrule from leaking is to keep the ferrule in the GC oven in a beaker and the temperature cycles will pre-shrink the ferrules so they are ready to go when you need to change one.

Thanks for your trick. I will be applying.

I did like your suggestions,
I opened the analyser chamber and saw column inserted into the interface too long. and i clipped.
And the results are as follows:
http://i.imgur.com/G9gjw8I.jpg

But, when I scan organochlorine pesticide standard, have 2 problem:
1, mass assignments are incorrect
example: 181 -> 180.8; 165 -> 165.1; 355 -> 354.8 …
2, There are a lot of ghost peaks
What could be the cause? how to troubleshoot? or incorrect tune file (inappropriate parameters)?
Hope to get help. Thanks a lot.

[Thuypham]

and I want to ask how to separate oxychlodane and heptachloro epoxide?
I using DB-5ms column. I have tried many temperature program but inseparable.

[James_Ball]

Your air looks very good, so you probably have no leaks now. The water is still a little high but it should bake out over time.

The masses you are seeing are probably the actual mass instead of the theoretical mass. I normally run a test sample, then adjust the mass in the quantitation method to match the mass the instrument actually measures.

Some of those compounds are really difficult to separate, you may want to try something like the Restek CLP Pest1 or Pest2 column and see if those separate them better. They are designed for separation when using ECD so they should give more separation than an MS column would.

[Thuypham]

Hi James_Ball,

Thanks for your response.

Changing the column is very difficult because i just have the columns : DB 5ms and HP 5ms UI.

I´ve tried the ways like this:
- decreasing the flow: 1 ml/min -> 0.5 ml/min
- decreasing the Scan speed: 1.562 (N=2) -> 781 (N=3), N=4
- decreasing the Step size: 0.1 -> 0.05
But the result is inseparable

Do you have any suggestions?
Thank you very much.

[James_Ball]

Scan speed and step size will only adjust the resolution of the mass spectra, it doesn't have any effect on the separation.

Sometimes you can use a faster or slower temperature ramp to help separate difficult peaks. Also faster flows can produce narrower peaks for better separation. Try adjusting initial temperature, initial temperature hold time, temperature ramp rates and see if that has any effect.

I had one instrument I was having trouble resolving some early eluting components so I tried longer hold time and slower ramps but they just kept getting closer together, then I went with a 1min initial hold and faster ramp and they actually separated more, it was a surprise to me but it worked.