GC/MS

[Analisis]

Hello,

We are using a 6890 with a 5975 MSD from Agilent. The foreline pump had trouble some time ago, we disassembled it and it functioned for about one month more. Last week the pump failed and I found that the oil level was down. I disassembled the pump again and put more oil. The pump apparently works ok because the pressure in the analyzer is 4.01e-5. But when we try to do an autotune, the equipment stops, saying that peakwidth exceeds the limits. Also you see only noise in the profile of the tune.

I am checking with manual tune masses 69, 18 and 28 doing a scan. I am monitoring the relationship between 18/69 and 28/69. This are in the order of 521 and 931 % after two hours of pumpdown start. I have several times checked for leaks in the system, but cannot find one that should be very large for the amounts stated. I think that the problem is the pump, what do you think? Somebody knows another form of testing it?
The vacuum system is a standard turbo pump.
Thank you,

Adolfo

[bhuvfe]

Adolfo,

Try to plug the column or directly the MS interface to decrease the
number of variables. Once you know that your air leak is not coming
from the inlet/column you can focus on the rest; i.e. o-rings of the
manifold, etc.
Agilent generally advise to use argon to detect leaks. But to be
honest I never tried myself ( I used solid CO2 as alternative).


You mentioned the oil level. Did you have a oil leak at the foreline pump
or was it sucked up into the MS?

Regards,

bhuvfe

[Analisis]

bhuvfe

Well I had an oil leak that is for sure. The oil was not going inside the analyzer. What can I do if that happens? (scared).

The oil leak was fixed, but now I want to know again if some oil went inside of the analyzer. Can you look at it or sometimes it could not be seen?

I have no form to plug column right now. So I will come back when I have the plug for the interface.

What I have done is to reduce the flow of the column to a minimum. I could also not simply shut down the flow because this site is too hot and the air condition doesn't function. Yes I know this site is a mess, but this is what we are trying to fix.

Ok. Thank you very much for your help.

This story continues.

Adolfo

[Analisis]

Hello bhuvfe,

Yes I have checked again and when I open the analyzer after venting, a cloud comes out of the analyzer. I supouse that it is vaporized oil that went inside of the chamber. Oh man. How can I fix this mess? I know I should clean it but I am not sure with wath solvent and if I should clean all the analyzer or only some parts. Can somebody please help.

Thank you for your help.

Regards,

Adolfo

[bhuvfe]

Adolfo,

A cloud is not supposed to get out. Did you release the vacuum after venting?

I guess that now you will have to clean the ion source (to start with..). Refer to the manual, genarally Agilent's manuals are quite good (and normally you will find a video in the installation CD).

If that doesn't help then there's probably the quadrupole to be cleaned.
That's normally for Agilent service engineers.

See also previous message viewtopic.php?t=9777&start=0&postdays=0 ... 61a3d675a1

If I were you I would clean the source and see if you can get away with it...


Regarding the flow. You can plug the column simply using a inlet septum.
You can plug a interface nut using another ferrule and a pin (such the one used to enlarge/adapt the hole in ferrules) if you don't have a ferrule without a hole.

Regards,
bhuvfe

[gcguy]

In my experience with a 5973 ms, 2 hours is no where near enough time to get rid of the air and water in the system. We leave it for at least 12 hours after venting before starting the tune process.

GCguy

[Analisis]

Hello,

Yes I have vented before open the analyzer and then I saw a cloud coming out of the analyzer. I have seen it before, but because there was so much dust in this ambient I thought it was dust and I was amazed for the quantity of it. However this time the site was cleaned and I have cleaned also the instrument, so there was no dust. I then saw again, and I could see how this cloud settled down and it looked as small water vapor, it settled down very fast. Also it seemed to me that I could smell the oil.
Well the ion source is no problem to clean. I myself worked as service engineer for the distributor in Guatemala, but I never had a backflush in 13 years work. I know that there is a procedure for the quad, but I don't have the procedure handy.

Well I think I will start with a source cleaning.

Thank you again for your valuable help.

Regards,

Adolfo

[gcguy]

We use an oil mist adsorber between the rough pump and the instrument to be on the safe side.

[Analisis]

Hello gcguy,

Ok thank you, can you help me, where can I get this oil mist adsorber? I figure that this comes between the hose and the pump? I am neofit in this.
What parts do I need?

Adolfo

[Analisis]

Hello,

I have checked the site of Scientific Instrument Service and I think they have what I need. This is the link to the site and part http://www.sisweb.com/vacuum/sis/coaxialt2.htm

Thanks again

Adolfo

[bhuvfe]

Adolfo,

The oil trap will take care of only a part of your problem (i.e. the oil but not why you have/see oil when venting the instrument....)

Is the foreline pump under the bench or anyway 1m (or more) lower than where the MS is?

I had several power failures and I never had oil backflush (maybe I was lucky...).

Regards,
bhuvfe

[Rick]

Hi!
What is the knid of mechanical oil you are using?
It is important to use a good oil type and quality because the foreline pump is hot when running........

[Analisis]

Hi,

I have cleaned the ion source. With the procedure that recommend Agilent. They recommend first to

1 Disassemble the ion source.
2 Collect the following parts to be cleaned:
• Repeller
• Interface socket
• Source body
• Drawout plate
• Drawout cylinder
• Ion focus lens
• Entrance lens
These are the parts that contact the sample or ion beam. The other parts
normally should not require cleaning.
3 If the contamination is serious, such as an oil backflow into the analyzer,
seriously consider replacing the contaminated parts.
4 Abrasively clean the surfaces that contact the sample or ion beam.
Use an abrasive slurry of alumina powder and reagent-grade methanol on a
cotton swab. Use enough force to remove all discolorations. Polishing the
parts is not necessary; small scratches will not harm performance. Also
abrasively clean the discolorations where electrons from the filaments
enter the source body.
5 Rinse away all abrasive residue with reagent-grade methanol.
Make sure all abrasive residue is rinsed way before ultrasonic cleaning. If
the methanol becomes cloudy or contains visible particles, rinse again.
6 Separate the parts that were abrasively cleaned from the parts that were
not abrasively cleaned.
7 Ultrasonically clean the parts (each group separately) for 15 minutes in
each of the following solvents:
• Methylene chloride (reagent-grade)
• Acetone (reagent-grade)
• Methanol (reagent-grade)

I have done the steps above took from the manual. When I cleaned with Methylene chloride it turned green. After the wash in ultrasound bath the pieces looked with spots and stains, so I cleaned again with same procedure and the stains remained there. So I looked after other methods and used one that found in www.sisweb.com. They used a mild detergent, water, acetone and methanol to wash the pieces. The results were better. So after the cleaning I assembled the source again. Wait for the stabilization and after some hours (5). I tried the autotune. The surprise is that the filaments were out. In the 5975 you can see the filament light when they are on through a glass window. But this time there was no light. So I vented again and checked with ohmeter filaments, etc. on the source. I assembled again and this time I tried the filaments only 1.5 hours after pumpdown complete. The filamnets went on. I let the equipment then four hours more stabilizing but when I tried to tune, the filaments doesn't light again. After some more intents, the massspec doesn't show any peaks and only shows a message telling that the MSD has not found any peak.

Do you have some ideas.

Thank you for your help and time.

Regards,

Adolfo

[Analisis]

I forgot to tell that we are using Inland oil 45. This is the one used in Edwards pumps. This is a Pfeiffer pump and Pfeifer told me that they recommend only P3. But other people told me that it is the same and that I can use it in a secure form.

Adolfo

[thohry]

What is the situation now, after you have cleaned the ion source?