Chromatography Forum

Dear chhubert,

The outlet pressure is VACUUM. And thank you so much for the column related information.
I performed REPEAT PROFILE in MANUAL TUNE, however, it showed very inconsistant peak widths and could not gave a results, then I interrupted it. After a while, I tried to do Spectrum Scan and no peaks found .....
Seems the problem is getting bigger now.

I capped off the MSD interface and the instrument is pumping down now. So how much vacuum should be considered ok in this case (5973N)? I am getting 1.5*10(-5) now after 1 hour pumping down.

Hi WL

There may be more than one problem.

If you see no peaks the filament might be blown. Change to the other filament in manual tune and try again,, but beware that if there is a significant air leak the filament life will be shortened.

The other possibility is that no tune compound (PTFBA) is reaching the MS - I had a 5973 where the solenoid valve on the PTFBA reservoir would stick closed. During a tune there would just be a very noisy baseline, and after a while the tune would stop becuase it could not get stable peak widths. The solution was to remove the valve from the MS manifold and tap it on the bench, then put it back on. Note that you can still hear the solenoid clicking even when it does not open.

Assuming that you can get an MS signal, to trace an air leak into the MS you need nothing more fancy than a butane gas cigarette lighter. Release short puffs of gas around the column connections, the O-ring seals e.g. on the transfer line, the vent valve and the side flap that you open when you want to access the works of the MS. If you see extra peaks when you puff, that is where the leak is. Trying to trouble shoot a small leak by looking at the vacuum gauge is not going to work.

Peter

Yang,
It may be at the range of 10-6 torr when the MSD is capped off but it is merely a pure conjecture 'cos I don't have the experience before. Anyway, can you do the autotune when the MSD is capped? If so, did you find any air leak (N2/O2, and H2O)? If not, the air leak may be on the GC side?
In tracing small leak in GC/MSD system it is done by adding small amount of methanol, isopropanol or argon on the leak position and monitoring the fragment or m/z ion by the MSD using the function "repeated profile" in the manual tune. Before using it, you have to set the ions to be monitored (e.g. m/z 40 for argon) in the Edit Parameter. For details, you may go the Agilent web site and get the information under the "technical support".
To my experience, small air leak would not stop MSD from tuning. Big leak is usually coming from o-ring on the MSD side plate. Check this out too.

I reinstalled the column the day before yesterday, and the MSD could autotune then. I didn' do anything special, but it worked. I don't know why.
I found N2 and O2 when the MSD interface was capped, but I am not sure if that is because the system was not well pumped down at that time.
I checked the leak using acetone (m/z=58) yesterday. Surprisingly, I did no detect any m/z 58 ion when I applied acetone to interface, column, septum and around slide plate.

Thank you all for the help. Finally, I figured out the reason that caused the autotune problem---the column part into the ion source is too long.