99% loss of sensitivity in GC/MS!!!

[MSCHemist]

Hi all I have a problem I can't see anything except major solvent peaks on my GC/MS

A bit of history: I started a new position and found a GC/MS Agilent 6890-5975 in terrbile shape due to the gas line plumbing. It had worn compression fittings cobled together with ptfe tape and a bad mix of brass ferrules on steel lines and mix and match cu and steel tubing. It had bad leak problems. I redid all the lines and got it very nicely air tight. It worked well for a while.

Last Friday I did some PM, I cleaned the source, replaced the liner, septa, gold deal, clipped the column on both ends, replaced the PV trap. The bottom filament is in rough shape but all in it looked fine when I got it back together. It passed the atune em volts are getting a bit high 2000-2200 but the peak splitting on the 69 and 219 dissappeared after I cleaned the source.

Anyways now the main issue. I ran our 3-mcpd calibration and found the peaks had poor resolution but the correct retention times and good linearity. However the peak areas were 1% of what they were several months ago. I reclipped the column (thought It may be positioned wrong or cut badly), checeked the autosampler, and reran the atune and it was fine. So I reran the calibration and the same result.

I didn't have a performance check standard handy so I made my own. I grabbed some reference standards
(2-methyl butyrate, R-Limonene, 2,3 heptane dione, 2-furyl methly ketone, and diethyl-l-tartarate, and made a 4 ppm / ug/ml solution in n-hexane and injected 1ul splitless. I only saw the solvent peak and nothing else.

I figure the column is probably toast from the leaks but is that enough to cause this. Please give me some suggestions. Thanks

[MSCHemist]

I thought it might be the split valve stuck open so stuck a flowmeter on the split vent and it read 0 ml/min for 1 min while doing a splitless injection so no help there.

[R13]

Correct (if it is correct indeed in the earlier runs) retention time but bad resolution would indicate bad column. However that should not decrease sensitivity significanly.

Did You take out and check syringe? Mabe the needle is plugged or the barel leaky so sample is not injected properly?

Did You check septum purge flow?

[MSCHemist]

Yep I took the syringe out and pipetted some acetone. It is fine. Also It rinses with acetone twice before and after every injection.

I didn't check the purge flow I just checked the split flow.

I kind of figured a bad column would give poor peak shape and resolution but not a poor signal.

[R13]

Just occured to me - are You sure your column does not go into ion source too far?

[MSCHemist]

Just occured to me - are You sure your column does not go into ion source too far?

Nope I installed it so that you could just barely see it protrude past the end of the transfer line ~1mm?

Could something be stuck inside the column, a piece of speta, a piece of graphite from the ferrule. Perhaps it is cracked inside the transfer line as anywhere else I'd see a leak.

[Don_Hilton]

If you saw the solvent peak at about the right time, the column is not plugged. A plugged column would run slowly.

If you have overtighten the ferrule on the transfer line end of the column, it is possible to crush the column.

YOu may get flow through the crushed column. But I expect to see an air leak. When I've crushed a column in a transfer line, the break has been toward the ouside of the instrument and a tug on the column revealed the break.

[MSCHemist]

Well I ordered a new column so we'll see if that fixes it. I'll also run some solvent through the injector before I hook it up. Agilent says remove the gold seal but I don't want to replace it again. Can I just drip it through with the gold seal still in?

[MSCHemist]

Could there be an issue with the transfer line. Is there anyway to check it as it just comes one when you pump down the MS I don't see any status or temperature displays.

[Don_Hilton]

A cold transfer line would mess things up; i t would be a cold spot in the column. I don't know about the '75, but on the earlier Agilent (HP) instruments, the transfer line was heated off of one of the GC's AUX ports.

[Peter Apps]

I thought it might be the split valve stuck open so stuck a flowmeter on the split vent and it read 0 ml/min for 1 min while doing a splitless injection so no help there.

Due to the way that back-pressure regulated inlets (which I am pretty sure the Agilent has) are plumbed, there should be flow out of the split vent all the time. Splitless transfer to the column is generated by rerouting the gas flows inside the plumbing - this makes it infuriatinlgy difficult to troubleshoot back-pressure regulated inlets.

Try a split injection - about a 50:1 split of a 1 mg/ml solution will give you 20 ng per peak which should be easily detectable.

Peter

[MSCHemist]

Not the transfer line it reads 280 (it's aux 1 on the 6890)

[Bigbear]

If you are seeing the solvent peak you should increase your filament off time.
When you say your compound peaks are 1% of what they used to be, are you comparing integrated peaks?

[MSCHemist]

If you are seeing the solvent peak you should increase your filament off time.
When you say your compound peaks are 1% of what they used to be, are you comparing integrated peaks?

Yep the peak heights are over 10 fold less as well.

[MSCHemist]

Still no change. At this point my only remaining hypothesis is a break in the column inside the transfer line or the column has active sites inside it somehow.