FID major failure

[drbmoney]

Hi all,

My first time posting and it is mainly out of desperation - I hope you kind folks can provide some guidance since I'm at a loss.

I've used this FID for going on 5 months now and it's been working fine. Two days ago, while performing a run by gas-sampling from our manifold, the signal shot up to what appears to be the maximum output for the detector (it runs off the screen in my analysis software) followed by a drop down to 0 (not previous baseline, but 0 units). After this, the signal increased back up to about 10^8 and has decayed to about 10^7, where it is oscillating between 0.75-1.25x10^7. I've never seen anything like this, but figured it must've been contamination. We installed a new piece of swage line in the manifold connected to this GC-MS and perhaps it wasn't cleaned prior so when I sampled via sampling valve, I introduced it to the column. No changes to makeup or fuel for the FID were made.

Since then, I've cleaned the collector plate/insulating disks/jet by sonicating in soapy water (no change), I've turned off helium makeup to the FID (no change), I've ruled out air/H2 since another GC using the same tanks and lines is working fine, I've disconnected the column entirely (no change). Right now, the output is about 1000-1200pA (I hear normal is about 0-20pA) and not really showing any signs of going down in the short term. I've baked the detector to 400 degrees C and likely stripped the section of column in there and only saw a minimal drop in the output. Has anyone ever seen anything like this? The interconnect spring doesn't appear damage, no parts appear visibly soiled.

I'm at a loss. Agilent says all signs are pointing to something with the electronics, which would suck. Any help would be super-duper-mega appreciated. Thanks!

[tkubowicz]

Hello

I'd recommend:

1.Plug detector with blank nut - if you don't have one put piece of column and block it with old septa
2.Set up detector temp to 300°C, Air to 300ml/min, Hydrogen to 30ml/min, leave makeup gas off
3.lit the flame
4.Check signal output

If you still have high signal try to switch electrometer off and see what is the output ("Config" - "detector"-"electrometer"- "off/on")

Regards

Tomasz Kubowicz

[rb6banjo]

I hate to say it but I fear that Agilent is correct. I'm afraid there's not much you're going to be able to do about this one.

[Peter Apps]

It does sound as if the electronics are sick - what happens if the flame is not lit ?

Peter

[drbmoney]

Just to update for the record: I've ordered a replacement jet, collector body and spring. I'll try replacing those parts (<$200) before having a tech come out ($1450+parts). We shall see!

[duck duck goose]

One piece of advice, do not clean an FID with soap and water. You should use something a little more volatile like methanol or IPA. First thing I would do is sonicate the parts in water to remove any possible soap contamination, then sonicate in methanol or IPA. After sonicating, put the parts in a beaker and cover with tin foil and put this in the oven at 100 degrees for 10 minutes then reassemble the FID.

I doubt this will work because when the signal typically maxes out it is almost always an electronics issue, but it might...

[MSCHemist]

Sounds like either a contamination or electrical short error. I've seen that after cleaning. As suggested use volatile solvents. It will go away not too long after heating up the detector. If it is an electrical issue either there is a piece of metal in there somewhere or one of the insulation rings is broken/deteriorated.

Use water saturated with citric acid to remove corrosion. Use normal solvent to clean like methanol or isopropanol. Acetone, hexanes, and other such solvents are ok for metal parts but don't get them on the insulation rings (similar to GC/MS source).

[drbmoney]

Hey guys, here's a positive update! It wasn't the electronics! After troubleshooting damn near everything we could think of, my PI and I sat down and ran through all the other possibilities and both of us went, "Hey, maybe the traps are bad?" And then we both said "Nah!" because how could the traps be causing a baseline profile that hits maximum and/or is operating well above the range of normal scans?

Lo and behold, it was the hydrocarbon/H2O trap on the hydrogen line. I guess it should have clued me in to this when I noticed that the initial signal when adjusting H2 flow was essentially linearly related to H2 flow before it started to drift, but I honestly didn't even consider this a possibility - sure, a slightly higher baseline, but completely destroyed, wobbly and >500pA? I guess I learned something new.

All the electronics are safe and sound, and I got a new jet and collector body out of it so hey. Also, I find it surprising that so many people are advocating against soapy water when Agilent's procedure calls for the parts to be sonicated in water with detergent in it. In any case, after the sonication I rinsed thoroughly with DI water, followed by methanol, followed by drying in the oven so I never really thought that was the issue. Now I'm just baking out the columns since all of the temperature cycling etc. led to a lot of column bleed, but the problem is essentially fixed. Check your traps, kids!