Rising baseline with FID - very tricky one

[Johnny Rod]

Really scratching my head over this one...

We have an Agilent 6890 which has been in use for a number of years, split/splitless inlets, FPD at the back and FID at the front. The front column is a GS-alumina PLOT 0.53 x 30m. We use this to determine hydrocarbons C1-C5 as follows:
oven 30C-150C ramp
Helium constant flow at 10ml/min
Split injection 10:1
injection via automated valve at 120C

All this time the baseline has been totally flat as the oven temperature increases. We have just moved this instrument to a new lab about 30mins away, and there is now a rise of 10pA as it goes from 30C to 150C (this starts to drop away a little as it maintains at 150). We've tried:

column conditioning
individual cylinders of all gases vs. pipelines (pipelines were checked for air and moisture before use anyway)
swapping column to back inlet
Swapping column for 0.5m of a DB5
changing liner and septum
Changing FID jet and sonicating collector and insulators in methanol
running without injecting

I haven't got a chromatogram uploaded yet but will try to do this later to illustrate. I've also searched on here for similar problems and found nothiing that has made any difference. Any thoughts? We have yet to take it back to the old lab out of mindless optimism, but we're nearly there!

[CE Instruments]

From your trials etc. It is not column, GSV, FID etc.

If the signal goes up with the temperature ramp it would tell me that there was something coming in with the carrier and as the volume increases to maintain the flow the contaminent also increases. Are you running any filters in the Helium line ? If so remove. Tried running with either Hydrogen or Nitrogen carrier?

[GasMan]

You mention that this is a new lab, so is this also a new gas supply installation? If so, were instructions given on what cleanliness you expected on the final installation and how this was done. Many years ago we moved into a new lab and gave instructions that the tubing should be cleaned to remove any flux used when making soldered connections. They used trichloro ethylene. It removed the flux but was of no help when we tried to run EC detectors!!

I would make sure that you have filters on the gas lines.

Gasman

[Johnny Rod]

Yep all new pipework and cylinders. Tried nitrogen instead. Agilent Big Universal Trap in helium line - tried without it, also tried a different cylinder as with the other gases, straight from regulator to back of the GC.

Chromatogram here:
http://www.johnnyrod.co.uk/chrom.jpg

How can I check the FID temperature? It is supposed to be 300C, which is rather high but that's what we've been using. Where can I stick a thermocouple to check it though?

Thanks for the input so far. Any other thoughts?

[asm]

Could you please upload an old chromatogram too?

This one you linked seems pretty normal to me.

I would just bet that the displayed peak is much smaller so the "hill" in the baseline in the baseline is just better to see.

[Johnny Rod]

I don't have one to hand but the baseline normally doesn't rise more than 0.1-0.2pA across the run. It's an inorganic stationary phase so there isn't really any bleed as you would see in a silicone/wax type phase

[CE Instruments]

Only ever seen this once but could explain your problem. We had an issue with one of our total Nitrogen systems, new install to existing lines. High Nitrogen background, no matter what we did. Traced to a moth getting into the gas line !

As its not your carrier, not the filter it must be the gas lines. Try a gas cylinder by the instrument on a short new line.

[Johnny Rod]

After much mucking around I ended up swapping the FID for one from another GC. This seems to have done the trick, so I guess the old one has somehow become contaminated - this slowly desorbs from teh base of it (somewhere) as the column oven warms it. Not exactly a definitive answer and maybe still points to the gas lines in the new lab. I'll have to try cleaning the old one and putting it into the other GC to see if it still has a problem.

As an aside, before swapping we did notice that the RTs had lengthened by around 15-20%. We also found that changing the make-up gas flow changed the RTs quite a bit. I have already replaced the FID jet so I would think there are no blockages, otherwise any thoughts on this? It is a dedicated capillary FID so 0.011" jet tip with 300ml/min air, 50 of H2, 25 of N2 and 10 of He going through it.

Thanks as ever for your assistance

[dblux_]

... It is a dedicated capillary FID so 0.011" jet tip with 300ml/min air, 50 of H2, 25 of N2 and 10 of He going through it.
...

In fact the stream of air doesn't go through the jet.

[Peter Apps]

After much mucking around I ended up swapping the FID for one from another GC. This seems to have done the trick, so I guess the old one has somehow become contaminated - this slowly desorbs from teh base of it (somewhere) as the column oven warms it. Not exactly a definitive answer and maybe still points to the gas lines in the new lab. I'll have to try cleaning the old one and putting it into the other GC to see if it still has a problem.

As an aside, before swapping we did notice that the RTs had lengthened by around 15-20%. We also found that changing the make-up gas flow changed the RTs quite a bit. I have already replaced the FID jet so I would think there are no blockages, otherwise any thoughts on this? It is a dedicated capillary FID so 0.011" jet tip with 300ml/min air, 50 of H2, 25 of N2 and 10 of He going through it. Is there a chance that the GC was changed from constant flow to constant pressure - that would give both a retention shift and a baseline drift. Swapping the FID should not have made a difference though

Thanks as ever for your assistance

[Johnny Rod]

dblux you're right, I just checked!

Peter, no, same method but checked it over again as well. If I can find some concrete info then I'll let you know.

[Peter Apps]

Another possibility - the clearance between the tip of the column and the inside of the FID jet might be quite small with a 0.53 mm column, and then a change in the make-up gas flow, a small difference in jet diameter, or a difference in how far the column is inserted into the jet could change the back-pressure on the column.

How about swapping jets between FID bodies ?, testing what happens if the column is installed with its tip below the tube of the jet ?, or deliberately changing how far into the jet the column is inserted ?

Peter

[Johnny Rod]

Hi Peter, I wondered about that, as the column flow is high for capillary. On fitting it this time I made sure it was pulled back more than a little from the jet end.

I did find that running the GC this morning, after the weekend, the baseline has slowly wandered up and down. The column oven is kept at 150C on standby with all heated zones hot, but with the FID unlit. Perhaps this means some more rubbish has come in from the gas lines, but they are a lot cleaner than when the GC was first moved here. Not yet had time to try cleaning and refitting the old FID to another instrument. We did try it on cylinder gases but it could be that whatever is in there and causuing the baseline rise is going to take a long time to bake off so simply changing the gases did not elucidate the problem as it was not going to go as quickly as it arrived. Perhaps.

[Johnny Rod]

A follow-up for this in case anyone else finds it useful or is curious. We found no leaks but it looks like some contamination from the new gas lines was the problem. I don't like ot be so vague but we don't know what it was. The replacement detector settled down in the end but the baseline noise values were something like 25% higher than pre-move. This seems to be slowly improving so am going to give it a few weeks and looka gain.

I took the old detector to bits and sonicated it (weldment and all) in hexane then methanol, to try to cover all eventualities. I fitted this to our GCMS, ran some samples, then moved the data across to Chemstation on the original GC. The noise figures are very similar to before though a little higher, so the cleaning seems to have worked. The other GCs we moved in were some time later and don't seem to have had any problems.

Frustrating to say the least! Thanks for your help.

[jdezeeuw]

You stated that the retention times changed when you changed make up gas?

That's should absolute not happen. That means that you have a restriction in the detector that allows to built up a back pressure. You need to find that restriction as it may also be part of your problem

jaap de zeeuw, Restek Corporation