Agilent 7890 FID signal loss

[kahmark]

We recently purchased a 7890A GC w/ an ECD and FID (with an IRGA signal). Our FID worked fine for the setup aqueous sample checkout. However, we then installed packed columns. Our FID signal decreased to a maximum signal strength of 10-15 pA for a 1.5 ppm CH4 std. MUCH lower than I anticipated. I should note that at first the signal was high (100's of pA) but I turned the column around to test something else and lost the signal strength. We use a PP-Q column. Is there a chance I shot column material into the jet? Should I clean it? My flows are as follows: H2 30, Air 400, N2 makeup 25, front inlet flow of 21.

Any suggestions?

THANK YOU AS ALWAYS FELLOW CHROMATOGRAPHERS!

[GasMan]

If you are using packed columns, you do not need such a high make up flow. The total carrier gas and makeup flow through the 7890 FID should be about 30 ml/min, so using a flow of 21ml/min, which for a 1/8 packed column is a little on the low side, would require a make up flow of 8-9 ml/min.

A signal change of 10-15 pA for 1.5ppm of methane would be normal.

Gasman

[kahmark]

Thank you Gasman.

I'll add the following measured info. We split our sample after injection, 8ml/min to the IRGA and 13.5 to the FID. FID out with only makeup and carrier is 39.4. Do you think I could be diluting the sample too much prior to the FID? Should I still backoff on the makeup flow?

Thanks

[Consumer Products Guy]

Make sure that the carrier flow is actually coming out through the detector, maybe you have a leak at the inlet.

Interesting, a spankin' brand-new technology GC yet using packed columns....

[thohry]

Probably with packed columns, the peaks are more broadened so peak heights are reduced compared to capillary ones

[kahmark]

Thanks Guy, we analyze greenhouse gases and the packed columns resolve these gases quite well. Capillary columns have a higher plate yield but the packed columns are true. Would you recommend a capillary column for such work? Why? As far as I know, we are one of a few research institutes to use this automated GC technology. I don't see the benefit of switching to capillary columns do you?

What would you expect for a peak area/ht for a 1.5 ppm sample of CH4?

Carrier flow is exiting the FID. What is the best way to check for leaks when using N2 as a carrier?

[Consumer Products Guy]

Carrier flow is exiting the FID. What is the best way to check for leaks when using N2 as a carrier?

Soap bubbles (e.g. "Snoop"). I'd try to route to the detector(s) individually first before you try splitting.

Packed columns are still used commonly for gas analysis.

[kahmark]

I should note that we split the FID injection with a Valco Y. Presumably, half goes to a Licor and half to the FID. So, I removed the split column and put in another without the split to the Licor. I did this to take out the potential problem of the split column out of the picture. I see no increase in sensitivity. Any ideas?

[aldehyde]

Carrier flow is exiting the FID. What is the best way to check for leaks when using N2 as a carrier?

Soap bubbles (e.g. "Snoop"). I'd try to route to the detector(s) individually first before you try splitting.

Packed columns are still used commonly for gas analysis.

Official Agilent policy is no snoop (which is funny since snoop can be ordered from Agilent). Use 50/50 isopropanol/water. It bubbles just as well, the surfactants in snoop can really screw things up.

[kahmark]

I agree, I ran into this last year with an ECD. Snoop screwed me up for days after it aspirated into the system near the detector. I use straight up iso. Thanks for that comment.

Still having issues. I wonder if some of the packed column moved into the jet?

[aldehyde]

I once had a weird FID issue where the signal suddenly went way up and didn't look like it was coming down. I looked at the jet and it was OK, restarted the instrument and then let it bake out for an hour and it was fine. The only real theory I had was that I had somehow pushed some of the material up out of the packed column but my coworkers were very skeptical so I'm not really sure what happened. This was a 6890 and it happened quite awhile ago so I don't remember many of the details but as far as I know having material come out of the column is pretty unlikely the way they are packed.

I did run into an issue one time where there was a small piece of broken capillary column sitting next to the jet. The FID signal was very high and jumpy, I found the piece and after that it settled down but its likely that if I had just let it sit for a few hours at high temperature all of the phase inside the little piece would eventually burn off and things would be OK again.

It is possible that you have a restriction in the jet which is causing a smaller than expected response. Changing the jet is very easy, if you haven't done it before follow these instructions:

Cool the FID to below 100 C, unscrew the 3 screws holding the top of the FID on to the base (this website shows an exploded parts view including the 3 screws I'm talking about http://www.cobertassoc.com/Agilent-6890 ... -Detec.gif just unscrew them and pull the top of the FID off as one piece). Use a nut driver to unscrew the jet and then tweezers to pull the jet out. Put a new jet in and use the nut driver to tighten it (don't have to go crazy getting it as tight as possible just make sure its securely in there). Place the top of the FID back on and make sure you have it assembled correctly by checking the signal, if you see something between 2.0-0.0 pA its on there correctly. If you see 800089898729 or some other large number starting with 8 you have placed the top of the FID onto the spring that transfers the signal to the electrometer and you need to wiggle and adjust the top until it goes into place correctly.

Once its back together and screwed in tight (those 3 screws help ground the signal and its important that they are tight) heat it up to 300+ and let it sit for awhile to burn off any finger oils or other stuff that is on the jet and the signal should come back down to what it was before. Once it is equilibrated properly run your standard again and see if its any better. If the peaks are still small the problem is 99% the sampling technique (ie the syringe), the inlet, or your Y connector.

Where did you get the standard you are using? Did you purchase it or make it yourself?

Agilent has two FID standards that are very useful for troubleshooting. They both come in blue boxes containing 3 amber ampules of hydrocarbons (C14, C15, C16) in hexane and you should have some from when the GC was installed. I'm not sure if the method would translate well to your packed column but if you have the column that came with the GC (HP-1 30 m x 0.25 mm x 0.25 um) you could install it and try the Agilent standard. If you try this it would be much easier to tell where the problem is and how severe it is.

Let us know if you need more help .