Lost sensitivity on both FID and TCD

[tonylorn]

Hi everyone,
I am a new GC user working with an old HP 5890 Series II with FID and TCD. We have a single injection port, and split the gas into two columns leading to the two detectors. This set up worked well for our mixed gas of hydrogen and chlorinated hydrocarbons, until one day, right under my nose, I saw the hydrogen signal disappeared on the TCD, together with the FID peaks.

I performed some trouble shooting procedures, but haven't confirmed the problem yet.
1. The FID and TCD still give non-zero, non-maximum signals (base lines). The baselines carry certain normal noises, and can jump up and down when the flow rate is changed.
2. I did not detect any leaks in the system. Flow rates look like normal.
3. While blocking the inlet purge line at the injection port, and injecting several micro liters of liquid methanol, the detectors still can not pick up any significant peak.

Overall, it looks like either nothing injected got into the column, or the TCD and FID lost sensitivity at the same time. It is a strange thing to me. Can anybody provide some insights? Thanks a lot!

[thohry]

Hi everyone,
I am a new GC user working with an old HP 5890 Series II with FID and TCD. We have a single injection port, and split the gas into two columns leading to the two detectors. This set up worked well for our mixed gas of hydrogen and chlorinated hydrocarbons, until one day, right under my nose, I saw the hydrogen signal disappeared on the TCD, together with the FID peaks.

I performed some trouble shooting procedures, but haven't confirmed the problem yet.
1. The FID and TCD still give non-zero, non-maximum signals (base lines). The baselines carry certain normal noises, and can jump up and down when the flow rate is changed.
2. I did not detect any leaks in the system. Flow rates look like normal.
3. While blocking the inlet purge line at the injection port, and injecting several micro liters of liquid methanol, the detectors still can not pick up any significant peak.

Overall, it looks like either nothing injected got into the column, or the TCD and FID lost sensitivity at the same time. It is a strange thing to me. Can anybody provide some insights? Thanks a lot!

You say the flowrates look like normal, how do you know that? Probably there a choke in the first part of the column (before the T). You can check by injecting methanol (or hexane) and see the signal goes up. If the retention time is longer than a normal solvent peak, then it is blocked.

[LDCCL]

did you check all flow rate was correct? 2, have you confirmed all gas proportation of fuel gas carrier gas, make up gas, were right, 3 are you sure no any leakage in the flow path, or blockage 4,did u check the filament if burnt or no power supplied, 5 are you sure the software is running well, do you met any abnormal situation compare than before, i would give you some suggestion only after you recheck it over

[tonylorn]

You say the flowrates look like normal, how do you know that? Probably there a choke in the first part of the column (before the T). You can check by injecting methanol (or hexane) and see the signal goes up. If the retention time is longer than a normal solvent peak, then it is blocked.

Hi thohry,
Thanks for the reply. I checked the flow rates with a pool of methanol (which gave a nice string of bubbles), and a digital flow rate meter at the outlets of TCD and FID. Thus, I'm sure the columns are still conductive. Could anything else be wrong?

[thohry]

You say the flowrates look like normal, how do you know that? Probably there a choke in the first part of the column (before the T). You can check by injecting methanol (or hexane) and see the signal goes up. If the retention time is longer than a normal solvent peak, then it is blocked.

Hi thohry,
Thanks for the reply. I checked the flow rates with a pool of methanol (which gave a nice string of bubbles), and a digital flow rate meter at the outlets of TCD and FID. Thus, I'm sure the columns are still conductive. Could anything else be wrong?

That way of checking the flowrate is not so correct. Have you checked the retention time yet ? Afterall, the things may stay in the inlet. Check carefully the liner, septume, etc...

[Ron]

How are you doing injections of the gas standard, with a syringe or with a sampling loop and valve? It appears that what is injected is not making onto the column if you still have the proper flow from the detector end of the column. Since the signal disappeared from both detectors simultaneously the problem has to be before the split to the detectors.

[chromatographer1]

Since it is unlikely that both detectors would fail at the same time then it appears that your sample is not getting to the detectors. So either you are not really injecting sample onto the columns or you have a leak.

Describe the column you are using, the pressure you require for the flow rates you are using, and the flow rate itself. Perhaps if you told us enough information it will become obvious what is wrong and someone will be able to determine your problem with some assurance.

My guess is that you may having a severe leak at the septum or that your Y is the most likely location of your leak.

Good luck,

Rodney George

[Ron]

I just reread one of you posts, and you measured the flow at the detector outlet. Did you make sure that all gases other than the carrier were turned off before you did this? I always prefer disconnecting the column and checking the flow directly at the end of the column.

[tonylorn]

That way of checking the flowrate is not so correct. Have you checked the retention time yet ? Afterall, the things may stay in the inlet. Check carefully the liner, septume, etc...

Hi thohry,
Unfortunately the flow meter is the best thing I have to check the flow rates. This 5890 II does not have the online flow reading function. I tried several liners and septa, and blocked the inlet purge line. I can not tell what the retention time is, because there is no peak at all for micro liters of methanol! I am a newbie in GC, but I think even there's a small leak in the system, I should have some peak, right? Is it possible to lose the entire methanol peak due to inlet parts problems?
I took the inlets apart, and nothing seemed blocked. The pressure at the column head also read OK (15 psi).
Thanks.

[thohry]

That way of checking the flowrate is not so correct. Have you checked the retention time yet ? Afterall, the things may stay in the inlet. Check carefully the liner, septume, etc...

Hi thohry,
Unfortunately the flow meter is the best thing I have to check the flow rates. This 5890 II does not have the online flow reading function. I tried several liners and septa, and blocked the inlet purge line. I can not tell what the retention time is, because there is no peak at all for micro liters of methanol! I am a newbie in GC, but I think even there's a small leak in the system, I should have some peak, right? Is it possible to lose the entire methanol peak due to inlet parts problems?
I took the inlets apart, and nothing seemed blocked. The pressure at the column head also read OK (15 psi).
Thanks.

With all the info you supplied, I can see the syringe is blocked or just simply that it can not suck in sample ! I have got this kind of problem several times.
Why? because once you still have some flowrate in the column (providing new liner, septum etc..), there must be at least one solvent peak.

[AICMM]

tonylorn,

2 Columns, packed or plot and what length? I am assuming two PLOT columns which usually would mean two 30 meter columns. Only 15 PSI? What is the flow out the end of each column if they are not hooked up to the detectors at all?

Best regards.