Why did my column head pressure drop?

[Carvone]

Hello all,

I would like to ask anyone that may be able to assist me. I am operating an HP 5890 GC with FID and a DB-5 capillary column (30mx0.25x0.25). My carrier gas is nitrogen. I have the flows set at 1ml/min for carrier, 40ml/min hydrogen and 375 ml/min for compressed air.

I was running splitless injection with 8 psi column head pressure, and this has been working great for months. I wanted to experiment a bit and try using split injection to see if I could get sharper peak. So I altered the method, and for the next sample run the injection port gauge for column head pressure dropped from 8 psi to basically 0 psi. I changed it back to splitless injection, and ran another sample, and the column head pressure did increase but only to 4 psi. I tried to increase the column head pressure by turning the knob, but it would not go back to 8 psi. I shut the GC down and tested the column at the detector end with solvent and gas bubbles were coming out of column end.

I do not know why the column pressure dropped, and how to get it back up to 8 psi again. Anyone that might know how to help, I would really appreciate any advice.

Much thanks,

[Don_Hilton]

Check for a loose column fitting or septum that needs replacing. Also make sure you have an adequate supply of carrier gas - the pressure to the back of the instrument is at least 20 pounds above that set for the maximum required for operation of the column - with flow going through the column. If you have the instrument any distance from the gas cylinder, you need the gas line to be 1/4" diameter tubing rather than 1/8" to avoid excessive pressure drop in the gas line.

[Don_Hilton]

Check for a loose column fitting or septum that needs replacing. Also make sure you have an adequate supply of carrier gas - the pressure to the back of the instrument is at least 20 pounds above that set for the maximum required for operation of the column - with flow going through the column. If you have the instrument any distance from the gas cylinder, you need the gas line to be 1/4" diameter tubing rather than 1/8" to avoid excessive pressure drop in the gas line.

[Carvone]

Thank you Don,

I will change the septum, and check for loose fittings. The pressure at the tanks is over 20 psi higher than what is set at the GC, and the tanks are immediately beside the GC instrument.

But what about the changing to split from splitless? Why would this have affected the column head pressure? Have you ever seen this before?

Much obliged,
Carvone

[Russ]

Did you change the port liner when you went from split from splitless? If so you may have a leak at the O-ring sealing the top of the port. Have you tried pressure testing the port? You might also check to see that the purge solenoid is working. You should hear a soft but audible "click" when it is turned on and off.

[Peter Apps]

Hi Carvone

A leak is the most likely explanation, and the septum is the most likely place for it. Also you need to have enough flow into the inlet to supply the column, septum purge and split. If the flow is too small the pressure bleeds away. I'm not sure why changing to splitless shuld have caused this problem, but I vaguely recall that there was something odd about the inlet plumbing on 5890s that used early versions of back pressure regulation

Measure the flows out through the split and septum purge outlets, and inject some lighter gas and see how long it takes to elute.

I have seen flow and pressure differences between split and splitless when the inlet liner had appreciable resistance to gas flow. Whether this would give you a zero pressure reading will depend on the plumbing.

Peter

[AICMM]

Carvone,

Before all this happened, were you running splitless/split? Meaning, you start with Purge A (or B) off and switch to Purge A on at something like 0.5 minutes? At what time did you turn Purge A or B on?

Also, when you tell Purge A or B to turn off and on do you hear the valve?

Best regards.

[Carvone]

Thank you all so much for your prompt attention Don, Russ, Peter and AICMM. I see your names quite frequently from viewing the GC forum, so I am happy to be receiving expert advice.

The original settings of the GC, when it was working properly, were just splitless not switching from Splitless to split after say one minute, although that splitless/split was what I changed it to to when the column head pressure dropped. The chromatogram did show that the peaks were much smaller and the void volume had increased from about 3 minutes to 5 minutes. The pressure at my tank is 70 psi and it is set at 36 psi at the GC valve, with the purge vent of the pyrolysis unit set at 15 ml/min. The split vent and purge vent both dropped to virtually nothing after I switched to the splitless/split set-up. And I don't remember hearing any noises when it switched over.

I will not be working on the GC to Monday now, so I will attempt the things you all have suggested then. I should mention it is a pyrolysis-GC set-up where the pyrolysis unit sits atop the injection port with a needle that pierces through the septum and a cap that holds it in, so it may not be a septum issue since it is not ever being taken out and pierced back in again.

I realize it is a leak somewhere, as a pressure drop almost certainly means that, I just can't figure out why me switching from splitless to splitless/split would disrupt this. I will try again on Monday, and inform you all of my findings. I am going to take the weekend to brush up on the ole' HP 5890 operation, service and reference manuals again.

Thanks again~!
Carvone

[Carvone]

For all those that replied previously and offered their experience, I want to say thanks, and the drop in column pressure issue has been solved.

It turned out that when I tried switching the diverter valve from sending carrier gas to the pyrolysis interface unit (which then went on to the GC inlet) to sending the carrier gas directly to the GC inlet, the column pressure increased again. This told me there was some blockage within the pyrolysis interface unit. I then disassembled the pyrolysis interface unit and cleaned out the tubing using high pressure air and a piece of capillary column. This did the trick, as column pressure was restored when I switched the diverter valve back to sending carrier gas to the pyrolysis interface unit.

I thought there might be some sort of blockage somewhere, but didn't think to check this last week. I am still not sure why switching from splitless to split/splitless would have coincided with this, but it suffices to say the problem has been solved.

Thanks to all,
Carvone

[AICMM]

Carvone,

Glad to see your problem is resolved. One comment to add though. I would strongly suggest you look at splitless/split. About a minute or so in depending on your pyrolysis process. Reason is that the chromatography will improve significantly as you sweep the injection port clear. If you cannot afford to do that because you pyrolyze in segments, may I suggest a low split flow? Perhaps 1:2 since that won't clobber your sensitivity but it will significantly improve your sensitivity. If sensitivity is going to be an issue we should discuss inlets and tricks there that will help while still allowing you to run small split.

In addition, with a 5890 in splitless all the time, you are running the valve very hot and subject to failure (in which case you are split at that point....)

Best regards

[Ron]

One thing to remember for the 5890 is that you have to have about 25 to 30 mL/min flow out the split vent to properly perform splitless injection. So using split mode will not increase the use of carrier gas if splitless injections are being performed properly.

[Carvone]

Thanks Ron for this suggestion. I had the split set at 15 ml/min (same as the pyrolysis interface purge vent), but will now try bumping it up to 25 ml/min. Hopefully this will help improve the chromatograms.

I wasn't running splitless trying to save carrier gas, but since I am injecting smoke in essence, I want it to be more sensitive detection. I am looking for qualititave identification based on a chemical fingerprint rather than quantitative analysis.

[Carvone]

Thank you AICMM,

I admit I did not notice your reply before as the email notification only told me about the reply from Ron.

I have since switched to Helium as carrier gas, increased my linear velocity to 30 cm/sec, and thanks to your advice started experimenting with splitless/split and split modes, and the pyrograms are much neater with a lower baseline, and better ability to discriminate between similar samples.

It sounds like you have experience with Pyrolysis-GC. Is that true? This is the first time I have used a Pyrolysis unit coupled with GC. I find it much more difficult to optimize, I must say!

Also, a sincere thanks to everyone for all their help and experience in this and all other threads. I am thinking about a GC and how the decisions I make affect it on a level like never before as a result of reading all the postings from so many knwoledgeable gas chromatographers. I am very glad I found this forum site.

Carvone