Gas phase analysis -Different peak areas for same gas stream

[DocHollywood]

Hello,
I have a 5890 series 2 GC with SRI software. Currently I am trying to calibrate the GC for methane concentration analysis in CO2. As the calibration gas I am using a 1% methane Helium balance tank. The lines are at room temperature.
When I tried to calibrate the GC recently with the methane gas tank, I got different peak areas for different runs with the same tank.
Thinking there might be pressure problems with the gas regulator I tried the same runs with a mass flow controller and still had peak area disparities. The only thing after the sample loop which might be a problem is that there is a metal bellows pump attached to the line. I leave it off when i am calibrating. Not sure what could cause the different peak areas. ??

Thanks
AD

[AICMM]

DocHollywood,

Are you allowing the sample to reach atmospheric pressure with every injection? A rotameter or a water flask at the exhaust of your sampling line is a good way to see if you are at atmospheric.

Best regards.

[DocHollywood]

Currently I am sampling a gas stream that I am flowing through a mass flow controller. I sample it at the pressure the gas stream flows, which shouldn't fluctuate that much because of the MFC.
What I am worried about (correct me if am wrong to worry about this?) is the time it takes for the solenoid valve to come on. Wouldn't it matter if lets say it comes on after 9 seconds, which is what it is set to come on at, that I took an extra second to press the "run program" tab after I inject the sample loop? If I sound like a novice, it's most likely because I am.
Thanks
Appreciated,
AD

[chromatographer1]

If the solenoid is for a sample gas shut-off valve, then you need to allow the gas loop to depressurize to atmospheric pressure (several seconds) before you operate the sampling valve. Otherwise you have a pressurized loop which contains more gas than it should. And do not expect that the MFC will accurately maintain a constant amount of gas in your loop, at least not constant enough to give you accurate sampling. Process GCs don't try to do that historically, they fill at a higher pressure and then stop flow so the loop pressure can equilibrate to atmosphere.

best wishes,

Rodney George
consultant

[DocHollywood]

Thanks chromatographer1,

I'm just talking about the solenoid valve for the split/splitless settings in the instrument. This Purge valve is currently set to OFF at 0.00 and ON at 0.15 (or 9 sec). When Off from what I understand the gas goes straight into the column in splitless mode. When it turns back ON it goes to split mode. Wouldn't how quick I press START PROGRAM on the GC then matter? Or am I totally off on this?
AD

[Bruce Hamilton]

I'll let others, more expert than I discuss possible causes, but I'd like to point out a couple of things I've found, which may helpful.

1. Splitless mode on a 5890 is usually much more variable than other modes. Do you need the potential extra sensitivity?.
2. The standard is CH4/He, but you're analysing CH4/CO2 - I'd expect significant differences in pressure for same flow - which could affect your calibration - but should still give repeatable areas for the standard.
3. The problem may not be that the Start + 9 sec is not constant ( it should be ), but the loop pressure is not at equilibrium before the start - as noted by others above.

I'd definitely want to place a simple Dreschel bottle with a couple of cm of water on the loop outlet ( no flowmeters or other junk ) to ensure constant, static pressure, injection volumes to start with, then move onto flowing systems after they give consistent results.

Bruce Hamilton

[DocHollywood]

Thanks,
I will try the advice given.
AD

[AICMM]

DocHollywood,

Please tell us how you have the loop configured to the splitless/split inlet. Assuming you have plumbed one leg of the valve into the gas line into the inlet, I would point out that, in splitless mode, most of your flow is running out the septum purge line and you are getting no improvement in sensitivity at all. Two things that would probably improve your situation. One, plumb the column directly into the valve and just do valve injections. If for some reason you cannot do that, try a 2:1 or 3:1 split injection, leaving the system split all the time. This way the flow path is predominately down the injection port which will improve your on-column transfer. (Also easy to try!)

Have you tried taking the bellows pump off an looking at the reproducibility?

Best regards.

[DocHollywood]

Thanks AICMM,

You all have given me great advice.
I have tried taking of the bellows. It makes no difference.
I will try the advice in the following order.

1)Make sure the pressure has stabilized to atmospheric with a bubbler at the end of the line.

2)Try changing the split

3) Plumb the column directly to the injection valve.

Thanks again
AD

[DocHollywood]

Guys,
Someone mentioned to me that it might be a faulty sampling valve. Before I re-plumb the GC to bypass the inlet port has anyone ever seen a faulty sample valve for gas phase analysis.
Also, can anyone explain to me why the solenoid valve needs to come on after 9 seconds? Does that mean I'm running in splitless mode for 9 seconds and then back to split?

Thanks
AD

[Ron]

If you have a sampling valve plumbed into the carrier in line you are always running in split mode no matter how the purge valve is set. Take a look at the diagram in the manual, and you will see that for a gas phase injection the sample is always split. I would set the instrument for a split injection, you don't need to have the purge valve switching and changing the flow path. I don't think that is your problem, but it might contribute.

[DocHollywood]

The problem is solved. I turned the solenoid on during all the runs. Before it was set --> Purge/Valve--> Time --> Off - 0 then On - 0.15 . Now it is set Time --> Off-0 On-0. I am getting consistent peak areas within 2% Max. I still do not get what the point of having different times of initiation for the solenoid valve. Either you want it on or off? Split or splitless?

[AICMM]

DocHollywood,

Now you are running split all the time. If you want to improve your sensitivity a bit, decrease the flow out your split vent (some) and you should get a bit better sensitivity. Don't take this to the extreme (like 0) because you need some flow out the split vent port.

Splitless/split works much better with a liquid injection way down in the inlet. You start splitless to get good sensitivity and then (usually about 0.5 min) you switch to split to sweep the inlet clean and improve the early chromatography. Thing is, it doesn't really apply with valve injections through the inlet.....

Good to hear your problem is resolved.

Best regards.

[DocHollywood]

Thanks AICMM and thank you to everyone that responded for sharing their extensive knowledge!
AD