Handling gas tight syinges

[karthic]

I am using this SUPELCO PLOT column for gas analysis in my samples which generally contain CO, CO2, H2, acetylene, ethylene and ethane. I am facing a trouble in calibrating the GC using the standard gas mixes (that I got from Airgas). I figured out that I have a sampling problem. I usually take 100 ul of my standard gases in a gas tight syringe ( Hamilton). Now, since the gases are high pressure in cylinder, I am getting pressurized gases in my syringe and thus my peak height is big, leading to erroneous calibration.

When I called the Hamilton folks, they asked me to vent the gases for a few seconds before injecting so as to get the gases at atmospheric pressure. I tried to double check that with by pressurizing the bottles (serum bottles) to 14.7 psig (1 atm) and injecting those samples into the GC but I am still getting different peak heights.

Also, in my fermentation experiments the pressure keep varying from 20 psig to 5 psig and I will have to analyze the gas consumption on an everyday basis.

Could you please suggest me a sampling technique which could solve the problem.

[MikeD]

An in-line gas sample valve with some pressure regulation on the inlet and the exit bleeding to atmosphere is the best way to introduce a gas sample at a consistent volume and pressure, as many will confirm here.

If you don't have a GSV what I would do if possible is to vent a small flow of calibration and test gases through flexible clear silicone tubing dipped in a small beaker of water to prove venting. Pierce the tubing with the syringe and draw a sample that way. There are other variations, depending on how much gas you have to play with. Then don't inject too quickly through the GC septum because the sudden pressure rise in the inlet can cause blow-back either around the needle or the plunger.

Apart from the larger uncertainty of injection size the problem in my experience with the manual syringe method for gases is that some entrained air always gets in. If you tried it with nitrogen containing say < 10 ppm oxygen you would probably detect extra oxygen equivalent to 1-2 % air contamination at best, whereas the GSV method should show better than 0.02 % air contamination

[chromatographer1]

I concur. Gas sampling syringes are not ideal and success is very dependent upon one's technique.

Some means of drawing a gas sample through a sampling system into and through a valve with a sampling loop that can be brought to ambient pressure without air contamination must be devised if an accurate measurement is ever to be attained.

I know it has and can be done.

A little common sense will be necessary. It may be necessary to have to purge gas available to 'clean' out the sampling system before it is refilled.

Worst case: experiment and develop a good technique in using the syringe.

best wishes,

Rod

[AICMM]

Karthic,

If I understand your post correctly, your fermentor has a positive pressure. If you install a gas sample valve (which we all recommend) and a shut off valve upstream and then you put the output line of the GSV in water, you will have an excellent sampling system. Open the upstream valve, watch the bubbling in the water, turn off the upstream valve, watch the bubbling stop, inject your valve and start your run. Very simple, very straightforward, and very effective. Much better chromatography. Pretty inexpensive to boot.

Best regards.

[karthic]

I really appreciate your suggestions. I will have to talk to my research coordinators about the gas sampling valve.
I am adding some more information about the fermentation process. I am working on serum bottles currently before I go to bench scale. I start with 20 psi on these bottles and I have biological catalyst which consume these gases. Thus, the pressure decreases over time.

Again. Thanks a lot.

[Bruce Hamilton]

OK, as noted above gas sampling valves are the way to go, whatever ths sample is in, and whatever the pressure. The normal flow through systems above work fine when you have large, or on-line, samples

Obviously, if your sample is precious/limited in size/low pressure, then the solution is to run the sampling loop on the GSV at reduced pressure.
The small sample size also reduces the effect of reduced sample pressure with each large sample, which may help other analyses on the same sample.

It's not that hard, you just need a vacuum pump, a vacuum sensor, a tee joint, and a toggle shutoff valve. I use a 0 - 1000 mm Hg low-internal-volume sensor ( from Sensym ). The critical aspect is knowing what the pressure in the loop is just prior to injection.

There is one trick worth noting. Small diameter sample loops are notoriously difficult to vacuum purge of heavy residual gases, so I use a small ( 1 cm3 ) SS chamber volume on the other side of the loop, and a six port sampling valve.

The chamber allows molecular streaming through the loop, and helps purge the loop, as well as giving a more stable sample pressure. The easiest way to make such a chamber is to use a 1/16" to 1/2" swagelok reducer combined with a 1/2" plug.

You hook the pump to the toggle valve, and the valve to the tee, with one arm of the tee going to your septum bottle, using a luer lock needle blocked by a septum, or preferably a small needle valve.

If you use the simple block by a septum method, you should then put that septum against the vial sepatum, and push through both septa when collecting samples. You can also get sample bottles with simple valves on them, which I prefer to use.

Never use the vial septum as the blocking septum, as Sod's law states the most valuable sample will be the one where you push the needle too far, and the whole sample is evacuated.

That means the evacuation sequence is :-
pump->toggle valve->Tee->pressure sensor->sample loop->1cm3 volume.

Withe the needle blocked by a septum and the sampling valve in the fill position, you pull a vacuum on the whole system, close the pump off, vacuum should stay at zero. Then gently push needle through into the vial, until you have a suitable pressure in the loop and withdraw into septum. Wait a couple of seconds for the pressure to stabilise, record pressure, and inject.

One advantage of this system is that you can easily adjust the sample pressure if your peaks are too small or large, and don't have to change loops so often.

Note that if you can use large samples, the flow through method mentioned by others with a small water head at the outlet is very accurate, without the need for a pressure sensor or vacuum pump. The above method is good for grab samples at near atmospheric pressures, and for small samples.

Please keep having fun,

Bruce Hamilton