Gas Phase Sample Injection

[MDB-NCSU]

I'm in the process of setting up a refurbished HP 5890 Series II GC with a TCD and an FID detector. I want to convert one of the injectors to do gas-phase sample injections, rather than having both be liquid. Samples are currently taken using a Valco 16-port valve with 10mL sample loops. The samples do not need to be heated before injection and are at ambient pressure.

My understanding is that I need to purchase a 6-port valve and install it just before the injector by cutting the carrier gas line. I'm a little confused about the rest of the plumbing though. Do I need to purchase another sample loop that spans two of the ports on the 6-port valve to load the sample onto before injection? Or could I just plumb the 16-port valve in place of that sample loop and advance the 16-port loop when its time for the next injection?

If I need the additional sample port, how do I verify that I've loaded my sample from the 16-port valve on the extra loop before injection?

If I can just use the 16-port valve in which the samples are collected, how can I verify that I won't be overloading my column with sample? Would I need to most likely use a split injection?

The intention of this work is to look at dried combustion products, so the species of interest are CO, CO2, O2, N2, and light hydrocarbons. I plan on using a Shincarbon ST 80/100 column with the TCD first, and see where I need to go from there. I am new to chromatography, so please excuse my ignorance. Additionally, this work has a shoestring budget, so I'm trying to keep the costs down as much as possible.

Thanks!
Myles

[chromatographer1]

Myles,

You said you have a 16 port sampling valve with 10 mL sample loops. Really? or is it 10 microliter sample loops?

Do you a dual liquid inject with a single valve now?

Sorry I am confused. The sample loop in a liquid sampling valve is in the rotor. I don't think a dual inject of both a gas and a liquid sample with one valve is possible, but new solutions are appearing all the time. See www.vici.com

Perhaps if you could give a drawing of the plumbing your situation might be clear.

Micropacked columns with a 1mm or 0.8mm can overload with a sample that would not be a problem with a 2.1mm ID column.

I hope your problems can be resolved.

best wishes,

Rod

[GasMan]

Does this diagram that can be found on the Valco site, http://www.vici.com/support/app/st_japp.php,solve your problem. Instead of the 4 stream valve shown, you would substitute your 16 stream valve.
10 ml loops seem a bit large these days, but this depends on what levels you are trying to measure. If you have high concentrations of oxygen and nitrogen, you will not get good separation on the Shincarbon column, despite what you might see on manufacturers web sites.

Gasman

[AICMM]

MDB-NCSU,

What is sounds like to me is that you have a multi-port valve that is collecting aliquots of the reactor gas for subsequent analysis. Rod is right, a picture in this case is really useful.

If that is the case, then I would still advise a downstream (of the multi-port) gas sampling valve to do gas injections on the GC. The big issues that I see, at that point, is how to get the sample from the 10 cc loop to the gas sampling valve. If the reactor has some pressure, then this will probably not be an issue since you could pressurize the 10 cc loops and then allow each of them to expand into the gas sampling valve (as long as you keep the lines short.) (Drop ball method or what I affectionately call bubble-bubble-bubble.)

In your case, I would not plumb into the current liquid injection ports at all, I would plumb the column straight to the valve (with a short piece of transfer line.) Two reasons: 1) you avoid having to worry about split ratios and what and 2) the column you are planning on is packed and probably best not plumbed to 5890 splitless/split inlet.

Also, remember water, always remember water when using a TCD and/or a ShinCarbon.

Sounds an awful lot like a project I worked with doing syngas (pyrolysis) gases.

Best regards,

AICMM

[MDB-NCSU]

Thanks guys for your quick responses!

Took me a while to figure out how to get images posted to a forum, but I think I figured it out.

Firstly, I actually have 5ml sample loops, not 10 as I thought at first. And AICMM is correct, I am collecting aliquots of my reactor gas for later analysis. They unfortunately are not pressurized.

Rod: Here is a picture of the two injectors currently installed on the gc.
http://i1121.photobucket.com/albums/l50 ... hoto31.jpg I believe the one on the top is the split/splitless injector, which will be used for another project utilizing liquid injections with an autosampler (removed for photo). The bottom injector is the one I would like to use and is for for a packed purged column per the front of the instrument http://i1121.photobucket.com/albums/l50 ... photo5.jpg

I am wondering if I can plumb my sample loops like this: http://i1121.photobucket.com/albums/l50 ... tvalve.jpg

or if I should plumb it like this: http://i1121.photobucket.com/albums/l50 ... tvalve.jpg in which case, how do I ensure that I get my sample from the 16-loop aliquots loaded onto the sample loop in the 6-port valve?

Gasman: the link to the valco site in your post seems to be broken for me. And yes, my sample is products of combustion, so they will be appoximately 70% N2, 10% O2, and 10% CO2 with other smaller concentrations of CO, and unburned HC's. I am interested firstly in separating CO and CO2, and hopefully O2 (as a measure of the completeness of the combustion) with some HCs (CH4, H2, maybe C2H6) using the TCD. I would then like to separate the HC's (which I can switch to the FID and another column if need be). I had previously spoken with a ResTek rep and he recommended the shincarbon column. I'm not tied to it by anymeans if there is a better column you can think of.

AICMM: I like the idea of keeping the injection as simple as possible with the plumbing directly into the column, but I'm worried about overloading the column with sample. I'm worried that the N2 (and O2) may swamp the rest of the species in my sample. I may not be able to avoid this without using a split injection or a smaller sample loop as in the second diagram shown above.

Thanks guys!
Myles

[GasMan]

1.Go to the Valco site at http://www.vici.com
2. Goto the top of the page and place your mouse over 'Products'.
3. Goto the column named 'Valco Valves' and click on Selectors
4. On the right side under 'More Information' click on 'Selector Applications'
5. In the resulting window, click on 'Collection of fractions at timed intervals for later analysis'.

I think this will do exactley what you are trying to do.

By the way, I do not work for Valco.

Gasman

[GasMan]

I have now had a chance to look at your pictures. The way that you need to plumb your system is as in your third picture posted. This is the same as the Valco site.

I noticed in your second picture posted that there is a valve plumbing diagram. Can you post that? Do you know if the GC at the moment is still plumbed up as in the diagram?

Gasman