Column for TCD

[shanbfh]

Hello GC experts
I need expert advice for choosing a column for TCD analysis. I want to analyze h2, ch4, o2, n2 and co2 from biogas samples and we have an agilent 7890 gc. we are currently using DB-624 column for FID (mainly ch4 and ethanol analyses). and we have an extra HP-5 column from agilent but i do not know if i can use for TCD?
could some one suggest please..`?

thanks in advance
sh

[muGC]

Agilent has a biogas GC solution: http://www.chem.agilent.com/Library/dat ... 9517EN.pdf

To be able to separate and detect all of the permanent gases you have listed, you need to have a complex valving system with mol sieve and plot columns. Using simply one column from an inlet to the TCD will not work.

[AICMM]

shanbfh,

If you can live without baseline resolution of O2 and N2 a Shincarbon will separate all of those components in one run. I would suspect that a Carboxen Plot would do so as well (in deference to Rodney.) The more you shoot the poorer your O2/N2 resolution. You will pay a price in all of the other stuff that will eventually come off the Shincarbon.

If you cannot live without baseline separation of the O2 and N2 then a mol-sieve in parallel with a porous polymer would separate everything to baseline. You will still pay the price with late eluters and the CO2 and water will pile up on the sieve (you can manage this if you are careful.)

Now the bad news. You probably won't be able to detect everything on your list with a single TCD (the forum is chocked with notes about this problem.) If you go for hydrogen, then you use argon carrier and your sensitivity to O2, N2 etc is not that great. You go for O2, N2 then you use helium carrier and you won't have fun measuring hydrogen. Your options are two TCD's on two channels (one argon, one helium) with two injectors or an HID with either a Shincarbon or parallel column configuration.

Look around in the chromatography forum and you will find lots of discussion of this problem.

Best regards,

AICMM

[muGC]

Basically you need a standard/refinery gas analyzer that many vendors sell. Otherwise you're going to get a false representation of your sample. Better to do it right the first time. The Agilent application I suggested is just a micro/fast version.

[shanbfh]

Hi guys,
thank you very much for you answers...i guess i have to read some more to understand your suggestions ..but it was of great help !!

thanks very much..i will ask for more suggestions..

br
shane

[AICMM]

shanbfh,

I would argue that if you know what you want to do, then you don't need to buy a packaged black box from someone. If you have more money (~20 - 30 K) than time or don't want to mess with it, then, a packaged RGA might be your better solution. (All of which begs the question, if you buy a black box, do you really know about your data anyway, but that's a topic for another day.)

You have a perfectly good chassis for a starting point for this application so, if it were me, I would make this chassis do what it needed to do. On the other hand, I work on GC's every day so...

Best regards,

AICMM

[GC3]

Shanbfh,

This overhaul involves valves, auxiliary oven, aux epc, heaters, and packed columns. I've built more permanent gas analyzers than I can count. Depending on the concentrations you can do everything on one TCD channel. It would save you a lot of money to modify your existing platform. If your serious about doing this give me a call.

[shanbfh]

Dear all,

thank you very much for your valuable information and time. we are going to decide upon this soon with the help of agilent GC expert ..I could talk to the tech guy with the help of some background info you guys gave me..

thanks so much for your contribution..

best regards
shan

[aidnai]

My 2 cents: I agree with one of the statements above that it is unlikely to be able to measure everything with one TCD, for the reasons stated -- whichever carrier you choose, you're going to have a hard time differentiating analytes with a similar mass to the carrier (i.e. helium/h2, argon/air). HID was suggested, but perhaps DID or PDID is more suitable.

Your other problem is separating everything without trashing your column -- The solution I have used in the past is actually columns in series, rather than parallel. You inject the sample onto a column (Porapak, PLOT or whatever). Non retained components (permanent gases) blast through and get to the second column, a molesieve. At this point you have a 4-port valve programmed that isolates the molesieve while the heavier stuff elutes (CO2, water, everything else). After everything is done eluting, you switch back to the molesieve and let the permanent gases elute.

As mentioned elsewhere, you don't want to load up a molesieve with CO2, moisture, etc. It will wreck the chromatography for everything else and require a lengthy bakeout (overnight is pretty standard).

VICI is a standard supplier of the 4-port valves along with actuators (either pneumatic or electronic). I've been buying packed columns from Ohio Valley, though there are many other good vendors I'm sure.

Good luck,