Trace Oxygen With 6890 TCD

[KeithD]

I am running a natural gas analysis, and I am concerned with air contamination of my samples. I am running Hydrogen as a carrier and have very poor resolution of oxygen at low levels. I recently changed the TCD to determine if the filaments had been degraded, the first shot with the new detector produced an oxygen peak with an area of 8600 cts. and a nitrogen peak with 56000 cts. The second shot produced no oxygen peak with a nitrogen peak of 55500 cts. Where did the oxygen go? Is there some reaction between oxygen and the detector. I have no leaks as I have done extensive checking. Should I change my carrier to Helium? I am at a loss and any suggestions would be helpful.

Thanks

[chromatographer1]

Hydrogen can react with heated metal components. It would be worth a try to see if helium will make the problem go away.

but are you using a Mole Sieve Column? or a porous polymer, or a diatomaceous earth type column?

How much are you injecting? What size of the column are you using?

A little more detail will be helpful in trying to answer your questions.

It may have been a contaminated (incompletely flushed before sampleing) gas syringe causing your one time oxygen peak.

best wishes,

Rod

[KeithD]

I am using a Mole Sieve Column, I am not sure about my sample loop or column length, I will go check on that. But I am shooting a 0.499% air ref. with an ethane balance at 1/2 ATM, -15 " of mercury. The sample introduction involves a vacuum being pulled on the sample loop, then the sample is charged to the system. Nitrogen is consistant between runs, Oxygen disappears. I am going to try switching over to helium in the morning. I was just curious if the temp. of my TCD could have any effect on the reactivity of oxygen with the filaments. Thanks for your help.

[chromatographer1]

With the hydrogen carrier you may be reducing active sites on your mole sieve support which then may react with oxygen with later injections. If you can shoot a large air sample before your analytical run (perhaps a separate valve) your problem may then go away if you choose to use hydrogen carrier.

Just an idea.

best wishes,

Rod

[KeithD]

Helium appears to have solved the problem of the disappearing oxygen. Thanks for your suggestions and help.

[tdaigle]

Good Job! Glad to see Helium worked out for you. Is it me, or do you think Agilent should take the initiative to post and send out frequent updates pertaining to such cases? From a marketing standpoint they may gain the respect and trust from consumers that have to trouble shoot these incidents using their own resources (time, money, productivity)?

[chromatographer1]

While I work for a competitor of Agilent I would not criticize them in a manner which they might consider to be unfair.

There are hundreds of applications for Gc in the industry and there are a lot of ways to ruin an otherwise excellent application by small changes which may not seen harmful by the end user.

To expect a vendor like Agilent to troubleshoot every possible modification is unfair in my opinion. A vendor sells tools which the user applies to their work as appropriately as possible. No one knows everything about everything so mistakes or problems do arise at times.

It is, of course, appropriate to ask the vendor for help with an application or to post a message on a forum for assistance. The good vendor usually does share information when it is available and assist their customers in solving their problems. That is one reason I work for a company that has excellent techical service TO ANYONE, not just their customers, and it is FREE, no charges or contract needed.

Just my two cents worth.

Rod

[mcdonar2000]

I know this is a chromatography forum, but why not save yourself some heartache and use a Teledyne oxygen meter? They can measure accurately down to 1 ppm in nitrogen, hydrogen, and most gases.

[CIL-BAHA]

http://www.sepsci.com/chromforum/viewtopic.php?t=2443
"
I was told TCD's MAJOR PROBLEM was keeping the dual flows of detector reference gas constant.
"
This may help explain the reactivity of the more 'unreactive' gas.

[arie2044]

Hi
I think you waste your GC and a lot of work for calibration & analysis instead for using a very simply solution that was proposed by mcdonor2000.
the accuracy, repeatability and sensitivity you achieve with Oxygen meter
you will not see on GC

Good luck

[Ron]

My understanding is that the oxygen is measure as an impurity during the analysis of natural gas. While the oxygen meter may give better resulys if oxygen is the only analyte of interest, the objective here seems to be to measure oxygen as part of the standard analytical procedure without going to a separate instrument.

[arie2044]

morning to all
The appearence of O2 in first injection can result from poor purging of the inlet system to GC, we have 20% oxygen in the air, so it must be purged out very well.

[GasMan]

Keith,

I have actually worked on this 'problem' and we found that the oxygen will react with the hydrogen on any hot metal suface. The problem is NOT only with Agilent TCD's, it will be seen on any TCD manufactured out of metal. The answer has been given to you, use helium as carrier gas.

Gasman

[tdaigle]

We are using Helium with good results. I'm not sure why or who went off on tangents. The problem was solved. Thanks everyone.

TD