Chromatography Forum

I'm new to using the TCD and have previously used a helium-flow method to measure CO, CO2, N2, CH4

I'm currently working on a project where I need to accurately measure hydrogen and carbon monoxide (CO2 and CH4 would be good to know but are not the focus). I realize that I should switch from helium-flow to either argon or nitrogen but I thought I'd ask for some input before I start working on critiquing a method. What method would you recommend I use? Flow? Column temp(s)? Detector Temp? etc.

I'm using a 100/120 Carbosieve S-II, 10ft, 1/8 in diameter, 2.1 mm SS column
and an Agilent 5975C series GC

I'm not sure off the top of my head what make/model the TCD detector is but I can find out if necessary.

I realize it could be easy (and perhaps more precise) to use helium flow for CO and another method and gas-flow for H2 but I am new to this and therefore unsure how easy it is to switch between the two. Can it be done on the fly? i.e. run one method then load the other and have the GC do the switch for me? If this is the best option, does anyone have an accurate but short method I could use for each? I ask because I'll have many samples to run and they are manual injections.

I'd prefer one method for both gases but am up for suggestions. Thanks for reading

My sense is that you will be much happier with Argon carrier for measuring CO, as opposed to N2 carrier.

Switching the carrier on a single column/detector is very time consuming as you have to wait for things to settle down. I would try to avoid that. If you can set up your GC to use two different carriers on two separate injector/column/detector combos, (i.e. 'front' and 'back' or 'A' and 'B'), you could look into that. In that case you could be a lot more flexible, for example, H2 carrier for CO, CO2, N2, CH4, etc on channel 'A', then N2 carrier for H2 on channel 'B'.

What concentration ranges are you measuring? We always did ppm level analysis of CO, CO2, CH4 using methanizer/FID.

The analysis can be very quick because retention times are going to be around 1-3 minutes and you can run it isothermal, i.e. the GC doesn't have to do anything in between runs.

The column you have is not the optimal one because while separations are huge the temperature programming will cause huge baseline issues.

What other gases are you expecting to see in your samples besides air? water, CO2, C2 or higher hydrocarbons?

The carbon based packings are better at separating CO from air and methane. I would prefer to use Carboxen 1000 or its equivalent.

To measure Hydrogen you need argon at a minimum and nitrogen would be much more sensitive ie better.
But nitrogen would give a poor response for CO (very bad idea) so a two carrier dual detector setup would best meet your needs.

Is this possible for you?

The carboxen 1000 column like its predecessor Carbosieve S-II does retain CO2 and the C2+ hydrocarbons strongly and these should be removed by heating the column to higher temperatures than is required for your analysis. This consumes time of course and may not be desirable.

The methanizer method uses hydrogen carrier and a FID and care must be exercised that you calibrate regularly so you can confirm it is working properly.

Accurate measurement with a TCD means you must use two different carriers, He or H2 for CO, and N2 for H2.

The use of Argon carrier is a compromise and you may wish to investigate to see if it will meet your needs for linearity and sensitivity.

best wishes,

Rod

The column you have is not the optimal one because while separations are huge the temperature programming will cause huge baseline issues.

What other gases are you expecting to see in your samples besides air? water, CO2, C2 or higher hydrocarbons?

If I change carrier gases to argon will it influence baseline issues? CO, CO2, and CH4 have been showing up fine in my analysis (large percent of the gas injected) but I'm guessing the baseline issue may cause my H2 resolution to be poor considering it is approx. 5% of the gas produced.

Besides hydrogen and carbon monoxide I am mainly concerned with measuring N2, CO2, and CH4. There may be other higher hydrocarbons present in very small amounts but are not being studied.

The carbon based packings are better at separating CO from air and methane. I would prefer to use Carboxen 1000 or its equivalent.

I will look into this, I am a graduate student but my adviser may go for this if I have a good argument for it.

To measure Hydrogen you need argon at a minimum and nitrogen would be much more sensitive ie better.
But nitrogen would give a poor response for CO (very bad idea) so a two carrier dual detector setup would best meet your needs.

Is this possible for you?

Accurate measurement with a TCD means you must use two different carriers, He or H2 for CO, and N2 for H2.

The use of Argon carrier is a compromise and you may wish to investigate to see if it will meet your needs for linearity and sensitivity.

Unfortunately, a dual detector setup is not possible. The other port is being used for the MS and I don't believe I can convince them to change it over and get another detector for me

If you can set up your GC to use two different carriers on two separate injector/column/detector combos, (i.e. 'front' and 'back' or 'A' and 'B'), you could look into that.

This would be ideal, unfortunately we only have one TCD detector and the other port is being used for the MS. I was hoping the GC could take care of switching the gases without having to cycle it down and manually change it.

What concentration ranges are you measuring? We always did ppm level analysis of CO, CO2, CH4 using methanizer/FID.

The analysis can be very quick because retention times are going to be around 1-3 minutes and you can run it isothermal, i.e. the GC doesn't have to do anything in between runs.

Pretty high concentrations. Ideally, the gas coming from a gasification system would be about 85% H2 & CO (also producing some CO2 and CH4). However, N2 is used as the carrier gas in that system so my samples commonly have something like 60-80% N2 and the rest is CO,CO2,CH4, and H2. Would is be possible to use an FID for high concentrations?

A FID would not be suitable for CO measurement, nor for H2 as well.

If you cannot cool your oven down so the H2 is separated from nitrogen then you should be using a Mole sieve 13X (not the 5A) column.

See Supelco web site for a isothermal chromatogram.

Rod

edited correction: 13X, not 31X]

sffitzge,

Sound like a broken record here, but perfect application for HID type detector since you want to measure hydrogen along with the other fixed gases in one run. HID can do this with just helium as the carrier, no need to change gases. Replaced a TCD for a customer who was doing very similar work - pyrolysis of sugar chaff to look at by-products. (Standard disclaimers apply - I make an HID as do other companies, perfectly willing to discuss this if you want to contact me at aicmm at flash.net)

Having said all that, keep in mind that when you switch carriers to go from hydrogen to CO or nitrogen analysis, you have to displace all of the previous carrier saturated into the column packing. This takes quite a while. On way to do it is to make your H2 analysis one day, switch in the evening and do the other analysis next morning.

If you stay with TCD, try putting a Swage type 3-way valve on the carrier line for your TCD channel. Then all you have to do is flip the valve, load a different method (reflecting the proper carrier) and wait (and wait) for it to equilibrate.

Best regards,

AICMM

At % levels of analyte, in particular hydrogen, am I missing why it wouldn't be worth trying TCD with helium carrier? Large concs for a HID?

At less than 10% levels hydrogen's linearity is non-linear, even one may say inside out, at the least.

If I remember correctly, around 7% it actually becomes NEGATIVE in response when helium carrier is used. That is why since the beginning of GC nitrogen was used as a TCD carrier gas.

Only in the last 15-20 years has argon been used and it is not superior to nitrogen for the detection and measurement of hydrogen, but it may allow other gases to be measured at the same time as hydrogen.

I posted a ' W ' or a ' M ' peak chromatogram of hydrogen some time ago on this forum.

best wishes,

Rod

Thanks for all the replies, I'll be changing the TCD over to argon today and running some standards to see if it will work for me.

I realize for detecting hydrogen that nitrogen may have been a better carrier gas. However, due to having one detector I had to find a way to make this work for H2, N2, CO, CO2, and CH4. If using argon does not work as well as we hope I may be able to convince them that we need a new column or to try a new detector.

Another compromise, rarely used by analysts, is to have a carrier gas composed of 10% hydrogen and 90% helium. This allows the TCD measurement of hydrogen at high levels and the other gases as well.

BUT

it is a bit esoteric, no? (oh the things we tried back in the 70s and the 80s to get the job done quickly-cheaply)

And it can be prone to all sorts of errors and difficulties, it is a solution for the poor in money but rich in time and wasted time in trouble-shooting. It also gives poor results at low % levels of hydrogen in the sample.

But I guess it was worth mentioning to a poor graduate student who is desperate.

best wishes,

Rod

Keeping the remarks of AICMM in mind,

By putting two columns in the oven, and putting valves to switch between the two, using one at a time and only using one carrier gas for each column when switched into the carrier path, it is possible to use two carrier gas on the same injector - detector train rather quickly, without purging the column completely with a new carrier gas.

This would allow you to analyze hydrogen with nitrogen carrier and the other gases with helium by changing columns as you change the carrier gas and reset the detector's baseline and polarity.

Just an idea as a cheaper solution.

best wishes,

Rod

This chromatogram of hydrogen shows what happens to its response when helium is used as a carrier. Initially the response is positive for hydrogen but as the concentration increases it DIVES down and becomes negative.
(at % levels) Notice the M shape of the peak. The front edge of oxygen peak is visible at the right side of the chromatogram.

This makes it impossible to accurately measure hydrogen using helium carrier. This is why nitrogen and sometimes argon is used. The response of other fixed gases is almost nil or has a much lower response using the other two gases as the carrier.

best wishes,

Rod

sffitzge

Well, are you going to leave us in the dark? How did argon work for you?

What were your results? Did you get enough sensitivity to perform an adequate measurement? at what levels?

Don't go away mad. Tell us the end of the story.

best wishes,

Rod

sffitzge

Well, are you going to leave us in the dark? How did argon work for you?

What were your results? Did you get enough sensitivity to perform an adequate measurement? at what levels?

Don't go away mad. Tell us the end of the story.

best wishes,

Rod

Sorry to leave you in suspense. Not mad at all from all the help I received. Just frustrated with the equipment I have to work with. We tried argon, but had problems dialing in the method for both hydrogen and CO so we decided a better route would be to buy a gas analyzer (separate from the GC) for the hydrogen and go back to helium carrier gas for CO, CO2, CH4, etc. since that method was working well from those gases.

I wish we could have stuck with the TCD and run with two columns instead but the MS is taking up our other port and my adviser had absolutely no intention of taking that offline for me.

Thank you for your help