Chromatography Forum

Dear Forum

We are trying to run an experiment that requires the detection of ppm of H2 released by some bacteria growing in anaerobic encapsulated flask medium.

25ul of the flask atmosphere is injected in a GC with a 5A Molecular Sieve column and detected by a TCD.
However the H2 peak is extreamely broad and large, nothing like we use to see in articles or literature, and the sensivity is extreamely bad. I can just detect H2 volumes above 2,5ul of pure H2.

Skipping the fact that something is broken or leaking - I've checked it already, the column and the TCD are new.

The parameters I'm using are:
injector (PKD) emperature - 110ºC
oven emperature - 70ºC
constant flow of N2- 10ml/ min
block emperature - 150ºC
Transfer Temperature - 140ºC
Filament Temperature - 250ºC
Make up gas Flow - 28ml/ min
Gain - 10x

Is there something realy wrong here ? Please help me?

Thanks

Your choice of the carrier gas is good as the thermal conductivity of hydrogen and nitrogen have a large difference.

Now some other considerations:

- I think your sample size is too small. 100+ ul is a good choice for your application. (I'd try 200-300 ul). This lowers your detection limit by about one order of magnitude
- Your filament temperature should be 300°C. As the sensitivity increases three-fold (i.e. double temperature -> 8 times more sensitive) this should help your detection limit.
- Lower your block temperature to 110°C. This way no water can condense and you gain another boost in sensitivity.
- Please check the manual of your TCD detector. The column flow of 10 ml/min and your make-up flow add up to almost 40 ml/min. This is too much gas. Try to increase your column flow to about 20 ml/min to help the separation and just don't use make-up. But please check beforehand if your detector can cope with these conditions. Otherwise your TCD could burn out!
- According to my experience your oven temperature is too high. Try 50°C instead.
- Check your attenuation setting and voltage setting.


And now some questions:
- How long is your column?
- Do you use a Macro-TCD or a Micro-TCD?


And please post (s)ome of your chromatograms.

I suspect you are using a capillary column.

If you are using a packed column you can inject a 1000µL or more.

The advice (from Hb] ) about reducing your makeup (why I suspect you have a capillary column) is good advice.

Describe your GC and its detector. This information will assist in giving you more good advice.

I also suspect you are not seeing hydrogen at all but something which is too big to be trapped in the sieve and is flowing through the column as a non-retained peak. Especially if you are measuring gases from a biological experiment.

It could even be water that overloads the capillary column's ability to absorb.

Isobutane elutes before nitrogen on a MS 5A and close to where H2 elutes.

Please share some details and the forum members will be glad to assist you.

best wishes,

Rod

Dear HBJ

I can't tell you in this moment much more details about the TCD, because there is no TCD manual (but it is the standart equipment from Thermo Unicam). I think the TCD it is a Macro one, so it doesn't work whitout make up gas flux (as far as I tested - the all equipment starts to Beep when the make is not working).

The column is also a standart one - Molecular Sieve (MS - 5A) with a lengh of 3m and a diameter of 1.8''

The picture here, is a H2 calibration Curve I did to test the equipment.
The way we do it is by diluting several known volumes of H2 in an Ar saturated atmosphere 12ml flask and injecting 25ul of this new atmosphere in the Trace GC Ultra (from Thermo Unicam)

[img][img]http://img1.putfile.com/thumb/2/3114225817.jpg[/img][/img]

The curves I selected are from 25ul sample injection taken from the 5ul H2 injected previously in 12ml Ar flask. The procedure was repeted with 10ul and 20ul of H2.

The peaks are not sharp and easely mistaken by anything else.

What am I doing wrong here?

Thanks so very much !! for you reply

Hi,

Just a thought, have you conditioned your new column? I seem to recall that every so often, I had to condition it by baking it at 320°C for several hours before use.

You would have to check that temperature.

Regards,

Ralph

@GOM: Bingo! You're right!

So Slick please condition your column. This could help you to get some real peaks. If it's possible you should disconnect the column from the detector. This way any dirt from the column can't enter the detector and contaminate the filaments or the cell.
If you separate the column from the detector it is VERY important to switch the detector completely off cause otherwise the detector's filaments may burn out!

Another hint for handling TCDs: ALWAYS let the carrier gas purge the detector for about 15 minutes before you heat/activate the TCD. This way, you prevent the filaments from being attacked by oxygen which can shorten their lifetime tremendously.

And I hope you let the TCD reach its equilibrium (about 2 hours until it reaches a flat, stable baseline).

Regarding your chromatograms this doesn't really look like peaks. So please condition your column and increase the column flow to about 30 ml/min. Every TCD I know (which are older HP/DANI) will be able to cope with these conditions. Furthermore your run time is far too short. A packed column is unable (AFAIK) to separate a gas mixture in under one minute.

According to http://www.discoverysciences.com/chroma ... etails=yes you can use a temperature of 35°C with that column and try to inject some air. Though nitrogen obviously won't produce a peak (as it is your carrier gas) you should at least see an oxygen peak.

P.S.: Your mol sieve will absorb (among others) the carbon dioxide in your room air. So condition your column once in a while to get rid of all these contaminants

P.S.S.: Do you have a different column (Porapak, Hayesep, ...) at hand?

All of the above comment are good.

Some additional comments.

Your makeup gas could be for the balancing filament? and so it is not for the sense filament. And yes it is required.

You should check the polarity of the detector so you can be certain that your hydrogen peak is positive, not negative (below the baseline).

And hydrogen will elute at 60 seconds for a 3m column at 15cc/min and 75°C.

best wishes,

Rod

Slick

Send me your email address and I will send you some chromatograms of hydrogen on a 3m MS 5A 1/8" column.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
techservice@sial.com

You should check the polarity of the detector so you can be certain that your hydrogen peak is positive, not negative (below the baseline).

Speaking of polarity, if the hydrogen standard is in an argon matrix (or was that just "Air" misspled?) won't the hydrogen and argon peaks go opposite directions with a nitrogen carrier?

(BTW, there's an excellent on-line reference for the physical properties of gases, including thermal conductivity, at Air Liquide's website.)

Conductivity Values for fixed gases
Weist Handbook of Chemistry and Physics

Carrier difference Helium Nitrogen Argon Hydrogen
Value
Hydrogen 471.1 95 405 426 0
Helium 376.1 0 310 331 -95
Nitrogen 65.7 -310.4 0 20 -405.4
Argon 45.5 -330.6 -20 0 -425.6
Oxygen 68.2 -307.9 3 23 -402.9
Methane 89.3 -286.8 24 44 -381.8
Carbon Monoxide 63.9 -312.2 -2 18 -407.2
Carbon Dioxide 43.8 -332.3 -22 -2 -427.3
Ethylene 55.0 -321.1 -11 10 -416.1
Ethane 58.3 -317.8 -7 13 -412.8
Propane 48.4 -327.7 -17 3 -422.7
Butane 43.4 -332.7 -22 -2 -427.7

If this does not paste correctly then I will gladly send out the spreadsheet to anyone who emails me.

Rod at rgeorge@sial.com

Conductivity Values for fixed gases
Weist Handbook of Chemistry and Physics
Carrier
Helium Nitrogen Argon Hydrogen
Hydrogen 471.1 95 405 426 0
Helium 376.1 0 310 331 -95
Nitrogen 65.7 -310.4 0 20 -405.4
Argon 45.5 -330.6 -20 0 -425.6
Oxygen 68.2 -307.9 3 23 -402.9
Methane 89.3 -286.8 24 44 -381.8
Carbon Monoxide 63.9 -312.2 -2 18 -407.2
Carbon Dioxide 43.8 -332.3 -22 -2 -427.3
Ethylene 55.0 -321.1 -11 10 -416.1
Ethane 58.3 -317.8 -7 13 -412.8
Propane 48.4 -327.7 -17 3 -422.7
Butane 43.4 -332.7 -22 -2 -427.7

If this does not paste correctly then I will gladly send out the spreadsheet to anyone who emails me.

The software for this forum does not appear to support tables, or HTML code.

However, the friendly folks that run this board are apparently also putting together a wiki. So, I put that information into an HTML table and posted it over there: link. (I'm not sure what the value in the first column is. I'm guessing it's the thermal conductivity expressed in some units, but I don't know the units.)

But posting the table there raises the question: Can the values for physical properties of a compound be copyrighted?

These values are taken from The Handbook of Chemistry and Physics - Weist

When the first value in each line are multiplied by 10^-6 the values are in dimensions of

cal*cm*cm^-2*s^-1*°C^-1

The reference temperature is 120°F or 48.9°C

And Michael, the copyrighted boiling point of water is 100°C or 212°F.

I sure hope scientific data is not copyrighted.

The second value in each line is the difference in conductivity of each gas when helium is used as a carrier.
The third value in each line is the difference in conductivity of each gas when nitrogen is used as a carrier.
The fourth value in each line is the difference in conductivity of each gas when argon is used as a carrier.
The fifth value in each line is the difference in conductivity when hydrogen is used as a carrier.

hope this helps if anyone is interested.

Rod

Looks like the pros have taken over

I'm curious if Slick's separation is working now...

[EDIT]
Rod, your data is really appreciated. I'll print it out and stick it to my old DANI box with a TCD.

Any updates?