Separation of Hydrogen and Oxygen

[0pfer]

Hi @ll,

In our working group we are researching the splitting of water. so in future we want to determine the amout of produced hydrogen- and oxygen-gas.
may be anyone of you has got some experiences with the separation of the coresponding peaks and can tell me the best conditions for the gc.
I'm especially interessted in suitable values for the oven temperature and flow rate.

my carrier gas is N2 and TCD to detect the gases.

thx a lot for your help!

[willnatalie]

We use a mole sieve column for some of the gas separation needed. This is on an RGA, so specifics are going to be slightly different.

Specific conditions (temp, flow) come with the column.

Hope this helps

Will

[0pfer]

ok, and what conditions are u using?

i have to say that i need to detect very low H2-concetrations (0.05 to 1 vol%).

right now the peaks i recieve are quite broad, although i use high flow rates (150 to 200 mL/min).

is there in general a relationship between the temperature and the shape of the peaks?

yesterday i tried a higher temperature (150°C) and i recieved very nice shaped peaks, but then it was completly impossible to seperate o2 from h2.

beside the temperature, are there some other parameters i could change to get sharper peaks?

[larkl]

How much sample are you injecting?

What type of column and column dimensions are you using?

0.05 to 1% is NOT a low concentration. You may be overloading hence the poor peak shape.

[0pfer]

my injection volume is usually 100µl or 200µl. i can not imaging that this is to much for a conc. of 0.05 vol% of H2, do u?

here are the informations about the column:

Name: RT-Msieve 5A, Thickness: 10 µm, Lenght: 30 m, Diameter: 0.53 nm

The GC i'm using: GCMS-QP2010 Series (by Shimadzu)

[MikeD]

One cause of broadened peaks on Mol Sieve 5A can be adsorption of water over time. If your samples are saturated with water vapour you would need to periodically bake out at 300degC for at least 3 hours.

I would expect sharp peaks at 150 degC since I don't think there is any retention other than dead volume at this high temperature.

Returning to the 5A column conditions - I don't have data for your dimensions but here are some for 25m x 0.32mm 5A at 5 ml/min column flow, split flow >50 ml/min, injecion vol 100uL, temp 50 degC.

hydrogen 3 mins
oxygen (and argon if present) 4 mins

However the thermal conductivity of oxygen is only fractionally greater than your nitrogen carrier (about +0.4%) so I would only expect a small noisy blip even with pure oxygen. Conductivity of argon is about half nitrogen so you get a small negative blip if argon present which may or may not cancel the oxygen.

All in all I am not sure what you are seeing when you say that hydrogen and oxygen are not separated. Try argon carrier and see what happens to the hydrogen and oxygen peaks. I don't think it's feasible to measure both hydrogen/oxygen on one column with nitrogen carrier.

[AICMM]

opfer,

Clarify for me, do you really have 200 mL/min of column flow? On a megabore PLOT column that is screaming. Or is that really column plus makeup (plus reference, etc....?) Column flow should probably be around 10-20 mL/min tops so make sure that what you are measuring as actually the flow out the back of the column. Remember as well that TCD is concentration sensitive so the addition of make-up comes at the price of sensitivity.

Re MikeD comment, I had my first opportunity to run air on a TCD with nitrogen carrier and I was shocked at how stinkin' un-sensitive it really is. Wow. I'm used to HID's where 20 uL of argon is a huge peak and 20 ppm of hydrogen is no problem.

Best regards.

[0pfer]

thank you very much.

its exactly as you both said. today i tried some air samples that show me that the sensitivity for oxygen is really very low.
so i come to the conclusion that its senceless to detect both hydrogen and oxygen with n2 as carrier gas, especially when the concentration of oxygen is very small.
this afternoon i will concetrate on the determination of small hydrogen amouts and therefor to figure out the best conditions.

@aicmm: i was talking about the total flow sorry. my column flow is between 25 and 30 mL/min

thx again 4 your help!

[chromatographer1]

You will find the separation of O2 and H2 improved if you don't use a capillary column.

Look at chromatograms on the internet using micropacked and packed columns. Restek, Supelco, Varian, Agilent, Grace, and others will have examples.

Rodney George
consultant