CO2, N2 and water gas sample analysis

[kasia_roro]

Thanks Rod, seems I am really ignorant... I will call Supelco and discuss about the new packed column like you suggested. The Porapak Q I got from Shimadzu together with GC I am not sure from where they purchase it. Do you mean that Porapak Q I have should work for the samples I have? Bellow you have chromatogram where you can see big peak from not resolved O2 and N2 with the little peak in a tail at 2.7 min which is from CO2 and at 6 min a peak which ether from water or methanol which I decomposed to water and CO2, 500 ul of gas sample injected manually. I always have this tailing doesn't matter which conditions (temp., flow) I am using. When I cooled down the sample with liquid nitrogen and then sample I got good separation of air and CO2. Therefore I thought and I was also told by a colleague that if I remove a water I will have separation of air and CO2 and here was the idea to use pre-column.


What about the H2 sample I was mentioning and the idea of using argon for both systems? If I go for the column you suggested you think I will not need pre-column?
Regards
Kasia

[chromatographer1]

Your late eluding peak is not water, it might be something in the C4 or Ethanol range.

With your fast temperature ramp, you are driving up your baseline from the residual air residing in the pores of the beads. Normally, there is a constant air 'bleed' from this residual sample but with your ramp you drive it out and the baseline bumps up..

Your injector is WAY TOO HOT, you are increasing your sample volume from heat expansion and may be sending sample backwards into cooler parts of your injector system. Read about 'flashback' or 'backflash' with GC injectors to learn more.

Drop your injector to 100C. Really. Since your water is already vapor why does it have to be 250C ?

Inject slowly, 0.2mL per sec.

Run the sample isothermally for 5 minutes before you ramp and see the difference you will have with your chromatography. Your 10ft (3m) column is plenty of column to handle this.

Let us know what you discover.

Rod

[AICMM]

kasia_roro,

The reason that Rod can be so confident about it not being water, by the way, is the peak shape is way, way too nice. Water is usually a really ugly peak (only exception I have found is on thick film PDMS type phase.)

Please clarify for me (since it seems the conversation goes back and forth.) The TCD injections are loop or syringe? Reason I ask is this, syringe injections, even with relatively small volumes (z.b. 0.5 ml) usually give much broader air peaks than valve injections do. (Sort of like what you posted here...) I think it has to do with surge in pressures and pushing sample in places it really should not be. Valve injections, on the other hand, typically give better peak shape. Saw it in person with trace hydrogen in ground water application I set up.

The other question I have is about flows. You have a 25 mL/min column flow rate but do you also add make-up to your TCD? It might not be a bad idea in this case since you do not seem to be looking for trace level analytes. You will loose sensitivity since TCD is concentration dependent but it might solve you peak width issues...

Best regards,

AICMM

[kasia_roro]

Dear Rod, I am aware that the injector temp. was far to high. This was just an example one of the few where I started with injection temp. 70C and for regular measurements I was using 120 but it doesn't change much. Which temp. program would be optimal to resolve air and CO2? I am afraid that with this big tailing it is not possible. Why when I cooled down my sample I could see good resolution of air and CO2? Has water something to do with it? I am not carrying much about the water peak I just want to be sure that I can evaluate amount of CO2 and that water is not damaging my column. I will try what you suggested next week and I will let you know. Do you think my column is fine with this big tailing?

Dear AICMM, I have both loop and syringe, at least I think so I have an auto-sampler for TCD line which is sampling with the loop. My reactor is connected over vale with GC, during sampling valve switch and fill 500 ul loop and then inject to column. If I sample manually, I am applying additional pressure of carrier gas (I guess this is the make-up gas then) and I am filling the loop by syringe (500ul gas sample). I was also trying with the higher flow rate, the tailing was smaller but the resolution as well.

As you see I tried everything and slowly I am running out of the ideas and that is why I thought that the packed column at least the one I have can not handle with water and therefore I wanted to change to PLOT system. Do you think my Porapak Q is damaged or something? Rod do you think that buying a Chromosorb 102 as you suggested will solve the problem or I need a pre-column anyway? What about argon as a carrier gas?

Sorry for being so annoying but I need to make the system working within two weeks and for the moment I am stucked.
Thanks for help
Kasia

[chromatographer1]

Your Q column may not have been good from the beginning. Why don't you show us an on scale chromatogram of a syringe of air alone, and the same chromatogram 'blown up'.

Argon will not help you except to spend more of your money.

Rod

[kasia_roro]

What do you mean by "blown up" ? Regarding argon, I was thinking about using the same carrier gas for both type of samples as I was mentioning already few times. We have two type of samples, when I work with CO2 samples I am using helium as a carrier gas. When my colleague needs to investigate H2 samples she uses nitrogen so we have two carrier gas on one GC, what is maybe not the best idea. Therefore I was thinking to change to argon because it is fine for both CO2 and H2, and then I will have at least whole time the same carrier gas. Any way each time we need to also change a packed column, so I dont know If then the change of carrier gas matters.
Regards
kasia

[chromatographer1]

When a chromatogram is plotted with less attenuation of the y-axis it is said to be 'blown up'. The chromatogram you are displaying is 'blown up.' The y-axis is from 0 to 2.5.

The other peaks are off-scale.

To see these off-scale peaks as ON SCALE ;eaks would be helpful, if not on-scale, then using an a scale of 0-200 instead of 0-2.5 perhaps.

best wishes,

Rod

Don't forget to look up the thermal conductivity of CO2 and Argon. Compare the difference between them with the difference of CO2 and helium or hydrogen, then determine if you can use argon as a carrier for an analysis of CO2 at the levels you wish to measure.