CO2, N2 and water gas sample analysis

[mikki11978]

Hi,

I have gas sample containing CO2, N2 and water vapor which I need to analyze. What type of column I should use? I have Hayesep Q 8ft? alltech is suggesting Hayesep DB...should I change my column to DB or 8ft Hayesep Q will do the job. Currently I have a mol sieve inside as well. Do have to remove mol sieve if H2O and CO2 are present in my sample?
I guess TCD detector is used for permanent gas analysis.

thanks, mikki

[Dank]

Yes a TCD will do fine for this kind of analysis. You can actually used a packed column to do both;a cheap affordable way. I use a a 27 1/2' 30% DC 200/500 using He as a carrier . I don't know how much moisture is present in your sample. If it is a liquid you could try to stripping using He and collecting the headspace. PM if you need info on where to get the column. We make them and should have some around for sale.

[mikki11978]

water is in a form of vapor. Cannot do much about it. How about molecular sieve. do i have to remove it? this GC already has mol sieve inside.

[chromatographer1]

Your Q column will work fine for N2, CO2, and water.

Better for the water peak shape would be R polymer column.

As long as you don't allow the water and CO2 to pass into your sieve column it will be fine. After many injections the MS will not have the separation as before.

I would suggest you keep your oven above 80 degrees. Reconditioning the MS column at 200 for a few hours daily will fix any issues.

best wishes,

Rod

[kasia_roro]

Hello,
I find this forum while searching for the solution to solve my problem with the water in my synthetic air+CO2+ e.g. acetaldehyde (or other organic molecule). May I will start from the begging: I have a Shimadzu GC 2014 with FID and TCD detector. GC/ TCD line is connected to photoreactor in a loop and then each e.g. 20 min the sample is taken by autosampler. For FID I have manual injection. In the reactor I am decomposing organic molecules like acetaldehyde in presence of photocatalyst in synthetic air in the gas phase. By FID detector I am evaluating acetaldehyde concentration and by TCD CO2 evolution. The problem is that for a TCD detector I have a Porapak Q column and big tailing of the air peak (N2 and O2 comes as a one big tailing peak) and in a tail is peak of CO2. When I froze out water (water vapor upto 20000 ppm range CO2 is in the same level) I got nice separation of air peak and CO2 peak. However when I am freezing out water there is a danger that I will also remove CO2 with water. I have to admit that I am a new user of new equipment without any experience in GC with some with HPLC. Could you please tell me if eg. Hayesep R could work or you have some better ideas.
Thank you in advance

Kasia

[chromatographer1]

You may be having sample introduction issues more than chromatography per se.

How much are you injecting? The shape of the loop and the diameter of the sampling valve all have an effect on your sample plug (the tailing etc).

You should have 0.030 inch ID valve port openings. and the loop if longer and narrower rather than shorter and thicker, will work better.

You may also need to change the ID of your column.

Water normally elutes much AFTER the CO2 peak on a Q column. R or D polymer will pull water back but if also pulls CO2 but to a smaller degree.

If you get a sharp plug injected you get sharp and well separated peaks for air, CO2, and water.

Tell us more details and perhaps we can help you.

best wishes,

Rod

[kasia_roro]

Rod thanks for fast replay
As I told my TCD is connected to the photoreactor and then I have injection by autosampler but for the experiments I mentioned I used a headspace vial i which I run the reaction and then I was sampling a gas sample by syringe and I injected manually. Injection on the column was 1ul. In case of the valve port openings I have no idea what you mean and where to check it, sorry... I have just such information on the delivery sheet: 221-VIED46UW sampling valve vici, including 6port 2-pos valve, m-electric 4"so 1/8".75mm, 330C/300psi gas, N60/T (ED46UWT), Sample loop 500 ul, UW type valve 1/8" ends with nuts and ferrules, 316SS (SL500UW) - do you understand something form this? otherwise I will call Shimadzu and ask them. ID of my column is 2.1 mm (0.08 inch) it's pack Porapack Q 80/100 3m column. In my case doesn't matter the parameters I have seen just tailing of big air peak, CO2 on the tail and no water. Just when I froze out the water (immersed few time the vile in liquid nitrogen) I have seen good separation of air and CO2 and no water peak. Water peak I could observed when I replaced Porapak Q by Carboxen 1000 but then water and CO2 were coming as a double peak and could not resolve it. My support person from Shimadzu told me that I am "killing" my Porapak Q column by water and that I will not get separation on this column if I have water in my sample. I also had a meeting with expert from Supelco and he told "GC is not happy with water and you are not happy with your GC" Therefore I was research if I can remove water before injecting on the column or somehow before loop but then it is always danger that I will remove CO2 as well. Do you think that I can use a Porapak Q for water separation? The other alternative I found are PLOT columns like PoraPlot Q or U or HayeSep D but this require reconstruction of GC.
Thanks for support
Kasia

[chromatographer1]

You have a packed column sampling valve with the larger ports.

I suspect you mistyped your sample volume you injected. It is not 1uL but 1mL, yes?

Your sample is a gas? Or are you injecting liquid? Very hard to believe you are getting tailing with 1uL of gas.

So I will assume your sample is gas and you inject 1mL.

Yes, you discovered that water and CO2 elute close together on the Carboxen.

Since I suspect your column is a metal column I would buy a 2-6ft metal column with a bored through 1/8" union and connect a precolumn to your 3m Porapak Q column. This precolumn I would fill with 10% Carbowax 20M or 1500 on Chromosorb PAW packing (80/100 or 60/80 or 100/120 should be fine for mesh size). This will cause the water to be held up and allow the fixed gases to proceed well ahead of the eluting water. I hope your flow rate is greater than 15 cc/min. While of course, a slower rate will improve your detection limit it will also give a larger amount of tailing of any peak due to the amount of dead space in your column-injector configuration. You could go all the way up to 60cc/min but 30-40 might be a better choice.

Good luck with your research.

Rod

[kasia_roro]

Yes I have a packed metal column. I didn't mistype I just mixed the lines 1 ul is for FID line. I have two line GC with two separate columns Rtx-Wax (Fused silica) 30m ID 0.25 connected to FID and second line Porapak Q 80/100 3m ID 2.5 mm for TCD. During manual injection on TCD I am injecting 500 ul to fill up the loop and this all goes on the column. I have a gas samples. I already asked Shimadzu to get smaller loop e.g. 250 ul.
Thanks for the hint with the precolumn, it was my idea as well but since I have already 2 columns in one oven it can be tricky to fit in one more (Shimadzu even suggest that it is possible that they will have to take off the capillary column for FID and this is not possible for me because I need it as well). Therefore I had the idea to get a HeyeSep D packed column which seems managing with water and replace the Porapak Q or you think it doesn't make sens? Do you think that I will then have any way the same problem with HeyeSep D like with Porapak Q? Sorry for probably stupid questions but I just started in this topic and still there is too many unknowns...
I just read on the forum that water is quite bad for silica based columns, what would you suggest to do to remove water before injection on Rtx-Wax column? I need to remove water in manner to don't remove organic pollutant like acetaldehyde.
Thanks again
Kasia

[chromatographer1]

Hayesep D and Porapak Q are essentially the same polymer bead. Unless your Q column has many voids from being poorly manufactured you will gain little or nothing in replacing it with a HS D column.

water is liquid form is not good for capillary columns But your capillary column will tolerate water for many months before you see any problems with it.

If you can remove physically water from a sample without removing the much more volatile acetaldehyde you are a better chemist than I, Kasia. Mole sieve? gel permeation? Addition of anhydrous salts?

Please publish how you may find a way to do that. I always love to learn something new.

best wishes,

Rod

[kasia_roro]

I see, in this case it seems that only pre-column can help me with the Porapak Q. Regarding the Rtx-Wax column I was rather thinking about some water trap or short precolumn with molecular sive. I have a gas samples and the water is a vapor in my samples and not so high concentration but I have quite noisy background from FID and therefore I would rather prefer to get rid of water on both columns. By the way you asked before about the flow rate I have 25 ml/min and the 100 ml/min is max I can go. I already tried all combination : different flow rate, different temp., different temp. programs and nothing helped for the separation on Porapak Q. Once I had just good separation of air and CO2 when I froze water out. Slowly I am running out of ideas...
Kasia

[jole]

hi everyone,

I am a freshman in the GC work. My boss asked me to use GC to anaylze the concentration of O2, N2 and water (vapor). I am wondering how to make the calibration curve for vapor? could some expert give me advices?

thank you in advance

jole

[chromatographer1]

Kasia,

You will soon come to the conclusion that you must separate the water with a precolumn. You can take the Q column out of the oven and remove a foot or so of the packing at the head and replace it with a glass wool plug (small amount) and then pack the head of the column with the Carbowax 1500 on Chrom Paw packing terminated with more glass wool. That should be enough to perform the separation you need as well as not crowding your oven.

Jole,

You really should start your own thread so Kasia's answers and yours do not become confused.

Only a Carboxen 1000 column will separate O2 and N2 whilie allowing you to measure water. Any mole sieve column you use will trap the water.

Once you know the humidity content of an air sample you can dilute the air with helium or whatever carrier gas you will be using for your analysis for your calibration curve.

Good luck,

Rod

[kasia_roro]

Rod thanks again for your hints. However it sound quite complicated to pack the column on my own. I prefer to go rather for some commercially available solutions because I am really beginner in this field and I don't want to destroy something. I just got information from Shimadzu that it is not a problem to put a pre-column to my GC but this also requires a reconstruction which is not much cheaper than changing a system from the packed column to the capillary PLOT type of columns. I know from literature that eg. HP-PLOT U or Q would do the job to separate CO2 from water in presence of air. Additional issue hear is that by this GC/TCD also H2 should be analyzed which will be a gas sample with air and water vapor. Therefore I was thinking to replace pack column with the PLOT column which could resolve all mentioned gases. At the moment I have as well two different carrier gases. I am using helium on Porapak Q for CO2 evaluation and colleague is using nitrogen on MS 5A packed column for H2. I was thinking to change as well carrier gas to have for both argon and then have one carrier gas and one column for both. What do you think about it? I hope I am not making fool of myself with this idea

Thanks
Kasia

[chromatographer1]

Your sample size will greater decrease with a capillary polymer bead column. You will not gain any separation unless your present packed column are really badly made. No one in process control would choose a capillary Q column over a packed column. The detection limit will INCREASE with a capillary column which is not the direction I am sure you wish to go. Why have a limit of 3000 ppm when you could have 5 ppm with a packed column? Of course, the cost will be 3-4 times higher for the capillary, and its usable life will probably be several times less.

I suspect your problem is the installation of the column or the column itself is poorly made. I really suspect (not seeing a chromatogram) that you have a badly made column with a huge void in it. If you bought a replacement it might only cost you $150. I would try another supplier than your present vendor. You could even ask them to put a precolumn of Carbowax 20m into your new porous polymer column if you really want to pull water away from your fixed gases. (2 feet of Carbowax 20M and 8 feet of Q packing.)

Alltech and Supelco are capable of selling you a column which will meet your needs. Supelco will even install screens which allows the void spaces usually caused by the glass wool terminations will now be filled with packing. This eliminates a good deal of dead volume.

Your description of water interfering with CO2 is MOST unusual. These are usually separated by minutes of flat baseline, MINUTES, not seconds. If you can download Supelco's packed column application guide

[ Bulletin 890 - Packed Column GC Application Guide (810 KB ) ]

look at figure 117, a 6ft of column of Chromosorb 102 (a replacement for Porapak Q and Hayesep D) and you will see water eluting at 3 minutes, CO2 at 1 min and N2 at 30 seconds.

This is what you should be seeing.

best wishes,

Rod

ps when you use argon carrier gas your sensitivity for nitrogen and CO2 will decrease as well. I suggest you see a chromatogram of someone else's work with similar hardware before you make such drastic changes. Why did your supplier construct your GC as it is now constructed if there were a better way to do your analysis? and simpler and less complicated?

Change and you will find out the reasons. You may not be happy with the results. Especially when all you had to do was to lower the oven temperature to 50°C and get a new $150 column.