environmental N2O: how to get rid of CO2?

[Mari]

Hi All

we use Agilent GC 6890 with SupelQ column for environmental (~300-350ppb) N2O ( split inj/oven/ECD - 150/35/320 oC, carrier He, make up P5), and it works fine with calibration N2O gas that has no CO2. When it comes to the real atmospheric air samples though, CO2 concentration is about three orders of magnitude higher and its peak comes out right before N2O, so N2O peak is totally lost. I know that for the systems with the packed columns CO2 trap pre-column or back-flash column could be used. I could no find anything like this for the capillary ones in the literature though. Help please! Thanks a lot

[larkl]

What kind of column was used on the packed instrument to strip the CO2 from the sample?

[Mari]

Some did it with Hayesep Q or N for the back-flush, others used ascarite (a type of sodium hydroxide coated silica) trap pre-column.

Can back-flush be used with capillary columns at all? Thanks!

[GasMan]

Mari,

For backflush to work, you would need a column that elutes the CO2 after the nitrous oxide. This will not happen with Porapak Q or N. The only column that will elute the nitrous oxide before the CO2 is Porapak T, but I have not seen any capillary columns with this packing.

Capillary columns can be backflushed, but with care. Problems can occur with the PLOT type columns. When you backflush, you reverse the flow in the column which means that the 'pressure' end of the column is suddenly at atmospheric pressure. This can give a pulse of flow that can loosen the column particles which can give problems in your switching valve.

Gasman

[Mari]

Thank you, GasMan!

well, I am really not experienced and advanced enough chromatographist to experiment with expensive columns, so I probably would not take that responsibility.

Any other way to remove CO2 from the samples, how about pre-columns? Same question: are they used with capillary system at all?
Thanks!

[larkl]

Haven't had too many problems valving with PLOT columns, it's a pretty common application (valving and PLOT). Scoring of the valves from the particles can occur. It wouldn't stop me from trying it. If you could find a column that'll retain the CO2, I'd set it up as a stripper column (and valve) upstream of the main column. Maybe you could use a packed stripper column with a megabore capillary main column. Interesting problem.

[GasMan]

Mari,

You may want to check out the posting 'Argon/CH4 carrier gas purification'. They are doing the same analysis, but using packed columns. I question why you are using a capillary column with a splitter inlet. If your sample is in the ppb range, why split and dilute it further.

It is possible to set up a 10 port valve with a packed pre-column which would hold back the CO2 and inject into the capillary inlet. The question I would have is what split ratio are you using and what column flow. If you can give me the serial number of the 6890 GC, I can find out what was built onto your GC at the time of delivery, then we can decide what you need.

Gasman

[geigor]

Hi Mari
I'm working on the same analysis but I use two packed column Unibeads 1S mounted on a 10 port valve: N2O peak is very well resolved with no interferences from other light gas. contact me if you need some more details

[Mari]

Thank you, GasMan

here is the info:

Agilent 6890N,
SN US10435046
Split ratio 10:1
Split flow 49.9 mL/min
Total flow 57.7 ml/min

We have split/splitless injection system and were told by service people that packed columns will not work on it.

[Mari]

Thank you, Geigor, I'll drop you a line by email

[Mari]

I was just recommended to try cryogenic oven cooling (LCO2 or LN2), though I'm not exactly sure how it will work. Will it trap CO2 but not N2O? N2O melting point is actually lower but not much.
I wonder how it will effect the column, and if it requires change of our temperature conditions. Any advice is kindly appreciated.

[GasMan]

Mari,

Your 6890 was supplied with a capillary inlet, a PTV inlet, the ECD and a NPD. There was no valving on the delivered GC and no oven cryo option. The oven cryo is designed to allow oven temperatures below ambient to be used. This can give better separation of low boiling compounds. The fact that you have a PTV on the 6890 with CO2 cooling, limits your choice of oven cryo. It must be the same as the PTV, which means you would have to use CO2 also for the oven. You would need to order the kit G1565A from Agilent.

It would be a good idea of yours to contact Geigor to find out how they are doing the analysis. Adding a 10 port valve will give better results ( I assume that you are using a syringe to inject the samples). It may be at first a more expensive solution, but in the long run it may be cheaper as you will not have the cost of CO2 cylinders.

Gasman

[Mari]

Thank you , GasMan,

no, actually, we are using 7694 Headspace, our samples are in the sealed vials. We don't mind investing more into this analysis if it worth it. The problem is that I have very little info about how cryocooling works.
Will LCO2 create subambient low enough to improve separation and extend retention times between CO2 and N2O? If someone can give me a link to any application, it would be highly appreciated! I did not find much in Agilent library except for the Installation Guide.
Thanks!

[GasMan]

Mari,

The oven cryo works as a special valve installed on the left side of the oven that injects liquid CO2 directly into the oven. To get some idea of how effective this would be, why not try cooling you oven down to room temp, and then injecting your sample. The 10' to 15'C change will give you an idea if this is the right way to go.

Gasman

[AICMM]

Mari,

Go down to your local grocery store or gas vendor and buy a 10 Kg block of dry ice and put it in a cooler with lots of newspaper insulation. Chuck a chunk of it into your oven with the oven off and when it gets to some lowish temp. try your run. A very crude form of cryo that allows you to evaluate the improvement that you might obtain by installing a CO2 cryo valve.

Best regards.