Detection of NOx using gas chromatography

[daffonce]

Hello all,
I am looking to measure NO2 and NO using gas chromatography. We have used uv spectometer and an FT IR and have not been satisfied with the results. The first is just not that responsive time wise, and the second the water band interferes with the NO2. So we were thinking of using our GC. We looked into using an NCD however it seems like they are for very low concentrations of NOx.

We are trying to measure the NO2, NO, and any possible N2O4 in a gas that contains ~5000 PPM of NO2, the balance of the working fluid is non condensing humid air. I spoke with a sales rep from agilent and he seemed to think a TCD detector would work well in detecting such high concentrations of these substances. Now my question is what column type should be used to split NOx from air, water vapor, and each other (i.e. separate NO NO2 and N2O4).

[chromatographer1]

The only data I have seen to separate these and not react with them (and that may or may not be true) is Porapak or Hayesep Q. It takes a fairly long column to separate NO2 from CO2 and CO from NO. I would suggest that you use glass or fused silica coated SS columns. Chromosorb 102 would be another choice.

I have not performed this separation and have no idea if the data is valid as it was collected many years ago. I do not have the reference or the person who did the orginal work.

I wish I could offer a solid solution. The only other suggestion I could offer would be to use a thick film methyl silicone or hydrocarbon grease capillary (backpressured to improve separation if possible) and subambient temperatures.

A mass spectrometer may be needed for this analysis.

best wishes,

Rod

[JGK]

Working at those levels you may be better off working with a commercially available NO/NOX analyzer.

[AICMM]

daffonce,

I have a customer who has done NO at less than 1/2 ppm. Did not do NO2 but that should be the easier of the two. Used a Hayesep P column set. Biggest issue is with consistent results initially. Lots of runs and the results are more consistent which makes some sense with a reactive component like NO. At higher levels I should think you will have less of an issue with this. Used an argon ionization detector but you should be able to do them with a standard PID.

Best regards.