Adapting a 5890-II for Sulfur Testing

[xuchemguy]

I represent a manufacturer of molecular sieves. Our product is used around the world for purification of natural gas streams, including de-sulfurization. I am interested in further testing of our products for sulfur applications, specifically H2S breakthrough along with COS byproduct catalysis monitoring. We are trying to adapt our GC for this experimentation. Speciation is vital.

Our procedure was originally written to be used as an in-line analysis via mass spectrometry, but we no longer have the capability to perform the procedure as written. We want to make this work with GC. This was done by my European counterpart who has since retired, and my experience with GC is minimal.

We plan to use analytical grade nitrogen balanced with levels of H2S (5-800ppm max), COS (11ppm max), and CO2 (2% max) to test our sieve in a separate apparatus. Then we will either take samples (likely gas bags), or find a way to do in line testing with GC if possible, and test for the species of interest.

Our GC is currently a 5890-II FID with a split/splitless injector, set up for manual injection with a PC for programming/data recording etc.
From what I have read I will need a FPD or PFPD, some gas sampling valves (aka rotary valves) and a way to control them, a GS-GasPro capillary column, and a gas syringe to inject the sample (unless there is a way to do an in-line sampling via the rotary valve(s)). Would I need to purchase a board for the FPD?

I need to know if it is feasible to obtain these parts and install them with any sort of ease myself with a quick turnaround (few months). Any suggestions on where to get these parts, or any other options are appreciated. Or would it be more realistic to just buy/trade for another GC with these capabilities at the ready?

Cheers,
-MS

[CE Instruments]

It is not sensible to try to mod your 5890, the cost of doing so will be significant vs the cost of a new GC. Far better to get either a new GC or a second hand one. Forget FPD, it won't see the CO2 and you are looking at break through not seeing them in a big back ground. Look to get a GC with DID which is a universal detector and senisitive for your samples. Sample introduction by valco gas sampling valve, should be fine to set it up to sample regularly say evey two minutes.

[gapeev]

Hello,
One way to approach this:

1. Get a standalone (or a suitable add-on) DID unit from VICI, for example.
2. Fit your existing GC with a gas sampling valve or get a used GC from ebay. You only need it for flow and oven control. (Non)working detector in not an issue.
3. Put it all together and voilà!

We have a GC-MS suitable for sulfur applications. If you would like to outsource this, let's talk!

[AICMM]

xuchemguy,

Totally doable, relatively easy if your 5890 was equipped with an FID. (Standard caveat, I make the detector to be cited.) Using argon mode DBD, you can inject the sample and see both H2S and (lesser COS) with selectivity over methane but starting to get response for ethane, ethylene and others. A Plot U would work really well, a carbowax would work pretty well. Since your balance gas is nitrogen, this application is gravy.

Yes you need a sampling valve to be consistent, yes you should have a consistent sampling technique (bubble, bubble for example), yes you could put this "on-line" with some work, and yes it can be done pretty inexpensively. I have to think about if you can still use your splitless/split inlet (probably pretty easy with a plot column but I still have to chew on it.)

Feel free to contact me aicmm at flash(dot)net if you want to discuss further.

Best regards,

AICMM