Like you, I found the one reactor setup had limitations. I transformed our Flash EA to a two reactor system with little difficulty. I set the system up so that I could switch between a two reactor combustion system and the HTC configuration. It’s a little simpler if you remove the HTC option altogether, but that’s your choice.
Basically, I rerouted the SS capillaries so the EA carrier gas goes through the two reactors, then flows to the EA water trap, and the REF line bypasses what would be the HTC reactor and goes straight to the trap/GC/TCD in the EA. I left the oxidation reactor on the right side so I could still switch back to an HTC configuration when needed. For the connection of the oxidation reactor to the reduction reactor, I used Fisher part number 350 08427 reactor coupling unions (~120 Euro). If your HTC reactor is not a bottom feed setup, I think the bottom coupling is the same as for the oxidation reactor. The top of the reduction reactor (outflow) has a different coupling than the bottom one, with a larger flange that is threaded to connect to a carousel. For some reason, I had a coupling that fit the reactor tube, with a ferrule connection to SS tubing on the other end, either leftover from the installation, or from the onsite graveyard of parts from previous labs…I’m sure it is easily available from Fisher, though.
I plumbed our system with three-way and four-way Swagelok valves so that I could switch between EA configuration and HTC configuration, with or without the bottom feed option in the HTC. While not the exact same issue as you, I did have problems with our HTC bottom feed o-ring (which fits into the carbon glassy tube) becoming too hot after each batch run and degrading. None of the Fisher engineers could solve the problem, so I switched to what works—the original top feed system, but kept the option for bottom feed just in case. Maybe it’s a similar problem as yours, with the hot zone migrating lower in the reactor?
There are a few particulars to my system, such as using the same zero blank autosampler for both HTC and EA work, but if you are going simple with just an EA combustion system, it should be straight forward.
Shoot me an email off list if you want pics or additional details.
If you run large numbers of C and N samples, the two reactor system is much, much better. I’ve never looked back after modifying.
Centre for Earth Evolution and Dynamics
Senior Engineer CLIPT Lab
University of Oslo
PO Box 1028 Blindern
0315 Oslo, Norway
E-mail: [log in to unmask]