I work with a type of spectrometer (made by Picarro) that uses near-infrared lasers to measure the stable isotopic composition of water vapor. The samples measured are typically liquid water samples that are vaporized at 110 °C and then carried by a flow of dry air through the measurement cavity. If the water samples contain dissolved organics that are also vaporized at 110 °C (e.g., methanol, ethanol), then the resulting spectral measurement can be impacted if those compounds also absorb light at similar wavenumbers.

For a sense of scale, we typically inject 1.8 uL of water, vaporize the sample, and then introduce the vapor over the course of about 10 minutes. The dry air carries the sample vapor through the system at ~40 mL/min.

The current solution involves using a small flow-through reactor held at ~400 °C with a body material of glass and a catalyst that efficiently combusts the organic compounds. This is placed in the instrument flow path between the vaporizer and the analyzer. However, the catalyst has a limited effective lifetime.

I am interested in making a replacement device that will resistively heat a reactor to between 600 and 1,000 °C, which should (given the carrier gas is air, thus plenty of O2) efficiently combust any organics without the need for a solid catalyst. The wide range of temperature is because I do not know a priori what temperature is necessary to effectively combust all organics. In practice, I'd end up using the lowest temperature necessary to match the performance of our existing device.

My main question is, thus: what would be the best reactor material that satisfies the following conditions: (a) able to be safely heated to 600-1000 °C, (b) not be damaged by flowing water vapor, and (c) not engage in any meaningful side reactions with the water vapor. A small (d): as I'm in an academic lab, the material ideally is also relatively cheap.

My immediate thought was to use alumina, but I am now aware that there are opportunities for reactions with water and some alumina polymorphs that might change the isotopic composition of the water during travel through the reactor. My second thought is to use quartz, which I believe may be a good option. However, I am hoping someone with more familiarity in materials science could advise.


1 Answer 1


I would recommend a platinum or platinum-iridium tube for several reasons:

  1. Platinum is non-reactive. Of course you are the chemist, so your judgement there is best - but my understanding is that it's your best bet in terms of metals.

  2. It's relatively inexpensive - not on a per-ounce basis, of course, but for a small tube it is only several hundred dollars despite the fact that no more than a handful of Pt tubes must be made each year. Sigma Aldritch sells pretty big ones for about $700 and other suppliers don't list prices but have smaller ones.

  3. It will be easy to work with. You should be able to bend the tube (which would be, I think, quite difficult for a specialty high-temperature glass). Cutting it will be extremely easy. You can attach it to low temperature components using compression fittings as long as you keep those fittings cool enough to minimize reactivity.

  4. It will conduct heat very well, many times better than glass. It will react almost instantly to changes in temperature.

That's not to say quartz or another glass/glass-ceramic is a bad choice - they will probably work well too.

  • 1
    $\begingroup$ Thanks for the reply! I think the design we have in mind wouldn't require the tube to have any bends, but you do make a good point that it would take to a compression fitting nicely. However, I would think that the more cost-effective tubes (like you link to on Sigma) have very thin walls (0.1 mm) that might actually make a compression fitting difficult. That, and I would probably err on the side of a longer hot zone. Looking at longer tubes (maybe 100 to 150 mm) with thicker walls (0.5 mm) really sees the price skyrocket. However, I'll keep Pt in mind and explore pricing a bit more. $\endgroup$ Jun 29 at 16:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.