Suppose I have a gas stream that I suspect to contain hydrogen in ~0.1-1% concentrations. Can anyone think of a way to detect that hydrogen in a
semi-quantitative way without taking samples to a GC? Surely there is some metal oxide that could be reduced or some solution that would undergo a
measureable change on exposure to a hydrogen containing flow but I can not think of any.
I think my only play is electrochemical measurement with Pt electrodes but that will be irritating to set up and standardize.
Any ideas?watson.fawkes - 15-10-2008 at 11:32
It's hard to have specific ideas for you without knowing what else is in the gas stream. A quite common way of detecting gasses is measuring the heat
generated by catalytic oxidation with a selective catalyst. All such devices I've seen have false positives and catalyst poisoning to worry about.
Detection thresholds can be quite low in the right circumstances.Fleaker - 15-10-2008 at 15:23
What else is in the gas stream? Perhaps try something with palladium sponge or foil or the like and get the amount of H2 adsorbed via mass. Just a
thought.Hang-Man - 16-10-2008 at 05:41
The gas stream will contain primarily argon and small quantities of low molecular weight organic fragments. No oxygen. The theoretical maximum amount
of hydrogen per experiment is about half a mol non-uniformly emitted over a 90 minute period in an argon flow rate of ~100 sccm.
I'm beginning to think the only way is to take gas samples throughout the run and have them run on the GC. We don't have thermal conductivity detector
however which I believe is the best way to detect hydrogen. I guess I'll have them sent out watson.fawkes - 16-10-2008 at 09:42
I don't know what your fixed-cost vs. variable-cost position is, but if you need to do this a lot and continuously you might consider a mass
spectroscopy. You'd heat a sacrificial fraction (quite small) of the gas stream with RF centered at the H<sub>2</sub> ionization energy
(to get preferential ionization) and then split the stream into three: non-ionized particles, H<sub>2</sub> ion, and the heavier ions.
This need only be partly quantitative, because all you care about is a single fraction. Calibration just for that fraction is substantially easier
than for everything. A calibration gas mixture can be mixed up from argon, hydrogen, and methane.unionised - 16-10-2008 at 09:44
At that concentration I think you could probably make a thermal conductivity detector from two broken light bulbs arranged in a Wheatstone bridge
circuit. Almost certainly worth a try.Twospoons - 16-10-2008 at 13:48
There's three for you, just by searching ' Hydrogen gas sensor' on google. The Figaro sensors are usually fairly cheap.Hang-Man - 16-10-2008 at 14:33
Mass spec with the resolution required to detect hydrogen is certainly out of my price point. I only need to detect it once for proof of concept;
that would pay for a GC with a TCD.
I googled the hell out of hydrogen detection but only got safety related items and binary detection devices. I'm not sure if those will work, most
hydrogen detectors are geared towards detection in air, as illustrated by the mode of operation of the figaro sensor:
"These elements are installed into a “Wheastone Bridge”. A variable resistor should be adjusted so that the bridge will produce a stable baseline
signal when in an environmnet free of combustible gases. When combustible gases are present, they will be combusted on the detecting element, causing
its temperature to rise. Accordingly the resistance of this element will increase"
Although I will look into them further, I don't think they'll be suitable for a flow containing just H2 Ar and hydrocarbon fragments. The second one
may work, I'll have to look into it.
If I bubble the solution through an electrochemical cell I should be able to oxidize some of the gas to protons using Pt electodes and measure the
current? Any electrochemists in the room?
[Edited on 16-10-2008 by Hang-Man]watson.fawkes - 16-10-2008 at 15:09
Quote:
Originally posted by Twospoons
The Figaro sensors are usually fairly cheap.
They're great little devices, but they all rely on selective,
catalytic oxidation in ordinary atmosphere with an excess of oxygen.watson.fawkes - 16-10-2008 at 15:16
Quote:
Originally posted by Hang-Man
Mass spec with the resolution required to detect hydrogen is certainly out of my price point. I only need to detect it once for proof of concept;
that would pay for a GC with a TCD.
If only once, then disregard my suggestion. I was interested because
it's only about half of a full mass spectrometer. If you can use GC readily, just go do that. The extra work involved in any specific solution would
seem to mitigate against anything particularly fancy, especially since an off-the-shelf solution for what you're looking for doesn't seem to exist.
Quote:
If I bubble the solution through an electrochemical cell I should be able to oxidize some of the gas to protons using Pt electodes and measure the
current? Any electrochemists in the room?
I doubt you'll be able to discriminate hydrogen from the other
organic compounds, as there are plenty of other bonds in your system with less energy that the diatomic hydrogen bond.
