Oscar asked about the "smoke detector" mounted low on the living room wall in our new house. You know, the one that emits a piercing shriek when you press the test button, conveniently located at toddler eye-level.
"That’s a carbon monoxide detector." Mark explained that it will go off if a certain poisonous gas builds up in our house, to warn us that we need to leave.
"How does it work?"
"Um, it…" I said. Then stopped. Then looked at Mark. "How does it work?" He didn’t know either. Google to the rescue!
Turns out there are three types of CO detectors, all of which rely on chemical reactions within the detector. Metal oxide semiconductor detectors contain heated tin oxide or platinum oxide, which reacts with CO. Biomimetic detectors contain gel-coated disks that darken in the presence of CO, triggering an alarm. Electrochemical CO detectors react with CO to generate electric current, again triggering an alarm.
What piqued my interest: This chemistry forum discussion describes the two competing reactions in a common platinum-oxide detector:
[1] PtO + CO —-> CO2 + Pt (exothermic)
[2] 2Pt + O2 —-> 2 PtO
According to the forum, the heat liberated in reaction [1] raises the temperature of a sensor which in turn triggers the alarm. Reaction [2] is the regeneration reaction. They would exist in equilibrium in any given atmosphere.
I’m guessing that the detector contains a high-surface-area sample of platinum, perhaps the sensor itself, on which there’s normally an oxide coating. If the CO concentration in the air rises, reaction [1] speeds up relative to reaction 2, the oxide coating starts to disappear, and the platinum gets warmer. If it falls again, reaction [1] slows down and reaction [2] speeds up, and the oxide coating reappears.
It looks at first glance that high levels of CO2 might tend to drive the first reaction back, but if the reaction is indeed very exothermic, then only a slight concentration of CO would drive it forward while it would require a large level of CO2 to drive it backward.
Seems to me that the humble CO detector would be a great basis for a textbook problem in kinetics, surface chemistry, or physical chemistry. Given the free energy tables and a few pieces of data, it should be straightforward to calculate (for example) the sensitivity of the reactions to the relative concentrations of CO, CO2, and O2. Why, you could even bring mass and energy transport into the picture. How fast can the molecules of CO diffuse from the bulk atmosphere to the surface? And how fast does the reaction generate heat, compared to the rate at which the sensor dissipates heat?
And what did Oscar think about it? I don’t know. He wandered away while his dad and I were discussing this very interesting problem.
bearing blog 