As I said on Twitter, I found out this week (1) that there are cheap variable resistors sensitive to acetone (or ethanol) and (2) that many people, even scientists, don’t think this is amazing.
Multi-atom molecules are hugely bigger than electrons, but hugely smaller than the bulk semiconductor. They shouldn’t be able to affect resistance.
Other technologies for interfacing chemistry and electronics tend to be based on light absorption (breathalysers, blood oxygen detectors, some DNA sequences) or on electrons/protons released in chemical reactions (other breathalysers, other DNA sequencers) or in past days on formation of ions in solution. It’s reasonably clear how these translate between the molecular scale and the electron scale.
Looking on Google, it seems that the details of gas-sensitive resistors aren’t completely understood (or weren’t in 2009). The basic idea is that oxygen atoms attach to the surface of semiconductor grains and suck electrons. Reducing gases such as acetone displace some of the oxygen atoms, and don’t suck electrons, leaving more electrons free for conduction. Since semiconductors can be tuned to have just barely enough free electrons to conduct, the surface effect of the oxygen suckage can be enough to matter.
The application that brought this up was measuring breath ketones for someone on a low-carb diet, with a DIY device that cost an order of magnitude less than the commercial version. That’s clever, but I’m still more impressed that there’s a way to measure acetone concentrations with a simple resistor.