The most widely distributed radioactive source
There are hundreds of millions of ionisation smoke detectors in the world, and each holds a small quantity of americium 241 — around 0.3 micrograms, less than a grain of salt, sealed in a foil of gold or silver.
Nothing else has put a radioactive source into so many ordinary rooms.
The physics is simple. Alpha particles from the source ionise the air between two electrodes in a small chamber, so a steady current flows. When smoke enters, its particles attach to the ions and neutralise them, the current falls, and the alarm triggers.
⚠️ The detector responds to combustion products before they are visible, which is what makes an ionisation alarm quick on a fast flaming fire.
And the other kind
Photoelectric detectors work differently: a beam of light and a sensor placed off its axis, so that smoke scattering the light triggers the alarm. No radioactive source at all.
The two respond to different fires. Ionisation alarms are faster on flaming fires with fine invisible particles; photoelectric alarms are faster on smouldering fires, which produce larger visible smoke particles and which are the more common way house fires kill.
⚠️ Photoelectric alarms are also less prone to false triggers from cooking, which matters because a false alarm is often followed by the battery being removed.
Fire services in several countries now recommend photoelectric alarms as the default, or dual-sensor units. This is a judgement about fire behaviour and human behaviour, not about the radiation, which is negligible either way.
Why this isotope
Americium 241 emits alpha particles at 5.5 megaelectronvolts, which ionise air efficiently, and only a weak gamma at 60 kiloelectronvolts, which is easily absorbed by the housing.
Its half-life of 432 years means the source strength does not change over the device’s life. A detector installed today will emit as much in fifty years as it does now, and if it fails it will not be because the americium ran out.
⚠️ And there is a surplus. Americium 241 is the decay product of plutonium 241, so it accumulates in spent reactor fuel and in stored plutonium — reprocessed plutonium has to be cleaned of it, which produces americium as a matter of course.
An element that does not exist in nature, produced as an unwanted by-product of weapons and power programmes, ended up in domestic ceilings because it happened to be the right tool.
The boy with the shed
David Hahn was a teenager in Commerce Township, Michigan, working towards an Eagle Scout badge in atomic energy. He decided to build a breeder reactor.
He collected americium from smoke detectors — buying broken ones by the box — thorium from camping lantern mantles, and radium from the luminous dials of old clocks bought at antique shops. He obtained beryllium from a chemical supplier by posing as a teacher.
His neutron source worked. His reactor did not: he assembled a pile of these materials in a potting shed in his mother’s garden and it produced no chain reaction, but it did produce rising radiation levels detectable across the property.
⚠️ He dismantled it himself when the readings alarmed him. Police stopped him with the components in his car boot, and the EPA eventually removed the shed as radioactive waste and buried it in Utah.
The account is often told as a story about a clever boy. It is more usefully a story about how much regulated material can be assembled by someone patient with a telephone.
Powering spacecraft
Plutonium 238 is the isotope of choice for radioisotope thermoelectric generators on deep space missions, and it is in short supply — production stopped in the United States for two decades.
The European Space Agency has been developing americium 241 as an alternative.
It is a worse fuel: about a quarter of the power per gram, because its half-life is longer and its decay less energetic, and it needs more shielding. It is also available, because it accumulates in stored civil plutonium and can be separated chemically without a reactor.
⚠️ The trade is mass against supply. A spacecraft powered by americium would be heavier for the same output, which is a real cost, but the material can be obtained without restarting a plutonium production line.