A bolometer is a sensor that detects radiation by absorbing it and measuring the resulting temperature change. Unlike sensors that respond only to visible light or specific wavelengths, a bolometer can pick up energy across an extraordinarily wide range of the electromagnetic spectrum, from infrared and microwaves to submillimeter waves. This makes it one of the most versatile detection instruments in physics, astronomy, and thermal imaging.
How a Bolometer Works
The core principle is simple: radiation hits a material, the material heats up, and that tiny temperature increase changes its electrical resistance. The bolometer then converts that resistance change into an electrical signal you can read. It’s essentially a very precise thermometer for light.
Every bolometer has the same basic architecture. An absorber captures incoming radiation and converts it to heat. A temperature-sensitive element (the sensor) sits in thermal contact with that absorber. Its resistance shifts as its temperature rises, and that shift is the measurement. A weak thermal link connects the whole assembly to a heat sink, which acts as a stable reference temperature. The thermal link is deliberately weak so that absorbed energy raises the sensor’s temperature noticeably before the heat slowly drains away to the sink. Without that bottleneck, the temperature rise would be too small and too brief to measure.
The absorber needs to be carefully engineered. To capture the maximum amount of incoming radiation, its electrical resistance per unit area is matched to the impedance of free space (about 377 ohms). The absorber is also designed with materials that spread heat quickly and evenly, so the sensor registers a clean, fast signal rather than a sluggish, blurred one.
Key Performance Measure: Noise Equivalent Power
Bolometer sensitivity is measured by something called noise equivalent power, or NEP. This is the smallest amount of incoming power the detector can distinguish from its own background noise. Lower is better. A silicon bolometer cooled to liquid helium temperatures approaches the theoretical sensitivity limits of thermal detectors, with NEP values in the range of hundreds of zeptowatts per root hertz. (A zeptowatt is a trillionth of a billionth of a watt.) Recent advances have pushed that even lower. A 2024 study published in Communications Physics reported a record NEP of about 20 zeptowatts per root hertz using a nanobolometer paired with a specialized amplifier, with response times as fast as 30 microseconds.
Cooled vs. Uncooled Bolometers
Bolometers split into two broad categories based on operating temperature, and the difference matters enormously for what they can do.
Cryogenic Bolometers
The most sensitive bolometers operate at extremely cold temperatures, sometimes below 1 kelvin (roughly negative 457°F). Cooling the sensor this far reduces thermal noise, which is the random jiggling of atoms that can mask faint signals. Cryogenic bolometers are the instruments of choice for detecting the faintest signals in the universe. They were first used to measure cosmic microwave background radiation, the afterglow of the Big Bang, from the South Pole in the early 1990s, with sensors cooled to 100 millikelvin. Silicon bolometers cooled to liquid helium temperature (4.2 kelvin) can detect radiation across a spectral range of 2 to 3,000 micrometers, covering far more of the spectrum than most competing technologies.
The tradeoff is practical. Cryogenic cooling requires bulky, expensive refrigeration systems. That makes these bolometers poorly suited for long-duration space missions or portable applications.
Uncooled Microbolometers
Microbolometers operate at or near room temperature. They’re far less sensitive than their cryogenic cousins, but they don’t need a cooling system, which makes them smaller, lighter, cheaper, and more durable. The most common type uses vanadium oxide as the temperature-sensitive material deposited onto a tiny suspended membrane. Arrays of thousands or millions of these microbolometers form the focal plane in thermal imaging cameras, where each microbolometer acts as a single pixel.
NASA has evaluated uncooled microbolometer technology as a potential replacement for cooled quantum-well detectors on future Landsat satellites, specifically because eliminating the cryocooler would reduce instrument mass, cost, and complexity while still delivering useful thermal imaging from orbit.
How Bolometers Compare to Other Detectors
Bolometers aren’t the only way to detect infrared and longer-wavelength radiation, but they fill a niche that other technologies can’t.
- Photon detectors (like those based on mercury cadmium telluride or quantum-well structures) respond to individual photons knocking electrons loose in a semiconductor. They’re fast and sensitive, but they only work across a narrow wavelength band, typically 3 to 16 micrometers. They also usually need cooling.
- Pyroelectric detectors generate a signal when their temperature changes rapidly, responding to the rate of temperature change rather than the temperature itself. They work at room temperature across a broad spectral range (1 to 1,000 micrometers), but they can only detect changing signals, not steady ones.
- Bolometers respond to total absorbed energy regardless of wavelength. A silicon bolometer covers 2 to 3,000 micrometers. They measure both steady and changing signals. Cryogenic versions achieve the highest sensitivity of any thermal detector, while uncooled versions are cheaper and easier to fabricate than photon detectors, with good uniformity across large arrays.
The choice between these technologies comes down to what you need. For broadband coverage and extreme sensitivity, bolometers win. For speed and narrow-band precision at mid-infrared wavelengths, photon detectors are better. For a cheap, room-temperature broadband option that only needs to detect changes, pyroelectric detectors work well.
Where Bolometers Are Used
Bolometers show up in a surprisingly wide range of fields. In astronomy, cryogenic bolometers are the primary sensors for studying the cosmic microwave background and for submillimeter-wave observations of cold dust and gas in distant galaxies. These measurements require detecting radiation so faint that only the most sensitive instruments on Earth can pick it up.
In everyday life, uncooled microbolometer arrays are the technology inside most thermal imaging cameras. These are used for night vision systems, building energy audits (spotting heat leaks through walls and windows), electrical infrastructure inspection, environmental monitoring, and firefighting. Search-and-rescue teams use handheld thermal cameras with microbolometer sensors to locate people in smoke, darkness, or rubble after disasters. Forest fire prevention systems use long-wavelength infrared bolometers to detect hotspots before they become visible flames.
Remote thermometers, including some of the non-contact forehead thermometers that became ubiquitous during the COVID-19 pandemic, rely on thermal detection principles closely related to bolometer technology. Security systems also use infrared bolometer arrays for perimeter monitoring and intrusion detection.
A Brief Origin Story
The bolometer was devised in 1881 by Samuel Pierpont Langley, an American astrophysicist who later became the third Secretary of the Smithsonian Institution (and a pioneer in powered flight). Langley built his bolometer to measure the infrared spectrum of the Sun with unprecedented sensitivity. His original instrument used thin platinum strips whose resistance changed with temperature. It was sensitive enough to detect the heat from a cow at a quarter mile away, a demonstration that captured public attention at the time. The basic operating principle, absorbing radiation, measuring a resistance change, has remained the same ever since, though modern materials and fabrication techniques have improved sensitivity by many orders of magnitude.