Radiative Cooling

last updated 2026-08-31

Physics / mechanism

Radiative cooling exploits thermal emission to move heat from a surface to a cold sink without pumps, compressors or working fluids. Two distinct engineering problems appear in the sources. The first is passive daytime radiative cooling at the Earth’s surface, where a coating must reject incoming solar energy while emitting its own thermal radiation to the sky. The second is heat rejection in vacuum, where radiation is the only available transport mechanism and the design variable is radiator area and mass.

For terrestrial coatings, one implementation is a disordered “photonic glass”: an optically thick, randomly packed layer of silica microspheres that scatters solar wavelengths strongly while the silica itself provides infrared emission. Once the coating is thick enough to operate deep in the multiple-scattering (diffusive) regime, both the spectral response and the net cooling performance become insensitive to microsphere diameter across 2 to 8 µm, despite measurable differences in the microscopic scattering properties. This weakens the usual assumption that particle-size optimisation is the critical design parameter and shifts the emphasis to layer thickness, packing and surface functionalisation.

The related solar-rejection problem for glazing is spectral rather than diffusive. Because more than 70% of incident solar energy lies below 1000 nm, useful cooling requires abrupt transitions from transparent to reflecting at both edges of the visible band, which conventionally implies tens of dielectric layers. An eight-layer planar stack has been demonstrated with sharp reflectance changes near 390 and 680 nm, giving above 70% visible transmittance with above 80% near-infrared and above 60% ultraviolet rejection.

In vacuum the constraint inverts: emissivity is easy, area is expensive. The ISS radiator system rejects only about 70 kW using roughly 325 m² of hardware at a reported $340-500M, which sets the scale of the area, mass and cost burden for any megawatt-to-gigawatt orbital compute platform ref. Separately, plasma physics uses “radiative cooling” for a different mechanism, in which radiation losses are dynamically significant enough to alter magnetic reconnection and particle acceleration; that regime is a research topic rather than a thermal-management technology.

Competitive landscape

ApproachSinkKey metric in sourcesStatus
Silica microsphere paintSkyCooling robust to 2-8 µm particle diameter in diffusive regimeOutdoor measurement, scalable disordered coating
Multilayer solar-rejecting glazingReduced heat gain rather than net emission>70% visible, >80% NIR and >60% UV rejection in 8 layersLab-realised structure
Space radiatorsVacuum~70 kW per ~325 m², $340-500M for the ISS system refFlight-proven but area- and cost-limited

Paints and glazings are complementary parts of a building envelope, not direct substitutes: the coating provides net radiative heat rejection from opaque surfaces, while the window stack suppresses solar gain through the transparent aperture. The vacuum case is a separate market where radiative cooling has no alternative, and where the sources treat thermal as the largest structural constraint on orbital data centres ref.

Evidence base

Frontier (open questions)

Synthesised 2026-08-31 from 6 KB sources by the resynth pipeline; citations are KB source slugs.

Recent mentions

Frontier questions