About 40% of the US annual energy consumption comes from buildings, largely in the form of heating and cooling. A promising way to reduce that consumption is designing building materials that let off heat when it’s hot outside and let in heat when it’s cold. One way to do that is passive radiative cooling. Materials employing that method block the Sun’s warming rays in the near-IR and then efficiently cool by emitting radiation in the mid-IR and long-wave IR, which passes easily through the atmosphere to the heat sink of space.

Radiative cooling, as the name suggests, serves warm climates well. But in cold months, the approach is counterproductive. Efforts to create a system that radiatively cools when hot but doesn’t when cold have thus far required energy input or other external activation. Now two independent research teams have developed material coatings that passively switch on and off their radiative cooling depending on the ambient temperature. They demonstrate the possible energy savings for windows and roofs.

The researchers used a layer of tungsten-doped vanadium dioxide, which transitions from an insulator to a metal at a temperature determined by the doping level. In the insulating state, the W-doped VO₂ is largely transparent to IR, and in the metallic state, the layer becomes highly absorptive. In addition to the transition temperature, the doping level adjusts how efficiently the VO₂ layer emits heat.

Yi Long of Nanyang Technological University in Singapore led one of the teams. She and her colleagues created a sample window of about 5 cm × 5 cm. They spin coated a commercial glass with W-doped VO₂ and a spacing layer. Optical transmittance measurements revealed efficient cooling in the form of a high long-wave IR emissivity of 0.61 at higher temperatures. (A perfect blackbody emitter, by comparison, would have a value of 1.) Below about 27.5 °C, that emissivity plummeted to 0.21.

To investigate the potential energy savings in practice, Long and her colleagues simulated single-story and 12-story buildings in climates around the world. The baseline annual energy consumption of a 12-story building with clear glass windows is up to 4668.6 MJ/m². Their dynamic windows saved up to 115.9 MJ/m² annually compared with the best energy-efficient windows without dynamic regulation and up to 324.6 MJ/m² compared with commercial low-emissivity glass.

Junqiao Wu of the University of California, Berkeley, and his colleagues tested how W-doped VO₂ might fare as a roof-tile coating. They settled on an IR-enhancing optical structure and an ideal W doping that, in laboratory spectral measurements, yielded emissivity as low as 0.2 for temperatures below the 22 °C transition and up to 0.9 for temperatures above it.

Wu and his colleagues also measured the sample’s temperature over the course of a sunny summer day on a rooftop in Berkeley. Compared with commercial roof coatings that have a consistent emittance of 0.9, theirs provided about the same cooling during the heat of the day and stayed about 2 °C warmer at night and in the early morning. For a single-family home in Baltimore, Maryland, the researchers predict their material should offer minimum annual energy savings of 22.4 MJ/m².

Moving forward, Long and her team plan to improve on the performance of their current composite. Meanwhile, Wu and his team intend to test larger-scale versions in real-world conditions. Beyond building materials, W-doped VO₂ could be useful for electronics, cars, or other objects that need shielding from extreme temperatures. (S. Wang et al., Science 374, 1501, 2021; K. Tang et al., Science 374, 1504, 2021.)