Mesoporous Aerogel-Based Infrared Metamaterials for Radiative Cooling of Windows

Heat transfer through windows accounts for 25-30% of a building’s overall heating and cooling demands, corresponding to approximately 5-6% of the 100 quadrillion BTUs of energy consumed annually in the United States. Improving the thermal management of windows, as well as opaque building elements, offers a significant opportunity for enhancing energy efficiency and decarbonization in the face of climate change. Recent work has demonstrated that vertically oriented surfaces such as windows and walls can optimally harness radiative cooling to the sky while simultaneously reflecting thermal radiation from the hot ground by exhibiting selectivity in their emissivity between 8 and 13 microns in wavelength. In this work, we introduce a novel approach for the radiative cooling of windows using mesoporous organo-silica aerogels engineered as infrared metamaterials to serve as optimal radiative cooling materials for windows. These aerogels take advantage of both the intrinsic phonon-polariton resonances of SiO2 as well as a metamaterial effect enabled by their high porosity to exhibit highly selective behavior in their emissivity within the 8-13 microns atmospheric transparency window, while simultaneously reflecting background radiation at other wavelengths. To optimize the aerogel films for maximum infrared-selective performance, we employ an effective medium theory simulation framework alongside machine learning-driven inverse design. Through simulations, we optimize the composition, porosity, and thickness for maximal infrared selective behavior and thus performance in both window and opaque applications. We incorporate experimental optical data, including refractive index and extinction coefficients, into our machine-learning models to enhance accuracy and ensure that the designs closely reflect the properties of the synthesized aerogels. Traditional effective medium models rely on bulk silica’s complex refractive indices, but in mesoporous structures, we discovered that the phonon-polariton resonances behave differently due to an increase in surface area and change in bond-angle distributions in the systems. To account for this, we developed a Lorentz oscillator model combined with machine learning techniques to perform a broad parameter sweep. The resulting model accurately captures the shifts in phonon resonances, validated through Raman spectroscopy of the synthesized aerogels. This modeling approach has enabled the synthesis of aerogels with selective emissivity within the atmospheric transparency window. We fabricated radiative cooling devices by bonding optically clear, 100 microns-thick aerogel samples to aluminum substrates using 5-micron-thick PDMS layers. Infrared spectroscopy of these devices revealed low absorption from 4-8 microns and high absorption beyond 8 microns. Outdoor experiments demonstrated the vertically oriented device was consistently ~1.5°C cooler than a broadband emitter over our daytime measurement window. Our findings suggest that these visibly transparent, yet infrared-selective thin-film aerogels offer a promising route to controlling heat flow through windows, potentially reducing the energy consumption of buildings.