In this project, the potential for passive cleaning of outdoor air using photocatalytic titanium dioxide particles was investigated. Small particles (nanoparticles) of anatase titanium dioxide can serve as photo-catalysts to assist in the oxidation of atmospheric pollutants such as oxides of nitrogen (NOx) and volatile organic compounds (VOCs). The energy enabling this catalysis is contained in the ultraviolet part of the solar spectrum. By-products may be produced, but the ultimate fate of the pollutant molecules is to be oxidized ("mineralized") to H2O, CO2, and acids such as HNO3 and H2SO4. Active indoor air cleaners using this technology are now commercially available. They can, for example, eliminate malodorous compounds in the air. Research and development in this area has been ongoing for more than a decade in Japan, and recently a major multi-institutional project has been completed in Europe. Despite the extensive research to date, the practical feasibility of passive photo-catalytic air cleaning is still a controversial subject. Some individuals claim that pollution can be halved with straightforward measures; others are unconvinced. This project has clarified the technology potential by focusing on measured values of catalytic activity. Catalytic activity is the rate at which air can be cleaned by a given area of photocatalyst. For example, one square meter of catalytic film or plate can clean a certain number of cubic meters of air per day. The units of catalytic activity are m d-1. Laboratory data show that a high-quality TiO2 catalyst has an activity of about 200 m d-1 for NOx, about 60 m d-1 for typical volatile organic compounds, and approximately 0 m d-1 (i.e., not useful) for carbon monoxide. Prior to widespread deployment of passive photo-catalytic air-cleaning technology, large-scale meteorological simulations are needed to validate deployment strategies.

Fluorescent cool dark surfaces stay cool in the sun by reflecting near-infrared (NIR) radiation and by actively re-emitting in the NIR spectrum some of the energy absorbed from visible sunlight. The fraction of incident solar energy rejected by reflection and fluorescence is the “effective solar reflectance”, or ESR, of the surface. It is challenging to measure ESR with a solar spectrometer or a solar reflectometer, the radiometric instruments most commonly used to measure the solar reflectance (SR) of specimens in the laboratory. We have tested a variety of calorimetric techniques for using temperature in the sun to interpolate the effective solar absorptance (1 − ESR) of a fluorescent test specimen from the known solar absorptances of non-fluorescent reference specimens. Our experiments show that averaging out noise in the temperature signal induced by variations in convection is key. We developed a computer-controlled rotary apparatus that compares the temperatures in the sun of up to six specimens. Trials on six different fluorescent specimens indicate that it can measure ESR with a repeatability of about 0.02. To maximize the ratio of signal to noise in temperature determination, and to facilitate calculation of the fluorescence benefit (ESR − SR), measurements should be performed with specimens facing the sun.

Albedo (solar reflectance) can be measured outdoors with a pyranometer or indoors with a hemispherical reflectometer or a spectrophotometer. The current study evaluates these methods and their applicability to roofing materials by measuring and comparing the ASTM E1918 (pyranometer), non-ASTM E1918A (alternative pyranometer), ASTM C1549 (reflectometer), and E903 (spectrophotometer) albedos of 10 roofing products, including three single-ply membranes, one asphalt shingle, three roofing aggregates, and three high-profile tiles. It uses full-size (4 m × 4 m) assemblies in the E1918 and E1918A trials; corrects the E1918 albedos to remove shadow and background errors; and evaluates C1549 and E903 albedos with two different irradiance spectra, one global horizontal and the other beam normal. E1918A albedos matched E1918 corrected (E1918_cor) albedos to within 0.036. Agreement between C1549 air mass 1 global horizontal (C1549_G1) albedo and E1918_cor albedo was within 0.015 for membrane and shingle coupons, and within 0.031 for aggregates, though the uncertainty in the latter agreement was up to 0.043. After minor corrections, C1549_G1 albedos of flat or slightly convex tile chips were 0.032–0.052 higher than the E1918_cor albedos of their corresponding high-profile tile assemblies because high-profile surfaces have concavities. Switching to the beam-normal albedo C1549_1.5E increased the C1549 albedos of the 10 tested products by 0.004–0.054, with the largest increases (C1549_1.5E – C1549_G1) accruing to spectrally selective cool colors. Using C1549_1.5E albedo to characterize a chip of a cool-colored tile compounded the error induced by representing a high-profile surface by a flat or slightly convex specimen.