Mass-produced chemicals – from plastics to fertilizers – often involve chemical reactions that require intense heat and/or pressure, and hence significant amounts of energy. A promising alternative is to deliver energy to drive these reactions in a more controlled way in the form of light, in a phenomenon known as photocatalysis. Several light-driven reactions have been demonstrated with up to 100x improvement in their speed and efficiency compared to purely thermal processes. Still, the underlying mechanisms in photocatalysis are poorly understood, making it harder to systematically develop efficient light-driven reactions. This research will combine massively parallel computer simulations at the Frontier and Aurora supercomputers with modern, quantummechanical theories to understand photocatalysts with unprecedented accuracy and generate new design principles. 

The calculations in this research will allow one to understand several key steps that dictate the processes of photocatalysis, including: 1) the absorption of light by a material (the photocatalyst); 2) the transfer of photon energy to the chemical reaction; and 3) the resulting evolution of the reaction with time. This research will study prototypical photocatalytic materials such as titanium dioxide and silver nanoparticles and derive connections between materials properties, such as their electronic properties and geometries, with photocatalytic efficiency. Such a connection is critical since materials that are good light absorbers are not necessarily good at driving chemical reactions. Our research has the potential to provide a standardized and systematic way of evaluating and engineering photocatalysts, which will greatly accelerate the development and deployment of this technology and address several of the DOE’s missions, such as understanding and optimizing the energy flow at the nanoscale.