This project develops accurate ab initio electronic structure methods that greatly reduce scaling with system size, to facilitate the application of these methods to heterogeneous catalysis for reactions that are important to the Department of Energy. The methods developed in this research will be generally applicable to heterogeneous catalysis and associated interfacial phenomena. The developed electronic structure methods will be interfaced with advanced methods in non-equilibrium statistical mechanics that will enable the study of diffusion processes in nanopores, as well as the chemical reactivity.
Computational investigations of heterogeneous catalysis require the explicit incorporation of thousands of atoms in the simulations. Consequently, such simulations require algorithms that are able to take effective advantage of pre-exascale and exascale computers that are expected to become available within the next 4-5 years. The electronic structure method development will focus on fragmentation methods, the development or new dataflow algorithms, the exploration of novel computer architectures that can potentially reduce the energy/power consumption without significant adverse effects on time-to-solution, the development of interoperability among three no-cost electronic structure programs, and the creation of a quantum chemistry data base (QCDB) that can be used by any electronic structure program, all enabled by machine learning paradigms.
This work is supported by the US Department of Energy, Office of Science, Basic Energy Sciences, Division of Chemical, Sciences, Geosciences, and Biological Sciences through a Computational Chemical Sciences project. Ames Laboratory is operated by Iowa State University under Contract DE-AC02-07CH11338.
Principal Investigator: Mark Gordon
Staff Scientists: Yong Han
Post Docs: Dipayan Datta