Kieron Burke of the University of California-Irvine is supported by an award from the Theoretical and Computational Chemistry program to carry out research on the development of improvements to density functional theory (DFT). The improvements focus on the development of new functionals including a kinetic energy functional, which would eliminate the need to compute orbitals.
Density functional theory has become an essential tool of electronic structure calculations in many fields of chemistry, but there exist so many empirical functionals that users are often bewildered about how to start using the method for their particular application. The primary objective of this project is to develop an entirely new methodology for developing appropriate density functionals using the semiclassical theories on which it is based. The approach uses a simple principle, that of asymptotic exactness as the number of electrons becomes large. The impact of Burke's work is expected to be broad and to have application in many fields of science. The impact of the work will be further broadened by the development of tutorials and distribution of Burke's "ABCs of DFT" online notes.
The project led to many publications in top scientific journals, both within the Burke research group and in collaboration with several other scientific groups. Each represents a fundamental step forward in applying density functional theory to atoms, molecules, and solids. Examples include Density functional partition theory, a new way to think about and calculate the properties of molecules and solids from effective fragment (e.g., atomic) calculations Potential functional theory, an interesting alternative to density functional theory, which uses the potential as the basic variable, and in which extremely accurate and systematic approximations can be derived. Electron affinities with density functional approximations: Accurate calculation of electron affinities is important for many biological species, but standard approximations appear to be flawed for these systems. We showed how such calculations can be trusted, and produced a better method for doing them. We calculated the ionization potential of neutral atoms as the atomic number goes to infinity. This showed that very simple density functional approximations become exact (in a very subtle sense) in this limit, supporting very deep work of applied mathematicians over the past several decades. Some will ultimately find their way into the codes used by thousands of researchers worldwide, to improve simulations in chemistry, physics, and materials science. Furtherwork was also performed on educational materials for improving both graduate and undergraduate education in density functional theory.
