Density Functional Theory of Electronic Structure

Temple University, Philadelphia PA

Investigators

Abstract

NONTECHNICAL SUMMARY The Division of Materials Research and the Chemistry Division contribute funds to this award that will lead to more accurate computer modeling of molecules, chemicals, and materials. To do this the PI will focus on the "glue" that binds one atom to another to form molecules and materials which has the technical name exchange-correlation energy. Making this energy smaller strengthens the "glue" because electrons avoid close approaches to other electrons which have the same electric charge. The density of electrons determines the exchange-correlation energy, but the exact formula is not known. Nevertheless, it is possible to use a computer to predict what molecules and materials can exist, and with what properties, by using an approximate formula. The PI has developed approximate formulas that lead to good predictions of the properties of many materials and molecules as compared with actual experiments. In this research the PI will develop even more accurate approximations for the "glue" that still lend themselves to efficient simulation of molecules, and materials on a computer. The PI's most recent approximate formula is called SCAN. As a feature, it shares all the attributes of the exact formula that are known from fundamental principles of quantum mechanics that are possible for an approximation like SCAN. Nevertheless, no formula of this type can be exact; approximations of this type introduce a spurious interaction of an electron with itself. A major goal of this project will be to develop a widely-useful correction to overcome this source of error for SCAN. The development of more accurate approximations for the "glue" that holds atoms together leads to better predictions for the properties of chemicals, molecules, and materials. These predictions can lead to the discovery of new materials with desired properties for a wide range of applications from building and construction, to sophisticated electronic devices, to biomaterials for medical applications, and more. This research enables better computer modeling of materials with potential impact on the Materials Genome Initiative. This award also helps support the PI's efforts to develop better ways to help educate more high-school physics teachers,and to involve undergraduate students in the research. TECHNICAL SUMMARY The Division of Materials Research and the Chemistry Division contribute funds to this award that supports research in Kohn-Sham density functional theory, the most widely-used method to calculate ground-state energies or energy differences, equilibrium nuclear positions, and electron densities in atoms, molecules, and solids. The theory is exact in principle, although in practice the density functional for the exchange-correlation energy must be approximated. In the preceding award period, the Perdew research group developed SCAN, a "strongly constrained and appropriately normed" functional that is more accurate for diversely-bonded systems than comparably-efficient approximations. Satisfying all 17 known exact constraints that a semilocal functional can, the SCAN meta-generalized gradient approximation could replace the widely-used Perdew-Burke-Ernzerhof generalized gradient approximation. In the current award period, extensive tests will be made for SCAN and for SCAN with a long-range van der Waals correction. These tests would include the formation energies which determine relative stabilities of molecules and solids from the elements in their standard states, large reference-data sets for sp-bonded and transition-element-containing molecules, the polyacetylene chain, the ground-state crystal structures of solids that have proven challenging to get correct for , fundamental band gaps of solids, surface energies and work functions of metals, and adsorption energies for molecules on surfaces, A refined self-interaction correction to SCAN would seek to preserve the excellent SCAN description of equilibrium bonds for weakly-correlated systems while improving the description of stretched bonds, charge transfers, and strongly-correlated systems. A SCAN-like additive correction to the fully-nonlocal random phase approximation would be developed. Proof would be sought for a hypothesized tight lower bound on the exchange energy of any spin-unpolarized density. Finally, exchange-correlation energy differences for selected systems would be analyzed in a way that could lead to a better understanding of electronic systems and their properties, and of the successes of SCAN. The development of more accurate approximations for exchange-correlation functions will result in better predictions for the properties of atoms, molecules, and materials and can lead to the discovery of new materials with desired properties for a wide range of applications from building and construction to sophisticated electronic devices to biomaterials for medical applications and more. This research enables better computer modeling of materials with potential impact on the Materials Genome Initiative. This award also helps support the PI's efforts to develop better ways to help educate more high-school physics teachers, and to involve undergraduate students in the research.