Development of Novel Exchange-Correlation Functionals & Applications

William Marsh Rice University, Houston TX

Investigators

Abstract

Gustavo Scuseria of Rice University is supported by an award from the Theoretical and Computational Chemistry program for research to develop more accurate and efficient approximations in density functional theory (DFT). The work is expected to significantly enhance DFT's applicability to ground and excited electronic states of both molecules and solids. The projects being pursued involve concepts such as local hybrids and local range separation. The new functionals being developed in the Scuseria group are being applied to carbon nanotubes and graphene nanoribbons, both of which are promising materials for nanotechnology applications. Computational quantum chemistry has provided essential tools for predicting chemical properties and interpreting new chemical phenomena. Kohn-Sham density functional theory (DFT) is the most widely used method in quantum chemistry. The strongest appeal of DFT is that it delivers usefully accurate predictions at a computational cost much lower than other techniques such as correlated wavefunction methods. Application of linear scaling techniques can make DFT affordable even for very large systems (both molecules and solids). This makes DFT the only method available for first-principles prediction of quantum effects such as conductance in these systems. While DFT is formally exact, in practice it requires approximating a large component of the total energy: the exchange-correlation interaction. DFT's accuracy, and ultimate usefulness, is limited by the quality of approximations to the exact exchange-correlation functional. Existing approximate functionals are reasonably accurate, but lose their edge over modern wavefunction methods for many properties. The PI's research group has been active in the development of novel exchange-correlation functionals for the last several years. Many of the functionals developed in his group (e.g., VSXC,TPSS, HSE) are widely used in academia, national laboratories, and industry. The current work is expected to have an equally broad impact and to contribute to the training of students at all levels.