The local approximations to exchange-correlation functionals that are widely used in Kohn-Sham density functional theory usually underestimate band gaps and molecular excitation energies, and therefore, it becomes necessary to use more expensive hybrid functionals or more empirical DFT+U functionals for accurate predictions and modeling of these properties. This work presents a meta-generalized gradient approximation (meta-GGA) called High Local Exchange 2017 (HLE17) and illustrates how it can be useful for obtaining accurate semiconductor band gaps and molecular excitation energies. Unlike the conventional way of using the DFT+U method, one does not need to determine new parameters for every property or system studied. The HLE17 functional builds upon our earlier work (HLE16) where we had shown that by increasing the coefficient of local exchange and simultaneously decreasing the coefficient of local correlation with a GGA, the band gaps and excitation energies could be significantly improved without significantly degrading the ground-state molecular energetic properties. However, for almost every database tested in this work, HLE17 shows improvement over HLE16, and the improvement is particularly notable for solid-state lattice constants. This new functional provides a strategy for calculating properties that are otherwise difficult to calculate by a local functional. (Figure Presented).
|Original language||English (US)|
|Number of pages||11|
|Journal||Journal of Physical Chemistry C|
|State||Published - Apr 6 2017|
Bibliographical noteFunding Information:
The authors thank Soumen Ghosh and Chad Hoyer for help with the excitation energy database. P.V. acknowledges a Richard D. Amelar and Arthur S. Lodge Fellowship. This work was supported in part by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under award number DE-SC0015997.
© 2017 American Chemical Society.