Harris-type van der Waals density functional scheme
Journal article, 2013

Biomolecular systems that involve thousands of atoms are difficult to address with standard density functional theory (DFT) calculations. With the development of sparse-matter methods such as the van der Waals density functional (vdW-DF) method [M. Dion et al., Phys. Rev. Lett. 92, 246401 (2004)], it is now possible to include the dispersive forces in DFT which are necessary to describe the cohesion and behavior of these systems. vdW-DF implementations can be as efficient as those for traditional DFT. Yet, the computational costs of self-consistently determining the electron wave functions and hence the kinetic-energy repulsion still limit the scope of sparse-matter DFT. We propose to speed up sparse-matter calculations by using the Harris scheme [J. Harris, Phys. Rev. B 31, 1770 (1985)]; that is, we propose to perform electronic relaxations only for separated fragments (molecules) and use a superposition of fragment densities as a starting point to obtain the total energy non-self-consistently. We evaluate the feasibility of this approach for an adaption of the Harris scheme for non-self-consistent vdW-DF (sfd-vdW-DF). We study four molecular dimers with varying degrees of polarity and find that the sfd scheme accurately reproduces standard non-self-consistent vdW-DF for van der Waals dominated systems but is less accurate for those dominated by polar interactions. Results for the S22 set of typical organic molecular dimers are promising.

Author

Kristian Berland

Chalmers, Applied Physics, Electronics Material and Systems Laboratory

Elisa Londero

Chalmers, Microtechnology and Nanoscience (MC2), Quantum Device Physics

Elsebeth Schröder

Chalmers, Microtechnology and Nanoscience (MC2), Quantum Device Physics

Per Hyldgaard

Chalmers, Applied Physics, Electronics Material and Systems Laboratory

Physical Review B - Condensed Matter and Materials Physics

1098-0121 (ISSN)

Vol. 88 4

Areas of Advance

Nanoscience and Nanotechnology (2010-2017)

Materials Science

Roots

Basic sciences

Infrastructure

C3SE (Chalmers Centre for Computational Science and Engineering)

Subject Categories

Condensed Matter Physics

DOI

10.1103/PhysRevB.88.045431

More information

Created

10/7/2017