In this paper, we present a localised molecular orbital (LMO) based methodology of the TDHF excited state calculation in which the computing time is linearly scaled by the size of the system. In the benchmark calculations using simple poly-acetylene molecules, the LMO-driven TDHF provides significant shorter CPU time than conventional TDHF calculations and near O(N) computing cost. We also present the techniques accelerating the convergence speed in Davidson iterative diagonalisation and show its usefulness. The methodology accurately reproduces the excited state properties of poly-acetylene molecules obtained in conventional TDHF calculations.
All Science Journal Classification (ASJC) codes
- Molecular Biology
- Condensed Matter Physics
- Physical and Theoretical Chemistry