| Co-Investigator(Kenkyū-buntansha) |
TAKETSUGU Tetsuya Ochanomizu University Faculty of Science, Associate Professor, 理学部, 助教授 (90280932)
NAKAJIMA Takahito The University of Tokyo, Graduate School of Engineering, Research Associate, 大学院・工学系研究科, 助手 (10312993)
NAKANO Haruyuki The University of Tokyo, Graduate School of Engineering, Lecturer, 大学院・工学系研究科, 講師 (90251363)
|
|---|
| Research Abstract |
The aim of this research project is to develop an ab initio molecular orbital program that can determine the electronic structure of large-scale systems, including more than 1000 atoms, with high accuracy and high performance. To achieve this aim, both pure theoretical and practical aspects are necessary : development of new methods in theory and new technology in algorithm and program. In the former aspect, our focus was particularly on the development of (1) new ab initio molecular orbital theory, especially multireference-based methods, (2) molecular orbital theory including relativistic effects, and (3) exchange and correlation functional in density functional theory ; and we succeeded in the development of MRMP with localized orbitals, MRMP with CASCI reference, quasi-complete active space SCF method, QCAS-SCF reference QDPT, RESC, third-order Douglas-Kroll method, highly efficient scheme of Dirac-Hatree-Fock and Dirac-Kohn-Sham methods, relativistic model potentials, parameter-free exchange functional, one-parameter-progressive correlation functional. In the latter aspect, we developed a molecular integral program, Spherica, based on high-performance integral method, and the programs of Dirac-Hatree-Fock and Dirac-Kohn-Sham methods based on the above-mentioned schemes. These are not programs simply coded according to conventional computational techniques, but programs obtained by the use of new technologies found after optimization of every computational scheme including molecular integrals.
|
