Search

The canonical molecular orbitals (CMO) calculation provides useful knowledge such as molecular orbitals, energy level, charge density and so on. As one of the properties obtained from the CMO calculation of protein results, the interaction energy analysis between amino acid residues and/or substrates is useful as molecular design and functional analysis. The energy decomposition analysis approaches are the well-known calculation methods for interaction energy analysis, however, the electronic structure calculation of large molecule is very expensive and hard to achieve. Therefore, we have implemented the interaction analysis method by using the energy density analysis (EDA) method. The calculation for the EDA method is as simple as the Mulliken population method but requires matrix elements for energy calculation. By adopting the grid-free method for the exchange-correlation term, we could apply the EDA method in the DFT scheme.

Although canonical molecular orbital (CMO) calculations for proteins are very useful for protein engineering, they are computationally expensive and difficult to achieve because proteins are large molecules. We had developed the third-generation density functional (3G DF) method, which performed electronic state calculation with high efficiency in parallel computer by performing only matrix operation during the SCF calculation. In this study, by applying the powerful computing platform of graphic processing unit (GPU) to our 3G DF calculation method, we constructed a computing environment that achieved the large-scale CMO computation with high efficiency.Although canonical molecular orbital (CMO) calculations for proteins are very useful for protein engineering, they are computationally expensive and difficult to achieve because proteins are large molecules. We had developed the third-generation density functional (3G DF) method, which performed electronic state calculation with high efficiency in parallel computer by performing only matrix operation during the SCF calculation. In this study, by applying the powerful computing platform of graphic processing unit (GPU) to our 3G DF calculation method, we constructed a computing environment that achieved the large-scale CMO computation with high efficiency.

Despite its importance, carrier conduction in electrical insulators is poorly understood. This work presents a computational study of hole conduction in single crystalline alkanes (n-C18H38 and n-C36H74). Hole mobilities are computed with the combination of molecular dynamics simulation, quantum chemical calculation, and the kinetic Monte Carlo method. The hole hopping rate is computed by the Fermi golden rule rate kernel without high temperature approximation. A strong correlation between the anisotropy of hole mobility and crystalline morphology is found. Hole mobilities in the direction of the c axis are more than an order of magnitude larger than those in the a − b plane. At room temperature, hole mobility is increased by roughly a factor of 10 due to the thermal motion of molecules. Computed anisotropic hole mobilities are in reasonable agreement with experimental values when the effect of dynamic disorder is taken into account. The results strongly indicate that hole transfer in crystalline alkane occurs in the phonon-assisted transport regime.