Computational efficiency for the simulation of bulk crystals and surfaces is highly desirable. In an effort to study semiconductor crystals, we present a self-consistent treatment for the simulation of silicon crystals and surfaces based on the combination of a siligen model and a semiempirical Hamiltonian method. An artificial atom called siligen is introduced for the application of the semiempirical method to finite-size silicon clusters. The calculated average bond energies for the saturated silicon clusters are between 2.045 and 2.568 eV, compared to the measured value of 2.31 eV. A simulated bulk silicon surface using siligens is introduced in order to examine variation of the bond strength between fluorine atoms and the simulated silicon (111)surface. It is found that bond strength computed from the simulated surface, with siligens, rapidly converges to a saturated limit as the number of surface layers increases, while a pure silicon (111)surface without siligens yields no satisfactory convergence.

No abstract

No abstract

No abstract

No abstract