Estudo sobre a aplicação do método Monte Carlo Simulated Annealing ao processo de geração de Bases Gaussianas Atômicas.

Título da Revista

ISSN da Revista

Título de Volume

Editor

Universidade Federal do Espírito Santo

Resumo

The Monte Carlo Simulated Annealing Method was used to generate adapted Gaussian basis sets to the Li, C, O, Na, Si, S, K, Fe, Ge, and Se atoms in the ground state. This atom set was chosen to represent, on average, atoms of the 2? , 3? and 4? periods of the periodic table, and a study on cooling schedules was performed. The Improved Generator Coordinate Hartree-Fock method was initially used to generate adapted basis sets which were used as a starting point for generating new Gaussian basis sets. Subsequently, the Monte Carlo Simulated Annealing Method with variational principle, applied to the Hartree-Fock wave function, were used to obtain the fully optimized Gaussian basis sets. The same procedure was realized using adaptive and non-adaptive cooling schedules, and all atomic basis were generated 20 times. A reliability analysis of the results and comparisons between the studied cooling schedules and results produced by other procedures of basis generation was performed. We found that the adaptive cooling schedule produced more accurate results in the calculation of total HF energy for the six atoms in our sample with greater atomic number. The non-adaptive schedule produced more accurate results in the calculations of total HF energy for the rst four atoms in our sample. Our results has showed that both cooling schedules has approximately the same precision in the calculation of the atomic total HF energy and the non-adaptive schedule is, on average, more e cient. More detailed researches are necessary in order to verify if the method which we propose at this work is appropriate to generate atomic basis able to be used in molecular calculations.

Descrição

Palavras-chave

Monte Carlo, Bases Gaussianas, Física Atômica e Molecular

Citação

REIS, Thiago Mello dos. Estudo sobre a aplicação do método Monte Carlo simulated annealing ao processo de geração de bases gaussianas atômicas. 2012. 82 f. Dissertação (Mestrado em Física) – Programa de Pós-Graduação em Física, Universidade Federal do Espírito Santo, Centro de Ciências Exatas, Vitória, 2012.

Avaliação

Revisão

Suplementado Por

Referenciado Por