摘要：

Some aspects of the current state of Quantum Chemistry are presented, along with a selection of results for large molecules using modest resources. Following a brief exposition of the fundamental approximations of ab initio calculations, the SCF, DFT and MP2 methods for calculating molecular wavefunctions and energies are introduced. The Resolution of Identity (RI) approach is described with a view to the treatment of large molecules (100 atoms and more). Scaling behavior of the various computational methods is compared; here, aluminium clusters represent a demanding application. The utility of analytical energy gradients is shown, in particular using the example of a theoretical study of the structural isomers of sulfur-bridged copper clusters. A strategy to compute electronic excitation energies for large molecules is sketched and applications to fullerenes and cadmium-selenide nanoclusters are presented. Developments in scientific computing hardware are considered, with emphasis on the emergence of PC's. The opportunities and difficulties inherent in the parallelization of quantum chemical code are also discussed, and the performance of parallel TURBOMOLE is presented.

展开