Promising insights in parallel grid-based algorithms for quantum chemical topology.

Abstract

Some straightforward improvements designed to make grid-based quantum chemical topology more efficient and faster are presented. The strategy focuses on both the evaluation of the scalar function over three-dimensional discrete grids and the algorithms aimed to follow and integrate gradient trajectories over the basin volumes together. Beyond the density analysis, we show that the scheme is quite suitable for the electron localization function and its complex topology. With this speed-up of the parallelized process used to generate 3d-grids, this new scheme is several orders of magnitude faster than the original grid-based method implemented in our laboratory (TopMod09). The efficiency of our implementation (TopChem2) was also compared with well-known grid-based algorithms designed to assign the grid points to basins. The performances, speed versus accuracy have been discussed on the basis of results obtained from selected illustrative examples.