Article ID Journal Published Year Pages File Type
10997994 Journal of Computational Science 2016 36 Pages PDF
Abstract
We have computed the most positive and most negative values of the electrostatic potentials on the surfaces of a series of molecules, some of which contain atoms from the second and third rows of Groups IV, VI and VII. Three different methods were used - Hartree-Fock and the density functional B3LYP and M06-2X - in conjunction with numerous basis sets of various sizes. Our objectives were (a) to assess the degrees to which the different procedures gave comparable results, and (b) to ascertain which procedures best combine reliability with computational efficiency. Our primary conclusions were: (a) For a given basis set, the three methods give essentially comparable results. (b) Hydrogens do not require either polarization or diffuse functions, not even when the potential of interest is on the hydrogen. (c) Polarization functions are essential, for first-row as well as higher-row atoms; diffuse functions are less important. (d) The recommended basis set for electrostatic potential calculations, of those investigated, is the 6-31G*. Larger basis sets are not needed for this purpose. (e) If the 6-31G* is precluded (e.g. by the size of the system or by the presence of fourth-row atoms), then the 3-21G* is likely to be an acceptable alternative.
Related Topics
Physical Sciences and Engineering Computer Science Computational Theory and Mathematics
Authors
, , , , ,