scholarly journals Straightforward and Accurate Automatic Auxiliary Basis Set Generation for Molecular Calculations with Atomic Orbital Basis Sets

Author(s):  
Susi Lehtola
1996 ◽  
Vol 8 (21) ◽  
pp. 3859-3880 ◽  
Author(s):  
Daniel Sánchez-Portal ◽  
Emilio Artacho ◽  
José M Soler

1992 ◽  
Vol 70 (2) ◽  
pp. 612-630 ◽  
Author(s):  
Walter J. Stevens ◽  
Morris Krauss ◽  
Harold Basch ◽  
Paul G. Jasien

Relativistic compact effective potentials (RCEP), which replace the atomic core electrons in molecular calculations, have been derived from numerical Dirac–Fock atomic wavefunctions using shape-consistent valence pseudo-orbitals and an optimizing procedure based on an energy-overlap functional. Potentials are presented for the third-, fourth-, and fifth-row atoms of the Periodic Table (excluding the lanthanide series). The efficiency of molecular calculations is enhanced by using compact Gaussian expansions (no more than three terms) to represent the radial components of the potentials, and energy-optimized, shared-exponent, contracted-Gaussian atomic orbital basis sets. Transferability of the potentials has been tested by comparing calculated atomic excitation energies and ionization potentials with values obtained from numerical relativistic Hartree–Fock calculations. For the alkali and alkaline earth atoms, core polarization potentials (CPP) have been derived which may be added to the RCEP to make possible accurate molecular calculations without explicitly including core-valence correlating configurations in the wavefunction. Keywords: model potentials, effective core potentials, transition metals, relativistic calculations.


Author(s):  
J.-P. Blaudeau ◽  
S. R. Brozell ◽  
S. Matsika ◽  
Z. Zhang ◽  
R. M. Pitzer

1992 ◽  
Vol 70 (2) ◽  
pp. 580-588 ◽  
Author(s):  
R. Custodio ◽  
J. D. Goddard ◽  
M. Giordan ◽  
N. H. Morgon

The Simplex method was used to define atomic and universal meshes using the integral discretization technique for the Griffin–Hill–Wheeler-Hartree–Fock equations. This technique represents a basis set as an exponential set of the form:[Formula: see text]For atoms, the minimum total energy criterion was employed. For the universal basis, three different procedures were tested: (a) defining the universal basis using information on the isolated atoms, (b) determining the universal Ω0(k) through atomic calculations and reoptimizing the ΔΩ(k) for different symmetries employing simultaneously a single atomic calculation as a reference point, and (c) optimizing the universal mesh using a statistical criterion such as the squares of the deviations of the total energy. The meshes obtained by the minimum total energy criterion or the squares of deviations of the total energy for the universal basis are accurate for the total energy but the weight functions are deficient in the valence region. Shifting the optimized Ω0(k) to [Formula: see text], fixing [Formula: see text], and reoptimizing ΔΩ(k) for each symmetry species produces a better description of weight functions at the expense of a less accurate total energy. In general, no significant statistical difference was observed for the various universal bases generated by procedures (a) and (b) or by (c) provided the shift correction was made to the latter. Application of these bases to diatomic molecules (N2, CO, P2, CS) showed that the universal bases are as accurate as those optimized for atomic systems. If the bases are transferred from atoms to molecules, the shift corrections to the weight functions of the atoms are not useful in molecular calculations. The almost equivalent molecular properties and the good total energies show that the best basis for molecular calculations is that optimized by procedure (c). Keywords: universal basis sets, integral discretization technique.


Sign in / Sign up

Export Citation Format

Share Document