Creating Optimal Coordinates for Drastic Cost Reductions in Rigorous Ab Initio Quantum Chemistry Computations

The accurate determination of molecular vibrations and zero-point vibrational energies (ZPVEs) has been a crucial facet of quantum chemistry for decades. As system size increases, computing these molecular properties rapidly approaches computational intractability for ab initio methods such as CCSD(T). The Concordant Mode Approach (CMA) was recently established as a highly accurate protocol for computing vibrational frequencies at a fraction of the cost.  While the accuracy of the CMA methodology is quite impressive, there is no systematic method for including off-diagonal force constants which eliminate CMA frequency errors. We present a groundbreaking a priori method, CMA-2A, for selecting off-diagonal force constants with varying levels of robustness.

The force constant matrix sparsity can be further increased by utilizing molecular symmetry. While almost all contemporary electronic structure packages employ symmetry techniques, they rarely have the sophistication to transform basis functions into sets of symmetry adapted linear combinations (SALCs) for nonabelian point groups. Thus, the computational cost reductions are often restricted to creating SALCs for the highest abelian subgroup of the molecule. Herein, we present the Python package MolSym, which can treat nonabelian molecular point groups. Various features of the program, such as the creation of SALCs for various chemically relevant coordinate systems, will be discussed.