Using an Iterative Eigensolver and Intertwined Rank Reduction to Compute Vibrational Spectra of Molecules With More Than a Dozen Atoms: Uracil and Naphthalene

Abstract

We use a direct product basis, basis vectors computed by evaluating matrix-vector products, and rank reduction to calculate vibrational energy levels of uracil and naphthalene, with 12 and 18 atoms, respectively. A matrix representing the Hamiltonian in the direct product basis and vectors with as many components as there are direct product basis functions are neither calculated nor stored. We also introduce an improvement of the Hierarchical Intertwined Reduced-Rank Block Power Method (HI-RRBPM), proposed previously in Thomas and Carrington, Jr. [J. Chem. Phys. 146, 204110 (2017)]. It decreases the memory cost of the HI-RRBPM and enables one to compute vibrational spectra of molecules with over a dozen atoms with a typical desktop computer.