Computational Implementation of Spectral Theory

The project objective is to implement, validate, and optimize an integrated computer software module suite for calculating molecular adiabatic electronic energies and quantum chemical wave functions using spectral theory methods. To this end, a computational kernel is scripted for high-performance calculations of the electronic structures of chemicals & materials made from H, B, C, N, O, and F atoms. The computational kernel comprises four integrated component modules which perform the following series of individual calculations:(i) phase-consistent degenerate atomic eigenstates employing large-scale configuration-interaction methods in both real Cartesian-Gaussian and Slater-orbital representations;(ii) diatomic wave functions and Hamiltonian \& metric matrices constructed in the antisymmetrized outer-product representation of phase-consistent atomic eigenstates; (iii) aggregate molecular metric matrices in the ordered atomic spectral-product representation required for enforcing overall electron antisymmetry; and (iv)aggregate molecular spectral-product Hamiltonian matrices assembled from the atomic, diatomic, and metric information, employing Wigner rotation matrices for all angular atomic coordinate dependencies. Representative electronic structure calculations are performed of H, B, C, N, O, and F atom-containing compounds employing Lanczos diagonalization methods and on-the-fly evaluations of molecular Hamiltonian matrix elements. The computational kernel software, documentation of the module suite operation, and a report of benchmarking results constitute the deliverable items from the project. BENEFIT: The computational kernel devised provides the basis for construction and operation of a diverse selection of applications software modules designed for specific chemical purposes, including but not limited to determinations of structural extremal-energy configurations, associated minimum energy paths connecting such points, molecular electronic expectation values, including charge and spin densities, dipole and transition moments, and vibrational force constants, to mention some representative examples, on-the-fly classical and quantum Monte-Carlo simulations of molecules, clusters, and condensed matter, and the free-energies of molecular association when integrated into appropriate statistical-mechanical code suites. Target problems addressed by such applications software modules based on the computational kernel include the rational design and synthesis of new chemicals and materials, fabrication of nanoscale electronic and optical devices, and development of physiologically relevant macromolecular structures in the presence of solvent for small-molecule drug design and pharmacological purposes. A potentially significant commercial market for this technology, to be developed under Phase III continuation projects, is provided by Air Force and other DOD, DOE and NASA technical needs; the requirements of new compound development in the chemical industry; and the large NIH related high-technology biomedical and pharmaceutical research communities focused on development of therapeutic drugs and in vivo nanoscale bioimaging devices.