Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The Enduring Relevance of Semiempirical Quantum Mechanics

Jonathan E. Moussa

TL;DR

Addressing the persistent cost-accuracy gap in atomistic modeling, the paper argues for the enduring relevance of semiempirical quantum mechanics (SQM) as a midrange paradigm between ab initio QM and MM. It advocates a modular view of SQM, unifying QM and SQM through Hamiltonian forms, and integrating ML within SQM alongside tighter QM-SQM software coupling. Key contributions include a formal deconstruction of SQM into Hamiltonian, total-energy, and solver components; the delineation of two-center Slater–Koster and four-center Coulomb matrix-element formalisms; and discussions of NOTCH, ML-corrected SQM, and hydrogen-cluster benchmarks to illustrate error budgeting. The work emphasizes task distributions, data-driven benchmarking, and standardized interfaces to foster a more integrated QM–SQM ecosystem, with potential midrange commercialization as a practical outcome.

Abstract

The development of semiempirical models to simplify quantum mechanical descriptions of atomistic systems is a practice that started soon after the discovery of quantum mechanics and continues to the present day. There are now many methods for atomistic simulation with many software implementations and many users, on a scale large enough to be considered as a software market. Semiempirical models occupied a large share of this market in its early days, but the research activity in atomistic simulation has steadily polarized over the last three decades towards general-purpose but expensive ab initio quantum mechanics methods and fast but special-purpose molecular mechanics methods. I offer perspective on recent trends in atomistic simulation from the middle ground of semiempirical modeling, to learn from its past success and consider its possible paths to future growth. In particular, there is a lot of ongoing research activity in combining semiempirical quantum mechanics with machine learning models and some unrealized possibilities of tighter integration between ab initio and semiempirical quantum mechanics with more flexible theoretical frameworks and more modular software components.

The Enduring Relevance of Semiempirical Quantum Mechanics

TL;DR

Addressing the persistent cost-accuracy gap in atomistic modeling, the paper argues for the enduring relevance of semiempirical quantum mechanics (SQM) as a midrange paradigm between ab initio QM and MM. It advocates a modular view of SQM, unifying QM and SQM through Hamiltonian forms, and integrating ML within SQM alongside tighter QM-SQM software coupling. Key contributions include a formal deconstruction of SQM into Hamiltonian, total-energy, and solver components; the delineation of two-center Slater–Koster and four-center Coulomb matrix-element formalisms; and discussions of NOTCH, ML-corrected SQM, and hydrogen-cluster benchmarks to illustrate error budgeting. The work emphasizes task distributions, data-driven benchmarking, and standardized interfaces to foster a more integrated QM–SQM ecosystem, with potential midrange commercialization as a practical outcome.

Abstract

The development of semiempirical models to simplify quantum mechanical descriptions of atomistic systems is a practice that started soon after the discovery of quantum mechanics and continues to the present day. There are now many methods for atomistic simulation with many software implementations and many users, on a scale large enough to be considered as a software market. Semiempirical models occupied a large share of this market in its early days, but the research activity in atomistic simulation has steadily polarized over the last three decades towards general-purpose but expensive ab initio quantum mechanics methods and fast but special-purpose molecular mechanics methods. I offer perspective on recent trends in atomistic simulation from the middle ground of semiempirical modeling, to learn from its past success and consider its possible paths to future growth. In particular, there is a lot of ongoing research activity in combining semiempirical quantum mechanics with machine learning models and some unrealized possibilities of tighter integration between ab initio and semiempirical quantum mechanics with more flexible theoretical frameworks and more modular software components.
Paper Structure (12 sections, 8 equations, 3 figures, 1 table)

This paper contains 12 sections, 8 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Models from components
  3. SQM model deconstruction
  4. Statistical model selection
  5. Hamiltonian model forms
  6. Two-center matrix elements
  7. Four-center matrix elements
  8. Ongoing work and challenges
  9. Role of machine learning
  10. Hydrogen cluster example
  11. Conclusions
  12. Biography

Figures (3)

  • Figure 1: Number of citations to the 20 most cited atomistic simulation engines, as estimated using Google Scholar. Color denotes the software's primary simulation type: QM (red), SQM (green), or MM (blue). The history of aggregate annual citations is plotted on a logarithmic scale with its components ordered by overall citation count.
  • Figure 2: Number of citations to the 3 most cited SQM software programs, as estimated using Google Scholar. Note that the DFTB+ estimate effectively includes an estimate of the DFTB software that preceded it prior to 2007.
  • Figure 3: Optimized Slater--Koster model parameters (a,b) and RMS energy errors relative to FCI/def2-QZVPP reference data (c,d) for the two tasks of calculating the ground-state energy of each distinct charge and spin sector (a,c) and calculating all distinct stationary-state energies in the single-occupancy sector (b,d) of H$_2$ and symmetric H$_3$.