Universal Multireference Diagnostics for High-Throughput Screening of Large Molecules and Solids

Objective

Computational chemistry is one of the most significant branches of modern chemical research. The favourable efficiency/cost ratio of density functional theory (DFT) has extended its applications to the prediction of molecular properties such as spectroscopic constants, nonlinear optical properties, and reaction mechanisms to drug discovery in pharmaceutical industries and materials design in supramolecular industries. Its importance has only increased with the resurgence of machine learning and the advent of artificial intelligence. Despite the widespread use of DFT, current density functional approximations can capture well the so-called nondynamic electron correlation, severely compromising the simulations whenever these multireference (MR) effects comes into play. Hence, the development of correlation diagnostics capable of identifying MR systems is critical for computational chemistry. Such diagnostics are valuable not only for high-throughput screening but also for individual molecular studies, where accurate simulation depends on the selection of an appropriate computational method. This need is even more evident in periodic systems, where technologically critical properties –such as magnetic behaviour, optical responses, and metal–insulator transitions– depend sensitively on its accurate treatment. UniMRD aims to develop universal MR diagnostics capable of predicting MR character across a broad range of molecules and materials, while requiring minimal computational resources and ensuring both reliability and wide applicability.