Abstract

Among wavefunction based methods, quantum diffusion Monte Carlo (DMC) and coupled cluster with single, double, and perturbative triple excitations [ CCSD ] are state-of-the-art approaches that have been shown to yield accurate interaction energies for small organic molecules. These methods provide valuable reference information for widely-used semi-empirical and machine learning potentials, especially where experimental information is scarce. As we look to compute larger molecules and materials, many recent developments have been made in coupled cluster and DMC methods that allow larger systems to be computed efficiently. In this work, we look for agreement between CCSD and DMC in larger molecules and find excellent agreement in most cases and some surprisingly large disagreements in a few. We disentangle some of the key approximations and sources of error and consider the way towards consistent predictions of non-covalent interactions between extended molecules.