Accurate helium-benzene potential: From CCSD(T) to Gaussian process regression.

Area of Science:

• Physical Chemistry
• Computational Chemistry
• Materials Science

Background:

• Accurate modeling of non-covalent interactions between helium and graphitic materials is crucial for understanding quantum phenomena.
• The helium-benzene complex is a fundamental prototype, but quantitatively reliable potential energy surfaces (PES) are computationally challenging.
• Existing models often lack the accuracy needed for advanced simulations.

Purpose of the Study:

• To develop a highly accurate, continuous, three-dimensional potential energy surface (PES) for the He-benzene system.
• To benchmark high-level computational methods for describing weak interactions in reduced dimensions.
• To investigate the solvation behavior of helium on benzene using advanced simulation techniques.

Main Methods:

• High-level coupled-cluster (CCSD(T)) calculations extrapolated to the complete basis set limit for benchmark energies.
• Symmetry-adapted perturbation theory to decompose interaction components (dispersion, exchange-repulsion).
• Multifidelity Gaussian process regression combining density functional theory and coupled-cluster data to construct the PES.
• Path integral Monte Carlo (PIMC) simulations for low-temperature helium solvation studies.

Main Results:

• Established benchmark interaction energies for He-benzene using rigorous quantum chemical methods.
• Developed a continuous, 3D PES with sub-cm-1 accuracy, validated against physical laws.
• Symmetry-adapted perturbation theory confirmed the dominance of dispersion and exchange-repulsion forces.
• PIMC simulations revealed distinct solvation behavior compared to Lennard-Jones potentials, especially in adsorption layer filling.

Conclusions:

• The developed PES provides a significant advancement for accurate simulations of helium interacting with graphitic systems.
• This work offers a benchmark for future studies on larger polycyclic aromatic hydrocarbons and graphene.
• The findings highlight the limitations of empirical potentials and the necessity of accurate ab initio methods for quantum fluid-surface interactions.