Multireference diffusion Monte Carlo reaches 2D materials

Area of Science:

• Quantum mechanics
• Condensed matter physics
• Computational chemistry

Background:

• Quantum confinement in 2D materials enhances electronic correlation effects.
• Accurate and efficient prediction of 2D material properties is a significant theoretical challenge.
• Density Functional Theory (DFT) approximations limit prediction reliability due to exchange-correlation functional dependence.

Purpose of the Study:

• To estimate the impact of correlation on the total energy of graphene.
• To compare the performance of Self-Healing Diffusion Monte Carlo (SHDMC) with selected CI (sCI) and quantum Monte Carlo methods.
• To validate SHDMC for challenging 2D materials.

Main Methods:

• State-of-the-art selected CI calculations.
• Quantum Monte Carlo extrapolated calculations.
• Self-Healing Diffusion Monte Carlo (SHDMC) for wavefunction generation and energy estimation at the Γ point for a graphene unit cell.

Main Results:

• SHDMC yields a compact, high-quality wavefunction for graphene, unlike basis set dependent quantum chemistry methods.
• The SHDMC wavefunction is superior to sCI in the same basis and significantly smaller (approx. 1000x fewer determinants).
• Extrapolated SHDMC results align well with complete basis set extrapolated sCI, validating its accuracy.

Conclusions:

• SHDMC provides a high-quality, computationally efficient alternative for electronic structure calculations in 2D materials.
• The method shows reduced basis set dependence compared to traditional quantum chemistry approaches.
• This work establishes SHDMC as a promising tool for future studies of complex 2D materials.