Local MP2 with Density Fitting for Periodic Systems: A Parallel Implementation.

Area of Science:

• Computational Quantum Chemistry
• Solid-State Physics
• Materials Science

Background:

• Accurate electronic structure calculations are crucial for understanding crystalline materials.
• Local second-order Møller-Plesset perturbation theory (LMP2) offers a computationally feasible approach for correlated electron energies.
• Parallel implementations are necessary to handle the computational demands of large periodic systems.

Purpose of the Study:

• To develop and present a parallel implementation for LMP2 energy evaluation in periodic, nonconducting crystalline systems.
• To address the unique challenges of parallelizing LMP2 for periodic systems, including symmetry considerations.
• To demonstrate the efficiency and scalability of the developed implementation on relevant solid-state quantum chemistry problems.

Main Methods:

• Implementation of LMP2 energy calculations utilizing a density-fitting approximation for two-electron repulsion integrals.
• Development of parallel algorithms tailored for periodic systems, incorporating translational and point symmetry.
• Benchmarking the implementation on representative crystalline systems, including a large metal-organic framework (MOF-5).

Main Results:

• Demonstrated good parallel efficiency for the LMP2 implementation on up to 54 processors.
• Successfully performed large-scale calculations on a 106-atom MOF-5 structure, representing a significant computational achievement.
• Validated the approach for challenging solid-state quantum chemistry problems where MP2 methods are critical.

Conclusions:

• The presented parallel LMP2 implementation is efficient and scalable for periodic crystalline systems.
• The method provides a valuable tool for investigating complex solid-state materials.
• The successful application to large systems like MOF-5 highlights the potential of this approach in materials discovery and design.