Achieving linear scaling in computational cost for a fully polarizable MM/continuum embedding

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Quantum Mechanics/Molecular Mechanics

Background:

• Accurate molecular simulations require sophisticated models combining quantum mechanics (QM) and molecular mechanics (MM).
• Polarizable models are crucial for describing electronic response in chemical systems.
• Continuum solvation models are essential for simulating condensed-phase environments.

Purpose of the Study:

• To present a novel, efficient implementation of a fully polarizable QM/MM/continuum model.
• To achieve linear scaling in computational cost and memory requirements for QM/MM/continuum simulations.
• To enable the simulation of large molecular systems with high accuracy.

Main Methods:

• Induced-dipoles polarizable force field.
• Conductor-like Screening Model (ddCOSMO) as a polarizable continuum.
• Self-consistent field (SCF) QM method.
• Fast Multipole Method (FMM) for the force field.
• Efficient iterative procedure for MM/continuum embedding.

Main Results:

• Achieved linear scaling of computational cost and memory requirements.
• Demonstrated full coupling of MM and continuum polarization.
• Enabled inclusion of tens of thousands of atoms in the classical layer with limited computational effort.
• Overcame limitations on molecular cavity shape.

Conclusions:

• The new implementation offers unprecedented efficiency for polarizable QM/MM/continuum simulations.
• This method significantly reduces computational barriers for studying large, complex molecular systems.
• The approach facilitates more accurate and feasible modeling of condensed-phase phenomena.