Configurational-Bias Monte Carlo Back-Mapping Algorithm for Efficient and Rapid Conversion of Coarse-Grained Water...

Area of Science:

• Computational chemistry
• Materials science
• Statistical mechanics

Background:

• Coarse-grained molecular dynamics (MD) simulations enable longer time and length scale studies but lose atomistic detail.
• Reverse transformation (back mapping) from coarse-grained to atomistic models is challenging, especially for systems with high diffusion barriers.
• Existing back-mapping methods often require extensive simulations or are limited to specific systems.

Purpose of the Study:

• To introduce a novel, efficient algorithm for the reverse transformation of coarse-grained models to atomistic representations.
• To demonstrate the generic applicability of the new algorithm, using a coarse-grained water model as an example.
• To compare the performance of the new algorithm against existing molecular dynamics (MD) and Monte Carlo (MC) based back-mapping techniques.

Main Methods:

• Development of the crystalline-configurational-bias Monte Carlo (C-CBMC) algorithm, extending the configurational-bias Monte Carlo (CBMC) method.
• Application of C-CBMC to back-map coarse-grained water models, including ice-liquid water interfaces and various ice polymorphs (hexagonal, cubic, stacking disorder).
• Comparative simulations using TIP4P/Ice model to evaluate C-CBMC against standard MD and MC back-mapping techniques.

Main Results:

• The C-CBMC algorithm rapidly finds optimal hydrogen-bonded configurations, significantly outperforming other methods in terms of simulation steps.
• C-CBMC yields lower energy structures for crystalline ice (0.05-0.1 eV/water molecule lower) compared to existing methods.
• Back-mapped configurations using C-CBMC show significantly improved global hydrogen positioning and reduced errors in radial distribution functions (RDFs) compared to MD and MC methods.

Conclusions:

• The C-CBMC algorithm is an efficient and effective approach for the reverse transformation of coarse-grained models to atomistic detail.
• This method shows particular efficacy for crystalline systems, overcoming limitations of standard force-field-based relaxation methods that can get trapped in local minima.
• The improved accuracy in structural and energetic properties demonstrates the value of C-CBMC for complex molecular simulations.