# An Extended Energy-Biased Aggregation-Volume-Bias Monte Carlo (EB-AVBMC) Method for Nucleation Simulation of a Reactive Water Potential

## TL;DR

This paper introduces an improved Monte Carlo method for simulating water nucleation using a reactive water model, enabling accurate calculation of liquid-vapor properties.

## Contribution

The paper extends the EB-AVBMC method to reactive water models with revised acceptance rules and constraints for bond topology.

## Key findings

- The modified EB-AVBMC method accurately calculates nucleation free energies and surface tension for water at 298.15 K.
- The method prevents biased sampling by considering intramolecular energy and avoiding deletion of dissociated water molecules.
- Results are consistent with those from rigid water models, showing the method's reliability and generalizability.

## Abstract

The aggregation-volume-bias
Monte Carlo (AVBMC) algorithm has been
widely used with empirical water models like TIP3P, SPC/E, TIP4P,
and TIP4P/2005 to study nucleation and vapor–liquid
properties, but its application to reactive water models remains underexplored.
Here, we present an extension of the energy-bias aggregation-volume-bias
Monte Carlo (EB-AVBMC) method for calculating nucleation free energies
and liquid–vapor properties, such as gas density and surface
tension, using a three-body reactive force field based on the Vashishta
potential functional form [Phys. Rev. B
1990, 41, 12197–12209]. Key modifications include
revised acceptance rules that consider the intramolecular energy of
the inserted/deleted molecule to prevent high acceptance probabilities
that could bias the sampling and constraints to avoid the deletion
of dissociated water molecules. These adjustments ensure valid bond
topology modifications. We demonstrate the method’s applicability
by studying water nucleation at 298.15 K, with varying cluster sizes,
and showing a free energy consistent with studies from rigid water
models. This approach is generalizable to other reactive water force
fields, offering a valuable tool for simulating reactive liquid–vapor
properties.