# Introducing the Coupled-Cluster Theory to the Amorphous World of Liquids and Their Thermodynamic Simulations

## TL;DR

A new simulation method called FrAMonC enables accurate thermodynamic predictions for amorphous molecular materials using quantum-chemical calculations.

## Contribution

FrAMonC introduces fragment-based ab initio Monte Carlo simulations for amorphous liquids, enabling high-accuracy predictions with minimal empirical inputs.

## Key findings

- FrAMonC achieves superior accuracy in predicting liquid-phase densities and thermal expansivities.
- The method enables the use of coupled-cluster theory for thermodynamic simulations of amorphous materials.
- Gas–liquid heat capacity differences are accurately modeled using FrAMonC compared to existing models.

## Abstract

Amorphous molecular materials are ubiquitous, spanning
drugs, semiconductors,
or solvents. Large predictive capabilities of quantum-chemical simulations
of structural and thermodynamic properties and phase transitions for
such amorphous materials have remained out of reach for a long time
due to the related immense computational costs. This work introduces
a novel fragment-based ab initio Monte Carlo (FrAMonC) simulation
technique to the amorphous realm of molecular liquids and glasses.
It aims at enabling thermodynamic simulations for amorphous molecular
materials based on direct ab initio sampling and at minimizing the
amount of a priori required empirical inputs for such simulations.
Focus on individual cohesive interactions within the bulk, and their
sampling from multiple first-principles potentials with a many-body
expansion scheme enables the use of very accurate electron-structure
methods for the most important cohesive features within the material.
Even the coupled-cluster theory, the direct use of which is unprecedented
for molecular simulations of thermodynamic properties for liquids,
then becomes applicable to the description of bulk amorphous materials.
Its incorporation in the proposed Monte Carlo simulations promises
very high computational accuracy. Bulk-phase equilibrium properties
at finite temperatures and pressures, such as density and vaporization
enthalpy, as well as response properties such as thermal expansivity
and heat capacity that are particularly challenging to predict accurately,
are the observables targeted in this work. Superior computational
accuracy of the introduced FrAMonC simulations is demonstrated for
most target properties (liquid-phase densities, thermal expansivities,
and gas–liquid differences in the heat capacities) when compared
with established classical or quantum-chemical models that are commonly
used to model such properties of bulk liquids.