High Accuracy Simulation Methods for Realistic Materials Modelling
Abstract
The main objective of this short study is to demonstrate the feasibility of using highly accurate methods, based on the quantum Monte Carlo (QMC) technique, to provide benchmarks energies to aid the construction of machine learning (ML) approaches to study complex and realistic material science systems. The properties of materials and how they interact with the environment is often modified by the adsorption of water and other molecules on their surfaces. In particular, the interaction of water with various surfaces, and the interaction of water with other molecules on surfaces is ubiquitous in several chemical and biological processes, not to mention that every surface in contact with Earth's atmosphere is exposed to water vapour. Interaction energies between water and other molecules on carbidic and other complex surfaces is important to an almost endless list of practical applications such as corrosion, lubrication, friction, catalysis, coatings, energy storage, and sensors to name a few. Given that water interacts weakly with graphitic surfaces and that there are many competing structures with similar energies, this is one of the most challenging systems for any modelling method, and for this reason an accurate reference atomic level approach is needed to make progress with understanding these systems. If the approach can be validated on such a complex system, it would provide a firm ground to build on for the study other difficult systems, such as 2 dimensional layered materials. To demonstrate the feasibility of our approach, our initial focus will be on water and its interaction with graphene, as this is a system for which we have considerable experience and made most of the ground work, demonstrating the high accuracy of the QMC method. We will generate large sets of configurations of water clusters interacting with a graphene layer and compute the QMC energies of these configurations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 26, 2023
- Accession Number
- AD1211809
Entities
People
- Andrea Zen
- Angelos Michaelides
- Dario Alfè
Organizations
- University College London