PiNNwall: Heterogeneous Electrode Models from Integrating Machine Learning and Atomistic Simulation

Thomas Dufils,Lisanne Knijff,Yunqi Shao,Chao Zhang

DOI: https://doi.org/10.1021/acs.jctc.3c00359

2023-10-01

Abstract:Electrochemical energy storage always involves the capacitive process. The prevailing electrode model used in the molecular simulation of polarizable electrode-electrolyte systems is the Siepmann-Sprik model developed for perfect metal electrodes. This model has been recently extended to study the metallicity in the electrode by including the Thomas-Fermi screening length. Nevertheless, a further extension to heterogeneous electrode models requires introducing chemical specificity, which does not have any analytical recipes. Here, we address this challenge by integrating the atomistic machine learning code (PiNN) for generating the base charge and response kernel and the classical molecular dynamics code (MetalWalls) dedicated to the modeling of electrochemical systems, and this leads to the development of the PiNNwall interface. Apart from the cases of chemically doped graphene and graphene oxide electrodes as shown in this study, the PiNNwall interface also allows us to probe polarized oxide surfaces in which both the proton charge and the electronic charge can coexist. Therefore, this work opens the door for modeling heterogeneous and complex electrode materials often found in energy storage systems.

Materials Science,Chemical Physics

What problem does this paper attempt to address?

The paper is primarily dedicated to addressing the issue of electrode material simulation in electrochemical energy storage systems, particularly focusing on the modeling challenges of complex and heterogeneous electrode materials. Specifically, the objectives of the paper include:

1. **Developing the PiNNwall Interface**: By integrating machine learning (ML) with atomic-scale simulations, generating heterogeneous electrode models that can reflect the chemical specificity of electrode materials.

2. **Extending the Siepmann-Sprik Model**: The existing Siepmann-Sprik model is mainly used to describe the behavior of ideal metal electrodes but cannot naturally handle the chemical heterogeneity of electrode materials. This study extends this model by introducing machine learning methods, enabling it to handle more complex electrode materials.

3. **Simulating Chemically Doped Graphene and Graphene Oxide Electrodes**: Using the developed PiNNwall interface, the study investigates the effects of different chemical dopings (such as nitrogen doping, epoxy groups, hydroxyl, and carboxyl groups) on the performance of graphene and its derivatives (such as graphene oxide) electrodes.

4. **Exploring Surface Acid-Base Chemistry**: By simulating the protonation/deprotonation behavior of carboxyl-containing graphene oxide surfaces under different pH conditions, the study explores how these surface chemical reactions affect electrode performance.

In summary, this paper aims to develop a new framework by combining machine learning techniques and molecular dynamics simulations to better understand and simulate the behavior of complex electrode materials in electrochemical energy storage systems, especially those with chemical heterogeneity. This will help in designing new high-performance electrode materials and further promote the development of electrochemical energy storage technology.