[Paper Review] An Introduction to Electrocatalyst Design using Machine Learning for Renewable Energy Storage

C. Lawrence Zitnick, Lowik Chanussot+15 more|arXiv (Cornell University)|Oct 14, 2020

Machine Learning in Materials Science35 citations

TL;DR

The paper introduces how machine learning can approximate density functional theory calculations to accelerate electrocatalyst design for hydrogen energy storage and methane synthesis, via the OC20 dataset.

ABSTRACT

Scalable and cost-effective solutions to renewable energy storage are essential to addressing the world's rising energy needs while reducing climate change. As we increase our reliance on renewable energy sources such as wind and solar, which produce intermittent power, storage is needed to transfer power from times of peak generation to peak demand. This may require the storage of power for hours, days, or months. One solution that offers the potential of scaling to nation-sized grids is the conversion of renewable energy to other fuels, such as hydrogen or methane. To be widely adopted, this process requires cost-effective solutions to running electrochemical reactions. An open challenge is finding low-cost electrocatalysts to drive these reactions at high rates. Through the use of quantum mechanical simulations (density functional theory), new catalyst structures can be tested and evaluated. Unfortunately, the high computational cost of these simulations limits the number of structures that may be tested. The use of machine learning may provide a method to efficiently approximate these calculations, leading to new approaches in finding effective electrocatalysts. In this paper, we provide an introduction to the challenges in finding suitable electrocatalysts, how machine learning may be applied to the problem, and the use of the Open Catalyst Project OC20 dataset for model training.

Motivation & Objective

• Motivate scalable, cost-effective renewable energy storage strategies (HES and methane synthesis).
• Explain the role of electrocatalysts in improving electrochemical reactions and reducing costs.
• Introduce density functional theory limitations and the need for ML-based approximations.
• Present the Open Catalyst Project OC20 dataset as a foundation for ML models in catalysis.

Proposed method

• Describe the challenges in electrocatalyst discovery and the relevance of DFT relaxations.
• Discuss how ML models can learn from DFT data to predict energies and geometries.
• Introduce the OC20 dataset as a resource for training ML models for adsorption and relaxation tasks.
• Outline potential ML model directions for predicting catalyst-adsorbate interactions at surfaces.

Experimental results

Research questions

• RQ1 How can ML approximate quantum mechanical calculations to accelerate catalyst discovery for renewable energy storage?
• RQ2 What dataset and tasks are suitable for training ML models to predict adsorption energies and relaxed geometries?
• RQ3 What is the potential impact of ML-accelerated catalyst design on the feasibility of HES and methane synthesis?

Key findings

• ML can potentially provide efficient approximations to DFT, enabling large-scale exploration of catalyst options.
• OC20 is proposed as a dataset to train models for adsorption and relaxation tasks relevant to catalyst screening.
• Efficient ML models could enable testing of millions of catalyst-adsorbate combinations that are impractical with full DFT alone.
• The discussion highlights the high computational cost of DFT and the value of ML in accelerating catalyst discovery.
• The paper frames hydrogen energy storage and methane synthesis as key application domains for ML-aided electrocatalyst design.