2021.10.14 Webinar: AI-accelerated discovery processes for next-generation battery materials

Presenter: Tejs Vegge, Technical University of Denmark, Department of Energy Conversion and Storage, 2800 Kgs. Lyngby, Denmark

Abstract: Secondary batteries play a crucial role in the green transition, but until now, the development of low-cost and high capacity electrode materials has been too slow. Understanding and controlling the complex and dynamic processes in the electrodes, electrolytes and battery interfaces hold the key to develop ultra-performant and sustainable batteries. Atomic-scale simulations have reached predictive accuracy in many aspects of the materials design, characterization and discovery process. Here, we provide a number of recent examples of how density functional theory (DFT) simulations supported by machine learning and cluster expansion techniques [1] can been used efficiently to identify the limiting thermodynamic, ionic and electronic transport mechanisms in, e.g., lithium rich electrode materials [2] and multivalent systems [3]. We also present recent results on accelerated discovery and inverse design of emerging battery materials using autonomous workflows for prediction of the thermodynamic and kinetic properties of multivalent electrodes [4] and anionic redox processes in Li-rich compounds [5]. We further introduce the “Battery Interface Genome – Materials Acceleration Platform” (www.BIG-MAP.eu) project, which is part of the large-scale European research initiative BATTERY 2030+ and focuses on AI-orchestrated acquisition and utilization of data from the full battery discovery cycle to accelerate the materials and battery discovery process. We discuss the development of AI-based methods for optimal experimental design [6] and generative deep learning models for prediction of the spatio-temporal evolution of battery interphases [7], using models that are trained on multi-sourced and multi-fidelity data from multiscale computer simulations, operando characterization, high-throughput synthesis and laboratory testing. Finally, we will give our perspectives on a path towards better and smarter batteries by combining AI with multisensory and self-healing approaches [8].

Date and time: Thursday 14th of October 2021 from 14.00 to 14.45 (CEST).

Language: English.

Participation: The webinar will take place using Microsoft Teams online platform. Participants will receive an email with a link upon registration by filling out the form below.

Extended Registration deadline: Wednesday 13th of October at 18:00 (CEST).


[1] Chang, et al. J. Phys: CM 31, 325901 (2019)

[2] Chang et. al, J. Mater. Chem. A 7, 16551 (2020)

[3] Li et al, Angew. Chemie 59, 11483 (2020)

[4] Bölle et al, Batteries & Supercaps 3, 488 (2020)

[5] Tygesen et al, npj Comp. Mater. 6, 65 (2020)

[6] Bhowmik et al,  Energy Storage Mater. 21, 446 (2019)

[7] Bhomik & Vegge, Joule 4, 717 (2020)

[8] Vegge et al, Adv. Energy Mater, 10.1002/aenm.202100362 (2021)

Permanent link to this article: http://batteriselskab.dk/arrangementer/2021-10-14-webinar-ai-accelerated-discovery-processes-for-next-generation-battery-materials.htm