Materials Chemistry and Catalysis

Developing Machine Learning Models to Understand Cathode-Electrolyte Interfaces in Solid-State Batteries

Rapid consumption of fossil fuels necessitates a shift to electric vehicles powered by advanced energy resources. Traditional lithium-ion batteries (LIBs), widely used since their introduction in the 1970s and commercialization in the 1990s, now require improvements to cater to growing energy demands, with increased safety, energy density, and longevity^[1]. New high-capacity electrodes are being explored, but safety concerns with liquid electrolytes impede their use. Solid-state electrolytes, offering greater safety and superior mechanical, chemical, and thermal properties, have emerged as a solution^[2]. Research on solid-state electrolytes, capable of matching or surpassing the conductivity of liquid counterparts, is paving the way for safer, more efficient all-solid-state batteries (ASSBs)^[3].
However, the development of ASSBs is hindered by interface issues^[4-5]. My research aims to utilize atomistic simulations to understand the cathode-electrolyte interface. For a complex cathode-electrolyte interface, conventional DFT and empirical forcefield methods cannot meet the requirements of both accuracy and efficiency. Machine learning force fields^[6-7], which have recently emerged as a solution for large-scale atomistic simulations, can enable us to simulate complex cathode-electrolyte interfaces. We will first generate a DFT reference database that contains structure, energy, and force, which will then be used as a training set for the atomic energy network (ænet) package. The resulting machine learning model can facilitate large-scale and accurate atomistic simulations to better understand the cathode-electrolyte interface.

References

[1] Etacheri, Vinodkumar, et al. Energy & Environmental Science 4.9 (2011): 3243-3262.

[2] Janek, Jürgen, and Wolfgang G. Zeier. Nature Energy 8.3 (2023): 230-240.

[3] Bresser, Dominic, et al., J. Power Sources. 382 (2018): 176-178.

[4] Chen, Rusong, et al. Chemical reviews 120.14 (2019): 6820-6877.

[5] Xiao, Yihan, et al. Nature Reviews Materials 5.2 (2020): 105-126.

[6] Artrith, Nongnuch, and Alexander Urban. Computational Materials Science 114 (2016): 135-150.

[7] Artrith, Nongnuch, Alexander Urban, and Gerbrand Ceder. Physical Review B 96.1 (2017): 014112.

C.V.

September 2022 – present

PhD candidate at the Materials Chemistry and Catalysis, Utrecht University, the Netherlands.
Supervised by: dr. Nong Artrith.

Education

September 2019 – July 2022

MSc Physics, Hunan University, People’s Republic of China.
MSc Thesis: Unveiling the Role of Li⁺ Solvation Structures with Commercial Carbonates in the Formation of Solid Electrolyte Interphase for Lithium Metal Batteries.