Page 75 - Read Online

Page 16 of 25 Meng et al. J. Mater. Inf. 2025, 5, 3 https://dx.doi.org/10.20517/jmi.2024.74

Figure 8. (A) “Five-step” screening strategy. Reprinted with permission from Ref. [116] . Copyright © 2021 American Chemical Society; (B)
“Four-step” screening strategy. Reprinted with permission from Ref. [118] . Copyright © 2023 The Authors. Energy & Environmental
Materials published by John Wiley & Sons Australia, Ltd on behalf of Zhengzhou University.

Figure 9. (A) The workflow of ML to explore the origins of catalysis. Reprinted with permission from Ref. [118] . Copyright © 2023 The
Authors. Energy & Environmental Materials published by John Wiley & Sons Australia, Ltd on behalf of Zhengzhou University; (B) ML
screening and descriptor building framework of the work by Zhang et al.; (C) Workflow of the ML screening process as well as the
number of selected candidates after each screening step of the work by Zhang et al. Reprinted with permission from Ref. [119] . Copyright ©
2021 Zhengzhou University. ML: Machine learning.

performance materials. By identifying trends in electrocatalytic properties and using reactivity descriptors to
forecast promising catalysts, we can realize the rational design of catalysts, especially for processes as
complicated as NRR.

Adsorption energy descriptors
The adsorption energy reflects the strength of the interaction between the reactant and the catalyst and is a
key indicator for selecting high-efficiency catalysts. When the adsorption energy falls within the optimal
range, it reduces the reaction energy barrier and, consequently, the U . Catalysts with moderate adsorption
L