Cargando…

Charting the Atomic C Interaction with Transition Metal Surfaces

[Image: see text] Carbon interaction with transition metal (TM) surfaces is a relevant topic in heterogeneous catalysis, either for its poisoning capability, for the recently attributed promoter role when incorporated in the subsurface, or for the formation of early TM carbides, which are increasing...

Descripción completa

Detalles Bibliográficos
Autores principales: Piqué, Oriol, Koleva, Iskra Z., Bruix, Albert, Viñes, Francesc, Aleksandrov, Hristiyan A., Vayssilov, Georgi N., Illas, Francesc
Formato: Online Artículo Texto
Lenguaje:English
Publicado: American Chemical Society 2022
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9880994/
https://www.ncbi.nlm.nih.gov/pubmed/36718273
http://dx.doi.org/10.1021/acscatal.2c01562
Descripción
Sumario:[Image: see text] Carbon interaction with transition metal (TM) surfaces is a relevant topic in heterogeneous catalysis, either for its poisoning capability, for the recently attributed promoter role when incorporated in the subsurface, or for the formation of early TM carbides, which are increasingly used in catalysis. Herein, we present a high-throughput systematic study, adjoining thermodynamic plus kinetic evidence obtained by extensive density functional calculations on surface models (324 diffusion barriers located on 81 TM surfaces in total), which provides a navigation map of these interactions in a holistic fashion. Correlation between previously proposed electronic descriptors and ad/absorption energies has been tested, with the d-band center being found the most suitable one, although machine learning protocols also underscore the importance of the surface energy and the site coordination number. Descriptors have also been tested for diffusion barriers, with ad/absorption energies and the difference in energy between minima being the most appropriate ones. Furthermore, multivariable, polynomial, and random forest regressions show that both thermodynamic and kinetic data are better described when using a combination of different descriptors. Therefore, looking for a single perfect descriptor may not be the best quest, while combining different ones may be a better path to follow.