Dataset Information

Structure of an Ultrathin Oxide on Pt₃ Sn(111) Solved by Machine Learning Enhanced Global Optimization.

ABSTRACT: Determination of the atomic structure of solid surfaces typically depends on comparison of measured properties with simulations based on hypothesized structural models. For simple structures, the models may be guessed, but for more complex structures there is a need for reliable theory-based search algorithms. So far, such methods have been limited by the combinatorial complexity and computational expense of sufficiently accurate energy estimation for surfaces. However, the introduction of machine learning methods has the potential to change this radically. Here, we demonstrate how an evolutionary algorithm, utilizing machine learning for accelerated energy estimation and diverse population generation, can be used to solve an unknown surface structure-the (4×4) surface oxide on Pt₃ Sn(111)-based on limited experimental input. The algorithm is efficient and robust, and should be broadly applicable in surface studies, where it can replace manual, intuition based model generation.

SUBMITTER: Merte LR

PROVIDER: S-EPMC9320988 | biostudies-literature | 2022 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Structure of an Ultrathin Oxide on Pt<sub>3</sub> Sn(111) Solved by Machine Learning Enhanced Global Optimization.

Merte Lindsay R LR Bisbo Malthe Kjaer MK Sokolović Igor I Setvín Martin M Hagman Benjamin B Shipilin Mikhail M Schmid Michael M Diebold Ulrike U Lundgren Edvin E Hammer Bjørk B

Angewandte Chemie (International ed. in English) 20220429 25

Determination of the atomic structure of solid surfaces typically depends on comparison of measured properties with simulations based on hypothesized structural models. For simple structures, the models may be guessed, but for more complex structures there is a need for reliable theory-based search algorithms. So far, such methods have been limited by the combinatorial complexity and computational expense of sufficiently accurate energy estimation for surfaces. However, the introduction of machi ...[more]

PMID: 35384213

Similar Datasets

Project description:Accurately modeling heterogeneous catalysis requires accurate descriptions of rate-controlling elementary reactions of molecules on metal surfaces, but standard density functionals (DFs) are not accurate enough for this. The problem can be solved with the specific reaction parameter approach to density functional theory (SRP-DFT), but the transferability of SRP DFs among chemically related systems is limited. We combine the MS-PBEl, MS-B86bl, and MS-RPBEl semilocal made simple (MS) meta-generalized gradient approximation (GGA) (mGGA) DFs with rVV10 nonlocal correlation, and we evaluate their performance for the hydrogen (H2) + Cu(111), deuterium (D2) + Ag(111), H2 + Au(111), and D2 + Pt(111) gas-surface systems. The three MS mGGA DFs that have been combined with rVV10 nonlocal correlation were not fitted to reproduce particular experiments, nor has the b parameter present in rVV10 been reoptimized. Of the three DFs obtained the MS-PBEl-rVV10 DF yields an excellent description of van der Waals well geometries. The three original MS mGGA DFs gave a highly accurate description of the metals, which was comparable in quality to that obtained with the PBEsol DF. Here, we find that combining the three original MS mGGA DFs with rVV10 nonlocal correlation comes at the cost of a slightly less accurate description of the metal. However, the description of the metal obtained in this way is still better than the descriptions obtained with SRP DFs specifically optimized for individual systems. Using the Born-Oppenheimer static surface (BOSS) model, simulations of molecular beam dissociative chemisorption experiments yield chemical accuracy for the D2 + Ag(111) and D2 + Pt(111) systems. A comparison between calculated and measured E 1/2(ν, J) parameters describing associative desorption suggests chemical accuracy for the associative desorption of H2 from Au(111) as well. Our results suggest that ascending Jacob's ladder to the mGGA rung yields increasingly more accurate results for gas-surface reactions of H2 (D2) interacting with late transition metals.

Dataset Information

Structure of an Ultrathin Oxide on Pt₃ Sn(111) Solved by Machine Learning Enhanced Global Optimization.

Publications

Structure of an Ultrathin Oxide on Pt<sub>3</sub> Sn(111) Solved by Machine Learning Enhanced Global Optimization.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Dataset Information

Structure of an Ultrathin Oxide on Pt3 Sn(111) Solved by Machine Learning Enhanced Global Optimization.

Publications

Structure of an Ultrathin Oxide on Pt<sub>3</sub> Sn(111) Solved by Machine Learning Enhanced Global Optimization.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets

Structure of an Ultrathin Oxide on Pt₃ Sn(111) Solved by Machine Learning Enhanced Global Optimization.