English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Single atomic sites often determine the functionality and performance ofmaterials, such as catalysts, semi-conductors or enzymes. Computing andunderstanding the properties of such sites is therefore a crucial component ofthe rational materials design process. Because of complex electronic effects atthe atomic level, atomic site properties are conventionally derived fromcomputationally expensive first-principle calculations, as this level of theoryis required to achieve relevant accuracy. In this study, we present a widelyapplicable machine learning (ML) approach to compute atomic site propertieswith high accuracy in real time. The approach works well for complexnon-crystalline atomic structures and therefore opens up the possibility forhigh-throughput screenings of nano-materials, amorphous systems and materialsinterfaces. Our approach includes a robust featurization scheme to transformatomic structures into features which can be used by common machine learningmodels. Performing a genetic algorithm (GA) based feature selection, we showhow to establish an intuitive physical interpretation of the structure-propertyrelations implied by the ML models. With this approach, we compute atomic sitestabilities of metal nanoparticles ranging from 3-55 atoms with mean absoluteerrors in the range of 0.11-0.14 eV in real time. We also establish thechemical identity of the site as most important factor in determining atomicsite stabilities, followed by structural features like bond distances andangles. Both, the featurization and GA feature selection functionality arepublished in open-source python modules. With this method, we enable theefficient rational design of highly specialized real-world nano-catalyststhrough data-driven materials screening.

Artificial intelligence real-time prediction and physical interpretation of atomic binding energies in nano-scale metal clusters