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

Reticular materials, including metal-organic frameworks and covalent organicframeworks, combine relative ease of synthesis and an impressive range ofapplications in various fields, from gas storage to biomedicine. Diverseproperties arise from the variation of building units$\unicode{x2013}$metalcenters and organic linkers$\unicode{x2013}$in almost infinite chemical space.Such variation substantially complicates experimental design and promotes theuse of computational methods. In particular, the most successful artificialintelligence algorithms for predicting properties of reticular materials areatomic-level graph neural networks, which optionally incorporate domainknowledge. Nonetheless, the data-driven inverse design involving these modelssuffers from incorporation of irrelevant and redundant features such as fullatomistic graph and network topology. In this study, we propose a new way ofrepresenting materials, aiming to overcome the limitations of existing methods;the message passing is performed on a coarse-grained crystal graph thatcomprises molecular building units. To highlight the merits of our approach, weassessed predictive performance and energy efficiency of neural networks builton different materials representations, including composition-based andcrystal-structure-aware models. Coarse-grained crystal graph neural networksshowed decent accuracy at low computational costs, making them a valuablealternative to omnipresent atomic-level algorithms. Moreover, the presentedmodels can be successfully integrated into an inverse materials design pipelineas estimators of the objective function. Overall, the coarse-grained crystalgraph framework is aimed at challenging the prevailing atom-centric perspectiveon reticular materials design.

Coarse-grained crystal graph neural networks for reticular materials design