Abstract:Reticular materials, including metal-organic frameworks and covalent organic frameworks, combine relative ease of synthesis and an impressive range of applications in various fields, from gas storage to biomedicine. Diverse properties arise from the variation of building units$\unicode{x2013}$metal centers and organic linkers$\unicode{x2013}$in almost infinite chemical space. Such variation substantially complicates experimental design and promotes the use of computational methods. In particular, the most successful artificial intelligence algorithms for predicting properties of reticular materials are atomic-level graph neural networks, which optionally incorporate domain knowledge. Nonetheless, the data-driven inverse design involving these models suffers from incorporation of irrelevant and redundant features such as full atomistic graph and network topology. In this study, we propose a new way of representing materials, aiming to overcome the limitations of existing methods; the message passing is performed on a coarse-grained crystal graph that comprises molecular building units. To highlight the merits of our approach, we assessed predictive performance and energy efficiency of neural networks built on different materials representations, including composition-based and crystal-structure-aware models. Coarse-grained crystal graph neural networks showed decent accuracy at low computational costs, making them a valuable alternative to omnipresent atomic-level algorithms. Moreover, the presented models can be successfully integrated into an inverse materials design pipeline as estimators of the objective function. Overall, the coarse-grained crystal graph framework is aimed at challenging the prevailing atom-centric perspective on reticular materials design.

An Invertible, Invariant Crystal Representation for Inverse Design of Solid-State Materials Using Generative Deep Learning

An invertible crystallographic representation for general inverse design of inorganic crystals with targeted properties

SLICES-PLUS: A Crystal Representation Leveraging Spatial Symmetry

MatterGPT: A Generative Transformer for Multi-Property Inverse Design of Solid-State Materials

Towards invertible 2D crystal structure representation for efficient downstream task execution

VQCrystal: Leveraging Vector Quantization for Discovery of Stable Crystal Structures

Generative Inverse Design of Crystal Structures via Diffusion Models with Transformers

Generative deep learning for the inverse design of materials

Generative Design of Crystal Structures by Point Cloud Representations and Diffusion Model

Inverse design of two-dimensional materials with invertible neural networks

Global Inverse Design Across Multiple Photonic Structure Classes Using Generative Deep Learning

3-D Inorganic Crystal Structure Generation and Property Prediction via Representation Learning

Coarse-grained crystal graph neural networks for reticular materials design

Inverse design of glass structure with deep graph neural networks

Attribute driven inverse materials design using deep learning Bayesian framework

Inverse Design of Crystals and Quasicrystals in a Non-Additive Binary Mixture of Hard Disks

Deep-learning-based inverse design model for intelligent discovery of organic molecules

Graph deep learning accelerated efficient crystal structure search and feature extraction

Physics guided deep learning for generative design of crystal materials with symmetry constraints

Machine Learning-Aided Crystal Facet Rational Design With Ionic Liquid Controllable Synthesis