Abstract:Crystal structure prediction involves a search of a complex configurational space for local minima corresponding to stable crystal structures, which can be performed efficiently using atom-atom force fields for the assessment of intermolecular interactions. However, for challenging systems, the limitations in the accuracy of force fields prevent a reliable assessment of the relative thermodynamic stability of potential structures, while the cost of fully quantum mechanical approaches can limit applications of the methods. We present a method to rapidly improve force field lattice energies by correcting two-body interactions with a higher level of theory in a fragment-based approach and predicting these corrections with machine learning. Corrected lattice energies with commonly used density functionals and second order perturbation theory (MP2) all significantly improve the ranking of experimentally known polymorphs where the rigid molecule model is applicable. The relative lattice energies of known polymorphs are also found to systematically improve with the fragment corrections. Predicting two-body interactions with atom-centered symmetry functions in a Gaussian process is found to give highly accurate results using as little as 10-20% of the data for training, reducing the cost of the energy correction by up to an order of magnitude. The machine learning approach opens up the possibility of more widespread use of fragment-based methods in crystal structure prediction, whose increased accuracy at a low computational cost will benefit applications in areas such as polymorph screening and computer-guided materials discovery.

Machine-learning accelerated structure search for ligand-protected clusters

Machine Learning Potential Aided Structure Search for Low-Lying Candidates of Au Clusters

Accelerated prediction of atomically precise cluster structures using on-the-fly machine learning

A Conformational Analysis Method for Understanding the Energy Landscapes of Clusters.

CASTELO: Clustered Atom Subtypes aidEd Lead Optimization -- a combined machine learning and molecular modeling method

Graph Attention Network for Global Search of Atomic Clusters: A Case Study of Agn (n = 14−26) Clusters

Machine-Learned Fragment-Based Energies for Crystal Structure Prediction

Electroactive and Stretchable Hydrogels of 3,4-Ethylenedioxythiophene/thiophene Copolymers

A Modified Dynamic Lattice Searching Method to Predict the Lowest Energy Structure of Clusters for the Au Clusters

Gold-Thiolate Nanocluster Dynamics and Intercluster Reactions Enabled by a Machine Learned Interatomic Potential

Machine learning hydrogen adsorption on nanoclusters through structural descriptors

Data-efficient machine learning for molecular crystal structure prediction

Efficient Cysteine Conformer Search with Bayesian Optimization

A Hybrid Machine Learning Approach for Structure Stability Prediction in Molecular Co-crystal Screenings

Long-range dispersion-inclusive machine learning potentials for structure search and optimization of hybrid organic-inorganic interfaces

Machine-learning accelerated annealing with fitting-search style for multicomponent alloy structure predictions

Accelerated Organic Crystal Structure Prediction with Genetic Algorithms and Machine Learning

Atomic Structure Optimization with Machine-Learning Enabled Interpolation between Chemical Elements

Accelerating the discovery of low-energy structure configurations: a computational approach that integrates first-principles calculations, Monte Carlo sampling, and Machine Learning

Machine-learning structural reconstructions for accelerated point defect calculations

Discovering Syntheses of Atomically Precise Metal Nanoclusters by Applying Machine Learning to a High‐Throughput Platform