Metadynamics sampling in atomic environment space for collecting training data for machine learning potentials

The universal mathematical form of machine-learning potentials(MLPs)shifts the core of development of interatomic potentials to collecting proper training data.Ideally,the training set should encompass diverse local atomic environments but conventional approaches are prone to sampling similar configurations repeatedly,mainly due to the Boltzmann statistics.As such,practitioners handpick a large pool of distinct configurations manually,stretching the development period significantly.To overcome this hurdle,methods are being proposed that automatically generate training data.Herein,we suggest a sampling method optimized for gathering diverse yet relevant configurations semi-automatically.This is achieved by applying the metadynamics with the descriptor for the local atomic environment as a collective variable.As a result,the simulation is automatically steered toward unvisited local environment space such that each atom experiences diverse chemical environments without redundancy.We apply the proposed metadynamics sampling to H:Pt(111),GeTe,and Si systems.Throughout these examples,a small number of metadynamics trajectories can provide reference structures necessary for training high-fidelity MLPs.By proposing a semiautomatic sampling method tuned for MLPs,the present work paves the way to wider applications of MLPs to many challenging applications.

Accelerated identification of equilibrium structures of multicomponent inorganic crystals using machine learning potentials

The discovery of multicomponent inorganic compounds can provide direct solutions to scientific and engineering challenges,yet the vast uncharted material space dwarfs synthesis throughput.While the crystal structure prediction(CSP)may mitigate this frustration,the exponential complexity of CSP and expensive density functional theory(DFT)calculations prohibit material exploration at scale.Herein,we introduce SPINNER,a structure-prediction framework based on random and evolutionary searches.Harnessing speed and accuracy of neural network potentials(NNPs),the program navigates configurational spaces 10^(2)–10^(3) times faster than DFT-based methods.Furthermore,SPINNER incorporates algorithms tuned for NNPs,achieving performances exceeding conventional algorithms.In blind tests on 60 ternary compositions,SPINNER identifies experimental(or theoretically more stable)phases for~80%of materials.When benchmarked against data-mining or DFT-based evolutionary predictions,SPINNER identifies more stable phases in many cases.By developing a reliable and fast structure-prediction framework,this work paves the way to large-scale,open exploration of undiscovered inorganic crystals.