Inverse design of composite metal oxide optical materials based on deep transfer learning and global optimization

Optical materials with special optical properties are widely used in a broad span of technologies, from computer displays to solar energy utilization leading to large dataset accumulated from years of extensive materials synthesis and optical characterization. Previously, machine learning models have been developed to predict the optical absorption spectrum from a materials characterization image or vice versa. Herein we propose TLOpt, a transfer learning based inverse optical materials design algorithm for suggesting material compositions with a desired target light absorption spectrum. Our approach is based on the combination of a deep neural network model and global optimization algorithms including a genetic algorithm and Bayesian optimization. A transfer learning strategy is employed to solve the small dataset issue in training the neural network predictor of optical absorption spectrum using the Magpie materials composition descriptor. Our extensive experiments show that our algorithm can inverse design the materials composition with stoichiometry with high accuracy.

[1]  Gang Sun,et al.  Artificial neural network based inverse design: Airfoils and wings , 2015 .

[2]  Yi Yang,et al.  Nanophotonic particle simulation and inverse design using artificial neural networks , 2018, Science Advances.

[3]  Hui Xiong,et al.  A Comprehensive Survey on Transfer Learning , 2021, Proceedings of the IEEE.

[4]  Harold J. Kushner,et al.  A New Method of Locating the Maximum Point of an Arbitrary Multipeak Curve in the Presence of Noise , 1964 .

[5]  Jianjun Hu,et al.  Generative adversarial networks (GAN) based efficient sampling of chemical composition space for inverse design of inorganic materials , 2019, npj Computational Materials.

[6]  Ingo Rechenberg,et al.  Evolutionsstrategie : Optimierung technischer Systeme nach Prinzipien der biologischen Evolution , 1973 .

[7]  Lalit M. Patnaik,et al.  Genetic algorithms: a survey , 1994, Computer.

[8]  A. Oganov,et al.  Search for stable cocrystals of energetic materials using the evolutionary algorithm USPEX. , 2020, Physical chemistry chemical physics : PCCP.

[9]  Alán Aspuru-Guzik,et al.  Inverse molecular design using machine learning: Generative models for matter engineering , 2018, Science.

[10]  J. Gregoire,et al.  Machine Learning of Optical Properties of Materials - Predicting Spectra from Images and Images from Spectra , 2018 .

[11]  T. Liao,et al.  Metaheuristic-based inverse design of materials – A survey , 2020 .

[12]  Todd G. Deutsch,et al.  Sunlight absorption in water – efficiency and design implications for photoelectrochemical devices , 2014 .

[13]  Olexandr Isayev,et al.  Deep reinforcement learning for de novo drug design , 2017, Science Advances.

[14]  Zongfu Yu,et al.  Training Deep Neural Networks for the Inverse Design of Nanophotonic Structures , 2017, 2019 Conference on Lasers and Electro-Optics (CLEO).

[15]  Y. Ikeda.  A New Method of Alloy Design Using a Genetic Algorithm and Molecular Dynamics Simulation and Its Application to Nickel-Based Superalloys , 1997 .

[16]  Christopher M Wolverton,et al.  Atomistic calculations and materials informatics: A review , 2017 .

[17]  Yuxin Cui,et al.  Limited Data Rolling Bearing Fault Diagnosis With Few-Shot Learning , 2019, IEEE Access.

[18]  V. Batista,et al.  Search for Catalysts by Inverse Design: Artificial Intelligence, Mountain Climbers, and Alchemists. , 2019, Chemical reviews.

[19]  Alok Choudhary,et al.  A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials , 2016 .

[20]  Jonathan A. Fan,et al.  Simulator-based training of generative neural networks for the inverse design of metasurfaces , 2019, Nanophotonics.

[21]  Shu Yang,et al.  Universal inverse design of surfaces with thin nematic elastomer sheets , 2017, Proceedings of the National Academy of Sciences.

[22]  Juan Li,et al.  Learning solutions to the source inverse problem of wave equations using LS-SVM , 2019, Journal of Inverse and Ill-posed Problems.

[23]  Alexander Y. Piggott,et al.  Fabrication-constrained nanophotonic inverse design , 2016, Scientific Reports.

[24]  Krishna Rajan,et al.  Application-Driven Data Analysis , 2009 .

[25]  A. Zunger.  Inverse design in search of materials with target functionalities , 2018 .

[26]  Alán Aspuru-Guzik,et al.  Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules , 2016, ACS central science.

[27]  Lena Maier-Hein,et al.  Robust near real-time estimation of physiological parameters from megapixel multispectral images with inverse Monte Carlo and random forest regression , 2016, International Journal of Computer Assisted Radiology and Surgery.

[28]  Karzan Wakil,et al.  Prediction of concrete strength in presence of furnace slag and fly ash using Hybrid ANN-GA (Artificial Neural Network-Genetic Algorithm) , 2020 .

[29]  Jianjun Hu,et al.  Convolutional Neural Networks for Crystal Material Property Prediction Using Hybrid Orbital-Field Matrix and Magpie Descriptors , 2019, Crystals.

[30]  Xingao Gong,et al.  Inverse Design of Materials by Multi-Objective Differential Evolution($IM^2ODE$) , 2014 .

[31]  Kyu-Tae Lee,et al.  A Generative Model for Inverse Design of Metamaterials , 2018, Nano letters.

[32]  Donald R. Jones,et al.  Efficient Global Optimization of Expensive Black-Box Functions , 1998, J. Glob. Optim..

[33]  Jinlan Wang,et al.  Accelerated discovery of stable lead-free hybrid organic-inorganic perovskites via machine learning , 2018, Nature Communications.

[34]  Tejs Vegge,et al.  Genetic algorithms for computational materials discovery accelerated by machine learning , 2019, npj Computational Materials.

[35]  Yuxin Cui,et al.  Transfer Learning with Deep Recurrent Neural Networks for Remaining Useful Life Estimation , 2018, Applied Sciences.

[36]  T. Kaewmaraya,et al.  The study of structural, morphological and optical properties of (Al, Ga)-doped ZnO: DFT and experimental approaches , 2019, Applied Surface Science.

[37]  Steven G. Johnson,et al.  Inverse design of large-area metasurfaces. , 2018, Optics express.

[38]  Alán Aspuru-Guzik,et al.  Optimizing distributions over molecular space. An Objective-Reinforced Generative Adversarial Network for Inverse-design Chemistry (ORGANIC) , 2017 .

[39]  Tom K Woo,et al.  Materials design by evolutionary optimization of functional groups in metal-organic frameworks , 2016, Science Advances.

[40]  Enrico Zio,et al.  A Bayesian Optimal Design for Accelerated Degradation Testing Based on the Inverse Gaussian Process , 2017, IEEE Access.

[41]  Douglas A. Keszler,et al.  Inverse Design of High Absorption Thin‐Film Photovoltaic Materials , 2013 .

[42]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[43]  Kris T. Delaney,et al.  Inverse Design of Bulk Morphologies in Multiblock Polymers Using Particle Swarm Optimization , 2017 .

[44]  Jasper Snoek,et al.  Practical Bayesian Optimization of Machine Learning Algorithms , 2012, NIPS.

[45]  Nantiwat Pholdee,et al.  Inverse problem based differential evolution for efficient structural health monitoring of trusses , 2018, Appl. Soft Comput..

[46]  F. Schüth,et al.  On the suitability of different representations of solid catalysts for combinatorial library design by genetic algorithms. , 2008, Journal of combinatorial chemistry.

[47]  Jelena Vucković,et al.  Inverse design in nanophotonics , 2018, Nature Photonics.

[48]  J. Gregoire,et al.  Synthesis, optical imaging, and absorption spectroscopy data for 179072 metal oxides , 2019, Scientific Data.

[49]  M. Khalid,et al.  Synthesis, crystal structure analysis, spectral IR, UV–Vis, NMR assessments, electronic and nonlinear optical properties of potent quinoline based derivatives: Interplay of experimental and DFT study , 2019, Journal of Saudi Chemical Society.