Please use this identifier to cite or link to this item:
Title: An invertible crystallographic representation for general inverse design of inorganic crystals with targeted properties
Authors: REN ZEKUN 
Tian, Siyu Isaac Parker
Noh, Juhwan
Oviedo, Felipe
Xing, Guangzong
Liang, Qiaohao
Zhu, Ruiming
Armin Gerhard Aberle 
Sun, Shijing
Liu, Yi
Li, Qianxiao
Jayavelu, Senthilnath
Hippalgaonkar, Kedar
Jung, Yousung
Buonassisi, Tonio
Keywords: general inverse design
solid-state materials
invertible crystallographic representation
generalized crystallographic representation
property-structured latent space
variational autoencoder
machine learning
generative model
Issue Date: 20-Dec-2021
Publisher: Cell Press
Citation: REN ZEKUN, Tian, Siyu Isaac Parker, Noh, Juhwan, Oviedo, Felipe, Xing, Guangzong, JIALI LI, Liang, Qiaohao, Zhu, Ruiming, Armin Gerhard Aberle, Sun, Shijing, WANG XIAONAN, Liu, Yi, Li, Qianxiao, Jayavelu, Senthilnath, Hippalgaonkar, Kedar, Jung, Yousung, Buonassisi, Tonio (2021-12-20). An invertible crystallographic representation for general inverse design of inorganic crystals with targeted properties. Matter. ScholarBank@NUS Repository.
Abstract: Realizing general inverse design could greatly accelerate the discovery of new materials with user-defined properties. However, state-of-the-art generative models tend to be limited to a specific composition or crystal structure. Herein, we present a framework capable of general inverse design (not limited to a given set of elements or crystal structures), featuring a generalized invertible representation that encodes crystals in both real and reciprocal space, and a property-structured latent space from a variational autoencoder (VAE). In three design cases, the framework generates 142 new crystals with user-defined formation energies, bandgap, thermoelectric (TE) power factor, and combinations thereof. These generated crystals, absent in the training database, are validated by first-principles calculations. The success rates (number of first-principles-validated target-satisfying crystals/number of designed crystals) ranges between 7.1% and 38.9%. These results represent a significant step toward property-driven general inverse design using generative models, although practical challenges remain when coupled with experimental synthesis.
Source Title: Matter
ISSN: 2590-2385
DOI: 10.1016/j.matt.2021.11.032
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
Accepted Manuscript.pdfAccepted version1.47 MBAdobe PDF



Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.