Innovations in Materials Design, Part I: The Design of Inorganic Compounds. Part II: Searching for New Electro-Optical, Ferro-Electric, Superconducting, and Semiconducting Materials.
Abstract
The use of computer learning strategies for predicting inorganic compounds which are believed promising as new electro-optical, ferro-electric, superconducting or semiconducting materials is explained. Prediction reliability utilizing these computer learning strategies is based on: (1) expert selection of example compounds, (2) expert assessment of data for computer learning, and (3) comparison of predictions which have been obtained using various feature sets. The classes of the inorganic compounds most promising for searching for new electro-optical, ferro-electric, superconducting, and semiconducting materials are directly based on the analysis of the application domains and the known data. The results of predicting the crystal structure types at normal pressure and room temperature for the compound with composition of AB2Se4 are presented. Types considered were chalcopyrite, Th3P4, CaFe2O4, Yb3S4, Yb3Se4, PbGa2Se4, NiCr2Se4, spinel, or olivine. Analysis of predictions showed that the structures resembling olivine and NiCr2Se4 are an inherent feature of the compounds with composition A(IV)B(II)2Se4. but the structure types Th3P4 and NiCr2Se4 are characteristic of compounds with composition A(II)B(III)2Se4. Prediction of the crystal structure types at standard conditions for compounds with composition ABX2 were also carried out. Types considered included chalcopyrite, a or b-NaFeO2, a-LiFeO2, or TISe.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1996
- Accession Number
- ADA319754
Entities
People
- N. D. Vashenko
- N. N. Kiselyova
- N. V. Kravchenko
- N. V. Petukhov
- V. P. Gladun