Регистрация / Вход
Прислать материал

An expert system for the design crystalline materials

Name
Alexander
Surname
Shevchenko
Scientific organization
Samara State Aerospace University
Academic degree
PhD in Chemistry
Position
Senior Researcher
Scientific discipline
Chemistry & Chemical technologies
Topic
An expert system for the design crystalline materials
Abstract
- Development of an expert system, containing information about the correlations between the chemical composition and geometric-topological properties of crystal structures.
- Testing of the expert system for the prediction of structures 1, 2 and 3-periodic coordination polymers.
- Services Package for remote access to information stored in SCTMS databases.
Keywords
topology of crystal structures, ligands, metal atoms, Voronoi polyhedra, big data, knowledge base, data meaning
Summary

One of the main components in the design of new metal-organic frameworks are the choice of the complexing atoms and the ligands, which determine, as a rule, the topological type of crystal structure. To simplify this procedure, we have created a knowledge base about the complexing atoms “Topological Types of Atoms” (TTA) and ligands “Topological Types of Ligands” (TTL). We used as data sources information about the known crystal structures stored in Cambridge Structure Database (F.H.Allen, Acta Crystallogr., 2002, B58, 380) and Inorganic Crystal Structures Database (A.Belsky, M.Hellenbrandt, V.L.Karen and P.Luksch, Acta Crystallogr., 2002, B58, 364). 
Only such crystal structures were taken into account, which satisfy the conditions listed below:
- the crystal structures contain metal atoms, and are not alloys or intermetallic compounds;
- statistical disordering of metal atoms and its environment are absent in the structure;
- if the structures were obviously erroneous interatomic distances or atomic volumes, incomplete or incorrect chemical composition of the compounds, they are excluded from consideration.
As a result, our sample contained 360,051 crystal structures with 692,924 metal atoms having 76 chemical species. The Voronoi polyhedron, the coordination figure and the chemical composition of the first coordination sphere have been calculated for all metal atoms. This information is assembled in the TTA knowledge base. To determine the shape of the coordination figure we developed and used a new algorithm based on angular fingerprints.
The knowledge base TTL stores for each ligand information about its chemical composition, the graph of interatomic bonds, the occurrence and the coordination type of ligand in the crystal structures. At present, the knowledge base TTL contains 161,887 entries. 
Now we are creating for data processing in the knowledge bases TTA, TTL, TTD and TTO (A. R. Oganov, ed. (2010). Modern Methods of Crystal Structure Prediction. Berlin: Wiley-VCH. ISBN 978-3-527-40939-6) software that you need to find the rules and laws of crystal structure building. The results will be used for the design of new crystalline materials.