Elements of artificial intelligence and mathematical modeling are already an integral part of chemistry and materials science. For their development, mathematical apparatus and models for describing “composition-structure-property” correlations are being formed. However, methods for the design of materials with a crystalline structure have begun to appear very recently, in contrast to organic chemistry, where they form the basis of planning for almost every synthesis. This is due to the fact that the formation of a periodic structure imposes stereochemical and topological restrictions that are not fully known. Systematization of information and the use of new descriptors calculated from the properties of extended networks make it possible to detect general and specific patterns that control the structural organization and properties of materials.

Structural models and approaches for their analysis make it possible to predict the composition, structure and properties of chemical compounds, such as crystals and materials based on them, which is together called materials design. Design is necessary to create new materials with promising properties, such as sorption; catalysis; magnetic susceptibility; optical effects; electrical conductivity; mechanical and chemical response; and ionic and electronic conductivity, etc. Understanding the influence of the geometry and composition of molecular building units, covalent bonds and noncovalent interactions to the final structure is important for uncovering the rules of self-assembly and developing the rational strategies for designing new materials. The goal of the current Research Topic is to cover promising, recent, and novel research trends in the design of extended networks for tuning functionality of materials.

These techniques give advantages in prediction structures and properties of materials. The computer design of materials requires use of language that is understandable for digital systems — this is language of numbers and logical operations. Therefore, structures and relations with properties should be represented in the form of numerical or symbolic descriptors and logical operations. Upon accumulating the data about descriptors and relations between them the black box (machine learning, knowledge databases etc.) methods can be applied for the fast prediction of vast number of materials. However, not all data are available only from experiment. Periodic structure simulations (DFT, MD, MC, QTAIM) can provide data about electronic structure and properties of large number of unavailable materials (also yet not synthesised) at wide range of thermodynamic conditions. Data from simulations are usually more diverse and complete that satisfies better the requirements for teaching a machine.

We welcome Original Research, Review, Mini Review and Perspective articles on themes including, but not limited to:
• Development of computational approaches for structural analysis and new descriptors calculation
• Application of the computational toolbox in the study of the extended networks of structures of functional materials
• Strategies for tunning physical and chemical properties of materials through constructing structures of specified architecture directly or by implication governing the property
• New or unusual structural patterns of materials such as metal-organic frameworks, covalent organic frameworks, polymers, hydrogen-organic frameworks, intermetallics, nets of noncovalent interactions, etc.
• Advanced synthetic techniques for obtaining chemical compounds with predetermined extended (periodic or polymeric) structure.

Submit your idea

Impact

Frontiers in Chemistry

Submit your idea

Impact