 |
Center for Nonlinear Studies |
Computational design of materials to meet specific technological applications is gaining an increasing importance. However, the ability to compute and screen a large number of compounds satisfying specific targets, but decoupled from the ability to process and understand the obtained data, is not necessarily leading to the discovery of new materials with improved features. Great benefits could be obtained from exploiting computational tools along with concepts coming from basic sciences to manage one-by-one small sets of calculations under the guidance of a well-trained and intuitive practitioner. By means of computational modeling I propose a series of new compounds with appealing features, ranging from oxygen deficient oxides to tunable band gap in graphitic materials with Dirac cones, demonstrating that they can be obtained upon careful inspection of the physical and chemical properties of the virtual sub-products. Assisted by machine learning algorithms, complex set of equations involved in model materials properties description are solved with great accuracy.
Host: Turab Lookman