Alessandro De Vita

Methods Development: multi-scale techniques for atomistic modelling, information-efficient force-learning ("Learn On The Fly" - LOTF) algorithms for molecular dynamics simulations on massively parallel computers. Ab initio and classical atomistic techniques, coarse graining techniques and Monte Carlo techniques to model the self-assembly of supramolecular nanostructures.

Systems/Problems Investigated: Materials Chemomechanics phenomena e.g., brittle fracture in covalent crystals and ionic materials, crack propagation instabilities and interaction with dislocations and impurities; fracture in rocks, stress corrosion phenomena, defect and fracture at oxide interfaces; dislocation motion in advanced metal superalloys, hydrogen embrittlement of steels. Electronic structure and charge-transfer behaviour in metal-organic interfaces, supramolecular selfassembly on metal surfaces, electronic structure and chemical processes in adsorbed molecules. Structure and dynamics of DNA films for nanomedicine diagniostic sensors. Fuels, lubricants and additives of industrial interest. 

Development and application of multi-scale "learning" techniques for molecular dynamics on massively parallel computers.

Materials chemomechanics, brittle fracture, stress corrosion and supramolecular self-assembly.