• elastic models
• glass transition
• supercooled liquids

When they are cooled or compressed, several systems such as liquids, mixtures, polymers, biomaterials, metals, and molten salts may avoid the crystallization, resulting in a metastable supercooled phase. A full understanding of the extremely complex phenomenology in supercooled liquids is still missing. First there is the issue of how crystallization can be prevented and how deeply the liquid can be supercooled. However by far the most interesting feature of supercooled liquids is the glass transition (GT): following a huge increase in the viscosity as the temperature decreases, the liquid freezes into a glass, a microscopically disordered solid-like state. Understanding the
extraordinary viscous slow-down that accompanies glass formation is one of the major open challenges in condensed matter physics.

During my Ph.D. period (January 2009 - December 2011), I worked on several projects, all connected with the aim of understanding from microscopic basis the relaxation processes in glass-forming liquids. In the light of recent works, particular attention has been addressed to the connection between fast vibrational dynamics on picosecond time scales and the slow relaxation. The first part of my work has been devote to deepen some interesting aspects of this result and to discuss its implications on other aspects of the supercooled liquid phenomenology such as the diffusion and the violation of the Stokes-Einstein relation. Then I focused on the issue of the repulsive interactions controlling the static and dynamics in viscous liquids, and the related topic of the density-temperature scaling. In the last part of my research activity, investigated the elastic models of the GT, which relate the huge slowing down of glass-forming systems with the increasing solidity.

The study of supercooled liquids is approached here from a numerical point of view. Due to these huge potentialities, in the last years, computer experiments played an increasingly important role in glass transition studies. By performing Molecular Dynamics (MD) simulations, we were able to study the dynamics on the microscopic level and to collect informations on every observable of interest with quite a high level of precision, while the same process in experiments would require much more effort. MD simulations allow us to test the validity of theoretical models, as in the case of the elastic models, in a fully controlled environment. During all the study, we have maintained a close connection with the "real world", by comparing, whenever possible, MD results with experimental ones.

To study the complex glassy phenomenology, the chosen prototype of viscous liquid is the simple beads and springs model for polymeric chains. Polymers play a central role in several studies on the GT because of their natural inclination to disorder: in most cases a polymer liquid, rather than crystallize in a regular lattice, reaches the amorphous glassy state.