ESPCI web site
Molecular Simulation : A New Way to Probe Phase Transitions in Soft Matter
Armand Soldera
Laboratoire de Physico-Chimie Moléculaire / Laboratory of Physical Chemistry of Matter, Département de chimie – Faculté des sciences, Université de Sherbrooke, Sherbrooke (Québec), J1K 2R1, Canada
Abstract
In recent years, an alternative way of doing science has emerged. With the increase in computational power and algorithm efficiency, numerical experiments on the frontiers of theory and experience are about to become a crucial part of a laboratory. In this lecture, the focus is made on molecular simulation approach. Its great advantage is that it allows to probe physical properties at the atomistic level. With this new insight, the molecular mechanisms underlying a macroscopic property can be grasped in a different way. New compounds with increased performance can also be proposed. However, its main challenge lies in establishing the link between the simulation performed at the molecular level, thereby using a limited number of atoms, and the macroscopic properties with its Avogadro number of atoms. This link is far from straightforward. Distinctive approaches must thus be proposed. To illustrate this, two kind of phase transitions in soft matter are envisioned in terms of molecular simulation. 1) Transitions between liquid crystal phases are first regarded. Thanks to the combined use of simulation and experimental phase diagrams, the deep connection between atomic structure and the very existence of the Smectic C phase can be revealed from an energetic viewpoint. 2) The glass transition in polymers is also considered in this lecture. In this case, the fact that the cooling rate used to get the simulated glass transition temperature (Tg) is extremely rapid, in order of 1011 more rapid than the experimental one, led us to compare atomistic simulation with overcranking used by high-speed cameras to reveal slow-motion. We argue that this analogy paves the way to interesting opportunities in the description of the glass transition from an atomistic viewpoint. More specifically, the transition from the amorphous state to the glass phase may be detailed in terms of the degrees of freedom freeze. We thus establish a relation between local dynamics and the classical dihedral potential energy diagram of a carbon-carbon bond. A direct link is proved between the ensuing activation energy and Tg, leading to a chemical vision of the glass transition.