Molecular Dynamics Simulation of Confined Glass Forming Liquids
- PDF / 258,174 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 59 Downloads / 210 Views
Molecular Dynamics Simulation of Confined Glass Forming Liquids Fathollah Varnik1 , Peter Scheidler1 , J¨org Baschnagel2 , Walter Kob3 , Kurt Binder1 1 Institut fu¨ r Physik, Johannes Gutenberg-Universit a¨ t, Staudinger Weg 7, D–55099 Mainz, Germany 2 Institut Charles Sadron, ULP, 67083 Strasbourg, France 3 Laboratoire des Verres, Universit´e Montpellier II, 34000 Montpellier, France ABSTRACT Two model studies are presented in order to elucidate the effect of confinement on glass forming fluids, attempting to study the effects of the interactions between the confining walls and the fluid particles. In the first model, short bead-spring chains (modelling a melt of flexible polymers) are put in between perfectly flat, structureless walls, on which repulsive potentials act. It is shown that chains near the walls move faster (in the direction parallel to the walls) than chains in the bulk. A significant decrease of the (mode-coupling) critical temperature with decreasing film thickness is found. In the second model, a binary Lennard-Jones liquid is confined in a thin film, whose surface has an amorphous structure similar to the liquid. Although, as expected, the static structural properties of the liquid are not affected by the confinement, relaxation times near the wall are much larger than in the bulk. Consequences for the interpretation of experiments are briefly discussed. MOTIVATION Understanding the glassy state of condensed matter and in particular the transition that leads from a supercooled fluid to the amorphous solid is one of the greatest challenges of our time. A particularly controversial concept is the idea of a characteristic “correlation length” that grows as the glass transition is approached. This length is supposed to measure the size of the regions over which cooperative structural rearrangements need to occur to allow the system to relax [1]. The concept of this length scale was a motivation to investigate how the glass transition in thin polymer films depends on their thickness D [2–5]. Here one made use of the key idea of finite size scaling at ordinary static (second order) phase transitions, that “D scales with ” [6], i.e., a shifted transition is reached when has grown to about the size D= . The first measurements (by ellipsometry) [2] could be done only for rather large molecular weights, and found a decrease of the glass transition temperature Tg for D nm. Although in the meantime there are many more experiments [3–5], many details are still unclear. An intriguing feature clearly is that for long entangled chains the motion near a surface may differ from the one in the bulk [7]. Due to the complexity of the problem it is desirable to separate the problem of the glass transition from another very difficult and incompletely understood problem, namely of whether and how polymers “reptate” [8]. This can be done by studying the glass transition of short polymer chains that are not entangled and which can be studied very well with computer simulations [9]. A further advantage of such si
Data Loading...
