Principal Investigator Peter Harrowell Project g34

Department of Chemistry, Machine VP

University of Sydney

Co- Investigators Alison Casey

Department of Chemistry, University of Sydney


The Role of Molecular Flexibility in Stabilizing Translational Order in Liquid Crystals

This project has addressed the role of purely steric interactions (i.e. short range repulsions) in realistic molecular models in stabilizing layered or smectic liquid crystal phases. Specifically, we have investigated whether the presence of flexible and rigid sections in a range of smectic-forming molecules could stabilize the layered phase by packing statistics alone, i.e. without requiring additional long-range anisotropic interactions.


What are the basic questions addressed?

We have investigated whether the presence of flexible and rigid sections in a range of smectic-forming molecules could stabilize the layered phase by packing statistics alone, i.e. without requiring additional long-range anisotropic interactions.

What are the results to date and future of the work?

The simulations demonstrate, for the first time, that smectic phases can be stabilized by the purely entropic contributions which drives the partial segregation of portions of the molecules with different flexibilities. This result clearly has important implications for the phase behaviour of a wide range of complex molecules such as block copolymers. The dependence of the nematic-smectic phase transition on chain length and rigidity was examined. When the bond angles of the chain were constrained, we found that a tilted smectic, known as smectic C was observed. A correlation was demonstrated between the mean end-to-end length of the flexible groups and the entropic contribution of the chains. An approximate theory of the transition was examined taking advantage of this relationship. Finally, in the course of the preliminary work in this project, we found a new form of layering transition in a liquid of the rigid cores alone. This order, which we called discotic smectic, is characterised by the long molecular axis lying in the layers, instead of normal to them as is the case in the regular smectic A.

What computational techniques are used and why is a supercomputer required?

Given the dual computational demands involved in simulating flexible molecules (requiring the sampling of the large space of internal configurations) and establishing the nature of phase transitions, these calculations have been carried out on a 3D lattice in order to reduce the number of configurations. Monte Carlo sampling was used. While this simulation technique does not, typically, take real advantage of a vector processor, we have been able get around 70% vectorization of our algorithm and substantial speed ups. This is largely due to the nature of the slow part of the code, the exact counting of all possible internal molecular configurations for each centre of mass configuration. This central portion of the algorithm is completely vectorized.