Principal Investigator Denis J Evans Project r23

Research School of Chemistry Machine CM

Co-Investigator David P Hansen

Research School of Chemistry

The Study of the Process of Gelation Using Parallel Molecular Dynamics Programs

The aim of this project is to simulate a suspension undergoing gelation, and to study the effect of an applied shear on this process. Aggregation and gelation of polymer liquids have been studied experimentally in some depth. There have also been several types of computer simulation techniques, including Monte Carlo methods, percolation and random walk type algorithms used to study this process. To date there have been no Molecular Dynamics studies of this process.

We hypothesise that gelation in colloids is a manifestation of spinodal decomposition. When a stable suspension `gels' the nature of the colloid interaction potential suddenly. changes (This is typically achieved by changing the pH of the solution). The interaction becomes slightly shorter ranged but much more strongly attractive. In thermo-dynamic terms, the system behaves as though it had been rapidly quenched to a low reduced temperature, typically below the triple point temperature of the changed system. Gelation is therefore an unusual form of spinodal decomposition. It is unusual because the quenched state lies inside the gas-solid coexistence region rather than the usual gas-liquid spinodal decomposition.

We are using nonequilibrium molecular dynamics (NEMD) to examine these processes. We are examining the effect which an applied shear has on the structure and dynamics of the processes. We are also studying the more usual gas-liquid spinodal decomposition. It should be possible to characterise the fractal dimension of the fluid and the solid.

These simulations involve the solving of non-Newtonian, thermostatted SLLOD equations of motion for a number of particles in a system. The forces between each particle are calculated every timestep and the equations of motion then solved to calculate the new position and momentum of each particle. Thermodynamic and transport properties of the system can then be calculated. We are continuing to develop potential functions and new molecular dynamics techniques in order to model the gelling process.

This project is closely coupled with experimental neutron scattering experiments of simple real suspensions being carried out at the NIST facility by Hanley and Straty. Experiments have also been carried out on the effect which shear has on these suspensions when the suspension is undergoing a gelation process. Hence, it is hoped that our simulations can be used to gain a detailed microscopic understanding of the major qualitative features of gelation (including shear induction of gelation).

What are the basic questions addressed?

What is the microscopic basis of the sol-gel transition in colloidal suspensions?

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

We are comparing experimental neutron diffraction results against our computer simulations. We have advanced a new hypothesis for the transition namely that it represents a peculiar form of spinodal decomposition, from a dense fluid into the gas/solid coexistence region of the phase diagram.

We need to simulate larger systems over a wider range of conditions to enable better comparisons with experiment.

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

Molecular dynamics techniques are used to solve Newtonian equations of motion under nonequilibrium planar Poiseuille flow.

The simulations involve numerical solutions to systems of equations which involve up to several thousand particles. Each run involves about 100 000 timesteps, and several runs are required for a sufficient set of results to be compiled. Thus it is essential to have powerful and fast computing to perform the many calculations involved.


A Generalised Heat flow algorithm, David P Hansen and Denis J Evans, Molecular Physics, 81, 767-779 (1994).

A Parallel Algorithm for Nonequilibrium Molecular Dynamics Simulation of Shear Flow on Distributed Memory Machines, David P Hansen and Denis J Evans, Molecular Simulation, 13, 375-393 (1994).

Thermal Conductivity of the Two dimensional Soft Disk Fluid, David P Hansen and Denis J Evans, Molecular Simulation, to appear.

Response Theory Analysis of a Thermodynamic Quench, Rebecca K Schmidt and Denis J Evans, Molecular Physics, 83, 9-17.