|
The Molecular Simulations Laboratory is currently carrying out computer simulations to investigate
different kinds of phenomena. Methods used to simulate matter range from a completely deterministic approach (in which all the molecules in the
system obey Newtonian mechanics) like Molecular Dynamics (MD), to the other extreme of the spectra; a stochastic
approach (random motion of the molecules based only on a probability distribution) like the Monte-Carlo (MC) method.
In MD simulations equations of motion of each particle (or molecule) are solved to trace the time evolution of
(usually) classical systems. Static (thermodynamic) and dynamic (transport) properties can then be obtained by
time-averaging appropriate statistical mechanical expressions. MC is based on the ensemble concepts of Gibbs which
employs random numbers. Since it is based on Markov-chain ensemble averaging concepts and does not treat real-time
motion, it cannot be used directly to study dynamic problems., although several indirect techniques have been developed
for using the MC method to solve dynamic problems.
We have used the above techniques to study transport properties like diffusion coefficients, permeation ratios,
separation factors, viscosity and thermal conductivity of the bulk solution, as well as those of the individual
components (solute and solvents). Other static properties like density profile, osmostic pressure etc. can also
be evaluated. In addition, our work has also resulted in developing a simple method for determining the activity
coefficients of liquids. Many of our results have been corroborated with experimental and theoretical models developed
by other groups. Our results should help manufacturers design more efficient membranes.
The program coding has been done using FORTRAN 77 and FORTRAN 90 modules. A variety of machines have been used
for compiling and executing the programs, including RS/6000, HP/9000, and SUN/Sparc workstations, IBM-CM/VMS mainframe,
and CRAY-YMP, CRAY-C90 supercomputers.
The projects conducted have been supported by grants from Petroleum Research Fund
(PRF), National
Science Foundation (NSF), NATO, the Department of Energy
(DoE). We actively conduct joint-research with the University
of Kent (UK) and Cornell Theory Center (USA).
We are also part of a newly established international consortium on "Conceptual
Learning of Science" (COLOS). COLOS has the objective of changing the way science is taught at the Undergraduate
level. We plan to develop computer modules using computer simulation that explain the basic postulates of thermodynamics
on a rudimentary molecular perspective. The COLOS Project is funded by Hewlett-Packard,
and other agencies are currenly being sought, to continue the same.
|
