|
The Molecular Simulations Laboratory is currently carrying out computer simulations to investigate membrane separation
processes (especially reverse and electro-osmosis for both liquids and gaseous solutions. A new algorithm has been developed by
us to study, at the molecular level, fluids in confined boundaries, such as semi-permeable membranes. Many issues
related to membranes can be resolved only if the molecular nature of the membrane is examined. Our research allows
us to address these issues. We have used the algorithm to succesfully obtain many macroscopic properties on systems
like aqueous and methanolic solutions of various electrolytes and many model systems.
The separation, concentration and purification of the chemical species present in a mixture is a major problem
in the chemical industry. Conventional mass separation techniques, in recent years, have been supplemented by a
class of processes which utilize semi-permeable membranes as separation barriers.
Separation in membrane processes are the result of differences in the transport rate of chemical species through
the membrane interphase. Reverse osmosis is one such process which can be used as a very efficient means for separation.
It can and has been used in water desalination and purification, waste-water treatment (especially water containing
chemical and radioactive contaminants, which are otherwise difficult to remove), as well as, many separation processes
related to food products and biotechnology.
The transport rate of different particles through the membrane is determined by the driving force on the components,
and their mobility and concentration within the interphase. The primary factor determining separation using reverse
osmosis is the size of the molecules and the molecular interactions of the components with the semi-permeable membrane.
Since these can differ significantly, they can be used as a means of separating comppounds that are otherwise difficult
to separate.
At present, the applications of membranes is still impaired by high costs and shortcomings in the membrane performance,
as well as the inability to predict accurately the degree of separation achieved. Although many theoretical methods
are available to explain and correlate experimental observations in reverse osmosis, they cannot be used to answer
some more fundamental questions. A more thorough understanding at the molecular level (using techniques like Molecular
Dynamics and Monte-Carlo simulation) is required to address issues like mass transfer and concentration polarization,
and their corresponding effects on membrane selectivity, permeability, and durability.
In the context of fluid state research, computer simulation has assumed an important role that complements both
theoretical and experimental studies. Modern theories for the properties of fluids attempt to calculate molecular
distribution functions for fluids whose molecules interact with pre-chosen intermolecular potential functions.
With the theoretical result for the distribution function, statistical mechanics is then used to average over appropriate
functions to obtain macroscopic properties of interest. Computer simulation allows us to obtain essentially exact
distribution functions for a particular potential model. Simulations are attractive because of the following reasons:
-
Theories can be tested directly by comparing simulation results with theoretical predictions using the same
potential models in both theory and simulation.
-
Studies of intermolecular potential functions for real fluids can be performed.
-
Calculation of the properties and behaviour of simple model fluids (hard sphere fluid,
Lennard-Jones fluid,
hard diatiomics etc.) can be cariied out. Such data serves to guide theoreticians by revealing those microscopic
features of a fluid that give rise to its macroscopic behaviour.
-
Study of the behaviour of matter under extreme or unusual conditions that may be difficult to realize in the
laboratory can be achieved with ease. Also, the number of experiments required to fully explore a system is not
constrained by capital cost required.
|
