|Severin, , Nikolai: Molecular Dynamics Simulations of Polymers and Micelles at Interfaces |
The aim of this work is the atomistic molecular dynamics (MD) simulation of (1) polyethylene - isotactic polypropylene interfaces and (2) cylindrical and half-cylindrical micelles of amphiphilic molecules at solid-liquid interfaces.
Molecular dynamics solves the classical equations of motions for a system of N particles interacting according to a potential energy field. In general the particles may represent different physical objects like satellites in the gravitation field of earth or atoms of a molecular system. The most important capability of MD is that during the dynamics simulation a system undergoes conformational and momentum changes so that different parts of the phase space accessible to the simulated molecular system can be explored. By providing mechanisms for controlling the temperature and pressure of the simulated systems, molecular dynamics also allows to generate statistical ensembles from which various energetic, thermodynamic, structural and dynamic properties can be calculated.
The drawback of MD simulation is the limitation in the time span which can be simulated and on the system size i.e. the number of particles and interactions involved in the calculation. The fast progress in the development of computers allows to simulate larger systems with a more detailed description of inter-particle interactions. However for the detailed description of atomic interactions the number of atoms included in a simulation is still restricted to several tens of thousands. Simulated times of atomistic molecular systems are limited by several nanoseconds, which is related to the fast vibrations of chemical bonds. These are very severe restrictions, on the other hand many molecular systems of particular interest can be simulated. For example it is possible to simulate the dynamics behaviour of small bio-molecules, which is important to understand processes in the cell. A proper simulation of the environment of a molecular system allows to increase greatly the range of applications for MD simulations.
It is possible to simplify molecular system in order to increase the time span of MD simulation. The simplifications may range from simple unification of group of atoms to more general simplifications. For example the methyl group is replaced by single particle with potentials and dipole moment that should represent the methyl group. This simplification allows to increase the number of particles involved in the simulation and increase a little bit the simulation time step. But the information about the coordinates of all atoms is important for the short range interactions. The example of more general simplification is representation of a polymer molecule as a flexible rubber rod, without detailed consideration of atomic structure. This simplification allows to increase greatly the time span of MD
2simulation, it allows even to simulate some long time processes like adsorption of molecules on a solid-liquid interface or process of micro/macro phase separation.
In addition to MD method there are several other computer simulation methods that allow simulation of molecular systems. In the Brownian Dynamic (BD) method the solvent for example is represented as a random forces on atoms. It is also important to mention the Monte Carlo (MC) simulation method. In general it was developed to predict equilibrium distribution of a system. It has several advantages and disadvantages. In many cases MC method samples conformational space faster than MD method but in general it is not designed to provide dynamic information on the system.
The restriction to several nanoseconds of the simulated time renders atomistic MD simulations dependent on the starting configuration of a system. Hence, atomistic MD simulations are bounded to experimentally established structures. On the other hand many experiments provide just a general picture of the molecular conformation and do not provide atomic coordinates and in its turn MD simulation may provide information on the atomic coordinates of a molecular system. That is why MD simulations are a powerful means for the investigation of structure and short time dynamics of molecular systems in combination with experimental techniques.
DiML DTD Version 2.0||
der Humboldt-Universität zu Berlin
|HTML - Version erstellt am:|
Thu Feb 24 14:14:32 2000