The molecular dynamics methodology

In reality, atoms and molecules in solid materials are far from static unless the temperature is low but even at 0 K, vibrational motion remains. For ionically conductive materials, atomic movement is the subject of major interest. allows us to simulate the dynamics of the particles in a well-defined system to gain greater insights into local structure and local dynamics - such as ion transport in solid materials. In an MD simulation, atomic motion in a chemical system is described in classical mechanics terms by solving Newton s equations of motion  

The basic idea of MD goes back to a classical Newtonian idea in physics - that if one knows the location of all the particles in the universe, and the forces acting between them, one is able to predict the entire future. In a normal MD simulation, this universe comprises only some thousands of atoms - or in extreme cases, millions - gathered in an MD simulation box. 

With Newton s equations, it is possible to calculate sequentially the locations and velocities of all particles in the system. This generates a sequence of snapshots which constitutes a movie of the simulated system on the atomic scale. Owing to the massive computer time necessary to solve these equations for a large number of particles, the movies are generally fairly short - usually within the nanosecond regime. All that is needed to solve the equations of motion are the masses of the particles and a description of the potentials, E. 

The solution of this set of equations is managed by a computer algorithm. The most common is the so-called leapfrog , which works stepwise by  

Since the analytical description of the potential energy surface, the force field, strictly determines the outcome of any MD simulation, it is necessary that this description is as precise as possible. The common methodology is thus to generate specific potentials for the simulated system. These can be generated and fine-tuned in two different ways empirically or non-empiricaUy. 

---

Liquids are difficult to model computationally because the individual molecules (or ions) that make up the liquid are not isolated (as in the gas phase), but are interacting with each other. These interactions are not symmetric and static (as in solids) but are randomized and dynamic. Thus, errors are introduced when attempts are made to extract a portion of the liquid for study. This problem can be maneuvered around by taking a large portion and by controlling the boundary conditions of the sample. The techniques for extracting a sample and for treating the boundary of a sample are well established within the molecular dynamics methodologies. 

---