Electrostatic interactions, computation

Once the models for the charge distributions are in hand, the electrostatic interaction is computed as the interaction between the sets of point charges or distributed nuiltipoles, and added to an atom-atom, exp-6 fonn that represents the repulsion and dispersion interactions. Different exp-6 parameters, often from [140. 

Pratt L R and Hummer G (eds) 1999 Simulation and theory of electrostatic interactions in solution computational chemistry, biophysics and aqueous solutions AlP Conf. Proc. (Sante Fe, NM, 1999) vol 492 (New York American Institute of Physics)... 

Procacci P, Darden T A, Paci E and Marchi M 1997 ORAC a molecular dynamics program to simulate complex molecular systems with realistic electrostatic interactions J. Comput. Chem. 18 1848-62... 

Paschual-Ahuir J L, E Silla, J Tomasi and R Bonaccorsi 1987. Electrostatic Interaction of a Solute with a Continuum. Improved Description of the Cavity and of the Surface Cavity Bound Charge Distribution. Journal of Computational Chemistry 8 778-787. 

The Born model is based on electrostatic interactions, dielectric permitivity, and orbital overlaps. It has the advantage of being fairly straightforward and adaptable to computational methods. The free energy for the polarization of the solute is expressed as... 

Molecular mechanics methods may work well or poorly for compounds containing alkali metals. The crucial factor is often how the force field computes charges for electrostatic interactions. 

As for the dielectric constant, when explicit solvent molecules are included in the calculations, a value of 1, as in vacuum, should be used because the solvent molecules themselves will perform the charge screening. The omission of explicit solvent molecules can be partially accounted for by the use of an / -dependent dielectric, where the dielectric constant increases as the distance between the atoms, increases (e.g., at a separation of 1 A the dielectric constant equals 1 at a 3 A separation the dielectric equals 3 and so on). Alternatives include sigmoidal dielectrics [80] however, their use has not been widespread. In any case, it is important that the dielectric constant used for a computation correspond to that for which the force field being used was designed use of alternative dielectric constants will lead to improper weighting of the different electrostatic interactions, which may lead to significant errors in the computations. 

The only problem with the foregoing approach to molecular interactions is that the accurate solution of Schrddinger s equation is possible only for very small systems, due to the limitations in current algorithms and computer power. Eor systems of biological interest, molecular interactions must be approximated by the use of empirical force fields made up of parametrized tenns, most of which bear no recognizable relation to Coulomb s law. Nonetheless the force fields in use today all include tenns describing electrostatic interactions. This is due at least in part to the following facts. 

Thermodynamic information can also be obtained from simulations. Currently we are measuring the differences in chemical potential of various small molecules in dimethylimidazolium chloride. This involves gradually transforming one molecule into another and is a computationally intensive process. One preliminary result is that the difference in chemical potential of propane and dimethyl ether is about 17.5 kj/mol. These molecules are similar in size, but differ in their polarity. Not surprisingly, the polar ether is stabilized relative to the non-polar propane in the presence of the ionic liquid. One can also investigate the local arrangement of the ions around the solute and the contribution of different parts of the interaction to the energy. Thus, while both molecules have a favorable Lennard-Jones interaction with the cation, the main electrostatic interaction is that between the chloride ion and the ether molecule. 

The structures of phases such as the chiral nematic, the blue phases and the twist grain boundary phases are known to result from the presence of chiral interactions between the constituent molecules [3]. It should be possible, therefore, to explore the properties of such phases with computer simulations by introducing chirality into the pair potential and this can be achieved in two quite different ways. In one a point chiral interaction is added to the Gay-Berne potential in essentially the same manner as electrostatic interactions have been included (see Sect. 7). In the other, quite different approach a chiral molecule is created by linking together two or more Gay-Berne particles as in the formation of biaxial molecules (see Sect. 10). Here we shall consider the phases formed by chiral Gay-Berne systems produced using both strategies. 

---