Electrostatic computer modeling software

D. B. Boyd, in Reviews in Computational Chemistry, Vol. 4, K. B. Lipkowitz and D. B. Boyd, Eds., VCFI Publishers, New York, 1993, pp. 229-257. Compendium of Molecular Modeling Software. UHBD is available from Molecular Simulations Inc. DelPhi is available from BIOSYM. MEAD is described by D. Bashford and K. Gerwert, J. Mol. Biol., 224, 473 (1992). Electrostatic Calculations of the pK, Values of lonizable Groups in Bacteriorhodop-sin. MEAD can be obtained by anonymous ftp from scripps.edu (192.42.82.27) or from the author, bashford scripps.edu. 

Computational chemistry takes model making to a yet higher level. Most modeling software also incorporates programs that identify the most stable geometry of a molecule by calculating the energies of possible candidate structures. More than this, the electron distribution in a molecule can be calculated and displayed as described in the boxed essay Electrostatic Potential Maps earlier in this chapter. 

In all simulations of clay mineral systems we apply periodic boundary conditions at constant pressure and temperature (constant NPT), This allows the system volume to change freely at 100 kPa (1 bar) external pressure and 298 K. Furthermore we employ Ewald summation to compute both electrostatic potentials and dispersive van der Waals interactions, and the simulations are fully dynamic, using the Discover module and Insight II graphical user interface of the MSI molecular modeling suite (MSI, 1997). The free energy perturbation technique is not implemented in this software per se so that many of the aforementioned calculations have to be performed with spreadsheet software (e.g., Microsoft Excel). 

As we learn about physical chemistry, we extend that understanding to new systems of interest, but the transition from simple molecules to more complicated systems is limited by our human abilities. We can imagine a butane molecule and imagine both how and why different conformers are preferred. When we tackle pentane or combinations of functional groups, however, it becomes difficult to simultaneously consider the relative effects of electrostatics, steric interactions, and bulk properties, such as solubilities. Fortunately, we can construct models in software that allow us to properly treat all of these interactions. The scientific field of molecular modeling or. more generally, computational chemistry is the practice of simulation of molecular systems with sufficient detail to address a question of interest. 

In MM, a molecule is modeled as a collection of masses (atoms) and springs (bonds), with additional forces added to describe other interactions such as hydrogen bonding, electrostatics, and dispersion forces. Although such simulations have been done using carefully constructed mechanical models [22], MM has been most successfully implemented computationally. The present discussion will focus on the MM3 method [13], since it is popular and is implemented in a number of software packages. Beware that not all implementations of MM3 provide thermochemical information. 

The choice of models is, at present, largely author dependent and not much influenced by an analysis of the published information. Sometimes, the available computer software restricts users to certain models (e.g., the 2-pK formalism is much easier to apply with the common codes than the 1-pK formalism). As is clear from the above statements, a general decision for a certain model would only be possible by disregarding opinions of certain authors. The choice of a model should therefore, as indicated in Section I, be largely associated to the objective of the modeling exercise and be considered in that way. If electrostatic features are of no interest whatsoever in a certain system, then why not use a nonelecfrostatic model However, in case a mechanistic interpretation of ion adsorption is the objective, then it would be difficult to understand why electrostatic interactions should be disregarded. 

---