Surface-constrained all-atom solvent

Hwang et al. studied the energetics and dynamics of the Sn2 reaction in water by a combination of the EVB method and free enei y perturbation calculation.33 In this study, the solvent molecules were treated by the surface-constrained all-atom solvent (SCAAS), while the reactants were described by the EVB. These investigators employed an elaborate mapping function to drive the reactant state (e,) to the product state ( 2). The free energy of activation... 

G. King and A. Warshel, /. Chem. Phys, 91,3647 (1989). A Surface Constrained All-Atom Solvent Model for Effective Simulations of Polar Solutions. 

A drastic reduction has been introduced by King and Warshel with their SCAAS (surface constrained all atoms solvent) model. In this model the space is divided into three regions by two concentric spherical surfaces. The outer region is treated as a continuum dielectric, the inner sphere contains M and some solvent molecules (water), the intermediate layer has a thickness sufficient to contain a layer of solvent molecules. Water molecules in both internal regions are described as dipoles. The position and the orientation of such dipoles describe a solvent coordinate Q, which is coupled to a spatial coordinate R to describe a reaction in a 2D space. MD simulations are used to define the time-dependence of both coordinates (Q(r)> and / (/)). This model introduces new and important features in the area of QM/MM methods, but little attention is paid to boundary conditions. 

The particular model used in the original simulation i- of this reaction was that of a Cl + CI2 like reaction as modeled by a LEPS potential energy surface. The barrier for this symmetric reaction was normally taken to be 20 kcal/mol (—33 kT at room temperature). Other simulations used 10 and 5 kcal/mol barriers. The reactants were placed in either a 50 or 100 atom solvent (Ar in the earliest simulations Ar, He, or Xe in the later work) with periodic truncated octahedron boundary conditions. To sample the rare reactive events, as described previously, this system was equilibrated with the Cl—Cl—Cl reaction coordinate constrained at its value at the transition state dividing surface (specifically, the value of the antisymmetric stretch coordinate was set equal to zero). From symmetry arguments, this constraint is the appropriate one (except in the rare case where the solvent stabilizes the transition state sufficiently such that a well is created at the top of the gas phase barrier). For each initial configuration, velocities were chosen for all coordinates from a Boltzmann distribution and molecular dynamics run for 1 ps both forward and backward in time. 

---