Implicit solvent models Subject

With this work Born took the step of using atomic concepts and parameters, mixing them with continuum ideas (implicit solvent models) to make correlations with bulk thermodynamics. Not only was this a successful calculation but it remains a common theme in much current work on the subject some 80 years later. Born s theoretical understanding still underpins the field today, whether approached by many-body approximations [6] or by computer simulations [7]. 

These observations are not new. A similar electronic correlation mechanism has been proposed as explanation for the underestimation of charge state transition energies of point defects in semiconductors. Further support for the interaction with the solvent band structure is provided by the comparison to implicit solvent models, which omit such interactions. Errors in the calculation of reduction potentials are significantly smaller for the very same GGA functionals (see, for example, [14]). The unfavorable comparison to implicit solvent models raises the question why allatom methods are used if the huge increase in computational costs only leads to deterioration of accuracy. The answer must be that there are situations where an allatom approach is required. The reactivity and transport of excess electrons in water almost certainly involve conducting states as intermediates. Localization of holes in water is still subject to debate but could possibly play a role in transport and reaction kinetics. Electrochemical interfaces are another example of systems for which interactions between localized and extended states are important. The DFTMD/... 

Equation (3.21) shows that the potential of the mean force is an effective potential energy surface created by the solute-solvent interaction. The PMF may be calculated by an explicit treatment of the entire solute-solvent system by molecular dynamics or Monte Carlo methods, or it may be calculated by an implicit treatment of the solvent, such as by a continuum model, which is the subject of this book. A third possibility (discussed at length in Section 3.3.3) is that some solvent molecules are explicit or discrete and others are implicit and represented as a continuous medium. Such a mixed discrete-continuum model may be considered as a special case of a continuum model in which the solute and explicit solvent molecules form a supermolecule or cluster that is embedded in a continuum. In this contribution we will emphasize continuum models (including cluster-continuum models). 

All the models above are finite in size and rely on explicit molecnlar additions subject to the same computational method as the species studied. Another way to take into account and model the real surrounding in an electrolyte is to use various continuum methods to implicitly mimic the effect of, e.g., the dielectric constant of the electrolyte. Popular since many years are different variants of the polarizable continuum methods (PCM) applicable to both ab initio and DFT methods and where parameters for a variety of different solvents exist, and with possibilities to tailor for special electrolytes. The use of continuum methods has demonstrated the importance of simulating solvent effects - especially the difference between the gas phase electronic energies and the free energies of solvation (AG) via PCM. The use of continuum methods can also be tweaked in various ways, e.g., in TD cycles to treat different dielectric constants for different parts of the cycle. 

---