molecule. Continuum solvation, implicit solvation, places the solute molecule in a
cavity in a continuous medium which simulates the sea of solvent molecules. There
is also a hybrid method, microsolvation with continuum solvation.
Microsolvation, explicit solvation. This is called explicit because computation-
ally, individual solvent molecules are placed around the solute molecule. “Sur-
rounding” the solute with solvent molecules might be putting it too strongly,
because, at least with routine quantum-mechanical calculations, few molecules of
solvent are used, typically about one to ten. In experimental reality, a solvent
molecule is surrounded, depending on its size, by a first solvation shell of about
six (for a monatomic ion) to probably hundreds or thousands of water molecules for
a protein or nucleic acid molecule. The first solvation shell is in turn solvated by
what we could call a second shell, and so on; when to cease considering the solvent
molecules can be problematic [ 11 ]. Actually, solvation calculations on large bio-
molecules in a bath of a large number of explicit water moleculeshavebeen
reported. These are typically done with molecular dynamics, which is outside the
scope of this book [ 12 ], using molecular mechanics force fields; such calculations
have been reviewed [ 13 ]. In our presentation of explicit solvation we will concen-
trate on quantum mechanical calculations in which a few solvent molecules are
used – literally microsolvation. Two examples of this technique are discussed:
1.The effect of microsolvation on the E2 and SN2 reaction F"þC 2 H 5 FþnHF.
Bickelhaupt et al. used DFT to study the reaction of fluoroethane with fluoride
ion, solvating the reactants with from zero (gas phase) to four hydrogen fluoride
solvent molecules [ 14 ]. HF is an unusual solvent, and presumably was chosen
rather than water because of its geometric simplicity and because it is, like
water, protic, although an HF molecule can hydrogen bond to only one acceptor
at a time. A virtue was made of the “artificiality” of the HF/F"acid/base system:
that HF is much more acidic than water and that fluoride with their basis set is
“artificially strong” was said to “lead to pronounced effects of solvation, facil-
itating interpretation.” These authors clearly recognized that a microsolvated
system of their type is not really a well-simulated condensed-phase system:
solvent molecules are rationed in the former. The purpose of the computations
was to obtain a qualitative understanding of the effect of solvent on these
synthetically useful reactions. One deficiency of microsolvation here was that
an unsolvated fluoride ion tended to be ejected, since with a limited number of
solvent molecules transfer of HF molecules from the attacking F"to the forming
F"was not favored, and this raises the activation energy. In “real solvation”,
which might be called macrosolvation, there is an abundance of solvent mole-
cules and all species can be adequately solvated.
Nevertheless, important features of real solvent reactions were reproduced by
microsolvation. The role of ion-molecule complexes, important in the gas phase,
decreased rapidly with introduction of solvent molecules, the reaction profile
becoming nearly unimodal (see Continuum solvation, below). Activation energies
for both E2 and SN2 processes increased due to stronger solvation of reactants
than of transition states (although in this work, because of imposed geometric
8.1 Solvation 523