Handbook of Solvents - George Wypych - ChemTech - Ventech!

13.1 Solvent effects on chemical reactivity 745
found between the solvent reorganizational energy accompanying the excitation of ruthenium(II)
cyano complexes and the solvent acceptor number.
In the basicity scales, on the other hand, complications by solvent structure are not as
obvious. If restriction is to aprotic solvents, as is usual, various scales though obtained under
different conditions, are roughly equivalent. 21,4 There is for instance a remarkably good
relationship between the DN scale (obtained in dilute dichloromethane solution, i.e., with
medium effects largely excluded) and the B scale (derived from measurements performed
with 0.4 M solutions of MeOD in the various solvents 4 ). The relationship between β and B,
on the other hand, separates out into families of solvents. 20 Donor measures for protic solvents
eventually are hard to assess and often are at considerable variance from one scale to
another. 28,29 To rationalize the discrepancies, the concept of “bulk donicity” was introduced 7
but with little success. Instead, the consideration of structure changes accompanying solvation
might better help tackle the problem.
Another suspect feature of the common method of interpreting solvent-reactivity correlations
is that it is notoriously done in enthalpic (electronic, bond-strength etc.) terms.
This way of thinking goes back to the Hughes-Ingold theory. However, many reactions in
solution are not controlled by enthalpy changes but instead by entropy. Famous examples
are the class of Menschutkin reactions and the solvolysis of t-butyl halides. Both these reaction
types are characterized by the development of halide ions in the transition state, which
can be considered as ion-pair like. In view of this, rate acceleration observed in good acceptor
(or, alternatively, highly polar) solvents seems readily explainable in terms of solvation
of the developing halide ion with concomitant carbon-halogen bond weakening. If this is
true, most positive activation entropies and highest activation enthalpies should be expected
to occur for the poor acceptor solvents. However, a temperature dependence study of the
t-butyl halide solvolysis revealed just the opposite. 17 This intriguing feature points to
changes in solvent structure as a major determinant of the reaction rate with the ionic transition
state acting as a structure maker in poor acceptor solvents, and as a structure maker in
the protic solvents.
It is rather ironic that the expected increase in rate with increasing solvent acceptor
strength is a result of the coincidence of two, from the traditional point of view, unorthodox
facts: (i) The intrinsic solvation of the developing halide ion disfavors the reaction via the
entropy term. However, (ii), the extent of that solvation is greater in the poorly coordinating
solvents (providing they are polarizable such as the aromatic solvents and the
polyhalogenated hydrocarbons). In keeping with this interpretation, the Menschutkin reaction
between benzyl bromide and pyridine is characterized by more negative activation
volumina (i.e., stronger contraction of the reacting system in going to the activated complex)
in poor acceptor (but polarizable) solvents. 30 The importance is evident of studying
temperature or pressure dependencies of solvent effects on rate in order to arrive at a physically
meaningful interpretation of the correlations.
Another problem with the interpretation of multiparameter equations such as [13.1.2]
arises since some of the parameters used are not fully independent of one another. As to this,
the trend between π* and α has already been mentioned. Similarly, the δ H parameter displays
some connection to the polarity indices. 31,32 Virtually, the various parameters feature just
different blends of more fundamental intermolecular forces (see below). Because of this,
the interpretations of empirical solvent-reactivity correlations are often based more on intuition
or preconceived opinion than on physically defined interaction mechanisms. As it