Halogen Bonding with Dihalogens and Interhalogens - Halogen Bonding: Fundamentals and Applications - page 96

culations with the Becke3LYP functional and LANL2DZ basis set as well as
the mPW1PW [184] functional. The latter adds diffuse and polarization func-
tions for halogen atoms to the LANL2DZ basis set [186].
The calculated optimized geometries were found to be in good agree-
ment with available experimental data. In general, the T-shaped hypervalent
structures were found to be of lower energy than the charge-transfer struc-
tures. This energy difference was greater with chlorine than bromine and
even smaller (or reversed) with iodine. Stronger electron donors also encour-
aged formation of hypervalent complexes. These results are consistent with
the Husebye model, where initial formation of the charge-transfer complex
isfollowedbyX-Ybondcleavagetogiveions(Fig.5pathways(a)and(c),
respectively).
The possibility that the hypervalent complex might be formed directly
from the starting fragments in an oxidative addition reaction (pathway (a ))
or by rearrangement of the charge-transfer complex (pathway (b)) was in-
vestigated using the negative eigenvalues progression reduction (NERP) [183]
and the intrinsic reaction coordinate (IRC) techniques [187]. Both methods
identified reasonable transition states for (b) (see, for example, adduct 6 ),
but not for (a ). In fact, attempts to find a saddle point for the conversion of
the starting fragments into the hypervalent complex inevitably identified only
the transition state for pathway (b), leading the authors to conclude that dir-
ect addition of the dihalogen to the electron donor is unfeasible. However,
pathway (b) is predicted to be lower in energy than the Husebye route (path-
ways (c) and (d)). The authors did note that solvent effects would be expected
to play a major role in the reaction.

3.5
π -Electron Donors

Other electron donor systems have received attention over the past decade.
In 1949, Benesi and Hildebrand observed an absorption band in solutions
of iodine in benzene that they attributed to a charge-transfer complex. The
structure(s) contributing to this band have been a matter of some debate ever
since. In this topic, the chapter of J.K. Kochi and coworker discusses these
adducts in the solid and in solution [185]. Figure 6 shows various geometries
that have been proposed. Grozema and coworkers examined the potential
energy surface of the benzene-I 2 system at the MP2 level, correcting for coun-
terpoise errors across the entire surface rather than only at energy minima, as
had been the case for earlier studies [188].