Symmetry Breaking by Solvation
A R T I C L E S
ion is calculated to stabilize an asymmetric structure.29 However,
the asymmetry is general, even in nonpolar solvents, such as
the CD2Cl2 here. Besides, a crystal is also a polarizable medium
with strong electric fields. We therefore had proposed that the
disorder of neighboring ions, especially in nonpolar solvents,
is what creates the asymmetry. Zwitterion 1 allows this proposal
to be rejected.
In summary, the observed isotope shifts of 1-18O (R ) butyl,
octyl) demonstrate that these zwitterions exist as a pair of
equilibrating tautomers, in both protic and aprotic solvents. This
is the same result as in many of our other studies of H-bonds.
It seems to be a general phenomenon that H-bonds are not
symmetric in solution, although we cannot exclude the pos-
sibility that symmetric H-bonds are a low-temperature phenom-
enon.30 Nevertheless, the asymmetry of the H-bonding in 1
shows that the previous cases of asymmetry cannot be attrib-
uted merely to the varying location of a counterion. Instead
the asymmetry seems to be a general feature of H-bonds in
solution.
Comments Regarding Solvatomers. Computer simulations
had suggested that interaction with an asymmetrically located
counterion can be responsible for stabilizing the asymmetric
H-bond, especially in nonpolar solvents.18 This interaction
cannot desymmetrize zwitterion 1, where the quaternary nitrogen
is on a symmetry axis. The experimental results therefore
suggest that asymmetry is inherent to all solutions, not through
the counterion, but through the disorder of interactions with
individual solvent molecules. In principle, the conformational
disorder of the alkyl chains might also contribute, but their
interactions are much weaker than those of the solvent molecules
that are closer to the carboxyls. Those solvent molecules are
continuously rearranging their dipole moments, so that the
instantaneous stabilization varies with time and with location.
Even a nonpolar solvent molecule, with no net dipole moment,
has local dipoles that reorient, or a polarizability that is
modulated, as the molecule tumbles around an H-bond. This is
in contrast to the organized environment found in crystals. The
disorder of solvation is a fundamental feature of solutions. It is
obvious, but has hardly been explored.
Although these results require a mixture of two tautomers,
rather than a single symmetric species, we cannot conclude that
the H-bond is described by a double-well potential. This may
be so, with the instantaneous solvation stabilizing one well more
than the other. The alternative is a single-well potential where
the solvation stabilizes an asymmetric structure, as in Figure 2.
(Another possibility, a double-well potential where the zero-
point energy lies above the barrier, is equivalent to a single-
well potential for this discussion.) As the solvation changes,
the hydrogen moves across the H-bond. In each of these
structures the H-bond is asymmetric, and the equilibrium
between them can be perturbed by isotopic substitution. These
structures can be called solvatomers, signifying isomers or
stereoisomers or (as here) tautomers that differ in solvation. This
is a more proper use of the term than an earlier designation
Figure 2. Equilibrating H-bond solvatomers, each with a single-well
potential describing energy vs bond-distance difference d(AH) - d(HB).
that was applied to species that differ in the type of solvent
molecules and thus are not isomeric.31
Comments Regarding the Role of Low-Barrier H-Bonds
in Enzymatic Reactions. We cannot exclude the possibility
that the active site of an enzyme is like a crystal, with an ordered
environment that permits a symmetric H-bond. Yet if the energy
of solvation is sufficient to prevent a symmetric structure, then
there is no special stabilization associated with them.
Although the strongest of H-bonds are symmetric, these two
characteristics do not necessarily parallel each other. One way
to lower the barrier to proton transfer and approach a symmetric
H-bond is to constrain the heavy atoms to proximity. Yet, it
must be recognized that this constraint does not strengthen the
H-bond but weakens it. If the constraint were relaxed, the species
would become more stable. Therefore, a symmetric or low-
barrier H-bond does not exhibit unusually high stability. Indeed,
we have suggested that enzymatic acceleration by low-barrier
H-bonds is due to relief of a destabilization, as had been
suggested by Jencks,32 rather than to any stabilization from the
H-bond itself.33
Implications Regarding the Generality of Solvation for
Breaking of Symmetry. A wide variety of situations have been
encountered where the local environment reduces symmetry.
One of the most familiar is in the theory of electron or proton
transfer, where reorganization energy must be provided to an
asymmetric system in order to achieve a symmetric configu-
ration that allows the electron or proton to transfer.34 A classic
example is NH3,35 where nitrogen inversion is subject to a
double-well potential. In the gas phase the nitrogen is delocalized
between the two wells. If it could be localized in one, it would
rapidly tunnel to the other. However, in any interacting solvent
the nitrogen is pyramidal, and the inversion barrier in substituted
derivatives can be measured.36
Another familiar set of examples depends on the selection
rules for IR and Raman intensities in centrosymmetric mol-
ecules. The paradigm of a symmetric H-bond is HF2-, but in
-
crystalline toluidinium HF2 the forbidden in-phase stretch
acquires IR intensity owing to a loss of centrosymmetry.37
Similarly, the bending mode and the antisymmetric stretch of
CS2, which are Raman inactive in isolation, become allowed in
the liquid.38 The Raman spectra of I3- solutions show transitions
(31) Gislason, B. I.; Strehlow, H. Aust. J. Chem. 1983, 36, 1941.
(32) Jencks, W. P. Catalysis in Chemistry and Enzymology; McGraw-Hill: New
York, 1969; pp 282ff.
(33) Perrin, C. L; Ohta, B. K. J. Mol. Struct. 2003, 644, 1.
(34) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta. 1985, 811, 265.
(35) Herzberg, G. Infrared and Raman Spectra of Polyatomic Molecules;
Molecular Spectra and Molecular Structure, Vol. II, Van Nostrand:
Princeton NJ, 1945; pp 221ff.
(29) Mavri, J.; Hodosˇcˇek, M.; Hadzˇi, D. J. Mol. Struct. (Theochem) 1990, 209,
421.
(30) Schah-Mohammedi; P.; Shenderovich, I. G.; Detering, C.; Limbach,
H. H.; Tolstoy, P. M.; Smirnov, S. M.; Denisov, G. S.; Golubev, N. S. J.
Am. Chem. Soc. 2000, 122, 12878. Shenderovich, I. G.; Burtsev, A. P.;
Denisov, G. S.; Golubev, N. S.; Limbach, H. H. Magn. Reson. Chem. 2001,
39, S91. Perrin, C. L.; Lau, J. S.; Ohta, B. K. Pol. J. Chem. 2003, 77,
1693.
(36) Saunders, M.; Yamada, F. J. Am. Chem. Soc. 1963, 85, 1882.
(37) Harmon, K. M.; Lovelace, R. R. J. Phys. Chem. 1982, 86, 900.
(38) Evans, J. C.; Bernstein, H. J. Can. J. Chem. 1956, 34, 1127.
9
J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006 11823