Chemistry Reference
In-Depth Information
encompassing all cases of atomic linkage, from a gas-phase water molecule to a
complex organometallic crystal? We assume here that chemical bonding is but one
effect of the electron distribution in the system and that there is no fundamental
distinction between intramolecular and intermolecular bonds.
Often, a chemical bond is unanimously recognized, but strange as it may seem,
there is no definition of a chemical bond that may go unique and unchallenged
throughout the chemical world. Are we chemists dwelling on an evanescent defini-
tion of the backbone concept of our science? In the current structural literature some
atomic pairs in crystals are a priori treated as chemical bonds, apparently without
the need of a definition, or of restrictions and conditions. Of particular concern is
the tendency to designate as a chemical bond any proximity of atom pairs on
subjective choice, under the assumption that the strength of a chemical bond can
be a continuum down to almost zero. Along with an oxymoron, one sees a logical
flaw and the danger of fostering more confusion than understanding.
The following is a tentative description of a chemical bond in terms of a few
basic principles:
1. A chemical bond is a localized phenomenon that brings about a stable linkage
between two atomic nuclei. A chemical bond must result from the interaction
between the electron distributions in the basins close to the two involved atomic
nuclei. The identification of the bond must not be sensitive to changes of the
chemical environment (e.g., by induction effects).
2. A chemical bond produces proximity of the two involved atomic nuclei, whose
separation must be less than the sum of commonly accepted random-contact
radii (if there is arguing about the exact value of the contact radii then the
connection is not a chemical bond). This is a necessary but not sufficient
condition. Especially in the intermolecular case, proximity between two nuclei
does not always imply a chemical bond.
3. A chemical bond is represented by a dissociation curve, as a function of separa-
tion between the two involved nuclei, with a well defined energy minimum at
R
with depth
E
*, the bond dissociation energy. Such a curve can be prepared by a
quantum chemical calculation on model molecular systems designed so that the
conditions set in points (1) and (2) are fulfilled.
E
* must be much larger than
3RT and the effect of a small change in internuclear separation must be signifi-
cantly higher than RT. The bond stretching vibration must be harmonic to a very
good approximation at a significant distance from the minimum.
In organic condensed media, and especially in crystals, one sees quite often some
recurring, close proximity of certain atom types. These phenomena are often called
chemical bonds, but it is easily checked that they do not conform to some or all of
the above requirements, mostly because the corresponding interaction energy curves
are too shallow. These preferences should not be called chemical bonds because
they are prone to shifting, alteration, or even sweeping out, upon minor changes in the
environment. Consider a Boltzmann distribution of molecular configurations in the
equilibrium liquid-state precursor, either a melt or a solution. Some configurations
(e.g., those in which a positively charged CH fragment is close to a negatively
