Chemistry Reference
In-Depth Information
Because of their widespread importance to so many areas of chemistry,
biology, and materials science, noncovalent
interactions are a crucial topic of
study. Moreover, a clear understanding of such interactions is indispensable
for the rational design of new drugs and organic materials. Unfortunately,
at present, many basic properties of noncovalent interactions—particularly
those involving
p
systems—remain unknown.
19
Experiments aimed at eluci-
dating the details of noncovalent interactions face a number of serious obsta-
cles. Often the chemical systems of greatest interest are complex ones (e.g., a
drug in the active site of a protein) where several different types of noncovalent
interactions may contribute simultaneously. Indeed, even in model systems
specifically designed to study a particular noncovalent interaction, it has often
proven challenging to separate the interaction of interest from unexpected
secondary interactions or solvent effects.
20-22
Gas-phase studies of bimolecular complexes afford more control, but
these can be quite challenging. Most noncovalent interactions between
small molecules feature small binding energies (on the order of a few kilo
calories per mole or less), often meaning that very low temperatures must
be used to avoid thermal dissociation of the complex. Additionally, it is fre-
quently necessary to use mass selection techniques to ensure that the sample
contains only the complex of interest and not larger or smaller clusters.
Furthermore, the potential energy surfaces for these systems tend to be
fairly flat, meaning that the complexes may be fluxional without a well-
defined structure. If the potential surface features two or more potential
minima, conversion between them will be easy and rapid except at very
low temperatures.
Due to these experimental difficulties, there are great opportunities for
the computational chemist to answer important questions about the funda-
mental nature of noncovalent interactions and how they influence particular
chemical systems. A significant advantage of computational studies is that
one can directly study prototype systems featuring the noncovalent interaction
of interest, in the absence of competing interactions or solvation effects. Com-
putational studies of noncovalent interactions have therefore become increas-
ingly popular over the past 5 years and have led to important insights. Until
now, these studies have been rather difficult to carry out because the most
commonly used computational chemistry techniques do not give reliable
results for noncovalent interactions. It is the purpose of this review to explain
how noncovalent interactions can be computed reliably using more robust
electronic structure methods, and then to describe what approximations
appear to be valid and helpful for speeding up the computations. Our focus
is specifically on
p
interactions, but in terms of which techniques are appropri-
ate to use, there are not large differences between these and other noncovalent
interactions. Thus, the review on weakly interacting clusters in Tschumper
23
is
directly relevant to the issues discussed here. For pedagogical reasons, we have
retained some overlap in the topics discussed in the two reviews.
p
