Revisiting the Hunter-Sanders Model for π–π Interactions

The “Hunter-Sanders model” (J.
Am. Chem.
Soc. 1990, 112, 5525) is foundational to many
chemists’ understanding of interactions between aromatic systems.
Carter-Fenk and Herbert (Chem. Sci., 2020, 11, 6758) recently upended that understanding by showing that
the driving force for aromatic systems to adopt parallel displaced
geometries arises from steric, not Coulombic, repulsion of the π-electron
clouds. Carter-Fenk and Herbert also claimed to show that the original
Hunter-Sanders potential fails to predict the geometries
of a range of parallel and T-shaped dimers. Closer inspection reveals
that the data supporting this latter claim are flawed. Correctly implemented,
the Hunter-Sanders potential provides qualitatively correct predictions
for these systems and performs particularly well for the T-shaped
benzene dimer. Moreover, it predicts the preferred displacement direction
for some stacked heterocyclic dimers and accurately captures the impact
of a diverse group of substituents on the benzene sandwich dimer.
This is inclusive of the fact that all substituents enhance stacking
interactions in this geometry. Ironically, for substituted benzene
dimers, the Hunter-Sanders potential provides data in accord with
our Local, Direct Interaction Model but in contrast with the so-called
“Hunter-Sanders model.” At the same time, the Hunter-Sanders
potential struggles to capture heteroatom and substituent effects
in parallel displaced geometries in which the heteroatom/substituent
is located over the other ring, leading to qualitatively incorrect
predictions of the preferred displacement direction of substituted
benzene dimers. Overall, many aspects of the Hunter-Sanders potential
are flawed; however, others appear qualitatively correct.