Chemical intuition suggests that anions and π-aromatic systems would repel each other. Typically, we think of cations as being attracted to electron-rich π-systems of aromatic rings, and the cation−π interaction, a well-established noncovalent interaction, plays an important role in nature. Therefore the anion−π interaction can be considered the opposite of the cation−π interaction. Computational studies of simple models of anion−π interactions have provided estimates of the factors that govern the binding geometry and the binding energy, leading to a general consensus about the nature of these interactions. In order to attract an anion, the charge distribution of the aromatic system has to be reversed, usually through the decoration of the aromatic systems with strongly electron-withdrawing groups. Researchers have little doubt about the existence of attractive anion−π interactions in the gas phase and in the solid state. The bonding energies assigned to anion−π interactions from quantum chemical calculations and gas phase experiments are significant and compare well with the values obtained for cation−π interactions. In solution, however, there are few examples of attractive anion−π interactions.

In this Account, I describe several examples of neutral molecular receptors that bind anions in solution either solely through anion−π interactions or as a combination of anion−π interactions and hydrogen bonding. In the latter cases, the strength of the anion−π interaction is indirectly detected as a modulation of the stronger hydrogen bonding interaction (enforced proximity). The dissection of the energy contribution of the anion−π interaction to the overall binding is complex, which requires the use of appropriate reference systems.

This Account gives an overview the experimental efforts to determine the binding energies that can be expected from anion−π interactions in solution with examples that center around the recognition of halides. The studies show that anion−π interactions also exist in solution, and the free energy of binding estimated for these attractive interactions is less than 1 kcal/mol for each substituted phenyl groups. The quantification of anion−π interactions in solution relies on the use of molecular recognition model systems; therefore researchers need to consider how the structure of the model system can alter the magnitude of the observed energy values. In addition, the recognition of anions in solution requires the use of salts (ion pairs) as precursors, which complicates the analysis of the titration data and the corresponding estimate of the binding strength. In solution, the weak binding energies suggest that anion−π interactions are not as significant for the selective or enhanced binding of anions but offer potential applications in catalysis and transport within functional synthetic and biological systems.