Halogen Bonding Propensity in Solution: Direct Observation and Computational Prediction

Abstract

Halogen‐bonded complexes are often designed by consideration of electrostatic potential (ESP) predictions. ESP predictions do not capture the myriad variables associated with halogen bond (XB) donors and acceptors; thus, binding interaction cannot be quantitatively predicted. Here, a discrepancy between predictions based on ESP energy difference (ΔVs) and computed gas phase binding energy (ΔEbind) motivated the experimental determination of the relative strength of halogen bonding interactions in solution by Raman spectroscopic observation of complexes formed from interacting five iodobenzene‐derived XB donors and four pyridine XB acceptors. Evaluation of ΔEbind coupled with absolutely‐localized molecular orbital energy decomposition analysis (ALMO‐EDA) deconvolutes halogen bonding energy contributions and reveals a prominent role for charge transfer (CT) interactions. Raman spectra reveal ΔEbind accurately predicts stronger interactions within iodopentafluorobenzene (IPFB) complexes than with 1‐iodo‐3,5‐dinitrobenzene (IDNB) complexes even though IPFB has similar electrostatics to IDNB and contains a smaller σ‐hole.Strongly activated halogen bond donors give rise to observable bound and unbound states in Raman spectroscopy when paired with good halogen bond acceptors. The two strongest halogen bond donors examined have almost identical electrostatic potentials, but differ vastly in theoretical binding affinity and experimental halogen bonding strength. The energy terms giving rise to this behavior are elucidated by energy decomposition analysis.