Non-covalent π interactions Surfing the π-clouds for Non-covalent Interactions : A comparative Study of Α renes versus Alkenes

A comparative study by NMR spectroscopy using molecular balances indicates that non-covalent functional group interactions with an arene dominate over those with an alkene and a π-facial intramolecular hydrogen bond from a hydroxyl group to an arene is favoured by ~1.2 kJ mol. The strongest interaction observed in this study is with the cyano group and analysis of the series of Et, CH=CH2, C≡CH and C≡N groups is indicative of a weak long range electrostatic interaction and a correlation with the electrophilicity of the C atom of the Y substituent. Changes in the free energy differences of conformers show a linear dependence on the solvent hydrogen bond acceptor parameter β. The vital role played by non-covalent interactions, and especially those involving the contribution of aromatic rings to chemical and biological recognition, continues to be a subject of intense research activity. Detailed understanding and quantifiable estimates of the strength, distance, and angular dependence of such intermolecular forces is now considered to be essential, not only for the understanding of protein-ligand interactions and hence for drug design, but also for the synthesis of new asymmetric ligands, catalysts and sensors. In essence, such phenomena as π-stacking, the behaviour of an aromatic ring as a hydrogen bond acceptor, or cation-π interactions can each be viewed as a single “Velcro like hoop and loop” of differing strength, with the combination of several of these then leading to overall binding and molecular recognition. Although a variety of techniques including structural database mining, measurement of gas phase complexes and computational modelling have all contributed to provide valuable insights, the use of designed molecular balances, relying on measurement of a conformational change, has proven to be a particularly powerful tool for obtaining data on the very small interaction energies involved. Moreover, such balances also allow the often dominant influence of solvation to be explored. The molecular torsion balance pioneered by Wilcox has provided the basic framework for many elegant studies which exemplify the quantitative power of this approach, and the results from a significant number of new molecular balances have been summarised in an insightful review by Cockroft. Within this area, we have previously introduced the dibenzobicyclo[322]nonane framework as a useful probe for the comparative study of arene-functional group interactions in solution through systematic variation of the two substituents Y and Z on the central carbon atom of the bridge and determination of the conformational population, up (U) or down (D), of the more electronegative substituent by NMR. (Figure 1). In this manner, interesting insights, such as the “preference” of a fluorine atom over a hydroxyl group for an aromatic ring, or the arene affinity of sulphur over oxygen were gained. It is important to recognise that this bicyclic scaffold is not a torsional balance but a top pan balance (or seesaw) since, for any given derivative, the influence of Y on the first aromatic ring is being measured against the counterbalancing interaction of Z with the second aromatic ring. Figure 1. Conformational equilibrium U / D of molecular balances In sharp contrast to the extensive body of work on aromatic systems however, relatively few studies have quantified noncovalent interactions involving the simplest fundamental π-system of all, an alkene. Thus, even although the existence of π-facial hydrogen bonding of a hydroxyl group to an alkene has been recognised through infrared dilution studies and X-ray crystallographic database mining, a quantifiable comparison of such arene versus alkene non-covalent functional group interactions has, to the best of our knowledge, not yet been made. Herein, we now present our preliminary results to establish a basis set with a particular focus on measuring the relative strengths of a π-facial intramolecular hydrogen bond to an arene versus an alkene. As emphasised in Figure 2 consideration of the requirements for such a comparison leads to the design of four different molecular balances in order that measurements can be made relative to an identical counterbalancing interaction. In the present study for example, a comparison of the OH-arene versus OH-alkene [∗] Dr A. E. Aliev, Dr. J. R. T. Arendorf, Dr. I. Pavlakos, Dr. R. B. Moreno, Dr. M. J. Porter, Prof. W. B. Mortherwell Department of Chemistry, University College London 20 Gordon Street, London WC1H 0AJ (UK) E-mail: a.e.aliev@ucl.ac.uk, w.b.motherwell@ucl.ac.uk Prof. H. S. Rzepa, Department of Chemistry, Imperial College London, South Kensington campus, London, SW7 2AZ. [∗∗] Support for this work from Leverhulme Trust is gratefully