The Cation Complexation Model Predicts the Experimental p-Facial Selectivity of 2-ax- and 2-eq-Substituted Cyclohexanones. A Detailed Ab Initio MO Investigation Veejendra K. Yadav, * Duraiswamy A. Jeyaraj ² and Rengarajan Balamurugan Department of Chemistry, Indian Institute of Technology, Kanpur 208 016, India Received 22 May 2000; revised 6 July 2000; accepted 20 July 2000 Abstract —The geometrical changes on complexation of the carbonyl oxygen with prototypical cations such as H + and Li + are in tune with the polarity features of the C–X bonds in 2-ax-X-cyclohexanones (XCl, F, SR, OR; RH, Me); the stereoelectronic effects apply. While the 2-ax-Cl- and 2-ax-SR-cyclohexanones are predicted to favor axial attacks, the 2-ax-OR- and 2-ax-F-cyclohexanones must capture a nucleophile predominantly from the eq-direction. The eq-selectivity of 2-ax-OR-cyclohexanone is in contradiction with the torsional model of Anh and Felkin. The Houk model will also fail due to the eminent steric interactions arising from the 2-ax-OMe function in the eq-TS. Neither the Anh–Felkin model nor the Cieplak model could be applied to 2-eq-MeO-cyclohexanone. The complexation model, however, predicts eq-selectivity in full agreement with experimental results. Second order perturbation theory analysis of the Fock matrix in NBO basis indicates that the role of the antiperiplanar effects is not as significant as perceived earlier by Cieplak and by Anh and Felkin.  2000 Elsevier Science Ltd. All rights reserved. Introduction The diastereofacial selection of substituted cyclohexanones is an area of current interest. 1 Several models have been proposed to explain the p-facial preferences. The Anh– Felkin model 2 considers the transition state (TS) to be electron-donating and, thus, a nucleophile is required to attack the carbonyl carbon from a direction that is antiperi- planar to an electron-attracting s bond at the a carbon. In contrast, the Cieplak model 3 assumes the TS to be electron- attracting and, thus, a nucleophile is required to attack anti- periplanar to an electron-donating s bond at the a carbon. These two models, thus, predict opposite selectivities for the same substituent type. Furthermore, despite being simple, these models have failed to adequately accommodate several experimental stereochemical results. The Houk model 4 is not general either because, for instance, an extra- polation of the true ax-selection of 2-ax-Cl-cyclohexanone to 2-ax-OR-cyclohexanone contradicts the experiments. The pictorial representations of these models are given in Fig. 1. The orbital distortion model by Frenking 5 and the polarized p-frontier molecular orbital (PPFMO) model by Dannen- berg 6 are less popular amongst organic chemists probably because their applications are tedious. The electrostatic control models by Chandrasekhar and Mehta 7a and Houk 7b are successful in many instances but fail at some. Tomoda 8 has developed an exterior frontier orbital extension (EFOE) model that relies on the total space available on both the faces of the carbonyl function. No results, however, have been published as yet on the selectivities of 2-ax- and 2-eq- X-cyclohexanones. We have reasoned that the above failures are likely due to the ignorance of the experimentally well-documented cation complexation and/or chelation effects 9,10 and, hence, the consequent geometrical changes in the substrate. We our- selves have been attracted to the study of facial selectivities Tetrahedron 56 (2000) 7581–7589 Pergamon TETRAHEDRON 0040–4020/00/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved. PII: S0040-4020(00)00663-3 Keywords: diastereoselection; 2-X-cyclohexanones; cation complexation; p-facial selection. * Corresponding author. Tel.: +91-512-597439; fax: +91-512-590007; e-mail: vijendra@iitk.ac.in ² Present address: Chemical Biology, Max-Planck Institute of Molecular Physiology, Dortmund, Germany. Figure 1. Pictorial representations of: (a) the Cieplak model; (b) the Anh– Felkin model; and (c) the Houk model for axial attack on cyclohexanone with LiH as the nucleophile. Nunucleophile.