Reverse anomeric effect and steric hindrance to solvation of ionic groups

C. L. Perrin
1995 Pure and Applied Chemistry  
The proportions of axial anomers of various glucosylamines (2) and their conjugate acids were determined by 1H NMR. The change upon N-protonation is small and can be accounted for by steric effects, without any "reverse anomeric effect." To test whether N-protonation changes the steric bulk of an imidazolyl group, ring-inversion equilibria of cis-N-(4-alkylcyclohexyl)imidazoles (3c, 4c) and of their conjugate acids were studied. The equatorial-to-axial free-energy change, A h or AlmH+, is 2.2 f
more » ... 0.1 kcal/mol. To measure relative sizes more precisely, an NMR method was developed. The ratio of acidity constants of cis and trans 4 could be determined from chemical shifts as a 1: 1 mixture was titrated. The cis isomer is 0.048 pK unit less basic, corresponding to a AA of 0.089 f 0.004 kcdmol, with protonated imidazolyl larger. To reinvestigate the effect of N-protonation on conformational equilibria in sugar derivatives, this N M R titration was applied to a mixture of a-and p-N-(glucosy1)imidazoles (5). The AA is -0.018 to -0.368 kcal/mol, exactly opposite to the reverse anomeric effect! This method has been applied to r-butylcyclohexanes 6 with acidic groups. Ions usually have a measurably lower preference for the axial position than do the corresponding neutrals. Introduction. The favored chair conformer of a monosubstituted cyclohexane has its substituent equatorial. Were the substituent axial, it would suffer destabilizing steric repulsions with the axial hydrogens at C3 and C5. The A value, or freeenergy difference between axial and equatorial conformers, as in eq 1, is ln([axial]/[equatorial]) (1) a quantitative measure of the effective size of a substituent (1). The topics addressed here involve how that conformational equilibrium may be modified by anomeric effects and by the need for solvation of ionic substituents. Anomeric effects, especially with charged substituents, represent significant puzzles regarding molecular structure. They are important for understanding conformations of carbohydrates, simple organic heterocycles, and nucleosides and for understanding the reactivity of such molecules, which often react via their protonated forms as intermediates. More generally, the conformational behavior of charged substituents is a less well understood aspect of conformational analysis, and a new method for precise assessment of the effect of the charge can provide insight into the old problem of steric hindrance to solvation. Reverse Anomeric Effect. Despite the generality of the anomeric effect (2) (the axial tendency exhibited by many electronegative groups at C 1 of a tetrahydropyran or related heterocycle), some cationic groups, as in N-(a-glycosy1)pyridinium ions, prefer the equatorial position (3). This preference is called a "reverse anomeric effect", contributing as much as 1-3 kcdmol to the stabilization of the equatorial form. However, all the substituents that produce this effect involve bulky aromatic rings, and the observations could be due simply to avoidance of prohibitive steric repulsions associated with placing that group axial. A clever experiment was the study of conformational preferences of irnidazolyl groups. Since protonation at the distant nitrogen is considered not to change the size of the group, an imidazolyl group provides its own control for steric factors. Indeed, on N-protonation or N-methylation of N-(tetra-0-acetyla-glucosyl) or -mannosyl)imidazole there is a shift toward the conformer with the imidazolyl group equatorial (4). More quantitatively, N-(tri-0-acetyl-a-xylopyranosy1)imidazole (1) exists as 65% equatorial conformer (1E) in CDCl3, whereas in the presence of trifluoroacetic acid the proportion increases to >95% (5). This is a substantial change, corresponding to a free-energy change of >1.4 kcdmol. If N-protonation does not change the size of the imidazolyl group, the shift of the equilibrium cannot be due to steric effects but must be presumed to be due to the positive charge. Such results have been accepted as the best evidence for the reverse momeric effect. However, it must be recognized that the populations were not determined 719
doi:10.1351/pac199567050719 fatcat:j6ixxarnljenzhhprw5347y6hi