Conformational analysis. Part 9. A lanthanide-induced shift (LIS) nuclear magnetic resonance investigation of conformational isomerism and structure in 1- and 2-naphthaldehyde, 9-acetylanthracene, and 9-anthraldehyde

Abstract

A lanthanide-induced shift n.m.r. investigation of conformational isomerism in the title molecules has been carried out, based upon the pseudo-contact shifts of all the proton and carbon nuclei in these molecules. The ¹³C diamagnetic shifts were removed by use of La(fod)₃, and the pseudo-contact shifts analysed to obtain the conformational preferences of the formyl and acetyl groups. The results obtained for 9-anthraldehyde were very dependent on the concentration of the substrate. Only the pseudo-contact shifts obtained from dilute (0.2M) solutions gave good agreement between the observed and calculated shifts, with ca. 30° for the torsion angle between the formyl group and the aromatic rings. As starting geometries in the LIS analyses both a crystal structure of 9-anthraldehyde and fully optimized MNDO geometries for each value of the torsional angle have been used. The precise results were dependent on the starting geometry used. The results obtained for 9-acetylanthracene showed a much less marked dependence on the substrate concentration. Two different starting geometries gave the same good solution (R_cryst. 0.07) for a torsional angle of ca. 70° of the acetyl group with respect to the anthracene rings. The analysis of 1-naphthaldehyde was again critically dependent on the geometrical model used. The standard geometry gave a solution (R_cryst. 0.018) at 28%E-conformation but only when H-8 was removed. This proton has a large LIS and the calculated value is critically dependent on the geometry of the aldehyde at H-8. Analysis of the data using as input a fully optimized geometry for both the E- and Z-conformers from MNDO calculations gives the best solution (R_cryst. 0.021) for 5%E-conformation, which now includes all the nuclei. On using the pseudo-contact shifts measured for more dilute solutions the minimum shifts to 15%E-conformation, without any appreciable change in the agreement factor. The use of a standard geometry as input for the LIS analysis of 2-naphthaldehyde gave a good solution (R_cryst. 0.011) at ca. 80%E-conformation. No significant effect of the concentration of the substrate was detectable.