MAGNETIC RESONANCE IN CHEMISTRY Magn. Reson. Chem. 2004; 42: 402–408 Published online 3 February 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1355 Structure elucidation of 11-amino-8-hydroxypenta- cyclo[5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11-lactam through selective acetylation and complete 1 H and 13 C NMR spectral assignment of the mono-, di- and triacetates F. J. C. Martins, 1 * A. M. Viljoen, 1 H. G. Kruger 1 and P. L. Wessels 2 1 Department of Chemistry, Potchefstroom University for CHE, Potchefstroom 2520, South Africa 2 Department of Chemistry, University of Pretoria, Pretoria 0002, South Africa Received 24 September 2003; Revised 6 November 2003; Accepted 12 November 2003 NMR techniques cannot unambiguously distinguish between 11-amino-8-hydroxypentacyclo[5.4.0.0 2,6 . 0 3,10 .0 5,9 ]undecane-8,11-lactam and 8-amino-11-hydroxypentacyclo[5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11-lactam, both of which are possible products during the reaction of pentacyclo[5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11- dione with Strecker reagents. Treatment of 11-amino-8-hydroxy-pentacyclo[5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane- 8,11-lactam with acetic anhydride at room temperature produced a monoacetate. With acetic anhydride containing sodium acetate, a triacetate was obtained at reflux temperature. Treatment with acetyl chloride and N,N-dimethylaniline produced a diacetate. High-field 1 H and 13 C NMR techniques were used in the structure elucidation and assignment of the different NMR resonances of these three acetylated compounds. Copyright  2004 John Wiley & Sons, Ltd. KEYWORDS: NMR; 1 H NMR; 13 C NMR; pentacycloundecane molecules; lactam derivatives INTRODUCTION There is considerable interest in the synthesis and chemistry of polycyclic cage molecules. 1,2 1 H and 13 C NMR spec- troscopy should be the methods of choice for the elucidation of the structures of these types of compounds. However, their 1 H NMR spectra are very complex. In addition to geminal and vicinal proton–proton couplings, long-range (over more than three bonds) proton–proton interactions are also operative, resulting in fairly broad unresolved reso- nances. The difficulty of assigning the NMR spectra of these compounds has been commented on by several authors. 2–5 Complete correlation between the 1 H NMR spectra and the structures proved especially hard to achieve so that in many cases it has been necessary to resort to single-crystal x-ray crystallography to determine the structures. We recently reported 6 that treatment of pentacyclo [5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11-dione (1) with an aqueous mixture of sodium cyanide, ammonium chloride and ammo- nia unexpectedly led to 11-amino-8-hydroxypentacyclo [5.4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11-lactam (2). Suitable crys- tals for x-ray crystallographic analysis could not be obtained, which left little choice of method for structural studies. The structure elucidation of 2 from 1 H and 13 C NMR data proved to be a major task. There is no way to distinguish unambigu- ously between the lactams 2 and 3, are both of which possible L Correspondence to: F. J. C. Martins, Department of Chemistry, Potchefstroom University for CHE, Potchefstroom 2520, South Africa. E-mail: cheamu@puknet.puk.ac.za products during the reaction mentioned above. In order to gain more information on the structure of 2,a 1 H and 13 C NMR investigation on the acetates of 2 was also conducted. 2 1 C NH OH O H 2 N 3 C NH NH 2 O HO O O 1 2 3 4 5 6 7 8 9 10 11 RESULTS AND DISCUSSION The infrared spectrum of the aminohydroxylactam 2 exhibits strong absorption bands in the O–H and N–H stretching vibration regions at 3380, 3295, 3290 and 3165 cm 1 .A carbonyl group absorption is registered at 1655 cm 1 . The mass spectrum shows a molecular ion at m/z 218 and is supportive of the molecular formula C 12 H 14 N 2 O 2 . Owing to the complex pattern of the 1 H NMR signals, a 500 MHz NMR investigation was conducted. Two sets of signals were observed in all the solvents used, which suggested the possibility that a lactim–lactam tautomeric equilibrium exists. This impression was strengthened by the observation that the ratio of the two sets of signal differ in different solvents. Hardly any change in the ratio occurred with Copyright  2004 John Wiley & Sons, Ltd.