No abstract

No abstract

The 13C NMR spectrum of the pentacyclic triterpene friedelin was fully assigned, taking into account signal multiplicities, correlations of the carbon lines with known proton resonances and Eu-induced shifts. T h i s work clarifies the situation prevalent in the literature, in which three separate groups have published contradictory information. In addition, the carbon and proton spectra of the naturally occurring 3c~-hydroxyfriedelan-2-one and its synthetic acetate were analysed. The different preferred conformations of the D/E rings for friedelin (boat-boat) and polpunonic acid (chair-chair) are rationalized.In the course of a recent study of the '"C NMR spectroscopy of polpunonic acid (2)l we needed, for comparison purposes, to have the corresponding data for friedelin (1). An inspection of the literature showed the situation to be confusing. The first full assignment of the carbon spectrum of friedelin was published in 1977 by Beierbeck et a1.: as part of a derivation of semi-empirical parameters for the calculation of '"C chemical shifts. Friedelin was only one of more than 80 substances examined, and no details were given on the spectral techniques used; it is assumed that the assignment was based mostly on the calculation of predicted values of the chemical shifts.Three years later, a paper by Gunatilaka et al." presented the 13C NMR data for friedelin and other friedelanones. They reported the use of off -resonance decoupled spectra (SFORD) and the inversionrecovery technique (PRFT) to determine multiplicities, and also lanthanide-induced shifts, but the observed A8 values were not given. Inspection of the data reveals, however, severe inconsistencies in comparison with the other friedelanones and friedelanediones in the paper, and also with what would be expected from basic chemical shift kn~wledge.~.' In addition, for many carbon signals even the multiplicities do not correspond to those implied by Beierbeck et aL2 In view of this not even the multiplicities of the carbon lines are a priori reliable, and we decided to make a careful study of the '"C NMR spectrum of friedelin in order to clarify the issue. EXPERIMENTAL AND RESULTSThe NMR spectra were recorded on a Bruker WH-270 spectrometer in the Fourier transform mode,

No abstract