Relations structure-odeur de bois de santal: recherche d'un modèle d'interaction fondé sur le concept d'hypermotif santalophore

Abstract

Structure-odour relationships for sandalwood odorants were studied for a set of 139 compounds (57 active and 82 inactive), essentially alkylcyclohexylalkanols, acyclic, norbornyl, campholenyl and decalin derivatives. Active compounds possess only a hydroxyl as functional group, but inactive compounds possess a hydroxyl, ketone, epoxide, aldehyde, ester or ether group. In general, these compounds have a flexible structure. The first step in the present study consists of a statistical evaluation of empirical rules concerning structure-odour relationships for sandalwood odorants, using discriminant analysis. After testing the empirical rules, the main structural elements responsible for sandalwood odour are defined from the previous rules. Structural elements which are favourable and those unfavourable to sandalwood odour are distinguished. Structural elements ( t-butyl group or an equivalent and hydroxyl group) found necessary to produce sandalwood fragrance constitute parts of santalophore patterns I and II. These patterns are built using Alchemy II program by superposition of strong sandalwood compounds possessing a relatively rigid structure. The distance between the hydroxyl oxygen atom and the quaternary carbon atom ( t-butyl) is found approximately 7.1 Å in pattern I and 6.5 Å in pattern II. Then patterns I and II cannot be fused in a unique santalophore pattern which is necessary if only one type of receptor exists for sandalwood odour. The difference between the distances observed in patterns I and II is not very important (0.6 Å). The molecules possessing pattern I and those possessing pattern II probably interact with different parts of the same receptor, which is slightly larger than the two patterns. The concept of superpattern is then introduced. The superpattern, defined as the envelope of santalophore patterns, is built by their superposition. It consists of a hydroxyl group and a bulky part, more important than those in patterns I and II. This bulky part is constituted of carbon and hydrogen atoms. Geometric characteristics of the superpattern cannot be described only by the distances between the structural elements. Thus, in order to predict sandalwood odour for a test set of 17 compounds, molecules were described by 3 types of geometric parameters: distances, bond angles and dihedral angles between hydroxyl oxygen atom and methyl groups of the bulky part. It was found that a complete discrimination between active and inactive molecules is obtained when two geometrical parameters are used simultaneously. The best discrimination is obtained with distances and dihedral angle parameters. An interaction model is discussed by combination of the superpattern with some hypotheses on the type of interactions involved. Thus the bulky part is considered responsible for interactions by dispersion forces and the hydroxyl group responsible for hydrogen bonding interactions. To test this hypothesis, the hydroxyl group of α-santalol and β-santalol were methylated. The compounds obtained have no sandalwood odour, as confirmed by a perfumer. We could then conclude the hydroxyl group is a hydrogen bonding donor, but it is interesting to note that some carbonyl compounds, such as decahydro β-naphtyl formate, which are only hydrogen bonding acceptors are described in the literature as possessing an odour reminiscent of sandalwood. These compounds, often used in compositions with a sandalwood character are under investigation to determine if they actually present the sandalwood note.