J. Soc. Cosmet. Chem. 22 ! 5-26 (1071) ¸ 1971 Society of Cosmetic Chemists of Great Britain Odor and optical activity E. T. THEIMER and MINA R. McDANIEL* Presented in part at the Symposium on "Perfumery", organised by the British Society of Perfumers and the Society of Cosmetic Chemists of Great Britain, at Eastbourne, Sussex, on 21st April 1970. Synopsis--The important question whether optical ANTIPODES of odorous compounds can be discriminated by the human OLFACTORY system, has as yet not been answered un- equivocally. Previous studies on the ODOR perception of d and l isomers provided no mechan- ism for eliminating the effect of trace impurities. In the present study, five compounds with fragrance odors were each prepared in four forms dextro a, dextro •, laevo a, laevo [•. Because of the synthetic routes used, the two a materials contained similar trace impurities, as did the two • materials. Ability to discriminate was determined by panel testing using both "lay" and "expert" subjects. The results showed that dextro and laavo forms could be distinguished in every case. The TECHNIQUES employed surmount the problem of the presence of unavoidablo trace IMPURITIES and produce results leading to con- clusions having statistica'ly high confidence factors. The importance of using panels was made clear by cases of individuals lacking discrimin- atory ability. Since this ability depended upon the compound being examined, as well as upon the individual, anosmias are postulated which may be limited to chirality and are thus called specific CHIRAL ANOSMIA. INTRODUCTION Molecular structure determines the quality and intensity of olfactory stimuli evoked by chemical compounds. However, the relationship between structure and odor, which has been studied for many classes of compounds, remains strictly empirical, although the importance of even minor structural differences in determining olfactory characteristics is well established. Thus geometrical isomers, whether ortho/meta/para, cis/trans, exo/endo, etc., are in most cases readily distinguished both as to quality and strength. On the other hand, the influence of optical rotation on odor remains *International Flavors & Fragrances, Union Beach, N.J. 07735, U.S.A. 15

16 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS relatively obscure, with conflicting statements reported in the literature both for threshold and quality, in contrast to the clearly established ability to distinguish such isomers by taste. This is due in part to the extreme sensitivity of the nose, which makes the presence of trace impurities, even below the limits of physical measurement, an important factor. Further, in these reports, the method of determining the odor differences, if any, is usually not elaborated, which is not conducive to confidence in the validity of the conclusions. Three methods have been used to obtain optical antipodes for olfactive testing, of which the first is the least, and the third the most convincing in reliability. Isolation or preparation from sources of optically active starting materials For example, d- and/-citronellols can be made from the respective d- and/-citronellals ex oil of Citronella Java and Java Lemon Oil, or from d- and 1-2, 6-dimethyl-2, 7-octadienes prepared by the thermal rearrangement of d- and/-pinanes which in turn derive from d- and l- alpha or beta pinenes. If, after such diverse routes, the final products, purified as well as possible, are indistinguishable in odor, the conclusion is compelling that optical activity does not influence the odor of citronellol. On the other hand, the ability of an expert panel to distinguish between the two can always be attributed, not to inherent differences between the d- and l- forms, but to the presence of trace impurities deriving from the starting materiMs or the processing. Resolution of the racemate In this case, the starting material being the same, any difference de- tected is more convincing. It may be argued, however, that because of physical differences in the intermediates used in the resolution, the end products may be of different purity, and again it may be the impurities which are responsible for odor differences. In the case of alpha-ionone, which has been resolved and evaluated twice by Sobotka et al (1) and by Naves (2), the conclusions reached are quite contradictory. In one case (1), the d and l forms are stated to be different in odor in the other (2), they are stated to be the same, but only one-tenth as strong as the racemate prepared by mixing the prepared