Floral odor analyses

D. Marsault, and N. Chaline. J. Pierre (INRA Rennes) contributed to the field work, L.J. Wadhams (lACR Rothamsted) to the floral odor analyses, and L. Malone (Horticultural and Food Research Institute, New Zealand) and A. Gouty (lACR Rothamsted) to the manuscript. Part of this work was funded by the EU in the Biotechnology Program of the 4th Framework, and by the CETIOM within the framework of an inter-institute study. 

Anselmi et al. (02JCS(P2)1525), very recently published the conformational analysis and dynamics of cA/ rafts-4-methylcyclohexyl tetrahydro-pyranyl ethers (cf. Scheme 12) and compared their structures with the floral odors of the compounds. The cis isomer 40(cis), endowed with a main white flower note, has an bent, oval molecular shape. The trans derivatives 40(trans) and 41, exhibit different odors, possess an extended structure of cylindrical molecular shape. Brenna et al. (02CJC714) reexamined the configuration/conformation of rose oxide analogues. However, the conformational analysis provided poorer results than published previously (78JPC303) without even citing the previous paper. 

Of the 10 constituents which represent nearly half the oil of neroH, only linalool (10) can be said to contribute direcdy to the characteristic aroma of orange flower oil. In 1977, IFF chemists performed an in-depth analysis of this oil and identified three simple terpenic compounds, each present at less than 0.01%, a-terpenyl methyl ether [1457-68-0] (31), geranyl methyl ether [2565-82-4] (32), andhnalyl methyl ether [60763-44-2] (33) (11). The latter two compounds possess green floral-citms aromas and have been known to perfumery for some time a-terpenyl methyl ether (31) has been called the orange flower ether by IFF chemists owing to its characteristic odor. 

Fig. 5.2. Gas chromatography (GC) and electroantennography (EAG) analysis of male Manduca sexta antennal responses to floral volatiles from the night blooming cactus Peniocereus greggii. The upper trace is a flame ionization detection (FID) chromatogram of floral headspace odors separated on a carbowax GC column, while the lower trace is a simultaneous recording of summed antennal action potentials elicited by individual compounds as they elute. The largest absolute responses followed methyl benzoate, methyl salicylate, and benzyl alcohol (peaks 3-5, respectively). Note the poor responses (circled) to benzaldehyde and benzyl benzoate (peaks 2, 6) and the disproportionately higher responses (bold arrows) to methyl salicylate and benzyl salicylate (peak 7) relative to their peak areas. Peak 1 is the internal standard (toluene) remaining unnumbered peaks are ambient contaminants.

In an attempt to simplify the complex mixture of chemicals present in extracts of labrusca grapes we have used 6% water deactivated Florisil developed with solvents of increasing polarity (ii). Made hy mixing diethyl ether at 0, 1, 3, 10, 30, and 100% with pentane, each solvent has approximately twice the polarity of the previous one (Ifi.). With Concord grape juice extracts the fraction with the most labrusca like odor character was the 3% fraction which had a very strong sweet floral characteristic. An analysis of this fraction (li) showed the presence of damascenone [l-(2,6,6-trimethyl-l,3-cyclohexadien-l-yl)-2-buten-l-one] a compound previously reported to be present in vinifera eranesf17) as well as other natural productsQfi.). 

The flavor profile analysis was performed as previously described [6]. The number of evaluated flavor attributes was reduced from six to five (sweet, green, floral, fruity, and sour), removing the smoky odor quality as it was not important for the tomatoes investigated here. The intensities of the odor attributes were scored on a category scale from 0 (not perceptible) to 3 (strongly perceptible) in increments of 0.5. 

---