Hedonic judgments of chemical compounds are correlated with molecular size

Abstract

Different psychophysical works have reported that, when a wide range of odors is assessed, the hedonic dimension is the most salient. Hence, pleasantness is the most basic attribute of odor perception. Recent studies suggest that the molecular size of a given odorant is positively correlated with its hedonic character. This correlation was confirmed in the present study, but further basic molecular features affecting pleasantness were identified by means of multiple linear regression for the compounds contained in five chemical sets. For three of them, hedonic judgments are available in the literature. For a further two chemical sets, hedonic scores were estimated from odor character descriptions based on numerical profiles. Generally speaking, fairly similar equations were obtained for the prediction of hedonic judgments in the five chemical sets, with R(2) values ranging from 0.46 to 0.71. The results suggest that larger molecules containing oxygen are more likely to be perceived as pleasant, while the opposite applies to carboxylic acids and sulfur compounds.

Keywords: VOC; hedonic valence; numerical odor profile; odor character descriptor; olfaction; pleasantness.

Figure 1.

Plot of fitted regression analysis of p₁ (loadings in the formation of the first principal component of Dravnieks’ Atlas) versus HT_D84 (hedonic tones proposed by [34]). The fitted curve corresponds to Equation (4). The four highest residuals (filled points) are moderate outliers.

Figure 2.

Results from the PCA performed with the database obtained by Amoore and Venstrom [39]. Prior to the analysis, variables were transformed for normality, and then mean-centered and scaled to unit variance. (A) Loading plot (p₂ vs. p₁) and (B) score plot (t₂ vs. t₁) for the first and second principal components. White squares represent the samples that were described with highest scores as floral, and so on, according to the caption (the pungent category does not appear because none of the samples was primarily described as pungent).