Prediction of odour detection thresholds using partition coefficients

A linear regression between odour detection thresholds and three basic physical properties (vapour pressure, water solubility and octanol/water partition coefficient) has been derived. The principles underlying this model are the transport process of the odorants from air to the olfactory neurons in the nasal epithelium. The regression was applied to odour detection thresholds measured from different authors separately, showing similar parameters. A general linear regression was obtained for odour detection threshold data from several authors and 121 chemicals. This regression allows a preliminary estimation of odour detection thresholds from literature data with fair accuracy. The coefficient of determination obtained, r(2) = 0.769, indicates that these physical properties account for 77% of the odour threshold variance. Copyright (C) 2011 John Wiley & Sons, Ltd.