Food Chemistry 54 (1995) I-7 Copyright 0 1995 Elsevier Science Limited Printed in Great Britain. All rights reserved 0308-81461951S9.50 Volatile compounds of rehydrated French beans, bell peppers and leeks. Part II. Gas chromatography/sniffing port analysis and sensory evaluation Saskia M. van Ruth,” Jacques P. Roozen,’ Jan L. Cozijnsen” & Maarten A. Posthund ’ Wageningen Agricultural University, Department of Food Science, PO Box 8129; ’ Wageningen Agricultural University, Department of Organic Chemistry, PO Box 8026, NL-6700 EC Wageningen, The Netherlands (Received 17 October 1994; revised version received and accepted 12 January 1995) The flavours of rehydrated diced French beans, red bell peppers and leeks were characterised by gas chromatography/sniffing port analysis (GCYSP) of volatile compounds released in a mouth model system, and by descriptive sensory analysis. Volatile compounds were identified by combined gas chromatography/mass spectrometry. In French beans, bell peppers and leeks, respectively, 10, 16 and 22 compounds possessed detectable odours. A common odour profile was shown in the three vegetables. It comprised each of the odour active compounds present in French beans: i.e. 2-methylpropanal (chocolate), 2/3_methylbutanal (choco- late), 2,3-butanedione (caramel, fatty), hexanal (grassy, bell pepper), 2-methyl-2- butenal (chemical), octanal (sweet, sickly/musty, grassy, rancid), 1 -octen-3-one (mushroom), dimethyl trisulphide (rotten, metal), I-octen-3-01 (fatty, sickly/ musty, mushroom) and one unknown compound (chemical, rotten, rancid). The three vegetables differed markedly in GCYSP patterns and in scores for sensory attributes. Use of nose-clips diminished the scores for attributes in sensory analysis. In principal component analysis, correlation of rehydrated vegetables with sensory attributes and volatile compounds showed considerable contribution of volatile compounds to the flavour of rehydrated vegetables. INTRODUCTION Foods have highly complex chemical compositions, containing both volatile and non-volatile substances. Their sensory responses may be significantly affected by their relative concentrations and the temperature of the product as consumed (Heath, 1981). Human taste-buds are capable of differentiating only four stimuli. As the nose is capable of discerning hundreds of different odours, it is not surprising that a major part of flavour research has dealt with analysis of volatile compounds (Hoff et al., 1978). An instrumental approach to characterise flavour can be regarded as a two-phase arrangement. The first phase involves representative isolation of volatile com- pounds, because gas chromatographic (GC) patterns are largely influenced by isolation procedures. Release rates of volatile compounds from a product depend upon partition coefficients of the compound, molecular interactions and ambient temperatures (Legger & Roozen, 1994). The second phase involves selection of volatile compounds relevant to the flavour. This implicates correct determination of the relevant flavour compounds from the whole range of volatiles present in a particular food product (Dirinck & De Winne, 1994). An interesting approach is sniffing the gas chromato- graphic effluent (GC/sniffing port analysis; GC/SP), in order to associate odour activity with the eluting com- pounds. Flame ionisation, flame photometric, electron capture, mass spectrometric, fourier transform, infrared and thermal conductivity detectors are not as sensitive as the human nose for many odorants (Acree & Barnard, 1994). The interest in determining the individ- ual contribution of volatile compounds present in foods, has led to a new generation of GCYSP techniques, which can be classified into four categories: (i) Dilution analysis methods for producing titre or potency values based on stepwise dilution to thresh- old, e.g. CHARM (Acree et al., 1984) and aroma extraction dilution analysis (Ullrich & Grosch, 1987). (ii) Response interval methods for recording time duration of perceived odours, and the number of assessors with odour perception, it also estimates a titre or potency (Linssen et al., 1993).