Premium
Flavour of natural and roasted Turkish hazelnut varieties ( Corylus avellana L.) by descriptive sensory analysis, electronic nose and chemometrics
Author(s) -
Alasalvar Cesarettin,
Pelvan Ebru,
Bahar Banu,
Korel Figen,
Ölmez Hülya
Publication year - 2012
Publication title -
international journal of food science and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.831
H-Index - 96
eISSN - 1365-2621
pISSN - 0950-5423
DOI - 10.1111/j.1365-2621.2011.02817.x
Subject(s) - electronic nose , chemometrics , flavour , principal component analysis , mathematics , food science , quantitative descriptive analysis , sensory analysis , flavor , chemistry , statistics , artificial intelligence , chromatography , computer science
Summary A total of eighteen natural and roasted hazelnut varieties (amongst which only Tombul variety is classified as prime quality), grown in the Giresun province of Turkey, were compared for their differences in descriptive sensory analysis (DSA), electronic nose (e‐nose) data and chemometrics. Differences in some descriptive of DSA between natural and roasted hazelnuts as well as within the varieties were observed. Although Tombul hazelnut was selected as one of the best varieties in terms of flavour attributes and received the highest intensities in general, no significant differences ( P > 0.05) existed among hazelnut varieties except in certain flavour attributes (‘after taste’ and ‘nutty’). DSA and e‐nose data of natural and roasted hazelnuts were also evaluated for discrimination using principal component analysis (PCA) and cluster analysis. Results of PCA using e‐nose data showed that extracted principal components explained 99.7% and 99.8% of the total variance of the data for natural and roasted hazelnut varieties, respectively. Both DSA and e‐nose can be used for discrimination of natural and roasted hazelnuts.