Premium
Fusion of near‐infrared and fluorescence spectroscopy for untargeted fraud detection of Chinese tea seed oil using chemometric methods
Author(s) -
Hu Ou,
Chen Jing,
Gao Pengfei,
Li Gangfeng,
Du Shijie,
Fu Haiyan,
Shi Qiong,
Xu Lu
Publication year - 2018
Publication title -
journal of the science of food and agriculture
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.782
H-Index - 142
eISSN - 1097-0010
pISSN - 0022-5142
DOI - 10.1002/jsfa.9424
Subject(s) - rapeseed , partial least squares regression , chemometrics , chemistry , sunflower oil , edible oil , chromatography , food science , mathematics , statistics
BACKGROUND This paper investigated the feasibility of data fusion of near‐infrared (NIR) and fluorescence spectroscopy for rapid analysis of cheap vegetable oils in Chinese Camellia oleifera Abel. (COA) oil. Because practical frauds usually involve adulterations of multiple known and unknown cheap oils, traditional analytical methods aimed at detecting one or more known adulterants are insufficient to identify adulterated COA oil. Therefore, untargeted analysis was performed by developing class models of pure COA oil using robust one‐class partial least squares (OCPLS). RESULTS The most accurate OCPLS model was obtained with fusion of standard normal variate (SNV)‐NIR and SNV–fluorescence spectra with sensitivity of 0.954 and specificity of 0.91. Robust OCPLS could detect adulterations with 2% (w/w) or more cheap oils, including rapeseed oil, sunflower seed oil, corn oil and peanut oil. CONCLUSION Fusion of NIR and fluorescence data and chemometrics provided enhanced capacity for rapid and untargeted analysis of multiple adulterations in Chinese COA oils. © 2018 Society of Chemical Industry