A new approach for discriminant analysis of adulterated milk is proposed based on two-dimensional (2D) hetero-spectral near-infrared (NIR) and mid-infrared (IR) correlation spectroscopy along with multi-way partial least squares discriminant analysis (NPLS-DA). NIR transmittance spectra and IR attenuated total reflection spectra of pure milk and adulterated milk with level of melamine varying from 0.03 to 3 g·L(-1) were collected at room temperature. The synchronous 2D hetero-spectral IR/NIR correlation spectra of all samples were calculated to build a discriminant model to classify adulterated milk and pure milk. Also, the NPLS-DA models were built based on synchronous 2D homo-spectral NIR/NIR and IR/IR correlation spectra, respectively. Comparison results showed that the NPLS-DA model could provide better results using 2D hetero-spectral IR/NIR correlation spectra than using 2D homo-spectral NIR/NIR and 2D IR/IR correlation spectra.
Keywords: Adulterated milk; Melamine; Multi-way partial least squares discriminant analysis (NPLS-DA); Two-dimensional hetero-spectral IR/NIR correlation spectroscopy.
Copyright © 2015 Elsevier B.V. All rights reserved.