Multivariate analysis of attenuated total reflection-Fourier transform infrared spectroscopic data to confirm the origin of honeys.

摘要: Fourier transform infrared (FT-IR) spectroscopy and chemometrics were used to verify the origin of honey samples (n = 150) from Europe South America. Authentic collected five sources, namely unfiltered Mexico in 2004, commercially filtered Ireland Argentina Czech Republic 2005 2006, Hungary 2006. Samples diluted with distilled water a standard solids content (70° Brix) their spectra (2500–12 500 nm) recorded at room temperature using an FT-IR spectrometer equipped germanium attenuated total reflection (ATR) accessory. First- second-derivative normal variate (SNV) data pretreatments applied spectra, which analyzed partial least squares (PLS) regression analysis, factorial discriminant analysis (FDA), soft independent modeling class analogy (SIMCA). In general, when wavelength range (6800–11 rather than whole spectrum was studied, higher correct classification rates achieved. An overall 93.3% obtained for honeys by PLS while FDA techniques correctly classified 94.7% samples. Correct classifications up 100% achieved SIMCA, but models describing some classes had very high false positive rates.