Analysis of milk odd- and branched-chain fatty acids using Fourier transform (FT)-Raman spectroscopy

Fourier transform (FT)-Raman spectra of pure C13:0, C15:0, C17:0, iso C14:0, iso C15:0, and ante C15:0 fatty acid methyl ester standards (FAMESs) and 75 milk fat samples from 6 different dietary experiments were acquired at room temperature (RT) and immediately after freezing at -80 °C (FT). The latter generally included much more well-defined and sharper scattering bands than those obtained at RT. Further, the spectra at FT revealed additional acuate bands in the vicinity of peculiar wavenumber regions, as well as an increase of Raman scattering intensity, which was sometimes associated with a shift of the peak. Partial least-squares (PLS) regression models based on either selected regions or the full spectra and using two pretreatment methods [multiplicative scatter correction (MSC, using raw spectra of milk fat only) and modified MSC (MMSC, a combination of pure FAMESs and milk fat spectra)] with cross-validation were used to evaluate the different types of milk fat FT-Raman spectra for the predictions of individual odd- and branched-chain fatty acids (OBCFAs) and their sums. In general, most individual (C15:0, ante C15:0, iso C17:0, and ante C17:0) and grouped (ODD, ANTE, and total OBCFAs) fatty acids were favorably (coefficient of determination, R(2) > 0.65) predicted using models with FT spectra only or a combination of RT and FT spectra (RFT), when compared to models with spectra analyzed at RT only. The results indicate the interest to use FT-Raman spectra collected at different temperatures for the prediction of narrow concentrations of saturated OBCFAs in milk fat.