Monitoring of milk rennet coagulation: Chemical and physical perspective using Raman spectroscopy

The present study applies Raman spectroscopy and chemometrics to monitor the milk renneting process. Raman sensor information (Laser 785 nm) was simultaneously retrieved with rheological measurements to predict the curd cutting time. Principal component analysis was conducted on Raman spectra to collect the relevant information on the coagulating process. Rheological cutting time was modelled using the time at the inflection point of first principal component scores, from which an R² of 0.901 was obtained for calibration purposes, while 0.872 resulted from a 6-fold cross-validation. For the first time, the response of the Raman sensor was also used to assess the dynamic evolution of signal intensity for every individual Raman shift in order to identify which chemical bonds and molecules are significantly involved in the renneting process. To this concern, the role of tryptophan, PO₄³⁻ groups, aspartic acid and phenylalanine results to be important in projecting the original spectra in the principal component subspace. Raman signals can be employed to study the physical behavior of renneting milk and to analyze the changes in chemical bonds, depending on the data pretreatment method applied, specifically in baseline correction. The presented results demonstrate that Raman spectroscopy can be successfully implemented to provide quantitative and qualitative knowledge of the milk coagulation process.