Optical prediction models of whey protein denaturation in thermally treated milk for the development of an inline sensor
- Autores: Heather Taterka
- Directores de la Tesis: Manuel Castillo Zambudio (dir. tes.)
- Lectura: En la Universitat Autònoma de Barcelona ( España ) en 2016
- Idioma: inglés
- Tribunal Calificador de la Tesis: Federico Harte (presid.), Reyes Pla Soler (secret.), Colette Fagan (voc.)
- Programa de doctorado: Programa de Doctorado en Ciencia de los Alimentos por la Universidad Autónoma de Barcelona
- Materias:
- Enlaces
- Resumen
An inline whey protein denaturation sensor would be of interest to the dairy industry to monitor milk batch variations and to achieve the highest quality products. It has been well-established that whey protein denaturation is a pH-dependent mechanism, in which proteins at lower pH values (pH 6.3) tend to form complexes with κ-casein on the surface of the casein micelle, and at higher pH values (pH 7.1) the preference is for unfolded whey proteins to for serum complexes, in general, with other denatured whey proteins. The objective of this PhD was to develop successful prediction models of whey protein denaturation variables utilizing an optical sensor set-up with the potential for inline implementation during thermal processing. The optical sensor system was developed with inline implementation in mind, with the goal being to measure the effects of temperature, pH and time on the changes in light scatter of thermal treated skim milk and relate these changes to the denaturation of whey proteins. Variables to be compared to the optical light backscatter response were particle size and the whey protein concentration of the three whey protein configurations that occur in milk after thermal treatment: native, micelle-bound and soluble aggregate whey protein. In the second and third experiments, tryptophan front-face fluorescence spectroscopy was also tested with the potential for sensor development and compared to light backscatter technology.
Results of the first experiment showed a relationship between light backscatter intensity and particle size, in particular at pH 6.3 whereas at pH 7.1 no notable changes in the light backscatter intensity or particle size were observed with an increasing in heat treatment temperature. In the second experiment, curves of LB and FFF intensity versus time at pH 6.3 resembled curves of particle size and bound whey protein, and their first-order kinetic rate constants were not statistically different. The third experiment included a range of fat percentages (<0.5%, 1.3% and 3.7%) and exhibited a noticeably greater amount of light scatter and larger particle size with increasing fat content. Model equations showed successful predictions of particle size as a function of light backscatter. In the second experiment, models of bound whey protein at pH 6.3 were best fit to models as a function of the light backscatter spectra, whereas soluble aggregate whey protein content showed best fit when using tryptophan fluorescence measurements. Light backscatter regions which corresponding to best-fit models for particle size and bound whey protein models were near the maximum intensity wavelength (540-600 nm) or included a ratio combination of a numerator value between 387-569 nm and denominator from 963-1033 nm. Front-face fluorescence models also exhibited good R2 values near the maximum intensity wavelength, however a ratio of numerator near 340 nm combined with a denominator around 390 nm yielded models with a better fit. An interesting finding was the relationship exhibited by particle size models as a function of light backscatter, which exhibited an exponential character using an equation with the intercept value similar to the initial particle size. Combined models over a range of pH values (6.3, 6.7 and 7.1) predicted particle size as a function of light backscatter, giving promise to the development of an optical inline backscatter sensor technology.
