Chemdeg, an R package for the analysis of foods isothermal degradation kinetics

Accurate parameter estimation is an important issue for the quantitative analysis of the degradation kinetics of foods and of foods components. Nonlinear regression of time-dependent data with model equations derived from chemical kinetics is the general approach to this purpose, but in some cases, above all when the data are noisy or experimental error has not been measured, it may lead to draw imprecise or even erroneous conclusions. We discuss these points and we describe a formal procedure for the appropriate evaluation of kinetic parameters. When standard nth-order kinetic models fail to explain experimental data the versatile Weibull equation is commonly used for data fitting. We discuss why in our opinion the Weibull model is not always suitable and we propose alternative models that are more consistent with the chemistry underlying the stability of foods and of their components. We have developed the R package chemdeg to objectively support the whole process of model selection and parameter estimation for the analysis of degradation kinetic data.