INNOVATIVE APPROACHES AND INSTRUMENTS IN MODELLING AND MONITORING THE SHELF LIFE OF PACKAGED PERISHABLE FOODS

Shelf life models are mathematical equations that describe the relationship between food, package, and environment. Shelf life models are useful to predict the shelf life of food, to design food packages and to provide useful information. The accuracy by which these relationships are measured influences the reliability of the models. It is also necessary to know the relationships among the variables that play a significant role in defining food quality. Many chemical, biological, and physical changes are dependent on moisture content or water activity in foods, whose changes may be dictated by the protection from the environment given by packaging in the distribution channel. Water activity is ruled by the moisture content in the food, which may change with time through the permeation process of water vapor across the packaging film. The equilibrium relationship between moisture content and water activity is explained by the sorption/desorption isotherm, which is obtained mostly by experiments determining moisture contents of the food samples equilibrated under different relative humidity conditions. Many attempts were conducted to express the equilibrium relationship by mathematical forms. The most widely used are the equations of Smith, Halsey, Henderson, Oswin, Inglesias, and Chirife. These equations are used in forecast models to determine the shelf life of moisture depending products. All these models assume that the relative moisture of the headspace of the packaging (which changes with the entrance of water vapour from the outside) is immediately in equilibrium with the product. This condition has not been verified experimentally and is based on theoretical considerations. The aim of this PhD thesis is the study of innovative approaches and instruments to model and monitor the shelf life of packaged perishable foods. The crucial part of this work is to verify the theoretical assumption of immediate equilibrium assumed by moisture depending prediction models. To achieve this result it has been developed and built an acquisition system based on moisture sensors, a recording unit and specific software to collect data. The system has been tested on different products and packages in order to define the proper materials and working conditions to define the experimental plan and the system settings. A model system based on “Fetta biscottata” (a typical Italian bakery product chosen because of its regular shape, geometry, and uniformity among different batches) packaged in rigid EPS trays of different volumes has been chosen. This model system allowed to study how different quantities of product, different headspace volumes, and different packaging solutions influence the equilibrium between the relative humidity of the headspace and of the product and how this differences influence the shelf life prediction. The model system studied in the present work has demonstrated that, in this case, the theoretical assumption generally used in the shelf life modelling, stating the immediate equilibrium between the headspace and the product relative moistures, is not verified. These relative moistures are not immediately in equilibrium but show an increasing difference correlated to the increase of the headspace volume/product weight ratio. This correlation has been studied and it has been described by a linear equation in the considered range. The application of a shelf life prediction model to the studied system has shown that a correction is needed to avoid an underestimation of the shelf life. The application of this correction will be particularly interesting for the shelf life modelling of moisture depending products characterized by packaging solution with high headspace volumes. In a second part of the work it has been studied the possibility to simplify the shelf life prediction for a category of products applying a unique shelf life model for all the products part of the category. The shelf life prediction is a labor and time intensive work, a simplification might be useful in real case studies, especially if the category includes high quantities of references. The study on 14 different types of “Fetta biscottata” has demonstrated that it is possible to simplify the work only with simplifications based on a good knowledge of statistical indexes and their correct application.