Drying of healthy foods : from mechanism to optimization

Convective drying is an effective post-harvest method to the extend shelf life of vegetables and to reduce the mass for transportation. The heat load during drying, however, affects the quality attributes negatively. Today, consumers in the industrialized world pay a raised attention on food quality, and especially to the nutritional value of food.&nbsp; This increased demand on quality has become a challenge for drying research, and to retain the nutritional value, mild drying conditions must be applied. However, at these conditions the energy efficiency to remove the water from the product by evaporation is low; often below 50%. Moreover, due to the growing global market for dried products, drying contributes more and more to the global energy consumption and CO₂-emission.&nbsp; Hence, there is a need for high energy efficient drying methods with low CO₂-emission. A straightforward solution to increase drying energy efficiency is high temperature drying, but these conditions are conflicting with the aim to retain nutritional components. To combine these two aims, i.e. energy efficient drying and retention of nutritional components, is a challenge for drying research.&nbsp; In this thesis work, the conflict between quality retention and energy efficiency is investigated for the drying of broccoli. The approach in this work is based on mechanistic driven drying modeling and optimization. The approach includes three crucial elements: 1) mechanistic driven model development, 2) model validation, and 3) mechanistic model assisted optimization. In the first parts of the thesis advanced mechanistic drying models are introduced. These are the Free Volume theory for moisture transport (<strong>Chapter 2</strong>), and the Flory-Huggins Free Volume theory to describe sorption isotherms (<strong>Chapter 3</strong>). The strength of these theories is that the mobility of water is based on the changes in physical state during drying (from rubbery to glassy state) and the mixing properties of water, biopolymers and solutes. These mechanistic models allow the extrapolation of the drying behavior to not experimentally validated conditions. Moreover, the model parameters have a physical basis and can directly be related to material properties.&nbsp; The drying behavior in broccoli is represented by 2D color maps to visualize the spatial distribution of moisture content, the progress of degradation of nutritional components in the product and shrinkage during drying. The influences of pre-treatments are also incorporated in the models. The second part of the thesis concerns experimental validation of the models from <strong>Chapter 2</strong> and <strong>Chapter 3</strong>. In <strong>Chapter 4 </strong>moisture transport during drying of broccoli is monitored with MRI (Magnetic Resonance Imaging) as a non-destructive technique. The results show the spatial distribution of moisture content, shrinkage and drying rate of differently pre-treated samples during drying. The images revealed non-Fickian diffusion behavior for fresh stalks. The non-Fickian diffusion is caused by the moisture transport resistance of the stalk skin which creates, together with shrinkage, center directed stress driven moisture transport. This phenomenon was absent in pre-treated broccoli samples for which the resistance for moisture transport in the skin was reduced.&nbsp;&nbsp; The drying rates for broccoli florets and stalks are derived from MRI data in <strong>Chapter 5</strong>. The Free Volume theory for moisture transport is validated on the average moisture content from the MRI experiments. The fitting parameters are the mass transfer coefficient and the self-diffusion coefficient of solids. The results quantify the enhanced drying rates of fresh and pre-treated samples due to the removal of the transport resistance in the skin and the changed cell structure. The influence of pre-treatments on the drying rate is in line with the results of <strong>Chapter 4</strong>. Comparison of experiments in a pilot dryer showed a good agreement with the drying behavior in the MRI device. <strong>Chapter 6</strong>concerns dynamic optimization to derive optimal drying trajectories that increase both energy efficiency and retention of nutritional components during drying of broccoli. For this step it was necessary to derive, from the spatial model, a drying model for the average moisture content and average value of the nutritional components. The kinetics for the degradation of glucosinolates and vitamin C (obtained from a parallel project) and the drying rate of broccoli are applied to calculate optimized drying trajectories for the control variables of temperature and air flow rate. The results have shown that with optimal trajectories the energy consumption can be halved, the vitamin C retention can be increased significantly, and the influence of drying on the degradation of glucosinolates is reduced to nearly zero. The optimized drying trajectories are plotted in an isokinetic temperature-moisture content state diagram which shows that the product areas with high degradation rates are circumvented. Finally, in <strong>Chapter 7 </strong>the contribution of the thesis work and the impact on drying research and the perspectives are discussed. The mechanistic driven drying modeling and optimization approach to produce healthy dried food is regarded as a fundamental approach which uses physical and chemical properties of the product. The advantage of the approach is the potential for application to a large range of processing conditions. The isokinetic temperature-moisture content state diagram, which gives a direct overview of possible pathways to retain heat sensitive components, is a powerful tool to support decision making in multi-objective problems in food process design. This thesis work is an important step in mechanistic modeling and optimization, but the end of this approach is not yet reached. Further adoption of the proposed methodology of monitoring and modeling transport phenomena and degradation of micronutrients in food matrices is believed to advance the quality of food products