Electrostatic separation for functional food ingredient production

<strong>Summary</strong> Dry fractionation is a promising alternative to wet extraction processes for production of food ingredients, since it uses hardly any water, consumes less energy and retains the native functionality of the ingredients. It combines milling and dry separation to enrich agro-materials in specific components such as protein. Electrostatic separation recently emerged as a novel dry separation process and it relies on electrostatic forces for separation. Though the potential of electrostatic separation to fractionate agro-materials has been demonstrated, the effectiveness in terms of purity and yield and the influence of process parameters on charging and separation of food ingredients have not yet been systematically studied. Therefore, the objective of this thesis was to gain better understanding of the charging and separation behaviour of model and agro-materials, provide insight in the critical factors for successful electrostatic separation and explore the potential of this separation method to different agro-materials. The charging step is critical to the effectiveness of electrostatic separation and is influenced by many factors. <strong>Chapter 2 </strong>presents characterization of the charging behaviour of single-component model particles in nitrogen gas flowing through aluminium tubes, using a lab-scale electrostatic separator. Polystyrene particles and wheat gluten were used as model particles. Higher gas velocities led to a higher specific charge by increasing the normal component of impact velocity. Smaller particles gained more specific charge than larger ones because of their higher surface to volume ratio and their sensitivity towards gas flow pattern changes. Longer charging tube lengths allowed more contact between the particles and the wall and therefore resulted in higher specific charge. The relative humidity of the nitrogen gas flow within the range 0 – 60% had no influence on the charging behaviour of both model particles. <strong>Chapter 3 </strong>demonstrates the potential of applying electrostatic separation to enrich arabinoxylans from wheat bran with the same lab-scale electrostatic separator<strong>. </strong>A combination of larger particle size, higher gas velocity and shorter charging tube was preferred for separation, because it sufficiently charged the particles while agglomeration was minimized. Electrostatic separation with the optimum setting achieved a similar enrichment in arabinoxylans (from 23% to 30% dry matter basis) as sieving does. However, the combination of electrostatic separation and sieving further improved the enrichment and resulted in a fraction with an arabinoxylans content of 43% dry matter basis, which is around the maximum achievable purity that can be reached by dry fractionation. To allow better defined charging and separation experiments, a bench-scale electrostatic separator was designed and constructed. With this custom-built separator, the charging and separation of model mixtures prepared from wheat gluten and starch were studied in<strong> chapter 4</strong>. The net charge of gluten-starch mixtures was not simply the sum of the charge of the two individual components, indicating that particle-particle interactions play an important role. We hypothesized that the formation of agglomerates between oppositely charged particles negatively influenced separation, which was supported by the fact that the dispersibility for mixtures of the two components was lower compared to that of individual components. We found that during electrostatic separation of mixtures, it is important to find the optimal condition that provides sufficient charge to charges, but avoids agglomeration between oppositely charged particles. This could be achieved by the combination of lower dosing rate and higher gas flow rate. <strong>Chapter 5</strong> reports on dry fractionation by combining milling and electrostatic separation with the custom-built bench-scale separator, providing an alternative to wet extraction of protein from lupine seeds. Relatively coarse milling was preferred because it disclosed sufficient protein bodies from the matrix, while avoiding poor dispersibility of the powder due to its very fine particle size. With the optimal settings of single-step electrostatic separation, a fraction with 57.3 g/100 g dry solids could be obtained. The protein content was further improved to 65.0% dry matter basis after two more separation steps, which is 15% higher than obtained by air classification. The yield of the protein enriched fraction was further increased by recycling the fractions from the filter bags, but this was accompanied by a decrease in protein content and vice versa. A significant shift towards better yield and purities was achieved by re-milling the flour that was not collected on the electrodes. A final fraction with a protein content of 65.1% dry matter basis and a yield of 6% was obtained, which recovered 10% of the protein in the original flour. <strong>Chapter 6</strong> explores the possibility of enriching dietary fibre from defatted rice bran by dry fractionation, where the custom-built bench-scale electrostatic separator was used. All three tested separation routes produced fibre-enriched fractions with similar yield (20 – 21 % of the milled flour) and fibre content (67 – 68 % dry matter basis), which recovered 42 – 48 % of the fibre from the original flour. The enriched fractions obtained by a two-step electrostatic separation process contained more small particles compared to the other two, which resulted in different functional properties. Compared to the total dietary fibre extracted by the enzymatic-gravimetric method, the enriched fractions by dry fractionation had a similar water retention capacity and oil binding capacity. This suggests that the fibre-enriched fractions by dry fractionation can be applied in foods and provide similar technological properties and physiological effects as the wet-extracted dietary fibre does. <strong>Chapter 7</strong> concludes the thesis with a general discussion on the main findings, based on which two schemes for protein enrichment and fibre enrichment were proposed. Subsequently the challenges to achieve a successful electrostatic separation for agro-material and up-scaling are discussed. Finally, the chapter ends with an outlook on future research. This thesis provided insight in the key factors for successful electrostatic separation. It demonstrated the potential of applying this separation method for functional ingredient production from different agro-materials and also gave directions for further improvement and scaling-up.