Dry fractionation and bioprocessing for novel legume ingredients

Dry fractionation can be used to prepare protein enriched legume ingredients through dry milling and dry separation. Depending on the structure and composition of the legume, air classification or electrostatic separation or their combination can be employed. However, during dry fractionation the protein fraction is also enriched in antinutritional factors (ANFs), which requires further mild processing to reduce their concentrations. Therefore, the aim of this dissertation was to develop a sustainable processing route combining dry fractionation and solid-state fermentation to process different legumes into functional protein-enriched ingredients with enhanced nutritional value. The focus is to seek an optimum balance between purity and yield by optimizing the operating conditions of electrostatic separation and explore the potential of dry fractionation for selected legume varieties. The differences in compositions between protein-rich, starch-rich, or oil-rich legumes and the impacts on the selection and configuration of dry separation were discussed. Subsequently, the effect of solid-state fermentation on improving the nutritional and functional properties of the dry-enriched fractions is evaluated. Finally, the potential use of the enriched and fermented legume ingredients for bread making is demonstrated. The changes in the nutritional, organoleptic, and techno-functional properties and their contributions to develop novel bakery products, ethnic foods, imitation products are assessed.

The results showed that soy protein enrichment was achieved by defatting, milling, and electrostatic separation. Both oil pressing and organic solvent effectively removed the oil from soybean, although oil pressing compacted the microstructure of soy meal visually. Moderate impact milling speed (3000 rpm) was observed to effectively liberate protein bodies from the cellular matrix whilst preventing agglomeration of small fragments. Electrostatic separation was evaluated using two different charging tube designs. A higher yield was found after separation with a spiral tube compared to that obtained with a charging slit. The spiral tube provides a longer residence time improving the charging and subsequent separation process. A maximum of 15% of protein enrichment was achieved during electrostatic separation (from 45 to 52 g protein per 100 g dry basis) having recovered 62% of total protein from the defatted soy flour.

Following up the previous study, the impact of charging tube materials and diameter on the separation performances of a gluten-starch model mixture and lupine flour is discussed. The results showed that gluten takes a positive charge and wheat starch takes a negative charge after contact with all the charging materials (aluminium, stainless steel, PTFE, Nylon). The charge of the gluten-starch mixture was however not the same as the sum of the charge of individual components. Additionally, the charge magnitude of pure materials did not reveal any relation to the in literature reported triboelectric series, which is probably related to different charging conditions. For the gluten-starch mixture, different protein enrichment was achieved with different charging materials. For lupine flour, the protein purity increased from 37 to ~65 g/100 g dry matter basis for all used tube materials. Tubes with different diameters showed the largest influence on the separation performance. Overall, the results suggested that particle-particle collisions may be primarily responsible for much of the charging of particles. This explains why charging experiments with pure components do not predict the separation behaviour during electrostatic separation, but it also implies that redesigning the charging system to maximize particle-particle collisions could lead to significantly better charging and thus separation.

Protein enrichment by a two-step dry separation process combining air classification and electrostatic separation was investigated for starch-rich legumes. Yellow pea, lentil, and chickpea were subjected to impact milling at optimized settings. Subsequently, starch granules and fibres were removed from proteins during air classification and electrostatic separation, respectively. This two-step process showed enrichment for pea and lentil but not for chickpea due to the smaller starch granules and higher fat content. Process optimization for pea showed that the pea fine fraction had an optimum balance between protein purity and yield when the air-classifier wheel speed was set at 8000 rpm. The subsequent electrostatic separation was optimized with two passes for pea fine fraction. By recycling fractions from collector bags for a second separation pass, we obtained a protein purity of 63.4% dry basis with a yield of 15.8 g/100 g, leading to a protein recovery of 18.0% from pea.

Chickpea was subjected to air classification and spontaneous solid-state fermentation (SSF). By daily back-slopping at 37 °C dominant autochthonous lactic acid bacteria (LAB) in chickpea flour and dry-enriched fractions were isolated, which included Pediococcus pentosaceus and Pediococcus acidilactici. LAB strains were selected based on their ability to metabolize α-galactosides (raffinose, stachyose, and verbascose).  In the first 24 h during SSF, the pH of chickpea doughs decreased from 6.6 to 4.2. After 72 h, the amount of raffinose and stachyose decreased by 88 and 99%, respectively. The content of phytic acid reduced by 17% and the total phenolic contents increased by 119%. The chickpea sourdoughs showed higher water holding capacity, but a decreased foaming ability. Changes in smell, texture, and colour were also observed.

In line with our previous study, the contents of α-galactosides were reduced in wheat bread fortified with fermented chickpea. Fortified bread had a redder and yellower crust, but fermentation slightly reduced browning during baking probably due to the conversion of reducing sugars during fermentation, and due to the lower pH. Crumb hardness increased as the fortification level increased. Bread crumb with chickpea fractions had a denser structure. The chickpea weakened the gluten network, which led to poorer gas retention and the formation of smaller gas cells in the crumb. No difference between fermented and non-fermented chickpea ingredients was observed. The sourdough bread exhibited better microbiological stability compared to that of unfermented chickpea bread and wheat bread.

The dissertation shows the potential for producing protein concentrates from different legumes by dry fractionation. Additionally, solid-state fermentation was applied to enhance the use of dry-enriched legume ingredients. Based on the findings of this study, further technology development and scaling-up of these processes are expected in the future.