| Summary: | In this research, up to 97% w/w and 65% w/w of the proteins in the laboratory and industrially defatted sunflower meals, respectively, were extracted using aqueous alkaline solutions at pH 10.0. Up to 84% of these soluble proteins were recovered as solids when the pH was lowered from 10.0 to 4.0 (isoelectric pH) using aqueous HCl acid. The amount of proteins recovered as solids decreased with decreasing concentrations of the acid. Three types of precipitator configuration were used in this study: the batch, mixed suspension mixed product removal (MSMPR), and the tubular precipitators. Factors affecting the particle size distributions and the solids yield were studied. The batch precipitator was found to produce precipitates with high solids concentrations (yield), but small mean panicle sizes $({<} 10\ \mu m)$ with narrow spread in the particle size distribution (PSD). The MSMPR precipitator produced low solids yields and medium particle sizes (ca. 10 to 20 $\mu m)$ with a wide spread in the PSD. High concentrations of the protein in the feed solution resulted in bimodal PSD. The tubular precipitator operating in turbulent flow regime produced precipitates with small particles and a narrow size distribution similar to the batch precipitator. In the laminar flow regime, the tubular precipitator produced large particles (aggregates) which increased in size with increases in the mean residence time and protein concentration in the feed solution. These sizes increased from ca. $10\ \mu m$ to ca. $40\ \mu m.$ The solids yield for the tubular precipitator increased with mean residence time at both flow regimes. Solids yields in the laminar flow regime were as high as those obtained from the batch precipitator. Solutions of the population balance equations describing the batch and MSMPR precipitators have been developed extensively for several precipitating systems. A model (population balance equation) equation describing the precipitation of sunflower proteins in the tubular precipitator was developed. The developed model equations were used to predict the experimental data (mean particle size) from the tubular precipitator. The differences between the model predictions and the experimental data were about 25% for particles with small mean sizes $({<}10\ pm)$ and less for large particles. With these model parameters, it is possible to predict the length of the tubular precipitator required to obtain precipitates with the desired PSD and solids concentration. The required inputs to the model are the precipitator geometry (diameter of the tubular precipitator and the length) and the operating conditions (flow velocity, protein feed concentration, and starting PSD at a known location close to the static mixers). (Abstract shortened by UMI.)
|
|---|
