Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Food Science and Human Nutrition

First Advisor

Patricia A. Murphy


Processing effects on soy storage proteins, glycinin and beta-conglycinin, as well as isoflavones and group B soyasaponins, protein-associated phytochemicals, were evaluated. It was hypothesized that the physicochemical characteristics of these molecules, such as protein native state, hydrophobicity, or ionization, would affect their partitioning during processing and, for the proteins, affect functional properties.;A pilot-scale soy protein fractionation process, yielding glycinin, beta-conglycinin, and an intermediate fraction comprised of a mixture of these proteins, was scaled up from 15 kg to 50 kg of defatted soy white flake starting material in order to produce larger quantities of protein products. The scaled-up process recovered 12% less of the initial protein in the product fractions compared to the 15-kg process, but 2.4 times more total protein were recovered in the scaled-up process product fractions. The insoluble fraction, a waste stream, retained 59, 67, and 75% of the protein, isoflavones, and saponins, respectively. The whey contained 6% of the protein and 12% of the isoflavones, but almost no saponins. The saponin to isoflavone ratio was about 2:1 in the beta-conglycinin, but about 1:1 in the glycinin. These data demonstrate differences in phytochemical partitioning, possibly due to physicochemical characteristics such as hydrophobicity and ionization. The isoflavone profile shifted from malonylglucosides toward aglucon forms, possibly due to the action of native beta-glucosidases.;Solubilization temperature and pH influenced protein and phytochemical recovery and partitioning during bench-scale soy protein isolate production. Increasing temperature from 25 to 60°C and pH from 8.5 to 10.5 increased extraction of phytochemicals from the soy flakes. Saponin concentration increased in the isolate. Neutralizing samples prior to phytochemical extraction significantly increased measured saponin concentrations in the isolate and measured isoflavone concentrations in the insoluble fraction. Increasing temperature and pH shifted isoflavone profile from malonylglucosides to glucosides and saponin profile from DDMP saponins to non-DDMP saponins.;Extraction pH, ethanol concentration, water-to-soy flake ratio, and temperature were evaluated during bench-scale protein fractionation for ability to improve protein and photochemical solubilization and recovery. A 10:1 water-to-flake ratio and 45°C were selected as optimized solubilization conditions. When compared to the control (15:1 water-to-flake ratio and 20°C) under pilot-scale conditions, the optimized method produced more beta-conglycinin, but at the expense of purity. The optimized method displayed increased saponin extraction from the soy flakes. A portion of the malonylglucoside isoflavones were converted to beta-glucoside and aglucon forms.;Protein products from the optimized and control pilot-scale fractionation processes were compared for functionality. Optimized process intermediate and beta-conglycinin fractions, which had higher surface hydrophobicity than the corresponding control fractions, exhibited significantly better emulsification activity compared to those for the control process. Because the beta-conglycinin fractions had a lower denaturation temperature than the glycinin fractions, and denaturation of the protein is required as a first step in gel formation, the beta-conglycinin fractions gelled at a lower temperature than did glycinin. The beta-conglycinin fractions made firmer gels than did glycinin fractions at pH 3 and 7. The intermediate fractions, which contained little native-state protein, demonstrated higher apparent viscosities than the soy protein isolate and other process fractions, probably due to the increased molecular interaction of the unfolded, non-native state protein.



Digital Repository @ Iowa State University,

Copyright Owner

David Alan Rickert



Proquest ID


File Format


File Size

244 pages