The Kjeldahl and Dumas methods for quantifying nitrogen content were compared using nine soybean products having protein contents ranging from 0.5 to 90%. In addition to comparing day-to-day variability of the Dumas method, differences between and variabilities of two Kjeldahl systems and Kjeldahl operators were also evaluated. The Kjeldahl method gave slightly, but significantly, lower values than did the Dumas method. Both the Kjeldahl and Dumas methods had equivalent variabilities (same SD about the means). The ratios between the means for the Kjeldahl and Dumas (K/D) protein values ranged from 0.66 to 1.03. The conversion equation y = −0.00536 + 0.97188x (R 2 = 0.9997) was developed and validated to convert from Dumas to Kjeldahl protein concentrations.

Glycinin and ␤-conglycinin have unique functionality characteristics that contribute important properties in soy foods and soy ingredients. Limited functionality data have been published for glycinin and ␤-conglycinin fractions produced in pilot-scale quantities. Protein extraction conditions were previously optimized for our pilotscale fractionation process to maximize protein solubilization and subsequent product recovery. Glycinin, ␤-conglycinin, and intermediate (mixture of glycinin and ␤-conglycinin) fractions were prepared using optimizedprocess (OP) extraction conditions (10:1 water-to-flake ratio, 45°C) and previous conditions termed Wu process (WP) (15:1, 20°C). Viscosity, solubility, gelling, foaming, emulsification capacity, and emulsification activity and stability of the fractionated proteins, and soy protein isolate (SPI) produced from the same defatted soy white flakes were compared to evaluate functional properties of these different protein fractions. Differential scanning calorimetry, sodium dodecylsulfate-polyacrylamide gel electrophoresis, and surface hydrophobicity data were used to interpret functionality differences. OP ␤-conglycinin had more glycinin contamination than did the WP ␤-conglycinin. OP and WP solubility profiles were each similar for respective glycinin and ␤-conglycinin fractions. Emulsification activities and stabilities were higher for OP ␤-conglycinin and OP intermediate fractions compared with respective WP fractions. ␤-Conglycinin and SPI emulsification capacities (ECs) mirrored solubility profile, whereas glycinin ECs did not. OP glycinin had a higher foaming capacity than WP glycinin. OP and WP intermediate fraction apparent viscosities trended higher than those of other protein fractions. ␤-Conglycinin dispersions at pH 3 and 7 produced firm gels at 80°C, whereas glycinin dispersions formed weaker gels at 99°C and did not gel at 80°C.

No abstract

Soy protein isolates (SPIs) are produced industrially using methods using various protein extraction temperatures and pH values. The effects of process temperature (25 °C or 60 °C) and pH (8.5, 9.5, or 10.5) on isoflavone and group B saponin extraction, partitioning, and profile during bench-scale SPI production were evaluated. Protein, isoflavone, and saponin extraction increased with increasing temperature and pH. Substantial quantities of isoflavones and saponins remained in the insoluble fraction waste stream. Isoflavones were also lost to the whey waste stream, whereas saponins were not detected in the whey. Neutralization of SPI samples at the time of analytical extraction increased measured isoflavone concentrations for the pH 8.5 process extraction treatments, whereas neutralization substantially increased measured saponin concentrations in SPIs for all process extraction treatments. Malonylglucoside isoflavones were primarily converted to ␤ ␤ ␤ ␤ ␤-glucoside forms as temperature and pH were increased, and these conditions caused conversion of ␣ ␣ ␣ ␣ ␣g, ␤ ␤ ␤ ␤ ␤g, and ␤ ␤ ␤ ␤ ␤a saponins to saponin V, I, and II forms, respectively. Analytical extraction pH should receive careful consideration when analyzing soy matrices for isoflavones and saponins.