The emulsifying properties of phosvitin dissolved in water and 0.1, 0.5 and 1.0 M NaCl were determined from pH 3 to 10. The change in its emulsifying activity (BA) with pH was slight but significant (~~0.05) and emulsion stability (ES) was relatively high (68-73%), except at pH 5 (17%) and 10 (48%). The EA of phosvitin was higher than that of bovine serum albumin (BSA) at pH 3 or 8 and ES was higher than BSA at all pH levels except at pH 5 and 10. Added NaCl decreased in the EA of phosvitin at pH 3 and 10 and decreased the ES between pH 3 and 9. Increased instability of emulsions resulted mainly in coalescence of oil droplets at NaCl B 0.5M. Salt increased the viscosity of phosvitin emulsion only at pH 3 but not at pH > 5. The viscosities of BSA emulsions were higher than those of phosvitin at pH 3, 5 or 8. effect of pH and sodium chloride on the EA, ES and viscosity of phosvitin emulsions, using BSA as a reference protein. The instability of the emulsions was described in terms of creaming and coalescence. Materials MATERIALS & METHODSPhosvitin and BSA were purchased from Sigma Chemical Co. (St. Louis, MO). Sodium dodecyl sulfate was from Bio-Rad Laboratories (Richmond, CA) and sodium chloride was from Fisher Scientific Co. (Fair Lawn; NJ). Corn oil (lOO%), manufactured by Best Foods Canada, Inc. (Etobicoke, Ontario), was from a local grocery store.

The effects of protein concentration (O.l-2.0%), oil volume fraction (O-17-0.67), mixing speed (lO,OOO-22,000 rpm) and mixing time (OS-8 min) on the emulsifying properties of phosvitin and bovine serum albumin (BSA) were compared. Emulsifying activity and emulsion stability increased with protein concentration, oil volume fraction and mixing. Effects of these variables were assessed quantitatively using an empirical equation. Mixing speed had the greatest influence and protein concentration had the least influence on emulsifying activity for both phosvitin and BSA. For emulsion stability, mixing speed had the greatest influence for phosvitin; oil volume fraction had the greatest influence for BSA. Phosvitin was a better emulsifier than BSA at pH 7.

Egg yolk powder was partly defatted using hexane, 95% ethanol, isopropanol, chloroform-methanol (2~1, v:v) hexane-isopropanol(77:23, w:w) and hexane-ethanol (77:23, wzw). Hexane-isopropanol was most efficient, extracting more than 50% of the lipid. Protein solubility and emulsifying activity of the extracted powder decreased after extraction. Hexane-isopropanol reduced emulsifying activity about 14% and protein solubility to 13.8% from 15.7%. Extraction with other solvents resulted in up to 41% loss of emulsifying activity; protein solubility decreased to as low as 3%. Ovalbumin in the water-soluble fraction of extracted egg yolk powder reduced its emulsifying effectiveness.

MATERIALS & METHODSPhosvitin in water at pH 7 had a denaturation temperature (Td) of 79.7 -t 1.4"C when heated at lO'C/min. When dissolved in O.lM and l.OM NaCl. the T, decreased to 77.7 -+ 1.2"C and 77.2 + 1.3"C. resoectivelv. MaterialsPhosvitin was purchased from Sigma Chemical Co. (St. Louis, MO). I-Menthol and a-naphthol were obtained from Fisher Scientific Co. (Fair Lawn, NJ). Indium was provided by DuPont Instruments (Wilmington, DE). , " / 1,I and in 10 and 20% sucrose there was no change in T,. Heat treatment of phosvitin solutions at 265°C led to decreased emulsifying activity (EA). The emulsion stability (ES) decreased when phosv&nsolutioni were heated at 70, 80 or 96°C for up to 60 min. The ES was not affected (p < 0.05) for phosvitin solutions after heating at 267.5"C for up to 60 min.

A model was developed consisting of a modified Plank's equation to estimate phase change time and two unsteady state cross-product heat transfer equations for estimating precooling and tempering times. It accurately predicted total time to proceed from an initial temperature above the freezing point to a final temperature of -18°C. A correction factor was developed and incorporated in the P term in Plank's equation to correct for the effect of initial and freezer medium temperatures and heat transfer coefficients. The model, tested over a broad range of freezing conditions, had a mean absolute error of 5.9% in predicted values relative to experimental values.