Abstract
Mathematical models for the prediction of antioxidant distribution in whole milk were derived and verified. These distribution equations allow prediction of the division of added antioxidant between the oil and water phases and the amount bound to the proteins of milk. Simple considerations of oil/water partitioning and protein binding are involved in the predictions. For systems where the oil is in the form of a continuous bulk phase, the agreement between experiment and theory is good. The protein binding exerts a marked influence on the distribution of antioxidants between the aqueous and oil phases. For emulsions, where the lipid is finely divided globules, the experimental results show higher than anticipated antioxidant content in the oil phase, suggesting sorption of the antioxidant at the greatly expanded oil/water interface. The results are consistent with the known concentration dependence of sorption phenomena and the size of the antioxidant molecule. The nature of the interfacial complex was not determined. Agreement between theory and experiment for antioxidant in the fat phase of whole milk powder was reasonable. The extent of antioxidant sorption at the fat/solid matrix interface of whole milk powders was not determined.