Starch addition in renneted milk gels: partitioning between curd and whey and effect on curd syneresis and gel microstructure.

Milk gels were made by renneting and acidifying skim milk containing 5 different starches, and then compressed by centrifugation to express whey and simulate curd syneresis during the manufacture of low-fat cheese. A series of 17 starches were examined, with 5 starches being selected for in-depth analysis: a modified waxy corn starch (WC), a waxy rice starch (WR), an instant tapioca starch (IT), a modified tapioca starch (MT), and dextrin (DX). Milks containing WC, WR, and DX were given a 72°C heat treatment, whereas those containing IT and MT had a 30-min treatment at 66°C that matched their optimum gelatinization treatments. Curd yields were calculated by weight, estimated starch content in whey was measured gravimetrically by alcohol precipitation, and starch retention in curd was calculated. Curd yields were 13.1% for the control milk (no added starch) and 18.4, 20.7, 21.5, 23.5, and 13.2% for the gels containing starches WC, WR, IT, MT, and DX, respectively. Estimated starch retentions in the curd were, respectively, 71, 90, 90, 21, and 1%. Laser scanning confocal microscopy was used to determine the location of the starches in the curd and their interaction with the protein matrix. Waxy corn, WR, and IT starches have potential to improve texture of low-fat cheese because they had high retention in the curd and they generated interruptions in the protein matrix network that may have helped limit extensive protein-protein interactions. Modified tapioca starch interfered with formation of the protein structure of the curd and produced a soft noncohesive gel, even though most (79%) of the MT starch was lost in the whey. Few distinct starch particles were present in the MT curd network. Dextrin was not retained in the curd and did not disrupt the protein network, making it unsuitable for use in low-fat cheese.