What cohesive forces are involved in the gelation of casein micelle dispersions at high concentration ?

Concentration operations like membrane filtration or evaporation play a key role in skimmed milk processing. All these processes involve the transition from a liquid to a gel of casein micelles, the major protein in milk. However, still little is known about this peculiar phase behavior at elevated concentrations. In recent works, we demonstrate that transition occurs when the micelles are getting into contact, i.e., at close-packing (C » 200 g/L) [1]. We also have evidences that the micelles stick one to another when in contact and form connections that are strong enough to make the gels partly "irreversible" [1-2]. This work aims at identifying the nature of these cohesive forces. Answering this crucial question could lie to the improvement of existing dairy processes as well as to the development of new operations. Our general approach consists in looking at the possibility for the gels of casein micelles to reswell or redisperse, as a function of time and in different conditions (specific solvents and/or temperature). Such "reversibility tests" allows getting direct information about the forces that oppose reswelling and redispersion; which are the forces we are interested in. Experimentally, this first consists in making concentrated gels of casein micelles. For that purpose, the osmotic stress technique was used (see [1]), and gels of concentrations ranging from ~200 to ~800 g/L were made. We thus covered different states of compression: at ~200 g/L the micelles are still undeformed but into contact (close-packing), while they have been compressed 4 times at ~800 g/L. The gels were then "decompressed" through immersion in various solvents (pH, ionic strength) and at various temperatures, thus allowing them to reswell, and redisperse (partly or totally). After different times of decompression, the residual gels were analyzed in terms of irreversibility (IRR, expressed as the mass of casein in the remaining gel over its initial mass). Meanwhile, the caseins and minerals that were released in the solvent during decompression were analyzed in terms of concentration and properties (such as the average size, as determined through Dynamic Light Scattering). Clearly, the results obtained give original and pertinent information about the forces responsible for the cohesion of the gels. As an example, when native solvent is used as the decompression solvent (= milk ultrafiltrate), we found that the gels obtained at ~200 g/L redisperse in a few hours and almost entirely (IRR = 20%). Conversely, the gels obtained at C > 300 g/L are close to be totally irreversible (IRR = 95%) at low decompression times and start to slowly redisperse after 16 hours and more. This suggests (i) that the micelles need to be directly compressed to form enough bonds between them, and (ii), that those bonds have a limited life-time. Quite surprisingly, our results also indicate that there is an "optimum" of cohesiveness at 400 g/L, a gel concentration at which IRR > 80%, even after 10 days of reswelling. In the presentation, we will also present the results obtained in other solvents and temperatures. Then we will give some first conclusions about the nature of the forces involved in the gelation process