Prediction of the glass transition temperature of water solutions: comparison of different models

The glass transition temperature (Tg) of a sample is an important parameter that determines its stability during storage. While Tg can be measured by a variety of methods, it is a time-consuming procedure, especially if the sample is to be kept at subzero temperatures, in anhydrous conditions, or if sampling a portion of the specimen for analysis is cumbersome. Hence, predicting rather than directly measuring Tg as a function of the content of the constituents of a blend, mixture, or solution can be a powerful tool. Two main models for predicting Tg have been proposed: Couchman-Karasz (C-K) and Gordon-Taylor (G-T) formalisms. However, many aspects of both are theoretical/terminological in nature, and substantial controversy exists about the various experimental approaches to measuring Tg as well. Here, we compare C-K and G-T formalisms, as well as related problems that arise from the variety of definitions and methods of measuring Tg. Water-trehalose solutions are used as an example for application of the analysis. However, the same conclusions can be expanded to any other solutions so thermodynamical parameters (Tg, DeltaCp, and k) of 20 other commonly used solutes are provided. Practical pitfalls related to determining water content, including experimental data on thermal gravimetry, are also discussed.