Predicting the glass transition temperature of bioactive glasses from their molecular chemical composition

A recently published paper (M.D. O'Donnell, Acta Biomaterialia 7 (2011) 2264-2269) suggests that it is possible to correlate the glass transition temperature (T(g)) of bioactive glasses with their molar composition, based on iterative least-squares fitting of published T(g) data. However, we show that the glass structure is an important parameter in determining T(g). Phase separation, local structural effects and components (intermediate oxides) which can switch their structural role in the glass network need to be taken into consideration, as they are likely to influence the glass transition temperature of bioactive glasses. Although the model suggested by O'Donnell works reasonably well for glasses within the composition range presented, it is oversimplified and fails for glasses outside certain compositional boundaries.