Predicting composition–property relationships for glass ionomer cements: A multifactor central composite approach to material optimization

Adjusting powder–liquid ratio (P/L) and polyacrylic acid concentration (AC) has been documented as a means of tailoring the handling and mechanical properties of glass ionomer cements (GICs). This work implemented a novel approach in which the interactive effects of these two factors on three key GIC properties (working time, setting time, and compressive strength) were investigated using a central composite design of experiments. Using nonlinear regression analysis, formulation–property relationships were derived for each property, which enabled prediction of an optimal formulation (P/L and AC) through application of the desirability approach. A novel aluminum free GIC was investigated, as this material may present the first clinically viable GIC for use in injectable spinal applications, such as vertebroplasty. Ultimately, this study presents the first series of predictive regression models that explain the formulation-dependence of a GIC, and the first statistical method for optimizing both P/L and AC depending on user-defined inputs.