Predicting the amorphous-phase composition during lyophilization

Abstract

The glass-transition temperature and the composition of the amorphous phase/maximally concentrated solution (classically referred to as T_g' and w_g', respectively) as function of added excipients are crucial for the design of lyophilization processes. Whereas the determination of T_g' can be accomplished easily using mDSC, the determination of w_g' poses challenges, since the experimental effort needs to be redone for each new excipient mixture (limited transferability of the results possible). In this work, an approach was developed which allows to predict w_g' for (1) single excipients, (2) given compositions of a binary excipient mixture, and (3) single excipients in aqueous (model) protein solutions using the thermodynamic model PC-SAFT and one experimental data point of T_g'. Sucrose, trehalose, fructose, sorbitol, and lactose were considered as single excipients. The binary excipient mixture consisted of sucrose and ectoine. The model protein was bovine serum albumin in combination with sucrose. The results reveal that the new approach can precisely predict w_g' in the systems considered, including the non-linear course of w_g' identified for different sucrose/ectoine ratios. The same applies to the course of w_g' as function of the protein concentration. This newly developed approach allows for the reduction of the experimental effort to a minimum.

Keywords: Freeze-concentration; Freeze-drying; Glass transition; PC-SAFT; Solubility prediction.