, 2007 and Kawabata et al., 2011). A higher degree of prediction and precision in decision making would enable more efficient drug product development and provide an early stage insight into the potential of solubility limited drug compounds to be processed into functional and stable dosage forms. In this context, it is necessary to develop methods that can predict the solid state behaviour of drug compounds during processing and manufacturing. Solid state alterations, in particular amorphization, often have significant influence on the performance

of a substance, impacting for instance mechanical properties (Ziffels and Steckel, 2010), dissolution (Lindfors et al., 2006 and Murdande et al., 2010) and bioavailability find more (Hancock and Parks, 2000). Amorphization is hence a strategy with high potential to increase bioavailability of compounds for which poor solubility is limiting intestinal absorption. However, as the inherent instability of the amorphous state limits production, handling and use of products based on amorphous compounds, research efforts are currently directed towards methods that stabilize the amorphous phase (Kearns

et al., 2008 and Laitinen et al., in press). Fundamental aspects governing the physical stability, i.e. the resistance of an amorphous compound to be transformed into its crystalline PI3K inhibitor state, has lately been in focus with the purpose

to obtain an increased understanding of the dynamics (Aso et al., 2001, Bhattacharya and Suryanarayanan, 2009, Singh and de Pablo, 2011 and Stukalin et al., 2009) and nucleation processes (Marsac et al., 2006 and Vyazovkin and Dranca, 2007). Thermodynamically the physical stability is governed by the to difference in Gibbs free energy between the amorphous and the crystalline states. Both nucleation rate and crystal growth is however also affected by the dynamics, i.e. the molecular mobility, of the amorphous phase. The glass transition temperature (Tg) has therefore been used as a reference temperature when determining glass-formation temperatures ( Corrigan et al., 2004 and Yamaguchi et al., 1992) and storage temperatures ( Hancock et al., 1995 and Schoug et al., 2009). However, the predictive capacity of Tg for physical stability has been shown to be poor, which is manifested, by for instance, the observation that compounds with similar Tg may have different amorphous stability ( Marsac et al., 2006), and that alterations in amorphous stability attained by variations in production settings not always are reflected in observable changes of Tg ( Yamaguchi et al., 1992 and Zhang et al., 2009). Some recent publications have described the use of statistical methodology to find other physicochemical properties that correlate with glass-forming ability and glass stability.