Investigation of preparation methods on surface/bulk structural relaxation and glass fragility of amorphous solid dispersions

Abstract

The objective of this study was to investigate the effect of preparation methods on the surface/bulk molecular mobility and glass fragility of solid dispersions. Solid dispersions containing indometacin and PVP K30 were chosen as the model system. An inverse gas chromatography method was used to determine the surface structural relaxation of the solid dispersions and these data were compared to those for bulk relaxation obtained by DSC. The values of τ^β for the surface relaxation were 4.6, 7.1 and 1.8 h for melt quenched, ball milled and spray dried solid dispersions respectively, compared to 15.6, 7.9 and 9.8 h of the bulk. In all systems, the surface had higher molecular mobility than the bulk. The glass fragility of the solid dispersions was also influenced by the preparation methods with the most fragile system showing the best stability. The zero mobility temperature (T₀) was used to correlate with the physical stability of the solid dispersions. Despite having similar T_g (65 °C), the T₀ of the melt quenched, ball milled and spray dried samples were 21.6, −4.2 and 16.7 °C respectively which correlated well with their physical stability results. Therefore, T₀ appears to be a better indicator than T_g for predicting stability of amorphous materials.

Introduction

The difficulty in delivery of poorly soluble drug substances can be addressed by use of amorphous forms, but there are concerns with respect to the stability of such systems, with a risk of crystallisation causing a slowing of dissolution rate as a function of storage time. Physical stability testing of amorphous forms is problematic as storage at stressed conditions can be misleading as the sample can move to the rubbery rather than glassy state, making extrapolation difficult. Real time storage is more reliable, but obviously it is time consuming. Therefore, some alternative methods have been used to try to predict the stability of amorphous materials. Among these methods, relaxation assessment has been conventionally used since it can provide useful information about the molecular mobility of the product and therefore its shelf-life.

Some studies on polymers have shown that the molecular mobility at the surface should be differentiated from the bulk (Tanaka et al., 1996, DeMaggio et al., 1997, Kajiyama et al., 1997), since surface mobility of polymers has important implications for friction, lubrication, adhesion and any applications involving polymer modification by way of coatings (Fakhraai and Forrest, 2008). The study of surface molecular mobility for pharmaceutical materials is also of importance because the surface properties can affect the surface wetting, diffusion and interaction between different materials. A number of studies have indicated the necessity of studying surface crystallisation of amorphous materials. For example, Wu and Yu (2006) found that the surface of indometacin crystallised at a faster rate than the bulk. Wu et al. (2007) investigated a method of using nano-coating to inhibit surface crystallisation of indometacin. Zhu et al. (2010) studied the crystal growth rates of amorphous griseofulvin below its glass transition temperature (T_g) and it was found that its surface crystallisation dominated the overall crystallisation kinetics of the amorphous powders. It was concluded by Zhu et al. (2010) that surface crystallisation should be distinguished from the bulk.

Relaxation studies may become a fast and useful way by which to predict the crystallisation tendency of amorphous materials and as such provide valuable information for use during formulation design and product manufacturing. There are several techniques that have been used for measurement of bulk relaxation such as dielectric relaxation spectroscopy, isothermal calorimetry, and differential scanning calorimetry (DSC) (Caron et al., 2010); however, none of these techniques can be used to study surface relaxation. In the pharmaceutical field, the lack of an efficient technique for fast determination of surface relaxation has limited the understanding of the molecular mobility at the surface. Inverse gas chromatography (IGC) has been widely used to measure isotherms, surface free energy, heat of sorption and surface heterogeneity (Newell et al., 2001, Roubani-Kalantzopoulou, 2004). IGC can also be used for some specific applications such as differentiating small but functionally significant levels of amorphous content between samples and measuring glass transition temperature as a function of relative humidity or temperature (Surana et al., 2003, Buckton et al., 2004). In a previous study, the authors have developed an inverse gas chromatography method using the retention volume of decane to study the structural relaxation at the surface of an amorphous solid dispersion system and it was found that IGC can be a potential powerful technique for surface molecular mobility measurement (Hasegawa et al., 2009).

The decane retention volume method was therefore applied in this study to investigate the surface relaxation of amorphous polyvinylpyrrolidone (PVP) and indometacin solid dispersions prepared using different methods (ball milling, spray drying and melt quenching). Previous studies have shown that the physical properties of an amorphous material could be greatly affected by the way it was prepared (Surana et al., 2004), but the understanding of the effect of preparation methods on solid dispersions is limited (Patterson et al., 2007). Moreover, the effect of preparation methods on the surface molecular mobility is not known. It is understood that amorphous materials can have varying amounts of molecular order, but that they lack the long range packing order that is a characteristic of a crystal. The hypothesis in this study is that since different preparation methods can generate amorphous materials of different structure, surface and bulk properties of the amorphous forms will vary independently to each other as a function of preparation method. The aim was to compare the surface relaxation, measured using IGC, with the relaxation of the bulk measured using DSC, and to see to what extent these two relaxation rates predicted the long term stability of otherwise identical amorphous materials made using different methods.

Apart from molecular mobility, other parameters, such as T_g are often used to attempt prediction of the physical stability of an amorphous solid. In order for an amorphous form to have good stability it is necessary for it to have a high T_g. When the storage temperature is too close to or even higher than the T_g, the glassy sample will transform into the rubbery state which gives rise to physical collapse and crystallisation (te Booy et al., 1992). Although a high T_g is important, it is still difficult to predict the stability of amorphous materials using T_g alone. Many studies have been conducted in the past by different researchers to find out some other useful predictors for good amorphous stability. Apart from molecular mobility which is considered as the main factor governing physical stability (Andronis and Zografi, 1997, DiMartino et al., 2000), the relationship of thermodynamic factors such as entropy, enthalpy and Gibbs free energy to amorphous stability was studied in some other cases (Zhou et al., 2002, Marsac et al., 2006). Zhou et al. (2002) found that together with molecular mobility, configurational entropy could help to predict the physical stability of amorphous pharmaceuticals. In another study, Marsac et al. (2006) found that nifedipine which had a larger enthalpic driving force for crystallisation and lower activation energy for nucleation crystallised more readily than felodipine. Graeser et al. (2009) studied configuration entropy in relation to amorphous stability above T_g, however due to the non-equilibrium nature of the glassy state, thermodynamic considerations cannot be easily applied below the T_g. As the name implies, the zero mobility temperature (T₀) is the temperature at which molecular mobility would effectively stop in an amorphous sample, and is the point where the temperature has been decreased sufficiently below T_g, that the configuration entropy will reach zero (Hatley, 1997). Therefore, T₀ would be the highest storage temperature for amorphous materials at which the crystallisation tendency would be minimised. In this study, the glass fragility of the solid dispersion system prepared by different methods was measured and the fragility values were used to determine T₀. The extent to which T₀ was able to predict the stability of the amorphous form was assessed by comparison to long term stability storage tests.