1
Introduction to Amorphous Solid Dispersions

George Zografi1 and Ann Newman2

1University of Wisconsin-Madison, Madison, WI, USA

2Seventh Street Development Group, Lafayette, IN, USA

1.1 Introduction

Over the years one of the major goals of synthetic chemists has been to provide the crystalline form of any active pharmaceutical ingredient (API) being introduced into pharmaceutical development. This is primarily because the symmetrical three-dimensional long-range order and the relatively tight packing of molecules in a crystal lattice most often ensure a high level of chemical purity and solid-state stability. At the same time, an API being developed for oral administration in a solid dosage form generally requires sufficient aqueous solubility upon contact with in vitro and in vivo dissolution media in order to obtain optimal rates of dissolution and acceptable oral bioavailability. The importance of aqueous solubility in affecting dissolution rates can be shown with the classical Noyes-Whitney equation [1]:

where dC/dt is the dissolution rate, kD is the dissolution rate constant (dependent on the stirring rate and the diffusion constant), A is the total surface area of the drug particles, Cs is the aqueous saturation solubility of the drug, and Ct is the concentration dissolved at time t. Based on this equation, it can be seen that all other factors being constant, the rate of dissolution is proportional to the surface area of the solute particle and to the solubility of the drug. Consequently, drugs with low aqueous solubility would be expected to exhibit low dissolution rates and, likely, poor oral bioavailability. The importance of the rate of dissolution and hence aqueous solubility in acting as a determinant of oral absorption was formally recognized with the establishment of the Biopharmaceutics Classification System (BCS) [2], where, as illustrated in Figure 1.1, the API is classified into four categories: classes 1 and 3 containing molecules with high aqueous solubility, and classes 2 and 4 containing molecules with low solubility; molecules in classes 3 and 4 also exhibit poor biological membrane permeability, another deterrent to drug absorption. Interestingly, over the past few decades there has been a significant increase in the number of APIs under development that have fallen into BSC classes 2 and 4 because of solubility problems. This decrease in dissolution of crystalline API appears to correlate with a corresponding increase in the number of API molecules in the development process that have larger average molecular weights, higher melting temperatures, and a higher degree of hydrophobicity than that observed in previous years. As a consequence, during the past few years, there has been a significantly increased effort to develop strategies that might serve to enhance the rate of dissolution of an API by means of formulation, chemical modification, or processing.

Figure 1.1 Biopharmaceutical classification systems (adapted from Ref. 2).

Based on Equation 1.1, we can conclude that there are two major factors that can be used as a basis for enhancing dissolution rates of poorly water-soluble crystalline APIs sufficiently to have some controllable influence on increasing oral bioavailability. These are the surface area of the solid exposed to the aqueous medium and the solubility of the solid in aqueous media. Strategies for enhancing dissolution can be divided further into (i) formulation and processing, (ii) chemical modification, and (iii) use of "high-energy" structurally disordered physical forms of the solid. Starting with the crystalline API, the formulator can simply reduce the particle size of crystalline materials to increase their specific surface area (area per unit mass). Very significant increases in dissolution rate, for example, have been attained by producing particles with diameters on the order of 100-300 nm. One also can increase dissolution rates by adding solubilizers to the formulation, such as surfactants, or complexing agents, such as cyclodextrins, which help to produce a supersaturated solution when the API encounters an aqueous medium. Surfactants can also act as wetting agents to improve access of the aqueous medium to hydrophobic API, thus effectively increasing the available surface area. High levels of supersaturation, upon contact with water, can also be obtained by dissolving the API in liquid lipid-based formulations and administering the product in hard or soft capsule form. Such an approach tends to produce a supersaturated solution upon exposure to aqueous dissolution media. Alteration of the API chemically by forming more highly water-soluble crystalline salts or cocrystals, when possible, can be a very efficient way of increasing dissolution rates as long as the dissolved form of the API can be maintained in a supersaturated state relative to that of the crystalline "free form" of the API itself. Finally, since the high lattice energy of an API crystal, as often reflected at high melting temperatures, can serve as an impediment to attaining adequate thermodynamic solubility, any approach that can change, reduce, or eliminate the crystal lattice energy should be able to enhance the apparent solubility. For example, liquid forms of molecules will generally exhibit greater solubility than their crystalline counterparts (supersaturation), all other factors being equal. Indeed, it is well known that higher energy "less-stable" polymorphic crystal forms of an API generally exhibit greater solubility than the most stable form. It has also been shown that disorder in the crystal lattice introduced as crystal defects can serve to increase dissolution from the defect sites relative to that from the less defective crystal. Consequently, it is not surprising that complete elimination of long-range three-dimensional order in the crystal by forming the amorphous form of an API can greatly enhance apparent solubility and rates of dissolution. Of course, since the amorphous state represents a high-energy form relative to the crystal, this approach can be useful only as long as a supersaturated solution of API can be maintained in the aqueous medium over the time period required for gastrointestinal absorption. Since the overall theme of this book deals with amorphous API-polymer solid dispersions designed to provide enhanced oral bioavailability by creating such supersaturation, it will be useful in this introductory chapter to review some of the important physicochemical characteristics of amorphous solids as single components and as mixtures of API with other formulation components that might be used to enhance oral bioavailability in drug products. A brief discussion of API-polymer amorphous dispersions, in particular, will serve as an introductory overview of various principles that will be applied in more detail throughout the rest of the book.

1.2 Formation of the Amorphous State and the Glass Transition Temperature

Let us first consider a single-component system such as an API in its most stable crystalline form. From a classical free energy-temperature diagram [3], as illustrated in Figure 1.2, we can observe a significant reduction in the free energy per mole of the crystal as the temperature of the sample is increased until we reach the melting temperature Tm where the crystal undergoes a spontaneous first-order conversion to the liquid form, with the liquid now in a lower free energy state. If the liquid is slowly cooled to below Tm, and there is sufficient time for nucleation and crystal growth to occur, the system will revert to the equilibrium state of the crystal. If, however, as seen in Figure 1.2, the liquid sample is cooled rapidly through Tm so as to kinetically avoid crystallization, the system will show no discontinuities at Tm and maintain the equilibrium properties of the liquid as a supercooled liquid that is metastable relative to the crystal. Upon further cooling and as the viscosity of the supercooled liquid increases and diffusive motions of the molecules decrease, equilibrium can no longer be maintained and a distinct discontinuity in the free energy-temperature diagram occurs with the formation of the unstable glassy state. This occurs at a distinct temperature, designated the glass transition temperature Tg, the value of which for a particular molecule under the same processing conditions is determined by the molecular weight, degree of polarity, and the effect of molecular shape on the closeness of molecular packing. For example, the more polar the solid or the higher the molecular weight, the greater the value of Tg, while the bulkier the shape of the molecule and poorer the packing, the lower the Tg. The value of Tg is experimentally determined most conveniently by using differential scanning calorimetry, where the heat capacity can be measured as the sample temperature is continuously changed at a constant rate from low temperatures to the melting temperature. Because of structural changes that bring about changes in the rate of molecular motions, the heat capacity generally undergoes a distinctly abrupt change at Tg, as illustrated in Figure 1.3. In general, it has been shown that the viscosity of an organic liquid at Tm is on the order of 10-2 Pas, while at Tg this value has increased to about 1012 Pas, a 14 order of magnitude change! Since this point of...