The supersaturation potential of poorly water-soluble compounds is of interest in the context of solubility enhancing formulations for enhanced bioavailability. In this regard, the amorphous "solubility", i.e., the maximum increase in solution concentration that can be obtained relative to the crystalline form, is an important parameter, albeit a very difficult one to evaluate experimentally. The goal of the current study was to develop new approaches to determine the amorphous "solubility" and to compare the experimental values to theoretical predictions. A group of six diverse model compounds was evaluated using the solvent exchange method to generate an amorphous phase in situ, determining the concentration at which the amorphous material was formed. The theoretical estimation of the amorphous "solubility" was based on the thermal properties of the crystalline and amorphous phases, the crystalline solubility, and the estimated concentration of water in the water-saturated amorphous phase. The formation of an amorphous precipitate could be captured transiently for all six compounds and hence the amorphous "solubility" determined experimentally. A comparison of the experimental amorphous "solubility" values to those calculated theoretically showed excellent agreement, in particular when the theoretical estimate method treated the precipitated phase as a supercooled liquid, and took into account heat capacity differences between the two forms. The maximum supersaturation ratio in water was found to be highly compound dependent, varying between 4 for ibuprofen and 54 for sorafenib. This information may be useful to predict improvements in biological exposure for poorly water-soluble compounds formulated as amorphous solid dispersions or other formulations that rely on supersaturation.

High energy solids, such as salts, co-crystals, or amorphous solid dispersions, have been widely used to generate supersaturated aqueous solutions and improve drug bioavailability. However, most research on solubility enhancing strategies has focused on the kinetics of dissolution, and there is relatively little comparison of the different degrees of supersaturation achieved by using different solid state forms of the same compound. Recent studies from our group have demonstrated that the maximum achievable supersaturation is dictated by the aqueous solubility of the amorphous form of the drug. Liquid−liquid phase separation (LLPS) occurs at concentrations above this value. Herein, it was hypothesized that the upper limit of supersaturation that can be achieved from dissolution of various high energy solids is also governed by the amorphous solubility. To test this hypothesis, the dissolution and supersaturation behavior of different solid forms of a model compound, CRH1, were investigated using a variety of techniques. With the exception of CRH1 crystalline free base, all solid forms generated supersaturated solutions. The extent of supersaturation, onset of crystallization time, and area under the curve increased significantly when a polymer with crystallization inhibitory properties was present in the dissolution medium or incorporated in the formulation (in the case of amorphous solid dispersions). In the presence of the polymeric crystallization inhibitor, several solid state forms, including the amorphous solid dispersion and the salts, dissolved to concentrations above the amorphous free form solubility and underwent LLPS, generating a drug-rich phase. Other solid state forms underwent crystallization prior to attaining the amorphous solubility and showed no evidence of LLPS (co-crystal and glass forms). These studies should aid in solid state form selection and formulation and help to understand how to achieve maximized supersaturation in vivo.

Quantitative analysis of solid-state processes from isothermal microcalorimetric data is straightforward if data for the total process have been recorded and problematic (in the more likely case) when they have not. Data are usually plotted as a function of fraction reacted (α); for calorimetric data, this requires knowledge of the total heat change (Q) upon completion of the process. Determination of Q is difficult in cases where the process is fast (initial data missing) or slow (final data missing). Here we introduce several mathematical methods that allow the direct calculation of Q by selection of data points when only partial data are present, based on analysis with the Pérez-Maqueda model. All methods in addition allow direct determination of the reaction mechanism descriptors m and n and from this the rate constant, k. The validity of the methods is tested with the use of simulated calorimetric data, and we introduce a graphical method for generating solid-state power-time data. The methods are then applied to the crystallization of indomethacin from a glass. All methods correctly recovered the total reaction enthalpy (16.6 J) and suggested that the crystallization followed an Avrami model. The rate constants for crystallization were determined to be 3.98 × 10(-6), 4.13 × 10(-6), and 3.98 × 10(-6) s(-1) with methods 1, 2, and 3, respectively.

Calorimetry is a general method for determination of the rates of zero-order processes, but analysis of the data for the rate constant and reaction enthalpy is difficult because these occur as a product in the rate equation so evaluation of one requires knowledge of the other. Three methods for evaluation of both parameters, without prior knowledge, are illustrated with examples and compared with literature data. Method 1 requires the reaction to be studied in two buffers with different enthalpies of ionization. Method 2 is based on calculation of reaction enthalpy from group additivity functions. Method 3 applies when reaction progresses to completion. The methods are applied to the enzymatic hydrolysis of urea, the hydrolysis of acetylsalicylic acid, and the photodegradation of nifedipine, respectively.

No abstract