Influence of particle size on solubility of active pharmaceutical ingredients

Abstract

The aqueous solubility of an active pharmaceutical ingredient (API) is an important property that requires evaluation during early development and prior to formulation of the final product. With general, experimental, solubility testing of different APIs, the question always arises as to whether particle size had been determined beforehand or not. All available literature suggests that particle size, for pharmaceutical powders, does not significantly affect equilibrium solubility. The dissolution rate will differ according to different particle sizes, but the overall results should be identical after equilibrium is established. This study was therefore planned to investigate as to whether different particle size fractions of the same API, dissolving at different rates, would all reach solubility equilibrium within 24 hours. Also, APIs from different solubility classes were investigated, because poorly soluble substances would most likely require a longer period of time to equilibrate. The time period of 24 hours was selected, because many published solubility studies report using that interval and is it the standard for our research group also. Available APIs were selected to determine the influence (if any) of particle size on their equilibrium solubilities and the time required for attaining that status. For the purpose of this investigation, five APIs were selected from compounds at our disposal in-house, ranging from freely soluble to poorly soluble in the order: chloroquine phosphate > pyrazinamide > mefloquine hydrochloride > closantel sodium > roxithromycin. Solubility studies were successfully completed on four of the five APIs selected. For closantel sodium, pyrazinamide and roxithromycin it was demonstrated that the 24 hour test period was sufficient for the attainment of equilibrium solubility, regardless of the particle size fractions tested. Surprisingly, the only API in this study for which 24 hours was an insufficient test period was mefloquine HCl, which was not the least soluble compound tested. Further testing would be required to clarify this anomaly. What was evident from the outcomes of this investigation was that although the ubiquitous 24 hour solubility test may work well in many cases, its suitability should be reviewed on a case-by-case basis and not just for the most poorly soluble compounds. Researchers testing solubility at temperatures lower than 37°C should be especially cautious of using a standardised test period, because equilibrium solubility would take longer to achieve with less energy available to the system.