Cosolvency approach for assessing the solubility of drugs in poly(vinylpyrrolidone)

Abstract

The log-linear cosolvency model was applied for estimating the solubility of four drugs: ritonavir, griseofulvin, itraconazole and ketoconazole in poly(vinylpyrrolidone) (PVP). Cosolvent mixtures consisted of PVP mixed in different proportions with N-ethylpyrrolidone, which served as the monomeric analogue of the repeating unit of the polymer. Solubility in the monomer–polymer mixtures was determined by HPLC. As the configuration of the solvating unit in the solvent mixture changed from entirely monomeric to increasingly polymeric, the solubility of the drugs decreased in a fashion that follows the log-linear cosolvency model. The linear relationship was used to obtain estimates for the solubility of the drugs in the different grades of PVP. The solubility of the drugs in PVP is low (from <1% to ∼15% w/w). Among the set of drug solutes, ritonavir exhibited the highest solubility in PVP (w/w). Mixing with the monomer is most favorable for griseofulvin among the four drugs. However, the detrimental effect of polymerization on its solubility is more pronounced than for ritonavir. The mixing of itraconazole with the monomer is more favorable than the mixing of ketoconazole. However, despite the molecular similarity between ketaconazole and itraconazole, the solubility of the latter is particularly affected by the polymeric configuration of the solvating unit, to the point of exhibiting differences in solubility resulting from the chain length of the grade of PVP used. The log-linear cosolvency model is a useful tool for estimating the solubility of the drugs in the polymer at room temperature, while providing quantitative information on the differences in mixing behavior of the four model compounds.