(152bs) Impact of Solvent Selection on Oiling-out in Pharmaceutical Crystallization

In the pharmaceutical industry, crystallization is an important step utilized to purify active pharmaceutical ingredients (APIs). With increasing numbers of APIs having increased structural complexity and hence poor solubility, undesired liquid-liquid phase separation (LLPS), also known as oiling out, are often encountered during crystallization. Oiling out of the API is often associated with product quality issues and process difficulties. Solvent switch is often used as an effective strategy to avoid oiling out. However, the selection of appropriate solvents is usually laborious or empirical. Therefore, in this study, we aim to understand the impact of solvent selection on API oiling out propensity.

A poorly soluble drug was selected as a model compound. We surveyed four different solvent systems, DMSO/H₂O, ACN/H₂O, MeOH/H₂O, and ethyl acetate/heptane, due to their varied performance in API crystallization. Oiling out was observed in the former two solvent systems, whereas the later two were used for crystallization development. Two different temperatures commonly used in API crystallization, room temperature and 55°C, were investigated. The crystalline solubility of the API was measured gravimetrically using the shake flask method. The oiling out limits, or LLPS phase boundaries, were determined by antisolvent titration. The maximum achievable degree of supersaturation was calculated by dividing the onset LLPS values with the drug’s crystalline solubility in the corresponding solvent system. Lack of crystalline form change of the API was confirmed using X-ray powder diffraction (XRPD), and the lack of crystallinity in the oiling out phase was confirmed using polarized light microscopy (PLM). For the ethyl acetate/heptane system at 55°C, onset LLPS values were not determined due to fast crystallization of the API.

The maximum achievable degree of supersaturation was found to be around one order of magnitude lower in the two oiling out solvents (DMSO/H₂O and ACN/H₂O) compared to the two crystallization solvents (MeOH/H₂O, and ethyl acetate/heptane). The maximum oiling out limit was found to be relatively insensitive to solvent choices, with the highest values of approximately 500 mg solute per gram solvent across all solvent systems. However, the crystalline solubility of the API was found to vary greatly in different solvents. The two oiling out solvent systems, DMSO and ACN, showed greater solubilization capacity of the crystalline API. Thus, a much narrower operation window is expected due to lower maximum achievable degree of supersaturation, and oiling out readily occurs when higher degree of supersaturation is generated. The crystalline solubility of the API in the two crystallization solvent systems was low, which enabled a high degree of supersaturation and a wide operation window for crystallization development. Oiling out would not occur unless the LLPS phase boundary was reached, and high supersaturation led to fast crystallization of the API.

In conclusion, this study revealed for the first time the mechanisms underlying the impact of solvent selection on oiling out propensity. The ability of the solvent to solubilize crystalline APIs appears to be the most important factor, whereas oiling out limits are less sensitive to solvent selection. Thus, for APIs with oil out risks, especially poorly soluble drugs, it is necessary to select solvents with less solubilization capacity to avoid oiling out and facilitate crystallization.