T–X phase diagram of salicylic acid–anthranilic acid with three crystalline solid solution phases and a co-crystal, resulting in variable enantiotropic transition temperature and a polymorphic co-existence domain.

Crystallization from solution is a key unit operation utilized across the synthetic scheme to remove impurities. However, little is still known of the underlying impurity purge mechanisms that are responsible for controlling the final purity of the product. Reported herein is the solubility-limited impurity purge mechanism in which the impurity exists as a separate solid phase with its own solubility. A mathematical framework is presented that describes the separation of the impurity in the solid and liquid phases based on the relative solubilities of the product and impurity, and initial impurity level. Three theoretical solubility-limited impurity purge mechanisms are derived that are confirmed experimentally using salicylic acid, ibuprofen, and acetaminophen as model compounds. A practical experimental test is introduced that is used to identify if the impurity is rejected by solubility-limitation and its corresponding type. Finally, development strategies are presented to remove impurities that are purged based on their solubilities.

The mechanisms of purging structurally similar impurities in solution crystallization have been evaluated using the model compound salicylic acid. Of the 11 added impurities, 3 showed appreciable entrapment in the solid phase: viz., salicylamide, anthranilic acid, and benzoic acid. X-ray powder diffraction (XRPD), differential scanning calorimetry (DSC), and the use of a previously reported solubility-limited impurity purge (SLIP) test have shown that the impurities are entrapped by a lattice incorporation mechanism. Impurities become integrated within the product crystals during the crystallization by forming terminal solid solutions. Most of the impurity entrapment was found to take place very early in the crystallization, immediately after seeding. The least entrapment occurred at the end of the crystallization, despite the mother liquor being enriched in impurities. These changes caused purity variations in the solids, which were not properly captured by the average value. A mathematic framework was developed to afford the material impurity distribution (MID), which represents the mass-based impurity profile across a material based on the SLIP test. It is shown that the level of impurities in the crystallized material is far from constant and in fact varies by orders of magnitude, in many cases by more than 20 times. These differences give rise to changes in the physical properties of salicylic acid, as exemplified by a reduction in crystallinity, a lower and broader melting event, and a doubling of solubility.

We report a ripening mechanism in crystallization of an organic compound that is observable above 1 μm and is different from Ostwald ripening. Salicylic acid crystallized in methanol and water in the presence of simulated structurally similar impurities resulted in impurity entrapment via the formation of solid solutions. Impurities were retained in the solids to a higher degree in the early part of the crystallization and decreased rapidly toward the end. This uneven impurity entrapment resulted in impurity gradients in individual crystals, which were detected and confirmed by Raman imaging. The impurity entrapment impacted the physical properties of salicylic acid, including its solubility and melting properties. An intra-particle thermodynamic driving force was thus found to exist during the crystallization, which caused selective dissolution of the dirtier cores of the crystal and recrystallization on the cleaner edges. These concomitant but opposing mass transfer processes during the crystallization were responsible for the formation of hollow crystals of salicylic acid. The ripening phenomenon is demonstrated experimentally in the beginning and end of an isothermal anti-solvent crystallization as well as during an agitated filter dryer operation.

No abstract

No abstract

No abstract

No abstract