Needle crystals can cause filtering and handling problems in industrial settings, and the factors leading to a needle crystal morphology have been investigated. The crystal growth of the amide and methyl, ethyl, isopropyl, and t-butyl esters of diflunisal have been examined, and needle growth has been observed for all except the t-butyl ester. Their crystal structures show that the t-butyl ester is the only structure that does not contain molecular stacking. A second polymorph of a persistent needle forming phenylsulfonamide with a block like habit has been isolated. The structure analysis has been extended to known needle forming systems from the literature. The intermolecular interactions in needle forming structures have been analyzed using the PIXEL program, and the properties driving needle crystal growth were found to include a 1D motif with interaction energy greater than −30 kJ/mol, at least 50% vdW contact between the motif neighbors, and a filled unit cell which is a monolayer. Crystal structures are classified into persistent and controllable needle formers. Needle growth in the latter class can be controlled by choice of solvent. The factors shown here to be drivers of needle growth will help in the design of processes for the production of less problematic crystal products.

In pursuit of a solvent free green alternative to solution based processes, we have applied the combined use of catalytic additives and polycrystalline powder templates for polymorph control of gas phase crystallization to a range of pharmaceuticals and related compounds. Complementary volatile additives have been found which can catalyse the sublimation of a range of typical active pharmaceutical ingredients, APIs. Sublimation temperatures are typically reduced by up to 20 °C and the process is accelerated. The use of polycrystalline powder templates for polymorph control has also been successfully applied in several cases. Temperature control at the sites of both sublimation and desublimation is often required. The absence of even traces of solvent in the polymorphs produced appears to give the samples higher stability than samples obtained by crystallization from solution. Complete polymorph control was achieved with the following APIs, carbamazepine (5 polymorphs), metaxalone (2 polymorphs), mefenamic acid (2 polymorphs), paracetamol (2 polymorphs) and ortho-, meta-and para-amino benzoic acids (1, 4 and 2 polymorphs respectively).

Cocrystallization from the gas phase in the presence of tailor-made additives proceeds with highly effective morphology control.

The cocrystallization of the fluoroquinolones ciprofloxacin (cip), norfloxacin (nor), and enrofloxacin (enro) with the α,ω-dicarboxylic acids glutaric acid (glu), adipic acid (adi), pimelic acid (pim), suberic acid (sub), azeliac acid (az), and sebacic acid (seb) resulted in 27 new molecular salts and ternary molecular ionic cocrystals of compositions A + B − , A 2 + B 2− , A 2 + B 2− B, and A + B − A. Depending on the solvent, different stoichiomorphs, solvates, or polymorphs were obtained. All salts and cocrystals contain the robust R 2 NH 2 +...− OOC or R 3 NH +...− OOC synthon but have different supramolecular ring motifs. Moderate solubility enhancements over the parent fluoroquinolones were observed. Salts in the ratio of 1:1 and 2:1 were also prepared by ball-milling. The milled sample nor/az (1:1) was shown to gel the GRAS (generally recognized as safe) solvent propylene glycol, and enro/sub (1:1) was shown to gel both propylene glycol and water. Dynamic rheology measurements confirmed that nor/az and enro/sub behave like viscoelastic materials and supramolecular gels.