Polymorphic co-crystals from polymorphic co-crystal formers: Competition between carboxylic acid⋯pyridine and phenol⋯pyridine hydrogen bonds

The recent literature has shown an increase in the number of co-crystals reported to be polymorphic, with at least 45 such systems identified thus far. The question of whether co-crystals, defined as any multicomponent neutral molecular complex that forms a crystalline solid, are inherently less prone to polymorphism than the individual components is shown to be untrue in four sets of polymorphic co-crystals. The co-crystal formers in this study, acridine, nicotinamide, 3-hydroxybenzoic acid, 2,4-dihydroxybenzoic acid, malonic acid, and pimelic acid, are all polymorphic in their unimolecular states and are shown to be dimorphic in the following combinations: (3-hydroxybenzoic acid)·(acridine) [1(I) and 1(II)], (2,4-dihydroxybenzoic acid)·(nicotinamide) [4(I) and 4(II)], (malonic acid) ·(nicotinamide) [5(I) and 5(II)], and (pimelic acid) ·(nicotinamide) [6(I) and 6(II)]. These co-crystals are assembled primarily using carboxylic acid and phenol hydrogen bond donors that hydrogen bond to pyridine N or amide carbonyl acceptors. Two different combinations of donors and acceptors are primarily responsible for the formation of polymorphs in 1 and 4, whereas conformational differences within the malonic and pimelic acid molecules lead to different packing arrangements using the same combination of hydrogen bonded interactions in 5 and 6. The 1:2 co-crystal of (3-hydroxybenzoic acid)·(acridine)₂ (2) displays both the phenol O-H⋯N hydrogen bond observed in 1(I) and the carboxylic acid O-H⋯N hydrogen bond observed in 1(II). In addition, a methanol solvate of (2,4-dihydroxybenzoic acid)·(nicotinamide) (3) is reported. DFT calculations show that the carboxylic acid⋯pyridine hydrogen bond is strongest and one of co-crystallization's most useful interactions.