Growth of the 2:1 cocrystal of benzoic acid with isonicotinamide on the surface of a benzoic acid crystal. Note that the crystals all grow at a similar angle.

A key aim of crystal engineering is the directed creation of materials with specific properties through the control of intermolecular interactions. For a given material, variations in physico-chemical properties are achieved by polymorphism, where the molecule packs in different crystal structures. The discovery of new polymorphs is often serendipitous and can be difficult to control, highlighted by the existence of disappearing polymorphs. 

An investigation into the selectivity of composition during co-crystallization was undertaken using the test system of benzoic acid and isonicotinamide. These molecules are simple molecules with three possible synthons, but the system is capable of forming complexes with different compositions. The 1:1 complex is well known and its crystal structure has been determined, while, until recently, the existence of a 2:1 complex was a matter of debate. The existence of this phase was confirmed during recent investigations into the epitaxial growth of molecular complexes upon single component substrates. The 2:1 complex only forms after formation of benzoic acid; the reason for this is that the 2:1 complex is templated by the surface of the benzoic acid crystal.

The single crystals of the 2:1 cocrystal are quite small, with dimensions of around 0.4 x 0.05 x 0.05 mm³, and as the component molecules are composed of light atoms only, the crystals diffract only weakly. A dataset was collected on the RAPID II using exposure times of 20 minutes, giving sufficient quality data for the structure to be easily solved and refined. 

Of particular importance in this example is the ability to face index the crystal. Calculations have been carried out on the benzoic acid "host" crystal to determine the faces that are most likely to template the formation of the 2:1 co-crystal; this work predicted that the long axis of the 2:1 co-crystal should be perpendicular to the 001 direction, and this has recently been confirmed experimentally.

Face indexing of the 2:1 cocrystal, illustrating the growth of the co-crystal along the 001 direction.