Predictive tools for cocrystal formation

The rate determining step for the production of new cocrystals is represented by the selection of the coformer suitable for supramolecular synthon formation. The trial and error approach must be substituted by a more efficient and ecofriendly methodology to minimize costs, time, and reagents. A prediction method should generally be applied as a first screening step to reduce the number of coformers to be tested experimentally. In this chapter, the state-of-the-art of the prediction tools developed for cocrystal formation is presented. The most widely procedures are: supramolecular synthon search, hydrogen bond propensity (HBP), hydrogen bond energy (HBE), conductor-like screening model for real solvent (COSMO-RS), Hansen solubility parameters (HSP), and molecular complementarity (MC). All these methods are discussed in detail with selected examples, together with a critical analysis of their advantages and disadvantages. Finally, new approaches that exploit machine learning algorithms are also discussed. In conclusion, the aim of this chapter is to offer new insights and to promote cooperative efforts in the fascinating field of cocrystal landscape.