Evaluating the Predictive Abilities of Protocols Based on Hydrogen-Bond Propensity, Molecular Complementarity, and Hydrogen-Bond Energy for Cocrystal Screening

Abstract

The successful synthesis of multi-component molecular solids such as cocrystals typically requires a large number of experiments before a new solid form of the target compound emerges with improved bulk properties. In order to focus and guide cocrystal synthesis we evaluated three different protocols (individually and combined) for predicting whether a cocrystal would form dur-ing attempted cocrystallizations between six active pharmaceutical ingredients and 42 potential coformers of the general interest. A blend of different methods gave the best results and a combination of protocols based on molecular complementarity (MC) and hydrogen-bond propensity (HBP), correctly predicted the outcome of 81% of experiments, whereas a combination of MC, and a protocol based on hydrogen-bond energies (HBE) delivered an overall success rate of 77% (including true positives and true negatives). The ease with which these methods can be implemented prior to large-scale experimental cocrystal screening, means that the choice of potential coformers may be narrowed effectively and the experimental search can be substantially optimized.