Computer-Aided Solvent Selection for Improving the Morphology of Needle-like Crystals: A Case Study of 2,6-Dihydroxybenzoic Acid

Abstract

We developed a computer-aided rational solvent selection procedure for improving the morphology of needle-like crystals and applied it to 2,6-dihydroxybenzoic acid (DHB) form 2 crystals. In our procedure, a simple morphology prediction model, the Bravais−Friedel−Donnay−Harker (BFDH) model, was first applied to identify the needle tips and the major faces parallel to the needle direction (the surrounding faces). After that, a more sophisticated model, the modified attachment energy (MAE) model which incorporates solvent effects, was applied to the needle tips and the surrounding faces to estimate the aspect ratios of the DHB crystals obtained from various single solvents. Identifying the key interactions between chemical groups on solvent molecules and chemical groups on different crystal faces via analysis of molecular dynamics trajectories and using MAE values for mixed solvents, we were able to propose a specific solvent mixture for improving morphology. Specifically, our computer-aided solvent selection procedure suggested that a mixture of toluene and diethylether (mole ratio 4:1) should give DHB crystals with the smallest aspect ratios. We then successfully confirmed this experimentally by slurry aging experiments. DHB crystals aged in the mixture mentioned above exhibit significantly reduced aspect ratios. The solvent selection procedure developed in this work goes significantly beyond trial and error and has the potential to aid in the rational selection of solvents for improving the morphology of other needle-like organic crystals.