Crystallisation of molecular materials is important for a range of industries including the pharmaceutical and agrochemical sectors. Prediction of the crystal growth and dissolution of such systems requires detailed knowledge of the thermodynamics, and ideally kinetics, of surface sites in contact with solvent. The most reliable method for obtaining the free energy of solvated kinks, step edges and terraces with molecular resolution is often via computer simulation using molecular dynamics, as this includes explicit solvent interactions and entropy changes, but comes at a significant computational cost. Here this approach is used to provide a series of benchmark standard free energies for dissolution of different types of surface sites at the paracetamol form I (001) interface with water. A modified GAFF force field is used that is calibrated to predict the solubility of paracetamol in good agreement with experiment. These reference free energies are then used to examine the accuracy of a faster, more approximate method that has been previously tested primarily for solubility and morphology prediction, namely the CrystalClear protocol. Results show that not only do both approaches have the same qualitative trends, with removal of the first molecule from the terrace at the aqueous interface having a significantly higher thermodynamic cost relative to removal from kink sites, but both approaches are in quantitative agreement to within the uncertainty of the molecular dynamics results for nearly all surface sites. The only site that exhibits a noticeable difference is for the removal of the final molecule from a step and occurs due to relaxation effects. The excellent agreement between the two methods examined here provides confidence that the CrystalClear approach represents a reliable and efficient means of predicting the thermodynamic stability of individual molecular binding sites at solvated interfaces, which can then be used for coarse-grained methods such as Monte Carlo simulation of crystal growth and dissolution.
No takes yet. Share an insight, caveat, or question.
Walisinghe et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: