Stephane Beilles

Preferential crystallisation and comparative crystal growth study between pure enantiomer and racemic mixture of a chiral molecule: 5-ethyl-5-methylhydantoin

Abstract It is shown that (±) 5-ethyl-5-methylhydantoin (12Hyd) can be separated at a preparative scale by means of the auto-seeded and polythermic preferential crystallisation in water, provided that a small proportion of wetting agent is used. The influences of enantiomeric purity, supersaturation and wetting agent during the crystal growth of 12Hyd in water are investigated. Large particles in the shape of single crystals obtained from unstirred racemic solutions and grown under smooth conditions of supersaturation exhibit unusual hourglass figures through {1 0 1} faces when observed under polarised light. Moreover, they contain almost no enantiomeric excess, which indicates that they are not true single crystals. This is in apparent contradiction with the possibility of resolving the racemic mixture by means of preferential crystallisation. Stereoselective dissolutions of these apparent single crystals shows that this results from a crystal growth mechanism based on the alternated 2D nucleation of homochiral domains along specific growth directions, leading to “lamellar polyepitaxy” phenomenon along {1 0 1} faces and responsible for the formation of hourglass figures by means of different types of crystal defects. Crystal structure analysis and molecular modelling tools allow to present some explanations consistent with these data.