Measuring induction times and crystal nucleation rates.

A large variation is observed in induction times measured under equal conditions in 1 ml solutions. Ruling out experimental errors, this variation originates from the nucleation process. The induction time distribution is explained by the stochastic nature of nucleation if the number of nuclei formed is approaching 1 per vial. Accurate heterogeneous crystal nucleation rates were determined from the induction time distributions on a 1 ml scale for racemic diprophylline in two solvents. The difference in nucleation behaviour in the two solvents originates from the energy barrier for nucleation, which is much higher in the solvent in which induction times are much longer. In addition the pre-exponential factor for the crystal nucleation rate in both solvents is rather low compared to predictions using Classical Nucleation Theory. Unfortunately, concentration and surface characteristics of the effective heterogeneous particles are not known which clouds a further molecular interpretation.