No abstract

In this paper, we report a process control strategy for the production of metastable R-form glycine crystals of a desired mean mass size by manipulating the alternating temperature profile and the final termination temperature. The seed crystals of the R-form glycine introduced were grown successfully to the size of the product with no fine crystals. Generation of the γ-glycine crystals (stable polymorph) was completely avoided. This crystallization method is flexible and easy to operate, because the alternating temperature profile can be determined on-site according to the transient supersaturation and particle count number data obtained from an in-situ ATR-FTIR spectrometer and an in-situ FBRM particle counter, respectively. The termination time or batch time was also determined on-site to a point that the residual supersaturation became zero. This on-site strategy-determination technique is expected to be applied widely for a variety of polymorphic systems other than the glycine-water system as a practical method for the selective crystallization of metastable polymorphs.

The growth-starting supercooling ∆T G , defined as a supercooling at which a crystal starts to grow on cooling, was measured for the (100) face of a potassium dihydrogen phosphate (KDP) crystal on which an Al (III) impurity had been adsorbed beforehand. When adsorption was made onto the nongrowing (100) face at a low supercooling value of ∆T A ) 0.6 °C, the measured values of ∆T G were successfully explained with a mathematical model considering the Langmuir equilibrium (or instantaneous) adsorption. On the contrary, smaller values of ∆T G were obtained when adsorption was made on the growing (100) face at a high supercooling value of ∆T A ) 6.0 °C. These smaller values obtained for the adsorption at ∆T A ) 6.0 °C are concluded to be due to an insufficient amount of the adsorbed Al (III) impurity, and the adsorption is shown to proceed more slowly onto the growing (100) face compared to the nongrowing face.

The growth rate of a crystal in the presence of impurity depends on the history of supercooling. This behavior is called growth rate hysteresis (GRH). In this paper, GRH is described by using a mathematical model. This mathematical model is devised by considering the pinning mechanism of Cabrera and Vermilyea, the two-dimensional Gibbs−Thomson effect on step movement and slow adsorption of impurity species on a crystal surface. The model explains, reasonably but qualitatively, experimental literature data on GRH that its magnitude becomes large as the supercooling-changing rate R is decreased or the impurity concentration c is increased. The model also shows a possibility that the reverse effect of these two factors (R and c) on GRH may occur in the range of their small values and it predicts the GRH behavior over a wide range of experimental conditions. Limitations of the model are also discussed.