Scale-up and scale down of batch crystallization

A semi-automated apparatus that characterizes turbidity for the measurement of solubility and can measure multiple samples of one ml each simultaneously was developed. Its reliability was demonstrated by solubility measurement of several organic compounds. The solubilities of glycine and mannitol in water, paracetamol in isopropyl alcohol and adipic acid in ethyl alcohol were measured. Comparisons of solubility temperatures of test compounds in these experiments and the literature showed that the error was within a five weight %. The apparatus developed offers substantial savings in material, time and labor for solubility measurement.; To estimate the solubility at temperatures above and below the measured range, activity coefficients of solutes were estimated and their dependence on temperature extrapolated over a wide temperature range. These extrapolated activity coefficients were then used to calculate the solubility of solutes of interest at temperatures outside the range of the measurements. Comparison of solubility values obtained from the method and the data from the literature shows good agreement. Extrapolation of solubility versus temperature data beyond the range of measurement temperatures leads to accurate estimates of solubility, if sufficient data are experimentally obtained.; To study the effect of scale-up on the crystal size distribution (CSD) in crystallization operation, cooling and antisolvent crystallization were performed in two different scales, at 50 ml and 500 ml vessels. Seeding was also performed in cooling and antisolvent crystallization. The final CSD were compared and simple calculation of hydrodynamic parameters was performed to identify the scale-up effect on mixing and CSD. Controlled cooling and antisolvent addition were applied and same tip speed was employed as scale-up criterion. The larger volume vessel produced larger crystals in all experiments.; From the calculation of hydrodynamic parameters, it was shown energy dissipation rates in impeller and bulk region were much smaller in the larger volume reactor. From kinetic information, such as nucleation points and induction times according to supersaturation generation methods, it was revealed that in a larger volume reactor, smaller numbers of nuclei are produced at lower supersaturation and thus the final crystals are larger compared to the those obtained from a small volume reactor, irrespective of the supersaturation generation methods.; From seeding, it was found that in both supersaturation generation methods, relatively small amounts of seeds are needed to consume the supersaturation generated in a larger volume reactor because supersaturation in a larger volume reactor is relatively low compared to that in a small volume reactor. In addition, seeding reduced the difference in the CSD between the different scale vessels showing that control of nucleation is a key element in controlling the CSD on scale-up.