(656c) Combined Cooling and Anti-Solvent Crystallization of L-Asparagine Monohydrate

Crystallization is a key unit operation especially in fine chemical and pharmaceutical industries to produce high value solids with desired purity and size distribution [1]. The cooling and anti-solvent crystallization are common operational modes used in industry and the optimization in such processes is traditionally sought with respect to optimal cooling profile, anti-solvent feed-rate profile, and seeding characteristics. Essentially, the control is implemented by manipulating the supersaturation trajectory during the course of crystallization due to its relationship with the fundamental crystallization phenomena such as nucleation, growth, etc. Cooling crystallization is advantageous when the solubility of the compound strongly depends on temperature.Anti-solvent crystallization is particularly advantageous when the solute to be crystallized has a solubility that is a weak function of temperature and/or the soluteâ??s thermal stability is low. Sometimes the solubility of a compound is significantly affected by both temperature and addition of an anti-solvent. Therefore, combining cooling and anti-solvent crystallization can be advantageous as it offers two manipulative variables that can be varied to exert greater control on the supersaturation profile and thereby on the outcome of the crystallization process [2]-[3].

 

The objective of this research is to perform a systematic study on how the cooling and anti-solvent addition should be combined using three open-loop conventional cooling policies and three feeding policies that are inspired by these well-known cooling policies [4]. Specifically, we perform a series of combined cooling/anti-solvent crystallization of L-asparagine monohydrate (LAM), an important amino acid, from its aqueous solution using isopropanol as anti-solvent. First, solubility of LAM in water-isopropanol mixture is determined gravimetrically as a function of temperature (25 to 45 °C) and weight fraction of isopropanol (0 to 80%). Next, combined cooling/anti-solvent crystallization experiments are conducted where a saturated aqueous solution of LAM at 45 °C is cooled using a given cooling policy from 45 to 25 °C in 2 h while isopropanol is added simultaneously with a given feeding policy such that its weight fraction changes from 0 to 40 wt% by the end of the experiment. All combined cooling and anti-solvent crystallization experiments recorded improvement in crystal size distribution characteristics and significant enhancement in the yield compared to only cooling crystallization.

 

Experiments with various combinations of cooling and feeding policies generate a wealth of data for understanding the dynamics of the process and estimating kinetic parameters for developing a predictive model. The concentration of the solute is measured at regular sampling intervals by high performance liquid chromatography and the final crystal size distribution is measured by laser diffraction. We develop a dynamic model based on population balance equation coupled with mass balance equation; and nucleation and growth kinetics are simultaneously estimated from experimental data using a nonlinear parameter estimation technique. The fully validated model is then used to determine the optimal temperature and anti-solvent feeding profiles. A comparison is then made among the optimal policy and conventional policies to understand the convenient and optimal path of combining the two modes of operation.

References:

[1] J. W. Mullin, Crystallization. Woburn, MA: Butterworth-Heinemann, 2001.

[2] Z.K. Nagy, M. Fujiwara, and R.D. Braatz, J. Process Control 18, 856â??864, 2008.

[3] C. Lindenberg, M. Krattli, J. Cornel, and M. Mazzotti, Cryst. Growth Des. 9, 1124â??1136, 2009.

[4] H. Hojjati, M. Sheikhzadeh, and S. Rohani, Ind. Eng. Chem. Res. 46, 1232-1240, 2007.