(190x) Enantioselective Crystallization Process Design From Predictive COSMO Calculations | AIChE

(190x) Enantioselective Crystallization Process Design From Predictive COSMO Calculations

Authors 

Zinke, R. - Presenter, Max Planck Institute for Dynamics of Complex Technical Systems
Stein, M. - Presenter, Max Planck Institute for Dynamics of Complex Technical Systems


The separation of enantiomers from a racemic mixture is of great interest to the chemical and the pharmaceutical industry. Enantiomers are non-superimposable mirror images of each other and thus have identical physical and chemical properties but may have different biological effects. Enantioselective crystallization is an attractive strategy since only one enatiomer crystallizes in its pure form whereas the other remains in solution. The rationalization of such a process is a challenge both to theoretical and experimental approaches.

The process design relies on an accurate evaluation of binary and ternary solubilities. COnductor-like Screening models, i.e. COSMO-RS [1] and (refined-) COSMO-SAC [2, 3] are efficient methods to calculate solubilities from a molecular perspective of both the solute and the solvent. They give the possibility to calculate chemical potentials via the determination of activity coefficients of molecules in solution and thus are further able to compute different equilibrium thermodynamic properties of mixtures. We are using these methods to design and rationalize enatioselective crystallization processes.

In [4] the authors describe a crystallization scheme to isolate and thus purify single enantiomers from an asymmetric mixture of stereoisomers. The corresponding process was adapted to demonstrate its enantioselective separation potential for the example of bicalutamide, the pharmaceutical ingredient of the drug Casodex(TM). The followed process is based on an appropriate choice of a pair of solvent and antisolvent. After identification of suitable candidates, the chiral separation of enantiomers of (R,S)-bicalutamide was performed. The experimental implementation of the enantioselective crystallisation was supported theoretically by using COSMO-SAC and refined COSMO-SAC calculations. COSMO-SAC was used to identify suitable antisolvent candidates and both COSMO-SAC approaches were used to calculate binary solubilities. The aim was to test the accuracy of calculated solubilities of these methods. In [4] COSMO calculations were only of supporting character and all experimentally relevant process steps were made on the basis of experimental results only.

We here report an enantioselective crystallization of bicalutamide based on in silico predictions from molecular solution. We used different implementations of COSMO models, and evaluated their accuracies in particular with regard to the possibility of designing the process. It is our aim to assess the accuracy of current COSMO implementations for the enantioselective crystallization of bicalutamide. Therefore the method COSMO-RS and the derived versions (refined-) COSMO-SAC of COSMO-based models were evaluated and compared in view of critical crystallization specific tasks. First, the search for a suitable antisolvent with reduced bicalutamide solubility was performed using COSMO methods. In a second step, model-based estimates of relevant thermodynamic properties required to perform the selective crystallization of single enantiomers were conducted. Finally, our in silico process design was verified in the lab. Bicaluatmide was used as a benchmark case for a pharmaceutical substance but also well-characterized chiral chemical compounds, such as mandelic acid, were tested. Limitations and inadequacies in the underlying theory, the accuracy of calculated molecular properties and their sensitivities to the choice of COSMO implementation were analyzed. 

[1] A. Klamt and G. Schüürmann, COSMO: A new approach to dielectric screening in solvents with explicit expressions for the screening energy and its gradient, J. Chem. Soc. Perkin Trans., 2 (1993) 799-805.

[2] S.-T. Lin and S. I. Sandler, A priori phase equilibrium prediction from a segment contribution solvation model, Ind. Eng. Chem. Res. 46, (2002) 899-913.

[3] S. Wang and S. I. Sandler, Refinement of COSMO-SAC and the applications, Ind. Eng. Chem. Res. 46, (2007) 7275-7288.

[4] H. Kaemmerer, M. J. Jones, H. Lorenz and A. Seidel-Morgenstern,  Selective crystallization of a chiral compound-forming system-Solvent Screening, SLE determination and process design, Fluid Phase Equilibria 296, (2010) 92-205.