Template of abstract for ISPC16, June 22-27 2003, Taormina (Italy) (place your title here, Times 14, bold, centred)

Obtaining pure enantiomers and co-crystals using continuous preferential crystallization: A process development investigation using population balance models

T. Vetter1,2, C.L. Burcham1, M.F. Doherty2

1Eli Lilly & Company, Indianapolis, IN-46281, United States

2Department of Chemical Engineering, University of California, Santa Barbara, CA-93106, United States

Keywords: Continuous crystallization, separation of enantiomers, formation of co-crystals, wet milling, process design

Designing processes that yield enantiomerically pure products is of key importance in the pharmaceutical industry where typically only one of the enantiomers shows the desired effect on a patient, while the other enantiomer is either not effective as a drug or even toxic. Since asymmetric synthesis may not yield high enough enantiomeric purity, one has to resort to separation processes that separate enantiomers. For large biomolecules that are administered in liquid form chiral chromatography is the method of choice. However, small molecular drugs are sold as solids that typically originate from a crystallization process. For these products it is therefore interesting to investigate crystallization processes that can separate enantiomers and deliver a solid product in a single process step [1,2].
For conglomerate forming systems of enantiomers, i.e., systems in which crystals are either pure S enantiomer or pure R enantiomer (but no crystals of mixtures of the R and S enantiomers exist), preferential crystallization has been shown to be an attractive processing variant [3,4]. The two key concepts that enable preferential crystallization processes are the dependence of mass uptake of solute into the crystalline phase on the surface area of the crystals, i.e., the amount of solute incorporated per unit time depends linearly on the total surface area of the crystals present in the suspension, and the existence of a metastable zone, i.e., the nucleation of new crystals is an activated process and does not readily happen at low supersaturations. It therefore becomes possible to obtain a larger quantity of the desired enantiomer by seeding a crystallization process with crystals of this enantiomer. As a batch process, this process is best carried out in a polythermal fashion as described by Levilain and Coquerel [4]. Continuous processing concepts that allow preferential crystallization processes to be operated at steady state have also been developed [5-8]. We have recently designed a continuous flowsheet that is capable of separating enantiomers of a conglomerate in a wide range of processing conditions through the use of wet mills for continuous seed generation, recycle streams to obtain high yields and dissolver units to dissolve nuclei of the undesired enantiomer, which was shown to increase the range of processing conditions in which pure enantiomers can be obtained [9]. In that work process simulations were carried out using population balance equation models that allowed identifying key parameters of the process.
In this contribution, we add to these concepts by using this flowsheet to the far broader class of materials that forms racemic compounds. By operating in the three-phase region (solution, crystals of the racemic compound and crystals of the desired enantiomer), the process can guarantee the production of the desired enantiomer in pure from when it is performed with an enriched feed stream. The same concept and flowsheet can be applied to produce co-crystals in a wide region of their phase diagram, which is in contrast to batch processes that are driven to the thermodynamic equilibrium and therefore have to operate in the usually narrow two phase region (co-crystals and solution) of the phase diagram.

Acknowledgments: This work was supported in part by a Lilly Innovation Fellowship Award to TV from Eli

Lilly and Company.

References:

[1] Jacques, J.; Collet, A.; Wilen, S.H., â??Enantiomers, Racemates and Resolutionsâ?, Wiley, NewYork, 1981.
[2] Schroer, J.W.; Wibowo, C.; Ng, K.M., AIChE J., 2001, 47, 369-387.
[3] Coquerel, G., â??Preferential Crystallizationâ? in â??Novel Optical Resolution Technologiesâ?, Springer,
Heidelberg, Germany, 2007.

[4] Levilain, G.; Coquerel, G., CrystEngComm, 2010, 12, 1983â??1992.

[5] Levilain, G.; Eicke, M.J.; Seidel-Morgenstern, A., Cryst. Growth Des., 2012, 12, 5396-5401. [6] Qamar, S.; Elsner, M.P.; Hussain, I.; Seidel-Morgenstern, A., Chem. Eng. Sci., 2012,71, 5-17. [7] Qamar, S.; Galan, K.; Elsner, M.P.; Seidel-Morgenstem, A., Chem. Eng. Sci., 2013, 98, 25-39.

[8] Chaaban, J.H.; Dam-Johansen, K.; Skovby, T.; Kiil, S., Org. Process Res. Dev., 2013, 17, 1010-1020. [9] Vetter, T.; Burcham, C.L., Doherty, M.F., manuscript in preparation.