Polymorphism is the existence of a chemical compound to adopt different crystalline arrangements. The different solid forms display a variation in solid-state properties such as crystal morphology, solubility, density, stability, and compressibility which in turn influence the handling and processing properties of the compound, as well as the product performance such as bioavailability and shelf-life in the case of pharmaceutical compounds. With the increasing structural complexity of high value-added products, systems with multiple polymorphs are more frequently encountered in the pharmaceutical industry and it is pertinent to have a consistent and reliable production process for the preferred polymorph to achieve feasible economic yield as well as for regulatory compliance. Crystals of the most thermodynamically stable form can be grown as long as sufficient process time is provided under appropriate crystallization conditions in terms of the temperature and stirring rate. It is generally more challenging to produce large crystals of the metastable form; the main challenge being to prevent the cross nucleation of the stable polymorph. The same holds true for pseudo-polymorphic systems, in which solvated crystals can form. This presentation considers the design and control of metastable crystals in an enantiotropic pseudo-polymorphic system in which L-phenylalanine crystals are produced, as a model system for other enantiotropic polymorphic and pseudo-polymorphic systems.

Industrial batch crystallization processes are typically designed to follow a temperature or solvent addition profile, determined through models of the nucleation and crystal growth kinetics. The kinetic information can be obtained from a succession of crystallization experiments [1, 2, 3], but this can be time consuming for complex crystallization systems with multiple polymorphs. A different approach, which does not require accurate kinetic data, is to operate the crystallizer based on the solution concentration measurement so that it follows a suitable supersaturation profile to minimize unwanted secondary nucleation, which would widen the crystal product size distribution [4, 5, 6]. This concentration control strategy has been shown to be less sensitive to variations in growth and nucleation kinetics and most practical disturbances for the cooling crystallization of a non-polymorphic system [7]. This approach also has been shown to facilitate the rapid development of batch antisolvent crystallization processes [8]. In recent years we have also applied this approach to monotropic polymorphic crystallizations [9].

This paper describes the application of this concentration control methodology to the pseudo-dimorph system of L-phenylalanine to preferentially crystallize the anhydrate form which is metastable below the transition temperature. Earlier work [10] aimed at the selective crystallization of this form utilized additives such as ammonium sulfate and dextrose to decelerate the transformation to the monohydrate form below the transition point, through lowering of the anhydrate form solubility or inhibition of mass transfer to the crystal surface of the monohydrate crystals. Another approach selectively crystallized the anhydrate form in reverse micelles [11]. Our control strategy is based on in-situ solution concentration measurements to follow an appropriate supersaturation profile setpoint to minimize secondary nucleation of both types. This represents a more generic approach that does not necessitate the use of additives or other system-specific properties of the compound.

This presentation shows results of applying feedback control to reproducibly and efficiently produce crystals of the metastable pseudo-polymorph. L-phenylanine is an enantiotropic pseudo-dimorph system [10]; the anhydrate form resembles rhombic platelets, while the monohydrate form has a needle-like morphology. In water, the anhydrate form is metastable below the transition temperature which is reported to be 37°C [10]. In this paper, a mixed solvent system consisting of water and 25% of 2-propanol is used, for which the transition temperature was found at the lower temperature of 27°C. The anhydrous form of L-phenylalanine is desirable for industrial purposes because of its ease of downstream processing. An earlier work [10] studied the anhydrate to monohydrate transformation; the transformation rate was quantified based on the change in polymorph composition obtained from XRD results of samples removed periodically. For anhydrate seeds with typical size of 100 microns, the induction time for the monohydrate form is reported to be 3 and 22 hours at 15°C and 30°C, respectively.

ATR-FTIR spectroscopy coupled with a calibration model constructed using chemometric techniques [12, 13, 14] was used to provide in-situ solution concentration measurement. Laser backscattering (known commercially as the Lasentec FBRM), which measures characteristics of the crystal size distribution in-situ, was used to monitor the extent of secondary nucleation. The seeded batch cooling crystallizations were implemented with the concentration control approach using different supersaturation profiles to obtain the most appropriate batch crystallization recipe for selectively growing anhydrate L-phenylalanine crystals.

References

1. Togkalidou, T; Tung, H. H.; Sun, Y.; Andrews, A. T.; Braatz, R. D. Ind. Eng. Chem. Res. 2004, 43, 6168-6181.

2. Worlitschek, J.; Mazzotti, M. Crystal Growth & Design 2004, 4, 891-903.

3. Miller, S. M.; Rawlings, J. B. AIChE J. 1994, 40, 1312-1327.

4. Feng, L. L.; Berglund, K. A. Crystal Growth & Design 2002, 2, 449-452.

5. Liotta, V.; Sabesan, V. Org. Process Res. Dev. 2004, 8, 488-494.

6. Gron, H.; Borissova, A.; Roberts, K. J. Ind Eng Chem Res 2003, 42, 198-206.

7. Fujiwara, M.; Chow, P. S.; Ma, D. L.; Braatz, R. D. Crystal Growth & Design 2002, 2, 363-370.

8. Zhou, G. X.; Fujiwara, M.; Woo, X. Y.; Rusli, E.; Tung, H. H.; Starbuck, C.; Davidson, O.; Ge, Z.; Braatz, R. D. Crystal Growth & Design 2006, 6, 892-898.

9. N. Kee, et al, Pharmaceutical Reviews, submitted.

10. Mohan, R.; Koo, K. K.; Strege, C.; Myerson, A. S. Ind. Eng. Chem. Res. 2001, 40, 6111-6117.

11. Yano, J.; Milhofer, H. F.; Wachtel, E.; Garti, N. Langmuir 2000, 16, 10005-10014.

12. Workman, J. J.; Mobley, P. R.; Kowalski, B. R.; Bro, R. Appl. Spec. Rev. 1996, 31, 73-124.

13. Mobley, P. R.; Kowalski, B. R.; Workman, J. J.; Bro, R. Appl. Spec. Rev. 1996, 31, 347-368.

14. Bro, R.; Workman, J. J.; Mobley, P. R.; Kowalski, B. R. Appl. Spec. Rev. 1997, 32, 237-261.