8705
2007, 111, 8705-8707 Published on Web 07/12/2007
Alteration of Polymorphic Selectivity through Different Crystallization Mechanisms Occurring in the Same Crystallization Solution Mingcan Xu and Kenneth D. M. Harris* School of Chemistry, Cardiff UniVersity, Park Place, Cardiff CF10 3AT, Wales ReceiVed: June 6, 2007
We report a case in which two different crystallization mechanisms occurring in the same crystallization experiment are found to yield different polymorphic outcomes. In particular, we focus on crystallization of glycine from neutral aqueous solution. Crystallization in the bulk solution gives only the metastable R-polymorph, as observed in previous studies, whereas crystallization by evaporation of a thin film of the solution on the walls of the crystallization vessel is found to give rise to the thermodynamically stable γ-polymorph, and furthermore produces an uncharacteristic crystal morphology for this polymorph. A detailed set of control experiments are described that elucidate mechanistic details relating to the latter crystallization process. The fact that crystallization on the walls of a crystallization vessel can yield a different polymorphic outcome from crystallization in the bulk solution in the same experiment has potentially much wider significance with regard to other polymorphic systems.
Polymorphism arises when a given type of molecule can form different crystal structures.1 Although different polymorphs have the same chemical composition, their solid-state properties are generally different as a consequence of their different crystal structures. In recent years, there has been a huge upsurge of activity in this field, driven both by fundamental scientific curiosity and by industrial necessity. In particular, much interest has been devoted toward finding and characterizing as many polymorphs as possible of specific molecules of interest (e.g., a drug substance or pigment), in order that the polymorph with the most desirable properties for a targeted application can be selected. Different polymorphic forms can be produced by a variety of strategies, including conventional crystallization from solution using different solvents, different crystallization conditions (e.g., different regimes of temperature and/or concentration), crystallization in the presence of additives that promote nucleation of a specific polymorph, or by an appropriate solidstate transformation (e.g., dehydration of a hydrate crystal phase). Understanding and controlling such crystallization processes are clearly of paramount importance in many aspects of polymorphism research. Here we demonstrate that different crystallization mechanisms occurring for the same crystallization solution can give rise to a different polymorphic distribution of the crystallization product. The work reported here has focused in particular on a crystallization mechanism that produces the γ-polymorph of glycine from neutral aqueous solution (conditions that normally promote the formation of the R-polymorph), although we emphasize that the implications of the results are of wider relevance within the polymorphism field in general (rather than being of specific relevance only within the domain of glycine polymorphism). Mechanistic aspects of the crystallization * To whom correspondence should be addressed. E-mail: harriskdm@ cardiff.ac.uk.
10.1021/jp074375j CCC: $37.00
process that lead to the formation of the γ-polymorph are elucidated through a detailed set of control experiments. Glycine has received considerable attention in polymorphism research.2-5 Under ambient conditions, three polymorphs (denoted R, β, and γ) of glycine are known,2a-d and another polymorph has also been observed at high pressure.2e Under ambient conditions, the γ-polymorph is thermodynamically stable, and the order of stability2a,3 is γ > R > β. Crystallization from supersaturated aqueous solution at neutral pH is found to lead reproducibly to the metastable R-polymorph. The stable γ-polymorph, on the other hand, is obtained from sufficiently basic (pH > ca. 8.9) or sufficiently acidic (pH < ca. 3.8) aqueous solutions, but is not obtained spontaneously from neutral aqueous solution under normal conditions.2b,3a These observations suggest that nucleation of the γ-polymorph does not compete successfully with nucleation of the R-polymorph in neutral aqueous solution. Other reported strategies to induce nucleation of the γ-polymorph include addition of electrolytes4a (e.g., sodium chloride, sodium fluoride, or sodium nitrate), application of polarized laser radiation,4b application of a dc electric field,4c or evaporation of microdroplets.4d In our crystallization experiments, a homogeneous aqueous solution (25 mL) of glycine (19.2 wt %) was prepared at 353 K and transferred to a glass beaker (100 mL). On cooling to ambient temperature, the solution becomes supersaturated and crystals form within the solution. Powder X-ray diffraction confirms that these crystals are the R-polymorph (Figure 1a), in agreement with previous reports in the literature discussed above. However, on leaving the solution for a sufficient length of time (typically several weeks), a substantial amount of crystals are found to grow on the walls of the beaker above the level of the crystallization solution (Figure 2a), spreading gradually upward and even growing on the outside surface of the beaker. The same phenomenon is observed whether deionized water or © 2007 American Chemical Society
8706 J. Phys. Chem. B, Vol. 111, No. 30, 2007
Figure 1. Powder X-ray diffraction patterns recorded for (a) crystals grown within the crystallization solution (monophasic sample of the R-polymorph), (b) the sample of crystals collected from the walls of the crystallization beaker (mixture of the R- and γ-polymorphs), and (c) the needlelike crystals separated by hand from the sample of crystals collected from the walls of the crystallization beaker (monophasic sample of the γ-polymorph).
Figure 2. (a) Formation of glycine crystals on the walls of a glass beaker, (b) crystals of the γ-polymorph collected from the walls of the beaker, and (c) formation of glycine crystals on a glass slide partially immersed in an aqueous solution of glycine inside a Teflon beaker.
distilled water is used, and whether the crystallization experiment is carried out under flowing nitrogen or under a static atmosphere of air. Importantly, powder X-ray diffraction (Figure 1b) indicates that the crystals formed on the walls of the beaker are a mixture of the R- and γ-polymorphs of glycine (usually containing an excess of the R-polymorph). It is important to reiterate (see above) that under normal conditions of crystallization from bulk neutral aqueous solution, the γ-polymorph is not obtained. The two types of crystals formed on the walls of the beaker are readily separated by hand based on their distinct morphologies. Thus, the crystals of the R-polymorph have the characteristic prismatic shape reported previously,2b,4a whereas the crystals of the γ-polymorph have a needlelike morphology. Powder X-ray diffraction patterns recorded separately for the two sets of crystals separated by hand indicate that they are monophasic samples of the R- and γ-polymorphs (Figure 1b and 1c). Single-crystal X-ray diffraction and polarized optical microscopy confirm that the needles are indeed single crystals. To our knowledge, this needlelike morphology is unprecedented for the γ-polymorph, for which the characteristic morphology is trigonal pyramidal crystals.4a Many of the needlelike crystals of the γ-polymorph appear to grow in “clumps” (Figure 2b), with the long-axes of the needles in the clump converging, at least approximately, toward a common point. Single-crystal X-ray diffraction and polarized optical microscopy confirm that the long axis of the needle morphology is the unique c-axis of the trigonal structure2d of the γ-polymorph. Although the observation of crystals growing on the walls of a crystallization vessel above a crystallization solution is not uncommon, to our knowledge there is no other reported example
Letters in which this process gives rise to a polymorph that is not formed in the bulk solution in the same crystallization experiment (and, furthermore, is not observed under conventional crystallization conditions from the same solvent). Because the glycine molecules that form the crystals on the walls of the beaker originate from the aqueous solution, transport of glycine from the solution to the walls of the beaker presumably occurs either via the vapor phase or by diffusion along the walls of the beaker. Although the equilibrium vapor pressure of glycine above an aqueous solution is very low,6 a vapor phase mechanism for transport of glycine cannot be discounted, as even a small equilibrium vapor pressure may be sufficient to transport glycine molecules to the walls of the beaker. However, if transport of glycine molecules does occur via the vapor phase, we should expect to observe crystal growth of a seed crystal suspended within the beaker above the crystallization solution. In separate control experiments, single crystals of known mass of the R- and γ-polymorphs were suspended above saturated aqueous solutions of glycine. After several weeks, the crystals growing on the walls of the beaker had reached a height significantly higher than the position of the seed crystal. However, in all cases, for seed crystals of both the R- and γ-polymorphs, there was no visible change in the size of the seed crystal, and there was no change (within 0.001 g) in the measured mass of the crystal. These observations indicate that transport of glycine molecules in the vapor phase is not a significant mechanism. An alternative mechanism involves diffusion of the aqueous glycine solution along the walls of the glass beaker, which clearly requires that the glass surface is wetted by the solution. For an aqueous solution, such diffusion is clearly promoted by a hydrophilic surface (such as the glass surface used here) and inhibited by a hydrophobic surface (such as Teflon). To assess this issue, crystallization of glycine was carried out in a Teflon beaker (100 mL). Crystallization within the aqueous solution was observed as normal (giving the R-polymorph), but no crystals were observed to form on the walls of the beaker. At this stage, a glass slide was partially immersed in the crystallization solution (Figure 2c). Over a period of time, glycine crystals were observed to form on the surface of the glass slide above the solution. However, at no stage were any crystals observed to form on the walls of the Teflon beaker. These results demonstrate that the wettability of the surface is critical for the transport of glycine molecules and subsequent crystallization on the walls of the beaker. The above evidence supports the view that the crystallization process involves diffusion of a thin film of saturated aqueous glycine solution along the walls of the beaker, followed by evaporation of water from the thin film and crystallization of glycine. A relevant question is whether there is a continuous throughput of solution to replenish the water lost by evaporation and to sustain the continued growth of glycine crystals. To assess this issue, the crystallization process was started as described above, and after crystals of glycine had begun to form on the walls of the beaker, several drops of an aqueous Prussian blue solution were added to the crystallization solution (Figure 3). Over several hours, the blue color was observed to rise up the walls of the beaker and to spread across the glass surface, demonstrating clearly that transport of the solution is sustained along the walls of the beaker to maintain the crystal growth process. As a final control experiment, the crystallization process was carried out under conditions in which a humid atmosphere was maintained above the solution to inhibit evaporation (both from
Letters
Figure 3. Glycine crystals growing on the walls of a glass beaker as a function of time after addition of several drops of aqueous Prussian blue to the crystallization solution (time from the first photo to the last photo ca. 2 days).
the bulk solution and from the thin film of solution on the walls of the beaker). No significant amounts of crystals were observed to form on the walls of the beaker (after periods of time comparable to those discussed above), confirming that evaporation of the thin film is an important component of the process for formation of crystals on the walls of the beaker. In summary, the crystallization process described here for the formation of the γ-polymorph of glycine from neutral aqueous solution involves the following components. First, the aqueous solution of glycine is drawn upward by wetting the surface of the glass beaker above the level of the bulk solution, forming a thin film of solution on the walls of the beaker. Second, evaporation of water from this thin film leads to crystallization of glycine. The aqueous solution in the thin film is continually replenished, resulting in substantial transport of glycine molecules from the bulk solution to the walls of the beaker. We note that crystals of the R-polymorph still grow as normal within the bulk solution, and that the γ-polymorph is formed only on the walls of the glass beaker above the level of the bulk solution. It is important to emphasize that the bulk solution is also in contact with the same type of glass surface (at the bottom of the beaker) and that evaporation of the bulk solution also occurs, but for the bulk solution these glass-solution and air-solution interfaces are well separated from each other. On the other hand, in the case of the crystallization mechanism that occurs on the walls of the beaker arising from thin film evaporation, the glasssolution and air-solution interfaces can both exert a direct influence on the same process. This may be an important factor underlying the different polymorphic outcome observed in the crystallization by thin-film evaporation. Clearly, an important issue for future research is to establish the role of the glass surface in this process and the extent to which specific glycinesurface interactions are responsible for mediating the nucleation of the γ-polymorph. Finally, it is relevant to note that a recent paper7 claims that the surface of liquid water is acidic (pH < 4.8), and to
J. Phys. Chem. B, Vol. 111, No. 30, 2007 8707 contemplate the possible connection between this claim and the results reported in the present paper on the formation of the γ-polymorph by thin-film evaporation. Given the wellestablished fact that the formation of the γ-polymorph of glycine is favored in sufficiently acidic (bulk) aqueous solution,2b,3a the claim that water surfaces are acidic may suggest that a crystallization process occurring close to the surface of an aqueous solution might also promote the formation of the γ-polymorph. While this suggestion is certainly in agreement with the observations reported here, we must nevertheless regard the possible connection between these observations and the claims reported in ref 7 as speculative at the present stage, but further work to substantiate the possible connection is clearly merited. Acknowledgment. We are grateful to Research Councils U.K. for financial support through a Basic Technology grant (Control and Prediction of the Organic Solid State), to Dr. Colan Hughes for discussions, and to Dr. Zolta´n Ga´l (Oxford Diffraction Ltd.) for assistance in carrying out single-crystal X-ray diffraction measurements. References and Notes (1) (a) Dunitz, J. D. Pure Appl. Chem. 1991, 63, 177. (b) Bernstein, J. J. Phys. D: Appl. Phys. 1993, 26, B66. (c) Dunitz, J. D. Acta Crystallogr., Sect. B 1995, 51, 619. (d) Bernstein, J. Polymorphism in Molecular Crystals; Oxford University Press: Oxford, 2002. (e) Davey, R. J. Chem. Commun. 2003, 1463. (f) Braga, D.; Grepioni, F. Chem. Commun. 2005, 3635. (g) Bernstein, J. Chem. Commun. 2005, 5007. (2) (a) Perlovich, G. L.; Hansen, L. K.; Bauer-Brandl, A. J. Therm. Anal. Cal. 2001, 66, 699. (b) Towler, C. S.; Davey, R. J.; Lancaster, R. W.; Price, C. J. J. Am. Chem. Soc. 2004, 126, 13347. (c) Jo¨nsson, P.-G.; Kvick, Å. Acta Cryst. B 1972, 28, 1827. (d) Kvick, Å.; Canning, W. M.; Koetzle, T. F.; Williams, G. J. B. Acta Cryst. B 1980, 36, 115. (e) Dawson, A.; Allan, D. R.; Belmonte, S. A.; Clark, S. J.; David, W. I. F.; McGregor, P. A.; Parsons, S.; Pulham, C. R.; Sawyer, L. Cryst. Growth Des. 2005, 5, 1415. (3) (a) Boldyreva, E. V.; Drebushchak, V. A.; Drebushchak, T. N.; Paukov, I. E.; Kovalevskaya, Y. A.; Shutova, E. S. J. Therm. Anal. Calorim. 2003, 73, 409. (b) Boldyreva, E. V.; Drebushchak, V. A.; Drebushchak, T. N.; Paukov, I. E.; Kovalevskaya, Y. A.; Shutova, E. S. J. Therm. Anal. Calorim. 2003, 73, 419. (4) (a) Bhat, M. N.; Dharmaprakash, S. M. J. Cryst. Growth 2002, 242, 245. (b) Garetz, B. A.; Matic, J. Phys. ReV. Lett. 2002, 89, 175501. (c) Aber, J. E.; Arnold, S.; Garetz, B. A.; Myerson, A. S. Phys. ReV. Lett. 2005, 94, 145503. (d) He, G.; Bhamidi, V.; Wilson, S. R.; Tan, R. B. H.; Kenis, P. J. A.; Zukoski, C. F. Cryst. Growth Des. 2006, 6, 1746. (5) (a) Shimon, L. J. W.; Vaida, M.; Addadi, L.; Lahav, M.; Leiserowitz, L. J. Am. Chem. Soc. 1990, 112, 6215. (b) Myerson, A. S.; Lo, P. Y. J. Cryst. Growth 1990, 99, 1048. (c) Weissbuch, I.; Leiserowitz, L.; Lahav, M. AdV. Materials 1994, 6, 952. (d) Gidalevitz, D.; Feidenhans, R.; Matlis, S.; Smilgies, D.-M.; Christensen, M. J.; Leiserowitz, L. Angew. Chem., Int. Ed. Eng. 1997, 36, 955. (e) Torbeev, V. Y.; Shavit, E.; Weissbuch, I.; Leiserowitz, L.; Lahav, M. Cryst. Growth Des. 2005, 5, 2190. (f) Weissbuch, I.; Torbeev, V. Y.; Leiserowitz, L.; Lahav, M. Angew. Chem., Int. Ed. 2005, 44, 3226. (g) Sun, X. Y.; Garetz, B. A.; Myerson, A. S. Cryst. Growth Des. 2006, 6, 684. (h) Hughes, C. E.; Hamad, S.; Harris, K. D. M.; Catlow, C. R. A.; Griffiths, P. C. Faraday Discuss., in press. (6) Wolfenden, R.; Andersson, L.; Cullis, P. M.; Southgate, C. C. B. Biochemistry 1981, 20, 849. (7) Buch, W.; Milet, A.; Va´cha, R.; Jungwirth, P.; Devlin, J. P. Proc. Nat. Acad. Sci. 2007, 104, 7342.
