Polymorphism in Piroxicam - Crystal Growth & Design (ACS


Polymorphism in Piroxicam - Crystal Growth & Design (ACS...

0 downloads 125 Views 400KB Size

CRYSTAL GROWTH & DESIGN 2004 VOL. 4, NO. 6 1091-1098

Communications Polymorphism in Piroxicam Agam R. Sheth,† Simon Bates,‡ Francis X. Muller,§ and David J. W. Grant*,† Department of Pharmaceutics, College of Pharmacy, University of Minnesota, Weaver-Densford Hall, 308 Harvard Street Southeast, Minneapolis, Minnesota 55455-0343, SSCI Inc., 3065 Kent Avenue, West Lafayette, Indiana 47906-1076, and GlaxoSmithKline Pharmaceuticals, P.O. Box 1539, 709 Swedeland Road, King of Prussia, Pennsylvania 19406-0939 Received April 2, 2004

ABSTRACT: Piroxciam is a polymorphic drug. However, reports on the number and nomenclature of the polymorphs of piroxicam and the complete hydrogen-bonding patterns of piroxicam molecules in the crystal forms are in conflict and are sources of confusion, which we attempt to clarify. The difference in energy of the two polymorphs, I and II, of piroxicam arises predominantly from the difference between their lattice energies, rather than between their conformational energies. The detailed hydrogen-bonding networks of the two polymorphs are described and compared. Despite stabilization of the polymorphs by hydrogen bonds, a loss of polymorphic memory was observed upon cryogrinding the two polymorphs, leading to differences in recrystallization behavior between amorphous piroxicam prepared from polymorphs I and II. Piroxicam, an enolic acid and an oxicam derivative, is a potent, long-acting, nonsteroidal, antiinflammatory drug.1-3 Piroxicam is employed as a model drug in the present investigation of the effect of processing on the phase changes of pharmaceuticals at the molecular level. Table 1 summarizes the known reports on polymorphs of piroxicam and shows inconsistencies of nomenclature, preparation, and properties. A detailed comparison of the simulated (i.e., theoretically calculated) and experimental powder X-ray diffraction (PXRD) patterns can confirm or refute that the reported molecular and crystal structures of the compound are the same as those of the crystalline powder under investigation. Two polymorphs of piroxicam, prisms (Figure 1a) and needles (Figure 1b), commonly crystallize from solution. Simple evaporative crystallization reveals the following solvent effects. (i) A solvent, such as acetone or dichloromethane, sometimes yield prisms and sometimes needles. (ii) A solvent, such as ethyl acetate or tetrahydrofuran, yields a mixture of prisms and needles. (iii) Benzene or toluene consistently yield prisms. (iv) The factors that affect the polymorphic form also affect the crystal habit; for example, the polymorph that crystallizes as prisms also crystallizes as short rods from dichloromethane and plates from dioxane. Thus, despite previous reports describing methods of crystallizing the different polymorphs of piroxicam,2-6 the nature of the polymorphic form that crystallizes from solution depends on an interplay of several factors. The factors include the polarity of the solvent, initial supersaturation, and, depending on the * To whom correspondence should be addressed. Tel: 612-624-3956. Fax: 612-625-0609. E-mail: [email protected]. † University of Minnesota. ‡ SSCI Inc.. § GlaxoSmithKline Pharmaceuticals.

crystallization method, rate of cooling of the solution or evaporation of the solvent. Furthermore, although not studied in the present work, impurities will also likely affect the polymorph crystallizing from solution. Two crystal structures of piroxicam, reference codes BIYSEH7 and KAFYAR,8 are indexed in the Cambridge Structural Database (CSD).9 The experimental PXRD pattern of the prisms closely matches that calculated from BIYSEH in 2θ angles [mean difference, ∆(2θ) ) 0.07 ( 0.03°] but not in 2θ intensities. Using DASH10 and an indexing program,11 we redetermined the crystal structure corresponding to BIYSEH at 25 °C and found that the PXRD pattern calculated from this redetermined structure agrees well with the experimental PXRD patterns in both 2θ angles and 2θ intensities at 25 °C. This form is here designated form I. Table 1 shows that form I5,12 has also been named β,8,19 cubic form4,6 (even though the crystal system is monoclinic), and form A.3 PXRD confirms that the starting material of piroxicam used here (Sigma Chemical Company, St. Louis, MO) is also form I. The redetermined crystal structure of form I at 25 °C has the same unit cell constants and molecular packing as BIYSEH, but the atomic positions are slightly displaced. The observed variation in the relative intensities between the various PXRD patterns of samples crystallized on different occasions from different solvents can be attributed to differences in crystallite size and to slight differences in atomic positions between the BIYSEH and the redetermined form I crystal structure at 25 °C. For the form that crystallizes as needles, several authors5,6,8,12,13,19 based their studies of the crystal and molecular structures on those of KAFYAR.8 However, we found that the experimental PXRD pattern of this needle form does not match the PXRD pattern calculated from KAFYAR. Furthermore, a recent crystal structure of

10.1021/cg049876y CCC: $27.50 © 2004 American Chemical Society Published on Web 09/01/2004

1092

Crystal Growth & Design, Vol. 4, No. 6, 2004

Communications

Table 1. Summary of the Reported Solid State Properties of Piroxicam Polymorphs year and ref

major findings reported

19822

Piroxicam exists in “two possible tautomeric forms.” The “best solvent for recrystallization is methanol from which piroxicam crystallizes in the form of small needles.” The crystal structure of piroxicam was described.

19827

19853

Piroxicam crystallizes in two forms, form A as white needles and form B as yellow prisms.

19864

Two “interconvertible” forms of piroxicam “needle form” and “cubic form” crystallize depending on the solvent and cooling rate. Reck et al. report a “new modification” of piroxicam, crystallized from methanol, called R-piroxicam. The form of piroxicam reported in 19827 was termed β-piroxicam. Reck et al. relabel the original R-piroxicam8 as R1, and identify a new form, R2. Vrecer et al. report four polymorphic forms of piroxicam, forms I, II, III, and IV. Forms I and II crystallize from solution, whereas forms III and IV were obtained from amorphous piroxicam, which was prepared by melt/quench cooling. Piroxicam exists in “two modifications”, R and β. The effect of solvent on the polymorph crystallizing from solution was reported.

19888

199013 19915

199119

199622

19996

200312

Amorphous piroxicam is prepared by the melt spinning method and crystallizes rapidly being “almost completed at room temperature within few hours from the amorphous preparation.” The crystalline forms obtained are different from the starting material. Two polymorphs of piroxicam are labeled “white needle crystals” crystallizing independent of the solvent and cooling rate and “yellow cubic crystals”, crystallizing in the presence of a “precipitating agent like water used.” Vrecer et al. report three polymorphs of piroxicam, forms I, II, and III. The polymorph crystallizing from solution depends on the crystallization rate and polarity of the solvent. The crystal structure of form II was solved.

piroxicam12 gives a calculated PXRD pattern that closely matches the experimental PXRD pattern of the needle polymorph of piroxicam in 2θ angles [∆(2θ) ) 0.04 ( 0.03°] and in 2θ intensities. This crystalline form is here designated piroxicam form II. Table 1 shows that form II5,12 has also been named R,19 R1,8,13 R2,8,13 needle form,4,6 and form B.3 Vrecer et al.12 recently reported a new polymorph of piroxicam, form III, whose crystallization we confirmed. The crystallization behavior reported above suggests that forms I and II are energetically similar. In general, the energy difference between polymorphs has contributions from the differences in conformational energy and lattice energy. Ab initio calculations using Gaussian 0314 provide the contribution from conformational energies. Forms I and II contain a single molecule of piroxicam in their respective asymmetric units. The energies of this molecule in the two polymorphs were calculated using TD-B3LYP/6-311+G(2d,p).15 The geometry used in this calculation was determined by optimizing only the hydrogen positions of the crystal structure employing B3LYP/6-31G+(d,p).15 A small energy difference was found between the piroxicam molecules directly removed from their respective crystal lattices (∆E ) 1.23 kcal/mol). While such a difference is typical

present comments The reported peaks in the PXRD pattern of piroxicam do not match those of the currently known polymorphs. The crystallization conditions for growing single crystals were not reported. This crystal structure (space group monoclinic P21/c) is indexed in the CSD as BIYSEH.7,9 The crystal structures were not solved, and the PXRD data were not reported. Because the space group of form A was reported to be monoclinic P21/c, it appears that form A corresponds to the polymorph, the crystal structure of which was solved in 1982.7 Form B is actually piroxicam monohydrate. The reported PXRD patterns do not match any other known form. R-Piroxicam is indexed in the CSD as KAFYAR. However, the PXRD pattern predicted from its crystal structure does not match any known form. The crystallization conditions for R-piroxicam were not reported.

The conclusions were based only on DSC data, not PXRD. The melting of piroxicam is accompanied by chemical degradation; hence, melt/quench cooling is not an optimal technique for preparing amorphous piroxicam. The hydrogen bond pattern of form R is described with respect to the crystal structure solved in 1988,8 even though that structure does not correspond to any known polymorph. The melting of piroxicam is accompanied by chemical degradation; hence, melt/quench cooling is not an optimal technique for preparing amorphous piroxicam.

These findings conflict with those in 19864 and 1991,5,19 which report the effect of solvent on the polymorph of piroxicam crystallizing from solution. These findings conflict with the authors previous findings in 19915 regarding the number and nomenclature of piroxicam polymorphs.

of differences in conformational energy between polymorphs,16,17 it is rather large for piroxicam molecules in forms I and II, which have similar conformations and the same configuration with respect to its rotatable bonds. Indeed, this energy difference vanished upon full geometry optimization using B3LYP/6-31G+(d,p). The initial energy difference might, therefore, have thus arisen from slight strains caused by residual errors in the heavy atom coordinates, which need not be so great as to cause such large energy differences. The existence of such an energy difference, which disappears upon further geometry optimization, has been reported for the conformational polymorphism of 5-methyl-2-[(2-nitrophenyl)amino]-3-thiophenecarbonitrile.17 Because of insignificant differences in the conformational energy of the piroxicam molecule in polymorphs I and II, differences in lattice energy provide the dominant contribution to the difference in energy between the polymorphs. It is likely that hydrogen bonds contribute significantly to the difference in lattice energies of forms I and II of piroxicam. To understand the polymorphism of piroxicam, we wish to describe and interpret the similarities and differences in the hydrogen bonding. Hydrogen bonds were modeled

Communications

Crystal Growth & Design, Vol. 4, No. 6, 2004 1093

Figure 1. Optical micrographs of piroxicam crystals. (a) Prisms of form I crystallized from toluene and (b) needles of form II crystallized from dichloromethane.

using positions of hydrogen and nonhydrogen atoms obtained from previously solved crystal structures of forms I7 and II12 and a maximum H-A distance of 3.0 Å and a minimum D-H-A angle of 90°. Hydrogen bonds with a H-A distance of >3.0 are considered too weak to influence the crystal structure and packing.16 Figure 2 shows the hydrogen bond network in forms I and II. Even though the crystal structure of form I has previously been solved,7 the complete hydrogen bond pattern has not yet been described. Conflicting reports18 describe the hydrogen bond pattern as an infinite chain7 or as dimers.19 Figure 2 shows that the hydrogen bond network of form I contains piroxicam dimers that extend along the x- and y-axes. The intramolecular N-H‚‚‚N hydrogen bond is newly reported here. The hydrogen bond network in form II, shown in Figure 2, is in agreement with that reported by Vrecer et al.12 and consists of piroxicam molecules arranged in infinite sheets. Each molecular sheet runs parallel to the 00l family of planes (perpendicular to the z-axis). Intermolecular O-H‚‚‚O hydrogen bonds link adjacent sheets. For comparing hydrogen bonding among polymorphs, hydrogen bond lengths and angles20 have been superseded by graph sets.16,21 Table 2 gives the unitary motifs (on the diagonals) and the binary graph sets (off the diagonal) of forms I and II. Figure 2 and Table 2 show that the graph sets of forms I and II have identical intramolecular unitary motifs, S(6) and S(5), corresponding to the O-H‚‚‚O and N-H‚‚‚N hydrogen bonds, respectively. Furthermore, forms I and II also comprise a common binary graph set, S22(9), describing the O-H‚‚‚O and N-H‚‚‚N intramolecular hydrogen bonds. The differences in the hydrogen bond network between forms I and II are reflected in the intermolecular unitary motif and other second level graph sets. Form I dimers are identified by the first level graph set R22(14) describing the N-H‚‚‚O intermolecular hydrogen bond, while form II chains are identified by the first level graph set C(7) that describes the O-H‚‚‚O intermolecular hydrogen bond. The binary graph sets in Table 2 are basic, i.e., of the lowest degree. It is possible to define larger graph sets of higher degrees. For example, R33(18) in the first section of Table 2 represents the basic binary graph set of form I for the intramolecular O-H‚‚‚O and the intermolecular N-H‚‚‚O hydrogen bonds; R44(22) is its complex binary graph set of the next higher degree.

Figure 2. Hydrogen bond patterns of forms I and II of piroxicam. Both forms contain only one molecule in the asymmetric unit; hence, the two molecules in the hydrogen bond patterns of forms I and II are related by symmetry.7,12 The graph set notation corresponds to that in Table 2.

Amorphous piroxicam shows a strong propensity to crystallize, so its preparation is challenging. Previously, authors have prepared amorphous piroxicam by the melt/ quench-cooled method5,22 and have studied its properties. While we also prepared X-ray amorphous piroxicam by this technique, we found that the melting of piroxicam is accompanied by chemical degradation, which we detected

1094

Crystal Growth & Design, Vol. 4, No. 6, 2004

Communications

Table 2. Graph Set Analysis of the Hydrogen Bonds in Forms I and II of Piroxicam Form I Hydrogen Bonds hydrogen bond intramolecular O-H‚‚‚O (a) intramolecular N-H‚‚‚N (b) intermolecular N-H‚‚‚O (c)

intramolecular O-H‚‚‚O (a)

intramolecular N-H‚‚‚N (b)

intermolecular N-H‚‚‚O (c)

S(6) S22(9) [S(6)R22(14)] ) R32(18)

S(5) R24(8)

R22(14)

intramolecular O-H‚‚‚O (a)

intramolecular N-H‚‚‚N (b)

intermolecular O-H‚‚‚O (c)

S(6) S22(9) C(7)[S(6)] ) C12(7)

S(5) C(7)[S(5)] ) C22(10)

C (7)

Form II Hydrogen Bonds hydrogen bond intramolecular O-H‚‚‚O (a) intramolecular N-H‚‚‚N (b) intermolecular O-H‚‚‚O (c)

by high-performance liquid chromatography (HPLC)23 and one-dimensional 800 MHz proton nuclear magnetic resonance (NMR) spectroscopy. Such degradation occurs to the extent of 2-8%, depending on the method used for melting, whether in an oil bath, on a glass slide, on a hot stage, or in differential scanning calorimetry (DSC). Because of chemical degradation, the use of melting to prepare the amorphous form is unsuitable, so another method is needed. This work exemplifies the importance of testing for chemical degradation when preparing amorphous phases of organic crystals by melting and quench cooling. Ball milling amorphizes piroxicam with no detectable chemical degradation, as shown by HPLC and NMR. However, regardless of the conditions of milling, time of milling, temperature of milling, or the intervention of intermittent “cool down” periods, the resulting solid always exhibits crystallinity. Such difficulties called for an alternative approach to obtain amorphous piroxicam by milling. We found that cryogenic milling24 of both form I and form II resulted in X-ray amorphous piroxicam (Figure 3), a material not previously described. Amorphous forms prepared by grinding usually contain seeds or nuclei, corresponding to memory of the respective

starting polymorph (especially when their crystal lattices are stabilized by hydrogen bonds), and are therefore susceptible to recrystallization of the respective starting polymorph, from which they were prepared. As described below, the behavior of amorphous piroxicam obtained by cryogrinding forms I (PAI) and II (PAII) differed. PAI and PAII were freshly prepared and heated immediately in a variable temperature X-ray diffractometer (VTXRD) and DSC. The VTXRD pattern of PAI at 30 °C in Figure 4 shows crystalline peaks as well as an amorphous halo, indicating that recrystallization has already begun. The pattern at 30 °C is very similar to that of form I, although certain peaks, characteristic of form III and marked by an asterisk, have low intensities. The disappearance of form III peaks between 30 and 70 °C suggests the conversion of form III f I. Presumably, PAI crystallizes predominantly as form I with traces of form III, which rapidly convert to form I. In the VTXRD patterns of form I between 90 and 160 °C, three peaks between 25 and 30° 2θ, indicated by arrows in Figure 4, gradually shift to lower angles. This peak shift suggests lattice expansion of form I due to an increase in temperature; that is, an increase in temperature causes certain planes with relatively small

Figure 3. Powder X-ray diffraction pattern of amorphous piroxicam, PAI, prepared by cryogrinding form I, and amorphous piroxicam, PAII, prepared by cryogrinding form II.

Communications

Crystal Growth & Design, Vol. 4, No. 6, 2004 1095

Figure 4. VTXRD patterns of PAI at the temperatures (°C) stated on the right. The asterices indicate the emergence of form III peaks, while the arrows indicate expansion of the crystal lattice of form I with increasing temperature.

Figure 5. VTXRD patterns of PAII at the temperatures (°C) stated on the right.

1096

Crystal Growth & Design, Vol. 4, No. 6, 2004

Communications

Figure 6. PDF transform (from 3 to 41 Å) of form I (red) and form II (blue) cryoground for 60 min.

d-spacings in the crystal lattice of form I to gradually move apart. As a result of this expansion, the PXRD pattern of form I at room temperature (25 °C) differs slightly from that at higher temperatures (160°), suggesting similar, but not identical, crystal structures. Utilizing the experimental PXRD pattern of form I at 160 °C and BIYSEH, we performed crystal structure refinement using the Rietveld Method in MAUD.25 The PXRD pattern calculated from this refined crystal structure agrees well with the experimental PXRD pattern of form I at 160°, both in diffraction angles and in intensities. This crystal structure, corresponding to that of form I at 160°, comprised the same molecular packing as that of BIYSEH but with slightly different cell dimensions. The details of crystal structures of form I at room temperature and at 160° will be reported subsequently. In the VTXRD patterns of PAII (Figure 5), the emergence of peaks at 30-50 °C corresponds to the recrystallization of PAII. The PXRD pattern of these crystals differs from that at 30 °C in Figure 4 in that it corresponds exclusively to form III; no peaks of form I are evident. Figure 5 shows a polymorphic transformation beginning at about 70 °C, corresponding to III f I transformation, characterized by the disappearance of form III peaks and the appearance of form I peaks. This conversion overlaps with the lattice expansion of the newly formed form I, evident by a peak shift between 80 and 140 °C. The PXRD pattern at 140 °C in Figure 5 corresponds to form I. Interestingly, when PAI and PAII, after their preparation, were allowed to relax at 25 °C and 0% relative humidity for several days and were

then heated, they ultimately converted to form I. This conversion appeared to proceed via a different sequence of events. DSC results (not shown) support the above VTXRD observations. The VTXRD patterns, including their addition and subtraction, and the DSC traces indicate fast crystallization kinetics followed by rapid and complex transformations. It is also likely that rapid conversions overlap, resulting in mixtures of polymorphs. Three polymorphs of piroxicam have thus far been established and confirmed, forms I, II, and III. When a new polymorph is discovered and characterized, it is suggested that its designation be the next increment in this system. Clearly, the nature of the amorphous forms PAI and PAII differs. The above observations indicate a complete loss of polymorphic memory of form II upon cryogrinding, perhaps leading to the production of a more stable amorphous phase, PAII, which can therefore be expected to crystallize to a different polymorph. Indeed, PAII thus formed has a strong propensity to relax to form III. On the other hand, at least some polymorphic memory of form I appears to be retained in PAI, which recrystallizes predominantly to form I, although some form III character may also be present. The pair wise distribution function (PDF) transform was used to probe the differences in recrystallization behavior of PAI and PAII. The PDF transform is a Fourier sine transform of the reduced structure factor representation of the PXRD data and is a powerful tool for studying crystalline-amorphous relationships.26 The PDF transform, when applied to PXRD data, gives an instantaneous picture of the atom-atom distances in the material being

Communications

Crystal Growth & Design, Vol. 4, No. 6, 2004 1097

Figure 7. PDF transform (from 19 to 49 Å) of form I (red) and form I cryoground for 60 min (yellow).

studied. The PDF transform of PAI and PAII (Figure 6) has the same structure at short correlation lengths. However, the long-range order of PAI and PAII was found to differ, indicating that the nature of the amorphous material depends on the polymorphic form from which it was prepared. Figure 7 shows that the longer range atom-atom interactions in PAI are similar to those in form I and could steer the recrystallization of PAI toward a crystalline form similar to form I. However, no correlation was found between the PDF transform of PAII and form II and only a weak correlation between the PDF transform of PAII and form I. Thus, PAI shows similarities to form I whereas PAII is different from form II. Despite stabilization of the crystal lattices of forms I and II by hydrogen bonds, PAI shows partial loss, while PAII shows complete loss, of polymorphic memory. The reduction of crystal size through grinding is manifest in the PDF transform as a dampening of the PDF peaks as the atom-atom distances come close to the average crystal size. The peak positions corresponding to atom-atom distances smaller than the average crystal size will be unaffected. PDF results will be discussed in greater detail in a subsequent manuscript. PAI and PAII appear to be strong glasses with a small change in heat capacity on heating. During heating in DSC, the glass transition (Tg) of PAII was not observed while that of PAI was weak and observed in only two of the three runs (extrapolated Tg onset 0.22 ( 0.94 °C). Also, PAI appears to be less stable than PAII at room temperature in view of the lower recrystallization temperature of PAI (onset, 48.01

( 2.38 °C; midpoint, 61.18 ( 0.76 °C) than PAII (onset, 63.12 ( 1.6 °C; midpoint, 70.92 ( 0.36 °C). Acknowledgments. We thank Professor Joel Bernstein for valuable comments, Professors F. Vrecer, M. Vrbinc, and A. Meden12 for kindly providing the cif file of their crystal structure, and Richard Bostwick at SPEX CertiPrep, Inc. for loan of the cryogenic mill. We also thank GlaxoSmithKline, through Dr. Fran Muller, for gifts of piroxicam and for the award of a studentship to A.R.S. Finally, we thank the University of Minnesota Supercomputing Institute for providing resources for computational modeling. Supporting Information Available: PXRD data of form I at 25 °C, form I at 160 °C, form II, and form III in xy format. PXRD pattern of form I at 25 °C averaged from several PXRD patterns of form I obtained from different samples. Experimental details. Details of the synthesis of forms II and III from form I. This material is available free of charge via the Internet at http:// pubs.acs.org.

References (1) The Merck Index, 13th ed.; Merck & Co., Inc.: Whitehouse Station, NJ, 2001; p 7588. (b) Majerus, P. W.; Broze, G. J.; Miletich, J. P.; Tollefsen, D. M. Goodman and Gilman’s The Pharmacological Basis of Therapeutics, 9th ed.; Hardman, J. G., Limbird, L. E., Eds.; McGraw-Hill: New York, 1996; pp 640-641. (2) Mihalic, M.; Hofman, H.; Kajfez, F.; Kuftinec, J.; Blazevic, N.; Zinic, M. Physicochemical and analytical characteristics of piroxicam. Acta Pharm. Jugosl. 1982, 32, 13-20.

1098

Crystal Growth & Design, Vol. 4, No. 6, 2004

(3) Kozjek, F.; Golic, L.; Zupet, P.; Palka, E.; Vodopivec, P.; Japelj, M. Physicochemical properties and bioavailability of two crystal forms of piroxicam. Acta Pharm. Jugosl. 1985, 35, 275-281. (4) Mihalic, M.; Hofman, H.; Kuftinec, J.; Krile, B.; Caplar, V.; Kajfez, F.; Blazevic, N. Piroxicam. Anal. Profiles Drug Subst. 1986, 15, 509-531. (5) Vrecer, F.; Srcic, S.; Smid-Korbar, J. Investigation of piroxicam polymorphism. Int. J. Pharm. 1991, 68, 35-41. (6) Csoka, G.; Balogh, E.; Marton, S.; Farkas, E.; Racz, I. Examination of the polymorphism of piroxicam in connection with the preparation of a new “soft-patch” type pharmaceutical dosage form. Drug. Dev. Ind. Pharm. 1999, 25 (6), 813816. (7) Kojic-Prodic, B.; Ruzic-Toros, Z. Structure of the antiinflammatory drug 4-hydroxy-2-methyl-N-2-pyridyl-2H-1λ6,2-benzothiazine-3-carboxamide-1,1-dioxide (piroxicam). Acta Crystallogr. 1982, B38, 2948-2951. (8) Reck, G.; Dietz, G.; Laban, G.; Gunther, W.; Bannier, G.; Hohne, E. X-ray studies on piroxicam modifications. Pharmazie 1988, 43, 477-481. (9) Cambridge Crystallographic Data Center. The Cambridge Structural Database; University Chemical Laboratory: Cambridge, United Kingdom, 1996. (b) Allen, F. H. The Cambridge Structural Database: A quarter of a million crystal structures and rising. Acta Crystallogr. 2002, B58, 380388. (c) Allen, F. H.; Motherwell, W. D. S. Applications of the Cambridge Structural Database in organic chemistry and crystal chemistry. Acta Crystallogr. 2002, B58, 407422. (d) Bruno, I. J.; Cole, J. C.; Edgington, P. R.; Kessler, M.; Macrae, C. F.; McCabe, P.; Pearson, J.; Taylor, R. New software for searching the Cambridge Structural Database and visualizing crystal structures. Acta Crystallogr. 2002, B58, 389-397. (10) Version 2.2; Cambridge Crystallographic Data Center: 2003. (11) SSCI Indexing Program, Version 1.3.4; SSCI, Inc.: 2004. (12) Vrecer, F.; Vrbinc, M.; Meden, A. Characterization of piroxicam crystal modifications. Int. J. Pharm. 2003, 256, 3-15. (13) Reck, G.; Laban, G. Prediction and establishment of a new crystalline piroxicam modification. Pharmazie 1990, 45, 257-259. (14) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill,

Communications

(15) (16) (17)

(18) (19)

(20) (21)

(22)

(23) (24)

(25) (26)

P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision A.1; Gaussian, Inc.: Pittsburgh, PA, 2003. Becke, A. D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98 (7), 56485652. Bernstein, J. Polymorphism in Molecular Crystals; Clarendon Press: Oxford, United Kingdom, 2002. Yu, L.; Stephenson, G. A.; Mitchell, C. A.; Bunnell, C. A.; Snorek, S. V.; Bowyer, J. J.; Borchardt, T. B.; Stowell, J. G.; Byrn, S. R. Thermochemistry and conformational polymorphism of a hexamorphic crystal system. J. Am. Chem. Soc. 2000, 122, 585-591. Taddei, P.; Torreggiani, A.; Simoni, R. Influence of environment on piroxicam polymorphism: vibrational spectroscopic study. Biopolymers (Biospectroscopy) 2001, 62, 68-78. Janik, M.; Malarski, Z.; Mrozinski, J.; Wajcht, J.; Zborucki, Z. Influence of solvent effect on polymorphism of 4-hydroxy2-methyl-N-2-pyridyl-2H-1,2-benzothiazine-3-carboxamide1,1-dioxide (piroxicam). J. Crystallogr. Spectrosc. Res. 1991, 21 (4), 519-522. Bernstein, J. Conformational polymorphism. III. The crystal and molecular structures of the second and third forms of iminodiacetic acid. Acta Crystallogr. 1979, B35, 360-366. Etter, M. C. Aggregate structures of carboxylic acids and amides. Isr. J. Chem. 1985, 25, 312-319. (b) Etter, M. C. Encoding and decoding hydrogen-bond patterns of organic compounds. Acc. Chem. Res. 1990, 23, 120-126. (c) Etter, M. C. Hydrogen bonds as design elements in organic chemistry. J. Phys. Chem. 1991, 95, 4601-4610. (d) Etter, M. C.; MacDonald, J. C.; Bernstein, J. Graph-set analysis of hydrogen-bond patterns in organic crystals. Acta Crystallogr. 1990, B46, 256-262. (e) Etter, M. C.; UrbanczykLipkowska, Z.; Zia-Ebrahimi, M.; Panunto, T. W. Hydrogenbond directed cocrystallization and molecular recongnition properties of diarylureas. J. Am. Chem. Soc. 1990, 112, 8415-8426. (f) Bernstein, J.; Davies, R. E.; Shimoni, L.; Chang, N. Angew. Chem., Int. Ed. Engl. 1995, 34, 1555. (g) Grell, J.; Bernstein, J.; Tinhofer, B. Graph-set analysis of hydrogen-bond patterns: some mathematical concepts. Acta Crystallogr. 1999, B55, 1030-1043. Redenti, E.; Peveri, T.; Zanol, M.; Ventura, P.; Gnappi, G.; Montenero, A. A study of the differentiation between amorphous piroxicam:β-cyclodextrin complex and a mixture of the two amorphous components. Int. J. Pharm. 1996, 129, 289-294. The United States Pharmacopoeia XXVII, USP27/NF22, 27th revision; The United States Pharmacopoeial Convention, Inc.: Rockville, MD, 2004; p 1499. Crowley, K. J.; Zografi, G. Cryogenic grinding of indomethacin polymorphs and solvates: Assessment of amorphous phase formation and amorphous phase physical stability. J. Pharm. Sci. 2002, 91 (2), 492-507. Lutteroti, L. Material Analysis Using X-ray Diffraction (MAUD), version 1.998; Department of Materials Engineering, University of Trento: Italy. Billinge, S. J. L.; Thorpe, M. F. Local Structure from Diffraction; Plenum Press: New York, 1998.

CG049876Y