Article pubs.acs.org/crystal
Furosemide Solvates: Can They Serve As Precursors to Different Polymorphs of Furosemide? Vasily S. Minkov,*,†,‡ Alina A. Beloborodova,† Valeri A. Drebushchak,†,§ and Elena V. Boldyreva†,‡ †
REC-008, Novosibirsk State University, ul. Pirogova, 2, Novosibirsk 630090, Russia Institute of Solid State Chemistry and Mechanochemistry Siberian Branch of the Russian Academy of Sciences (SB RAS), ul. Kutateladze, 18, Novosibirsk 630128, Russia § Institute of Geology and Mineralogy Siberian Branch of the Russian Academy of Sciences (SB RAS), pr. Koptyuga, 3, Novosibirsk 630090, Russia ‡
S Supporting Information *
ABSTRACT: In the present work, four solvates of furosemide with tetrahydrofuran S(THF), 1,4-dioxane S(DIOX), N,N-dimethylformamide S(DMF), and dimethylsulfoxide S(DMSO) were crystallized, the crystal structures were solved for S(THF), S(DIOX), and S(DMF). The existence of S(THF) and S(DMSO) was not reported before; for the previously known S(DIOX) and S(DMF), the crystal structures remained unsolved. The detailed structural analysis of furosemide containing crystal structures showed that the molecule of furosemide has a high conformational lability because of the rotations of the sulfamoyl and furanylmethylamino fragments. Some of the furosemide conformations were shown to be stabilized by the intramolecular N− H···Cl hydrogen bond. Desolvation of the four solvates was studied by TG and X-ray diffraction and was shown to give different products depending on the precursor and particle size: the desolvation of powder of S(THF) and the large crystals of S(THF), S(DIOX), and S(DMF) gives Form-III of furosemide, whereas powders of S(DIOX), S(DMF), and S(DMSO) give Form-I.
1. INTRODUCTION
solvates is that only very few solvents are harmless and are allowed by pharmacopea. More often than being applied as such, solvates are used as precursors to produce solvent-free forms. The most important technical and industrial usage of solvates12 is related to such issues as purification,13 preparation of a specific polymorph with high chemical and physical purity,14−16 or particle size control because the desolvation of stoichiometric solvates often leads to products with a narrow and homogeneous particle size distribution.17 Varying conditions of desolvation, one can produce samples of the same polymorph differing in particle size, morphology, type and concentration of defects. This method is a variant of a more general method of using various precursors to modify the properties of solid materials obtained in the course of solid-state reactions.18−21 In some cases, decomposition of different precursors can give different polymorphs. This is widely used for a long time to obtain different polymorphs of inorganic compounds, for example of oxides applied as catalysts.22,23 In the last years the same technique was used to produce different polymorphs of drugs, some of which cannot be easily obtained by direct crystallization from solutions.15,24−27 Recently it was found
The importance of polymorphism of molecular crystals is hard to overestimate, especially when dealing with compounds used as materials or drugs. Different polymorphs of a drug substance may have different properties related to their manufacturing, therapeutic usage, or storage (density, hygroscopicity, melting points, thermal stability, solubility, rate of dissolution, surface free energy, toxicity, bioavailability, tabletting, etc.).1,2 Different polymorphs, solvates, and cocrystals can be patented, and this opens the way for a competition with brand drugs. Since the energies of different polymorphs are sometimes very close, producing desirable crystalline forms is quite a challenge and can also be complicated by the phenomena of concomitant polymorphism (when several polymorphs crystallize simultaneously from the same batch), or erratic and poorly reproducible (when crystallization gives different polymorphs even at seemingly identical experimental conditions).3 As examples, one can mention glycine,4 chlorpropamide,5,6 and sulfathiazole.7 Solvates are also considered as potentially promising forms that can serve as an alternative to solvent-free drugs. In many cases, it is easier to obtain solvates as a pure phase, since concomitant polymorphism of solvates is rather rare.8 It is equally rare that a solvate and a solvent-free form crystallize from the same batch.9−11 A limitation of practical usage of drug © 2013 American Chemical Society
Received: August 19, 2013 Revised: December 8, 2013 Published: December 20, 2013 513
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
Table 1. Crystal Data, Data Collection, and Structure Refinement Parameters for Furosemide Solvates chemical formula molecular weight crystal system space group T (K) a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å3) Z/Z′ Dcalcd (g cm−3) μ (mm−1) crystal size (mm) no. of measured independent reflns observed [I > 2σ(I)] reflns Rint θ values (deg) range of h, k, l
R [F2 > 2σ(F2)] wR (F2) no. of parameters S Δρmin, Δρmax (e Å−3)
furosemide tetrahydrofuran solvate
furosemide 1,4-dioxane solvate
furosemide N,N-dimethylformamide solvate
C12H11ClN2O5S·C4H8O 402.85 monoclinic P21/n 150 10.9529(5) 8.3816(3) 20.0071(9) 90 95.024(4) 90 1829.65(13) 4/1 1.462 0.359 0.17 × 0.25 × 0.38 20116 4916 3849 0.0374 θmin= 2.04, θmax= 29.15 h = −14−14 k = −9−11 l = −27−27 0.0349 0.0825 278 1.055 −0.393, 0.293
C12H11ClN2O5S·C4H8O2 418.85 triclinic P1̅ 150 7.9759(5) 10.6923(7) 10.9369(7) 86.247(5) 79.053(5) 80.548(5) 902.74(10) 2/1 1.541 0.371 0.06 × 0.24 × 0.38 11250 4839 3733 0.0358 θmin= 1.90, θmax= 29.17 h = −10−9 k = −14−14 l = −14−15 0.0345 0.0818 301 1.016 −0.544, 0.334
C12H11ClN2O5S·C3H7NO 403.84 monoclinic P21/c 150 12.1430(5) 12.4050(4) 24.3562(11) 90 99.520(3) 90 3618.3(3) 8/2 1.483 0.364 0.16 × 0.18 × 0.52 28595 7988 5577 0.0506 θmin= 1.70, θmax= 27.10 h = −15−14 k = −15−15 l = −31−31 0.0429 0.0819 578 0.969 −0.397, 0.320
forms of furosemide for which crystal structures remained unknown, namely a high temperature phase, and another phase obtained by rapid crystallization from acetone. Several solvates of furosemide with N,N-dimethylformamide and 1,4-dioxane were reported in 1990,32 but their crystal structures were not solved.
that 2,4,6-triethyl-1,3,5-tris(phenoxymethyl)-benzene can form 10 different solvates which lead to the formation of 3 different polymorphs on desolvation.11 At the same time, three solvates having common dimeric or catemeric hydrogen bonded motifs and one hydrate with completely different molecular interactions of aripiprazole invariantly give the same polymorph of pure drug;28 three different isostructural solvates of tetroxoprim also give the same polymorph on desolvation whereas desolvation of another one results in two different polymorphs.29 Thus, it is not yet possible at the present stage to predict a priori if different precursor solvates for a given drug will give different or identical desolvated products, and additional studies are needed to make any generalization and to get knowledge relevant for the selected compound. The aim of the present study was to crystallize various solvates of furosemide, to check whether these solvates can be used as precursors for producing different polymorphs of pure furosemide on their subsequent decomposition upon heating, and to search any correlation between the crystal structures of the solvates and on the furosemide polymorphs produced by desolvation. Furosemide is a loop diuretic which is widely used in the treatment of congestive heart failure and edema. Though the first crystal structure of furosemide has been published in 1978 by Lamotte,30 crystal structures of the other two polymorphs were reported in 2010 only by Babu.31 Attempts to characterize different crystalline modifications of furosemide including some solvates by FTIR and terahertz spectroscopy, X-ray powder diffraction and thermal analysis were made already in 1990 by Matsuda32 and in 2009 by Ge.33 Nevertheless, there are two
2. EXPERIMENTAL SECTION 2.1. Materials. Furosemide solvates with low boiling point solvents, namely S(THF) and S(DIOX), were crystallized from the saturated solutions in the corresponding solvent by slow evaporation in a refrigerator at about 2−3 °C. Colorless plate-shape crystals of S(THF) and block-shape crystals of S(DIOX) were kept under the corresponding mother liquors in closed vessels in a refrigerator, in order to prevent their decomposition. Since the high boiling point solvents, N,N-dimethylformamide and dimethylsulfoxide giving the furosemide solvates S(DMF) and S(DMSO), respectively, evaporate poorly at ambient conditions, water was used as an antisolvent to precipitate crystals from the corresponding saturated solutions. Crystals of S(DMF) had the shape of needles, whereas the solvate S(DMSO) crystallized as white powder not suitable for single-crystal X-ray diffraction. Crystals of all solvates lose their solvent molecules easily on storage at ambient conditions: solvates with low boiling point solvents decay over several hours, whereas those with high boiling point solvents decay over several days. In all the solvates the stoichiometric ratio between furosemide and solvent molecules was estimated as 1:1 by means of the single crystal X-ray diffraction, where it was possible (solvates S(THF), S(DIOX), S(DMF)), and by thermogravimetric analysis for all the solvates but S(DMSO). The composition of S(DMSO) was estimated by precise weighting the masses before and after desolvation in a vacuum chamber: three experiments gave the value of ∼1.3 molecules of dimethylsulfoxide per one molecule of furosemide (however, the true ratio should be 514
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
crystal data (CCDC No. 761769−76177231) deposited at the Cambridge Crystallographic Database.41 2.3. Thermogravimetric Analysis. was performed using Netzsch TG-209 with heating rate of 3 K/min for both types of samples: powder and large crystals (except S(DMSO)). Solvate S(THF) was prepared as large single crystal blocks and measured three times (52.55, 51.63, and 37.21 mg) and as powder with a mass of 25.07 mg. The mass of a single crystal and powder of solvate S(DIOX) was 5.00 and 31.95 mg. In the cases of furosemide solvates with high boiling point solvents, bulk sample with several crystals of S(DMF) with total mass of 10.93 mg, as well as powder with mass of 14.63 mg were measured. Experiments with desolvation of solvate S(DMSO) were carried out using precipitated fine powder with total mass of 74.50 mg. The sample holder of TG-209 thermobalance possesses in-built thermocouple that allows one to detect the temperature of the sample in situ (c-DTA option). It makes possible to measure mass loss as a function of sample temperature, not of furnace temperature as in typical TG design. This peculiarity of TG-209 thermobalance turned out to be critical in the discussion of the results obtained for the large crystal of S(THF).
reduced, to give 1:1 stoichiometry, because some mother liquor still remained in the fine powder sample before the desolvation process and could not be eliminated completely without vacuum; we cannot exclude with absolute certainty that the sample might represent also a slightly wetter 0.5:1 stoichiometry). For thermogravimetric analysis, we used (i) large crystals of well-crystallized furosemide solvates with all three dimensions of crystals larger than 1 mm and (ii) fine powders. Powder of S(DMSO) was precipitated during crystallization, powder samples of all other solvates were obtained by gentle grinding the large crystals in a mortar with considerable amount of mother liquor added, to protect the samples from desolvation (powder X-ray diffraction analysis did not reveal even traces of phases of pure furosemide in the powders of solvates, including those prepared by grinding of large solvate crystals, see Supporting Information and Figure 5). Grinding without adding mother liquor lead to partial desolvation, especially in case of S(THF) and S(DIOX). Since the decomposition of furosemide solvates with high boiling point solvents was time-consuming and hampered at ambient pressure, a high-temperature vacuum chamber was used for 36 h at 20 mbar and 60 °C for S(DMF), and at 80 °C for S(DMSO) to obtain the products of desolvation. After decomposition all samples were neatly moved onto a cuvette with no mechanical treatment (grinding or pressing) for further phase analysis using a powder X-ray diffractometer (desolvated large crystals preserved their shape and were analyzed as whole massive samples in reflection mode). 2.2. X-ray Diffraction. Single-crystal X-ray diffraction experiments were carried out by rotation method using a Stoe IPDS-II diffractometer equipped with a molybdenum X-ray tube (λ = 0.71073 Å), an image plate detector, a flat graphite monochromator and an Oxford Cryostreams cooling device (stability of gas flow temperature is ±0.1 K). All single-crystal X-ray diffraction experiments were performed at 150 K to reduce the decomposition of samples and increase the quality of data set. Preliminary experiments performed at ambient temperature showed the cell parameters of three solvates being very close to that at low temperature, thus phase transitions related to discontinuous changes in the cell parameters or crystal system do not occur on cooling down to 150 K. All crystals were also covered by low viscosity CryoOil (MiTeGen) to protect them additionally from environment during data collection. The data collection, indexing and integration of the reflections were performed using Stoe X-Area software package, data reduction was performed using Stoe X-Red software.34 Crystal structures of furosemide solvates were refined using SHELXL35,36 implemented in the X-Step32 shell.37 In case of S(THF) the solvent was found to be disordered over two positions with site occupancy of 0.709(7) and 0.291(7). All H atoms were found in a difference Fourier map and their positions were refined freely, with Uiso(H) = 1.5Ueq(C) for terminal methyl H atoms and 1.2 Ueq(parent atom) otherwise. In the difference maps for the S(THF) with cis-orientation of an amino group in a sulfamoyl fragment there were no residual peaks with intensities higher than 0.14 e·Å−3 near the nitrogen atom what excludes the possibility of the existence of another orientation of this group. The positions of Hatoms in the disordered tetrahydrofuran ring were refined using a riding model, with default distances of secondary C−H = 0.99 Å. The parameters characterizing data collection and refinement, as well as crystal data are summarized in Table 1. Mercury,38 ORTEP-3,39 and PLATON40 were used for visualization, analysis, and inspection of the crystal structures. The structural data for furosemide solvates are deposited as CIFs at the Cambridge Crystallographic Database (CCDC No. 955882−955884), and can be downloaded freely from the following site http://www.ccdc.cam.ac.uk. Phase analysis of desolvated products were performed using Bruker D8 Discover with GADDS powder X-ray diffractometer equipped with a copper X-ray tube (λ = 1.5418 Å) and the HI-STAR area detector, scanning 2θ range of 4−45°. All samples were found to be fine crystalline powders without a substantial preferred orientation of particles in massive blocks obtained from large solvate crystals. Powder patterns of all desolvation products were compared with those of the three known polymorphs of furosemide calculated from the single-
3. RESULTS AND DISCUSSION 3.1. Crystallization of Furosemide Solvates and Their Crystal Structures in Relation to Molecular Conformations. Crystallization of furosemide by slow evaporation (for low boiling point solvents) or by using water as an antisolvent (for high boiling point solvents) from saturated solutions in a number of aprotic solvents such as tetrahydrofuran, N,Ndimethylformamide, dimethylsulfoxide, and 1,4-dioxane resulted in the formation of the corresponding furosemide solvates. The structures of S(THF), S(DIOX), and S(DMF) were solved from single crystal X-ray diffraction data. S(DMSO) could not be obtained as single crystals suitable for single crystal X-ray diffraction analysis, and attempts of crystal structure solution from X-ray powder diffraction data were not successful yet. Crystallization from other aprotic solvents, namely, ethyl acetate, acetonitrile, and acetone did not give solvates, but furosemide Form-I. For a comparison, according to the published data,32 the crystallization of furosemide from most organic protic solvents results in the formation of different polymorphs of the drug: methanol, ethanol and isopropanol lead to crystallization of Form-I, whereas Form-III crystallizes from n-butanol (here and below the labeling of the furosemide polymorphs corresponds to that in ref 31). Solvates S(THF) and S(DIOX) crystallize in the monoclinic P21/n and triclinic P1̅ space groups, respectively, with one molecule of solvent and furosemide in the asymmetric unit, whereas S(DMF) crystallizes in the monoclinic P21/c space group with two molecules of furosemide and N,N-dimethylformamide in the asymmetric unit (Table 1). Thus, all three solved crystal structures are centrosymmetric. The asymmetric units of these solvates with corresponding labeling scheme are shown in Figure 1. The conformations of furosemide molecules, as well as the crystal packing, are different in all the solvates. The conformational lability is typical for furosemide molecules in all the known furosemide-containing crystal structures (11 total, including 3 polymorphs, 4 cocrystals, 1 salt41 and 3 solvates obtained in this work) because of the presence of the furanylmethylamino (−NH−CH2−C4H3O) and sulfamoyl (−SO2−NH2) fragments (Figures 1 and 2). At the same time, some molecular fragments seem to be very robust. For example, in all the structures the benzene ring, secondary amino and carboxyl groups, as well as chlorine and sulfur atoms 515
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
Figure 2. (a) Overlaid conformations of furosemide molecule in its solvates S(THF), S(DIOX), and S(DMF) colored by blue, red, and green (dark green relates to A molecule, the first molecule in asymmetric unit, light green to B). Hydrogen atoms of the benzene ring and of the methylene −CH2 group are omitted for clarity. (b) Schematic illustration of furosemide conformer variety has been revealed in a crystalline state so far. The τi labels correspond to the three torsion angles characterizing the conformation of a molecule, arrows show valence bonds along which rotation occurred. The G and T letters are abbreviations for gauche and trans conformations, upper left numbers point to the τi torsion angle describing the conformation. Atom colors are the same as in Figure 1
the values of 2.679(36) Å and 137(5)° averaged over all the known furosemide-containing structures. The conformation of the sulfamoyl and furanylmethylamino fragments can be described by three torsion angles: C4−C5− S1−N2 (τ1), C2−N1−C8−C12 (τ2), and N1−C8−C12−O3 (τ3) (Figures 1 and 2). The latter two torsion angles characterize the rotation of the furanylmethylamino fragment along the N1−C8 bond and of the furanyl ring along C8−C12 bond. Another important structural feature is the orientation of NH2 in the sulfamoyl group with relation to SO2. It can be roughly described in terms of syn and anti. The results of the conformational analysis of the furosemide molecule in a crystal are summarized in Table 2. In the present solvates the sulfamoyl group is in gauche+ conformation, whereas the furanylmethylamino fragment is prone to adopt different orientations from gauche− in S(DIOX) to gauche+ in S(THF) including an intermediate trans conformation in S(DMF) (see overlaid furosemide molecules in Figure 2). The gauche+ conformation of the sulfamoyl group is characteristic for all the previously reported furosemide-containing structures with the only exception of the furosemide Form-I where both molecules adopt trans conformation (two molecules in the asymmetric unit have TG−G− and TG−G+ conformations). The DFT computations for different conformers of isolated furosemide N-methyl derivative with furanyl substituted for H-atom showed that the conformer with trans conformation of the sulfamoyl group is the least favorable one, and its energy is about 4 kcal·mol−1 higher than that of the gauche conformer.31 However the crystal structure of Form-I as a whole is thermodynamically more stable as compared to Form-II and Form-III. It is clearly seen from Table 2 that the gauche+
Figure 1. Asymmetric unit of S(THF) (a), S(DIOX) (b), and S(DMF) (c) showing the atom-numbering scheme. Displacement ellipsoids are drawn at the 50% probability level.
lie practically in one plane. The value of the C6−C1−C7−O2 torsion angle, characterizing the twisting of the carboxyl group relative to the benzene ring around the C1−C7 bond, is not less than 169.3(2)° (the value for furosemide and pentoxifylline cocrystal acetone solvate).42 The planar conformation of the carboxyl group in the furosemide molecule can be supposed to be stabilized by the intramolecular N−H···O H-bond between the secondary amino and carboxyl groups. The values of the N···O distance and the N−H···O angle of this H-bond in S(THF), S(DIOX), and S(DMF) are in a good agreement with 516
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
Table 2. Several Torsion Angles τ1 (C4−C5−S1−N2), τ2 (C2−N1−C8−C12), τ3 (N1−C8−C12−O3), and Orientation of NH2 Group in the Sulfamoyl Group Characterizing Conformation of the Furosemide Molecule in Different Crystal Structures crystal structure S(THF) S(DIOX) S(DMF) furosemide Form-I
31
furosemide Form-II31a furosemide furosemide furosemide furosemide furosemide
31
Form-III pentoxifylline cocrystal42 pentoxifylline water cocrystal42 pentoxifylline acetone cocrystal42 caffeine cocrystal43a
furosemide cytosine salt a
43
τ1 (deg)
τ2 (deg)
τ3 (deg)
61.40(14) 64.06(15) 65.6(2) 62.6(2) 163.2(4) 166.0(4) 78.6(4)
79.43(18) −86.43(19) −179.7(2) −169.0(2) −61.4(6) −84.0(6) 162.2(4)
55.7(3) 66.7(3) 62.8(4) 69.0(3) 66.4(4)
91.3(4) −67.5(4) −75.2(5) 89.5(4) −71.8(5)
60.3(2)
171.8(2)
51.98(17) 59.80(18) −55.2(3) −59.9(3) −57.6(5) 68.2(6) 80.0(6) −97.0(10) −60.0(4) −65.5(4) −69.3(5) −93.4(4) 119.9(5) −59.0(10) 156.6(2)
orientation of NH2 cis trans trans trans trans trans trans gauche trans trans gauche trans
conformation G+G+G+syn G+G−G+anti G+TG−anti G+TG−anti TG−G−anti TG−G+anti G+TG+anti G+TG−anti G+G+G−anti G+G−G−syn G+G−G−anti G+G+G−anti G+G−Tsyn G+G−G−syn G+TTanti
A crystal structure has only one molecule in an asymmetric unit but the furanyl ring is disordered over two orientations.
Table 3. Geometric Parameters of Hydrogen Bonds in Crystal Structures of S(THF), S(DIOX), and S(DMF) hydrogen bond
D−H (Å)
H···A (Å)
N1−H2···O2 N2−H4···Cl1 O1−H1···O6 N2−H3···O2 N2−H4···O5
0.863(19) 0.87(2) 0.90(2) 0.84(2) 0.87(2)
1.945(19) 3.097(18) 1.67(2) 2.09(2) 2.05(2)
N1−H2···O2 N2−H4···Cl1 O1−H1···O6 N2−H3···O2 N2−H4···O5
0.83(2) 0.85(2) 0.83(2) 0.84(2) 0.85(2)
1.97(2) 3.02(2) 1.81(2) 2.08(2) 2.35(2)
N1a−H2a···O2a N1b−H2b···O2b N2a−H4a···Cl1a N2b−H4b···Cl1b O1a−H1a···O6b O1b−H1b···O6a N2a−H3a···O6a N2b−H3b···O6b N2a−H3a···O6b N2b−H3b···O6a
0.82(3) 0.86(3) 0.79(3) 0.80(3) 0.93(3) 0.86(3) 0.87(3) 0.87(3) 0.87(3) 0.87(3)
2.01(3) 1.97(3) 3.01(3) 2.96(3) 1.65(3) 1.73(3) 2.66(3) 2.68(3) 2.13(3) 2.09(3)
D···A (Å) S(THF) 2.6697(16) 3.2629(14) 2.5557(15) 2.9135(17) 2.9153(17) S(DIOX) 2.6661(18) 3.4571(15) 2.6318(16) 2.9053(17) 3.181(2) S(DMF) 2.656(2) 2.669(2) 3.430(2) 3.374(2) 2.582(2) 2.586(2) 3.042(3) 3.057(3) 2.973(3) 2.934(3)
conformation of sulfamoyl group and anti orientation of amino group are more frequent (such combination is also preferable for an isolated molecule of furosemide derivative in the gas phase31); this conformation can be supposed to be stabilized by the second intramolecular N−H···Cl H-bond. We believe that weak N−H···Cl H-bonds are formed in S(DIOX) and S(DMF): the H···Cl distance is about 3.00 Å (what is only 0.1 Å longer than the sum of van der Waals radii) and the N− H···Cl angle is near 115°. In S(THF) the amino group is in the eclipsed cis conformation, and though the N···Cl distance is short (3.2629(14) Å), the values of the H···Cl distance and N− H···Cl angle are equal to 3.097(18) Å and 93.3(13)°, respectively, what might be considered as the presence of a very weak H-bond, if it exists at all. In other crystal structures with 1G+ and anti, particularly trans conformation of the amino
∠D−H···O (deg)
symmetry code of an acceptor
140.7(16) 93.3(13) 170(2) 165.8(18) 172.7(18)
intramolecular intramolecular x, y, z 1/2 − x, −1/2 + y, 1/2 − z 1 − x, 1 − y, 1 − z
140.7(18) 113.8(16) 174(2) 165(2) 164.8(18)
intramolecular intramolecular x, y, z 2 − x, −y, −z 1 − x, 1 − y, −z
135(2) 138(2) 116(2) 115(2) 173(3) 174(3) 108(2) 107(2) 164(2) 164(2)
intramolecular intramolecular intramolecular intramolecular x, y, z 1 + x, y, z x, y, z x, y, z 1 − x, −1/2 + y, 1/2 − z 1 − x, 1/2 + y, 1/2 − z
group, an intramolecular N−H···Cl H-bond should be taken into consideration (with the H···Cl distances less than the sum of van der Waals radii and the values of N−H···Cl angles higher than 120°), with the exception of the structure of furosemide Form-II, where the H···Cl distance is somewhat longer (3.14(5) Å). In the crystal structures of the furosemide cocrystals with pentoxifylline and caffeine the amino group adopts a staggered gauche conformation, and there is no possibility to form this H-bond, though the N···Cl distances are quite short: 3.364(3) and 3.322(3) Å, respectively. Similarly, there is no way to form this intramolecular H-bond in the crystal structure of furosemide Form-I, since the sulfamoyl group is in trans conformation. The furanylmethylamino fragment has two axes for rotation: along the N1−C8 and the C8−C12 bonds. Flexible rotation of 517
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
And finally, the presence of a relatively short N−H···Cl H-bond (N···Cl of 3.257(2) Å) in the salt of furosemide and cytosine leads to a very significant elongation of the N−H···O H-bond between furosemide molecules, up to 3.349(9) Å.43 The second H-atom of the NH2 group participates in the formation of N−H···O H-bonds with the carboxyl group of the neighboring furosemide molecule in S(THF) and S(DIOX). In S(THF) N2−H3···O2 H-bonds account for the formation of infinite chains of furosemide molecules along the crystallographic axis b, whereas in S(DIOX) the bond of this type forms finite two-member chains. The carboxyl group participates in formation of O1−H1···O6 H-bonds with the O-atom of a solvent molecule. These H-bonds are quite short for S(THF) and S(DMF) with the values of the O1···O6 distance equal to 2.5557(15) and 2.582(2) Å (2.586(2) Å for the second molecule in S(DMF)), respectively, and is longer in solvate S(DIOX), 2.6318(16) Å. Notably, the second O-atom of 1,4dioxane molecule is not involved in any kind of H-bonding in S(DIOX). In the case of S(DMF) furosemide molecules are not linked to each other by H-bonds directly, H-atoms of every donor are involved into the formation of intramolecular N2i−H4i···Cl1i and N1i−H2i···O2i H-bonds or intermolecular H-bonds only with the carbonyl group of N,N-dimethylformamide molecule what makes this crystal structure unique as compared with others. Different H-bonding patterns account for different molecular packing in the crystal structures of three solvates (Figure 3). In S(THF) and S(DIOX) the solvent molecules (tetrahydrofuran and 1,4-dioxane, respectively) are arranged in infinite channels stretching along the crystallographic axis b and a, correspondingly. In contrast, the crystal structure of S(DMF) is layered, and the solvent molecules are located between the layers of furosemide molecules. 3.2. Desolvation of Furosemide Solvates and the Control of Polymorphism of Furosemide. The solvates containing the molecules of the low-boiling point solvents in their crystal structures, S(THF) and S(DIOX), lose tetrahydrofuran and 1,4-dioxane quite easily (Figure 4). Nevertheless, the curve of the mass decrease of the crystal of solvate S(THF) with the size of ∼2 × 2 × 3 mm3 on heating has a plateau giving evidence that the desolvation proceeds in two stages (the same behavior has been detected every time for large crystals with characteristic size of more than 1 mm). Powder sample of S(THF) desolvated in one stage, thus the desolvation of S(THF) shows a pronounced kinetic size-effect. The precise inspection of the large crystal of S(THF) after partial desolvation (the sample was taken out of the TG-chamber immediately after the mass loss curve felt down to the first plateau when the temperature of a sample was about 345 K) showed that the crystal has a polycrystalline crust of the furosemide Form-III. Under this crust the crystal of the solvate remained transparent without any substantial changes. Thus, a desolvated layer formed at the surface hinders further release of solvent molecules, unless the temperature increases more. Such phenomena are well-known for thermal decomposition of solids, especially if a sample is a relatively large crystal.46−49 As a recent example to which one can refer,50 on thermal desolvation of dimethylsulfoxide solvate of cortisone acetate the reaction stopped having reached some degree of conversion, and was resumed only after increasing temperature when the internal pressure of desolvated vapor reached a critical value accompanied by explosive disruption of the
this fragment accounts for the conformational diversity of the furosemide-containing crystal structures. In Figure 2a, different conformations of this fragment in the crystal structures of S(THF), S(DIOX), and S(DMF) are overlaid; all the possible furosemide conformers in different crystal structures known up to now are schematically presented in Figure 2b. In most crystal structures the furanyl ring is not involved in any H-bonds or π···π interactions, what increases the flexibility of this fragment. H-bonds in which the furanyl ring acts as an acceptor are present only in the cocrystal of furosemide and pentoxifylline hydrate (the values of the N···O distance and the N−H···O angle in N−H···O H-bond are equal to 3.071(6) Å and 155°, respectively42) and in the salt of furosemide with cytosine (the N−H···π bond with the corresponding values of 3.394 Å and 160°43). In both cases the sulfamoyl group acts as a donor for the H-bonding. Thus, in 10 out of 11 crystal structures containing 13 conformers in the asymmetric unit, including those with the disordered furanyl ring where the sulfamoyl group is 1G+ with τ1 torsion angle of ∼+60°, the furanylmethylamino fragment can adopt all three possible conformations (defined by τ2) with roughly the same occurrence: gauche + , gauche − , and intermediate trans, designated as 2G+, 2G−, and 2T in Figure 2b and Table 2. The orientation of the furanyl ring itself (characterized by τ3) can be different (3G+, 3G−, 3T) if τ2 takes 2 − G and 2T conformations (thus possible conformations are G+G−G+, G+G−G−, G+G−T, G+TG+, G+TG−, G+TT), but only 3 + G and 3G− when τ2 correspond to 2G+ (i.e., G+G+G+ and G+G+G− conformations). Interestingly, all other conditions remaining the same, the τ3 is prone to adopt the values characteristic for gauche− conformation: 3G− is found for 9 out of 15 conformers present in the crystals (Table 2). Molecular conformations are related not only to intramolecular, but also to intermolecular H-bonding (Table 3). In S(THF) and S(DIOX) sulfamoyl group forms a ring R22(8) motif by means of two N2−H4···O5 H-bonds (see refs 44 and 45 for graph set notation) which are quite similar to those in the polymorphs I and III of furosemide31 and in the salt of furosemide with cytosine.43 Analysis of furosemide-containing crystal structures reveal that the strengthening of the intramolecular N2−H4···Cl1 H-bond leads to the lengthening of intermolecular N2−H4···O5 H-bond due to the weakening of donor properties of the NH2 group already bound to the Cl acceptor. In S(THF) the orientation of NH2 does not allow to form a proper N−H···Cl H-bond. At the same time, the ringforming H-bond is the strongest among all structures containing the R22(8) motif with a short N···O distance of 2.9153(17) Å and the N−H···O angle of 172.7(18)°. In the two polymorphs of furosemide, Form-I, without the N−H···Cl Hbond, and Form-III, where a Cl-atom forms H-bond with a Hatom of the NH2 groups not involved in forming the ring motif, the N−H···O H-bonds also relatively short: with the values of the N···O distances and N−H−O angles equal to 2.978(5) Å and 171(5)° (first molecule in Form-I), 2.999(5) Å and 173(6)° (second molecule in Form-I), 3.028(4) Å and 162(3)° (Form-III).31 In S(DIOX) the N2−H4···Cl1 H-bond is relatively weak (N···Cl of 3.4571(15) Å and N−H···Cl of 113.8(16)°), but already starts to effect the N2−H4···O5 Hbond with the same proton, namely, it results in the elongation of the N···O distance up to 3.181(2) Å as compared to S(THF). In S(DMF) H4 atom participates only in the N−H··· Cl H-bond what leads to shortening the N···Cl distance further to 3.430(2) and 3.374(2) Å for A and B molecules, respectively. 518
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
Figure 4. Thermogravimetric curve of decomposition of different furosemide solvates: (a) S(THF), (b) S(DIOX), and (c) S(DMF). Black lines correspond to large crystals, whereas red lines correspond to powders. (d) Mass loss of S(DMSO) heated 6 times. The lines are marked with numbers in order of the runs. Each line starts from 100% and the final mass loss of the sample is the sum of all 6 runs (∼80%). The results were corrected for blank run. Figure 3. Fragments of the crystal packing in S(THF) (a), S(DIOX) (b), and S(DMF) (c). Dashed blue lines represent hydrogen bonds.
with the enthalpy of decomposition large enough to cool down the sample together with the crucible. This endothermic event breaks the line of the mass loss and looks like the 90° angle drop-off. The total change of mass for both samples of S(THF) is 17.4%, what corresponds to the complete desolvation with the furosemide: tetrahydrofuran ratio of 1:1. Large crystals and powder of S(DIOX) desolvated both in one step; the loss of about 21.3% of mass corresponds to the loss of one molecule of 1,4-dioxane per each molecule of furosemide, what is in a good agreement with the structural data (Figure 4). The desolvation of the furosemide solvates containing molecules of high boiling point solvents, that is, N,N-
particles (also called as pop corn effect) and a sudden endothermic event. For a large crystal of S(THF), the mass first decreases at about 6.5%, and then at 10.9%. After formation of a product crust and reaching the necessary temperature the second stage of desolvation occurs in a very narrow temperature range near 355 K. At the same time Figure 4a shows clearly that the temperature of the large crystalline sample even decreases during this process. This result was observed because of the TG-209 design (see section 2.3) and could be interpreted as resulting from a very fast decomposition 519
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
dimethylformamide and dimethylsulfoxide, occurred rather reluctantly. The mass of a bulk crystal sample of S(DMF) decreased continuously from room temperature to approximately 355 K, and from 375 to 430 K, with two overlapping steps over the interval of 355−375 K with the mass loss 3.2%. These steps of the mass loss can account for the formation of the product core at the surface on decomposition, similarly to what has been observed for large crystals of S(THF). Powder sample of S(DMF) desolvated completely in a quite narrow range of 350−370 K with total mass loss of about 17.5% what corresponds to furosemide: N,N-dimethylformamide ratio of 1: 1. At the beginning on heating from 293 to 316 K a gradual mass decrease of ∼1.14% corresponds to drying the powder sample. When attempting to separate drying of the powder sample of S(DMSO) from its desolvation, we were forced to interrupt heating many times by cooling the sample down to room temperature because a very fast mass loss started abruptly. Heating was started 6 times (up to 345, 345, 395, 403, 423, and finally up to 463 K, Figure 4d). Only smooth lines of gradual mass loss were detected without pronounced steps. The final mass loss of the sample in all the 6 runs was ∼80%, obviously indicating that the sample was not only desolvated but also decomposed. Thus, thermal desolvation of bulk crystals of S(DMF) and powder of S(DMSO) was accompanied by the decomposition: by the end of the TG experiments the samples looked as burnt, though the melting temperature of pure furosemide with its subsequent decomposition is considerably higher, 490 K.32 Since the crystals of S(DMF) and powder of S(DMSO) desolvate poorly at atmospheric pressure, they were desolvated under vacuum: reduced pressure made it possible to decrease the temperature and avoid the decomposition. Desolvation of powder samples of S(DIOX) and S(DMF) is quite similar to that reported earlier.32 The desolvation products of different samples of the furosemide solvates were further analyzed by X-ray powder diffraction. Selected X-ray powder patterns are shown in Figure 5. The large crystals of S(THF), S(DIOX), and S(DMF) preserved their shape after the desolvation, and these pseudomorphs were almost as hard to break as the initial single crystals. These samples after desolvation were not subject to any grinding and were X-rayed as such (the samples of desolvated S(THF) and S(DIOX) were taken from the sample chamber immediately after the thermogravimetric analysis, the samples of desolvated S(DMF) were obtained using a vacuum chamber equipped with a heating device because of its decomposition during TG, see above). However, the X-ray powder patterns looked as if the samples contained finedispersed powders of furosemide without any large crystallites and with no preferred orientation (Figure 6). Formation of such polycrystalline pseudomorphs was previously reported on decomposition of inorganic and coordination compounds,22 as well as in the course of polymorphic transformations of drugs, for example, for the monoclinic polymorph of paracetamol after the temperature-induced phase transition in a single crystal of the orthorhombic polymorph,51 and for different polymorphs of sulfathiazole.52 The scheme of transformation of different furosemide solvates samples on thermal desolvation is shown in Figure 7. It is interesting that different solvates of furosemide as well as different samples of the same solvate can give different polymorphs of pure furosemide on desolvation. The complete loss of solvent molecules from large crystal samples of S(THF),
Figure 5. Powder patterns of desolvated samples of furosemide solvates: (a) purple 1 corresponds to the desolvation product of powder sample of S(THF), blue 2, black 3, and red 4 correspond to the desolvation products of the large crystals of S(THF), S(DIOX), and S(DMF) respectively. All these samples contain exclusively furosemide Form-III (theoretical powder patterns colored by magenta 5 and calculated from single crystal data31). (b) Blue 7, black 8, and red 9 correspond to the desolvation products of powder samples of S(DMSO), S(DMF), and S(DIOX), respectively. All of them correspond to furosemide Form-I (theoretical powder patterns colored by magenta 10 and calculated from single crystal data30). Purple 6 is related to powder pattern of S(DMSO) before desolvation. The presence of a phase of furosemide Form-III in desolvated powder sample of S(DIOX) is marked by (*).
S(DIOX), and S(DMF) resulted in the formation of furosemide Form-III, whereas powder samples of S(DIOX), S(DMF), and S(DMSO) as well, give furosemide Form-I (though after desolvation of powder S(DIOX) there is some phase of furosemide Form-III). Previously reported32 formation of pure Form-I on desolvation of powders of solvates S(DIOX) and S(DMF) agrees with the present results. In our experiments, furosemide forms I and III obtained by desolvation of different samples of furosemide solvates were preserved during storage at least for 6 months. This differs somewhat from the fact reported in,32 that the storage of Form-III results in its transformation into Form-I. We cannot make a definite conclusion on the reasons of this discrepancy since the storage conditions in32 were not defined clearly (only “high humidity” was mentioned). Temperature did not induce any polymorphic transformations between forms I and III, at least within the range of all desolvation experiments. According to,32 both forms I and III transform to the high-temperature phase on heating up to ∼415 K, and this phase, in turn, transforms to Form-I on reverse cooling. Since different phases were obtained in a narrow temperature range, the formation of Form-I from III is very unlikely. Nevertheless we found that sometimes 520
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
Figure 6. Photos of large crystal of S(THF) before thermogravimetric measurments (a) and its pseudomorph after complete desolvation (b) with corresponding X-ray powder diffraction frame in 2θ range of 5−25° (c), showing the fine powder sample without single spots from big crystallites of furosemide Form-I phase and also absence of preferable texture.
sufficient to make any comments on their origin rather than a general statement of the role of mechanical stress relaxation (any transformation in the solid state) and diffusion limitations (desolvation or solid + gas reactions). The precise analysis of furosemide solvates crystal structures reveals the presence of the second intramolecular N−H···Cl Hbond in the furosemide molecule. This H-bond contributes to stabilizing the gauche+ conformation of sulfamoyl group.
■
ASSOCIATED CONTENT
* Supporting Information S
Figure 7. Scheme of transformation of different samples of furosemide solvates into furosemide Form-I and Form-III on thermal desolvation. Red arrows correspond to samples of large crystals, whereas blue dashed arrows correspond to powder samples.
Experimental X-ray powder patterns of ground samples of S(THF), S(DIOX), and S(DMF), and calculated X-ray powder patterns of corresponding furosemide solvates from single crystal X-ray diffraction; crystal structures of S(THF), S(DIOX), and S(DMF) at 150 K in CIF format. This material is available free of charge via the Internet at http://pubs.acs.org.
grinding could erratically lead to a partial transformation of Form-III into Form-I, and this is why we avoided using grinding after desolvation.
■
4. CONCLUSIONS The solvates of furosemide with tetrahydrofuran, 1,4-dioxane, N,N-dimethylformamide, and dimethylsulfoxide could be considered as promising precursors to obtain different polymorphs of furosemide through the thermal desolvation. Desolvation of large crystals of the solvates with characteristic dimensions more than 1 × 1 × 1 mm3 S(THF), S(DIOX), and S(DMF) leads to the formation of the furosemide Form-III, while powders with characteristic diameter of particles ∼1−10 μm of S(DIOX), S(DMF), and S(DMSO) give the furosemide Form-I instead. The desolvation with the total molar volume loss of about 25% gives fine powders as pseudomorphs preserving the shape of precursor crystals. An attempt to find similarities in the precursor crystal structures of solvates and the corresponding product polymorphs of furosemide after desolvation was not successful. As structural analysis has shown, the furosemide molecule can adopt several different conformations due to the rotation of the furanylmethylamino and sulfamoyl fragments, as well as the changes in the orientation of amino group. The molecular packing, structural motifs and furosemide-furosemide interactions in S(THF), S(DIOX), S(DMSO), and furosemide forms III and I are different. It is very interesting that not only solvate form can influence the desired polymorph of furosemide, but also the particles size of the solvate. The size of molecular crystals is known to influence the outcome of the solid-state transformations. Sometimes, no transformation at all is observed in powder samples, whereas larger crystals undergo a phase transition.51−53 Unfortunately, the number of examples of size effects described for molecular crystals is small and not
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge financial support from the Integration Projects No. 108 of the SB RAS, the Ministry of Education and Science of the Russian Federation (agreement No. 14.B37.21.1093), the Programmes of the Praesidium of the RAS (project No. 24.38) and the Department of Chemistry and Materials Sciences of the RAS (project No. 5.6.4).
■
REFERENCES
(1) Bernstein, J. Polymorphism in Molecular Crystals; Clarendon Press/International Union of Crystallography Monographs on Crystallography: Oxford, U.K., 2002. (2) Braga, D.; Grepioni, F.; Maini, L.; Polito, M. Struct. Bonding (Berlin) 2009, 132, 25−50. (3) Bernstein, J.; Davey, R. J.; Henck, J. O. Angew. Chem., Int. Ed. Engl. 1999, 38, 3440−3461. (4) Boldyreva, E. V.; Drebushchak, V. A.; Drebushchak, T. N.; Paukov, I. E.; Kovalevskaya, Yu. A.; Shutova, E. S. J. Therm. Anal. Calorim. 2003, 73, 409−418. (5) Drebushchak, V. A.; Drebushchak, T. N.; Chukanov, N. V.; Boldyreva, E. V. J. Therm. Anal. Calorim. 2008, 93, 343−351. (6) Chesalov, Yu. A.; Baltakhinov, V. P.; Drebushchak, T. N.; Boldyreva, E. V.; Chukanov, N. V.; Drebushchak, V. A. J. Mol. Struct. 2008, 891, 75−86. (7) Bingham, A. L.; Hughes, D. S.; Hursthouse, M. B.; Lancaster, R. W.; Tavener, S.; Threlfall, T. L. Chem. Commun. 2001, 7, 603−604. 521
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522
Crystal Growth & Design
Article
(44) Etter, M. C.; MacDonald, J. C.; Bernstein, J. Acta Crystallogr. B 1990, 46, 256−262. (45) Bernstein, J.; Davis, R. E.; Shimoni, L.; Chang, N.-L. Angew. Chem., Int. Ed. Engl. 1995, 34, 1555−1573. (46) Garner, W. E. Chemistry of the Solid State; Butterworths: London, 1955. (47) Boldyrev, V. V.; Bulens, M.; Delmon, B. The Control of the Reactivity of Solids; Elsevier: Amsterdam, 1979. (48) Brown, M. E.; Dollimore, D.; Galwey, A. K. Reactions in the Solid State; Elsevier: Amsterdam, 1980. (49) Petit, S.; Coquerel, G. Jt. Eur. Days Equilib. Phases 2009, No. 00016. (50) Petit, S.; Mallet, F.; Petit, M.-N.; Coquerel, G. J. Therm. Anal. Calorim. 2007, 90, 39−47. (51) Boldyreva, E. V.; Drebushchak, V. A.; Paukov, I. E.; Kovalevskaya, Yu. A.; Drebushchak, T. N. J. Therm. Anal. Calorim. 2004, 77, 607−623. (52) Boldyreva, E. V.; Drebushchak, T. N.; Drebushchak, V. A. The 4th Pharmaceutical Powder X-ray Diffraction Symposium (PPXRD-4), Book of Abstracts, Barcelona, Spain 21−24 February, 2005; International Centre for Diffraction Data: Newton Square, PA, 2005, p. 50. (53) Minkov, V. S.; Drebushchak, V. A.; Ogienko, A. G.; Boldyreva, E. V. CrystEngComm 2011, 13, 4417−4426.
(8) Bacchi, A.; Carcelli, M.; Chiodo, T.; Mezzadri, F.; Nestola, F.; Rossi, A. Cryst. Growth Des. 2009, 9, 3749−3758. (9) Basavoju, S.; Aitipamula, S.; Desiraju, G. R. CrystEngComm 2004, 6, 120−125. (10) Samas, B.; Seadeek, C.; Campeta, A. M.; Chekal, B. P. J. Pharm. Sci. 2011, 100, 186−194. (11) Bhattacharya, S.; Saha, B. K. Cryst. Growth Des. 2013, 13, 606− 613. (12) Bechtlov, B.; Nordhoff, S.; Ulrich. J. Cryst. Res. Technol. 2001, 36, 1315−1328. (13) Wirth, D. D.; Stephenson, G. A. Org. Process Res. Dev. 1997, 1, 55−60. (14) Schmidt, A. C.; Niederwanger, V.; Griesser, U. J. J. Therm. Anal. Calorim. 2004, 77, 639−652. (15) Rocco, W. L.; Morphet, C.; Laughlin, S. M. Int. J. Pharm. 1995, 122, 17−25. (16) Braun, D. E.; Kahlenberg, V.; Gelbrich, T.; Ludescher, J.; Griesser, U. J. CrystEngComm 2008, 10, 1617−1625. (17) Griesser, U. J. In Polymorphism in the Pharmaceutical Industry; Hilfiker, R., Ed.; Wiley-VCH: Weinheim, Germany, 2006, pp 211− 233. (18) Feitknecht, W. Kolloid Z. 1934, 68 (2), 128−132. (19) Hüttig, G. In Handbuch der Katalyse; Schwab, G. M., Ed.; Springer: Wien, 1943, pp 322−357. (20) Hedvall, J. A. Solid State Chemistry; Elsevier: Amsterdam, 1966. (21) Boldyreva, E. V. In Implications of Molecular and Materials Structure for New Technologies; Howard, J. A. K., Allen, F. H., Shields, G. P., Eds.; Kluwer Academic Publishers: Dordercht, the Netherlands, 1999; NATO Science Series, Series E: Applied Sciences, Vol. 360, pp 151−174. (22) Oswald, H.-R.; Reller, A. Pure Appl. Chem. 1989, 61 (8), 1323− 1330. (23) Dambournet, D.; Belharouak, I.; Amine, K. Chem. Mater. 2010, 22 (3), 1173−1179. (24) Landgraf, K.-F.; Olbrich, A.; Pauluhn, S.; Emig, P.; Kutscher, B.; Stange, H. Eur. J. Pharm. Biopharm. 1998, 46, 329−337. (25) Rubin-Preminger, J. M.; Bernstein, J. Cryst. Growth Des. 2005, 5, 1343−1349. (26) Suitchmezian, V.; Jeß, I.; Näther, C. Int. J. Pharm. 2006, 323, 101−109. (27) Surov, O. V.; Voronova, M. I.; Smirnov, P. R.; Mamardashvili, N. Z.; Shaposhnikov, G. P. CrystEngComm 2012, 14, 533−536. (28) Braun, D. E.; Gelbrich, T.; Kahlenberg, V.; Tessadri, R.; Wieser, J.; Griesser, U. J. Cryst. Growth Des. 2009, 9, 1054−1065. (29) Caira, M. R.; Bettinetti, G.; Sorrenti, M. J. Pharm. Sci. 2002, 91, 467−481. (30) Lamotte, J.; Campsteyn, H.; Dupont, L.; Vermeire, M. Acta Crystallogr. B 1978, 34, 1657−1661. (31) Babu, N. J.; Cherukuvada, S.; Thakuria, R.; Nangia, A. Cryst. Growth Des. 2010, 10 (4), 1979−1989. (32) Matsuda, Y.; Tatsumi, E. Int. J. Pharm. 1990, 60, 11−26. (33) Ge, M.; Liu, G.; Ma, S.; Wang, W. Bull. Korean Chem. Soc. 2009, 30 (10), 2265−2268. (34) X-AREA and X-RED; Stoe & Cie GmbH: Darmstadt, Germany, 2007. (35) Sheldrick, G. M. Program for Refinement of Crystal Structures; University of Göttingen: Göttingen, Germany, 1997. (36) Sheldrick, G. M. Acta Crystallogr. A 2008, 64, 112−122. (37) X-STEP32; Stoe & Cie GmbH: Darmstadt, Germany, 2000. (38) Macrae, C. F.; Edgington, P. R.; McCabe, P.; Pidcock, E.; Shields, G. P.; Taylor, R.; Towler, M.; van de Streek, J. J. Appl. Crystallogr. 2006, 39, 453−457. (39) Farrugia, L. J. J. Appl. Crystallogr. 2012, 45, 849−854. (40) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7−13. (41) Allen, F. H. Acta Cryst. B 2002, 58, 380−388. (42) Stepanovs, D.; Mishnev, A. Acta Crystallogr. C 2012, 68, o488− o491. (43) Goud, N. R.; Gangavaram, S.; Suresh, K.; Pal, S.; Manjunatha, S. G.; Nambiar, S.; Nangia, A. J. Pharm. Sci. 2012, 101 (2), 664−680. 522
dx.doi.org/10.1021/cg401257w | Cryst. Growth Des. 2014, 14, 513−522