Solid–Solid Transition between Hydrated Racemic ... - ACS Publications

All the solid state studies indicate that the RS-H → RS-A transition is fully reversible under ... Patrizia RossiPaola PaoliLaura ChelazziLuca Conti...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/crystal

Solid−Solid Transition between Hydrated Racemic Compound and Anhydrous Conglomerate in Na-Ibuprofen: A Combined X‑ray Diffraction, Solid-State NMR, Calorimetric, and Computational Study Patrizia Rossi,† Eleonora Macedi,† Paola Paoli,*,† Luca Bernazzani,‡ Elisa Carignani,‡ Silvia Borsacchi,‡ and Marco Geppi‡ †

Dipartimento di Ingegneria Industriale, Università degli Studi di Firenze, via S. Marta 3, 50139 Florence, Italy Dipartimento di Chimica e Chimica Industriale, Università di Pisa, via Risorgimento 35, 56126 Pisa, Italy



S Supporting Information *

ABSTRACT: A combined X-ray diffraction (XRD), solid state nuclear magnetic resonance (SSNMR), differential scanning calorimetry (DSC), and modeling approach has been applied to study the solid−solid transition of ibuprofen sodium salt between the hydrated racemic compound (RS-H) and the anhydrous conglomerate (RS-A). For comparison, the dihydrate → anhydrous transformation of the sodium salt of the pure S-enantiomer of ibuprofen was also investigated by means of SSNMR and DSC. All the solid state studies indicate that the RS-H → RS-A transition is fully reversible under different experimental conditions [temperature, pressure (ambient and vacuum), and type of atmosphere (N2, air, and static dry by P2O5)]. The static and dynamic disorder affecting the isobutyl fragment in RS-H, already observed by SSNMR, has been further investigated by single crystal XRD and computational techniques. On these grounds, a model for the dihydrate → anhydrous solid−solid transformations is proposed.



INTRODUCTION Solid state analysis plays a key role in the characterization of pharmaceutical materials since it can help to understand phenomena such as amorphicity, polymorphism, pseudopolymorphism, disorder in the crystalline state, intermolecular interactions, solvation, and other related issues.1−6 In this context, thermal, spectroscopic, and diffraction techniques are very popular analytical methods, and in particular solid state NMR (SSNMR), X-ray diffraction (XRD, both from singlecrystal and microcrystalline powder), and differential scanning calorimetry (DSC) have been proven to be excellent complementary methods.7−9 In fact, while X-ray crystallography needs high quality single crystals to get a direct picture of a molecular species and its arrangement in the crystal packing, SSNMR, without the need for single crystals, provides detailed molecular information, such as spatial proximities, disorder, solvation, and number of independent molecules.10,11 Furthermore, both SSNMR and X-ray powder diffraction can provide information about phase distribution (polymorphism, pseudo-polymorphism) when applied to mixtures or about phase changes when phase transformations are investigated. Thermal analysis techniques, such as DSC, can be used to characterize polyphasic substances to distinguish between enantiotropic and monotropic systems and to determine heats and temperatures related to phase transitions and details about water sorption−desorption processes, solvation, decomposition, just to name a few. Finally, crystal structure prediction methods, Monte Carlo and molecular dynamics simulations can provide useful details for understanding phase transformations © 2014 American Chemical Society

and disorder in organic crystal structures at the molecular level.12 Herein we report the combined use of SSNMR, DSC, and Xray crystallography (by using both single crystal and microcrystalline powder) and computational studies to investigate the solid state behavior of ibuprofen sodium salt (see Scheme 1). Ibuprofen (α-methyl-4-(isobutyl) phenylacetic acid) is a nonsteroidal anti-inflammatory drug (NSAID) widely used in the treatment of rheumatic disorders, pain, and fever.13−15 The corresponding dihydrate sodium salt, which crystallizes from water, is more water-soluble and absorbed into blood plasma more quickly than ibuprofen.16 For both ibuprofen and its Scheme 1. Schematic drawing of the Ibuprofen Anion with the Labelling Used for Carbon Atoms in the Analysis of SSNMR Data

Received: January 29, 2014 Revised: March 3, 2014 Published: March 21, 2014 2441

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

collect the solid (S)-Na-ibuprofen dihydrate (S-H) (265 mg, 68%). The microcrystalline powder suffers from significant preferred orientation, which was mitigated by collecting diffraction data with a spinning capillary (see Figure S1 in Supporting Information). The microcrystalline powder is isomorphic with that reported by Zhang et al.25 given that the experimental powder X-ray diffraction (XRPD) pattern is superimposable with the theoretical pattern obtained with the Mercury 3.1 program26 by using the crystallographic data retrieved from the single crystal experiment.25 Because of the preferred orientation affecting the samples, at variance with RS-H (vide infra) accurate measures of the Bragg reflection intensities in the hot chamber were precluded, and no temperature-resolved XRPD experiments were carried out. Samples Preparation for Single Crystal X-ray Analysis. Crystals of RS-H were obtained by dissolving 125 mg (0.6 mmol) of the commercial sodium salt in distilled water (0.5 mL) and then adding acetonitrile (5 mL). The solution was covered and cooled to 273 K. Single crystals were obtained as colorless thin plates. Crystals were stored in contact with their mother liquor, making sure the container was well closed. All the attempts to obtain single crystals of the dehydrated species of (R,S)-Na-ibuprofen and (S)-Na-ibuprofen (RS-A and S-A, respectively) suitable for X-ray diffraction analyses were unsuccessful. Samples Preparation for SSNMR. The anhydrous samples (RS-A and S-A) were obtained by drying the dihydrate forms (RS-H and S-H, respectively) in an oven at 363 K for 12 h. Rehydrated RS-rH and SrH were obtained from the corresponding anhydrous forms by exposure to atmospheric humidity for 5 h. Single Crystal X-ray Diffraction (SCD). Single crystals were mounted in air on glass fibers. Intensity data for RS-H were collected on an Oxford Diffraction Excalibur diffractometer using a Mo Kα radiation at room temperature (RS-Hrt), 150 K (RS-H150), 120 K (RS-H120), and 15 K (RS-H15). In all cases, data collection was performed by using the program CrysAlis CCD,27 data reduction was carried out with the program CrysAlis RED,27 and absorption correction was performed with the program ABSPACK in CrysAlis RED.27 The structures were solved by using the SIR-97 package28 and subsequently refined on the F2 values by the full-matrix least-squares program SHELXL-97.29 In all cases, the non-hydrogen atoms, except those of the isobutyl moiety where disorder occurs, were anisotropically refined. At 298, 150, and 120 K, the disorder affecting the CH(CH3)2 moiety was modeled by three fragments (a, b, and c). However, while at room temperature for each carbon atom, namely, C11, C12, and C13, three different positions were found, at 150 and at 120 K the a, b, and c models share the methyl group labeled C12. Finally, at 15 K only two orientations (a and b) of the isobutyl group were found that, as for the 150 and 120 K cases, share C12. All the hydrogen atoms of the ibuprofen anion were introduced in calculated positions and refined in agreement with the atoms to which they are bound. Finally, the hydrogen atoms of the water molecules were found in the Fourier difference map. Table S1 in Supporting Information reports the crystal data and refinement parameters. Powder X-ray Diffraction (XRPD). XRPD measures were carried out at room temperature by using a Bruker D8-Advance diffractometer (Cu Kα radiation, 40 kV × 40 mA), equipped with a Bruker Sol-X energy dispersive X-ray detector, scanning range 2θ = 3−30°, 0.03° increments of 2θ and a counting time of 1 s/step. Dehydration tests on RS-H at room temperature were carried out by introducing the sample in a closed hot XRPD chamber together with P2O5. An XRPD pattern was immediately recorded, and then further patterns were collected after 40, 60, 75, 145, 160, and 190 min. Temperature-resolved experiments (in the range 298−483 K) were performed with an Anton Paar HTK 1200N hot chamber mounted on a Panalytical X’Pert PRO automated diffractometer (Cu Kα radiation, 40 kV × 40 mA), equipped with the PIX-CEL solid state fast detector. The scanning range was 2θ = 3−50° with a 1 s/step counting time and 0.03° increments of 2θ. The temperature variation rate was 10 K min−1, and after the target temperature was reached the sample was kept for 10 min at that temperature before proceeding with data collection. Measures were performed both in air and under N2

dihydrate sodium salt, whose tolerability and safety profiles are comparable, the desired pharmacological effects reside almost exclusively in the S-enantiomers. Nevertheless the racemic compounds are actually used as drugs. Incidentally, the other enantiomer is partially converted into S-ibuprofen in humans. The chiral behavior of the ibuprofen free acid has been the subject of numerous studies,17−19 and it has been reported that the racemic compound or “true racemate” (“any homogeneous solid composed of equimolar amounts of enantiomeric molecules”, following the IUPAC rule) is much more stable than the corresponding racemic conglomerate (or racemic mixture, that is, “a mixture of equimolar amounts of enantiomeric molecules present as separate solid phases”, following the IUPAC rule). In particular, the true racemate form is the only stable solid form, and, as a consequence, (R,S)ibuprofen has a dimorphic (true racemate/conglomerate) behavior of the monotropic type.20 On the other hand, it is known that racemic sodium ibuprofen, (R,S)-Na-ibuprofen, in the solid state can be found both as dihydrate racemic compound (RS-H, hereafter) and as anhydrous racemic as well as anhydrous conglomerate (RS-A, hereafter) forms.21−23 In particular, Zhang et al. performed separate studies of RS-H and RS-A forms. They found that RSH is stable at ambient conditions as racemic compound, while three anhydrous forms, called α, β, and γ, have been observed and classified as racemic compounds (α and β) and racemic conglomerate (γ):21 α and β are observed by solidification of the melt, while the latter is the most stable form of RS-A. This finding is in keeping with the observation that racemic conglomerates are more common among salts with respect to the parent organic acids (or bases), as provided by a survey of more than 500 organic chiral compounds by Jacques et al.24 Moreover, Censi et al.23 recently studied the thermal dehydration of the racemic compound RS-H into RS-A, and they stated: “the water removal provoked a perturbation into the crystal leading to a less organized and less crystalline structure, where the disorder degree is higher”. With this in mind, we applied our combined approach (XRD, DSC, and SSNMR) to the solid state analysis of the racemic sodium ibuprofen dihydrate phase (RS-H), focusing our attention in particular on its transformation to RS-A under different experimental conditions: temperature, pressure (ambient and vacuum), and type of atmosphere (N2, air, and static dry by P2O5). We characterized in detail the RS-A form, also formulating a model for the transition mechanism. Ab initio and molecular dynamics simulations were also used to study the conformational space of the ibuprofen anion. Eventually, the SSNMR and DSC analyses were also extended to the investigation of the dehydration process of the pure Senantiomer.



MATERIALS AND METHODS

Active Pharmaceutical Ingredients. (R,S)-Na-ibuprofen was purchased from Sigma Aldrich (CAS number 31121-93-4) and used after exposure to atmospheric humidity, which is known to transform RS-A into RS-H racemic compound.21 (S)-Na-ibuprofen dihydrate (S-H) was synthesized in house by the neutralization of sodium hydroxyde and (S)-ibuprofen. In particular, sodium hydroxide (58.6 mg, 1.46 mmol) was dissolved in distilled water (0.2 mL), and then the basic solution was slowly added to a solution of (S)-ibuprofen (purchased from Sigma Aldrich, CAS number 51146-56-6) (302 mg, 1.464 mmol) in acetone (2 mL) at room temperature. A white crystalline solid promptly precipitated. The mixture was stirred for 50 min, cooled to 278 K, and finally filtered to 2442

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Table 1. Dihedral Angle Values (°) Defining the Conformational Isomers Found by XRD in the Solid State Structure of RS-Ha

a

rt stands for room temperature, while the other temperatures are explicitly indicated.

Figure 1. ORTEP3 view of the asymmetric unit of RS-H (rt). Displacement ellipsoids are drawn at 15% probability level. The three positions for the disordered isobutyl fragment are shown. H atoms have been omitted for the sake of clarity. atmosphere. In addition, measures at room temperature under a N2 flow were performed to study the dehydration process. A Panalytical X’Pert PRO automated diffractometer with Cu Kα radiation and X’celerator detector equipped with an Anton Paar TK 450 chamber was used to collect data in the range 298−120 K in vacuum. Solid State NMR (SSNMR). 13C CP MAS experiments were performed on a Varian InfinityPlus 400 spectrometer, equipped with a 3.2 mm CP MAS probe, working at a Larmor frequency of 400.03 and 100.59 MHz for 1H and 13C, respectively. The 1H 90° pulse duration was 2.0 μs. A spinning frequency of 10 kHz was used for the acquisition of 13C CP MAS spectra, where CP was performed by using a linear ramp and a contact time of 1 ms, and 1H nuclei were decoupled by a SPINAL-64 sequence30 with a nutation frequency of 100 kHz. Differential Scanning Calorimetry (DSC) and Thermogravimetric Analysis (TGA). DSC experiments were performed with a Perkin-Elmer Pyris Diamond differential scanning calorimeter. The measurements were performed in the temperature range 273−523 K at 10 K/min, under nitrogen as purging gas in open aluminum pans. The sample masses of both RS-H and S-H ranged from 4 to 5 mg. A TA Instruments Thermobalance model Q5000IR was employed for thermogravimetric measurements. The experiments were performed at a rate of 10 K/min, from 303 to 573 K under nitrogen flow (25 mL/ min) as purging gas. The amount of sample in each TG measurement varied between 2 and 4 mg. Molecular Modeling. Geometry optimizations (MM) and molecular dynamics (MD) simulations were performed on each of the conformational isomers found in the X-ray crystal structure of RSH. All calculations were made by using the CHARMm31 Force Field. MM calculations were performed on each species by using the Smart Minimizer energy minimization procedure implemented in Accelrys Discovery Studio 2.1,32 and before starting the MD simulations, the geometry of each compound was further optimized using the steepest descent and conjugate gradient algorithms. MD simulations were carried out at 40 and 300 K, both in vacuum as well as in an implicit water model; water calculations were performed mimicking the solvent by using a distance-dependent dielectric constant of 80. In the molecular dynamics simulations, the time step was 1 fs for all runs, equilibration time = 200 ps, production time = 2000 ps, and snapshot

conformations were sampled every 20 ps. The programs used for the MD and the energy minimization were the Standard Dynamics Cascade and Minimization protocols, and trajectories were analyzed by the Analyze Trajectory protocol, all implemented in Discovery Studio 2.1. GAUSSIAN03 (Rev. C02)33 was used for the ab initio computational studies. In all cases, the level of theory was HF-SCF, and the basis set was 6-311G(d,p).34 The Berny algorithm was used.35 The reliability of the stationary points was assessed by the evaluation of the vibrational frequencies. Geometry optimizations were performed on the three conformational isomers a, b, and c. The potential energy surface (PES) was explored by relaxed PES scans about the dihedral angles d1 and d2 (Table 1), which define the orientation of the isobutyl chain (dihedral starting values −120 and −60°; increment size 30°; number of points 3 and 6, for d1 and d2 respectively). A total of 18 scan points resulted.



RESULTS Single Crystal X-ray Diffraction and Modeling Studies. The X-ray data collected at temperatures ranging from room temperature to 15 K show no changes in the crystalline form of RS-H (see Table S1 in Supporting Information), which is substantially identical to that reported by Zhang (X-ray data collected at 173 K, Table S2 in Supporting Information).25 This indicates that the dihydrate phase of the racemic sodium ibuprofen is stable from 15 K up to room temperature. However, while Zhang modeled the disorder affecting the isobutyl chain by two fragments (a and b, hereafter) with occupancy factors 0.554 and 0.446, respectively, the data collected for the present work at room temperature, 150 K, and 120 K show a more complex picture. In these cases, the rotational disorder affecting the isobutyl group has been modeled over three positions. As a result, three conformational isomers, a, b, and c (see Table 1 and Figure 1) were found having different populations. In details, refinement with the disordered model gave a ratio of 0.46(2):0.35(2):0.19(2) at room temperature for the occupancy factor of a, b, and c, respectively; the ratio was determined to be 2443

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Figure 2. (a) 2D scatter plot of the d1 and d2 dihedral angles (°) from CSD data; (b) 2D scatter plot of the d1 and d2 dihedral angles (°) from MD trajectory (in vacuum, T = 300 K).

Figure 3. Experimental XRPD patterns of RS-A obtained at rt in P2O5 and in N2 flow.

intermolecular contacts in the crystal, thus strengthening the hypothesis of a quite complex structural disorder picture. Molecular Dynamics simulations performed on the isolated ibuprofen anion at different temperatures provide hints about the dynamic behavior of the isobutyl chain. The conformational space sampled by the side chain at 300 K (Figure 2b) reproduces well the dihedral angle distribution found in the solid state (CSD data, Figure 2a), while, as expected, at 40 K the isobutyl arm remains trapped within a limited region of the conformational space. To get a rough idea of the energy barrier separating the isomers, the potential energy surface (PES) nearby the three conformational isomers found in the solid state (T ≥ 120 K) was explored by relaxed PES scans about d1 and d2 (HF/6-311G(d,p) model chemistry). The a and b conformers appear isoenergetic, each one located in the PES minimum, and the c species is in a local minimum 12 kJ/mol higher in energy. Within this model, the isomers interconversion should not require more than 16 kJ/mol. The energy values obtained by PES refer to an isolated ibuprofen anion, and therefore they cannot be interpreted as precise quantitative estimations of the real energies in the solid state. Nevertheless,

0.58(1):0.27(3):0.16(3) and 0.57(1):0.30(4):0.13(4) at 150 and 120 K, respectively. Finally at 15 K, no residual peaks corresponding to atoms of the minor orientation could be detected and the a and b conformational isomers accounted for all the disorder (occupancy factors 0.542(5) and 0.458(6), respectively). Although these two-three conformations reproduce well the X-ray electron density map, the disorder affecting the isobutyl group could involve a much more complex distribution of conformers, as provided by recent SSNMR findings36 and as suggested by a survey of the Cambridge Structural Database (CSD, v. 5.34 updates May 2013)37 and the results from modeling (vide infra). In fact, the orientations characterizing the a and c isomers represent only the 5.6% of the structures deposited in the CSD (Figure 2a). Moreover, both conformers, at variance with b, undergo a large conformational change upon optimization of the isolated anion (force field and ab initio methods, see Materials and Methods), which leads to a further conformational isomer (−73°/ +55° for d1 and d2, respectively), isoenergetic with b (d1 = −105.3°, d2 = −54.8°). Finally, the isobutyl chains, which define hydrophobic channels, are not involved in significant 2444

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

under these experimental conditions. Finally, only a partial dehydration is observed on cooling to 120 K in a vacuum; in other words both the RS-H and RS-A species coexist in these experimental conditions and after heating back to room temperature (see Figure S5 in Supporting Information). To get more insights about the solid−solid transformation of the racemic sodium ibuprofen leading to the enantiomer resolution, temperature-resolved XRPD analyses in the range 298−483 K in N2 atmosphere were performed. During heating, RS-H undergoes a dehydration process that starts at the beginning of the experiment (as already reported the nitrogen flow promotes the removal of the two water molecules) and ends at 333 K: it can be seen how the characteristic diffraction peaks of RS-H are progressively replaced by those of RS-A, which was found to be the dehydrated conglomerate γ phase (Figure 4). The γ phase is then stable till fusion (473 K). Finally, several attempts were carried out to determine the structure of RS-A by XRPD data, but all were unsuccessful. Differential Scanning Calorimetry. Figure 5 reports the experimental DSC curves of S-H (a) and RS-H (b), respectively. The first heating DSC trace shows, for both samples, an intense rather broad peak at around 353 K (348 K in the case of the pure S-enantiomer). This peak is associated with the dehydration process, which is known to occur in this temperature range.25 From TGA, it was seen that this process is associated with a mass loss of 13.4%, corresponding to two molecules of water per Na-ibuprofen. The enthalpic effect is stronger in the case of the racemic sample (about 400 J/g with respect to 270 J/g observed for the S-enantiomer). Normalizing the heat effect per mole of water ripped out from the crystal lattice, molar enthalpies of 37 and 55 kJ are derived for the pure enantiomeric and the racemic samples, respectively. The first value approximately matches the enthalpy of vaporization of water (about 41 kJ/mol at 348 K), but the second is about 50% higher, indicating that an additional aliquot of energy is needed to perform the RS-H → RS-A transition with respect to the S− H → S-A one. A small peak at about 376 K is observed in the case of S-H but not for RS-H. This could be attributed to minor readjustment of the lattice following the loss of water. Finally, S-A undergoes melting at about 505 K with a corresponding enthalpy of fusion of about 117 J/g of the dehydrated compound (S-A). As far as RS-H is concerned, above the dehydration event, no significant signal is observed during the first heating until the sample approaches melting. In this case, a weak shoulder at about 467 K precedes the main melting peak whose maximum lies at about 473 K. These features nearly exactly match the observation by Zhang et al.21 On the basis of four thermodynamic criteria first formulated by Jacques38 and then extensively applied by Li et al.1 to several chiral pharmaceuticals, these authors concluded that the peak at 473 K (472 K in their work) corresponds to the melting of a racemic eutectic conglomerate (γ-form). The first cooling curve, of both the samples, shows a quite sharp crystallization peak with a hysteresis of about 21 and 44 K in the case of the pure enantiomer and the racemic mixture, respectively. As expected, in the second heating curves of both samples, the peak ascribed to the dehydration is not present. We notice that on second heating the melting point of the pure Senantiomer does not change with respect to the first heating,

we believe that this approximation is plausible for the isobutyl fragment, because it is shown from the crystal structure that the hydrophobic part of the molecule is not involved in significant intermolecular interactions in the crystal. Microcrystalline Powder X-ray Diffraction. The solid− solid transition occurring in the racemic sodium ibuprofen during the water sorption−desorption process under different experimental conditions (temperature, vacuum, air, inert atmosphere) was tracked by X-ray powder diffraction measurements. As a preliminary test, a fresh packet of the racemic sodium ibuprofen was opened, and immediately a XRPD pattern was collected at room temperature in air; then powder patterns were recorded every 30 min for 4 h. As expected, the salt (sold as anhydrous) hydrates quickly and after about 1 h the anhydrous (RS-A) → hydrate (RS-H) transformation is almost completed. The XRPD pattern of the hydrate phase compares well with the theoretical one obtained from the single-crystal data of the racemic sodium ibuprofen dihydrate phase collected at room temperature (RS-Hrt) (see Figure S2 in Supporting Information). On the other hand, dehydration of RS-H toward RS-A (vide infra) can be easily achieved in about 3 h by storing the sample at room temperature (rt) with P2O5 as provided by repeating XRPD data collection at regular intervals (see Figure S3 in Supporting Information). A faster dehydration (about 2 h) is even observed when the sample is stored at rt under a N2 flow (see Figure S4 in Supporting Information). Incidentally, the same treatment (rt, N2 flow) applied to a single crystal of RS-H immediately damages the crystal that stops diffracting (SCD experiment). In both cases (P2O5, N2 atmosphere), the XRPD patterns of the resulting dehydrated sample reveal that the same form is obtained (Figure 3), which also corresponds to the γ phase, after Zhang,21 of the racemic conglomerate (see Table 2 listing the characteristic peaks of RS-H and RS-A). The anhydrous sample quickly hydrates once stored in air at rt giving the RS-H phase. Thus, the water sorption−desorption process appears reversible when dealing with the microcrystalline sample, while no single crystal−single crystal water removal can be achieved, at least Table 2. Characteristic XRPD Peaks of the Dihydrate and Dehydrate (γ Phase) Species of Racemic Sodium Ibuprofen at rt 2θ (deg) RS-H

RS-A (γ phase)

3.81 4.00 11.21 12.00 16.00 16.54 17.12 17.35 17.67 17.91 17.93 18.18 18.85 18.96 19.33 19.48 20.06 2445

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Figure 4. Temperature-resolved XRPD analysis of RS-H between 298 and 343 K under N2 flow showing the gradual transformation into RS-A (γ phase).

spectrum as S-H, clearly indicating that by exposing the anhydrous sample to air the stable dihydrate S-H form is easily reobtained. On the contrary, the spectra of S-H and S-A are significantly different, although the line width of the signals in the two spectra is approximately the same. For both samples, several carbons give rise to doublets rather than to singlets, indicating that two nonequivalent molecules are present in their crystallographic asymmetric units. For S-H, this observation is in agreement with XRD data previously reported by Zhang et al.25 While in the spectrum of S-H only the methyl signals (carbons m, n, and o, see Scheme 1) are split into doublets, in the case of S-A the signals of carbons a, b, c, i, and o (see Scheme 1) are all split, with splittings larger than those observed for S-H. This indicates that for S-A there is a stronger difference in chemical environment between the two molecules of the asymmetric unit, particularly for the isopropionic fragment. In Figure 7, the spectra of the dihydrate, anhydrous and rehydrated forms of (R,S)-Na-ibuprofen are shown. As already observed for (S)-Na-ibuprofen, the spectrum of RS-rH is the same as that of RS-H, indicating that, also in this case, the dehydration/hydration process is reversible, as already observed in the XRPD experiments. Comparing the spectra of RS-H and RS-A, we can first notice that, while in the spectrum of RS-H only one signal is present for each chemically inequivalent carbon, in the spectrum of RS-A several signals are split into doublets. Again, this indicates that one and two molecules are present in the crystallographic asymmetric unit of RS-H (in agreement with single crystal XRD data) and RS-A, respectively. It must be noticed that all the signals of RS-A appear broader (on average by about 30%) than those of RS-H and all forms of (S)-Na-ibuprofen. This suggests that, even though RS-A is crystalline, it exhibits a larger degree of structural disorder with respect to the other forms. Incidentally, the same conclusion can be drawn by comparing the RS-H and RS-A XRPD patterns: the latter show broader and more poorly defined peaks.

Figure 5. Experimental DSC curves of S-H (a) and RS-H (b). In both graphics, the curves relative to the first heating (top), first cooling (middle), and second heating (bottom) are reported. The experiments were performed at a rate of 10 K/min.

while in the case of the racemic mixture the thermogram is the same as that obtained by Zhang et al.,21 which was explained by the coexistence of α, β, and γ phases. Solid State NMR. 13C CP MAS spectra were recorded for all the samples and are reported in Figures 6−8. The chemical shifts of all the signals and the spectral assignment, performed on the basis of that of RS-H,39 are reported in Table 3. In Figure 6, the spectra of S-H, S-A, and S-rH are shown. First, it is worth noticing that the S-rH sample shows the same 2446

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Figure 6. 13C CP MAS NMR spectra of (a) S-H, (b) S-A, and (c) S-rH. For all the experiments, 4000 transients were accumulated with 5 s of recycle delay. Spinning sidebands are marked with asterisks.

Figure 7. 13C CP MAS NMR spectra of (a) RS-H, (b) RS-A, and (c) RS-rH. For spectrum (a) 12 000 transients were accumulated, while for spectra (b) and (c) 4000 transients were accumulated. In all three experiments, a recycle delay of 5 s was used. Spinning sidebands are marked with asterisks.



By comparing the spectra of S-A and RS-A (Figure 8), it is possible to first notice a quite different line broadening for all resonances. By applying a suitable artificial line broadening to the spectrum of S-A, we could reproduce quite nicely the spectrum of RS-A, with the partial exception of the structure of the peaks ascribed to the protonated aromatic carbons (d−g, see Scheme 1) at about 130 ppm and the different relative intensities of the two peaks at about 20 ppm, due to carbon o (see Scheme 1). This suggests that the crystal structures of S-A and RS-A are probably very similar, supporting the XRPD findings,21 apart from the quite larger structural disorder or RSA, already discussed. Small local differences in the structure seem to involve both the aromatic and the methyl fragments. This is a further indication of the conglomerate nature of RS-A.

DISCUSSION

The results above-reported mainly focus on two aspects: (a) the interpretation of the disorder seen by single crystal XRD techniques for the isobutyl group in RS-H at different temperatures ranging from 15 K to room temperature; (b) the characterization and possible the structural model of the RS-H → RS-A and S−H → S-A solid−solid phase transition. The first aspect has been partially discussed in previous papers dealing with variable-temperature SSNMR characterizations. In particular, it was observed that around room temperature the isobutyl group does experience fast (with characteristic time of about 10−10 s) interconformational dynamics involving rotations about the phenyl-CH2 (d1) and CH2−CH (d2) single bonds (dynamic disorder).40 On the 2447

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Figure 8. Comparison of the 13C CP MAS NMR spectra of (a) S-A, (b) S-A with an artificial line broadening of 80 Hz, and (c) RS-A.

Table 3. Spectral Assignment of 13C CP MAS NMR Spectra of the Four Samples S−H, S-A, RS-H, and RS-Aa δ (13C)/ppm

a

nucleus

S-H

S-A

RS-H

RS-A

a b c d−g h i l m−n o

183.9 142.9 139.7 129.9−128.1 45.4 50.0 31.1 23.7−23.2 18.8−17.8

185.2−183.9 144.4−140.9 139.5−139.1 130.5−129.3−128.1 45.6 51.3−50.1 31.2 23.5 20.3−18.3

184.4 142.8 139.7 129.8−128.4 45.5 49.5 31.1 23.5 17.3

185.2−184.2 144.5−140.7 139.1 130.3−129.4−128.1 45.6 50.9−49.9 31.1 23.5 20.1−18.0

Chemical shifts values are affected by an experimental error of ±0.1 ppm. Carbon atoms labelling refers to Scheme 1.

Table 4. H-Bond Interactions in the Crystal Packings of RS-H and S-H from ref 25a

Geometric parameters were taken from ref 25 for RS-H too in order to compare data collected at the same temperature (173 K). 1: In RS-H 1 = −x + 1, −y, −z; in S−H 1 = x, y, z. 2: In RS-H and in one-half of asymmetric unit of S−H 2 = x + 1, y − 1, z; in the second half of the asymmetric unit of S−H 2 = x − 1, y + 1, z. 3: In RS-H and in one-half of asymmetric unit of S−H 3 = x + 1, y − 1, z; in the second half of the asymmetric unit of S−H 3 = x − 1, y + 1, z. 4: In RS-H and in one-half of asymmetric unit of S−H 4 = x + 1, y, z; in the second half of the asymmetric unit of S−H 4 = x − 1, y, z. bInteraction present in RS-H and in one-half of the asymmetric unit of S-H. cInteraction present in the other one-half of the asymmetric unit of S-H. a

disorder).36 The XRD and modeling results here reported agree with the NMR findings. Indeed, the energy barriers separating the a, b, and c isomers appear too low to make them separable in the gas phase at room temperature (MD and relaxed PES scan data).41 In other words, in the isolated molecule the rotation around d1 and d2 appears quite easy, and different

other hand, by lowering temperature, these motions have been seen to progressively slow down and freeze, reaching characteristic time values on the order of seconds at about 40 K. The freezing of the motions leaves the isobutyl groups of different molecules in different conformations leading to an overall distribution of conformations in the sample (static 2448

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Figure 9. Crystal packing of the RS-H and S-H salts highlighting the hydrophilic zone acting as a zip between the anions.25

Scheme 2. Schematic Representation of the Proposed Mechanism for the S−H → S-A (a) and RS-H → RS-A (b) Transformations

forms. First of all, the crystal packings of RS-H and S-H, as already pointed out by Zhang and co-workers, 25 are comparable. In fact, cell dimensions, density, and packing arrangement of the ibuprofen anions, sodium cations, water molecules as well as the H-bond pattern are very similar (Table 4, Figure 9, and Table S2 in Supporting Information). As for the latter, the interactions involving the water molecules in the two crystal structures appear equivalent in terms of number, type (H-bond donor and/or acceptor, Na−H2O interactions), and geometry (bond distances and angles). In particular, in both crystal structures the ibuprofen anions, which are bridged by the sodium cation through the carboxylate group, form monodimensional chains along the a-axis direction. In RS-H, whose asymmetric unit contains one ion pair and two water molecules, each chain is built up by a single enantiomeric species, and the two chains containing opposite enantiomers are paired together through sodium cations and water molecules, while in S-H all the chains are obviously homochiral. In RS-H chains containing opposite enantiomers are symmetry

conformational isomers can be populated. On the other hand, the disorder affecting the isobutyl chain in the crystal of RS-H reflects the fast motion of the side chain. In fact, the temperature dependence of the isomer populations suggests that, at least at temperature higher than 120 K, the disorder is basically dynamic in nature:42 as the temperature increases, the populations of the three orientations found for the isobutyl chain in the solid state tend to equilibrate. On the contrary, at very low temperature (T = 15 K) the presence of only two populated conformers (a and b) seems to indicate the existence of a static disorder. In addition, the disappearance of the minor conformational isomer c implies that a conformational rearrangement takes place in the solid through rotations about d1 and d2, probably also favored by the lack of significant intermolecular contacts involving the isobutyl side arm in the packing (vide infra). As far as the second issue is concerned, to understand the mechanism of dehydration occurring in both RS-H and S-H, it is helpful to start comparing the solid state behavior of the two 2449

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

anions (see, for example, refs 43 and 45). As a consequence, we postulate that the departure of the water molecules makes the crystal collapse, leading, possibly, to a transient disordered phase. This is in agreement with the occurrence of damages always observed for RS-H single crystals when subjected to water removal by nitrogen flow, in the attempt of obtaining good RS-A single crystals for X-ray diffraction. Considering that in the dihydrate phase the hydrophilic tails of the ibuprofen anions belonging to differently chiral chains are kept together through hydrophilic channels including water molecules and sodium ions, the transition can be tentatively modeled by a mechanism involving the loss of water molecules and the opening of the channel followed by the breakage of the quite strong O-carboxy−Na+ interactions, which indeed would correspond to an additional enthalpy for this process with respect to the analogous transition in the optically pure species. By contrast, we may suppose that the hydrophobic tails of chirally opposite ibuprofen anions easily separate due to their almost irrelevant intermolecular interactions. The easy rotation about the phenyl-CH2 (d1) and CH2−CH (d2) single bonds (SSNMR, single crystal XRD and modeling data) supports well this assumption. Given that both the SSNMR spectrum and the XRPD pattern of RS-A suggest a less ordered structure with respect to the dihydrate one, we can hypothesize that this is due to a transient disordered state occurring during the dehydration of RS-H (Scheme 2b). As a consequence, the daughter anhydrous phase which recrystallizes does not maintain the structural features of the dihydrate parent (RS-H), and in fact the conglomerate phase is obtained, as suggested by the SSNMR and XRPD findings,21 indicating that the crystal structures of S-A and RS-A are closely related (except for the already cited larger structural disorder affecting RS-A). On the other hand, in both the asymmetric units of RS-A (SSNMR data) and S-A (SSNMR data) two independent ion pairs are present (one for RS-H, SSNMR, and single crystal XRD data). Given that this solid−solid transformation occurs in the absence of water we must assume that the particles have sufficient mobility to rearrange in homochiral crystals. Finally, the complete reversibility of the sorption−desorption process does not imply the reversibility of the mechanism. As a matter of fact, this kind of solid state transformation of a racemic compound into a conglomerate, or vice versa, is quite an uncommon case.46−51

related by the inversion center and held together by the crystallization water molecules. The latter, together with the sodium cations and the carboxylate groups, define a hydrophilic buffer zone that acts as a sort of zipper connecting the anions having opposite chirality. A very similar arrangement can be recognized in S-H, whose unit cell contains two independent ibuprofen anions. In this case, homochiral chains, related by a pseudo-2-fold screw axis, are zipped up by the bridging water molecules. The close resemblance between the RS-H and S-H crystal lattices, due to the sodium−oxygen interactions which drive the building of the crystals,25 parallels well their similar bulky behavior on heating: in both cases, the loss of the two crystallization water molecules occurs in a single step, as provided by thermogravimetric analysis and differential scanning calorimetry. However, in RS-H the molar enthalpy associated with the loss of water is about 50% larger than in SH, indicating the presence of a somewhat different process. Indeed, this might be explained by the fact that the dehydration of RS-H also results in the transformation of a racemic compound in a conglomerate (accordingly observed by XRPD, SSNMR, and DSC), which requires a more complex rearrangement of the molecules in the solid state with respect to what happens in S-H. In the latter, a zip-mechanism43 for the dehydration process appears quite reasonable (see Scheme 2a). As a consequence, the diffusion of the water molecules out of the crystal, which most probably occurs through the hydrophilic channels, causes a smooth deformation of the zip that keeps the homochiral ibuprofen chains together. In other words, we can postulate that the Na+−−OOC interactions, even if weakened, are maintained during the solid−solid transformation and drive the rearrangement of the ion pairs in the final S-A form. The postulated mechanism agrees with the calorimetric data: vaporization of water at 348 K, followed by a minor crystal readjustment a 376 K. As a consequence, the anhydrous phase should be quite similar to the parent dihydrate one (S-H), and accordingly in both of them two independent ion pairs are still present in the asymmetric unit (results from single crystal XRD for S-H25 and SSNMR for S-H and S-A, respectively). Interestingly, the SSNMR data indicate that in SA the two independent anions mainly differ in the isopropionic fragment as found in the crystal structure (single crystal XRD data, CSD Refcode: ASUBUL)44 of the anhydrous sodium naproxen, whose skeleton resembles that of ibuprofen. In particular, the different orientation of the carboxylate group in the two independent naproxen anions has been ascribed to the interactions between the O-carboxy and the sodium atoms. On this ground, we can postulate for the sodium cations in S-A a coordination sphere similar to that observed in the salt with the (S)-naproxen (Scheme 2a). As for RS-H, Censi et al.23 report that the water removal follows a D3 mechanism, that is, a diffusion of the detached water molecules through hydrophilic channels, followed by the rearrangement of the ibuprofen and sodium ions. In other words, the sodium ibuprofen dehydration occurs via the breakage of the sodium ibuprofen−water bonds and the rearrangement of the anhydrous crystal. However, Censi et al. did not mention the formation of the conglomerate species, which occurs when a complete dehydration is achieved, as in our case. Because of the heterochiral and homochiral composition of RS-H and RS-A, respectively, a quite complex and articulated molecular mechanism must be envisaged to account for this transformation, which implies a solid−solid resolution process, that is, the separation between R and S



CONCLUSIONS In this paper, we presented a combined application of XRD, SSNMR, and calorimetric techniques that allowed us to characterize the different phases, either in racemic or optically pure samples, involved in the transition between dihydrate and anhydrous species of sodium ibuprofen. All the experimental results indicated that this transition is fully reversible. Under all the mentioned conditions, the dehydration of the racemic sample consists of a transition between the dihydrate racemic compound and the anhydrous conglomerate. This transition, clearly observed by XRPD, SSNMR, and DSC, obviously involves a big rearrangement of the whole crystal, following the lack of water molecules, which is quite rare. Giving that in the dihydrate phase the carboxylate hydrophilic tails of ibuprofen anions belonging to chains having opposite chirality are kept together through sodium ions and water molecules, the model proposed for the transition involves the diffusion of the water molecules out of the crystal and the rupture of the O-carboxy− Na+ interactions which results in the collapse of the crystal 2450

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

(8) He, Q.; Zhu, J.; Gomaa, H.; Jennings, M.; Rohani, S. J. Pharm. Sci. 2009, 98, 1835−1844. (9) Giron, D. J. Therm. Anal. Calorim. 2002, 68, 335−357. (10) Geppi, M.; Mollica, G.; Borsacchi, S.; Veracini, C. A. Appl. Spectrosc. Rev. 2008, 43, 202−302. (11) Carignani, E.; Borsacchi, S.; Bradley, J. P.; Brown, S. P.; Geppi, M. J. Phys. Chem. C 2013, 117, 17731−17740. (12) (a) Marković, M.; Milić, D.; Sabolović, J. Cryst. Growth Des. 2012, 12, 4116−4129. (b) Cruz-Cabeza, A. J.; Da, G. M.; Jones, W. Phys. Chem. Chem. Phys. 2011, 13, 12808−12816. (c) Delaney, S. P.; Pan, D.; Galella, M.; Yin, S. X.; Korter, T. M. Cryst. Growth Des. 2012, 12, 5017−5024. (d) Copley, R. C. B.; Barnett, S. A.; Karamertzanis, P. G.; Harris, K. D. M.; Kariuki, B. M.; Xu, M.; Nickels, E. A.; Lancaster, R. W.; Price, S. L. Cryst. Growth Des. 2008, 8, 3474−3481. (13) Rainsford, K. D. Ibuprofen. A Critical Bibliographic Review; Rainsford, K. D., Ed.; Taylor & Francis: London, 1999; pp 1−24. (14) Rainsford, K. D. Ibuprofen. A Critical Bibliographic Review; Rainsford, K. D., Ed.; Taylor & Francis: London, 1999; pp 145−275. (15) Rao, P.; Knaus, E. E. J. Pharm. Sci. 2008, 11, 81−110. (16) Sörgel, F.; Fuhr, U.; Minic, M.; Siegmund, M.; Maares, J.; Jetter, A.; Kinzig-Schippers, M.; Tomalik-Scharte, D.; Szymanski, J.; Goeser, T.; Toex, U.; Scheidel, B.; Lehmacher, W. Int. J. Clin. Pharmacol. Ther. 2005, 43, 140−149. (17) Freer, A. A.; Bunyan, J. N.; Shankland, N.; Sheen, D. B. Acta Crystallogr. 1993, C49, 1378−1380. (18) Shankland, N.; Wilson, C. C.; Florence, A. J.; Cox, P. J. Acta Crystallogr. 1997, C53, 951−954. (19) Hansen, L. Kr.; Perlovich, G. L.; Bauer-Brandl, A. Acta Crystallogr. 2006, E62, e17−e18. (20) Rietveld, I. B.; Barrio, M.; Do, B.; Tamarit, J.-L.; Céolin, R. J. Phys. Chem. B 2012, 116, 5568−5574. (21) Zhang, G. G. Z.; Paspal, S. Y. L.; Suryanarayanan, R.; Grant, D. J. W. J. Pharm. Sci. 2003, 92, 1356−1366. (22) Lee, T.; Chen, Y. H.; Wang, Y. W. Crys. Growth Des. 2008, 8, 415−426. (23) Censi, R.; Martena, V.; Hoti, E.; Malaj, L.; Di Martino, P. J. Therm. Anal. Calorim. 2013, 111, 2009−2018. (24) Jacques, J.; Leclercq, M.; Brienne, M. J. Tetrahedron 1981, 37, 1727−1733. (25) Zhang, Y.; Grant, D. J. W. Acta Crystallogr. 2005, C61, m435− m438. (26) Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.; McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de Streek, J.; Wood, P. A. J. Appl. Crystallogr. 2008, 41, 466−470. (27) CrysAlisPro; Agilent Technologies: Santa Clara, CA, 2011. (28) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. J. Appl. Crystallogr. 1999, 32, 115−119. (29) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112−122. (30) Fung, B. M.; Khitrin, A. K.; Ermolaev, K. J. Magn. Reson. 2000, 142, 97−101. (31) Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.; Swaminathan, S.; Karplus, M. J. Comput. Chem. 1983, 4, 187−217. (32) Accelrys, Inc., San Diego, CA. (33) 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.;

lattice, hence the additional enthalpy for this process with respect to the analogous transition in the optically pure species. This transient disordered state evolves into the anhydrous conglomerate phase, whose crystalline form is obviously completely different from the parent hydrated form (SSNMR and XRPD data) but very similar to that of the enantiomerically pure anhydrous sodium ibuprofen salt (SSNMR and XRPD data). The hypothesis that the transition proceeds through a disordered intermediate state is in agreement with the observation of a less ordered structure of RS-A with respect to RS-H as indicated by the line width of the SSNMR spectrum and the XRPD pattern. A possible interpretation is that the ordering of RS-H is destroyed by the formation of the intermediate phase and not completely recovered by the formation of RS-A. As a side result, the disorder of the isobutyl fragment in RSH, previously observed by single crystal XRD, and recently characterized as “dynamic” (fast interconformational motion) and “static” (distribution of frozen conformations) disorder at high and low temperatures, respectively, by SSNMR,36 has been further investigated by single crystal XRD data analysis and computational techniques. The latter results are in agreement with SSNMR findings and contribute to suggest a model for the racemate−racemic conglomerate solid state transition observed.



ASSOCIATED CONTENT

* Supporting Information S

Tables with crystal data and structure refinement parameters, XRPD patterns. This material is available free of charge via the Internet at http://pubs.acs.org. Crystallographic information files are also available from the Cambridge Crystallographic Data Center (CCDC) upon request (http://www.ccdc.cam.ac. uk, CCDC deposition numbers 983970−983973).



AUTHOR INFORMATION

Corresponding Author

*E-mail: paolapaoli@unifi.it. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Authors thank Dr. Andrea Ienco (ICCOM-CNR, FirenzeItaly), Dr. Lucia Maini, and Dr. Laura Chelazzi (Università di Bologna-Italy) for their help in temperature-resolved XRPD measures. CRIST (Centro di Cristallografia Strutturale, Università di Firenze-Italy) where the single-crystal X-ray analysis as well as all the rt XRPD experiments were carried out is also thanked.



REFERENCES

(1) Li, J. Z.; Zell, M. T.; Munson, E. J.; Grant, D. J. W. J. Pharm. Sci. 1999, 88, 337−346. (2) Yu, L.; Reutzel, S. M.; Stephenson, G. A. Pharm. Sci. Technol. Today 1998, 3, 118−127. (3) Bugay, D. E. Adv. Drug Delivery Rev. 2001, 48, 43−65. (4) Vippagunta, S. R.; Brittain, H. G.; Grant, D. J. W. Adv. Drug Delivery Rev. 2001, 48, 3−26. (5) Stephenson, G. A.; Forbes, R. A.; Reutzel-Edens, S. M. Adv. Drug Delivery Rev. 2001, 48, 67−90. (6) Altamura, M.; Guidi, A.; Jerry, L.; Paoli, P.; Rossi, P. CrystEngComm 2011, 13, 2310−2317. (7) Tishmack, P. A.; Bugay, D. E.; Byrn, S. R. J. Pharm. Sci. 2003, 92, 441−474. 2451

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452

Crystal Growth & Design

Article

Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004. (34) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80, 3265−3269. (35) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem. 1996, 17, 49−56. (36) Concistré, M.; Carignani, E.; Borsacchi, S.; Johannessen, O. G.; Mennucci, B.; Yang, Y.; Geppi, M.; Levitt, M. H. J. Phys. Chem. Lett. 2014, 5, 512−516. (37) Allen, F. H. Acta Crystallogr. 2002, B58, 380−388. (38) Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates and Resolutions; John Wiley & Sons: New York (reprint ed. 1991, reissued 1994 with corrections: Krieger Publishing Co.: Malabar, FL); pp 32, 93−100, 131−146, 228−229. (39) Geppi, M.; Guccione, S.; Mollica, G.; Pignatello, R.; Veracini, C. A. Pharm. Res. 2005, 22, 1544−1555. (40) Carignani, E.; Borsacchi, S.; Geppi, M. J. Phys. Chem. A 2011, 115, 8783−8790. (41) Nasipuri, D. Stereochemistry of Organic Compounds: Principles and Applications; New Academic Science: London, 2012 (42) Harada, J.; Harakawa, M.; Ogawa, K. CrystEngComm 2009, 11, 638−642. (43) Amharar, Y.; Petit, S.; Sanselme, M.; Cartigny, Y.; Petit, M.-N.; Coquerel, G. Crys. Growth Des. 2011, 11, 2453−2462. (44) Kim, Y-s.; VanDerveer, D.; Rousseau, R. W.; Wilkinson, A. P. Acta Crystallogr. 2004, E60, m419−m420. (45) van Eupen, J. Th. H.; Elffrink, W. W. J.; Keltjens, R.; Bennema, P.; de Gelder, R.; Smits, J. M. M.; van Eck, E. R. H.; Kentgens, A. P. M.; Deij, M. A.; Meekes, H. M.; Vlieg, E. Crys. Growth Des. 2008, 8, 71−79. (46) Levkin, P. A.; Schweda, E.; Kolb, H.; et al. Tetrahedron: Asymmetry 2004, 15, 1445−1450. (47) Mercier, N.; Barres, A. L.; Giffard, M.; et al. Angew. Chem. 2006, 118, 2154−2157. (48) Ros, F.; Molina, M. T. Eur. J. Org. Chem. 1999, 11, 3179−3183. (49) He, Q.; Rohani, S.; Zhu, J.; Gomaa, H. Crys. Growth Des. 2010, 10, 5136−5145. (50) Giovannini, J.; Céolin, R.; Perrin, M. A.; Toscani, S.; Louër, D.; Leveiller, F. J. Phys. IV 2001, 11, 93−97. (51) Yoshizawa, K.; Toyota, S.; Toda, F. Chem. Commun. 2004, 1844−1845.

2452

dx.doi.org/10.1021/cg500161e | Cryst. Growth Des. 2014, 14, 2441−2452