2015 年 63 巻 11 号 p. 858-865
Different crystal packing of hydrates from anhydrate crystals leads to different physical properties, such as solubility and stability. Investigation of the potential of varied hydrate formation, and understanding the stability in an anhydrous/hydrate system, are crucial to prevent an undesired transition during the manufacturing process and storage. Only one anhydrous form of T-3256336, a novel inhibitor of apoptosis (IAP) protein antagonist, was discovered during synthesis, and no hydrate form has been identified. In this study, we conducted hydrate screening such as dynamic water vapor sorption/desorption (DVS), and the slurry experiment, and characterized the solid-state properties of anhydrous/hydrate forms to determine the most desirable crystalline form for development. New hydrate forms, both mono-hydrate and hemi-hydrate forms, were discovered as a result of this hydrate screening. The characterization of two new hydrate forms was conducted, and the anhydrous form was determined to be the most desirable development form of T-3256336 in terms of solid-state stability. In addition, the stability of the anhydrous form was investigated using the water content and temperature controlled slurry experiment to obtain the desirable crystal form in the crystallization process. The water content regions of the stable phase of the desired form, the anhydrous form, were identified for the cooling crystallization process.
Approximately one-third of pharmaceutical solids have been reported to crystallize in the hydrate forms.1,2) The solid-state characterization of hydrate is essential for pharmaceutical development because the hydrate form has different physicochemical properties such as solubility and stability from its anhydrous form.3–5) Therefore, it is important to discover the possible hydrate forms in early drug development and understand the anhydrous/hydrate phase transformations to ensure undesired conversions do not occur during the development process.6)
There are several approaches for investigation of the anhydrous/hydrate phase transformation, such as dynamic water vapor sorption/desorption isotherm (DVS), storage under high humidity, and slurry experiments in water-content-controlled organic solvent mixtures.7–10) Cui and Yao explained that slurry experiments using water and organic solvent mixtures under water-content-controlled-conditions were rapid and reliable hydrate screening approaches.1) Some applications of slurry experiments with water and organic solvent mixtures to investigate the stability in the anhydrous/hydrate system of theophylline and ampicillin, carbamazepine were also reported.11–13) Water content is related to water activity, defined by the activity coefficient and mole fraction of water.
Hydrate formation is represented by the following equilibrium (Eq. 1), previously described by Grant and Higuchi3,14):
 | (1) |
 | (2) |
T-3256336 is a highly potent and selective inhibitor of apoptosis (IAP) protein antagonist15) and only the anhydrous form was obtained from the synthetic compounds by the medicinal chemists (Fig. 1). In this study, we conducted the hydrate screening such as DVS, the slurry experiment for the identification of the hydrate forms and the characterization of the solid state properties of the anhydrous/hydrate forms. New hydrate forms, the mono-hydrate and hemi-hydrate forms, were discovered by the slurry experiment and the solid state properties of new hydrate forms were characterized by powder X-ray diffractometry (PXRD), thermogravimetry (TG), differential scanning calorimetry (DSC), single crystal X-ray structure analysis and Raman spectroscopy. We determined the anhydrous form as the desired form for the development from the perspective of the stability and the physicochemical properties. In addition, the stability dependence of the anhydrous form on the water content and temperature was investigated using the water-content and temperature controlled slurry experiment in order to obtain the anhydrous form in the crystallization process.
The anhydrous form of T-3256336 was obtained from the Medicinal Chemistry Research Laboratories at Takeda Pharmaceutical Co., Ltd. Salts and solvents were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan).
MethodsDVS MeasurementThe hygroscopicity of the anhydrous form was determined using an Automated Water Sorption Analyzer (DVS-1, Surface Measurement Systems Ltd., London). The sample (10 mg) was exposed to 0 to 95% relative humidity (RH) at 25°C with step changes in 5% RH increments.
Slurry ExperimentThe slurry experiment for hydrate screening was performed in acetonitrile (MeCN)–water (1 : 1, v/v) and methyl tert-butyl ether (MTBE) saturated with water. MeCN and MTBE were used for hydrate screening because T-3256336 forms solvates in other organic solvents except for MeCN and MTBE. The suspensions of T-3256336 in these solvents were prepared at the concentration of 100 mg/mL. After 1 week and 2 weeks slurry equilibration at room temperature, each suspension was filtrated and the obtained solids were characterized by Raman microscopy.
To investigate the stability dependence on water activity for the anhydrous form at various temperatures, the slurry experiment controlling the water-content and the temperature was conducted by following the procedure. Excess of the anhydrous form was added to a mixture of water and acetone with a wide range of water content values. The suspension was equilibrated at various temperatures (5, 25, 40°C) for 3 d to create saturated solutions. After centrifugation, the supernatant was seeded with the anhydrous, hemi-hydrate, and mono-hydrate forms to enhance the nucleation process and ensure an excess of solid for each form. The suspension was then stirred at three different temperatures (5, 25, 40°C) for 7 d. The solid phase was also sampled and analyzed with Raman spectroscopy.
Exposure of Freeze Dried T-3256336 to Various Relative HumiditiesApproximately 0.9 mg of each crystalline powder of the anhydrous, hemi-hydrate and mono-hydrate forms were added to 90 mg of freeze dried T-3256336 as seed crystals. They were stored in the controlled relative humidity (0, 33, 53, 75, 93% RH) chamber at 25°C for 91 d. The control sample was stored in a vessel in which water was displaced by vapor diffusion of MTBE. The relative humidities used in this study were prepared by using saturated solutions of various salts in water.16)
PXRDPXRD patterns were collected using a RINT UltimaIV powder X-ray diffractometer (Rigaku Corp., Tokyo, Japan) with CuKα radiation, generated at 40 kV and 50 mA. The drug substances were placed on a silicone plate at room temperature. The data was collected from 2 to 35° (2θ) at a step size of 0.02° and scanning speed of 6°/min.
Thermal AnalysisThermogravimetry was performed using a TGA/DSC1 system (Mettler-Toledo International Inc., Greifensee, Switzerland). A TG curve was obtained in an open aluminum pan using ca. 0.60 mg of the drug substance and a heating rate of 5°C/min under nitrogen gas flow at 40 mL/min. The differential scanning calorimetry was measured using a DSC1 system (Mettler-Toledo International Inc.). A DSC thermogram was obtained in a crimped aluminum pan using ca. 0.25 mg of drug substance and a heating rate of 5°C/min under nitrogen flow at 20 mL/min.
Karl Fischer TitrationThe water contents of the hydrate forms were observed with the Karl Fischer AQ-7 (Hiranuma Sangyo, Ibaraki, Japan). Hydranal Aqualyte RS and Aqualyte CN were used as a generator electrolyte and a counter-electrolyte, respectively. Karl Fischer titration was performed under 32 to 34% RH at 30°C.
Raman SpectroscopyRaman spectra were collected with RXN systems (Kaiser Optical Systems, Ann Arbor, MI, U.S.A.) at room temperature, utilizing a laser wavelength of 785 nm and laser power of 400 mW as an excitation source and an air-cooled CCD detector. A 20-fold objective lens with probe system was used to collect the spectra. The spectra were acquired with 4 cm−1 spectral width and 10 s exposure.
Single Crystal X-Ray Structure AnalysisSingle crystal X-ray diffraction data was collected on a Rigaku R-AXIS RAPID (Rigaku, Tokyo, Japan) with CuKα radiation at −173°C. The N–H and O–H hydrogen atoms were placed on the difference Fourier map and refined isotropically. The non-hydrogen atoms were refined anisotropically, and other hydrogen atoms were placed in ideal position and refined using a riding model. The crystal structure was solved using the Crystal Structure crystallographic software package (Rigaku) and refined by SHELXL-97.17)
We conducted DVS measurement in order to discover the hydrate form because DVS is known as an easy and fast hydrate screening methodology.1,10) The hygroscopicity of the anhydrous form given in Fig. 2 showed a 1.1% weight increase from 5 to 95% RH. No change in the PXRD pattern of the anhydrous form after DVS measurement indicated no hydrate formation occurred during DVS.
The slurry experiment was reported to be a rapid and efficient hydrate formation screening.1,18) The water activity plays an important role in the slurry experiment for hydrate formation. We performed the slurry experiment using two solvents with different water activities, MeCN–water (1 : 1, v/v) and MTBE saturated with water as solvents in the slurry experiment to investigate the hydrate form. The water activity of MeCN–water (1 : 1, v/v) and MTBE saturated with water was higher than 0.9 and closely 0, respectively.
Two crystal forms showing the different Raman spectra from the anhydrous form were obtained after 2 weeks of the slurry experiment (Fig. 3). In the Raman spectra, the anhydrous form displayed a sharp band at 1646 and 1656 cm−1, the crystal form obtained from MeCN–water (1 : 1, v/v) at 1652 cm−1, and the crystal form obtained from MTBE saturated with water at 1672 cm−1. The high-intensity peak patterns of Raman spectra in the shift region of 1650 to 1800 cm−1, attributed to C=O stretching vibrations, were completely different for each crystal form because of the various hydrogen bond states. These results indicated that two crystals obtained from the slurry experiment could be new hydrate forms.
The solid state characterization of the anhydrous form and two hydrates obtained from the slurry experiment was conducted using PXRD, TG, DSC and single crystal X-ray diffraction.
The PXRD patterns of the anhydrous form and the obtained two crystals are presented in Fig. 4. Each crystal form had its own characteristic patterns, which allows for easy and clear identification of the crystalline solid form.
The thermal behavior of each crystal was studied by TG/DSC, as illustrated in Fig. 5. The anhydrous form showed a melting peak at 177°C, without weight loss upon heating. The DSC thermograms of the two hydrate forms showed a broad endothermic peak over a wide range, starting at room temperature, because of the dehydration process. The weight loss of the crystal form obtained from MeCN–water (1 : 1, v/v) and MTBE saturated with water corresponding to the DSC endotherm around room temperature was 2.6 and 1.4%, respectively. In addition, the water content of crystal form obtained from MeCN–water and MTBE saturated with water determined by Karl Fischer titration was 2.8 and 1.4%, respectively. These values were almost consistent with the stoichiometric values calculated for the mono-hydrate and hemi-hydrate forms of T-3256336 (2.9 and 1.5%, respectively). Therefore, the crystal form obtained from MeCN–water (1 : 1, v/v) and MTBE saturated with water were found to be the mono-hydrate and hemi-hydrate forms, respectively.
Single crystals of the anhydrous and hemi-hydrate forms were grown using slow-evaporation in MTBE and a mixture of MTBE and heptane, respectively. The crystallographic data of these crystals are summarized in Table 1.
| Anhydrous form | Hemi-hydrate form | |
|---|---|---|
| Empirical formula | C31H45F2N5O5 | C31H45F2N5O5·1/2H2O |
| Formula weight | 605.2 | 614.2 |
| Temperature (°C) | −173 | −173 |
| Crystal system | Orthorhombic | Tetragonal |
| Space group | P21212 | P43212 |
| a (Å) | 4.812 | 10.816 |
| b (Å) | 20.258 | 10.816 |
| c (Å) | 31.881 | 54.702 |
| α (°) | 90 | 90 |
| β (°) | 90 | 90 |
| γ (°) | 90 | 90 |
| Volume (Å3) | 3108 | 6399.6 |
| Z | 4 | 8 |
| Calculated density (g/cm3) | 1.294 | 1.276 |
| R1 | 0.066 | 0.075 |
| wR2 (all data) | 0.172 | 0.197 |
The space group of the anhydrous form was orthorhombic P21212 with four molecules in the asymmetric unit. Molecules in the anhydrous form formed one-dimensional chains along the axis through N–H…O hydrogen bonds (Fig. 6a). The hemi-hydrate form crystallizes in the tetragonal system as the space group P43212 with unit-cell parameters a=b=10.8, and c=54.7 Å. Water molecules form N–H…O and O–H…O bonds in the hydrate form, interacting with the N–H of carboxamide and O of the ethoxy group of T-3256336 (Fig. 6b). There was a half water molecule per T-3256336 molecule in the unit cell, which was in good agreement with the weight loss of the hemi-hydrate form seen in the DSC endotherm.
Roy et al. reported that the anhydrous form, whose density and packing efficiency are smaller than those of the hydrate form, immediately transformed to hydrate forms at room temperature.19) The densities of T-3256336 anhydrous and hemi-hydrate forms were almost same. And the anhydrous form of T-3256336 did not transform to hydrates and was stable in solid state at room temperature. These results were not contradicting the report by Roy et al.
Stability Dependence on Water-Activity for Anhydrous and Hydrate FormsThe dehydration of the mono-hydrate and hemi-hydrate forms was observed around 60°C (Fig. 5), which indicated the possibility of dehydration in the manufacturing process and the storage conditions. Therefore, we determined the anhydrous form as the desirable development form of T-3256336 in terms of the solid state stability. To ensure the undesired hydration of the anhydrous form do not occur during the manufacturing process and the storage, we investigated the stable region of the anhydrous form using the slurry experiment with different water content at 25°C.
A sharp band at 1646 and 1656 cm−1 shifted to 1652 cm−1, which was a characteristic peak of the mono-hydrate form, with water content above 8.0%v/v. When water content below 7.0%v/v at 25°C, the T-3256336 anhydrous form was more stable compared to the hydrates (Table 2).
| Water content (%v/v) | 5°C | 25°C | 40°C |
|---|---|---|---|
| Crystal form | |||
| 0.00 | Anhydrous form | Anhydrous form | N.T. |
| 0.08 | Anhydrous form | Anhydrous form | N.T. |
| 0.20 | Anhydrous form | Anhydrous form | N.T. |
| 0.45 | Anhydrous form | Anhydrous form | N.T. |
| 0.70 | Anhydrous form | Anhydrous form | N.T. |
| 1.0 | Anhydrous form | Anhydrous form | N.T. |
| 1.8 | Hemi-hydrate form | Anhydrous form | N.T. |
| 3.0 | Hemi-hydrate form | Anhydrous form | N.T. |
| 4.0 | Hemi-hydrate form | Anhydrous form | N.T. |
| 6.0 | Mono-hydrate form | Anhydrous form | N.T. |
| 7.0 | N.T. | Anhydrous form | N.T. |
| 7.5 | N.T. | Hemi-hydrate form | N.T. |
| 8.0 | N.T. | Mono-hydrate form | Anhydrous form |
| 10 | N.T. | Mono-hydrate form | Anhydrous form |
| 14 | N.T. | Mono-hydrate form | Mono-hydrate form |
| 17 | N.T. | Mono-hydrate form | N.T. |
| 38 | N.T. | Mono-hydrate form | N.T. |
| 50 | N.T. | Mono-hydrate form | N.T. |
| 100 | N.T. | Mono-hydrate form | N.T. |
N.T.: Not tested.
A sharp band at 1646 and 1656 cm−1, which were characteristic peak of the anhydrous form, shifted to 1672 cm−1, which was characteristic peak of hemi-hydrate at a water content of 7.5%. These results indicated that the anhydrous form was changed to the hemi-hydrate form at water content from 7.0 to 7.5%v/v and the transformation from the hemi-hydrate form to the mono-hydrate form was occurred at water content above 8.0%v/v.
Described as Eq. 2, the hydration state depends on water activity in the surrounding medium such as the crystallization solution and the vapor pressure. The equilibrium water activity of the anhydrous/hydrate system in a slurry experiment has a direct correlation with the equilibrium relative humidity in solid state at the same temperature, because water activity is the ratio of water vapor pressure in a hydrate system to the water vapor pressure of pure water.7) Water content (8.0%v/v) for the anhydrous to the mono-hydrate form transformation at 25°C corresponds to a water activity of 0.74,20,21) suggesting that solid state transformation of the anhydrous to the mono-hydrate form could occur at more than 74% RH.
To study the relationship of T-3256336 between the equilibrium water activity in a slurry experiment and the equilibrium relative humidity in solid state at the same temperature, we stored the freeze dried T-3256336 at various relative humidities (Fig. 7). The anhydrous form was obtained in a vessel where the water is displaced by vapor diffusion of methyl tert-butyl ether at 25°C for 91 d. The mono-hydrate form obtained from MeCN–water (1 : 1, v/v) in the slurry experiment was completely crystallized at high humidity (93% RH) for 91 d, and a small amount of the mono-hydrate form was detected at 75% RH (Figs. 8, 9). Crystallization of the hemi-hydrate form obtained from MTBE saturated with water in the slurry experiment was not observed at any relative humidity.
The mono-hydrate formation was observed for 7 d by exposure of the freeze-dried sample to 95% RH, and a very small amount of the mono-hydrate form was detected at 75% RH for 91 d. The hydration at 75% RH was much slower compared to that at 95% RH, and the mono-hydrate formation was not complete even after storage at 75% RH for 91 d. Although T-3256336 was freeze dried to increase surface area and reduce crystallinity, the equilibrium water activity was not consistent with the equilibrium relative humidity. The activation energy for the solid state transformation is generally higher compared to the solvent mediated transformation because the water molecules migration through the crystal lattice and the change in the crystal structure are required in the solid state transformation. The different activation energy between the solid state transformation and the solvent mediated transformation causes the inconsistency between the equilibrium water activity and the equilibrium relative humidity.
The freeze dried amorphous compound did not completely convert to the mono-hydrate form at equilibrium relative humidity, which indicated that it is difficult to suggest a correlation of the stability in an anhydrous/hydrate system between a slurry experiment and solid state for all compounds, even if the conversion rate was enhanced by increasing the surface area and reducing the crystallinity.18)
Effect of Temperature on Anhydrous/Hydrate Transition PhaseCrystallization from solution by cooling is one of the most common crystallization approaches in addition to anti-solvent crystallization. The crystal phase depends on the temperature and pressure; the effect of temperature on the crystal phase must be understood to use cooling crystallization. In order to investigate the effect of temperature on the stable phase of the anhydrous and hydrate forms, slurry experiments under different temperatures (5, 40°C) were conducted.
The transition of the anhydrous to the hemi-hydrate form and from the hemi-hydrate form to the mono-hydrate form occurred from 1.0 to 1.8%v/v and from 4.0 to 6.0%v/v water content, respectively, at 5°C (Table 2). The anhydrous form transformed to the mono-hydrate form in the slurry experiments with water content of 10 to 14%v/v at 40°C (Table 2).
Higher temperature leads to increasing water content of the anhydrous/hydrate transition phase. Hydration occurs via a hydrogen bond between the drug and water molecule, and hydrogen bond strength depends on temperature. A higher temperature leads to weaker hydrogen bond strength and, thus, water content for the hydration is increased with increasing temperature.6,20) The phase diagram of equilibrium water content and temperature in the anhydrous form, the hemi-hydrate and mono-hydrate forms indicated that the stable phase of the mono-hydrate form was estimated to diminish at temperature over 25°C.
These results demonstrated that water content should be kept below 1.0%v/v for cooling crystallization of the anhydrous form, which provides useful information for the manufacturing process.
DVS and the slurry experiment were conducted for the hydrate screening of T-3256336. We succeeded to discover new hydrate forms of T-3256336, the mono-hydrate and hemi-hydrate forms, by the slurry experiments with water and organic solvent mixtures. The mono-hydrate and hemi-hydrate forms were fully characterized by PXRD, TG, DSC, Raman spectroscopy and single-crystal structure analysis and the anhydrous form was determined as the desirable development form due to a low dehydration temperature of the hydrate forms. The investigation of the influence of water content and temperature on the stability of the anhydrous form using the water-content-controlled slurry experiments revealed the crystallization condition to obtain the anhydrous form in the manufacturing process.
We would like to thank Motoo Iida of Medicinal Chemistry Research Laboratories for the single crystal X-ray diffraction data analysis. We are also grateful to Satoru Asahi of Drug Metabolism and Pharmacokinetics Research Laboratories and Glenn Steven of Analytical Development Laboratories for giving us an opportunity of this research. Thanks are given to the colleagues of Material science & Physicochemical profiling at Analytical Development Laboratories of Takeda Pharmaceutical Company Limited for their helpful discussions.
The authors declare no conflict of interest.
