Crystalline forms of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide

1A

A 160 L glass-lined vessel was charged with 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide (3.72 kg, 10.32 mol) and methanol (80.8 L). The contents were heated to reflux (65° C.) and authentic 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide Form 1 seeds (109 g, 302.5 mmol) charged as a slurry in methanol (1.9 L). The Form I seeds were obtained via slurry to slurry turnover in refluxing methanol (30 volumes). The mixture was stirred at 275 rpm and aged for 14.5 h. The slurry was cooled from 65° C. to 20° C. over 12 h and then aged at 20° C. for 6 h.

The batch was filtered and the wet cake was washed with methanol (5.4 L). The solid was dried at 40° C. for 23 h in a vacuum oven with a nitrogen bleed. The material was then passed through a Co-Mill to break up any lumps and 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide Form 1 (2.21 kg, 6.14 mol) was obtained as a white powder.

1B

Crude API (150 g) was purified and obtained as a hydrate. The hydrate was then stirred in methanol at reflux with 5% form 1 seed for 14 hours. The slurry was sampled at reflux for XRPD, which indicated a small amount of form 2 was present (<5%). The slurry was cooled to 20° C. and wet milled for 30 minutes at 5200 rpm using an IKA Ultra Turrax wetmill. It was then returned to reflux for another 16 hours.

After sampling at reflux, XRPD indicated form 2 was still present, albeit at a significantly reduced level. So the slurry was cooled to 20° C. and wet milled again for 30 minutes at 5200 rpm. It was then returned to reflux for another 16 hours.

This time XRPD indicated the material was 100% form 1. The slurry was cooled to 20° C. over 12 hours and then aged at 20° C. for 6 hours. The material was reconfirmed as form 1 before filtration and drying. A total of 109 g form 1 API was obtained as an off-white powder (82% recovery corrected for assay).

1C

1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide (52 g) was diluted with N,N-dimethylacetamide (170 mL) and methanol (52 mL). The resulting slurry was stirred at a temperature between 20 and 25° C. for 5 to 10 min at 400 rpm. The mixture was heated to 60° C. When the temperature reached 52° C., the mixture turned clear and became a solution.

A and additional methanol (61 mL) was added to the solution which was further stirred at 60° C. for 5 to10 min. Seeds (260 mg or 0.5% of total api) of Form 1 of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide were added and a slurry immediately formed. Agitation was increased to 500 rpm and methanol (661 mL) was added over 10 h. The obtained slurry was aged at 60° C., and then cooled to 0° C. over 5 h. The slurry was milled for 60 min, aged for 2 hours at 0° C. and then filtered. The isolated product was washed with methanol (2×160 mL) and then dried under vacuum at room temperature for 2 hours followed by drying in a vacuum oven at 50° C. for 16 hours. 46.6 g of Form 1 crystals obtained.

Form 2 was obtained as described in WO 2018/108969.

3A

Form 2 (0.2 g) was slurried in water (10 mL) at RT for 14 days. The suspension was then filtered and vacuum dried at ambient condition (0.18 g isolated). Hydrate formation was confirmed by PXRD.

3B

Form 2 (0.1 g) was slurried in 70/30 (v/v) MeOH/H2O (2 mL) at RT for 1 day. The suspension was then filtered and vacuum dried at ambient condition (X g isolated). Hydrate formation was confirmed by PXRD

Each of these samples of Forms 1 and 2 and the hydrate were characterized as described below:

The X-Ray Powder Diffraction (XRPD)

X-ray powder diffraction studies are widely used to characterize molecular structures, crystallinity, and polymorphism. The X-ray powder diffraction patterns of Form 1 and Form 2 and of the hydrate were generated on Bruker AXS D8 Advance with a LYNXEYE XE-T detector in reflection mode.

Solid State NMR

In addition to the X-ray powder diffraction patterns described above, Form 1, 2 and hydrate samples were further characterized based on their carbon-13 solid-state nuclear magnetic resonance (NMR) spectrum. The carbon-13 spectrum was recorded on a Bruker AV400 NMR spectrometer operating at a carrier frequency of 400.14 MHz, using a Bruker 4 mm H/F/X BB triple resonance CPMAS probe. The spectrum was collected utilizing proton/carbon-13 variable-amplitude cross-polarization (VACP) at 80 kHz, with a contact time of 3 ms. Other experimental parameters used for data acquisition were a proton 90-degree pulse of 100 kHz, SPINAL64 decoupling at 100 kHz, a pulse delay of 30.0 s, and signal averaging for 2900 scans. The magic-angle spinning (MAS) rate was set to 13 kHz. A Lorentzian line broadening of 30 Hz was applied to the spectrum before Fourier Transformation. Chemical shifts are reported on the TMS scale using the carbonyl carbon of glycine (176.70 ppm) as a secondary reference.

Differential Scanning Calorimetry (DSC)

DSC data were acquired using TA Instruments DSC Q2000 or equivalent instrumentation. A sample with a weight between 1 and 6 mg is weighed into an open pan. This pan is placed in the sample position in the calorimeter cell. An empty pan is placed in the reference position. The calorimeter cell is closed and a flow of nitrogen is passed through the cell. The heating program is set to heat the sample at a heating rate of 10° C./min to a temperature of approximately 275° C. When the run is completed, the data are analyzed using the DSC analysis program in the system software. The observed endo- and exotherms are integrated between baseline temperature points that are above and below the temperature range over which the endotherm or exotherm is observed. The data reported are the onset temperature, peak temperature and enthalpy.

Physical Characterization of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide Crystalline Form 1

FIG. 1 shows the X-ray powder diffraction pattern of Form 1. Form 1 exhibited characteristic diffraction peaks corresponding to d-spacings of 12.7, 6.4, and 4.5 angstroms. Form 1 was further characterized by the d-spacings of 5.0, a doublet peak at 4.4 and 4.2 angstroms and another doublet peak at 4.0 and 3.9 angstroms. Form 1 was even further characterized by the d-spacings of a peak set at 3.6, 3.4, and 3.1 angstroms.

Form 1 was characterized based on its solid-state carbon-13 nuclear magnetic resonance (NMR) spectrum (see FIG. 2 ). The carbon-13 spectrum was recorded on a Bruker AV400 NMR spectrometer operating at a carrier frequency of 400.14 MHz, using a Bruker 4 mm H/F/X BB triple resonance CPMAS probe. The spectrum was collected utilizing proton/carbon-13 variable-amplitude cross-polarization (VACP) at 80 kHz, with a contact time of 3 ms. Other experimental parameters used for data acquisition were a proton 90-degree pulse of 100 kHz, SPINAL64 decoupling at 100 kHz, a pulse delay of 30.0 s, and signal averaging for 2900 scans. The magic-angle spinning (MAS) rate was set to 13 kHz. A Lorentzian line broadening of 30 Hz was applied to the spectrum before Fourier Transformation. Chemical shifts are reported on the TMS scale using the carbonyl carbon of glycine (176.70 ppm.) as a secondary reference.

Characteristic carbon-13 isotropic chemical shifts for Form 1 are observed at 30.61, 32.60, 53.4022, 59.59, 67.21, 71.93, 88.50, 96.35, 104.47, 120.39, 121.95, 131.45, 153.31, 161.41, 163.36, and 166.70 ppm.

FIG. 3 is a typical DSC thermogram of the crystalline Form 1. The DSC thermogram is characterized by a melting endotherm with an extrapolated onset temperature of 256.9° C., a peak temperature of 257.2° C. and enthalpy of 127.1 J/g.

Physical Characterization of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide Crystalline Form 2

FIG. 4 shows the X-ray powder diffraction pattern of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide Crystalline Form 2 exhibited characteristic diffraction peaks corresponding to d-spacings of 13.9, 11.2, and 7.0 angstroms. Form 2 was further characterized by the d-spacings of a doublet peak at 6.0 and 5.9 angstroms and another doublet peak at 5.3 and 5.2 angstroms.

Form 2 was characterized based on its solid-state carbon-13 nuclear magnetic resonance (NMR) spectrum (see FIG. 5 ). The carbon-13 spectrum was recorded on a Bruker AV400 NMR spectrometer operating at a carrier frequency of 400.14 MHz, using a Bruker 4 mm H/F/X BB triple resonance CPMAS probe. The spectrum was collected utilizing proton/carbon-13 variable-amplitude cross-polarization (VACP) at 80 kHz, with a contact time of 3 ms. Other experimental parameters used for data acquisition were a proton 90-degree pulse of 100 kHz, SPINAL64 decoupling at 100 kHz, a pulse delay of 17.0 s, and signal averaging for 5916 scans. The magic-angle spinning (MAS) rate was set to 13 kHz. A Lorentzian line broadening of 30 Hz was applied to the spectrum before Fourier Transformation. Chemical shifts are reported on the TMS scale using the carbonyl carbon of glycine (176.70 ppm.) as a secondary reference.

Characteristic carbon-13 isotropic chemical shifts for Form 2 are observed at 29.39, 30.67, 32.06, 52.40, 54.80, 59.92, 65.57, 67.64, 71.01, 72.30 88.52, 92.33, 93.02, 103.96, 121.32, 122.79, 130.85, 152.10, 162.16, 163.05, 164.45, and 168.12 ppm.

FIG. 6 is a typical DSC thermogram of the crystalline Form 2. The DSC thermogram is characterized by a melting endotherm with an extrapolated onset temperature of 251.8° C., a peak temperature of 252.9° C. and enthalpy of 124.6 J/g.

Physical Characterization of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide crystalline hydrate form

FIG. 7 shows the X-ray powder diffraction pattern of 1-[(3R,4S)-4-cyanotetrahydropyran-3-yl]-3-[(2-fluoro-6-methoxy-4-pyridyl)amino]pyrazole-4-carboxamide crystalline hydrate form exhibited characteristic diffraction peaks corresponding to d-spacings of 12.1, a triplet peak at 7.0, 6.9 and 6.7 angstroms and a peak at 5.1 angstroms. The hydrate form was further characterized by the d-spacings of 5.5, 4.1, 3.2 and 3.1 angstroms.

The hydrate was characterized based on its solid-state carbon-13 nuclear magnetic resonance (NMR) spectrum. The carbon-13 spectrum was recorded on a Bruker AV400 NMR spectrometer operating at a carrier frequency of 400.14 MHz, using a Bruker 4 mm H/F/X BB triple resonance CPMAS probe. The spectrum was collected utilizing proton/carbon-13 variable-amplitude cross-polarization (VACP) at 80 kHz, with a contact time of 3 ms. Other experimental parameters used for data acquisition were a proton 90-degree pulse of 100 kHz, SPINAL64 decoupling at 100 kHz, a pulse delay of 1.6 s, and signal averaging for 8000 scans. The magic-angle spinning (MAS) rate was set to 13 kHz. A Lorentzian line broadening of 30 Hz was applied to the spectrum before Fourier Transformation. Chemical shifts are reported on the TMS scale using the carbonyl carbon of glycine (176.70 ppm.) as a secondary reference.

Characteristic carbon-13 isotropic chemical shifts for the hydrate are observed at 29.80, 30.52, 54.35, 60.67, 65.45, 70.62, 89.08, 92.52, 102.17, 121.96, 123.48, 131.05, 152.21, 163.46, 164.53, and 168.40 ppm.

FIG. 9 is a typical DSC curve of the crystalline hydrate form. The DSC curve is characterized by two endotherms and one exotherm. The first endotherm with an extrapolated onset temperature of 27.4° C., a peak temperature of 52.1° C., and enthalpy of 46.6 J/g is due to dehydration. The exotherm with an extrapolated onset temperature of 164.1° C. is due to solid phase transition to Form 2. The endotherm with an extrapolated onset temperature of 253.9° C., a peak temperature of 255.1° C. and enthalpy of 124.6 J/g is due to melting of Form 2.

Relative Thermodynamic Stability of Form 1, Form 2 and Hydrate

Form 1 and Form 2 are monotropically related. Form 1 is more stable than Form 2. Hydrate is the most stable form in aqueous solutions.

Processability Assessment for Form 1 and Form 2

Both Form 1 and Form 2 are physically and chemically stable after 12 months of storage at 40° C./75% relative humidity (RH) and 55° C. There is a risk of Form 2 conversion to hydrate at the typical wet granulation temperature when seeds of hydrate are present. Small scale crystallization process studies demonstrated that Form 1 can be consistently delivered.

Form 1 remained as Form 1 after slurried in MeOH for 3 weeks at room temperature (RT). When ˜10 mg each of Form 1 and Form 2 were slurried in 1 mL of MeOH at 50° C., full conversion to Form 1 was observed after 4 days.

When Form 2 was slurried in 2-Me-THF, Ethyl Acetate, MeOH or MeOH/toluene (70:30 v/v) at 50° C., seeded with Form 1 at 14 days and equilibrated for an additional 14 days (28 days total), full or conversion to Form 1 was observed. Partial conversion was observed after 28 days in Ethyl Acetate and Isopropyl Acetate at 25° C. This is evidence that Form 1 is more thermodynamically stable than Form 2.

The lower solubility of Form 1 than Form 2 also indicated that Form 1 is the more stable form:

Conversion to Hydrate Risk

Slurry experiments in water containing 5% SLS were used to assess the risk of form conversion during wet granulation. Typical wet granulation conditions include drying at 60 to 70° C. for 1.5 to 2 hours.

Slurries containing 20 mg/mL of either Form 1 or Form 2 in 5% sodium lauryl sulfate (SLS) aqueous solution were seeded with 5% hydrate seeds and monitored for form change with PXRD over 24 hours at room temperature and at 80° C. No hydrate peak growth observed for Form 1 samples, indicating that Form 1 is stable for 24 hours in the presence of the hydrate at RT and 80° C. While no hydrate peak growth was observed for the Form 2 samples at 80° C., hydrate peak growth was observed for Form 2 samples at RT. This indicates that Form 2 is not stable in the presence of hydrate seeds at RT. These results indicated that the risk of form conversion to hydrate is lower for Form 1 than for Form 2 when hydrate seeds are present.