Crystalline forms of a purine derivative

Abstract

The present invention relates to new crystalline forms of a purine derivative which exhibits excellent anti-tumor activity. The invention also relates to a pharmaceutical composition containing said crystalline forms as an active ingredient, and use thereof in the prevention or treatment of disease. The invention further relates to a process for preparing the crystalline forms.

Claims

1. A crystalline form of compound (I), ##STR00003## wherein said compound is in the form of the free base and which is characterized by an x-ray powder diffraction pattern having two or more diffraction peaks at 2[theta] values selected from 7.530.2, 9.600.2, 10.220.2, 11.290.2, 11.660.2, 12.260.2, 12.620.2, 13.170.2, 14.060.2, 14.850.2, 15.150.2, 15.570.2, 16.990.2, 17.680.2, 18.300.2, 18.390.2, 18.630.2, 18.970.2, 19.320.2 and 20.200.2.

2. The crystalline form of claim 1, which is further characterized by a differential scanning calorimetry trace recorded at a heating rate of 20 C. per minute which shows a maximum endothermic peak at a temperature between about 130 C. and about 140 C., or a differential scanning calorimetry trace substantially in accordance with that shown in FIG. 6.

3. A pharmaceutical composition comprising the crystalline form of claim 1 and a pharmaceutically acceptable carrier, diluent or excipient.

4. A method for the treatment of a proliferative disorder, wherein the proliferative disorder is cancer, said method comprising administering a pharmacologically effective amount of a crystalline form according to claim 1 to a subject in need of thereof.

5. A process for preparing crystalline free base of compound (I), ##STR00004## wherein said compound is in the form of the free base and which is characterized by an x-ray powder diffraction pattern having two or more diffraction peaks at 2[theta] values selected from 7.530.2, 9.600.2, 10.220.2, 11.290.2, 11.660.2, 12.260.2, 12.620.2, 13.170.2, 14.060.2, 14.850.2, 15.150.2, 15.570.2, 16.990.2, 17.680.2, 18.300.2, 18.390.2, 18.630.2, 18.970.2, 19.320.2 and 20.200.2, wherein said method comprises the steps of crystallising compound (I) in free base form from methyl t-butyl ether (MTBE).

Description

(1) The present invention is further described with reference to the following figures, wherein:

(2) FIG. 1 shows the X-ray powder diffraction pattern of Form E of the L-tartrate salt of compound (I), as obtained by Example 5.2. The diffraction pattern was obtained by irradiation of the crystalline product using a PANalytical diffractometer.

(3) FIG. 2 is a DSC thermogram of Form E of the L-tartrate salt of compound (I) obtained using a PerkinElmer DSC4000 at a heating rate of 20 C..Math.min.sup.1. Peak max is observed at 182.40 C.

(4) FIG. 3 shows the X-ray powder diffraction pattern of Form D of the L-tartrate salt of compound (I) as obtained by Example 4. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS C2 GADDS.

(5) FIG. 4 is a DSC thermogram of Form D of the L-tartrate salt of compound (I) as obtained by Example 4, using Mettler DSC823e at a heating rate of 20 C..Math.min.sup.1. Peak max is observed at 151.83 C.

(6) FIG. 5 shows the X-ray powder diffraction pattern of crystalline free base (Form A) of compound (I) as obtained by Example 1. The diffraction pattern was obtained by irradiation of the crystalline product using a PANalytical diffractometer.

(7) FIG. 6 shows a DSC thermogram (lower trace) of crystalline free base (Form A) of compound (I) as obtained by Example 1, using a PerkinElmer Pyris 6 at a heating rate of 20 C..Math.min.sup.1 (peak max is observed at 137.63 C.). The upper trace shows TGA for the same salt measured using a Pyris 1 TGA at a heating rate of 20 C..Math.min.sup.1. No significant mass loss was observed up to 250 C., onset observed at 369.39 C.

(8) FIG. 7 shows the X-ray powder diffraction pattern of the citrate salt (Form F) of compound (I) as obtained by Example 2. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS C2 GADDS.

(9) FIG. 8 shows a DSC thermogram (lower trace) of the citrate salt (Form F) of compound (I) as obtained by Example 2, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 150.83 C. and 187.83). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(10) FIG. 9 shows the X-ray powder diffraction pattern of the benzenesulfonic acid salt (Form G) of compound (I) as obtained by Example 3. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS C2 GADDS.

(11) FIG. 10 shows a DSC thermogram (lower trace) of the benzenesulfonic acid salt (Form G) of compound (I) as obtained by Example 3, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 153.00). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(12) FIG. 11 shows the X-ray powder diffraction pattern of the phosphate salt (Form B) of compound (I) as obtained by Example 7. The diffraction pattern was obtained by irradiation of the crystalline product Bruker AXS C2 GADDS.

(13) FIG. 12 is a DSC thermogram (lower trace) of Form 1 the phosphate salt (Form B) of compound (I) as obtained by Example 7, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 84.95 and 123.93). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e with a heating rate of 10 C..Math.min.sup.1.

(14) FIG. 13 shows the X-ray powder diffraction pattern of the phosphate salt (Form C) of compound (I) as obtained by Example 6. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS C2 GADDS.

(15) FIG. 14 is a DSC thermogram of Form 2 of the phosphate salt (Form C) of compound (I) as obtained by Example 6, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 86.50 and 132.17). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e with a heating rate of 10 C..Math.min.sup.1.

(16) FIG. 15 shows an isotherm plot for Form A (weight change % vs RH %) as measured using a Hiden Isochema moisture sorption analyser.

(17) FIG. 16 shows a kinetic plot of weight change (%) of Form A vs time and relative humidity %.

(18) FIG. 17 shows XRPD analysis of Form A post GVS.

(19) FIG. 18 shows XRPD analysis of Form A pre and post solubility study.

(20) FIG. 19 shows the X-ray powder diffraction pattern of the hydrochloride salt (Form H) of compound (I) as obtained by Example 8. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(21) FIG. 20 shows the X-ray powder diffraction pattern of the hydrochloride salt (Form I) of compound (I) as obtained by Example 9. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(22) FIG. 21 shows the X-ray powder diffraction pattern of the hydrobromide salt (Form J) of compound (I) as obtained by Example 10. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(23) FIG. 22 shows the X-ray powder diffraction pattern of the mesylate salt (Form L) of compound (I) as obtained by Example 12. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(24) FIG. 23 shows the X-ray powder diffraction pattern of the maleate salt (Form M) of compound (I) as obtained by Example 13. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(25) FIG. 24 shows the X-ray powder diffraction pattern of the gentisate salt (Form O) of compound (I) as obtained by Example 15. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(26) FIG. 25 shows the X-ray powder diffraction pattern of the fumarate salt (Form P) of compound (I) as obtained by Example 16. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(27) FIG. 26 shows the X-ray powder diffraction pattern of the L-malate salt (Form Q) of compound (I) as obtained by Example 17. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(28) FIG. 27 shows the X-ray powder diffraction pattern of the L-malate salt (Form R) of compound (I) as obtained by Example 18. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(29) FIG. 28 shows the X-ray powder diffraction pattern of the bromide salt (Form K) of compound (I) as obtained by Example 11. The diffraction pattern was obtained by irradiation of the crystalline product using Bruker AXS D8 Advance.

(30) FIG. 29 shows a DSC thermogram (lower trace) of the HBr salt Form K of compound (I) as obtained by Example 11, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 74 and 95 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(31) FIG. 30 shows a DSC thermogram (lower trace) of the mesylate salt Form L of compound (I) as obtained by Example 12, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 132 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(32) FIG. 31 shows a DSC thermogram (lower trace) of the maleate salt Form M of compound (I) as obtained by Example 13, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 123 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(33) FIG. 32 shows a DSC thermogram (lower trace) of the gentisate salt Form O of compound (I) as obtained by Example 15, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 99 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(34) FIG. 33 shows a DSC thermogram (lower trace) of the fumarate salt Form P of compound (I) as obtained by Example 16, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 143 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(35) FIG. 34 shows a DSC thermogram (lower trace) of the L-malate salt Form Q of compound (I) as obtained by Example 17, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 92 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(36) FIG. 35 shows a DSC thermogram (lower trace) of the L-malate salt Form R of compound (I) as obtained by Example 18, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 114 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(37) FIG. 36 shows a DSC thermogram (lower trace) of the HCl salt Form H of compound (I) as obtained by Example 8, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 68, 87, 104 and 148 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(38) FIG. 37 shows a DSC thermogram (lower trace) of the HCl salt Form I of compound (I) as obtained by Example 9, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 120 and 139 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(39) FIG. 38 shows a DSC thermogram (lower trace) of the HBr salt Form J of compound (I) as obtained by Example 10, using Mettler DSC823e at a heating rate of 10 C..Math.min.sup.1 (peak max observed at 142 C.). The upper trace shows TGA analysis for the same salt using a Mettler TGA/SDTA 851e at a heating rate of 10 C..Math.min.sup.1.

(40) The present invention is further described with reference to the following non-limiting Examples.

EXAMPLES

Instrument and Methodology Details

(41) X-Ray Powder Diffraction (XRPD)

(42) All XRPD patterns referred to herein are obtained using copper K-alpha radiation. As used herein, XRPD values as described in the accompanying specification, figures or tables refer to approximate values. Where the reference is to XRPD values listed in the tables, this refers to the 2-theta values, independent of any other parameters listed in the tables, such as peak intensity or the like.

(43) XRPD on the phosphate (Forms B, C), citrate (Form F), benzenesulfonic acid (Form G) and L-tartrate (Form D) salts were carried out using a Bruker AXS C2 GADDS diffractometer as described below.

(44) XRPD on the hydrochloride (Forms H, I), hydrobromide (Forms J, K), mesylate (Form L), maleate (Form M), gentisate (Form O), fumarate (Form P), and L-malate (Forms, Q, R) salts were carried out using a Bruker AXS D8 Advance diffractometer as described below. XRPD on free base (Form A) compound (I) and the L-tartrate salt (Form E) was carried out using PANalytical diffractometer as described below.

(45) Bruker AXS C2 GADDS

(46) X-Ray Powder Diffraction patterns were collected on a Bruker AXS C2 GADDS diffractometer using Cu K radiation (40 kV, 40 mA), automated XYZ stage, laser video microscope for auto-sample positioning and a HiStar 2-dimensional area detector. X-ray optics consists of a single Gbel multilayer mirror coupled with a pinhole collimator of 0.3 mm. The beam divergence, i.e. the effective size of the X-ray beam on the sample, was approximately 4 mm. A - continuous scan mode was employed with a sample detector distance of 20 cm which gives an effective 2 range of 3.2-29.7. Typically the sample would be exposed to the X-ray beam for 120 seconds. The software used for data collection was GADDS for WNT 4.1.16 and the data were analysed and presented using Diffrac Plus EVA v 9.0.0.2 or v 13.0.0.2.

(47) Bruker AXS D8 Advance

(48) X-Ray Powder Diffraction patterns were collected on a Bruker D8 diffractometer using Cu Ka radiation (40 kV, 40 mA), -2 goniometer, and divergence of V4 and receiving slits, a Ge monochromator and a Lynxeye detector. The instrument is performance checked using a certified Corundum standard (NIST 1976). The software used for data collection was Diffrac Plus XRD Commander v2.5.0 and the data were analysed and presented using Diffrac Plus EVA v 11.0.0.2 or v 13.0.0.2. Unless otherwise stated, XRPD patterns collected on this instrument were used to produce the XRPD peak lists.

(49) Samples were run under ambient conditions as flat plate specimens using powder as received. Approximately 10 mg of the sample was gently packed into a cavity cut into polished, zero-background (510) silicon wafer. The sample was rotated in its own plane during analysis. The details of the data collection are: Angular range: 2 to 42 2 Step size: 0.05 2 Collection time: 0.5 s.step-1
Ambient Conditions

(50) Samples run under ambient conditions were prepared as flat plate specimens using powder as received without grinding. Approximately 1-2 mg of the sample was lightly pressed on a glass slide to obtain a flat surface.

(51) Non-Ambient Conditions

(52) Samples run under non-ambient conditions were mounted on a silicon wafer with heat conducting compound. The sample was then heated to the appropriate temperature at ca. 10 C..Math.min.sup.1 and subsequently held isothermally for ca 1 minute before data collection was initiated.

(53) PANalytical X'Pert PRO

(54) X-Ray Powder Diffraction patterns were collected on a PANalytical diffractometer using Cu K radiation (45 kV, 40 mA), goniometer, focusing mirror, divergence slit (), soller slits at both incident and divergent beam (4 mm) and a PIXcel detector. The software used for data collection was X'Pert Data Collector, version 2.2f and the data was presented using X'Pert Data Viewer, version 1.2d.

(55) Samples were run under ambient conditions and analysed by transmission foil XRPD, using the powder sample as received. Approximately 2-5 mg of the sample was mounted on a 96 position sample plate supported on a polyimide (Kapton, 12.7 m thickness) film. Data was collected in the range 3-40 2 with a continuous scan (speed of 0.146/s). Samples were oscillated2 mm in the x plane at a speed of 2 mm.Math.s.sup.1 throughout data collection to maximise particle sampling and minimise preferred orientation effects.

(56) Nuclear Magnetic Resonance (NMR)

(57) .sup.1H NMR spectra were collected on a Bruker 400 MHz instrument equipped with an autosampler and controlled by a DRX400 console. Automated experiments were acquired using ICONNMR v4.0.4 (build 1) running with Topspin v 1.3 (patch level 8) using the standard Bruker loaded experiments. For non-routine spectroscopy, data were acquired through the use of Topspin alone. Samples were prepared in d6-DMSO, unless otherwise stated. Off-line analysis was carried out using ACD SpecManager v 9.09 (build 7703).

(58) Differential Scanning Calorimetry (DSC)

(59) DSC studies on the phosphate (Forms B, C), citrate (Form F), benzenesulfonate (Form G), L-tartrate (Form D), hydrochloride (Forms H and I), hydrobromide (Form J and K), mesylate (Form L), maleate (Form M), gentisate (Form O), fumarate (Form P), L-malate (Form Q and R) salts were carried out using a Mettler DSC 823e as described below.

(60) DSC studies on free base (Form A) compound (I) were carried out using PerkinElmer Pyris 6 DSC described below.

(61) DSC studies on the L-tartrate salt (Form E) were carried out using PerkinElmer DSC 4000 DSC described below.

(62) Mettler DSC 823e

(63) DSC data were collected on a Mettler DSC 823e equipped with a 50 position autosampler. The instrument was calibrated for energy and temperature using certified indium. Typically 0.5-3 mg of each sample, in a pin-holed aluminium pan, was heated at 10 C..Math.min.sup.1 from 25 C. to 350 C. A nitrogen purge at 50 ml.Math.min.sup.1 was maintained over the sample. The instrument control and data analysis software was STARe v9.10.

(64) PerkinElmer Pyris 6 DSC/DSC 4000

(65) DSC data was collected on a PerkinElmer Pyris 6 DSC or DSC 4000. The instrument was verified for energy and temperature calibration using certified indium. A predefined amount of sample (in mg) was placed in a pin holed aluminium pan and heated at 20 C..Math.min.sup.1 from 30 C. to 320 C. The instrument control and data analysis was Pyris Software v9.0.1.0174.

(66) Thermo-Gravimetric Analysis (TGA)

(67) Mettler TGA/SDTA 851e

(68) TGA data were collected on a Mettler TGA/SDTA 851e equipped with a 34 position autosampler. The instrument was temperature calibrated using certified indium. Typically 5-30 mg of each sample was loaded onto a pre-weighed aluminium crucible and was heated at 10 C..Math.min.sup.1 from ambient temperature to 350 C. A nitrogen purge at 50 ml.Math.min.sup.1 was maintained over the sample. The instrument control and data analysis software was STARe v9.10.

(69) Pyris 1 TGA

(70) TGA data was collected on a Pyris 1 TGA equipped with a 20 position autosampler. The instrument was calibrated using certified indium. 6.329 mg of the sample was loaded onto a pre-weighed aluminium crucible and was heated at 20 C..Math.min.sup.1 (or 40 C..Math.min.sup.1) from ambient temperature to 500 C. A nitrogen purge at 20 ml.Math.min.sup.1 was maintained over the sample. The instrument control and data analysis was Pyris Software v9.0.1.0174.

(71) Polarised Light Microscopy (PLM)

(72) Samples were studied on a Leica LM/DM polarised light microscope with a digital videocamera for image capture. A small amount of each sample was placed on a glass slide, mounted in immersion oil and covered with a glass slip, the individual particles being separated as well as possible. The sample was viewed with appropriate magnification and partially polarised light, coupled to a false-colour filter.

(73) Microscopy

(74) Samples were studied on a Leica DME polarised light microscope with a digital video camera for image capture. A small amount of the sample was placed on a glass slide and covered with a glass slip, individual particles being separated as well as possible. The sample was viewed with appropriate magnification (10/0.22) and fully polarised light to assess crystallinity.

(75) Hot Stage Microscopy (HSM)

(76) Hot Stage Microscopy was carried out using a Leica LM/DM polarised light microscope combined with a Mettler-Toledo MTFP82HT hot-stage and a digital video camera for image capture. A small amount of each sample was placed onto a glass slide with individual particles separated as well as possible. The sample was viewed with appropriate magnification and partially polarised light, coupled to a false-colour filter, whilst being heated from ambient temperature typically at 10-20 C..Math.min.sup.1.

(77) Gravimetric Vapour Sorption (GVS)

(78) SMS DVS Intrinsic

(79) Sorption isotherms were obtained using a SMS DVS Intrinsic moisture sorption analyser, controlled by SMS Analysis Suite software. The sample temperature was maintained at 25 C. by the instrument controls. The humidity was controlled by mixing streams of dry and wet nitrogen, with a total flow rate of 200 ml.Math.min.sup.1. The relative humidity was measured by a calibrated Rotronic probe (dynamic range of 1.0-100% RH), located near the sample. The weight change, (mass relaxation) of the sample as a function of % RH was constantly monitored by the microbalance (accuracy0.005 mg). Typically 5-20 mg of sample was placed in a tared mesh stainless steel basket under ambient conditions. The sample was loaded and unloaded at 40% RH and 25 C. (typical room conditions). A moisture sorption isotherm was performed as outlined below (2 scans giving 1 complete cycle). The standard isotherm was performed at 25 C. at 10% RH intervals over a 0.5-90% RH range.

(80) Method Parameters for SMS DVS Intrinsic Experiments:

(81) TABLE-US-00003 Parameters Values Adsorption - Scan 1 40-90 Desorption/Adsorption - Scan 2 85 - Dry, Dry - 40 Intervals (% RH) 10 Number of Scans 2 Flow rate (ml .Math. min.sup.1) 200 Temperature ( C.) 25 Stability ( C. .Math. min.sup.1) 0.2 Sorption Time (hours) 6 hour time out

(82) The sample was recovered after completion of the isotherm and re-analysed by XRPD.

(83) Hiden Isochema Moisture Sorption Analyser (IGAsorp)

(84) Sorption isotherms were obtained using a Hiden Isochema Moisture Sorption Analyser (IGAsorp) controlled by IGAsorp Systems Software V6.50.48. The sample temperature was maintained at 25 C. by the instrument controls. The humidity was controlled by mixing streams of dry and wet nitrogen, with a total flow rate of 250 ml.Math.min.sup.1. The instrument was verified for relative humidity content by measuring three calibrated Rotronic salt solutions (10-50-88%). The weight change of the sample as a function of % RH was monitored by microbalance (accuracy0.005 mg). A defined amount of sample was placed in a tared mesh stainless steel basket under ambient conditions. A full experimental cycle consisted of two scans (sorption and desorption) at a constant temperature (25 C.) and 10% RH intervals over a 10-90% RH range (90 minutes for each humidity level).

(85) Water Determination by Karl Fischer Titration (KF)

(86) The water content of each sample was measured on a Mettler Toledo DL39 Coulometer using Hydranal Coulomat AG reagent and an argon purge. Weighed solid samples were introduced into the vessel on a platinum TGA pan which was connected to a subaseal to avoid water ingress. Approx 10 mg of sample was used per titration and duplicate determinations were made.

(87) Thermodynamic Aqueous Solubility

(88) Aqueous solubility was determined by suspending sufficient compound in HPLC grage water to give a maximum final concentration of 10 mg.Math.ml.sup.1 of the parent free-form of the compound. The suspension was equilibrated at 25 C. for 24 hours then the pH was measured. The suspension was then filtered through a glass fibre C filter into a 96 well plate. The filtrate was then diluted by a factor of 101. Quantitation was by HPLC with reference to a standard solution of approximately 0.1 mg.Math.ml.sup.1 in DMSO. Different volumes of the standard, diluted and undiluted sample solutions were injected. The solubility was calculated using the peak areas determined by integration of the peak found at the same retention time as the principal peak in the standard injection.

(89) HPLC Method Parameters for Solubility Measurements

(90) TABLE-US-00004 Type of Method Reverse phase with gradient elution Column: Phenomenex Luna, C18 (2) 5 m 4.6 mm Column Temperature ( C.) 25 Standard Injection (l): 1, 2, 3, 5, 7, 10 Test Injections (l): 1, 2, 3, 10, 20, 50 Detection: 260, 80 Wavelength Bandwidth (mm): Flow Rate (ml .Math. min.sup.1): 2 Phase A: 0.1% TFA in water Phase B: 0.085% TFA in acentonitrile Time (min) % Phase A % Phase B Timetable: 0.0 95 5 1.0 80 20 2.3 5 95 3.3 5 95 3.5 95 5 4.4 95 5

(91) Analysis was performed on an Agilent HP1100 series system equipped with a diode array detector and using ChemStation software vB.02.01-SR1.

(92) Chemical Purity Determination by HPLC

(93) Purity analysis was performed on an Agilent HP1100 series system equipped with a diode array detector and using ChemStation software vB.02.01-SR1.

(94) HPLC Method Parameters for Chemical Purity Determinations

(95) TABLE-US-00005 Sample Preparation: 0.5 mg .Math. ml1 in acetonitrile: water 1:1 v/v Column: Phenomenex Luna C18 (2) 150 4.6 mm 5 m Column Temperature ( C.): 25 Injection l): 5 Detection: 255, 90 Wavelength Bandwidth (mm): Flow Rate (ml .Math. min1): 1 Phase A: 0.1% TFA in water Phase B: 0.085% TFA in acentonitrile Time (min) % Phase A % Phase B Timetable: 0 95 5 25 5 95 25.2 95 5 30 95 5
Ion Chromatography (IC)

(96) Data were collected on a Metrohm 761 Compaction chromatography (for cations) and a Metrohm 861 Advanced Compaction chromatography (for anions) using ion Chromatography Net software v2.3. Accurately weighed samples were prepared as stocksolutions in DMSO and diluted 1:9 with either DMSO or water prior to testing. Quantification was achieved by comparison with standard solutions of known concentration of the ion being analysed.

(97) HPLC Method Parameters for Anion Chromatography

(98) TABLE-US-00006 Type of method Anion exchange Column: Metrosep A Supp 5 - 250 (4.0 250 mm) Column Temperature ( C.): Ambient Injection (l): 20 Detection: Conductivity detector Flow Rate (ml .Math. min.sup.1): 0.7 Eluent: 3.2 mM sodium, carbonate, 1.0 mM sodium hydrogen carbonate in 5% aqueous acetone.
pKa Determination and Prediction
Determination.

(99) Data were collected on a Sirius GlpKa instrument with a D-PAS attachment. Measurements were made at 25 C. in aqueous solution by UV and in methanol water mixtures by potentiometry. The titration media was ionic-strength adjusted (ISA) with 0.15 M KCl (aq). The values found in the methanol water mixtures were corrected to 0% co-solvent via a Yasuda-Shedlovsky extrapolation. The data were refined using Refinement Pro software v2.2.

(100) Prediction

(101) Prediction of pKa values was made using ACD pKa prediction software v11.

(102) Log P Determination

(103) Data were collected by potentiometric titration on a Sirius GlpKa instrument using three ratios of octanol:ionic-strength adjusted (ISA) water to generate Log P, Log Pion, and Log D values. The data were refined using Refinement Pro software v2.2. Prediction of Log P values was made using ACD v11 software.

(104) Compound Preparation

(105) Compound (I) may be prepared in accordance with the methodology described in WO 2008/122767 (Cyclacel Limited).

(106) Alternatively Compound (I) may be prepared via the following procedure:

(107) A solution of (4,6-dimethylpyridin-3ylmethyl)-(2-fluoro-9-isopropyl 9H-purin-6-yl)-amine (30 g), (2R,3S)-3-amino-pentan-2-ol (29.5 g) and DIEA (33.0 mL) in ethylene glycol (270 mL) was heated at 125 C. under nitrogen overnight. A further 0.5 equivalents of (2R,3S)-3-amino-pentan-2-ol (4.9 g) was added and the reaction stirred for an additional 6 hours. Analysis by HPLC indicated 1.9% (4,6-dimethylpyridin-3ylmethyl)-(2-fluoro-9-isopropyl 9H-purin-6-yl)-amine remained. The reaction was therefore left to stir at 125 C. overnight. Analysis by HPLC now indicated only 0.35% (4,6-dimethylpyridin-3ylmethyl)-(2-fluoro-9-isopropyl 9H-purin-6-yl)-amine remained. The reaction was therefore cooled to room temperature and added to ethyl acetate (2460 mL). Water (1320 mL) was added and the phases separated. The aqueous phase was extracted with ethyl acetate (22460 mL) and the combined organics were washed with water (22460 mL), dried over MgSO.sub.4, filtered and stripped. Purification by flash column chromatography (1500 g silica, 3% MeOH in DCM as eluent) gave the desired product as a white solid. Drying in a vacuum oven overnight gave compound (I) in 59% yield (22.3 g, JCCA824). 1H NMR confirmed the identity of the product and HPLC gave a purity of 99.16%.

Example 1

Crystallisation of Free Base Compound (I) to Give Form A

(108) Compound (I) was crystallised from MTBE by the following method. MTBE (2 vol) was added to compound (I) and heated to reflux. The mixture was held at reflux for 30-60 minutes before the temperature was reduced to 50 C. (held for 2 hours). The suspension was allowed to cool slowly to room temperature before being filtered and rinsed with MTBE (31 vol). The solids were dried in vacuum oven at 40 C. for 8 hours to afford the desired crystalline free base (mass recovery 84.5%, LC purity 97.4%).

(109) XRPD information on Form A is found in Table 1.

(110) Gravimetric Vapour Sorption on Form A

(111) 11.254 mg of sample was placed in a tarred mesh stainless steel basket under ambient conditions. A full experimental cycle consisted of two scans (sorption and desorption) at a constant temperature (25 C.) and 10% RH intervals over a 40-90% range (180 minutes for each humidity level). The mass increase of approximately 0.09 mg (0.8%) and the facile uptake and loss relative to humidity level indicates a non-hygroscopic sample that dampens with surface moisture only. FIG. 15 shows an isotherm plot for Form A (weight change % vs RH %), whereas FIG. 16 shows a kinetic plot of weight change % vs time and relative humidity %. Analysis of the sample post GVS run showed no change by XRPD (FIG. 17).

(112) Thermodynamic Aqueous Solubility of Form A

(113) Aqueous solubility was determined by suspending sufficient compound in HPLC grade water to give a maximum final concentration of 10 mg.Math.ml.sup.1 of the parent free-form of the compound. The suspension was equilibrated at 25 C. for 24 hours. The suspension was then filtered through a filter into an HPLC vial. The filtrate was then diluted by an appropriate factor. Quantification was executed by HPLC with reference to a standard solution of 0.5 mg in 1 mL acetonitrile/water (1:1). Different volumes of the standard, diluted and undiluted sample solutions were injected. The solubility was calculated using the peak areas determined by integration of the peak found at the same retention time as the principal peak in the standard injection. The aqueous solubility was determined to be 0.329 mg/mL. Analysis of Form A post solubility study showed no change by XRPD (FIG. 18).

Example 2

Preparation of Citrate Salt (Form F) of Compound (I)

(114) Compound (I) (100 mg, 0.25 mmol, 1 equiv), citric acid (49 mg, 0.26 mmol, 1.02 equiv) and ethyl acetate (1 ml, 10 vol) were charged to a vial and stirred under ambient conditions for 24 hoursa small lump of sticky solid remained undissolved. The mixture was stored in a shaker on a heat/cool cycle (60 C./RT, 4 h) for 68 h. The resultant white precipitate was isolated by vacuum filtration, washed with EtOAc (2500 l, 25 vol) and dried in a vacuum oven at 30 C. for 16 hours to yield the citrate salt as a white solid (53 mg, 40% yield).

(115) The .sup.1H NMR spectrum of the citrate salt was consistent with structure and a set of diastereotopic peaks was present corresponding to the citrate anion. .sup.1H NMR analysis also confirmed the presence of residual ethyl acetate. XRPD analysis confirmed the material to be crystalline. DSC analysis showed two endothermic events: a sharp peak, corresponding to the melt, with an onset of 145 C. and a broad event with an onset of 165 C. TGA analysis showed no weight loss before or during the melt, followed by decomposition above 180 C., confirming the second endothermic event in the DSC is likely to be decomposition. TGA analysis also proved the material is not a solvate.

(116) A sample of the product was stored in a humidity chamber at 25 C. and 94% RH for 3 days, after which time the material was slightly tacky, though it had not deliquesced. XRPD analysis of this material showed the same pattern as obtained for the initial product, but with a significant amorphous halo. A second sample was stored in a humidity chamber at 25 C. and 75% RH for 70 hours, after which time the sample appeared unchanged and the XRPD pattern obtained for this material was consistent with that of the original product. GVS analysis conformed that no hydrate is formed at high RH, though the material proved to be hygroscopic above 70% RH. XRPD information on the citrate salt is found in Table 3.

(117) Summary of Results for Citrate Salt:

(118) TABLE-US-00007 Onset of melt HPLC purity Aqueous solubility 145 C. 99.9% >15 mg .Math. ml.sup.1

Example 3

Preparation of Benzenesulfonic Acid Salt (Form G) of Compound (I)

(119) Compound (I) (100 mg, 0.25 mmol, 1 equiv), benzenesulfonic acid (41 mg, 0.26 mmol, 1.02 equiv) and tert-butyl-methylether (1 ml, 10 vol) were charged to a vial and under ambient conditions and stirring was initiated, the mixture never went into solution. The mixture was stirred for 30 minutes, after which time the sample contained a significant quantity of sticky solid. The mixture was stirred for a further 23.5 hours and showed no change. The mixture was stored in a shaker on a heat/cool cycle (60 C./RT, 4 h) for 68 h. The resultant white precipitate was isolated by vacuum filtration, washed with TBME (3500 l, 35 vol) and dried in a vacuum oven at 30 C. for 16 hours to yield the benzenesulfonic acid salt as a white solid (65 mg, 47% yield).

(120) The 1H NMR spectrum of the benzenesulfonate salt clearly showed aromatic peaks corresponding to the benzenesulfonate. However, a large proportion of the peaks corresponding to compound (I) appear as broadened multiplets, probably due to the presence of the sulfonic acid group, making successful assignment non-trivial. A significant quantity of residual TBME was also visible, though this is believed to be unbound solvent rather than a solvate. XRPD analysis confirmed the material to be crystalline. DSC analysis showed only one endothermic event with an onset of 147 C., and hot stage microscopy confirmed this event to be a melt. TGA analysis exhibited a weight loss of 1.1% between 75 and 120 C. that was not accompanied by an endotherm in the DSC and is attributed to the loss of unbound solvent, proving the material to be non-solvated. Deliquescence of a sample of the product was observed after storage in a humidity chamber at 25 C. and 94% RH for 2 hours. A second sample was stored in a humidity chamber at 25 C. and 75% RH for 70 hours, after which time the sample was slightly tacky. The XRPD pattern obtained for this material was consistent with that of the original product, but with a slightly larger amorphous halo.

(121) XRPD information on the benzenesulfonic acid salt is found in Table 4.

(122) Summary of Results for Benzenesulfonic Acid Salt:

(123) TABLE-US-00008 Onset of melt HPLC purity Aqueous solubility 147 C. 98.9% >20 mg .Math. ml.sup.1

Example 4

Preparation of L-Tartrate Salt (Form D) of Compound (I)

(124) Compound (I) (500 mg, 1.26 mmol, 1 equiv.), L-tartaric acid (193 mg, 1.28 mmol, 1.02 equiv) and ethyl acetate (5 ml, 10 vol) were charged to a flask and stirred under ambient conditions for 2 hours, precipitation occurred inside 1 hour. The white precipitate was isolated by vacuum filtration, washed with EtOAc (30.5 ml, 21 ml) and dried in a vacuum oven at 40 C. for 16 hours to yield the L-tartrate salt as a white solid (565 mg, 82% yield).

(125) The .sup.1H NMR spectrum of the tartrate salt was consistent with structure and exhibited a singlet at 4.31 ppm corresponding to the tartrate anion. XRPD analysis confirmed the material to be crystalline. DSC analysis showed a single endothermic event with an onset of 147 C. that was confirmed, by hot stage microscopy, to be a melt. TGA analysis showed no weight loss before or during the melt, followed by decomposition above 200 C., showing that this material is not a solvate.

(126) A sample of product was stored in a humidity chamber at 25 C. and 94% RH for 70 hours, after which time the material was slightly tacky, though it had not deliquesced. XRPD analysis of this material showed the presence of some peaks corresponding to those in the pattern acquired for the original product, though a significant amorphous halo was also visible, due to the uptake of water. A second sample was stored in a humidity chamber at 25 C. and 75% RH for 4 days, after which time the sample was unchanged. The XRPD pattern obtained for this material was consistent with that of the original product. GVS analysis confirmed that no hydrate is formed at high RH, though the material proved to be hygroscopic above 70% RH.

(127) XRPD information on the L-tartrate salt (Form D) is found in Table 2.

(128) Summary of Results for L-Tartrate Salt (Form D):

(129) TABLE-US-00009 Onset of melt HPLC purity Aqueous solubility 147 C. 99.4% >20 mg .Math. ml.sup.1

Example 5

Preparation of L-Tartrate Salt (Form E) of Compound (I)

Example 5.1

(130) (a) A suspension of Form D L-tartrate salt of compound (I) (1.0 g) in ethanol (12 ml) was heated at reflux. Acetonitrile (3 ml) was added portion wise over 30 minutes. After this addition, a solution was not obtained. Further portions of ethanol (4.5 ml) and acetonitrile (1 ml) were added until a solution was obtained. The solution was polish filtered (hot) then cooled to room temperature at a rate of 10 C./hour (crystallisation initiated at 65 C.). After stirring at room temperature overnight, the resulting solid was filtered, washed with cold ethanol (5 ml) and pulled dry. Further drying in a vacuum oven at 50 C. yielded the desired product as a white crystalline solid (0.725 g, 73%). 1H NMR analysis confirmed a 1:1 salt and XRPD confirmed Form E.

(131) A solubility study at 25 C. indicated a solubility for Form E of 43.905 mg/ml) 24 hour incubation in water with constant agitation). The remaining solid was dried and XRPD analysis showed no differences in powder pattern after the aqueous slurry. The pH of the solution post solubility was pH 5.

(132) A TGA study for Form E was also carried out (held at 100 C. for 24 hours). The results indicated the material was stable at this temperature and XRPD analysis of the material post TGA stability study showed no differences in powder pattern.

(133) XRPD information on the L-tartrate salt (Form E) is found in Table 7.

Example 5.2

(134) A suspension of Form D L-tartrate salt of compound (I) (10.2 g) in ethanol (120 ml) was heated to 65 C. Acetonitrile (20 ml) was added and the suspension heated at reflux for 10 minutes after which time a solution was obtained. The solution was cooled to room temperature over 2-3 hours with crystallisation initiating at 50 C. The resulting suspension was stirred at room temperature overnight. The resulting solid was filtered, washed with ethanol (10 ml) and pulled dry. Further drying in a vacuum oven at 50 C. yielded the desired product as a white crystalline solid (8.76 g, 88%). 1H NMR analysis confirmed a 1:1 salt and XRPD confirmed Form E.

Example 5.3

Slurry Conversion

(135) Form E of the L-Tartrate salt of compound (I) was also prepared by slurry conversion from four different solvents (ethyl acetate, IPA, IMS or acetonitrile). A 1:1 mixture of Form by weight of D: Form E L-Tartrate salt (200 mg total) was heated at 45 C. over 48 hours in 2 ml of solvent prior to filtration and analysis. Form E was produced in each slurry (purity98%).

Example 5.4

Seeding

(136) A suspension of Form D L-tartrate salt compound (I) (10.2 g) in ethanol (120 ml) was heated to 65 C. Acetonitrile (20 ml) was added and the suspension heated at reflux for 10 minutes. The mixture was polish filtered through HPLC filter frits. No precipitation was observed in process. The material was then cooled from reflux and seeded at 70 C. with Form E L-tartrate salt (as prepared above), cooling at a rate of 10 C. every 1.5 hours. The first seed dissolved completely and seeding was repeated at 60 C. The seed remained and the solution changed to show a very faint opaque phase. Crystallisation began at approximately 50 C. An isolated yield of 80% was obtained.

Example 5.5

Formation from Free Base of Compound (I)

(137) CYC065 free base Form A (0.2 g) was dissolved in ethanol (9 vol, 1.8 mL) and heated at reflux. A solution of tartaric acid (1 eq, 0.076 g) in water (1.7 vol, 0.34 mL)/ethanol (1 vol, 0.2 mL) was added dropwise maintaining the temperature at reflux. The resulting solution was then polish filtered before cooling to 70 C. A seed of Form E was added giving a cloudy solution. The batch was stirred at 70 C. for 1 hour before cooling to room temperature. After stirring at room temperature for 2 hours, the solid was filtered, washed with ethanol (20.5 mL) and pulled dry. Further drying in a vacuum oven at 50 C. yielded CYC065-L-tartrate salt Form E as a white solid (0.2 g, 72%). 1H NMR confirmed a 1:1 salt and HPLC indicated a purity of 97.97%. XRPD and DSC confirmed Form E.

Example 6

Preparation of the Phosphate Salt (Form C) of Compound (I) from Ethanol

(138) A solution of phosphoric acid in water (85% w/w) (192 l, 1.67 mmol, 1.02 equiv) was added to a stirring solution of compound (I) (650 mg, 1.64 mmol, 1 equiv) in ethanol (6.5 ml, 10 vol) at RT in a cool water bath over a period of 2 minutes, the mixture remained a light yellow solution throughout the addition. The water bath was removed and the mixture was stirred at RT for 2 hours and the resultant white precipitate was isolated by vacuum filtration, washed with ethanol (21.3 ml, 22 vol) and dried in a vacuum oven at 40 C. for 16 hours to yield the phosphate salt as a white solid (641 mg, 66% yield).

(139) The XRPD pattern obtained for the product was similar, though not identical to, the corresponding material crystallised from IPA (see Example 7). These results suggest that crystallisation from ethanol yields a different polymorph of the resultant phosphate salt than crystallisation from IPA. DSC analysis showed two broad endothermic events with onsets of 67 and 125 C. The first endotherm was comparable with the first event observed in the DSC traces of the material formed in IPA (see Example 7). Interestingly, the second endotherm had a significantly higher onset (123 C.) than the comparable material obtained from IPA (116 C.), again suggesting the presence of a different solid form.

(140) The .sup.1H NMR spectrum obtained was consistent with the structure and showed a minor amount of residual ethanol.

(141) TGA analysis showed a weight loss of 2.9%, between 45 and 98 C., corresponding to the first endotherm in the DSC, thus confirming that the first endotherm observed in the DSC was due to the loss of bound solvent. At this stage, it was not confirmed whether this solvent loss was due to the loss of IPA or water, or a mixture of both. TGA analysis showed no weight loss associated with the second event observed in the DSC, followed by decomposition above 220 C. Hot stage microscopy confirmed the second endotherm observed in the DSC to be a melt.

(142) GVS analysis was undertaken at 25 C. and showed the material is not hygroscopic as a gradual uptake of 0.5% by weight of water between 60 and 90% RH was observed. A weight loss of 0.5% was observed between 10 and 0% RH, which is lower than the weight loss observed in the TGA.

(143) Variable temperature XRPD analysis was undertaken on a sample of Form C in order to elucidate if the loss of solvent observed in the DSC resulted in a change of form. The sample was heated to each temperature, held for 3 minutes to allow equilibration and an XRPD pattern was collected. The material was taken to 100 C., in 10 degree increments, a pattern was collected, the sample was held at 100 C. for 15 minutes and a second pattern was collected. The sample was cooled back to RT and a final reference pattern was obtained.

(144) Subtle changes were observed between the XRPD patterns acquired at RT, both before and after cooling, and the pattern obtained at 100 C. This suggests that loss of solvent does bring about a change in form and that this process is reversible in the presence of air. DSC analysis of the sample after it had undergone VT XRPD analysis exhibited two broad endotherms, with onsets of 67 and 124 C., and was analogous to the DSC trace obtained for the sample prior to undergoing the VT XRPD experiment. This result confirms that the loss of solvent is reversible in air, therefore implying that the product exists as a hydrated form under ambient conditions and not as an alcoholic solvate.

(145) In order to add further weight to this hypothesis a heat/cool/heat DSC analysis was undertaken. In this experiment, a sample of product was heated to 100 C., held for 10 minutes, cooled to 30 C., held for 5 minutes and then re-heated to 250 C. As anticipated the first heat showed the desired endotherm corresponding to the loss of water. The second heat only exhibited a single endotherm, corresponding to the melt of the anhydrous form, therefore confirming that the loss of water is irreversible under an inert atmosphere of nitrogen.

(146) XRPD information on the phosphate salt prepared from ethanol (Form C) is found in Table 5.

(147) Summary of Results for Phosphate Salt Prepared from Ethanol:

(148) TABLE-US-00010 % H.sub.2O Onset of Onset of Wt loss HPLC Aqueous Karl dehydration melt TGA purity solubility Fischer 68 C. 123 C. 2.8 99.6% >20 mg .Math. ml.sup.1 3.5

(149) Removal of all residual ethanol was achieved by drying in a vacuum oven either at 60 C. for 24 hours (76 mg scale, Sample A) or at 50 C. for 68 hours (346 mg scale, Sample B). .sup.1H NMR, HPLC purity, XRPD and DSC analyses all confirmed that the dry materials were analogous to the starting product and had not degraded during the drying process. TGA and Karl Fischer analyses were used to calculate the equivalents of water.

(150) TABLE-US-00011 Wt % H.sub.2O Onset of Onset of loss HPLC Equiv Karl dehydration melt TGA purity H.sub.2O Fischer Sample A 67 C. 122 C. 2.5 99.6% 0.83 N/A Sample B 70 C. 123 C. 2.4 N/A 0.82 3.0

Example 7

Preparation of the Phosphate Salt (Form B) of Compound (I) from Propan-2-ol

(151) A solution of phosphoric acid in water (85% w/w) (147 l, 1.28 mmol, 1.02 equiv) was added to a stirring solution of compound (I) (500 mg, 1.26 mmol, 1 equiv) in isopropanol (5 ml, 10 vol) at RT in a cool water bath over a period of 2-3 minutes. A sticky white material formed into a disk during the addition. The mixture was stirred vigorously for 30 minutes, after which time the mixture had become a thick yellow sludge and stirring was poor. A further 750 l of IPA was added and the mixture was stirred for a further 1 h. The resultant white precipitate was isolated by vacuum filtration and washed with IPA (21.5 ml, 23 vol) and dried under suction. Two distinct types of material were clearly visible in the cakea dry white powdery solid round the outside and a sticky off white solid in the centre. The dry solid was isolated to yield the phosphate salt as a white solid (110 mg, 15% yield).

(152) The .sup.1H NMR spectrum of the product was consistent with structure, a minor amount of residual IPA was also noted. The XRPD confirmed that the product was in crystalline form. The DSC trace exhibited two events with onsets of 67 and 116 C. and the TGA exhibited a weight loss of 2.6%, between 45 and 98 C., corresponding to the loss of bound solvent during the first endothermic event in the DSC. At this stage it was not confirmed whether this solvent loss was due to loss of IPA or water, or a mixture of both. The TGA showed no weight loss associated with the second event in the DSC, followed by decomposition above 220 C. Hot stage microscopy confirmed the second endotherm observed in the DSC to be a melt.

(153) XRPD information on the phosphate salt prepared from propan-2-ol (Form B) is found in Table 6.

(154) Summary of Results for Phosphate Salt Prepared from Propan-2-Ol:

(155) TABLE-US-00012 Onset of Onset of Wt loss HPLC dehydration melt TGA purity 67 C. 116 C. 2.6% 99.5%

(156) Variable temperature XRPD analysis was undertaken on the product in order to elucidate if the loss of solvent observed in the DSC resulted in a change of form. The sample was heated to each temperature, held for 3 minutes to allow for equilibration and an XRPD pattern was collected. The material was taken to 100 C., in 10 degree increments, a pattern was collected then the sample was held at 100 C. for 15 minutes and a second pattern was collected. The sample was cooled back to RT and a final reference pattern was obtained.

(157) There are only very subtle changes in the XRPD patterns obtained at RT and at 100 C., although they do appear to be reversible as the differences disappear again upon cooling to RT.

(158) Given the suggestion, from the VT XRPD analysis, that the product might be a hydrate that can reversibly lose then reacquire water with heating, a heat/cool/heat DSC experiment was undertaken under an inert atmosphere of nitrogen. A sample of product was heated to 100 C., held for 10 minutes, cooled to 30 C., held for 5 minutes and then re-heated to 250 C. The first heat showed, as anticipated, the loss of solvent. However, the second heat showed no endothermic event corresponding to the loss of solvent therefore showing that this process is irreversible under an inert atmosphere of nitrogen.

(159) To investigate whether the phosphate salt exists as a hydrate under ambient conditions that can be reversibly dehydrated at temperature in the presence of air, a series of TGA experiments were undertaken to clarify how quickly the material rehydrates under ambient conditions. A series of samples were taken to 100 C. then cooled to 25 C. under nitrogen using the TGA equipment. The samples were stored under ambient conditions (RT in air) for varying times and then the TGA experiment was re-run. As a comparison a heat/cool/heat experiment was undertaken under an inert atmosphere of nitrogen using the TGA; a sample was heated to 100 C., cooled to 25 C. and heated to 100 C. for a second time.

(160) These results showed that the phosphate salt readily rehydrates under ambient conditions (assumed to be approximately 25 C. and 40% RH). Within 30 minutes an 8 mg sample of product in an open TGA pan had fully rehydrated. This result is promising for any potential scale up as it shows that even if, during the drying process, some, or all, of the water is lost from the sample, it will rehydrate upon storage under ambient conditions.

(161) A sample of product was stored in a humidity chamber at 25 C. and 94% RH for 88 hours, after which time no deliquescence was observed. The XRPD pattern obtained for this material matched those acquired for the phosphate salt crystallised from ethanol (see Example 6), rather than that of the parent compound. The DSC analysis exhibited two discreet melts, corresponding to the crystalline form produced by crystallisation from IPA and that produced from ethanol, with the melt of the ethanol form appearing to be the dominant event. GVS analysis was undertaken at 25 C. and showed the material not to be hygroscopic as a gradual uptake of only 1.5 weight % of water was observed above 80% RH. Interestingly upon taking to 0% RH only 0.5 weight % of water was lost, compared to >2% observed in the TGA. The XRPD pattern obtained for this material matched those acquired from product crystallised from ethanol, rather than that of the parent compound. The DSC analysis exhibited broadening of the endotherm corresponding to the melt, though it still appeared as a single event.

(162) It was postulated that complete loss of water may only occur at elevated temperature, thus a second GVS experiment was run at 40 C. At 40 C. the product exhibited a gradual uptake of 2 weight % of water above 80% RH, again proving the material is not hygroscopic. On this occasion upon taking to 0% RH the material lost 2.6 weight % of water, which is in close agreement with the mass loss observed in the TGA. The XRPD pattern obtained for this material matched those acquired from product crystallised from ethanol, rather than that of the parent compound. The DSC analysis exhibited two discreet melts corresponding to the crystalline form produced by crystallisation from IPA and that produced from ethanol, with the melt of the ethanol form appearing to be the dominant.

Example 8

Preparation of Hydrochloride Salt (Pattern 1; Designated Form H) of Compound (I)

(163) HCl (37 wt % solution in water) (88 l, 0.88 mmol, 1 equiv) was added dropwise to a solution of free base compound (1) (350.22 mg, 0.88 mmol, 1 equiv) in TBME (17.5 ml, 50 relative volumes) at RT with swirlinga sticky white solid formed instantaneously. The sample was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h at each) for 63 h. The mixture had concentrated to approximately original volume during the maturation. The resultant solid was isolated by vacuum filtration and washed with TBME (25 ml). The solid was dried under suction and in a vacuum oven at 30 C./3 mbar for 20 h to yield the hydrochloride salt as a white solid (158.69 mg, 42% based on mono-chloride salt formation).

(164) Summary of Data for Hydrochloride Salt (Form H) of Compound (I)

(165) TABLE-US-00013 Analysis Result .sup.1H NMR Changes consistent with salt formation, different to HCl Pattern 2 (Form I) 2.7% residual TBME (0.1 equiv) XRPD Consistent with HCl Pattern 1(Form H) VT-XRPD Conversion to a new pattern >80 C., sample became amorphous >100 C. IC 2.2 equiv of chloride DSC Four broad endotherms with onset temperatures at 51.4, 83.7, 97.1 and 144.0 C. (H 3, 1, 8 and 18 respectively) TGA 3.4% wt loss from 45 to 105 C. equates to 0.8 equiv water 2.6% wt loss from 130 to 170 C. - loss of TBME? HSM Loss of birefringence at 60 C., melt/formation of a gum accompanied by loss of solvent from 85 C. Karl 8.6% water equates to 2.3 equiv water Fischer HPLC 98.8% (largest % imp 0.46 @ 1.11 RRT) AQ 75 mg .Math. ml.sup.1, pH 2.2 Solubility 40 C./ 30 min - Deliquescence 75% RH

(166) XRPD information on the hydrochloride salt (Form H) of compound (I) is found in Table 8 and FIG. 19.

(167) XRPD analysis of the product appeared to suggest formation of HCl Pattern 1 (Form H), however, the ion chromatography results were inconsistent with those obtained previously for HCl Pattern 1 (Form H) that suggested it to be a mono-chloride. These results were unexpected, and do not match the observation that HCl Pattern 2 (Form I, mono chloride salt) shows conversion to HCl Pattern 1 (Form H) upon storage at 40 C./75% RH.

(168) In an attempt to confirm the formation of a bis-HCl salt, a sample of the hydrochloride salt (Form H) (25.99 mg) was slurried in 3% water in EtOAc (1.3 ml, 50 relative volumes) and the mixture was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h) for 72 h. A colourless gum was obtained. In two further attempts to verify the stoichiometry of the isolated material, samples of the hydrochloride salt (Form H) were washed with 3% water in EtOAc or TBME (3750 l) and dried under suction. Unfortunately, in each case a sticky solid was isolated instead of the desired solid.

(169) VT-XRPD analysis of the product suggested conversion of HCl Pattern 1 (Form H) to a new pattern (HCl Pattern 4) above 80 C. followed by complete loss of crystallinity above 100 C. Visual analysis of the sample used for VT-XRPD showed the formation of a foam at elevated temperatures, once crystallinity had been lost, indicative of a loss of volatile material (possibly TBME or excess HCl). TGA shows a weight loss of 3.4% from 45 to 105 C., equivalent to 0.8 equiv H.sub.2O and greater than the observed levels of TBME. The weight loss could indicate the presence of a solvate that desolvates to the different form observed by VT-XRPD. Indeed, HSM showed a loss of birefringence around 60 C. that could indicate desolvation followed by a melt, or the formation of a gum due to the presence of silicone oil, accompanied by the loss of solvent from 85 C. These results suggest HCl Pattern 1 (Form H) material to be a solvate, however due to the presence of TBME by NMR and two weight losses by TGA it cannot be verified whether the material is a TBME solvate or hydrate.

Example 9

Preparation of Hydrochloride Salt (Pattern 2; Designated Form I) of Compound (I)

(170) HCl (37 wt % solution in water) (95 l, 0.95 mmol, 0.9 equiv) was added dropwise to a solution of free base compound (1) (399.72 mg, 1.01 mmol, 1 equiv) in EtOAc (12 ml, 30 relative volumes) at RT with swirlinga sticky white solid formed instantaneously. The sample was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h at each) for 63 h. The resultant solid was isolated by vacuum filtration and washed with EtOAc (23 ml). The solid was dried under suction and in a vacuum oven at 30 C./3 mbar for 20 h to yield the hydrochloride (Form I) as a yellow solid (151.71 mg, 35% based on mono-chloride salt formation).

(171) Summary of Data for Hydrochloride Salt (Form I) of Compound (I)

(172) TABLE-US-00014 Analysis Result .sup.1H NMR Changes consistent with salt formation, different to HCl 1 (Form H) 5.0% residual EtOAc (~0.2 equiv) XRPD Partially crystalline, consistent with HCl Pattern 2 (Form I) Matched HBr Pattern 1 (Form K) VT-XRPD No change up to 120 C., melt by 140 C. IC 1.0 equiv of chloride DSC Two overlapping endotherms 97.7 and 129.7 C. (H 12 and 50 J .Math. g.sup.1) TGA 3.6% wt loss from 100 to 140 C., loss of EtOAc? HSM Extended melt with an onset of 109 C. some degradation above 125 C. HPLC 99.1% (largest % imp 0.33 @ 1.07 RRT) AQ >79 mg .Math. ml.sup.1, pH 4.4 Solubility 40 C./ 7 d - Sticky solid, XRPD showed conversion to HCl Pattern 1 75% RH (Form H)

(173) XRPD information on the hydrochloride salt (Form I) of compound (I) is found in Table 9 and FIG. 20.

(174) Analysis of the product confirmed HCl Pattern 2 (Form I) material to be a mono-chloride salt that appears to be isostructural with HBr Pattern 1. HSM shows a melt corresponding to the broad endotherms observed in the DSC, whilst VT-XRPD shows no change in form before the melt. These results suggest the material to be non-solvated, thus the residual solvent observed in the NMR is likely to be trapped within the crystal lattice and is lost upon melting, as observed in the TGA. Storage of HCl Pattern 2 (Form I) material at 40 C./75% RH showed it to be unstable under elevated levels of relative humidity.

(175) Analysis of HCl Pattern 1 (Form H)

(176) HCl Pattern 2 (Form I) showed conversion to HCl Pattern 1 (Form H) upon storage at 40 C./75%. This result may indicate that HCl Pattern 1 (Form H) is likely to be a mono-chloride salt. Proton NMR and TGA analyses were undertaken on this sample in an attempt to clarify the properties of HCl Pattern 1 (Form H) material, as the analysis sample from Example 8 was complicated by the presence of residual TBME and possibly excess HCl. Proton NMR confirmed the absence of residual TBME, whilst TGA showed a stepped weight loss of 4.6%, equivalent to 1.2 equiv of water. These results suggest HCl Pattern 1 (Form H) material to be a mono-hydrated mono-chloride salt of compound (I).

Example 10

Preparation of Hydrobromide Salt (Pattern 1; Designated Form J) of Compound (I)

(177) HBr (48 wt % solution in water) (77 l, 0.45 mmol, 0.9 equiv) was added dropwise to a solution of free base of compound (1) (200.46 mg, 0.50 mmol, 1 equiv) in EtOAc (10 ml, 50 relative volumes) at RT with swirlinga sticky white solid formed instantaneously. The sample was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h at each) for 18 h. The resultant sticky solid was isolated by vacuum filtration and washed with EtOAc (22 ml). The solid was dried under suction and in a vacuum oven at 30 C./3 mbar for 14 h to yield the hydrobromide salt as a yellow solid (86.54 mg, 40% based on mono-bromide salt formation).

(178) Summary of Data for Hydrobromide Salt (Form J) of Compound (I)

(179) TABLE-US-00015 Analysis Result .sup.1H NMR Changes consistent with salt formation 5% residual EtOAc (~0.2 equiv) XRPD Partially crystalline, consistent with HBr Pattern 1 (Form J) Matched HCl Pattern 2 (Form I) IC 1.1 equiv of bromide DSC Broad endotherm 138.2 C. (H 35 J .Math. g.sup.1) HSM Melt observed with an onset of 119 C. TGA 2.2% wt loss from 105 to 150 C., loss of EtOAc? HPLC 99.0% (largest % imp 0.34 @ 1.07 RRT) AQ >30 mg .Math. ml.sup.1, pH 4.7 Solubility 40 C./ 15 h -Deliquescence 75% RH 7 d - Sticky solid, XRPD showed conversion to a new Pattern; HBr 2 (Form K)

(180) XRPD information on the hydrobromide salt (Form J) of compound (I) is found in Table 10 and FIG. 21.

(181) Analysis confirmed HBr Pattern 1 (Form J) to be a partially crystalline mono-bromide salt that appears to be isostructural with HCl Pattern 2 (Form I). HSM and DSC analyses suggest a melt with an onset of approximately 138 C. As with HCl Pattern 2 (Form I), it is proposed that the material is non-solvated and that the weight loss observed in the TGA is due to the loss of solvent trapped within the crystal lattice upon melting. Storage at 40 C./75% RH showed conversion to a new PatternHBr Pattern 2 (Form K).

Example 11

Preparation of Hydrobromide Salt (Pattern 2; Designated Form K) of Compound (I)

(182) A new XRPD pattern corresponding to the HBr salt was obtained upon storage of HBr Pattern 1 (Form J) at 40 C./75% relative humidity. HBr Pattern 1 (Form J) showed deliquescence after 15 h, but crystallisation was observed upon extended storage of 7 days.

(183) Summary of Data for Hydrobromide (Form K) of Compound (I)

(184) TABLE-US-00016 Analysis Result XRPD Different to HBr pattern 1 (Form J) - HBr Pattern 2 (Form K) (inconsistent with HCl 1 (Form H)) DSC Broad endotherm 51.0 C. (H 70 J .Math. g.sup.1) followed immediately by a sharp endotherm 90.3 C. (H 75 J .Math. g.sup.1) TGA 3.0% wt loss from 45 to 70 C. equates to 0.8 equiv water IC 1.0

(185) Analysis suggests HBr Pattern 2 (Form K) to be a mono-hydrated mono-bromide salt, exhibiting dehydration with an onset of approximately 50 C. Further scale up and analysis of this material may be beneficial to further confirm these observations.

Example 12

Preparation of Mesylate Salt (Form L) of Compound (I)

(186) Methanesulfonic acid (59 l, 0.91 mmol, 0.9 equiv) was added dropwise to a stirring solution of free base compound (1) (400.08 mg, 1.0 mmol, 1 equiv) in TBME (20 ml, 50 relative volumes) at RTa sticky white solid formed instantaneouslyand the mixture was stirred at RT for 16 h. A precipitate appeared to have formed and an aliquot was taken and isolated by vacuum filtration. A sticky solid was obtained from the isolated sample that did not improve with drying. The remaining sample was stirred at RT for a further 8 h, heated to 40 C. and stirred for a further 1.5 h, then cooled to RT and stirred for 14 h. An aliquot was taken and the solid was isolated by vacuum filtration yielding a sticky solid. THF (1 ml) was added to the remaining mixture and the sample was stirred for 3 h. The mixture was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h at each) for 63 hevaporation was observed. TBME (10 ml) was added and the solid was isolated by vacuum filtration and washed with TBME (22 ml). The solid was dried under suction and in a vacuum oven at 30 C./3 mbar for 20 h to yield the mesylate salt (Form L) as a white solid (256.73 mg, 51% based on mono-mesylate salt formation).

(187) Summary of Data for Mesylate (Form L) of Compound (I)

(188) TABLE-US-00017 Analysis Result .sup.1H NMR 1 equiv methanesulfonic acid, 0.4% TBME XRPD Consistent with Mesylate Pattern 1 IC 0.7 equiv methanesulfonic acid DSC Sharp endotherm onset 126.1 C. (H 86 J .Math. g.sup.1) TGA 0.2% wt loss from 110 to 135 C. associated with endotherm - loss of TBME HSM Melt observed with an onset of 121 C. HPLC 98.8% (largest % imp 0.45 @ 1.11 RRT) AQ 137 mg .Math. ml.sup.1, pH 4.53 Solubility 40 C./ 15 h -Deliquesced 75% RH 7 d - No change

(189) XRPD information on the mesylate salt (Form L) of compound (I) is found in Table 11 and FIG. 22.

(190) Analysis suggested Mesylate Pattern 1 to be a non-solvated mono mesylate salt with a melt at approximately 126 C. The aqueous solubility was determined to be 140 mg.Math.ml.sup.1, the highest of the salts obtained during the course of this work. The sample was found to be hygroscopic as deliquescence was observed inside 15 h at 40 C. and 75% RH.

Example 13

Preparation of Maleate Salt (Form M) of Compound (I)

(191) Free base compound (I) (400.85 mg, 1.01 mmol, 1 equiv), maleic acid (117.01 mg, 1.01 mmol, 1 equiv) and TBME (20 ml, 50 relative volumes) were charged to a flask and stirred at RT for 16 h. A fine yellow suspension along a few lumps of yellow solid was observed. An aliquot was taken, the solid was isolated by vacuum filtration, dried under suction and analysed by XRPD (consistent with Maleate Pattern 1). The remaining mixture was stirred at RT for a further 8 h. The solid was isolated by vacuum filtration, dried under suction and in a vacuum oven at 30 C./3 mbar for 15 h to yield the maleate salt (Form M) as a yellow solid (279.67 mg, 54% yield based on mono-maleate salt formation).

(192) Summary of Data for Maleate (Form M) of Compound (I)

(193) TABLE-US-00018 Analysis Result .sup.1H NMR ~1 equiv maleic acid (peak overlaps with Compound (I)), 1.3% TBME XRPD Consistent with Maleate Pattern 1 IC 1.1 equiv maleic acid DSC Sharp endotherm, onset 116.1 C. (H 75 J .Math. g.sup.1) HSM Melt observed with an onset of 112 C. TGA 11.1% wt loss from 135 to 195 C. (~0.4 equiv Maleic acid) HPLC 99.0% (largest % imp 0.31 @ 1.11 RRT) AQ 73 mg .Math. ml.sup.1, pH 3.9 Solubility 40 C./ 7 d - Sticky solid, XRPD consistent with Maleate 75% RH Pattern 1

(194) XRPD information on the maleate salt (Form M) of compound (I) is found in Table 12 and FIG. 23.

(195) Analysis suggested Maleate Pattern 1 to be a non-solvated mono maleate salt with a melt at approximately 116 C. Material giving Maleate Pattern 1 appears to be hygroscopic above 75% RH, deliquescence was observed at 25 C./97% and a sticky solid, though consistent with Maleate Pattern 1 by XRPD, was obtained at 40 C./75% RH.

Example 14

Preparation of Gentisate Salt (Form N) of Compound (I)

(196) Free base compound (I) (400.27 mg, 1.01 mmol, 1 equiv), gentisic acid (155.25 mg, 1.01 mmol, 1 equiv) and EtOAc (8 ml, 20 relative volumes) were charged to a flask at RT. The mixture was stored in a shaker on a heat/cool cycle (40 C./RT, 4 h at each) for 63 h to give a light yellow solution. The mixture was cooled to 5 C. for 24 hno precipitation was observed. The mixture was cooled to 18 C. for 24 ha light yellow solid was formed. The solid was isolated by vacuum filtration using glassware pre-cooled to 18 C. Upon isolation and warming towards RT the sample became a yellow oil. The damp sample was dried in a vacuum oven at 30 C./3 mbar for 20 h to yield the gentisate salt as a yellow foam (160.61 mg, 29% based on mono-gentisate salt formation).

(197) Summary of Data for Gentisate (Form N) of Compound (I)

(198) TABLE-US-00019 Analysis Result .sup.1H NMR 1.1 equiv Gentisic acid, 11.8% EtOAc (~0.5 equiv) XRPD Amorphous mDSC Tg 29 C. (limits 14 and 42 C.)

(199) Formation of the desired partially crystalline Gentisate Pattern 1 was not observed, instead formation of an amorphous mono-gentisate salt was observed. In an attempt to crystallise the material samples were matured in a small range of solvent and the results are summarised in the table below.

Example 15

Preparation of Gentisate Salt (Form O) of Compound (I)

(200) The amorphous gentisate salt (Form N) as prepared in example 14 (15 mg) was dissolved in acetonitrile (20 vol) to give a solution at room temperature. The solvent was allowed to evaporate slowly to half the original volume to give crystalline gentisate salt (Form O).

(201) Summary of Data for Gentisate (Form O) of Compound (I)

(202) TABLE-US-00020 Analysis Result .sup.1H NMR 1.0 equiv Gentisic acid, 5.3% MeCN (0.5 equiv) DSC Endotherm, onset 92.1 C. (H 54 J .Math. g1) TGA 1.7% wt loss from 85 to 115 C.

(203) The XRPD peak list and labelled XRPD diffractogram for Gentisate Pattern 2, using Bruker AXS C2 GADDS instrument, can be found below in Table 13 and FIG. 24 respectively.

(204) These results suggest Gentisate Pattern 2 (Form O) material to be a mono Gentisate acid salt containing approximately 0.5 equiv of acetonitrile that is lost in conjunction with an endothermic event. Further analyses to elucidate if the sample is a hemi-solvate or contains trapped solvent were not possible due to the minor amount of material produced.

Example 16

Preparation of Fumarate Salt (Form P) of Compound (I)

(205) Free base compound (I) (400.18 mg, 1.01 mmol, 1 equiv), fumaric acid (116.71 mg, 1.01 mmol, 1 equiv) and EtOAc (4 ml, 10 relative volumes) were charged to a flask and stirred at RT for 16 h. A very thick white suspension was obtained. EtOAc (2 ml) was added and an aliquot was taken. The solid was isolated by vacuum filtration, dried under suction and analysed by XRPD (consistent with Fumarate Pattern 1). The remaining mixture was stirred at RT for a further 3 h. The solid was isolated by vacuum filtration, washed with EtOAc (32 ml), dried under suction and in a vacuum oven at 30 C./3 mbar for 15 h to yield the fumarate salt as a yellow solid (261.35 mg, 50% yield based on mono-fumarate salt formation).

(206) Summary of Data for Fumarate (Form P) of Compound (I)

(207) TABLE-US-00021 Analysis Result .sup.1H NMR 1.0 equiv fumaric acid, no residual solvent XRPD Consistent with Fumarate Pattern 1 DSC Sharp endotherm 139.5 C. (H 102 J .Math. g1) TGA 16.3% wt loss from 170 to 250 C. (~0.7 equiv Fumaric acid) HSM Melt observed with an onset of 135 C. HPLC 99.1% (largest % imp 0.28 @ 1.1 RRT) AQ 54 mg .Math. ml.sup.1, pH 4.1 Solubility 40 C./ 7 d - No visual change, XRPD consistent with 75% RH Fumarate Pattern 1 GVS Non-hygroscopic, reversible uptake of <0.4 wt % from 0 to 90% RH

(208) XRPD information on the fumarate salt (Form P) of compound (I) is found in Table 14 and FIG. 25.

(209) Analysis suggested Fumarate Pattern1 to be a non-solvated mono fumarate salt with a melt at approximately 140 C. The aqueous solubility was determined to be approximately 50 mg.Math.ml.sup.1 and was amongst the lowest of the salts obtained during this work. GVS analysis confirmed the material to be non-hygroscopic with less than 0.4 wt % water adsorbed from 0 to 90% RH.

Example 17

Preparation of L-Malate Salt (Form Q) of Compound (I)

(210) Free base compound (I) (400.19 mg, 1.01 mmol, 1 equiv), L-malic acid (135.35 mg, 1.01 mmol, 1 equiv) and EtOAc (20 ml, 50 relative volumes) were charged to a flask and stirred at RT for 24 h. Incomplete dissolution of the acid was observed. The mixture was heated to 40 C. and stirred for 1.5 h, then cooled to RT and stirred for a further 14 h. The mixture was cooled to 5 C. for 72 hno major precipitation was observed. The mixture was cooled to 18 C. for 30 hsome precipitate had formed. The sample was stirred at RT for 1 h to yield a mobile suspension. The solid was isolated by vacuum filtration, washed with EtOAc (23 ml), dried under suction and in a vacuum oven at 30 C./3 mbar for 14 h to yield the L-malate salt as a white solid (325.99 mg, 61% yield based on mono-L-malate salt formation).

(211) Summary of Data for L-Malate (Form Q) of Compound (I)

(212) TABLE-US-00022 Analysis Result .sup.1H NMR 1.0 equiv malic acid, 0.6% residual EtOAc XRPD Consistent with L-malate pattern 1 DSC Endotherm 80.9 C. (H 30 J .Math. g.sup.1) TGA 18.7% wt loss from 150 to 250 C. (~0.7 equiv Malic acid) HSM Melt observed with an onset of 83 C. HPLC 98.8% (largest % imp 0.52 @ 1.1 RRT) AQ 92 mg .Math. ml.sup.1, pH 4.3 Solubility 40 C./ 15 h - Deliquescence 75% RH 7 d - White solid, XRPD showed crystallisation of a new pattern; L- malate Pattern 2 (Form R)

(213) XRPD information on the L-malate salt (Form Q) of compound (I) is found in Table 15 and FIG. 26.

(214) Analysis suggested L-malate Pattern1 material to be a non-solvated mono malate salt with a melt at approximately 81 C. L-malate Pattern 1 material appeared to be hygroscopic, as deliquescence occurred within 15 h at 40 C./75% RH. However, upon extended storage at 40 C./75% RH (72 h) crystallisation of a new solid form was observed. XRPD analysis confirmed this material to be inconsistent with L-malate Pattern 1 (Form Q) and was given the identification of L-malate Pattern 2 (Form R).

Example 18

Preparation of L-Malate Salt (Form R) of Compound (I)

(215) A sample of L-malate Form Q (75.45 mg) was stored at 40 C./75% RH for 72 hdeliquescence was observed inside 2 h and crystallisation had occurred inside 72 h. The sample was dried in a vacuum oven at 40 C./3 mbar for 18 h to yield the L-malate (Form R) as a white solid. NMR analysis of the sample of Form R formed during the storage of Form Q at 40 C./75% RH confirmed it to be a mono-malate salt. Interestingly, DSC analysis appeared to suggest the material to be non-solvated with a higher melting point (103.5 C.) than L-malate Form Q.

(216) Summary of Data for L-Malate (Form R) of Compound (I)

(217) TABLE-US-00023 Analysis Result .sup.1H NMR 1.1 equiv Malic acid, no residual solvent DSC Endotherm 103.5 C. (81 J .Math. g.sup.1)

(218) Formation of a larger quantity of L-malate Pattern 2 was attempted. A sample of compound (75.45 mg) was stored at 40 C./75% RH for 72 hdeliquescence was observed inside 2 h and crystallisation had occurred inside 72 h. The sample was dried in a vacuum oven at 40 C./3 mbar for 18 h to yield the product as a white solid.

(219) Summary of Data for L-Malate (Form R) of Compound (I)Scale Up

(220) TABLE-US-00024 Analysis Result .sup.1H NMR 1.1 equiv Malic acid, no residual solvent XRPD Consistent with L-malate Pattern 2 (Form R) DSC Endotherm 103.6 C. (H 84 J .Math. g.sup.1) TGA 17.5% wt loss from 175 to 255 C. (~0.6 equiv Malic acid) GVS Reversible uptake of ~2.5 wt % H.sub.2O from 0 to 90% RH

(221) XRPD information on the L-malate salt (Form R) of compound (I) is found in Table 16 and FIG. 27.

(222) These results appear to confirm L-malate Pattern 2 (Form R) to be an anhydrous mono L-malate salt. GVS analysis showed the material to be less hygroscopic than the benzensulfonate and citrate salts.

(223) Various modifications and variations of the described aspects of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes of carrying out the invention which are obvious to those skilled in the relevant fields are intended to be within the scope of the following claims.

(224) TABLE-US-00025 TABLE 1 XRPD peaks for crystalline free base (Form A) of compound (I) Pos. Height FWHM d-spacing Rel. Int. Tip width [2Th.] [cts] [2Th.] [] [%] [2Th.] 7.5313 21162.64 0.0768 11.73852 100.00 0.0921 9.6026 846.92 0.0768 9.21066 4.00 0.0921 10.2275 400.15 0.1023 8.64925 1.89 0.1228 11.2954 1863.94 0.1023 7.83384 8.81 0.1228 11.6652 300.53 0.1023 7.58631 1.42 0.1228 12.2672 3812.02 0.1023 7.21534 18.01 0.1228 12.6242 497.83 0.1023 7.01205 2.35 0.1228 13.1780 953.85 0.1023 6.71859 4.51 0.1228 14.0653 4092.77 0.1023 6.29672 19.34 0.1228 14.8535 1458.15 0.0768 5.96431 6.89 0.0921 15.1515 343.64 0.0768 5.84765 1.62 0.0921 15.5775 2894.50 0.1279 5.68868 13.68 0.1535 16.9914 2108.17 0.1023 5.21838 9.96 0.1228 17.6862 1501.97 0.1279 5.01490 7.10 0.1535 18.3040 644.51 0.0591 4.84701 3.05 0.0709 18.3954 1212.49 0.0768 4.82314 5.73 0.0921 18.6301 1666.18 0.1023 4.76289 7.87 0.1228 18.9784 1639.81 0.1279 4.67626 7.75 0.1535 19.3292 475.31 0.1023 4.59219 2.25 0.1228 20.2061 1067.39 0.1023 4.39483 5.04 0.1228

(225) TABLE-US-00026 TABLE 2 XRPD peaks for L-tartrate salt (Form D) of compound (I) Angle Intensity 2-Theta % 3.82 51.2 7.57 100.0 8.12 14.6 10.53 17.1 11.39 15.0 12.00 15.3 13.54 27.4 15.15 52.6 16.35 20.2 16.88 41.0 17.37 21.7 18.51 24.8 18.98 20.6 19.77 69.9 21.06 31.6 22.70 55.2 23.47 77.1 24.66 43.7 28.73 59.2

(226) TABLE-US-00027 TABLE 3 XRPD peaks for citrate salt (Form F) of compound (I) Angle Intensity 2-Theta % 5.14 21.4 7.73 17.5 10.24 30.8 12.70 19.8 13.06 23.2 14.42 10.0 15.30 22.7 15.98 11.4 16.74 23.9 17.24 33.1 18.05 42.3 19.04 35.8 20.23 17.0 21.04 22.7 22.45 40.3 22.75 37.9 24.01 82.5 25.43 100.0 26.51 29.7 27.48 24.0 28.77 28.8 29.71 27.0

(227) TABLE-US-00028 TABLE 4 XRPD peaks for benzenesulfonate salt (Form G) of compound (I) Angle Intensity 2-Theta % 5.72 17.5 11.45 22.8 11.79 24.3 15.56 100.0 16.57 61.1 18.04 63.9 19.14 52.7 20.02 71.1 21.05 57.5 22.80 85.1 23.16 86.6 24.44 68.9 25.40 49.7 28.74 54.7

(228) TABLE-US-00029 TABLE 5 XRPD peaks for phosphate salt from ethanol (Form C) of compound (I) Angle Intensity 2-Theta % 6.49 100.0 8.91 14.7 9.75 16.8 10.52 15.6 13.03 7.1 15.44 11.7 16.27 68.6 17.85 15.5 18.29 20.7 19.52 72.1 20.02 32.5 21.11 20.7 22.80 21.1 24.92 47.1 28.33 18.6 29.41 21.4

(229) TABLE-US-00030 TABLE 6 XRPD peaks for phosphate salt from IPA (Form B) of compound (I) Angle Intensity 2-Theta % 6.46 100.0 8.88 15.3 9.67 14.6 10.47 13.3 12.78 9.5 15.33 9.3 16.12 48.1 16.82 9.8 18.13 22.5 19.38 46.8 19.95 19.2 20.97 17.9 24.11 17.4 24.83 28.7 26.54 14.7 28.11 16.8

(230) TABLE-US-00031 TABLE 7 XRPD peaks for L-tartrate salt (Form E) of compound (I) Pos. Height FWHM d-spacing Rel. Int. Tip width [ 2Th.] [cts] [ 2Th.] [] [%] [ 2Th.] 6.6675 15483.76 0.0768 13.25733 100.00 0.0921 8.2340 241.15 0.1023 10.73824 1.56 0.1228 9.7722 479.22 0.1023 9.05118 3.09 0.1228 11.9598 926.89 0.1023 7.40005 5.99 0.1228 12.3792 494.24 0.0768 7.15029 3.19 0.0921 13.0632 4104.92 0.0768 6.77739 26.51 0.0921 13.3777 2386.00 0.1023 6.61876 15.41 0.1228 13.9359 413.30 0.0768 6.35490 2.67 0.0921 14.9035 1349.55 0.1023 5.94439 8.72 0.1228 15.4032 975.17 0.0768 5.75266 6.30 0.0921 15.9507 949.23 0.1023 5.55642 6.13 0.1228 16.2665 488.77 0.1023 5.44926 3.16 0.1228 16.5423 792.08 0.1023 5.35902 5.12 0.1228 17.3614 2687.54 0.1023 5.10799 17.36 0.1228 17.5690 1410.91 0.1023 5.04809 9.11 0.1228 17.8630 201.26 0.1023 4.96566 1.30 0.1228 19.6395 1756.56 0.0768 4.52032 11.34 0.0921 19.8636 777.97 0.0768 4.46982 5.02 0.0921 20.1195 549.42 0.1023 4.41355 3.55 0.1228 20.7288 1423.91 0.1279 4.28518 9.20 0.1535 21.1373 389.18 0.1279 4.20327 2.51 0.1535 21.5804 674.89 0.1535 4.11797 4.36 0.1842 22.5683 459.02 0.1535 3.93989 2.96 0.1842 22.9541 780.05 0.1279 3.87454 5.04 0.1535 23.2869 904.34 0.1023 3.81992 5.84 0.1228 23.5693 1652.40 0.1535 3.77478 10.67 0.1842 24.0730 899.56 0.1535 3.69692 5.81 0.1842 24.6316 316.32 0.1791 3.61434 2.04 0.2149 25.2971 1357.36 0.1535 3.52074 8.77 0.1842 26.3772 346.67 0.1023 3.37898 2.24 0.1228 27.0905 141.69 0.1023 3.29160 0.92 0.1228 27.6723 474.86 0.1023 3.22371 3.07 0.1228 27.9727 708.87 0.1535 3.18977 4.58 0.1842 28.9051 262.52 0.1535 3.08896 1.70 0.1842 29.2843 136.18 0.1535 3.04982 0.88 0.1842 30.0801 73.71 0.1535 2.97092 0.48 0.1842 30.4059 137.17 0.1279 2.93982 0.89 0.1535 31.9006 27.79 0.1535 2.80541 0.18 0.1842 34.4898 70.18 0.2047 2.60050 0.45 0.2456

(231) TABLE-US-00032 TABLE 8 XRPD peaks for Hydrochloride (Form H) of compound (I) Angle/2-Theta Intensity/% 5.6 100.0 8.6 8.0 9.5 6.2 10.9 10.2 11.2 12.2 12.7 20.5 13.0 8.9 14.3 5.9 16.0 10.9 17.3 8.0 17.7 9.4 18.8 15.9 19.1 9.8 20.3 8.5 20.7 9.4 22.9 8.1 23.6 10.1 24.5 10.7 25.0 8.4 25.5 9.9 25.8 10.2 26.4 8.4 29.1 10.2

(232) TABLE-US-00033 TABLE 9 XRPD peaks for Hydrochloride (Form I) of compound (I) Angle/2-Theta Intensity/% 4.9 41.7 6.4 100.0 7.5 65.5 12.1 19.0 14.4 30.3 19.8 25.1 21.5 31.0 23.4 26.2 25.7 28.8

(233) TABLE-US-00034 TABLE 10 XRPD peaks for Hydrobromide (Form J) of compound (I) Angle/2-Theta Intensity/% 6.4 100.0 7.2 57.0 12.0 20.5 14.4 38.7 17.1 22.7 19.6 35.3 21.4 37.9 25.5 40.4

(234) TABLE-US-00035 TABLE 11 XRPD peaks for Mesylate (Form L) of compound (I) Angle/2-Theta Intensity/% 6.3 100.0 7.9 6.4 12.5 18.6 13.4 3.7 14.6 9.5 15.9 5.4 16.5 26.4 17.5 23.5 18.1 18.6 18.7 4.1 19.3 5.5 20.0 12.9 20.6 10.5 20.9 8.3 21.7 20.2 22.6 18.0 23.8 9.8 24.5 5.6 25.1 10.5 25.5 7.0 26.1 10.4 27.5 5.1 29.1 4.9 29.7 6.3 30.3 4.2

(235) TABLE-US-00036 TABLE 12 XRPD peaks for Maleate (Form M) of compound (I) Angle/2-Theta Intensity/% 3.8 21.7 7.6 100.0 8.5 28.0 10.8 5.6 11.4 8.6 12.2 37.2 15.2 48.5 15.8 11.1 17.0 10.0 18.0 18.1 18.8 13.5 19.4 24.1 20.3 8.9 21.6 6.0 22.6 6.2 23.6 23.2 24.3 22.0 24.8 20.5 26.0 14.7 27.2 10.6 27.9 16.2 28.2 8.6 28.8 4.8 29.9 5.9 30.2 5.0 31.7 5.1 32.7 4.2 33.2 4.7

(236) TABLE-US-00037 TABLE 13 XRPD peaks for Gentisate (Form O) of compound (I) Angle/2-Theta Intensity/% 6.318 20.5 12.161 21.3 12.449 19.6 13.13 59.5 14.414 21.2 14.831 18.5 16.369 17.2 17.122 18.1 18.787 25.7 19.485 19 20.419 17.2 23.371 100 23.769 54.1

(237) TABLE-US-00038 TABLE 14 XRPD peaks for Fumarate (Form P) of compound (I) Angle/2-Theta Intensity/% 3.8 14.1 7.7 100.0 8.1 14.6 8.8 44.6 10.2 11.2 11.3 9.9 13.1 45.4 15.2 28.7 15.5 16.7 16.5 27.1 17.7 60.0 19.1 13.7 19.6 10.9 20.0 6.0 20.9 5.0 21.5 5.4 21.9 11.2 22.7 9.9 23.3 28.6 23.8 25.4 24.1 18.3 25.0 20.0 25.3 9.6 26.7 17.7 27.9 8.8 28.9 8.9

(238) TABLE-US-00039 TABLE 15 XRPD peaks for L-malate (Form Q) of compound (I) Angle/2-Theta Intensity/% 6.7 39.7 8.6 60.0 9.3 44.4 11.0 19.7 12.7 100.0 13.6 41.4 14.1 37.1 15.1 47.5 15.8 27.5 16.5 26.9 17.8 38.2 18.7 56.4 19.5 41.9 19.8 43.9 21.2 28.2 22.5 33.2 23.5 29.4 24.9 37.3 25.7 42.3

(239) TABLE-US-00040 TABLE 16 XRPD peaks for L-malate (Form R) of compound (I) Angle/2-Theta Intensity/% 6.773 100 9.851 10.7 12.191 6.9 13.364 39.3 13.587 37.3 14.145 6.3 15.883 17.5 16.44 22.9 17.331 22.2 17.753 17.8 19.733 27.6 20.118 13.4 20.496 29.8 20.837 15.7 21.299 12.4 22.231 29.3 23.273 28.7 23.828 11.5 24.19 15 24.606 11.8 25.182 14.7 25.673 13.6 26.026 23.3 26.31 13.9 26.905 10.2 27.779 9.9 28.775 17.9 31.109 8.9 32.352 7 33.239 9.3

(240) TABLE-US-00041 TABLE 17 XRPD peaks for bromide (Form K) of compound (I) Angle/2-Theta Intensity/% 5.732 100 16.386 32.1 17.675 31.8 18.369 46.3 19.633 45.9 20.542 59 24.136 85.3 25.272 59.5 25.976 46.4 28.119 45.2