Preparation of psilocybin, different polymorphic forms, intermediates, formulations and their use

Abstract

This invention relates to the large-scale production of psilocybin for use in medicine. More particularly, it relates to a method of obtaining high purity crystalline psilocybin, particularly, in the form of Polymorph A. It further relates to a method for the manufacture of psilocybin and intermediates in the production thereof and formulations containing psilocybin.

Claims

1. Crystalline psilocybin characterized by X-ray powder diffraction (XRPD) peaks at 11.5±0.1, 12.0±0.1, 14.5±0.1, 17.5±0.1, and 19.7±0.1 °2θ, wherein the crystalline psilocybin has a chemical purity of greater than 98% and no single impurity of greater than 1% as determined by high-performance liquid chromatography (HPLC) analysis.

2. The crystalline psilocybin of claim 1, further characterized by XRPD peaks at least one of 20.4±0.1, 22.2±0.1, 24.3±0.1, or 25.7±0.1 °2θ.

3. The crystalline psilocybin of claim 1, further characterized by a XRPD diffraction pattern that is substantially the same as shown in FIG. 7a.

4. The crystalline psilocybin of claim 1, having a chemical purity of greater than 99%.

5. The crystalline psilocybin of claim 1, further characterized by a water content of <0.5% w/w as determined by Karl Fisher Titration.

6. The crystalline psilocybin of claim 1, having less than 1% psilocin as determined by HPLC analysis.

7. The crystalline psilocybin of claim 1, having less than 0.5% psilocin as determined by HPLC analysis.

8. The crystalline psilocybin of claim 1, having less than 1% phosphoric acid as determined by .sup.31P nuclear magnetic resonance (.sup.31P NMR).

9. The crystalline psilocybin of claim 1, having less than 0.5% phosphoric acid as determined by .sup.31P NMR.

10. The crystalline psilocybin of claim 1, further characterized by a <0.5% w/w loss in the thermogravimetric analysis (TGA) thermogram between 25° C. and 200° C.

11. The crystalline psilocybin of claim 1, further characterized by an endothermic event in a differential scanning calorimetry (DSC) thermogram having an onset temperature of between 205° C. and 220° C.

12. The crystalline psilocybin of claim 1, further characterized by an endothermic event in a DSC thermogram having an onset temperature of between 145° C. and 155° C.

13. The crystalline psilocybin of claim 1, further characterized by one or more of the following: a) a loss on drying of no more than 2% w/w; b) residue on ignition of no more than 0.5% w/w; c) assay (on a dry basis) of 95-103% by weight as measured by HPLC; d) residual solvent content of no more than 3000 ppm methanol, 5000 ppm ethanol, 720 ppm tetrahydrofuran (THF), and 890 ppm toluene, as measured by high resolution gas chromatography (HRGC); e) Inductively Coupled Plasma Mass Spectrometry (ICP-MS) elemental analysis of: i. no more than 1.5 ppm cadmium (Cd); ii. no more than 1.5 ppm lead (Pb); iii. no more than 4.5 ppm arsenic (As); iv. no more than 9.0 ppm mercury (Hg); v. no more than 15 ppm cobalt (Co); vi. no more than 30 ppm vanadium (V); vii. no more than 60 ppm nickel (Ni); viii. no more than 165 ppm lithium (Li); and ix. no more than 30 ppm lead (Pd).

14. The crystalline psilocybin of claim 1, comprising crystals with a size ranging from 50 to 200 microns.

15. A pharmaceutical composition comprising a therapeutically effective amount of the crystalline psilocybin of claim 1 and a pharmaceutically acceptable excipient.

16. The pharmaceutical composition of claim 15, wherein the pharmaceutical composition is a capsule.

17. The pharmaceutical composition of claim 15, wherein the pharmaceutical composition is a tablet.

18. The pharmaceutical composition of claim 15, comprising about 1 mg to about 40 mg of the crystalline psilocybin.

19. The pharmaceutical composition of claim 15, comprising about 1 mg of the crystalline psilocybin.

20. The pharmaceutical composition of claim 15, comprising about 25 mg of the crystalline psilocybin.

21. The pharmaceutical composition of claim 15, wherein the crystalline psilocybin has a chemical purity of greater than 99%.

22. The pharmaceutical composition of claim 15, wherein the crystalline psilocybin has content of <0.5% w/w as determined by Karl Fisher Titration.

23. The pharmaceutical composition of claim 15, having less than 1% psilocin as determined by HPLC analysis.

24. The pharmaceutical composition of claim 15, having less than 1% phosphoric acid as determined by .sup.31P NMR.

25. The pharmaceutical composition of claim 19, wherein the crystalline psilocybin has a chemical purity of greater than 99%.

26. The pharmaceutical composition of claim 19, having less than 1% psilocin as determined by HPLC analysis.

27. The pharmaceutical composition of claim 19, having less than 1% phosphoric acid as determined by .sup.31P NMR.

28. The pharmaceutical composition of claim 20, wherein the crystalline psilocybin has a chemical purity of greater than 99%.

29. The pharmaceutical composition of claim 20, having less than 1% psilocin as determined by HPLC analysis.

30. The pharmaceutical composition of claim 20, having less than 1% phosphoric acid as determined by .sup.31P NMR.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which:

(2) FIG. 1 is a schematic of the reaction taught in JNP;

(3) FIG. 2 is a schematic of the Stage 1 reaction of one aspect of the present invention;

(4) FIG. 3 is a schematic of the Stage 2 reaction of one aspect of the present invention;

(5) FIG. 4 is a schematic of the Stage 3 reaction of one aspect of the present invention;

(6) FIG. 5 is a schematic of the Stage 4 reaction of one aspect of the present invention;

(7) FIG. 6 is a schematic of the Stage 5 reaction of one aspect of the present invention;

(8) FIG. 7a is a XRPD diffractogram of Polymorph A (GM764B);

(9) FIG. 7b is a XRPD diffractogram of Polymorph A′ (JCCA2160F);

(10) FIG. 7c is a XRPD diffractogram of Polymorph B; (JCCA2160-F-TM2);

(11) FIG. 7d is a XRPD diffractogram of a Hydrate A (JCCA2157E);

(12) FIG. 7e is a XRPD diffractogram of an ethanol solvate (JCCA2158D);

(13) FIG. 7f is a XRPD diffractogram of product obtained during development of the process (CB646-E) (top)—compared to the diffractograms Polymorph A′ (JCCA2160F) (middle) and Polymorph B (JCCA2160-TM2) (bottom);

(14) FIG. 8a is a DSC and TGA thermograph of Polymorph A (GM764B);

(15) FIG. 8b is a DSC and TGA thermograph of Polymorph A′ (JCCA2160F);

(16) FIG. 8c is a DSC thermograph of Polymorph B (GM748A);

(17) FIG. 8d is a DSC and TGA thermograph of Hydrate A (JCCA2157E);

(18) FIG. 8e is a DSC and TGA thermograph of ethanol solvate (JCCA2158D);

(19) FIG. 9 is a form phase diagram showing the inter-relationship of form in water-based systems;

(20) FIG. 10 is a 1H NMR spectrum of Psilocybin; (Read alongside assignment Example 7);

(21) FIG. 11 is a .sup.13C NMR spectrum of Psilocybin; (Read alongside assignment Example 7);

(22) FIG. 12 is a FT-IR Spectrum of Psilocybin;

(23) FIG. 13 is a Mass Spectrum of Psilocybin;

(24) FIG. 14 is a numbered structural formula of Psilocybin;

(25) FIG. 15 is a temperature solubility curve for Psilocybin in water;

(26) FIG. 16a is a micrograph showing crystals obtained by controlled cooling;

(27) FIG. 16b is a micrograph showing crystals obtained by uncontrolled cooling drying;

(28) FIG. 17 is a form phase diagram showing the inter-relationship of forms in different solvent systems;

(29) FIG. 18 is XRPD diffractogram—Pattern C for solids isolated at 25 and 50° C.;

(30) FIG. 19 is XRPD diffractograms—Patterns D, E and F for solids isolated at 25 and 50° C.;

(31) FIG. 20 is a comparison of the XRPD diffractograms acquired for the solids isolated from the equilibration of amorphous Psilocybin in solvents A to H;

(32) FIG. 21 is a comparison of the XRPD diffractograms acquired for the solids isolated from the equilibration of amorphous Psilocybin in solvents I to P; and

(33) FIG. 22 is a comparison of the XRPD diffractograms acquired for the solids isolated from the equilibration of amorphous Psilocybin in solvents R to Y.

DETAILED DESCRIPTION

(34) In contrast to the prior art, the present invention sought to produce psilocybin at a commercial large scale, in amounts or batches of at least 100 g, and more preferably at least 250 g, levels 1 log or 2 logs higher than the levels described in JNP, which describes a “large” scale method to producing gram quantities on a 10 g scale.

(35) To demonstrate the many significant development steps from JNP, the description below sets out details of experiments and investigations undertaken at each of the process stages, which illustrate the selections made to overcome the numerous technical problems faced, in producing psilocybin (7) to GMP at a large scale (including the various intermediates (2-6)) starting from 4-hydroxyindole (1).

(36) Reference to a particular numerical value includes at least that particular value, unless the context clearly dictates otherwise. When a range of values is expressed, another embodiment includes from the one particular value and/or to the other particular value. Further, reference to values stated in ranges include each and every value within that range. All ranges are inclusive and combinable.

(37) When values are expressed as approximations, by use of the antecedent “about,” it will be understood that the particular value forms another embodiment.

(38) As used herein, the singular forms “a,” “an,” and “the” include the plural.

(39) The term “about” when used in reference to numerical ranges, cut-offs, or specific values is used to indicate that the recited values may vary by up to as much as 10% from the listed value. As many of the numerical values used herein are experimentally determined, it should be understood by those skilled in the art that such determinations can, and often times will, vary among different experiments. The values used herein should not be considered unduly limiting by virtue of this inherent variation. Thus, the term “about” is used to encompass variations of ±10% or less, variations of ±5% or less, variations of ±1% or less, variations of ±0.5% or less, or variations of ±0.1% or less from the specified value.

(40) As used herein, “treating” and like terms refer to reducing the severity and/or frequency of symptoms, eliminating symptoms and/or the underlying cause of said symptoms, reducing the frequency or likelihood of symptoms and/or their underlying cause, delaying, preventing and/or slowing the progression of diseases and/or disorders and improving or remediating damage caused, directly or indirectly, by the diseases and/or disorders.

(41) The following abbreviations have been used herein: DSC—Differential Scanning Calorimetry RT—room temperature TBME—methyl tert-butyl ether TGA—Thermogravimetric Analysis THF—tetrahydrofuran wrt—with respect to XRPD—X-Ray Powder Diffraction

Example 1

Stage 6 Crystallisation Process and Resulting Polymorphs

(42) Experimental to Produce Form A′:

(43) 1.0 g of crude Psilocybin was charged to a 25 mL flask. Water (12.8 mL/16 volumes based on activity of input material) was added. The mixture was agitated and heated to 80° C. A dark brown solution with visible undissolved solids was obtained. The mixture was polish filtered through a warmed 0.45 μm filter into a hot 25 mL flask. The undissolved solids were removed to give a dark brown solution. The solution was re-equilibrated at 75° C. and then cooled slowly (10° C./hour) to ambient temperature. The resulting pale brown solution was equilibrated at ambient temperature for 16 hours. The suspension was cooled to 5° C. prior to isolation of the solid by vacuum filtration. The filter cake was washed with water (0.8 mL/1 volume) and dried in vacuo at 50° C. for 16 hours. Yield of 75%, chemical purity 99%, NMR assay >98%.

(44) The procedure above was repeated with 14 volumes (11.2 mL) of water. Yield of 69%, chemical purity 99%, NMR assay >98%.

(45) In both cases, dissolution of crude Psilocybin was achieved at ca. 75° C. On gradual cooling, precipitation was observed at ca. 60° C.

(46) In both cases, psilocybin Polymorph A′ was produced, confirmed by XRPD (diffractogram consistent with FIG. 7b) and DSC (thermogram consistent with FIG. 8b).

(47) Experimental to Produce Form A:

(48) 94 g of crude Psilocybin obtained from the Stage 5 process (about 93% pure by HPLC with about 4% pyrophosphate impurity) was subject to an aqueous re-crystallisation as set out below:

(49) The protocol used sufficient water (12 volumes), a rapid agitation rate (450 rpm) and a controlled cooling profile (10° C./hr).

(50) Psilocybin (94.0 g) (CB650E) was charged to a 2 L flask. Water (902 ml, 12 volumes based upon activity of input material) was added. The mixture was agitated and heated to about 78° C. A dark brown solution with visible undissolved solids was obtained. The mixture was polish filtered through a 1.2 μm in-line filter into a hot 5 L flask fitted with an overhead stirrer (450 rpm). The undissolved solids were removed to give a clarified dark brown solution. The solution was re-equilibrated at about 75° C. for 15 minutes and then cooled slowly (10° C./hour) to ambient temperature. The solution was seeded with Psilocybin Hydrate A (GM758A-XRPD diffractogram consistent with FIG. 7d) following maturation in water) at 68° C.-70° C. The resulting pale brown suspension was equilibrated at ambient temperature for about 16 hours. The suspension was cooled to 5° C. for one hour prior to isolation of the solid by vacuum filtration. The filter cake was washed with water (282 mL, 3 volumes) and dried in vacuo at about 50° C. for 30 hours.

(51) The process was completed successfully with a yield of 75% achieved. Chemical purity of the solid was confirmed as 99.3%. Analysis of the solid by XRPD post drying for 30 hours showed Polymorph A (FIG. 7a). A characteristic perturbation was observed at ca ˜17 °2θ, such as 17.5° 2θ, and was pronounced in the bulk material.

(52) Solid State Characterisation of Polymorph a and Polymorph A′

(53) The DSC and TGA thermograms (FIG. 8a) obtained for Polymorph A were comparable to the DSC and TGA thermograms obtained for Polymorph A′ (FIG. 8b). The TGA thermograms (FIG. 8a) obtained for Polymorph A and Polymorph A′ show no weight loss prior to decomposition. This suggested that the difference between the XRPDs obtained for Polymorph A (FIG. 7a; perturbation present at ca. ˜17 °2θ) and for Polymorph A′ which was obtained at small scale (FIG. 7b; perturbation not present) was not due to excess hydration.

(54) Microscopy of the solid (FIG. 16a) shows rod shaped crystals with good uniformity with a size range of between 50 and 200 micron.

(55) The XRPD diffractogram obtained for Polymorph A′ does not demonstrate a perturbation at ca. ˜17 °2θ to the same extent as Polymorph A. The perturbation in the XRPD diffractogram at ca. ˜17 °2θ is more pronounced for psilocybin produced at large scale (compared to that obtained at small scale) and was unexpected. Applicant has demonstrated that the Hydrate A is the only polymorphic form that exists across a range of temperatures with no diffraction peak in the 17 °2 theta region (see FIG. 7M. This strongly suggests a collapse of Hydrate A upon dehydration to yield Polymorph A or A′ that varies with scale and that Polymorph A is the true form with Polymorph A′ formed at a small scale being atypical.

(56) To test the robustness of this theory and to demonstrate a return to Polymorph A, a small portion of the bulk was re-dried following another soak in water (to reproduce Hydrate A). A small sample (250 mg—psilocybin Polymorph A) was equilibrated in water (10 vols) for one hour. The suspension was filtered and analysis of the damp solid confirmed that Hydrate A had been generated (FIG. 7d), no perturbation at 17 °2 theta. The material was dried in vacuo for 16 hours and the solid reassessed by XRPD. Polymorph A′ material was confirmed by XRPD (FIG. 7b) with the reduction in the XRPD perturbation noted. Additional drying of the original bulk solid and ageing at ambient temperature did not change the XRPD diffractogram of the solid. The two solid versions obtained, the XRPD diffractograms for Polymorph A and Polymorph A′ are virtually identical other than the ˜17.5° 2 theta peak. The thermal properties are also identical. The distinction between the XRPD diffractograms for Polymorph A versus Polymorph A′ is subtle and both polymorphs kinetically convert to the hydrated state very rapidly.

(57) Additional experiments were performed to ascertain if the differences in the XRPD diffractograms for Polymorph A and Polymorph A′ were due to the larger scale crystallisation process delivering solids of a larger particle size that subsequently did not dry as effectively and caused the change, or whether the habit and size difference of the crystalline solid was the cause. Psilocybin Polymorph A (polymorphic form confirmed prior to experiment) was ground via a mortar and pestle and assessed by XRPD. No change in polymorph was observed. Another portion (51 mg) was charged with water (<1 mL) and assessed damp to confirm that the hydrate was formed. Both lots were dried in vacuo at 50° C. for ca. 18 hours and re-assessed by XRPD. The ground sample remained as Polymorph A. The hydrated sample after dehydration was shown to be Polymorph A′ (i.e., no reflection at ˜17.5 °2θ). This suggested that size/habit alone were not the sole reason for the original reflection peaks.

(58) TGA assessment revealed that the input lot demonstrated a small mass loss (0.139% weight) by ca. 70° C. The particle size reduced and subsequently dried solid demonstrated a greater mass loss of 0.343 wt % by ca. 75° C. whereas the hydrated and dried solid demonstrated the smallest mass loss of 0.069 wt % by ca. 80° C. The particle size reduced and subsequently dried solid was held at 80° C. for 10 minutes (past the point of mass loss by TGA) but assessment by XRPD revealed no change from the input, meaning that low levels of hydration and partial swelling of the crystalline lattice were not the cause of the variation

(59) It is possible to generate Polymorph A′ via the hydration of Polymorph A and subsequent drying of the isolated solid on a small scale.

(60) Psilocybin Polymorph A and Polymorph A′, ca. 60 mg each, were charged with water, 0.2 ml, to deliver Hydrate A from both lots. Half of each Hydrate A was dried in vacuo at 25° C. for ca. 17% hours and the remainder of each Hydrate A was dried at ambient temperature under a stream of N.sub.2 for ca. 17% hours. The solids were isolated following drying and assessed by XRPD. XRPD assessment of the solids isolated from the Polymorph A input confirmed that Hydrate A was successfully generated and that the solids dried to give Polymorph A′ from both drying methods. XRPD assessment of the solids isolated from the Polymorph A′ input confirmed that Hydrate A was successfully generated and that the solids dried to remain as Polymorph A′ from both drying methods.

(61) On the small scale investigated, Polymorph A and Polymorph A′ will dry to give Polymorph A′ via conversion to Hydrate A.

(62) Psilocybin Polymorph A (100 mg) was particle size reduced via mortar and pestle grinding. The ground lot was subject to two different drying regimes in order to assess whether reducing the particle size affected the dehydration of the sample. The first sample was held at 80° C. for 10 minutes and the second sample held at 110° C. for 10 minutes. Both solids were assessed by XRPD which revealed that Polymorph A was retained. It was considered whether the ground lot in the prior isothermal stresses were not held at 110° C. for long enough to impact the form and so a portion of the ground lot was dried in vacuo at 110° C. for ca. 24 hours. Assessment by XRPD revealed a subtle change in form with the Polymorph A reflections at ca. 17 still present but at a slightly reduced intensity.

(63) It was concluded that Polymorph A would not readily convert to Polymorph A′ via particle size reduction and/or drying at high temperature.

(64) Methodology

(65) Stability assessments of Psilocybin, not containing the pyrophosphate impurity, indicated that at temperatures in excess of 80° C., the level of the Stage 3 intermediate impurity (psilocin) generated by hydrolysis of Psilocybin is of concern. For example, when a 83 mg/mL Psilocybin aqueous solution is heated to 90° C. and analysed by HPLC at 1, 2 and 4 hours, the level of the Stage 3 impurity were determined as 0.28, 1.82 and 7.79 area % respectively. In comparison, when 50 mg of psilocybin is dissolved in water (1.2-1.8 ml; volume sufficient to maintain a solution) and heated to 70, 75 and 80° C. for 4 hours, the level of the Stage 3 impurity was determined, by HPLC, as 0.53, 0.74 and 2.30 area % respectively. The recrystallization heats the crude Psilocybin to between 75° C. and 80° C. in order to achieve dissolution, and polish filtration. The immediate cooling of the solution limits the level of Psilocybin hydrolysis by reducing the residency time of the material to excessive temperature.

(66) Further trial re-crystallisation's of Psilocybin were conducted introducing the following variations:

(67) Varying the volumes of water used;

(68) Varying agitation;

(69) Having a controlled cooling profile;

(70) Having a rapid (uncontrolled) cooling profile.

(71) Using smaller volumes of water (as little as 12 volumes) did not hinder the re-crystallisation process and dissolution of Psilocybin was achieved at a temperature to enable a polish filtration step. Different cooling rates were shown to result in different crystal size distributions; a slow controlled cool at ca 10° C. per hour produced a relatively larger and more even average crystal size (FIG. 16a) whereas a rapid cooling profile delivered smaller crystals (FIG. 16b). A controlled cooling profile is preferred and this was reflected in an improved purity for the controlled cool.

(72) Using the process resulted in Psilocybin of 99.3% chemical purity, with a 75% yield. Thermal characteristics of the solid corresponded with those desired. Differences in the XRPD diffractogram of the dry solid have suggested that the drying profile may be important in determining how the Hydrate A collapses to give the preferred solid form. Polymorph A has been demonstrated to be stable under accelerated stability testing conditions for 12 months.

(73) Experimental

(74) Stage 5 was charged to vessel under N.sub.2 followed by water (approx. 12-15 vol based on active Stage 5). The mixture was heated to about 80° C. to achieve dissolution and polish filtered through a 1.2 μm in-line filter into a clean new flask heated to 80° C. The agitation rate was set to a high level (450 rpm) and the solution equilibrated at 70-75° C. The solution was cooled to ambient temperature, at approx. 10° C./hour, seeding with Psilocybin Hydrate A (0.001×stage 5 charge) at 68-70° C. The suspension was held at ambient temperature overnight then cooled to approx. 5° C. and held for 1 hour. The suspension was filtered, washing with water (2-3 volumes based upon active charge of Stage 5). The pure Psilocybin was dried in vacuo at 50° C. Crystalline material psilocybin (Polymorph A or Polymorph A′ dependent on scale) was obtained, for example using 94 g input of psilocybin yielded Polymorph A and using 1 g input of psilocybin yielded Polymorph A′. Typically, batch sizes of greater than 5 g deliver Polymorph A, while batch sizes less than 5 g deliver Polymorph A′.

(75) The differences from JNP and the benefits can be summarised as follows:

(76) i) This additional crystallisation step gives rise to a defined crystalline form—Polymorph A (or A′).

(77) ii) Heating to about 80° C., for a short period, has the advantage that solubility is maximised (and hydrolysis avoided), which ensures good yields.

(78) iii) At about 70-80° C. polish filtration can be used to remove insoluble impurities. This is best achieved using an in-line filter—typically about 1.2 μm. This ensures good chemical purity.

(79) iv) Using a high agitation rate (typically about 450 rpm) ensures speedy dissolution allowing the time at which the solution is kept at 80° C. to be minimised, thus avoiding increased levels of the Stage 3 intermediate impurity formed by hydrolysis of Psilocybin.
v) The provision of controlled cooling, typically cooling at about 10° C. per hour, delivers a more uniform crystal size and maintains form as crystalline Hydrate A.
vi) Seeding the solution at about 70° C. with Psilocybin Hydrate A facilitates crystallisation as the Hydrate A.
vii) The crystals are washed in water and dried at about 50° C. to maximise purity and deliver Polymorph A or A′ depending on scale.

Examples 2 to 6

Stages 1 to 5 Production of Psilocybin

(80) The following Examples illustrate significant developments from the process described in JNP, and illustrated in FIG. 1, as described herein before.

Example 2

Stage 1 (FIG. 2)

(81) The stage 1 conditions in JNP used 1.1 eq Ac.sub.2O, and 1.2 eq Pyridine in the solvent DCM. The reaction was found to be complete (99% product, 0% SM) after stirring overnight. The reaction mixture was washed with water and concentrated in vacuo giving a brown oil. In the literature, the oil was taken up in EtOAc and concentrated by evaporation, giving precipitation of solids at low volume.

(82) Investigation

(83) However, in the Applicants hands precipitation of solids from EtOAc was not observed. Precipitation of solids was encouraged by trituration with heptane, however this would not form a scalable process. The solids were collected giving high purity stage 1 product (75% yield, >95% pure by NMR).

(84) While the reaction worked well, the isolation procedure required further development in order that an easy to handle solid could be obtained. It was hoped that isolation of the solids by filtration would then also offer a means of purification.

(85) The reaction was first trialed in EtOAc to see if precipitation of solids could be encouraged allowing isolation directly from the reaction mixture. However, the reaction profile in EtOAc was found to be less favourable than in DCM and therefore the reaction was abandoned.

(86) Applicant washed out the pyridine from the DCM reaction mixture, as it was believed this may be preventing re-crystallisation of the product. The reaction was repeated (completion of 0.4% SM, 98.7% product by HPLC) and the reaction mixture washed with 20% citric acid to achieve pH 2/3, removing pyridine and then saturated NaHCO.sub.3 (aq) to avoid low pH in the evaporation steps. The organics were dried and a solvent swap to heptane was carried out giving precipitation of stage 1. The solids were collected by filtration yielding pure stage 1 after drying in vacuo (87% yield, >95% purity by NMR).

(87) Stability trials were carried out that confirmed the reaction mixture was stable overnight when stirred with 20% citric acid, and also saturated NaHCO.sub.3. The product was found to be stable when oven dried at 40° C. and 60° C.

(88) Scale Up

(89) The stage 1 reaction was successfully scaled up processing >100 g of 4-hydroxyindole. The reaction progressed as expected and was worked up to give stage 1 product (93% yield, ˜98% NMR purity).

(90) GMP Raw Material Synthesis

(91) A large scale stage 1 reaction was carried out to supply GMP starting material (processing >500 g of 4-hydroxyindole). The reaction proceeded as expected giving consumption of SM by HPLC (99.2% product, <0.1% SM). The reaction was worked up using the established procedure to give stage 1 product after drying (94% yield, 99.1% by HPLC, 99% NMR assay).

(92) It was also noted in development that the stage 1 procedure was effective in removing minor impurities present in some batches of 4-hydroxyindole. The low level impurities present in 4-hydroxyindole were completely removed after the stage 1 reaction providing clean material in high yield (89%) and purity (99% by HPLC, 99% by NMR assay).

(93) Experimental

(94) 4-hydroxyindole (1 eq. limiting reagent) was charged to a vessel under N.sub.2 followed by DCM (dichloromethane; 6 vol based on 4-hydroxyindole charge). The reaction was cooled to 0-5° C. and pyridine added (1.2 eq) dropwise at 0-5° C. Acetic anhydride (1.1 eq) was added dropwise at 0-5° C. and the reaction warmed to 20-25° C. for 1-1.5 hrs and stirred at 20-25° C. for a further 3 hours. The reaction was sampled and analysed for completion. The reaction was then washed three times with 20% aqueous citric acid solution (3×3 vol based on 4-hydroxyindole charge) and once with sat. NaHCO.sub.3 (3 vol based on 4-hydroxyindole charge). The DCM solution was dried over MgSO.sub.4 and filtered and the DCM layer concentrated to half volume by distillation. Heptane (6 vol based on 4-hydroxyindole charge) was added and further DCM was removed by distillation until full precipitation of the Stage 1 had occurred. The reaction was cooled to 15-25° C. and the solids collected by filtration, washing with heptane (1 vol based on 4-hydroxyindole charge) dried under vacuum overnight at 60° C.

(95) The differences from JNP and the benefits can be summarised as follows:

(96) i) Applicant washed out the pyridine using citric acid at a pH of about 2-3. This facilitated improved isolation and crystallisation. In practice the DCM phase is separated and the aqueous citric acid phase discarded.

(97) ii) An additional wash in sodium bicarbonate resulted in further improvement.

(98) iii) A solvent swap to heptane improved solid precipitation maximising yield and resulting in reproducible high purity Stage 1.

Example 3

Stage 2 (FIG. 3)

(99) Step i—Acid Chloride Formation

(100) Formation of reactive Intermediate 2A by reaction of the stage 1 product with oxalyl chloride (1.5 eq) was initially trialed in a mixture of TBME and THF (6 vol/1 vol) to determine if it was a viable alternative to volatile and highly flammable Et.sub.2O as used in the literature. The reaction gave completion after ˜18 hours with a similar solubility profile to Et.sub.2O (stage 1 in solution, precipitation of stage 2A).

(101) As the acid chloride intermediate is prone to hydrolysis, leading to variable analytical results a more robust sample make up and analysis was developed in which the reaction was quenched into THF/NMe.sub.2 (to give stage 2) and then analysed by HPLC.

(102) The ratio of TBME and THF was optimised to give the highest purity and yield of intermediate with a preferred ratio of TBME:THF of 6:1 chosen for scale up. Other ratios of TBME:THF may be used.

(103) A scale up reaction was carried out using the preferred solvent mixture (1 vol THF, 6 vol TBME) but with the oxalyl chloride addition carried out at 30-35° C. The resulting solution was then heated at 40° C. for 2.5 hrs giving a completion with ˜1% stage 1 product remaining. Carrying out the addition hot to maintain a solution ensures a high reaction rate and gave an improved level of completion with a much shorter reaction time (2.5 hrs vs overnight). The product was still seen to precipitate at temperature after ˜15 min and no detrimental effect on the reaction profile was observed by HPLC.

(104) As the stability of Intermediate 2A was not known, telescoping of material through to stage 2 was attempted rather than isolating the intermediate and risking degradation (hydrolysis). The reaction profile was complex with multiple components present at a low level. TBME was added and the precipitate collected. However, this was also found to be a complex mixture by HPLC/NMR.

(105) Due to the poor reaction profile it was deemed necessary to isolate Intermediate 2A to allow for purification away from excess oxalyl chloride. The reaction was repeated and the yellow precipitate collected by filtration and washed with TBME to remove the excess oxalyl chloride (80% yield). NMR analysis confirmed the product to be of sufficient purity (˜95% by NMR). However despite storing under nitrogen, some decomposition was noted over the following days giving partial hydrolysis including de-protection of the acetate group.

(106) In order to try and reduce the potential for hydrolysis of the intermediate acid chloride during isolation, further investigations into a telescoped procedure were carried out. It was found that by allowing the reaction mixture to settle the TBME liquors could easily be decanted and then the residual solids washed with further portions of TBME in a similar manner. This allowed purification of Intermediate 2A away from excess oxalyl chloride whilst minimising exposure to moisture.

(107) It was felt that some reaction yield may be lost due to partial solubility of the intermediate in the THF/TBME mixture. This was confirmed by adding heptane to the decantation liquors which gave precipitation of further solids. To limit this solubility, a heptane (8 vol) addition was made prior to decantation. Rather than washing the solids with TBME, heptane was also used for the washes (3×6 vol) which maximised the yield while maintaining the high purity of the intermediate. This methodology was successfully scaled up and is the preferred process.

(108) Step ii—Reaction with Dimethylamine

(109) The literature (Synthesis, 1999, 6, 935-938; D. E. Nichols) suggested HNMe.sub.2 gas was effective for this transformation. However to simplify large scale processing this was substituted for either solid HNMe.sub.2.HCl, with an additional excess of base, or a solution of HNMe.sub.2 in THF. JNP uses HNMe.sub.2 in the presence of excess base (pyridine).

(110) Initially isolated Intermediate 2A was used to optimise the reaction with dimethylamine via a series of trial reactions (see Table 11).

(111) TABLE-US-00011 TABLE 11 Stage 2 reaction optimisation trials Dimethylamine HPLC # Source Base Solvent completion Isolated solids 1 2M Pyridine THF/TBME 80% product, 64% yield, HNMe.sub.2/THF 1.3 eq ~70% pure 1.2 eq by NMR, 2 HNMe.sub.2•HCl K.sub.2CO.sub.3 THF/H.sub.2O 70% product 40% yield 1.5 eq ~98% by NMR. 3 2M Pyridine Et.sub.2O 81% product 63% yield, HNMe.sub.2/THF 3.6 eq ~90% by NMR. 1.33 eq 4 2M N/A Et.sub.2O 93% product 72% yield, HNMe.sub.2/THF ~98% by NMR. 2.9 eq

(112) The literature supplied conditions with pyridine (#1) were trialed along with a similar reaction in Et.sub.2O (#3), a biphasic reaction using Me.sub.2NH.HCl and aqueous K.sub.2CO.sub.3 (#2) and a reaction with excess 2M Me.sub.2NH in THF (#4). The major component by HPLC was the desired product in all cases with the conditions using aqueous base and excess Me.sub.2NH being generally much cleaner than those with pyridine. While significant hydrolysis product was seen in all cases this was thought to be the result of unreacted Intermediate 2A which was quenched during sample makeup for HPLC analysis. The reactions were worked up by addition of water and then the organic solvent was removed in vacuo giving precipitation of solids.

(113) The reaction with excess amine showed a much improved impurity profile which translated into a higher yield (72% vs 63%) and purity (98% vs 90%). This approach limited the water content of the reaction and therefore minimised the opportunity for hydrolysis to occur. Purification was also expected to be more facile due to the absence of pyridine in the isolated solids. For these reasons the conditions with excess HNMe.sub.2 as base were chosen for scale up.

(114) The reaction in Et.sub.2O gave a clean (#4) profile. However, to facilitate large scale processing it proved advantageous to switch to a less volatile solvent, such as TBME. This would facilitate telescoping the acid chloride into this reaction. For these reasons it was chosen to carry out the reaction in TBME using excess 2M Me.sub.2NH in THF.

(115) It was believed that the addition of water would aid the workup by solubilising the HNMe.sub.2.HCl salts that were present and resulted in a very thick mixture and slow filtrations. This was trialed. However, when water was added to the reactions in TBME and THF a poor recovery was obtained with analysis of the liquors showing additional impurities and extensive acetate de-protection (phenol product). Further development of the purification was therefore required.

(116) Purification Development

(117) It was desirable to develop a purification strategy that would remove the hydrolysis product and other impurities observed. It was also desirable to include water in the crystallisation to reduce the salt content of the isolated material (assumed HNMe.sub.2.HCl). To this end a series of 15 solvents and solvent mixtures were screened (100 mg scale, 10 vol solvent, heat cycling to 60° C.).

(118) TABLE-US-00012 TABLE 12 Stage 2 purification trials HPLC Purity (hydrolysis imp of Solvent Observations Recovery Intermediate 2A) TBME Slurry 47 mg* 96.4% (2.8%) DCM Slurry 60 mg  99.5% (0.5%) Toluene Slurry 30 mg* 95.8% (3.4%) EtOAc Slurry 66 mg* 97.8% (1.6%) iPrOAc Slurry 40 mg* 97.3% (2.0%) IPA Slurry 65 mg  99.4% (0.4%) EtOH/H.sub.2O, 1:1 Slurry 62 mg  99.4% (0.5%) MeCN/H.sub.2O, 1:1 Partial solution at RT 30 mg  99.0% (0.7%) Solution at 60° C. Acetone/H.sub.2O, 1:1 Slurry at RT 51 mg  99.4% (0.5%) Solution at 60° C. THF/H.sub.2O, 1:1 Partial solution at RT No n/a Solution at 60° C. precipitate Heptane Slurry 34 mg* 95.0% (3.6%) MIBK Slurry 65 mg  97.9% (1.4%) MEK Slurry 60 mg  99.6% (0.3%) Cyclohexane Slurry 41 mg* 94.4% (4%)   Xylenes Slurry 56 mg* 96.1% (3%)   *Recovery not representative due to thick suspensions and solids adhering to the glass vial.

(119) From the solvents screened acetone/water gave a re-crystallisation with little observed solubility at room temperature. Since this was an aqueous system it had the advantage of helping to purge Me.sub.2NH.HCl from the solids.

(120) The acetone/water re-crystallisation was scaled up. A solution was obtained at temperature (5 vol acetone, 1 vol water) prior to addition of further water (4 vol) and the mixture cooled to RT giving crystallisation (62% recovery, >99% HPLC purity). This process was subsequently scaled up further with addition of more water to aid the recovery (in total 5 vol acetone, 10 vol water, 78% recovery).

(121) The process was scaled up further (30 g) and the crude solids taken through the re-crystallisation procedure. While product purity was high, there was a drop in yield (56% yield, 99% by NMR assay, 99.4% by HPLC).

(122) In order to improve the recovery, the amount of water added was further increased from 10 vol to 15 vol. This maintained product purity at greater than 99% and gave a higher recovery on a small scale (90% recovery, 56-70% previously observed). However, scale up of this amended procedure again gave a low recovery (58% yield). Therefore, due to the issues encountered when scaling up the re-crystallisation, an alternative means of purification was sought based on the original slurry screen that was carried out (Table 12 above).

(123) Redevelopment of the purification strategy took place using material isolated from a large scale, stage 2, reaction. The reaction progressed as expected to give crude product after oven drying (70% by NMR assay, 79% active yield). To remove the significant salt component (presumed to be HNMe.sub.2.HCl) a portion was water slurried at RT. After drying this gave 75% recovery (95% by NMR assay) showing this to be an effective means of reducing the salt content. HPLC purity remained unchanged at ˜93%. A method was then sought to increase the chemical purity of the solids.

(124) From the initial screen both EtOH:H.sub.2O, IPA and acetone:H.sub.2O appeared to give high purity product with a good recovery and so these solvent systems were chosen for further investigation. Input purity was 92.7% with the main impurities at levels of 1.4%, 1.0% and 0.8%.

(125) TABLE-US-00013 TABLE 13 Further purification development (slurry at 40° C. 1 hr, filtration at RT, 1 vol wash) HPLC purity (main impurity levels) Solvent system Recovery Product Imp 1 Imp 2 Imp 3 Acetone/Water 67% 98.8% 0.7% 0.2% 0.1% 6/16 vol 1:1 EtOH:H.sub.2O 79% 98.0% 0.9% 0.5% 0.2% 10 vol IPA 90% 97.5% 0.8% 0.8% 0.3% 10 vol IPA 92% 97.2% 1.0% 0.7% 0.3% 5 vol 4.5:2.5 IPA:H.sub.2O * 79% 98.7% 0.6% 0.4% 0.1% 7 vol 6:1 IPA:H.sub.2O   82% ** 98.0% 0.9% 0.6% 0.2% 7 vol * The reaction was initially run in 1:1 IPA:H.sub.2O at 5 vol. However it became too thick to stir and so a further 2 vol of IPA was added. ** The mixture was thick and the solids present were very fine making filtration difficult with some solids beating the filter.

(126) The results of these trials suggested that good recoveries were possible from these systems, particularly those based on IPA. EtOH:H.sub.2O gave a marginally better impurity profile than IPA alone; however the recovery was not as good (79 vs 90%). The impurity profile with IPA was greatly enhanced by the presence of water (98.7% vs ˜97.5%) however this led to a lower recovery (79 vs 90%). This suggested a certain level of water solubility for the compound. A final trial in IPA and EtOH:Water was conducted with reduced water volumes to see if a balance could be found that provided high purity and a recovery of ˜90%. While this system improved the yield the filtration was slow and therefore further solvent mixtures were also evaluated.

(127) TABLE-US-00014 TABLE 14 Examination of further solvent mixtures (500 mg scale, slurry at 40° C. 1 hr, filtration at RT, 1 vol wash) HPLC purity (impurity levels) Solvent system Recovery Product Imp 1 Imp 2 Imp 3 IPA 460 mg (92%) 97.6% 1.1% 0.7% 0.3% 10 vol 1:3 EtOH:H2O 451 mg (90%) 96.2% 1.4% 0.8% 0.5% 5 vol 1:2 EtOH:H2O 440 mg (88%) 97.5% 1.3% 0.7% 0.4% 5 vol 1:1 EtOH:H2O 375 mg (75%) 97.9% 1.0% 0.7% 0.3% 5 vol 2:1 EtOH:H2O 354 mg (71%) 97.6% 1.0% 0.6% 0.4% 5 vol 3:1 EtOH:H2O 369 mg (74%) 97.9% 3.9% 0.5% 0.3% 5 vol

(128) The 5 vol EtOH/Water slurries were very thick and not easily handled. Since the purity of the solids was comparable from all the trials (slight variations are likely due to the quality of the filtration and wash), the 100% IPA conditions were scaled up as they offered a high recovery and the resulting suspension was easily handled.

(129) An initial scale up of the preferred slurry gave (92% recovery) with HPLC purity of 96.4% (Impurity levels of 1.2%, 0.7%, 0.4%). Liquors analysis showed them to be enriched in all of the main impurities—72% (8.6%, 3.8%, 3.4%) by HPLC. This was deemed a suitable purification method offering a high recovery and the material was use tested in the following stage to ensure tracking and removal of the impurities was achieved downstream (>99% at stage 3, no single impurity >0.5%).

(130) The slurry proved scalable when remaining crude stage 2 material (70% assay) was water slurried to remove inorganics, and then slurried in IPA to give material of improved purity (97% by NMR assay, 76% yield for stage 2, 96.8% by HPLC, impurities at 1.1%, 0.8%, 0.4%).

(131) GMP Raw Material Synthesis

(132) A large scale, stage 2 reaction was carried out to supply the GMP campaign. The reaction progressed as expected to provide crude product that was water slurried and filtered to provide stage 2 that was 93% pure by HPLC. This was further slurried in 8 vol IPA and filtered to give stage 2 product (93.7% by HPLC, 92% assay, 66% active yield). Since the purity obtained was lower than that observed during the development campaign, a use test was conducted which confirmed that high purity stage 3 obtained was suitable for onward processing (GMP raw material).

(133) A second batch was carried out under identical conditions to give crude product which after a water slurry was 90% pure by HPLC. This material was subsequently water slurried and purified by IPA slurry to give 384 g of stage 2 product (93.0% by HPLC, 91% NMR assay, 60% active yield).

(134) A third batch was carried out to resupply the GMP synthesis. The crude product was successfully purified by a water then IPA slurry to deliver stage 2 (79% yield) with an increased purity when compared to previous batches (97.3% by HPLC, 96% NMR assay).

(135) Experimental

(136) Stage 1 (1 eq. limiting reagent) was charged to a vessel under N.sub.2 followed by THF (1 vol wrt stage 1 charge) and TBME (6 vol wrt stage 1 charge). Oxalyl chloride was then added dropwise to the vessel (1.5 eq) allowing the exotherm to initially raise the temperature to 35-40° C. and then applying cooling as required to maintain 35-40° C. Immediately following the addition the reaction was heated to 40° C. and stirred for 2-6 hours. The reaction was sampled and analysed for completion, then cooled to RT and heptane (8 vol wrt stage 1 charge) added giving precipitation of further solids. The reaction was stirred for 10 min and then the solids were allowed to settle. The majority of the solvent was decanted from the solid which was then washed twice with heptane (2×6 vol wrt stage 1 charge), decanting in a similar manner after each wash. The solids were then sampled and analysed. TBME was charged to the vessel (4 vol wrt stage 1 charge) to give a yellow slurry which was cooled to −20° C. using a dry ice/acetone bath. A 2M solution of Me.sub.2NH in THF (2 eq) was added dropwise to the vessel over ˜15 min maintaining the temperature at −20° C. to −10° C. The reaction mixture was allowed to warm slowly to RT and stirred overnight. Further Me.sub.2NH can be added at this point if required. The reaction was sampled and analysed for completion. The reaction was filtered, washing with heptane (2×2 vol wrt stage 1 charge) and the isolated solids dried at 60° C. under vacuum. The crude stage 2 was slurried in water (8 vol wrt stage 1 charge) for 2-18 hours and then filtered, washing with water (2 vol wrt stage 1 charge). The solids were dried at 60° C. under vacuum to obtain crude stage 2 with <2% w/w water (determined by Karl Fischer titration (KF)). The crude stage 2 was slurried in IPA (10 vol) for 2-18 hrs and then filtered, washed with IPA (1 vol wrt mass of crude Stage 2) and oven dried under vacuum at 60° C.

(137) The differences from JNP and the benefits can be summarised as follows:

(138) Step 1

(139) i) Firstly, the use of a THF/TBME solvent system in place of diethyl ether was less volatile and flammable.

(140) ii) Secondly, the addition of oxalyl chloride was conducted at an elevated temperature, heated to 40° C., giving rise to improved solubility and preventing entrapment of stage 1 product in the precipitate. It also provided a high reaction rate, improved levels of completion and shorter reaction times.
iii) Thirdly, the Intermediate 2A was isolated to allow for purification away from excess oxalyl chloride.
iv) Fourthly, heptane was added to help precipitate Intermediate 2A.
Step 2
v) By ensuring the amine was used in excess much improved purity and yields were obtained, due to minimal water being present, and hence reduced hydrolysis.
vi) Finally, the use of water and IPA slurries provided good purity of the Stage 2 product.

Example 4

Stage 3

(141) An initial stage 3 trial reaction was carried out using purified stage 2 material (>99% by HPLC) and the supplied reaction conditions. The stage 2 input material was found to be largely insoluble in THF, and so rather than adding a solution of stage 2 to LiAlH.sub.4 the reverse addition was carried out. 4 eq of LiAlH.sub.4 was used as a 1M solution in THF with the addition made at 20-25° C., over ˜2 hrs. At this point 10% product was observed with several intermediate species present. The reaction was heated at reflux for ˜7 hrs to give 93% product conversion (by HPLC). The reaction was worked up to give crude stage 3 product (˜90% by HPLC, ˜90% by NMR, ˜87% corrected yield).

(142) A trial reaction was carried out in which the LiAlH.sub.4 charge employed was successfully reduced (3 eq vs 4 eq). It was hoped this would benefit the work up by reducing the quantity of Li and Al salts generated. After prolonged heating at reflux (10-18 hrs) the reaction intermediate was largely consumed (2-3% remaining) with ˜95% product by HPLC.

(143) Workup Development

(144) Although the first trial reaction was successfully worked up using Rochelles salt, the volumes employed were very high (˜100 vol) and this procedure would not form a viable process for scale up. A variety of alternative workup procedures were examined in order to try and reduce the volumes required and aid removal of Li/Al salts.

(145) A reduced volume quench was trialed with EtOAc and then Rochelles salt. Grey solids were present as a thick paste which settled to the bottom of the flask. While filtration failed, the liquors could be decanted and the solids re-slurried in THF/EtOAc to extract the product. An aqueous workup was then carried out and the product isolated by concentration. This yielded product of good purity (90-95% by NMR) in good yield (94% uncorrected for purity). However the process was not readily amenable to scale up.

(146) A reaction was quenched by addition of EtOAc and then sat. Na.sub.2SO.sub.4 in the presence of anhydrous Na.sub.2SO.sub.4 to act as a binding agent. The reaction gave granular solids which could be readily filtered. An aqueous workup was then carried out and the product isolated by concentration. A good yield was obtained (˜94% uncorrected for purity) but the product contained higher levels of the main impurity by NMR (10% vs 2-4% usually observed).

(147) A reaction was quenched with 20% AcOH at 0° C. leading to the formation of a gel which could not be filtered. The reaction was abandoned.

(148) A reaction was quenched with EtOAc and then 20% citric acid to give solids which could be separated by filtration. The liquors were concentrated to obtain the product. While this procedure was slightly lower yielding (˜77% uncorrected for purity), the product was of very high purity (>95%).

(149) A further reaction was quenched by the addition of EtOAc and then water (3 mL per g of LiAlH.sub.4 in THF). A gel formed which could not be readily filtered and the reaction was abandoned.

(150) Finally, a reaction was quenched by the Fieser method. An addition of water (1 mL per g of LiAlH.sub.4) was made then 15% NaOH (1 mL per g of LiAlH.sub.4) and finally water (3 mL per g of LiAlH.sub.4). This gave solids which could be filtered from the reaction mixture. The liquors were then partitioned and concentrated in vacuo (87% yield, 90-95% by NMR).

(151) These Experiments are summarised in Table 15 below:

(152) TABLE-US-00015 TABLE 15 Summary of alternative workup conditions Workup # procedure Yield Purity 3.1 EtOAc 2.8 g (94% uncorrected) 90-95% by NMR quench Rochelles salt (reduced vol.) 3.2 EtOAc 2.8 g (94% uncorrected) ~80-85% by quench in NMR, 10% imp. presence of Na.sub.2SO.sub.4 3.3 20% AcOH Emulsion (reaction to waste) quench 3.4 EtOAc i) 529 mg (71% uncorrected) >95% purity by quench ii) 2.3 g (77% uncorrected) NMR 20% citric acid 3.5 Water Emulsion (reaction to waste) quench 3.6 Water/NaOH i) 615 mg (83% uncorrected) 90-95% by NMR quench ii) 2.6 g (87% uncorrected)

(153) Both quenches with citric acid and NaOH gave solids that could be readily filtered from the reaction mixture and required minimal solvent volumes. While the conditions with NaOH were higher yielding (˜10%), the product obtained from this procedure was less pure and would likely require further purification before use in the next stage. The lower yield with citric acid was likely due to some precipitation of the product citrate salt. This had a purifying effect with clean product obtained directly after concentration. These conditions were chosen for scale up and it was hoped that further optimisation of the citric acid charge would enable clean product to be isolated in high yield from this process.

(154) The reaction was repeated with a slightly reduced citric acid charge in order to try and maximise the recovery. This reaction yielded product in 57% yield with a further 20% yield obtained by re-slurry of the filter cake in THF (both samples 97.7% by HPLC).

(155) The reaction was scaled up. However during the EtOAc quench, where the reaction was previously seen to thicken, the reaction gummed in the flask to form a thick mass which restricted mixing. While the addition of citric acid then led to the usual slurry/gel this did not represent a viable process. This reaction was worked up with the filter cake re-slurried in THF to maximise the recovery giving 76% active yield, 95.0% by HPLC.

(156) The reaction was repeated in order to develop a better quench and avoid the gum formation seen with EtOAc. A portion was quenched by addition of acetone which led to a readily stirred suspension/emulsion with no sign of thickening. The citric acid treatment was then carried out to give a filterable mixture. This quench was successfully carried out on the remainder of the reaction and worked up to provide crude product in good yield (71% assay, 82% corrected yield, 98.0% by HPLC).

(157) After the quench the reaction mixture was generally found to be pH 8/9. As part of the workup optimisation process different pHs were investigated. A reaction was split for workup with half receiving a slightly reduced citric acid charge (to obtain pH 11/12 after quench) and the other half taken to pH 7 by addition of further citric acid. The pH 11 reaction was worked up to give material of 85% NMR assay (73% yield) with the pH 7 reaction giving 60% NMR assay (62% yield). It was clear from this result that obtaining the correct pH after quench was critical in order to give a >70% yield. By reducing citric acid charge only slightly (still approx. 2 vol of 20% citric acid) an approx. 8% increase in yield was obtained. With this information in hand the pH of future reactions was monitored during the quench in order to ensure the mixture remained strongly basic.

(158) Purification Development

(159) A purification screen was carried out using 100 mg portions of crude Psilocin product which were slurried in 10 vol of solvent with heat cycling to 60° C. The slurries were cooled to RT over the weekend and then any solids collected by filtration. Stability to acid and base was also tested with the view to carrying out an acid/base workup. The results of the screen are presented in Table 16 below:

(160) TABLE-US-00016 TABLE 16 Psilocin purification screen. (Input purity and 3 main impurity levels: 90.2%, 3.8%, 0.9%, 0%). Recovery HPLC Purity (and Solvent Observations (approx.) main impurity levels) MeOH Solution at RT n/a n/a EtOH Solution at 60° C. 35 mg 97.2%, 0.4%, 1.1%, 0.8% Precipitate at RT IPA Slurry at 60° C. 51 mg 97.6%, 0.5%, 0.5%, 1.1% MeCN Solution at 60° C. 46 mg 96.8%, 0.6%, 0.4%, 1.6% Precipitate at RT EtOAc Slurry at 60° C. 58 mg 97.1%, 0.9%, 0.2%, 1.3% .sup.iPrOAc Slurry at 60° C. 58 mg 98.2%, 0.8%, 0.2%, 0.4% Toluene Slurry at 60° C. 70 mg 93.3%, 3.6%, 0.2%, 0.8% Heptane Slurry at 60° C. 77 mg 91.3%, 3.8%, 0.2%, 0.7% Acetone Solution at 60° C. 30 mg 97.7%, 0.4%, 0.3%, 0.9% Precipitate at RT MEK Solution at 60° C. 24 mg 97.3%, 0.5%, 0.5%, 1.2% Precipitate at RT MIBK Slurry at 60° C. 49 mg 97.4%, 0.6%, 0.2%, 1.3% THF Solution at RT n/a n/a TBME Slurry at 60° C. 67 mg 95.5%, 2.0%, 0.1%, 1.5% DCM Solution at RT n/a n/a 1M HCl Solution at RT n/a 83.7%, new imp 8% 1M KOH Slurry at RT (black) n/a 89.1%, 5.4%, 0.9%, 2.2%

(161) The first of the three highlighted impurities corresponded with the most stable reaction intermediate that is observed at ˜70%, when the LiAlH.sub.4 addition is complete (requiring refluxing to convert to product). The third impurity was not present in the input and appeared to be generated during the slurry procedure. Of the solvents that remained as a slurry, .sup.iPrOAc gave the highest purity. Several re-crystallisations were found with MeCN having the potential to remove impurities during the crystallisation and having a recovery which had the potential to improve during development. Some degradation was observed in both acid and base with the KOH sample rapidly turning black.

(162) Purification of the crude stage 3 material was scaled up using the two most promising solvents (MeCN and .sup.iPrOAc). The solvent volumes were reduced to a minimum in order to improve the recovery. The results of these trials are presented in Table 17 below.

(163) TABLE-US-00017 TABLE 17 Further development of MeCN/.sup.iPrOAc purification Recovery HPLC Purity (and Solvent Observations (%) main impurity levels) MeCN (5 vol) Recryst in 5 vol 512 mg 97.6%, 0.7%, 0.8%, 0% (51%) .sup.iPrOAc (3 vol) Slurry in 3 vol 706 mg 95.8%, 1.3%, 0.6%, 0% (71%)

(164) A re-crystallisation was obtained from MeCN in 5 vol and a hot slurry was achieved in .sup.iPrOAc at 3 vol (both at 75° C.). The recovery from MeCN was again poor despite reduced volumes, however the product was of very high purity (>>95% by NMR). The recovery from .sup.iPrOAc was better with a large increase in product purity when analysed by NMR (˜95%).

(165) Although HPLC and NMR purity of material from the iPrOAc slurry was high, a low assay value (85% by NMR assay) was observed. In order to improve the assay value of the material, as well as remove colour, (all materials obtained so far were strongly purple, green or brown) purification by silica pad was investigated.

(166) Crude Psilocin (71% assay, 98.0% by HPLC) was passed through 4 eq of silica eluting with THF. An 80% recovery of an off white solid with slightly improved HPLC purity (98.4%) and assay value (˜82% assay) was obtained. This proved to be an effective means of increasing the product assay value and was therefore included as part of the reaction workup.

(167) A series of .sup.iPrOAc/anti-solvent slurries (Table 18) were then performed using the silica treated input (100 mg per slurry) to try and improve the recovery, whist maintaining chemical purity (input purity 98.4%).

(168) TABLE-US-00018 TABLE 18 Results of .sup.iPrOAC/anti-solvent additions. Solvent system Recovery HPLC purity .sup.iPrOAc 78% 99.7% 5 vol 1:1 .sup.iPrOAc:Heptane 70% 99.6% 5 vol 1:1 .sup.iPrOAc:TBME 83% 99.7% 5 vol 1:1 .sup.iPrOAc:Toluene 84% 99.4% 5 vol .sup.iBuOAc 80% 99.6% 5 vol

(169) Since all purity values were comparable, two solvent systems were chosen for scale up based on the highest recoveries obtained. The two favoured slurries (TBME and Toluene as anti-solvent) were scaled up (1.0 g per slurry) to better assess the recovery.

(170) TABLE-US-00019 TABLE 19 Scale up of favoured purification methods Solvent system Recovery HPLC purity 1:1 .sup.iPrOAc:TBME 79.9% 99.1% 5 vol 1:1 .sup.iPrOAc:Toluene 79.4% 99.6% 5 vol

(171) Both of these options provided material of >99% HPLC purity at ˜80% recovery and, combined with a silica pad, appear to provide an effective means of purification for the Psilocin product. Further colour was removed into the liquors during the slurry giving Psilocin as a white solid. All impurities were effectively removed to below 0.5%. The .sup.iPrOAc:TBME slurry was chosen for scale up as this used non-toxic ICH class 3 solvents.

(172) Scale Up

(173) The developed stage 3 conditions were scaled up and the reaction progressed to give a completion of 94.4% product with 2.9% of the reaction intermediate present by HPLC after overnight at reflux (typical of the process). After a silica pad Psilocin was obtained in 83% purity by NMR assay, 66% active yield, 97.0% by HPLC. This material was further purified by slurry in iPrOAc/TBME to give material 100% by NMR assay in 62% yield and 99.7% by HPLC.

(174) Due to the crude yield from the reaction being lower than expected (66% vs ˜75%) the filter cake and silica pad were reinvestigated in order to try and recover additional material. However, this was unsuccessful.

(175) The lower than expected yield may have been due to decomposition of the product during workup, although previous stress tests had indicated the material to be stable under the conditions used. To investigate this further the reaction was repeated. The crude product was isolated before the silica pad and additional stress test samples taken to confirm degradation of the product was not occurring during workup.

(176) The reaction progressed as expected to give completion (93.7% product, 2.9% intermediate) and was concentrated yielding crude material (77% NMR assay, 66% active yield). The filter cake was re-slurried in THF/MeOH but no significant Psilocin was isolated. In order to try and displace any product that was coordinated to the aluminium salts, further citric acid was added to take the pH to 4 (from pH 8) and the cake re-slurried in THF, but again no significant Psilocin was isolated. Mass balance was not obtained from the reaction with the 66% active yield closely matching what was previously obtained. This batch was purified by silica pad and slurry in .sup.iPrOAc/TBME to give a 62% yield of high purity material (99.8% by HPLC).

(177) Despite the solvent volumes employed being relatively high and a silica pad being required for removal of aluminium and lithium species, the process was still well suited to the required scale.

(178) The stage 3 reaction was further scaled up to process. The reaction proceeded as expected to give completion after 18 hours (˜91% product, ˜3% reaction intermediate remaining). Workup by silica pad and slurry gave a 57% yield of high purity Psilocin (>99% by HPLC, 99% NMR assay, 0.35% w/w water Karl Fischer).

(179) Experimental

(180) Stage 2 (1 eq. limiting reagent) was charged to the vessel followed by THF (5 vol wrt stage 2 charge). The mixture was cooled to 0° C. and a 1M THF solution of LiAlH.sub.4 (3 eq) added dropwise over 30-45 min maintaining the temperature at 0-20° C. Following the addition, the reaction was stirred for 30 min at 10-20° C. and then heated to reflux and stirred for ˜16 hrs. The reaction was sampled and analysed for completion, cooled to 0° C. and quenched by dropwise addition of acetone (9.3 eq.) at 0-30° C. followed by a 20% aq citric acid soln (1.9 vol wrt stage 2 charge) at 0-30° C. The pH of the addition was monitored to ensure that it remains at pH>11 and the addition was stopped early if required. The resulting suspension was stirred for 1 hr and filtered, washing with THF (2 vol wrt stage 2 charge) to remove Li and Al salts. The filter cake was slurried in THF (12.5 vol wrt stage 2 charge) for ˜1 hr and filtered, washing with THF (5 vol wrt stage 2 charge) to recover product from the Li and Al salts. The combined organics were dried over MgSO.sub.4 and filtered. The filtrate was evaporated in vacuo until approximately 10 volumes remained (wrt stage 2 charge) and this solution was applied to a silica pad (3 eq wrt stage 2 charge). The silica pad was eluted with THF and the product fractions were combined and evaporated to dryness in vacuo. The crude stage 3 (Psilocin) was slurried in 1:1 iPrOAc:TBME (5 vol wrt mass at step 18) for 2-18 hrs, filtered, washing with TBME (2.5 vol wrt mass at step 18) and dried in vacuo at 40° C. to isolate pure Psilocin.

(181) The differences from JNP and the benefits can be summarised as follows:

(182) i) Firstly, Applicant, whilst using THF as a solvent, quenched the reaching using acetone. This lead to a suspension/emulsion without thickening.

(183) ii) Secondly, Applicant quenched with citric acid maintaining a basic pH, typically about 11. The pH control ensured high yields were obtained.

(184) iii) Thirdly, following purification by silica pad, to remove residual Li/Al salts, eluting with THF, a iPrOAc:TBME slurry provides a highly purified product which was then dried.

Example 5

Stage 4

(185) Initially the literature conditions were used to process a 2.58 g sample giving ˜88% conversion to Intermediate 4A when analysed by HPLC. The product was purified by the addition of aminopropyl silica and filtration through Celite. The resulting green oil (5.5 g) was slurried in DCM giving benzyl transfer and precipitation of the zwitterionic stage 4 (4.1 g, 70% yield, ˜95% by NMR).

(186) Step i

(187) Initial development at this stage was focused on finding an alternative to .sup.nBuLi that was easier to handle and ideally did not introduce further lithium into the synthesis. An initial screen of alternative conditions was carried out including the following bases: Li.sup.tBuO, K.sup.tBuO, NaH, NaHMDS, and NaNH.sub.2. All of the reactions gave product with NaHMDS performing as well as .sup.nBuLi. All of the reactions became very thick with gelling observed and overhead stirring was recommended for efficient stirring.

(188) The initial screen suggested NaHMDS would be a suitable alternative to .sup.nBuLi (81% conversion to product/Intermediate 4A). These conditions were scaled up to 1.5 g alongside a reference reaction with .sup.nBuLi. Overhead stirring was used in both cases.

(189) TABLE-US-00020 TABLE 20 Comparison of .sup.nBuLi and NaHMDS Timepoint HPLC—.sup.nBuLi HPLC—NaHMDS 1 hr, −30° C. 6.6% St 3, 78% Int 4A, <1% St 3, 78% Int 4A, 1% St 4 4% St 4 2 hr, 0° C. 6.5% St 3, 77% Int 4A, <1% St 3, 76% Int 4A, 1% St 4 4% St 4 Crude product 4.89 g 4.38 g <1% St 3, 2% Int 4A, <1% St 3, 5% Int4A, 60% St 4 68% St 4 Abbreviations used in the table: St 3 = Stage 3, Int 4A = Intermediate 4A, St 4 = Stage 4

(190) The reaction profile obtained in both cases was very similar with the NaHMDS reaction giving consumption of stage 3. Both reactions were filtered on Celite to remove a white precipitate and concentrated. By NMR excess benzyl protons were present in both cases (especially in the example with .sup.nBuLi) and the isolated yield was >100%. The NaHMDS conditions proved successful giving a favourable reaction profile and were chosen for further scale up. However, workup and purification development was required.

(191) Step ii

(192) HPLC data indicated the material isolated from the trials above using NaHMDS and .sup.nBuLi had rearranged to give zwitterionic stage 4 upon concentration. Purification of this material away from the benzylphosphoric acid by-products and other impurities was attempted by slurry in a number of solvents.

(193) TABLE-US-00021 TABLE 21 Trial purification of crude stage 4 product Solvent Mass recovered HPLC Purity DCM White solid 84% St 4 EtOH Gum — EtOAc Gum — IPA White solid 88% St 4 Toluene Gum — TBME White solid 62% St 4 MIBK Gum — MeCN Gum — Acetone White solid 86% St 4 Abbreviations used in table: St 4 = Stage 4 *filtration poor

(194) White solids were obtained from several solvents however the solids obtained from DCM and TBME turned to a pale purple gum when stored over the weekend. Those obtained from IPA and acetone remained as free flowing white solids on storage suggesting that the stability of these solids was likely to be higher and that they would allow for easier handling.

(195) The slurries in IPA and acetone were scaled up to 1 g. However, gumming was noticed immediately on addition of the solvent. The gum was slowly dispersed by vigorous stirring and eventually showed signs of crystallisation, with a white slurry forming after an overnight stir. However, this process was not suitable for scale up. Solids were isolated in good yield with IPA providing the highest purity.

(196) THF was also investigated as this had advantages in that it was also the reaction solvent. However, when this was trialed initial gum formation was again observed (isolated ˜80% yield, ˜92% by HPLC). In order to try and avoid the gum formation and give a more controlled crystallisation the crude stage 4 was first solubilised in a low volume of DMSO (2 vol). THF was then added to this (10 vol) and the solution stirred over the weekend. This slowly gave precipitation of the product which was collected by filtration and washed with THF to yield stage 4 (86% yield) with 96% HPLC purity (>95% by NMR).

(197) As the THF crystallisation was successful and it was previously noted that complete conversion to zwitterionic stage 4 occurred during concentration of the reaction liquors (THF/EtOAc) at 40° C., it was hoped that the changes of solvent could be avoided and the product crystallised directly by stirring out the reaction mixture at 40° C.

(198) Two 4 g NaHMDS reactions were carried out with both reactions reaching completion with ˜80% conversion to Intermediate 4A. One reaction was diluted with EtOAc and the other with THF and both were filtered to remove phosphate by-products. In order to further reduce the phosphate impurity levels a brine wash was carried out and the organics dried and concentrated to 10 vol. These solutions were stirred overnight at 40° C. to give conversion to, and precipitation of, stage 4 (˜1% stage 3, ˜0.2% Intermediate 4A, ˜82% stage 4). The solids were collected by filtration giving 8.03 g (88% yield) from EtOAc/THF and 5.65 g (62% yield) from THF. The brown/grey solids obtained from EtOAc/THF were of lower purity (˜90% by HPLC, 78% assay) when compared to the white solids obtained from THF (97% by HPLC, 88% assay). Analysis of the aqueous layer from the THF reaction showed product to be present and additional losses were incurred to the final THF filtrate.

(199) Due to the higher purity obtained from THF, this solvent was investigated further in order to optimise the recovery. The brine wash was omitted due to product losses to the aqueous layer and the reaction mixture was further concentrated after the reaction to minimise losses during the final filtration step. This new procedure was trialed on a 75 g scale with portions of the reaction mixture concentrated to 8 vol and 6 vol. Upon filtration no difference in yield was noted between the two portions with an overall yield of 140.4 g (90% by NMR assay, 74% active yield, 90% by HPLC).

(200) Impurity Tracking

(201) Three main impurities were observed in the isolated product, with identities for two of these species proposed based on MS data.

(202) The debenzylated impurity (typically ˜2-5% by HPLC) was shown to give psilocybin during the following hydrogenation and could therefore be tolerated at a higher level. The main observed impurity in the isolated stage 4 (typically ˜5-8% by HPLC) was the anhydride impurity. This was tracked through the subsequent hydrogenation and shown to be readily removed by re-crystallisation from water as the highly soluble pyrophosphorate impurity that results from debenzylation. The other main observed impurity (m/z 295.2 observed by LCMS) was found to be controlled to less than 2% by limiting the reaction temperature (below −50° C.) and was not observed in Psilocybin after hydrogenation.

(203) The impurity profile of the 140 g batch produced above showed 90.0% stage 4, 6.4% anhydride impurity, 0.2% N-debenzylated impurity and 1.2% of the m/z 295.2 impurity.

(204) GMP Synthesis

(205) The first large scale stage 3 batch (544 g input) was completed using the established procedure to give 213.5 g (53% yield, 99% by HPLC). A second batch (628.2 g input) was also processed successfully to give 295.2 g (66% yield, 99% by HPLC).

(206) Some variability in yield at this stage was noted over 3 large scale batches (57%, 53% and 66%). This is probably a consequence of minor differences in the workup and quench procedure.

(207) Experimental

(208) Stage 3 was charged to a vessel followed by THF (15 vol wrt stage 3 charge) and cooled to ≤−50° C. using a dry ice/acetone bath. 1M NaHMDS solution in THF (1.13 eq) was charged maintaining a temperature of 5-45° C., target <−50° C. The reaction was stirred for 30 minutes at −60 to −50° C. Tetrabenzylpyrophosphate (2.26 eq) was charged to the reaction in a single portion followed by additional THF (20 vol) while maintaining the reaction temperature <−30° C. The reaction was warmed to 0° C., over 1.5-2 hours and sampled for completion. The reaction was filtered to remove phosphate salts washing with THF (8 vol). The filtrate was concentrated until 6-8 vol remains and stirred overnight at 40° C. to convert Intermediate 4A to stage 4 product. The reaction was sampled for completion and then filtered and the solid washed with THF (2 vol). The stage 4 product was dried in a vacuum oven at 40° C.

(209) The differences from JNP and the benefits can be summarised as follows:

(210) Step i

(211) i) Firstly, sodium hexamethyldisilazide was introduced to support deprotonation. This proved an effective alternative to Butyl Lithium, which was easier to handle, and did not introduce further lithium into the reaction.

(212) ii) Secondly, by diluting the reaction with THF, a much higher purity Intermediate 4A was obtained.

(213) iii) Thirdly, by controlling the reaction temperature at below −50° C., undesirable mz 295.2 observed by LCMS was controlled to levels of less than 2%.

(214) Step ii

(215) iv) Fourthly, by monitoring levels of stage 4A impurities, particularly the N-debenzylated Stage 4 (Table 7) and anhydride Stage 4 (Table 7), a pure product can be produced reproducibly.

(216) v) The intermediate stage 4A to stage 4 conversion can be carried out in the reaction solvent, avoiding the need for time consuming solvent swaps.

(217) vi) Finally, the obtained solid is washed with THF and oven dried to obtain stage 4.

Example 6

Stage 5

(218) Catalyst poisoning was noted during development of this stage and a charcolation step can be included in the process when required to prevent incomplete hydrogenation. However, charcolation is not routinely required.

(219) After sparging with hydrogen for 3 hours typical reactions showed high levels of completion (>90% product, 3-5% SM remaining). A small amount of water was added to aid solubility and after sparging with hydrogen for a further 1 hour, consumption of stage 4 was achieved.

(220) A successful reaction was worked up by filtration, followed by evaporation to remove methanol, leaving the product as a thick suspension in water. Ethanol was added and the solid filtered to give Psilocybin in 69% yield. .sup.1H NMR confirmed the identity of the product but indicated a minor related impurity was present. LCMS analysis indicated a purity of 95.2% with the major impurity (4.1%) being identified as the pyrophosphoric acid impurity. (Table 7) deriving from the anhydride impurity at stage 4. It was later shown that this impurity was effectively purged during the final product re-crystallisation (Stage 6).

(221) A further reaction was then carried out using stage 4 material from the finalised THF workup which was 88.0% pure by HPLC and contained 7.3% N-debenzylated stage 4 (converted to product), with none of the anhydride impurity. Again completion was noted and the reaction worked up as previously to give Psilocybin in 46% yield. The low yield was believed to result from precipitation of the product during the catalyst filtration step. .sup.1H NMR confirmed the identity of the product and HPLC indicated a purity of 98.9%.

(222) Further development of the reaction conditions was carried out to optimise the water volume employed and minimise product losses during the filtration step. After the reaction, a solution was obtained by addition of 10 volumes of water with heating to 40° C. This allowed for removal of the catalyst by filtration without incurring product losses on the filter.

(223) Some stage 3 was generated by hydrolysis during the reaction and workup with levels of approx. 1-2.5% appearing to be typical of the process. A reduction in the stage 3 level was demonstrated during the final product re-crystallisation.

(224) Scale Up

(225) The large scale stage 4 batch (non-GMP) was processed as a single batch (148 g active input). Consumption of stage 4 was achieved with 88% product and 0.9% stage 3 resulting from hydrolysis. The anhydride impurity (6.4%) was completely converted to the corresponding pyrophosphoric acid impurity (5.2%).

(226) The large scale hydrogenation was filtered and concentrated to yield 109 g of crude product after stripping back from ethanol to reduce the water content (˜71% by NMR assay, 86% yield).

(227) Experimental

(228) 10% Pd/C (˜50% water wet, type 87L, 0.1×stage 4 charge) was charged to a vessel under N.sub.2 followed by Methanol (20 vol wrt stage 4) and Stage 4. The N.sub.2 was replaced with H.sub.2 and the reaction was stirred under H.sub.2 (atmospheric pressure) for 1-2 hours. The reaction was sampled for completion and then water was added (10 vol wrt stage 4) maintaining a temperature of <25° C. The mixture stirred for a further 1-2 hours under H.sub.2 (atmospheric pressure). The reaction was sampled and checked for completion.

(229) If the reaction was incomplete, H.sub.2 was recharged and the reaction continued for a further 1-12 hr until completion was observed. The reaction was then placed under N.sub.2 and warmed to 40° C. and held for 15-45 minutes. The reaction was filtered through celite to remove catalyst, washing with methanol (13.3 vol wrt stage 4 charge) and water (6.7 vol wrt stage 4 charge). The filtrate was concentrated in vacuo, azeotroping with ethanol to remove water until a solid was obtained. The differences from JNP and the benefits can be summarised as follows:

(230) Primarily, the reaction is monitored for levels of intermediates by HPLC, using relative retention times (RRT) and completion controlled with intermediates being present at less than 0.2%. The stage 5 pyrophosphoric acid impurity is also carefully monitored to confirm that it can be controlled in the final re-crystallisation.

(231) The final Stage 6 process is as described in Example 1.

Example 7

Testing Methodology and Protocols

(232) To test for purity etc the following methodology/protocols were employed. 7.1 NMR

(233) .sup.1H and .sup.13C NMR spectra of Psilocybin in D.sub.2O were obtained using 400 MHz spectrometer. Chemical shifts are reported in ppm relative to D.sub.2O in the .sup.1H NMR (□=4.75 ppm) and relative to MeOH (□=49.5 ppm), which was added as a reference, in the .sup.13C NMR the spectrum. Literature values for Psilocybin are reported in JNP. Analysis of Psilocybin by NMR gave data that was consistent with the structure and consistent with that reported in the literature with only minor variations in chemical shifts for protons near the ionisable groups which is expected as the zwitterionic nature of the compound makes the material very sensitive to small changes in pH.

(234) The .sup.1H NMR and .sup.13C NMR data are outlined below and the spectra are shown in FIGS. 10 and 11.

(235) .sup.1H NMR Data (400 MHz, D.sub.2O): 2.79 (s, 3H), 3.18 (t, J=7.4 Hz, 2H), 3.31 (t, J=7.4 Hz, 2H), 6.97 (d, J=8.0 Hz, 1H), 7.08 (s, 1H), 7.10 (t, J=8.0 Hz, 1H), 7.19 (d, 8.2 Hz, 1H).

(236) .sup.13C NMR Data (400 MHz, D.sub.2O (+ trace MeOH): 22.3 (1×CH.sub.2), 43.4 (2×CH.sub.3), 59.6 (1×CH.sub.2), 108.4 (1×CH), 108.6 (1×C), 109.5 (1×CH), 119.1 (d, .sup.3J.sub.P-H=6.7 Hz, 1×C), 123.3 (1×CH.sub.2), 124.8 (1×CH), 139.3 (1×C), 146.3 (d, .sup.2J.sub.P-H=6.7 Hz, 1×C)

(237) 7.2 FT-IR

(238) Data was collected on a Perkin Elmer Spectrum Two™ Spectrometer with UATR Two accessory. Analysis of Psilocybin (Batch: AR755) by FT-IR spectroscopy gave a spectrum (FIG. 12) that is consistent with the proposed structure. The broad peak at 3244 cm.sup.−1 is typical of an amine salt. The remainder of the peaks are in the fingerprint region and therefore can't be assigned individually.

(239) 7.3. Mass Spectrometry

(240) The mass spectrum of Psilocybin (AR755) was obtained on a Bruker Esquire 3000 plus Ion Trap Mass Spectrometer and was concordant with the structure. The mass spectrum (FIG. 13) showed a main peak at m/z=284.8 and 568.1 that corresponded to (M+H).sup.+ and (2M+H).sup.+ of Psilocybin. This implied the molecular ion has m/z 284 corresponding to the molecular formula of Psilocybin (C.sub.12H.sub.17N.sub.2O.sub.4P) (FIG. 14).

(241) 7.4 Residue on Ignition

(242) The residue on ignition method follows the pharmacopeia method with one adjustment. Inconsistent results were obtained when the crucible was heated to 500° C. and it is believed this is due to low volatility of the phosphate residues that are generated. The temperature was therefore increased to 800° C. for Psilocybin and consistent and accurate results were obtained.

(243) 7.5 HPLC—Assay and Purity Determination

(244) The HPLC method used for assay, chemical purity and quantifying the impurities of Psilocybin is a gradient HPLC-UV method and the conditions are outlined in Table 22. External standards are used for quantification. Approximately 1 mg/mL of Psilocybin was dissolved in Purified Water:MeOH (95:5). Sonicate to fully dissolve.

(245) Purity by HPLC is calculated in the basis of area % and is correlated against a known retention time standard.

(246) Assay by HPLC is calculated on an anhydrous basis, based on weight % versus a standard of known purity and composition.

(247) TABLE-US-00022 TABLE 22 Typical HPLC Conditions for Identification, Purity and Assay Parameter Conditions System Agilent 1100 series liquid chromatograph or equivalent Column XBridge C18, 4.6 × 150 mm; 3.5 μm (Ex; waters PN: 186003034) Flow Rate 1.0 ml.min.sup.−1 Injection Volume 5 μl Detection UV @ 267 nm Column Temperature 30° C. Mobile Phase A—Purified Water:Methanol:TFA (95:5:0.1) B—Methanol:Purified Water:TFA (95:5:0.1) Gradient Time (mins) % A % B 0 100 0 2 100 0 15 0 100 20 0 100 22 100 0
7.6 Residual Solvent Content by HRGC

(248) The HRGC method for quantifying residual solvents is a headspace method and is described in Table 23 below:

(249) TABLE-US-00023 TABLE 23 Typical Residual Solvent GC Method Parameter Conditions System Agilent 6890/7890 HRGC or similar Column DB-624 60 m × 0.32 mm, 1.80 μm film thickness (or equivalent) Oven Program 40° C. (hold for 15 min) then ramp (20° C.. min.sup.−1) to 200° C. (hold 5 min) Headspace Parameters Oven Temp 125° C. Loop Temp 140° C. Transfer Line Temp 150° C. Split Ratio 10:1 Injector temperature 200° C. Detector temperature 250° C., FID Head pressure 15 psi, constant pressure Carrier gas Nitrogen Column flow 2.0 ml. min.sup.−1 @ 40° C. Internal Standard 1,2-Difluorobenzene

(250) Levels of the following solvents and reagents are determined: Methanol, Ethanol, THF and Toluene.

(251) 7.7 Melting Point by DSC

(252) DSC data was collected on a PerkinElmer Pyris 6000 DSC (or similar). The instrument was verified for energy and temperature calibration using certified indium. The sample was weighed (typically 0.5 to 3.0 mg) into a pin-holed aluminium sample pan. The pan was crimped with an aluminium pan cover. The pan was heated at 20° C./min from 30 to 300° C. with a purge of dry nitrogen at 20 mL/min. During the melting point procedure, each batch of Psilocybin Polymorph A or A′ exhibited two endothermic events the latter; the first of which was attributed to solid-solid transition of Polymorph A or A′ to Polymorph B, and the second of which was attributed to melting of Polymorph B.

(253) 7.8 Polymorphism by XRPD

(254) The solid state form of Psilocybin is determined by XRPD. XRPD diffractograms were collected on a diffractometer (such as a PANalytical X'Pert PRO or equivalent) using Cu Kα radiation (45 kV, 40 mA), θ-θ goniometer, focusing mirror, divergence slit (½″), soller slits at both incident and divergent beam (4 mm) under ambient conditions. The data collection range was 3-35 °2θ with a continuous scan speed of 0.2°s.sup.−1. The resulting diffractogram is compared to that of a reference diffractogram of Polymorph A or A′ to ensure that it is concordant (FIG. 7a or 7b respectively).

(255) 7.9 Thermo-Gravimetric Analysis (TGA)

(256) TGA data was collected on a PerkinElmer Pyris 1 TGA (or similar). The instrument was calibrated using a certified weight and certified Alumel and Perkalloy for temperature. A predefined amount of sample (typically ca. 5 mg) was loaded into an aluminium crucible, and was heated at 20° C./min from ambient temperature to 400° C. A nitrogen purge at 20 mL/min was maintained over the sample.

(257) 7.10 Loss on Drying

(258) Determine in duplicate the loss on drying of the sample using a 1 g portion, accurately weighed, dried at 70° C., under vacuum to constant weight.

(259) Calculation:

(260) $\%\mathrm{Loss}\mathrm{on}\mathrm{Drying}=\frac{\left(W_{\mathrm{INITIAL}}-W_{\mathrm{FINAL}}\right)}{W_{\mathrm{SAMPLE}}}\times 100\mathrm{Where}:W_{\mathrm{INTITIAL}}=\mathrm{Initial}\mathrm{weight}\mathrm{of}\mathrm{dish}\mathrm{and}\mathrm{sample}\mathrm{prior}\mathrm{to}\mathrm{drying}\left(g\right)W_{\mathrm{FINAL}}=\mathrm{Final}\mathrm{weight}\mathrm{of}\mathrm{dish}\mathrm{and}\mathrm{dried}\mathrm{sample}\left(g\right)W_{\mathrm{SAMPLE}}=\mathrm{Weight}\mathrm{of}\mathrm{sample}\left(g\right)$

Example 8

Forced Degradation Studies

(261) Psilocybin drug substance was stressed under various conditions in solution and in the solid state to provide samples to evaluate analytical method selectivity.

(262) The forced degradation study was performed on Psilocybin; based on the requirements of ICH Q1A(R2). Testing under stressed conditions has provided information on potential degradation pathways and the intrinsic stability of Psilocybin. The optimised analytical method employed demonstrated specificity to Psilocybin; it was shown to be suitable and changes to identity, purity and potency of the product can be detected using this method. The method used has also been shown to be free from interferences from likely impurities and degradation products in accordance with ICH Q2(R1) (Validation of Analytical Procedures) with reference to specificity. Therefore, the HPLC method is deemed suitable for determining purity of Psilocybin and related impurities.

(263) The control sample of Psilocybin was stable in solution over the study period (study period was 7 days for non-photostability samples). Psilocybin degraded slowly when heated in solution producing psilocin as the major impurity. Psilocybin was also stable under acid conditions at room temperature. However, at 60° C. a slow and steady degradation was observed producing psilocin as the main impurity. Psilocybin was slightly unstable at room temperature in the presence of base with slow degradation to a range of impurities over the study period. Only very low levels of impurities were formed under the peroxide conditions with the overall purity dropping by ˜0.5%. In the solid state, slow chemical degradation was noted (3 days at 150° C.) predominantly producing psilocin (stage 3) as an impurity. Psilocybin was stable under photostability conditions both as a solid and when in solution.

(264) Stability Studies

(265) Stability studies were undertaken with two batches of Psilocybin as shown in Table 24.

(266) TABLE-US-00024 TABLE 24 Drug Study number/ Substance Site of Storage Intended time Study start Lot No. Packaging Manufacture Lot Use Condition points/Study Status ON- GM764B Double food Onyx Ref. Std. 2-8° C. 1, 3, 6 months YXSTAB0138 grade Scientific ongoing Polythene 25° C./60% RH 1, 3, 6 months bags. Outer ongoing Polythene 40° C./75% RH 1, 3, 6 months container ongoing ON- 170231 Double food Onyx Clinical 2-8° C. 1, 3, 6, 9, 12, YXSTAB0139 grade Scientific 18, 24, 36 months Polythene ongoing bags. Outer 25° C./60% RH 1, 3, 6, 9, 12, Polythene 18, 24, 36 months container ongoing 40° C./75% RH 1, 3, 6 months ongoing

(267) Samples were double bagged in food grade polythene bags and sealed in an outer polythene container and placed on storage at 2-8° C., 25° C./60% RH and 45° C./75% RH, a desiccant bag is included between the inner polythene bags to prevent moisture uptake. Tests for appearance, water content, purity and assay were carried out.

(268) The protocols for the two studies are shown in Table 25 and Table 26.

(269) The one month and three months stability data for batch GM764B are detailed in Table 27 and Table 28 below. The one, three, six, nine and twelve month stability data for GMP batch 170231 are provided in Table 29, Table 30, Table 31, Table 32 and Table 33 respectively below.

(270) TABLE-US-00025 TABLE 25 Onyx Stability Trial Protocol Sheet Product: Psilocybin Onyx trial number: ONYXSTAB0138 Batch number: GM764B Trial due start date: 10 MAR. 2017 Test method: N/A Date of manufacture: 6 FEB. 2017 Additional 1200 mg of material required in each container. information: Packaging Double polythene bagged lined contained within 300 ml HDPE container components: (food grade). Insert a desiccant bag between the two polythene bags. Test parameters Appearance Routine tests Assay (Anhydrous basis) by .sup.1H-NMR Water Content by loss on drying Chemical Purity/Impurities by HPLC Months 1 3 6 Spares Total 2° C.-8° C. X X X 2 5 25° C./60% RH X X X 2 5 40° C./75% RH X X X 0 3 Date due off 10 APR. 2017 10 JUN. 2017 10 SEP. 2017 13

(271) TABLE-US-00026 TABLE 26 Onyx Stability Trial Protocol Sheet Product: Psilocybin Onyx trial number: ONYXSTAB0139 Batch number: 170231 Trial due start date: 31 MAR. 2017 Test method: SS/PSILOCYBIN/ Date of manufacture: 27 FEB. 2017 Additional 2200 mg of material required in each container. information: Packaging Double polythene bagged lined contained within 300 ml HDPE container components: (food grade). Insert a desiccant bag between the two polythene bags. Test parameters Appearance Routine tests Assay (on a dry basis) by HPLC Water Content by loss on drying Chemical Purity/Impurities by HPLC Timepoint 1 month 3 months 6 months 9 months 12 months 2-8° C. X X X X X 25° C./60% RH X X X X X 40° C./75% RH X X X Date due off 31 APR. 2017 30 JUN. 2017 30 SEP. 2017 31 DEC. 2017 31 MAR. 2018 Timepoint 18 months 24 months 36 months Spares Total 2-8° C. X X X 2 10 25° C./60% RH X X X 2 10 40° C./75% RH 1 4 Date due off 30 SEP. 2018 31 MAR. 2019 31 MAR. 2020 24

(272) TABLE-US-00027 TABLE 27 One Month Stability Data for Batch GM764B Test Specification Limit T = 0 T = 1 month T = 1 month T = 1 month Condition N/A N/A 2° C.-8° C. 25° C./60% RH 40° C./75% RH Appearance For information only. An off white solid. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible Free from visible contamination contamination contamination contamination Assay by .sup.1H-NMR For information only.   97% .sup.w/.sub.w   99% .sup.w/.sub.w   98% .sup.w/.sub.w   96% .sup.w/.sub.w Water content by loss For information only.  0.86% .sup.w/.sub.w  0.35% .sup.w/.sub.w  0.20% .sup.w/.sub.w  0.14% .sup.w/.sub.w on drying Chemical Purity By For information only. 99.24% 99.24% 99.22% 99.23% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT 0.86  0.05%  0.05%  0.05%  0.05% RRT 1.46  0.05%  0.09%  0.10%  0.10% RRT 1.59 (Psilocin)  0.37%  0.35%  0.34%  0.34% Total Impurities  0.76%  0.76%  0.78%  0.77%

(273) TABLE-US-00028 TABLE 28 Three Month Stability Data for Batch GM764B Test Specification Limit T = 0 T = 3 month T = 3 month T = 3 month Condition N/A N/A 2° C.-8° C. 25° C.-60% RH 40° C.-75% RH Appearance For information only. An off white solid. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible Free from visible contamination contamination contamination contamination Assay by H-NMR For information only.   97% .sup.w/.sub.w   97% .sup.w/.sub.w   99% .sup.w/.sub.w   97% .sup.w/.sub.w Water content by loss For information only.  0.86% .sup.w/.sub.w  0.26% .sup.w/.sub.w  0.08% .sup.w/.sub.w  0.14% .sup.w/.sub.w on drying Chemical Purity By For information only. 99.24% 99.31% 99.27% 99.26% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT 0.86  0.05% LT LT LT  0.05%  0.05%  0.05% RRT 1.46  0.05%  0.10%  0.09%  0.10% RRT 1.59 (Psilocin)  0.37%  0.37%  0.36%  0.37% Total Impurities  0.76%  0.69%  0.73%  0.74%

(274) TABLE-US-00029 TABLE 29 One Month Stability Data for Batch 170231 Test Specification Limit T = 0 T = 1 month T = 1 month T = 1 month Condition N/A N/A 2° C.-8° C. 25° C./60% RH 40.sup.o C./75% RH Appearance For information only. An off white solid. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible Free from visible contamination contamination contamination contamination Chemical Purity By For information only.  99.28%  99.20%  99.16%  99.17% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT 1.49   0.06%   0.05%   0.05%   0.06% RRT 1.62 (Psilocin)   0.39%   0.36%   0.37%   0.36% RRT 1.70   0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 1.89 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 2.45 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurities LT 0.05%   0.22%   0.39%   0.42%   0.41% Total Impurities   0.72%   0.80%   0.84%   0.83% Assay by HPLC For information only  98.65% .sup.w/.sub.w  98.76% .sup.w/.sub.w  97.98% .sup.w/.sub.w  98.52% .sup.w/.sub.w (on a dry basis) Water content by loss For information only.   0.32% .sup.w/.sub.w   0.27% .sup.w/.sub.w   0.17% .sup.w/.sub.w   0.19% .sup.w/.sub.w on drying

(275) TABLE-US-00030 TABLE 30 Three Month Stability Data for Batch 170231 Test Specification Limit T = 0 T = 3 months T = 3 month T = 3 month Condition N/A N/A 2° C.-8° C. 25° C./60% RH 40° C./75% RH Appearance For information only. An off white solid. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible Free from visible contamination contamination contamination contamination Chemical Purity By For information only.  99.28%  99.30%  99.31%  99.17% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT0.69 LT 0.05%   0.05% LT 0.05% LT 0.05% RRT 1.49   0.06%   0.05%   0.05%   0.06% RRT 1.62 (Psilocin)   0.39%   0.37%   0.36%   0.39% RRT 1.70   0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 1.89 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 2.45 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurities LT 0.05%   0.22%   0.22%   0.27%   0.34% Total Impurities   0.72%   0.70%   0.69%   0.79% Assay by HPLC For information only  98.65% .sup.w/.sub.w  98.45% .sup.w/.sub.w  99.46% .sup.w/.sub.w  98.64% .sup.w/.sub.w (on a dry basis) Water content by loss For information only.   0.32% .sup.w/.sub.w   0.17% .sup.w/.sub.w   0.01% .sup.w/.sub.w   0.19% .sup.w/.sub.w on drying

(276) TABLE-US-00031 TABLE 31 Six Month Stability Data for Batch 170231 Test Specification Limit T = 0 T = 6 months T = 6 months T = 6 months Condition N/A N/A 2° C.-8° C. 25.sup.o C.-60% RH 40.sup.o C.-75% RH Appearance For information only. An off white solid. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible Free from visible contamination contamination contamination contamination Chemical Purity By For information only.  99.28%  99.20%  99.19%  99.12% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT0.69 LT 0.05%   0.06%   0.06%   0.06%   0.07%   0.07%   0.08% RRT 1.49   0.06% LT 0.05% LT 0.05% LT 0.05% RRT 1.62 (Psilocin)   0.39%   0.35%   0.34%   0.38% RRT 1.70   0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 1.89 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 2.45 LT 0.05% LT 0.05% LT 0.05% LT 0.05% Impurities LT 0.05%   0.22%   0.32%   0.34%   0.36% Total Impurities   0.72%   0.80%   0.81%   0.88% Assay by HPLC For information only  98.65% .sup.w/.sub.w  97.97% .sup.w/.sub.w  98.04% .sup.w/.sub.w  100.10% .sup.w/.sub.w (on a dry basis) Water content by loss For information only.   0.32% .sup.w/.sub.w   0.06% .sup.w/.sub.w   0.32% .sup.w/.sub.w   2.26% .sup.w/.sub.w on drying

(277) TABLE-US-00032 TABLE 32 Nine Month Stability Data for Batch 170231 Test Specification Limit T = 0 T = 9 months T = 9 months Condition N/A N/A 2° C.-8° C. 25° C.-60% RH Appearance For information only. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible contamination contamination contamination Chemical Purity By For information only.  99.28%  99.16%  99.16% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT0.69 LT 0.05% LT 0.05% LT 0.05% LT 0.05% LT 0.05% RRT 1.49   0.06%   0.07%   0.05% RRT 1.62 (Psilocin)   0.39%   0.06%   0.06% RRT 1.70   0.05%   0.37%   0.37% Impurity at RRT 1.89 LT 0.05% LT 0.05% LT 0.05% Impurity at RRT 2.45 LT 0.05% LT 0.05% LT 0.05% Impurities LT 0.05%   0.22% LT 0.05% LT 0.05%   0.34%   0.35% Total Impurities   0.72%   0.84%   0.84% Assay by HPLC For information only  98.65% .sup.w/.sub.w  97.53% .sup.w/.sub.w  98.12% .sup.w/.sub.w (on a dry basis) Water content by loss For information only.   0.32% .sup.w/.sub.w   0.21% .sup.w/.sub.w   0.10% .sup.w/.sub.w on drying

(278) TABLE-US-00033 TABLE 33 Twelve Month Stability Data for Batch 170231 Test Specification Limit T = 0 T = 12 months T = 12 months Condition N/A N/A 2° C.-8° C. 25° C./60% RH Appearance For information only. An off white solid. An off white solid. An off white solid. Free from visible Free from visible Free from visible contamination contamination contamination Chemical Purity By For information only.  99.28%  99.25%  99.25% HPLC Impurities by HPLC: For Information only. (Quote all GT 0.05%) RRT0.69 LT 0.05% LT 0.05% LT 0.05% RRT 1.49   0.06% LT 0.05% LT 0.05% RRT 1.62 (Psilocin)   0.39%   0.37%   0.37% RRT 1.70   0.05% LT 0.05% LT 0.05% Impurity at RRT 1.89 LT 0.05% LT 0.05% ND Impurity at RRT 2.45 LT 0.05% LT 0.05% LT 0.05% Impurities LT 0.05%   0.22%   0.38%   0.38% Total Impurities   0.72%   0.75%   0.75% Assay by HPLC For information only   98.65% .sup.w/.sub.w  99.63% .sup.w/.sub.w  98.97% .sup.w/.sub.w (on a dry basis) Water content by loss For information only.   0.32% .sup.w/.sub.w   0.49% .sup.w/.sub.w   0.61% .sup.w/.sub.w on drying

(279) Over the first 12 months of the ICH stability study Psilocybin has proven to be chemically stable under low temperature (2-8° C.), ambient (25° C./60% RH) and accelerated (40° C./75% RH) conditions. There has been no change in the appearance and HPLC analysis has also remained consistent. The water content has varied in all samples, due to the initial impact and then aging of the desiccant bags used in the study.

Example 9—Experimental to form Hydrate A

(280) Psilocybin (200 mg) was charged to a crystallisation tube followed by deionised water (4 ml). The mixture was equilibrated at 25° C. for 2 hours before the solid was isolated by vacuum filtration. The material was split into two equal portions. One portion was not subjected to further drying to give Hydrate A, lot GK2, by XRPD and DSC (diffractogram and thermogram consistent with FIG. 7d and FIG. 8d respectively).

Example 10—Experimental to Form Polymorph B

(281) Psilocybin Polymorph A (250 mg) was charged to a round bottom flask, heated to 173° C. using an oil bath and held at temperature for 5 minutes. The solid was cooled to ambient temperature and isolated to give lot GK3 with a recovery of 93%. Analysis by XRPD and DSC revealed lot GK3 to be Polymorph B (diffractogram and thermogram consistent with FIG. 7c and FIG. 8c respectively).

Example 11—Solid State Investigations

(282) A number of polymorphism investigations were completed. A summary of the solid forms found is shown in FIG. 17. The majority of the forms found were derived from solvent perturbation; in some cases stoichiometric solvates were isolated and in other cases non-stoichiometric solvates.

(283) Slurries of Polymorph A

(284) Solvent mediated equilibrations of Psilocybin Pattern A were conducted as a primary route into modification of the solid form and to visually assess the solubility of the material in a range of 24 solvents between 25 and 50° C.

(285) Psilocybin Pattern A (40 mg) was dosed into tubes at room temperature and solvents as listed in Table 34 were added in aliquots of 0.4 ml (10 vol.) to a total volume of 1.2 ml (30 vol.) and observations noted. The mixtures were agitated constantly. Heat cycling was conducted as follows: 50° C. for 18 hours, cool over 2 hours to 20° C., mature for 4 hours, heat to 50° C. for 4 hours, cool to 20° C. over 2 hours, mature for 18 hours. A repeat 50° C.-20° C. cycle over 24 hours was conducted and the following applied:

(286) Isolation post heating to 50° C. where solids were sufficient=A series

(287) Isolation post cooling to 20° C. where solids were sufficient=B series

(288) All isolated solids were dried in vacuo at 50° C. for 24 hours and analysed by XRPD. The observations are provided in Table 34.

(289) The API was largely insoluble in the solvents and solvent mixtures tested in 30 volumes at 50° C. resulting in heavy suspensions. Water did solubilise Psilocybin at 50° C.

(290) TABLE-US-00034 TABLE 34 Tabulated observations for heat cycling slurry maturations and using Pattern A blend as input Obs. Obs. Obs. 20° C., 20° C., 20° C., Obs. XRPD XRPD Entry Solvent 0.4 ml 0.8 ml 1.2 ml 50° C. A series B series 1 Cyclohexane Susp. Susp. Susp. Susp. A A 2 Chlorobenzene Susp. Susp. Susp. Susp. A A 3 2-Chlorobutane Susp. Susp. Susp. Susp. A A 4 Benzotrifluoride Susp. Susp. Susp. Susp. A A 5 Anisole Susp. Susp. Susp. Susp. A A 6 Nitromethane Susp. Susp. Susp. Susp. C C 7 CPME Susp. Susp. Susp. Susp. A A 8 Heptane Susp. Susp. Susp. Susp. A A 9 TBME Susp. Susp. Susp. Susp. C A 10 MIBK Susp. Susp. Susp. Susp. A A 11 MEK Susp. Susp. Susp. Susp. A A 12 iPrOAc Susp. Susp. Susp. Susp. C C 13 EtOAc Susp. Susp. Susp. Susp. A A 14 Toluene Susp. Susp. Susp. Susp. A A 15 THF Susp. Susp. Susp. Susp. A A 16 CHCl.sub.2 Susp. Susp. Susp. Susp. A A 17 MeOH Susp. Susp. Susp. Susp. D D 18 EtOH Susp. Susp. Susp. Susp. E E 19 IPA Susp. Susp. Susp. Susp. F F 20 MeCN Susp. Susp. Susp. Susp. C A 21 Water Susp. Susp. Susp. Solution n/a A 22 4:1 EtOH/water Susp. Susp. Susp. Susp. A A 23 4:1 THF/water Susp. Susp. Susp. Susp. A Hydrate A 24 4:1 IPA/water Susp. Susp. Susp. Susp. A C

(291) Results:

(292) In the figures (FIG. 18 and FIG. 19), “250” denotes isolation of the solid at 25° C. and “50 C” denotes isolation of the solid at 50° C. For example, GM832-20_50_A9 represents GM832 entry 20 (MeCN) isolated at 5000.

(293) 50° C. Slurries

(294) Entries 1, 2, 3, 4, 5, 7, 8, 10, 11, 13, 14, 15, 16, 22, 23, 24: XRPD diffractogram broadly consistent with Polymorph A, but with an additional peak of varying intensity at 18 °2θ.

(295) Entries 6, 9, 12, 20: XRPD diffractogram acquired for the isolated solids were broadly consistent (see FIG. 18) with additional peaks at 10 °2θ and 13.2 °2θ observed for some samples. This XRPD diffractogram was designated Pattern C. There is no chemotype correlation between the solvents (CH.sub.3NO.sub.2, TBME, iPrOAc and CH.sub.3CN).

(296) Entry 17: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern D.

(297) Entry 18: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern E.

(298) Entry 19: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern F.

(299) 25° C. Slurries:

(300) Entries 1, 2, 3, 4, 5, 7, 8, 9, 10, 11, 13, 14, 15, 16, 20, 21, 22: XRPD diffractograms are all similar to that acquired for Polymorph A.

(301) Entries 6, 12, 24: XRPD diffractograms acquired for the isolated solids were broadly consistent (see FIG. 18) with Pattern C.

(302) Entry 23: XRPD analysis showed Hydrate A had formed.

(303) Entry 17: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern D.

(304) Entry 18: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern E.

(305) Entry 19: XRPD diffractogram acquired had multiple diffraction peaks (FIG. 19). The XRPD diffractogram was designated Pattern F.

(306) Analysis of Results:

(307) The XRPD diffractograms for the solids isolated at 25° C. are broadly the same as for the XRPD diffractograms acquired for the solids isolated at 50° C.

(308) Patterns D, E and F were derived from alcohols (MeOH, EtOH and IPA). Solvated states were postulated considering an Ethanol Solvate was previously isolated during development. The XRPD diffractograms for the Ethanol Solvate are not identical, however, given that exact solvent level variation may deliver varying states of order within the lattice, the comparison between these XRPD diffractograms provides a strong hypothesis that these more significant phase variations are invoked by solvent entrainment.

(309) XRPD patterns D, E and F (FIG. 19) are all dissimilar to the XRPD diffractogram for Hydrate (FIG. 7d).

(310) Direct comparison of the XRPD diffractograms acquired for the MeOH, EtOH and IPA derived solids (Patterns D, E and F—FIG. 19) isolated at the two temperatures shows conforming diffractograms; the two MeOH diffractograms are similar while the EtOH and IPA are directly comparable.

(311) DSC analysis was performed on the isolated solids, and where sufficient sample was available, TGA. The solids that delivered Patterns D, E and F all features endotherms at ca. 170-180° C. but otherwise proffered distinct thermal profiles. TGA analysis for the MeOH slurry isolated solid showed one protracted mass loss from ca. 25-190° C. (3.1%). A stoichiometric methanol solvate would require 10.3% weight solvent. TGA analysis of the EtOH slurry isolated solid showed two distinct mass loss steps. The first one occurred before 100° C. (0.3% weight) is considered to be due to water, and the second larger loss (11.5% weight) due to solvent. A stoichiometric ethanol solvate requires 13.9% weight solvent. TGA analysis of the IPA slurry isolated solid featured two distinct mass loss steps. The first mass loss before 100° C. (0.4% weight) is considered to be due to water, while the second larger mass loss (13.9% weight) is considered to be due to residual solvent. A stoichiometric IPA solvate requires 17.5% weight solvent.

(312) Slurries of Amorphous Psilocybin

(313) In order to generate amorphous material a small sample of Psilocybin (0.5 g) was dissolved in water (0.5 L, 1000 vol.), polish filtered and lyophilised. Psilocybin was recovered as an off white fibrous material (lot MC1368A; 412 mg, 82%, XRPD amorphous).

(314) To assess visually solubility of the amorphous API and to induce form modification, a series of slurry maturations were performed as follows:

(315) Psilocybin (15 mg) was charged to tubes. Solvent was then added at ambient temperature (20° C., 0.3 ml, 20 vol.) and the suspensions agitated. Observations were made. After 1 hour of stirring, samples were heated to 45° C. for 18 hours and observations made. Samples were heated to 50° C. for 8 hours and observations were made. The samples were agitated for 72 hours at 25° C. and subject to a final heat cycle, prior to isolation. Observations are shown in Table 35.

(316) TABLE-US-00035 TABLE 35 Observations of amorphous Psilocybin during heat cycling slurry maturation and form fate Obs. at Obs. at Obs. at XRPD Entry Solvent 20° C. 45° C. 50° C. Data A Cyclohexane Susp. Susp. Susp. Semi-crystalline B Chlorobenzene Susp. Susp. Susp. Semi-crystalline C Chlorobutane Susp. Susp. Susp. Pattern B D Benzotrifluoride Susp. Susp. Susp. Semi-crystalline E Anisole Susp. Susp. Susp. Semi-crystalline F Nitromethane Susp. Susp. Susp. Pattern B G CPME Susp. Susp. Susp. Semi-crystalline H Heptane Susp. Susp. Susp. Semi-crystalline I TBME Susp Susp. Susp. Semi-crystalline J MIBK Susp. Susp. Susp. Semi-crystalline K MEK Susp. Susp. Susp. Semi-crystalline L iPrOAc Susp. Susp. Susp. Semi-crystalline M EtOAc Susp. Susp. Susp. Semi-crystalline N Toluene Susp. Susp. Susp. Similar to Solvate A O THF Susp. Susp. Susp. Semi-crystalline P Chloroform Susp. Susp. Susp. Similar to Pattern E R MeOH Susp. Susp. Susp. Semi-crystalline S EtOH Susp. Susp. Susp. Pattern D T IPA Susp. Susp. Susp. Pattern B U Acetonitrile Susp. Susp. Susp. Amorphous V Water Susp. Susp. Susp. Similar to Pattern A W 4:1 EtOH/water Susp. Susp. Susp. Similar to Pattern D X 4:1 THF/water Susp. Susp. Susp. Similar to Pattern A Y 4:1 IPA/water Susp. Susp. Susp. Similar to Pattern A

(317) Results

(318) The majority of solvents returned a solid that was considered to be semi-crystalline (predominantly amorphous with a notable reflection at ca. 18 °2θ).

(319) Truly amorphous was returned from equilibration in MeCN.

(320) Polymorph B was returned from chlorobutane, nitromethane and IPA (FIGS. 20 and 22).

(321) Pattern D, which was isolated from MeOH in the Polymorph A slurry experiments discussed above, was returned from the EtOH equilibration whereas MeOH in this study returned a semi-crystalline solid.

(322) Solids similar to Pattern A were recovered from water, THF:Water and IPA:Water (4:1).

(323) A solid similar to Pattern D was recovered from EtOH:Water (4:1), supporting the finding of the isolation of Pattern D from EtOH alone.

(324) A solid similar to Pattern E was recovered from Chloroform.

(325) From none of the solvents investigated was true Polymorph A or A′ isolated following extended equilibration and thermal maturation of amorphous Psilocybin.

Example 12—Formulation Development

(326) An initial series of experiments were conducted using formulations as set out in Table 36 below. The objective was to identify suitable single filler or combination fillers for large scale formulation.

(327) TABLE-US-00036 TABLE 36 Batch No (% w/w) APL-117- APL-117- APL-117- Material Name 6085-01 6085-02 6085-03 Psilocybin 1.0 1.0 1.0 Microcrystalline 91.5 49.5 81.5 cellulose Ph 102 Pregelatinised Starch — 45.0 — (Starch 1500) Compact Cel MAB — — 10 Hydroxypropyl 3.0 3.0 3.0 Cellulose (Klucel EXF) Sodium Starch 3.0 3.0 3.0 Glycolate Colloidal silicon 0.5 0.5 0.5 Dioxide Magnesium Stearate 1.0 1.0 1.0 (Vegetable Derived) Sodium Stearyl Fumarate — — — TOTAL 100.0 100.0 100.0

(328) The outcome, in terms of key physiochemical properties—Material flow, Blend Uniformity, and Content Uniformity are set out in Table 37 below:

(329) TABLE-US-00037 TABLE 37 Strength Material flow Blend Content Batch No (mg) (Carrs Index) Uniformity uniformity APL-117- 1.0 19.1 TOP = 127.9 % Label 6085-01 Middle = 106.4 claim = 92.4 Bottom = 104.5 AV = 7.9 Mean = 112.9 % RSD = 10.9 APL-117- 1.0 19.1 TOP = 115.9 % Label 6085-02 Middle = 106.6 claim = 95.2 Bottom = 106.1 AV = 5.9 Mean = 109.6 % RSD = 4.9 APL-117- 1.0 22.4 TOP = 105.0 % Label 6085-03 Middle = 101.4 claim = 96.3 Bottom = 98.7 AV = 4.6 Mean = 101.7 % RSD = 3.8

(330) Whilst batch (APL-117-6085-03) showed good blend uniformity across different sample analysed (TOP, MIDDLE and BOTTOM) and very good content uniformity its flow property (based on Carr's index) were towards the high end and it was predicted that the formulation would not accommodate higher drug loads.

(331) For this reason, a number of alternative formulations were trialed. The objective was to consider other filler combinations with the aim of improving the powder flow as well as achieving good blend uniformity and content uniformity.

(332) Formulations containing less Compactcel MAB and higher amount of glidant compared to Batch 3 (APL-117-6085-03) were also trialed

(333) These further formulations are set out in Table 38 below.

(334) TABLE-US-00038 TABLE 38 Batch No (% w/w) APL-117- APL-117- APL-117- Material Name 6085-05 6085-06 6085-07 Psilocybin 1.0 1.0 5.0 Microcrystalline cellulose Ph 102 — 89.0 85.0 Pregelatinised Starch (Starch 45.0 — — 1500) Compact Cel MAB — 5.0 5.0 Microcrystaline Cellulose 49.5 — — CEOLUS UF 702 Sodium Starch Glycolate 3.0 3.0 3.0 Colloidal silicon Dioxide 0.5 1.0 1.0 Sodium Stearyl Fumarate 1.0 1.0 1.0 TOTAL 100.0 100.0 100.0

(335) The results for these Batches are shown in Table 39 below:

(336) TABLE-US-00039 TABLE 39 Strength Material flow Blend Content Batch No (mg) (Carrs Index) Uniformity uniformity APL-117- 1.0 20.9 TOP = 130.0 % Label 6085-05 Middle = 105.4 claim = 88.3 Bottom = 107.2 AV = 16.5 Mean = 114.2 % RSD = 12.6% APL-117- 1.0 20.0 TOP = 107.0 % Label 6085-06 Middle = 96.2 claim = 96.2 Bottom = 95.5 AV = 10.5 Mean = 99.6 % RSD = 6.5 APL-117- 5.0 24.3 TOP = 91.5 % Label 6085-07 Middle = 94.2 claim = 96.0 Bottom = 94.8 AV = 11.9 Mean = 93.5 % RSD = 7.0

(337) APL-117-6085-05 failed to achieve good blend uniformity, and also failed on content PG-8T, uniformity criteria.

(338) APL-117-6085-06 and APL-117-6085-07 both exhibited improved powder flow, but the blend uniformity for both formulations was poorer than APL-117-6085-03.

(339) As a consequence, Applicant looked at modified excipients and more particularly silicified fillers with different particle sizes. These formulations are set out in Table 40 below:

(340) TABLE-US-00040 TABLE 40 Batch No (% w/w) Material Name APL-117-6085-11 APL-117-6085-12 Psilocybin 5.0 1.0 Prosolv 50 10.5 15.5 Prosolv 90 80.0 79.0 Sodium Starch Glycolate 3.0 3.0 Colloidal silicon Dioxide 0.5 0.5 Sodium Stearyl Fumarate 1.0 1.0 TOTAL 100.0 100.0

(341) Prosolv is a silicified microcrystalline cellulose, and the two variants were selected to determine if particle size affected outcome. Compared to standard microcrystalline cellulose (typical size range, depending on sieving, is 80-350 microns) the Prosolv has a finer particle size distribution, and that gives an increased surface area. The increased surface area it was hypothesised might provide superior flow and increased compaction together with improved content uniformity and stability in the formulation. The ratio of Prosolv 50 and Prosolv 90 was to produce a particle size distribution across both finer and coarser particles.

(342) The results are set out in Table 41 below

(343) TABLE-US-00041 TABLE 41 Strength Material flow Blend Content Batch No (mg) (Carrs Index) Uniformity uniformity APL-117- 5.0 24.3 TOP = 103.4 % Label 6085-11 Middle = 100.4 claim = 94.1 Bottom = 100.2 AV = 6.0 Mean = 101.5 % RSD = 2.0 APL-117- 1.0 21.1 TOP = 101.9 % Label 6085-12 Middle = 98.4 claim = 100.5 Bottom = 99.9 AV = 5.8 Mean = 100.1 % RSD = 3.8%

(344) It can be seen that the key parameters of content uniformity (greater than 90%, and in fact greater than 94%) and AV (less than 10, and in fact less than 7) are excellent as is the consistency in blend uniformity (greater than 95% allowing for error).