Use of a Curtius Rearrangement as part of the Multikilogram

Dec 4, 2017 - Commercial route definition for a glucokinase activator called for a reevaluation of the synthesis and processes used to access multikil...
17 downloads 15 Views 461KB Size
Subscriber access provided by Gothenburg University Library

Full Paper

Use of a Curtius Rearrangement as part of the Multikilogram Manufacture of a Pyrazine Building Block Alan Steven, and Phillip Hopes Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.7b00340 • Publication Date (Web): 04 Dec 2017 Downloaded from http://pubs.acs.org on December 4, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Organic Process Research & Development is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Use of a Curtius Rearrangement as part of the Multikilogram Manufacture of a Pyrazine Building Block Alan Steven*,† and Phillip Hopes‡ †

Pharmaceutical Technology & Development, AstraZeneca, Charter Way, Macclesfield, SK10 2NA, United Kingdom



Cyton Biosciences Ltd, 68 Macrae Road, Bristol, BS20 0DD, United Kingdom [email protected]

ACS Paragon Plus Environment

Organic Process Research & Development 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Paragon Plus Environment

Page 2 of 17

Page 3 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

ABSTRACT Commercial route definition for a glucokinase activator called for a reevaluation of the synthesis and processes used to access multikilogram quantities of 2-amino-5methylpyrazine. After the consideration of different options, a variation of the Curtius rearrangement used by the medicinal chemistry route was selected for further development. The formation of an acyl azide for the Curtius rearrangement required a process safety control strategy to be put in place. The process developed was used to successfully deliver multikilogram quantities of 2amino-5-methylpyrazine in an overall yield of 68%, starting from 5-methylpyrazine-2carboxylic acid.

KEY WORDS acyl azide, Curtius rearrangement, glucokinase activator, process safety, pyrazine

ACS Paragon Plus Environment

Organic Process Research & Development 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 17

INTRODUCTION 2-Amino-5-methylpyrazine (1) is a potential regulatory starting material for the synthesis of AZD1656 (2), a glucokinase activator that has been assessed in Phase IIb clinical trials for the treatment of Type II diabetes.1-3 Delivering 2-amino-5-methylpyrazine (1) of high quality is important as it is introduced in the final chemical step used to synthesise the API (Scheme 1). Early experiences with the use of the medicinal chemistry route to deliver multikilogram quantities of the material will be described. Alternative routes will then be discussed. The development and scaleup of the option selected, a variant of the medicinal chemistry route with its use of the Curtius rearrangement, will then be provided.

Scheme 1. Structure of AZD1656 (2) Me Me

N N

N N

O O

N H O

OMe

N

N H2 1

N

N O

AZD1656 (2)

EARLY EXPERIENCES WITH MEDICINAL CHEMISTRY ROUTE When AZD1656 passed into development, the medicinal chemistry approach to the 2-amino5-methylpyrazine (1) building block was initially retained.

This approach involved

converting 5-methylpyrazine-2-carboxylic acid (3) to Boc carbamate 4 using a Curtius rearrangement (Scheme 2). As part of this first scaleup campaign, 72.5 kg of 2-amino-5methylpyrazine (1) were delivered with minor modifications.

Attempts to convert 5-

methylpyrazine-2-carboxylic acid (3) to the acyl azide 5 via the acid chloride, mesylate or isobutyl mixed anhydride were unsuccessful, either from the point of view of conversion or

ACS Paragon Plus Environment

Page 5 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

byproduct formation, leading to the continued use of diphenylphosphoryl azide (DPPA).4-7 This reagent allows carbonyl activation, and azidation to be combined into a single operation as part of Curtius rearrangements used to deliver multikilogram quantities of material.8,9 The use of a large excess of tert-butanol was retained due to its only modest ability at trapping the isocyanate arising from the Curtius rearrangement of the acyl azide intermediate 5. This brought with it the inconvenience of melting the tert-butanol prior to charging and the use of warm (>30 °C) water in the condenser to prevent blockages caused by its freezing. Whilst our medicinal chemistry colleagues had used neat trifluoroacetic acid to deprotect Boc carbamate 4, we successfully reduced the molar excess of TFA to 20% whilst using water as the bulk solvent.

Finally, passing the initially produced 2-amino-5-methylpyrazine (1)

through a plug of silica gel was used to remove coloured impurities.

Scheme 2. Medicinal chemistry-based approach to 1 Me

N

N a

HO

N O

(60 %)

N3

N O

3

N Boc N H

N

Me

6

N

b (72 %) H N 2

Me

N 1

a) i) DPPA,

O

5

4

Me

N

Me

t BuOH, iPr EtN, 2

PhMe, ii) NaOH (aq); b) TFA, water

N NH N H

N N

Me

Our orientation with this sequence alerted us to a number of minor drawbacks. Small quantities (0.1–0.2 % wt/wt) of alcohol congeners in the tert-butanol trapped the isocyanate more efficiently than the tert-butanol itself, leading to the contamination of Boc carbamate 4

ACS Paragon Plus Environment

Organic Process Research & Development 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 17

with related substances. Boc carbamate 4 also showed some thermal lability under the conditions of the Curtius rearrangement. This led to the premature formation of a small quantity of 2-amino-5-methylpyrazine (1) which was unproductively trapped as a urea impurity 6 by the isocyanate intermediate.

2-Amino-5-methylpyrazine (1) also had an

appreciable water solubility. This meant multiple (4) extractions with n-propyl acetate were required when working up the deprotection reaction, in order to minimize losses to the aqueous phase. A dilute process which generated a large amount of waste was the result.

CONSIDERATION OF ALTERNATIVE APPROACHES As demands for the drug substance increased, it became apparent that the route shown in Scheme 2 was eligible for re-evaluation. The hydrolysis of the isocyanate intermediate generated over the course of the Curtius rearrangement of 5-methylpyrazine-2-carboxylic acid (3) was recognised as offering the opportunity to streamline the synthesis, avoiding the need to route via a carbamate intermediate. This was not explored, however, as the isolation of a carbamate intermediate was recognised as offering a useful means of rejecting impurities. In addition, the use of water in an isocyanate hydrolysis was recognised as hindering the recovery of the hydrophilic 2-amino-5-methylpyrazine (1). The elaboration of 5-methylpyrazine-2-carboxamide

to

5-methylpyrazin-2-amine

(1)

using

potassium

hypochlorite (i.e. a Hofmann rearrangement), whilst successful, was again not progressed due to the impact of the aqueous reaction conditions on the isolation of 5-methylpyrazin-2-amine (1). The oxidation of 2,5-dimethylpyrazine was ideated, but not explored, due to concerns over the contamination of commercial grades of 2,5-dimethylpyrazine with regioisomeric material that would compromise the impurity profile of the API.10

ACS Paragon Plus Environment

Page 7 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

The Taylor condensation of aminomalononitrile tosylate 7 with isonitroso ketone 8 to form pyrazine-N-oxide 9 was one option that was successfully demonstrated (Scheme 3).11 The double reduction of this material afforded 2-amino-5-methylpyrazine (1).12,13 This route was potentially much cheaper than one that started from 5-methylpyrazine-2-carboxylic acid (3), though cost of goods pressures were insufficient to force its development. This option also had the drawbacks of generating hydrogen cyanide and requiring elevated pressures (40 psi) for the second step. Tentative attempts at the iron(III)-promoted formation of the pyrazine Noxide of 2-amino-5-methylpyrazine (1) from the condensation of aminoacetonitrile and the aldoxime of acetoacetaldehyde were also studied but produced messy reaction profiles.14

Scheme 3. Taylor condensation approach to 1

NC

NH3 OTs 7

CN +

(63 %)

O HON

NC

N

H 2N

N O 9

8

b (69 %) H 2 N

Me

a

N

Me

N 1

a) MeOH, water; b) H2, Pt-C (5 %wt/wt), Darco KB-B, MeOH, AcOH

The option that was progressed retained the use of a Curtius rearrangement but switching the

tert-butyl carbamate used in the medicinal chemistry approach to the benzyl carbamate (Scheme 4). The deprotection of the latter by hydrogenolysis with a supported metal catalyst that could be screened off at the end of the reaction was envisaged as eschewing the need for an aqueous workup, making it easier to isolate the water-soluble 2-amino-5-methylpyrazine (1) in high yield. Gratifyingly, and as might have been expected, the substitution of the tert-

ACS Paragon Plus Environment

Organic Process Research & Development 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

butanol by the more reactive benzyl alcohol led to the formation of the corresponding carbamate product 10 in a cleaner fashion.

Azeodrying toluene solutions of the 5-

methylpyrazine-2-carboxylic acid and benzyl alcohol prior to their use prevented any isocyanate hydrolysis that would have led to the formation of urea 6 as a byproduct. This protocol also eliminated the risk of moisture hydrolysing the DPPA reagent to volatile (bp 37

°C) and explosive hydrazoic acid. Scheme 4. Curtius approach to 1

a 3

5 b

10

1

a) i) DPPA, toluene, iPr2NEt, 13-17 °C, ii) BnOH, 85-90 °C; b) H2, Pd-C (E196 NN/W, 5%), MeOH, NaOH

SCALEUP OF REDEVELOPED ROUTE Inhouse knowledge from Carius tube testing of a 30 % wt/v solution of DPPA in toluene, in the presence of mild and stainless steel, showed the material to be thermally stable to temperatures well above the boiling point associated with plant scale operation.15 However, the intention to utilise a Curtius rearrangement for the route used to supply commercial quantities of AZD1656 required the addressing of safety concerns around the handling of other azide-containing process streams associated with the chemistry shown in Scheme 4.16,17 In order to control the amount of potentially explosive sodium azide that could end up in benzyl carbamate 10 as a byproduct, equimolar amounts of DPPA and acid 3 were employed and a reaction conversion criterion set at no more than 5 % residual starting material.

ACS Paragon Plus Environment

Page 8 of 17

Page 9 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

Isothermal calorimetry using a Mettler RC1 calorimeter indicated a heat of reaction for the addition of the DPPA to a solution of a solution of benzyl carbamate and N,Ndiisopropylethylamine of –77.2 kJ/mol 3, corresponding to an adiabatic temperature rise of 30 °C. Once the addition had completed, accelerating rate calorimetry flagged that selfheating, leading to a thermal runaway, could occur at ambient plant temperatures. The adiabatic temperature rise for the corresponding decomposition was calculated as 69 °C. Whilst this temperature rise, on top of the maximum temperature of the synthesis reaction (47

°C for a starting reaction temperature of 17 °C), could not be achieved in full due to the use of toluene as solvent (bp 111 °C), a control strategy for the accompanying evolution of gas was still required.

Given that the manufacture was to take place in an equipment train that has to deal with multiple products, a preventative approach was targeted as the basis of safety.18 This involved controlling the addition of the DPPA so that it took place over at least 2 h, and could be halted in the event of a failure of either cooling or agitation, as well as limiting the temperature of the batch to no more than 17 °C. Crucially, this temperature was below the reaction temperature (21–23 °C) that was evaluated as giving a 24 h window before the decomposition of acyl azide 5 reached its maximum rate if held under adiabatic conditions.19 Heat accumulation at the end of the addition was evaluated as corresponding to an adiabatic temperature rise of 4 °C. When taken in conjunction with the adiabatic decomposition exotherm (69 °C), a cooling failure could still result in the decomposition, and the accompanying gas evolution, reaching its maximum rate below the boiling point of the solvent. As the maximum rate of gas evolution under thermal runaway conditions had not been determined, a fast dump into aqueous sodium hydroxide solution was built into the control strategy in the event of the batch temperature exceeding 30 °C.20

ACS Paragon Plus Environment

Organic Process Research & Development 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Isothermal calorimetry yielded a heat of reaction for the Curtius rearrangement of acyl azide 5 of –306 kJ/mol 3. The corresponding adiabatic temperature rise when its solution (without prior heating) was added to a benzyl alcohol solution at 85 °C was calculated as being 21 °C. More pressing than controlling this heat was the need to ensure the controlled disengagement of the nitrogen generated by the reaction. As shown in Figure 1, when the batch temperature was kept above 85 °C and the addition of the acyl azide solution took place over 2 h, gas generation was pro rata.21 After the addition, DSC analysis indicated that decomposition would only take place well above the boiling point of the batch.

Figure 1. Heat and gas generation during the addition of the acyl azide solution to a solution of benzyl alcohol The hydrogenolysis of benzyl carbamate 10 to 2-amino-5-methylpyrazine (1) proved relatively straightforward to develop. The exothermicity of the reaction associated with the

ACS Paragon Plus Environment

Page 10 of 17

Page 11 of 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Organic Process Research & Development

catalyst of choice, palladium on charcoal catalyst E196 NN/W (supplied by Evonik Industries AG), was controlled by limiting the hydrogen pressure to just one atmosphere. At the end of each hydrogenolysis, an impurity was detected whose 1H NMR spectrum indicated the presence of methylene signals at δH = 3.6 and 3.7, one of which showed an nOe to the methyl group (Scheme 5). This was attributed to the overreduction of 2-amino-5-methylpyrazine (1) to dihydropyrazine 11. This pathway was postulated as being promoted by trace quantities of acidic impurities which could protonate the pyrazine ring. It was duly discouraged by hydrogenolysing in the presence of a catalytic amount (1 mol%) of sodium hydroxide. This control was used in tandem with stopping the hydrogenolysis as soon as the amount of residual benzyl carbamate dropped below 0.10 LCAP.

Filtering off the hydrogenolysis

catalyst under nitrogen, whilst necessary for reasons of safety, also eliminated the risk of oxidised palladium species leaching into the reaction liquors. The palladium content of the filtrate was routinely