Use of a Curtius Rearrangement as Part of the Multikilogram Manufacture of a Pyrazine Building Block

Article abstract of DOI:10.1021/acs.oprd.7b00340

Commercial route definition for a glucokinase activator called for a reevaluation of the synthesis and processes used to access multikilogram quantities of 2-amino-5-methylpyrazine. After 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 2-amino-5-methylpyrazine in an overall yield of 68%, starting from 5-methylpyrazine-2-carboxylic acid.

Full text of DOI:10.1021/acs.oprd.7b00340

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
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Use of a Curtius Rearrangement as part of the Multikilogram Manufacture
of a Pyrazine Building Block
*,
†
‡
Alan Steven and Phillip Hopes
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
Pharmaceutical Technology & Development, AstraZeneca, Charter Way, Macclesfield,
SK10 2NA, United Kingdom
‡
Cyton Biosciences Ltd, 68 Macrae Road, Bristol, BS20 0DD, United Kingdom
Alan.Steven@astrazeneca.com
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ABSTRACT
Commercial route definition for a glucokinase activator called for a reevaluation of the
synthesis and processes used to access multikilogram quantities of 2ꢀaminoꢀ5ꢀ
methylpyrazine.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
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 2ꢀ
aminoꢀ5ꢀmethylpyrazine in an overall yield of 68%, starting from 5ꢀmethylpyrazineꢀ2ꢀ
carboxylic acid.
KEY WORDS
acyl azide, Curtius rearrangement, glucokinase activator, process safety, pyrazine
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
2ꢀAminoꢀ5ꢀmethylpyrazine (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
1ꢀ3
the treatment of Type II diabetes. Delivering 2ꢀaminoꢀ5ꢀmethylpyrazine (1) of high quality
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
is important as it is introduced in the final chemical step used to synthesise the API (Scheme
). Early experiences with the use of the medicinal chemistry route to deliver multikilogram
1
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
N
N
O
Me
O
N
OMe
N
H
N
N
O
N
N
NH2
1
O
AZD1656 (2)
EARLY EXPERIENCES WITH MEDICINAL CHEMISTRY ROUTE
When AZD1656 passed into development, the medicinal chemistry approach to the 2ꢀaminoꢀ
5
ꢀmethylpyrazine (
converting 5ꢀmethylpyrazineꢀ2ꢀcarboxylic acid (
rearrangement (Scheme 2). As part of this first scaleup campaign, 72.5 kg of 2ꢀaminoꢀ5ꢀ
methylpyrazine ( ) were delivered with minor modifications. Attempts to convert 5ꢀ
methylpyrazineꢀ2ꢀcarboxylic acid ( ) to the acyl azide 5 via the acid chloride, mesylate or
1
) building block was initially retained. This approach involved
3
) to Boc carbamate using a Curtius
4
1
3
isobutyl mixed anhydride were unsuccessful, either from the point of view of conversion or
ACS Paragon Plus Environment

Page 5 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
4
ꢀ7
byproduct formation, leading to the continued use of diphenylphosphoryl azide (DPPA).
This reagent allows carbonyl activation, and azidation to be combined into a single operation
8,9
as part of Curtius rearrangements used to deliver multikilogram quantities of material. The
use of a large excess of tertꢀbutanol was retained due to its only modest ability at trapping the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
isocyanate arising from the Curtius rearrangement of the acyl azide intermediate 5. This
brought with it the inconvenience of melting the tertꢀbutanol 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 2ꢀaminoꢀ5ꢀmethylpyrazine (
through a plug of silica gel was used to remove coloured impurities.
1)
Scheme 2. Medicinal chemistryꢀbased approach to
1
N
Me
(
N
N
Me
a
HO
N3
N
60 %)
O
O
3
5
N
N
Me
N
Me
b
(72 %)
Boc N
H
H N
N
2
4
1
Me
N
t
i
a) i) DPPA, BuOH, Pr EtN,
2
PhMe, ii) NaOH (aq);
b) TFA, water
N
NH
O
N
N
H
6
Me
N
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 tertꢀbutanol trapped the isocyanate
more efficiently than the tertꢀbutanol itself, leading to the contamination of Boc carbamate
4
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
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 2ꢀaminoꢀ5ꢀmethylpyrazine (
impurity by the isocyanate intermediate. 2ꢀAminoꢀ5ꢀmethylpyrazine (
appreciable water solubility. This meant multiple (4) extractions with ꢀpropyl acetate were
1
) which was unproductively trapped as a urea
6
1
) also had an
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
n
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 reꢀevaluation. The hydrolysis of the isocyanate intermediate
generated over the course of the Curtius rearrangement of 5ꢀmethylpyrazineꢀ2ꢀcarboxylic
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 2ꢀaminoꢀ5ꢀmethylpyrazine ( ). The elaboration of
1
5ꢀmethylpyrazineꢀ2ꢀcarboxamide to 5ꢀmethylpyrazinꢀ2ꢀamine (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 5ꢀmethylpyrazinꢀ2ꢀamine
(1
). The oxidation of 2,5ꢀdimethylpyrazine was ideated, but not explored, due to concerns
over the contamination of commercial grades of 2,5ꢀdimethylpyrazine with regioisomeric
1
0
material that would compromise the impurity profile of the API.
ACS Paragon Plus Environment

Page 7 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The Taylor condensation of aminomalononitrile tosylate
pyrazineꢀ ꢀoxide was one option that was successfully demonstrated (Scheme 3). The
double reduction of this material afforded 2ꢀaminoꢀ5ꢀmethylpyrazine (
potentially much cheaper than one that started from 5ꢀmethylpyrazineꢀ2ꢀcarboxylic acid (
7 with isonitroso ketone 8 to form
11
N
9
12,13
1
).
This route was
),
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
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
Nꢀ
oxide of 2ꢀaminoꢀ5ꢀmethylpyrazine ( ) from the condensation of aminoacetonitrile and the
1
14
aldoxime of acetoacetaldehyde were also studied but produced messy reaction profiles.
Scheme 3. Taylor condensation approach to 1
NC
NH OTs
3
NC
H N
N
Me
7
CN
a
+
(
63 %)
N
O
O
2
HON
8
9
N
Me
b
(
69 %) H N
N
2
1
a) MeOH, water; b) H , Pt-C (5 %wt/wt),
2
Darco KB-B, MeOH, AcOH
The option that was progressed retained the use of a Curtius rearrangement but switching the
tertꢀbutyl 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 waterꢀsoluble 2ꢀaminoꢀ5ꢀmethylpyrazine
(1
) in high yield. Gratifyingly, and as might have been expected, the substitution of the tertꢀ
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
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 5ꢀ
methylpyrazineꢀ2ꢀcarboxylic acid and benzyl alcohol prior to their use prevented any
isocyanate hydrolysis that would have led to the formation of urea
protocol also eliminated the risk of moisture hydrolysing the DPPA reagent to volatile (bp 37
C) and explosive hydrazoic acid.
6 as a byproduct. This
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
°
Scheme 4. Curtius approach to
1
a
3
5
b
10
1
i
a) i) DPPA, toluene, Pr NEt, 13-17 °C, ii) BnOH,
2
8
5-90 °C; b) H , Pd-C (E196 NN/W, 5%), MeOH,
2
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
15
temperatures well above the boiling point associated with plant scale operation. 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
16,17
other azideꢀcontaining process streams associated with the chemistry shown in Scheme 4.
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 were employed
3
and a reaction conversion criterion set at no more than 5 % residual starting material.
ACS Paragon Plus Environment

Page 9 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
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
diisopropylethylamine of –77.2 kJ/mol , corresponding to an adiabatic temperature rise of
C. Once the addition had completed, accelerating rate calorimetry flagged that selfꢀ
heating, 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
C), a control strategy for the accompanying evolution of gas was
N,Nꢀ
3
30 °
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
°
°
°
toluene as solvent (bp 111
still required.
°
Given that the manufacture was to take place in an equipment train that has to deal with
18
multiple products, a preventative approach was targeted as the basis of safety. 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
19
decomposition of acyl azide
5 reached its maximum rate if held under adiabatic conditions.
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
20
control strategy in the event of the batch temperature exceeding 30 °C.
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
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
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
generation was pro rata. 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 2ꢀaminoꢀ5ꢀmethylpyrazine (1) proved
relatively straightforward to develop. The exothermicity of the reaction associated with the
ACS Paragon Plus Environment

Page 11 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
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
1
each hydrogenolysis, an impurity was detected whose H NMR spectrum indicated the
H
presence of methylene signals at δ = 3.6 and 3.7, one of which showed an nOe to the methyl
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
group (Scheme 5). This was attributed to the overreduction of 2ꢀaminoꢀ5ꢀmethylpyrazine (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 <10 ppm, and that of isolated 2ꢀaminoꢀ5ꢀmethylpyrazine (1) was below
<
0.5 ppm. The small quantities (0.10−0.15 LCAP) of dihydropyrazine that did form in spite
of the use of base in the reaction could be oxidised back to 2ꢀaminoꢀ5ꢀmethylpyrazine (1) by
passing a stream of 1% oxygen in nitrogen through the batch. This was at the expense of the
introduction of colouration into the batch.
Scheme 5. Overreduction side reaction
1
nOe
11
Summary
The manufacture of 2ꢀaminoꢀ5ꢀmethylpyrazine (1) has been achieved using a set of processes
that rely on the implementation of a Curtius rearrangement. The safe operation of the process
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
relied on the controlled additions of DPPA to acid 3 and the acyl azide 5 so formed to benzyl
alcohol. By routing via a benzyl carbamate instead of a tertꢀbutyl carbamate, the overall
yield for the processes was able to be improved from 43 % in early scaleup deliveries to 68
%
. The process mass intensity (62 kg/kg 1) for the revised processes had been reduced by 56
, largely due to the removal of an aqueous workup from the step used to form 1. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
%
processes described herein were used in two separate manufacturing campaigns, each of
which produced circa 200 kg of material.
Experimental Section
General. NMR spectra were recorded on a Bruker instrument with tetramethylsilane (TMS)
as internal reference. Chemical shifts are expressed in ppm (δ) relative to TMS, coupling
constants (J) are in Hz. Infrared spectra were recorded with diamond ATR sampling on an
FTIR spectrometer using powdered samples. Melting points were recorded using a Q2000
differential scanning calorimeter from TA Instruments. High resolution mass spectra were
recorded on a Waters Synapt UPLCꢀMS instrument. Unless stated otherwise, commercial
grade materials were used.
Preparation of benzyl carbamate 10. CAUTION: This step used nonmetallic, glassꢀlined
22
equipment with inert transfer lines. CAUTION: The toluene solution of the acyl azide
intermediate 5 poses a thermal and pressure hazard. It should not be isolated from solution as
it can potentially decompose. A solution (solution A) of 5ꢀmethylpyrazineꢀ2ꢀcarboxylic acid
(3) (170 kg, 1.23 kmol) and diisopropylethylamine (322 L, 1.85 kmol) in toluene (510 L) was
vacuum distilled with a jacket temperature of 75 °C so that toluene (170 L) was removed. A
solution (solution B) of benzyl alcohol (382 L, 3.69 kmol) in toluene (2,040 L) was distilled
at atmospheric pressure with a jacket temperature of 130 °C so that 340 L of toluene were
removed.
With the temperature of solution A being maintained at 13−17 °C, a toluene
ACS Paragon Plus Environment

Page 13 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
solution of diphenylphosphoryl azide (1140 L at 29.7 % wt/v, 1.23 kmol) was added
(CAUTION: EXOTHERM if addition is not controlled, CAUTION: reagent can cause severe
burns and eye damage) over a minimum of 2 h. Rinsing through was performed with
additional toluene (68 L). CAUTION: The solution should be quenched into a solution
consisting of 2.2 kg of 1 M sodium hydroxide solution for every 1 kg of acyl azide solution
should the bulk temperature of the latter exceed 30 °C. The temperature of solution B was
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
then adjusted to 86−88 °C, and the preꢀprepared acyl azide solution was added over 2 h,
washing through with additional toluene (68 L). At the end of the addition, the reaction was
aged for a further 60 min, before sodium hydroxide solution (225 L of 29 % wt/wt diluted
with 1428 L water) was added over a minimum of 10 min. The reaction mixture was
maintained for a further 60 min. before the slurry was filtered. The cake was washed with
water (1,020 L) and methanol (340 L) before being dried below 55 °C. This afforded benzyl
carbamate 10 as a white solid (226 kg, 75.6%). The pH of the liquors and cake wash were
checked as being >12 before they were transferred into nonmetallic containers for disposal.
ꢀ1
1
6
Mp 191.4 °C. IR 1719, 1567, 1353, 1226, 1054, 759, 744, 706 cm . H NMR (DMSOꢀd ,
5
3
00 MHz) 10.45 (br, 1H), 8.93 (s, 1H), 8.20 (s, 1H), 7.42–7.32 (m, 5H), 5.18 (s, 2H), 2.41 (s,
1
3
H). C NMR (DMSOꢀd6, 125.7 MHz) 153.6, 147.7, 146.7, 141.7, 136.5, 134.3, 128.7 (2C),
+
128.4, 128.2 (2C), 66.5, 20.4. HRMS elemental calculated for C H N O (MH ): 244.1086;
13
14
3
2
found: 244.1080.
Preparation of 2-amino-5-methylpyrazine (1). A suspension of benzyl carbamate 10
(259.6 kg, 1.07 kmol) in methanol (1300 L) was prepared. Sodium hydroxide solution (1.54
kg of 29 % wt/wt, 11.2 mol) was added. A slurry of Pd/C catalyst E196 NN/W (5 % wt/wt
Pd on dry basis, 14.2 kg paste) in methanol (100 L) was added, and washed through with
additional methanol (15 L). The headspace was purged with hydrogen gas.
The
hydrogenolysis was performed with a headspace hydrogen pressure of 1 atm at 20 °C, leading
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to the dissolution of suspended solids. When the amount of residual benzyl carbamate 10
was <0.10%, the headspace was purged with nitrogen. A slurry of Norit SX Ultra charcoal
(13.0 kg) in methanol (100 L) was added and washed through with more methanol (15 L).
After stirring for 45 min, the solution was sequentially passed through a pad of arbocel
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(dicacel 4), a 1
ꢀm filter and then a 0.45 ꢀm filter, washing through with additional methanol
(285 kg). If necessary, a 1% oxygen in nitrogen gas mixture was passed through the filtrate
23
until there was no overreduced product remaining according to an HPLC check. A vacuum
distillation was then performed at 45 °C so as to remove 1440 L of methanol. A put (389 L
toluene added) and take (440 L) vacuum distillation was then performed at 45 °C. Three
more put (3 x 133 L toluene) and takes (3 x approximately 145 L of distillate) were
performed (CARE: 2ꢀaminoꢀ5ꢀmethylpyrazine (1) sublimes). The batch was then cooled to 5
°C at 10 °C/h and stirred at this temperature for a further 2 h, over which time the insoluble
liquid 2ꢀaminoꢀ5ꢀmethylpyrazine (1) layer crystallised. The slurry was filtered and the cake
washed with toluene (2 x 133 L) which had been prechilled to 0 °C. The cake was vacuum
dried below 40 °C in a biconical dryer or an agitated pan dryer and milled. This process
afforded 2ꢀaminoꢀ5ꢀmethylpyrazine (1) as a white solid (104.3 kg, 89.6%). Mp 115.7 °C. IR
ꢀ1
1
3
323, 3145, 1658, 1540, 1485, 1387, 1373, 1189, 1029, 875 cm . H NMR (DMSOꢀd
6
, 400
13
MHz) 7.79 (1H, d,
J= 1.4 Hz), 7.75 (1H, br s), 6.07 (2H, br s), 2.23 (3H, s). C NMR
(DMSOꢀd , 100.6 MHz) 154.0, 140.4, 139.3, 131.1, 19.6. HRMS elemental calculated for
6
+
C
5
H
8
N
3
(MH ): 110.0718; found: 110.0708.
ASSOCIATED CONTENT
1
13
Supporting information available: H NMR spectra, C NMR spectra, infrared spectra and
DSC traces for benzyl carbamate 10 and 2ꢀaminoꢀ5ꢀmethylpyrazine (1).
AUTHOR INFORMATION
ACS Paragon Plus Environment

Page 15 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Corresponding Author
Eꢀmail: Alan.Steven@astrazeneca.com
ORCID Alan Steven: 0000ꢀ0002ꢀ0134ꢀ0918
Notes
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The authors declare no competing financial interest.
Acknowledgement
We are grateful to Ajinomoto OmniChem (Wetteren) for performing the manufacture and
gathering process safety data, and Paul Gillespie (AstraZeneca) for assisting with the
interpretation of these data.
References
1. McKerrecher, D.; Waring, M. J. Prog. Med. Chem. 2013, 52, 1−43.
2. Waring, M. J.; Clarke, D. S.; Fenwick, M. D.; Godfrey, L.; Groombridge, S. D.;
Johnstone, C.; McKerrecher, D.; Pike, K. G.; Rayner, J. W.; Robb, G. R.; Wilson, I.
Med. Chem. Commun., 2012, 3, 1077–1081.
3
. Processes for making AZD1656 are covered in the following patents. a) Butters, M.;
Crabb, J.; Hopes, P.; Patel, B. Process for preparation of benzoyl amino heterocyclic
compounds. WO 2010092387, August 19, 2010. b) Bowden, S. A.; Hoile, D. P.;
Lövqvist, K. Crystalline polymorphic form 631. WO 2010092386, August 19, 2010. c)
Briggner, L.ꢀE.; Lindfors, L.; Smith, C. M. Pharmaceutical formulations. WO
2012007758, January 19, 2012.
4. Shioiri, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc. 1972, 94, 6203–6205.
5. Nebuloni, M.; Cardillo, P. Journ. Calorim., Anal. Therm. Thermodyn. Chim. 1986, 17
,
,
398–9.
6
. Bräse, S.; Gil, C.; Knepper, K.; Zimmermann, V. Angew. Chem., Int. Ed. 2005, 44
188–5240.
5
7. Liang, H. Synlett 2008, 16, 2554–2555.
8
. For semiꢀbatch operation, see: (a) Yue, T.ꢀY.; McLeod, D. D.; Albertson, K. B.; Beck, S.
R.; Deerberg, J.; Fortunak, J. M.; Nugent, W. A.; Radesca, L. A.; Tang, L.; Xiang, C. D.
Org. Process Res. Dev. 2006, 10, 262–271. (b) Beliaev, A.; Wahnon, J.; Russo, D. Org.
Process Res. Dev. 2012, 16, 704–709. (c) Grongsaard, P.; Bulger, P. G.; Wallace, D. J.;
Tan, L.; Chen, Q.; Dolman, S. J.; Nyrop, J.; Hoerrner, R. S.; Weisel, M.; Arredondo, J.;
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Itoh, T.; Xie, C.; Wen, X.; Zhao, D.; Muzzio, D. J.; Bassan, E. M.; Shultz, C. S. Org.
Process Res. Dev. 2012, 16, 1069–1081. (d) Tudhope, S. R.; Bellamy, J. A.; Ball, A.;
Rajasekar, D.; AzadiꢀArdakani, M.; Meera, H. S.; Gnanadeepam, J. M.; Saiganesh, R.;
Gibson, F.; He, L.; Behrens, C. H.; Underiner, G.; Marfurt, J.; Favre, N. Org. Process
Res. Dev. 2012, 16, 635–642. (e) Deerberg, J.; Prasad, S. J.; Sfouggatakis, C.; Eastgate,
M. D.; Fan, Y.; Chidambaram, R.; Sharma, P.; Li, L.; Schild, R.; Müslehiddinoğlu, J.;
Chung, H.ꢀJ.; Leung, S.; Rosso, V. Org. Process Res. Dev. 2016, 20, 1949–1966. (f)
Smith, M. J.; Lawler, M. J.; Kopp, N.; McLeod, D. D.; Davulcu, A. H.; Lin, D.;
Katipally, K.; Sfouggatakis, C. Org. Process Res. Dev. 2017, 21, 1859–1863.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
9. For the use of flow chemistry to derisk process safety liabilities associated with the
Curtius rearrangement, see: (a) Carter, C. F.; Lange, H.; Ley, S. V.; Baxendale, I. R.;
Wittkamp, B.; Goode, J. G.; Gaunt, N. L. Org. Process Res. Dev. 2010, 14, 393–404. (b)
Filipponi, P.; Ostacolo, C.; Novellino, E.; Pellicciari, R.; Gioiello, A., Org. Process Res.
Dev. 2014, 18, 1345–1353. (c) Marsini, M. A.; Buono, F. G.; Lorenz, J. C.; Yang, B.ꢀS.;
Reeves, J. T.; Sidhu, K.; Sarvestani, M.; Tan, Z.; Zhang, Y.; Li, N.; Lee, H.; Brazzillo, J.;
Nummy, L. J.; Chung, J. C.; Luvaga, I. K.; Narayanan, B. A.; Wei, X.; Song, J. J.;
Roschangar, F.; Yee, N. K.; Senanayake, C. H. Green Chem. 2017, 19 (6), 1454–1461.
1
0. The processing of 5ꢀmethylpyrazineꢀ2ꢀcarboxylic acid (3) to 2ꢀaminoꢀ5ꢀmethylpyrazine
1) using the sequence shown in Scheme 4 was found not to meaningfully reject
(
impurities arising from pyrazineꢀ2ꢀcarboxylic acid in the starting material. This led to an
acceptance criterion (provisionally 0.20 LC area percent) being placed on the level of
pyrazineꢀ2ꢀcarboxylic acid in 5ꢀmethylpyrazineꢀ2ꢀcarboxylic acid (3).
1
1. Taylor, E. C.; Perlman, K. L.; Sword, I. P.; SequinꢀFrey, M.; Jacobi, P. A. J. Am. Chem.
Soc. 1973, 95, 6407−6412.
1
1
2. AlbanezeꢀWalker, J.; Zhao, M.; Baker, M. D.; Dormer, P. G.; McNamara, J. Tet
.
Lett. 2002, 43, 6747 6750.
−
3. Diaz, F.; Farouz, F. S.; Holcomb, R.; Kesicki, E. A.; Ooi, H. C.; Rudolph,
A.; Stappenbeck, F.; Thorsett, E.; Gaudino, J. J.; Fischer, K. L.; Cook, A. W. Heteroaryl
urea derivatives useful for inhibiting chk1. WO 2006105262, Oct 05, 2006.
1
4. Itoh, T.; Maeda, K.; Wada, T.; Tomimoto, K.; Mase, T. Tetrahedron Lett. 2002, 43
,
9287–9290.
15. Craven, A. Inst. Chem. Eng. Symp. Ser. 1982, 102, 97–111.
1
1
1
6. The DPPA was produced inhouse. Wolff, O.; Waldvogel, S. R. Synthesis 2004,
305.
8, 1303–
1
7. Ende, D. J.; DeVries, K. M.; Clifford, P. J.; Brenek, S. J. Org. Process Res. Dev. 1998,
82–392.
2,
3
8. A protective basis of safety for use with dedicated equipment could have involved
appropriate ventꢀsizing for the vessel feed tank and reactor in order to prevent
overpressurization.
ACS Paragon Plus Environment

Page 17 of 17
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
19. A period of 24 hours for Time to Maximum rate assessments is an industry standard on
the basis that 24 hours should be enough time for a runaway reaction to be detected and
dealt with before it reaches its maximum rate.
20. The quench reaction was determined as being thermoneutral.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21. An alternative process design would have involved slowly adding DPPA to a solution of
acid substrate that was being held above the thermal trigger point for the Curtius
rearrangement. This should have limited the inventory of the potentially hazardous acyl
azide intermediate, which should have rearranged rapidly as soon as it had formed. a)
Yue, T.ꢀY.; McLeod, D. D.; Albertson, K. B.; Beck, S. R.; Deerberg, J.; Fortunak, J. M.;
Nugent, W. A.; Radesca, L. A.; Tang, L.; Xiang, C. D. Org. Process Res. Dev. 2006, 10
62. b) Anjeh, T. E.; Fevig, T. L.; Lawson, J. P.; Walker, D. P.; Wishka, D. G. Exoꢀ(
butyl 2 (+))ꢀ2ꢀaminoꢀ7ꢀazabicyclo[2.2.1]heptaneꢀ7ꢀcarboxylate, intermediates, and
,
2
tꢀ
R
process to prepare and isolate them. WO 2004074292, Sep 02, 2004.
22. Wear, J. O. J. Chem. Educ. 1975, 52, A23.
23. The experimentally determined limiting oxygen concentration below which toluene is
not flammable is 10.4 % v/v. Thus a 1 % v/v oxygen concentration in the purge gas
stream presents no potential for supporting combustion. Osterberg, P.M.; Niemeier, J.K.;
Welch, C.J.; Hawkins, J.M.; Martinelli, J.R.; Johnson, T.E.; Root, T.W.; Stahl, S.S Org.
Process Res. Dev., 2015, 19, 1537–1543.
ACS Paragon Plus Environment