Multikilogram-Scale Preparation of AZD4635 via C-H Borylation and Bromination: The Corrosion of Tantalum by a Bromine/Methanol Mixture

Article abstract of DOI:10.1021/acs.oprd.8b00342

An efficient route to AZD4635 has been developed utilizing the Suzuki-Miyaura reaction of a boronate ester prepared by C-H borylation on a multikilogram scale. Preparation of the cross-coupling partner using bromine/pyridine/methanol has highlighted the incompatibility of this reagent/solvent combination with tantalum, which is commonly used in the construction and repair of standard manufacturing vessels.

Full text of DOI:10.1021/acs.oprd.8b00342

Subscriber access provided by YORK UNIV
Full Paper
Multi-kilo scale preparation of AZD4635 via C-H borylation and
bromination: The corrosion of Tantalum by a bromine/methanol mixture
James J. Douglas, Bradley Adams, Helen Benson, Karl Broberg, Paul M. Gillespie, Oliver Hoult, Ameer K.
Ibraheem, Sophie Janbon, Guillaume Janin, Chris D. Parsons, Ralph C. Sigerson, and David J. Klauber
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00342 • Publication Date (Web): 20 Dec 2018
Downloaded from http://pubs.acs.org on December 20, 2018
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 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 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.
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 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Multi-kilo scale preparation of AZD4635 via C-H
borylation and bromination: The corrosion of
tantalum by a bromine/methanol mixture
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
1
1
1
1
2
James J. Douglas, Bradley W. V. Adams, Helen Benson, Karl Broberg, Paul M. Gillespie,
1
3
1
1
1
Oliver Hoult, Ameer K. Ibraheem, Sophie Janbon, Guillaume Janin, Chris D. Parsons,
1
1
Ralph C. Sigerson, David J. Klauber, *
1Early Chemical Development, Pharmaceutical Sciences, IMED Biotech Unit, AstraZeneca,
Macclesfield, UK
2Process Safety Group, Drug Substance Manufacture, Pharmaceutical Technology &
Development, Global Operations, AstraZeneca, Macclesfield, UK
3Production & Integrity Assurance, Intertek, Manchester Technology Centre, Units 12-15
Longwood Road, Manchester, M17 1PZ
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 25
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
TOC GRAPHIC
Me
Me
Br /MeOH reaction conditions found to
2
corrode tantalum reactor components
N
N
N
Br
N
N
Cl
N
Cl
Bpin
N
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
NH2
NH2
Ir catalysed C-H borylation
conducted on >5 kg scale
AZD4635
API
F
F
ACS Paragon Plus Environment
2

Page 3 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ABSTRACT: An efficient route to AZD4635 has been developed utilizing Suzuki-Miyaura
reaction of boronate ester (5) prepared by C-H borylation on multi-kilo scale. Preparation of
the cross-coupling partner (4) using bromine/pyridine/methanol has highlighted the
incompatibility of this reagent/solvent combination with tantalum, commonly used in the
construction and repair of standard manufacturing vessels.
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
KEYWORDS: materials of construction; tantalum; bromine; methanol; C-H borylation;
Suzuki-Miyaura
INTRODUCTION
1
A R antagonist AZD4635 (6) (Scheme 1) is an investigational new drug currently in Phase IIa
2
A
clinical trials for the treatment of solid tumors. Prior to in-licence of AZD4635 by
AstraZeneca, clinical supplies had been delivered using a developed version of the discovery
2
,3
4
route (Scheme 1). The 3-amino-1,2,4-triazine based API is prepared in a convergent fashion
via Suzuki-Miyaura coupling of aryl bromide (4) with pinacol boronate ester (5). These
intermediates are respectively prepared by bromination of 3-amino-1,2,4-triazine (3) and
iridium-mediated borylation of commercially available 2-chloro-6-methyl-pyridine (7).
3-Amino-1,2,4-triazine (3) is prepared by condensation of 1-(4-fluorophenyl)-2,2-dihydroxy-
ethanone (1) with aminoguanidine bicarbonate (2).5
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 25
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
C H3
N
O
C H3
H2N
H2N
Cl
B
C H3
N
H CO3
2
5
O
C H3
C H3
N
NH2
H3C
H O
OH
O
2
N
N
Br
N
N
N
N
N
N
Cl
N
API
AcOH, MeOH
THF
Br , Pyridine
QuadrasilMP
2
Pd(OAc)
2
NH2
NH2
NH2
2-MeTHF/H O
MeOH
55%
Amphos, K CO
2
2
3
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
F
66%
F
F
THF/H O
2
F
42%
1
3
4
EtOH/MTBE/MeOH
6
59%
Scheme 1. Prior synthetic route to AZD4635 (6)
In advance of scale-up manufacture at AstraZeneca to deliver API for additional clinical supply
and formulation development, assessment of the in-licensed process highlighted four key areas
for improvement: (i) operability of the bromination stage to prepare 4, minimizing operator
exposure to bromine; (ii) the modest 59% yield of 6 from the Suzuki-Miyaura coupling; (iii)
the use of scavenging to remove palladium in the API; and (iv) the lack of polymorph control
in the final crystallisation of the API necessitating re-processing and low overall yield.
RESULTS AND DISCUSSION
3
-Amino-1,2,4-triazine 3 was prepared with minor modifications to the in-licensed process.
The process performed reliably on kilo lab scale (7 kg input 1) with batch yields of 65˗67% to
deliver a total of 26 kg of 3.
Development of bromination to prepare 4. Bromination of 3 was previously accomplished by
6
treatment with excess bromine (2.50 mol eq) in the presence of pyridine (3.00 mol eq) in
methanol (8.00 rel vol) at 25°C. During the previous campaign this process was run on 2 kg
scale (input 3) to deliver 4 in 55% overall yield across 11 batches. The use of bromine in
ACS Paragon Plus Environment
4

Page 5 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
methanol poses safety concerns due to a potential background reaction of bromine with the
7
solvent. Following process safety assessment, it was deemed the current procedure could be
run safely without risk of uncontrollable exotherm up to 100 L scale, provided that the process
concentration of bromine was not increased above 10% v/v.
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 original process charged neat bromine via a pump, an option not available in the
AstraZeneca kilo lab facility without significant operational modification. Furthermore,
following the end of the reaction, solid product 4 was directly filtered from the
bromine/pyridine/methanol slurry, this process could not be operated safely in the kilo lab due
to potential operator exposure to unreacted bromine vapours when using open filtration
equipment. In order to accommodate the bromination reaction, a method to safely charge
bromine and quench the unreacted excess prior to isolation was developed. The exotherm
upon bromine addition was controlled by charging aliquots of bromine via a nitrogen flushed
charge bag to the cooled heterogeneous slurry of 3 in methanol/pyridine. Bromine was
sourced in 1 kg bottles which allowed scaling such that an entire bottle could be opened inside
an inert charge bag, the full contents charged to the reactor, and the bottle safely disposed.
This reduced the number of operations performed with bromine and eliminated the
requirement for any measurement by the process operator. Approximately 1 h after the
bromine charge, HPLC showed an approximately 9095% consumption of 3. A further 21 h
was required to progress to >98% consumption of 3, possibly due to the precipitation of product
4
,8 along with unreacted starting material 3. This observed precipitation, typically within 1 h
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 25
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
of the bromine addition, presented a significant mixing challenge on laboratory scale, but was
not an issue at 100 L kilo lab scale due to more efficient agitation.
Sodium thiosulfate was identified as a suitable reagent to quench any unreacted bromine at
the end of the reaction, as indicated by visual decolourization of the deep red reaction mixture
to a light cream (Figure 1). Due to the low solubility of sodium thiosulfate in methanol, it was
charged as a solution in water to the thick reaction slurry causing a significant exotherm.
Control of the exothermic quench was important, as at temperatures >25C the formation of
an unidentified by-product was observed. 0.34 mol eq (relative to 3) of sodium thiosulfate was
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
required to quench the excess bromine, consistent with the expected molar balance. This
suggests that bromine is not significantly consumed by side reactions, for example with the
methanol solvent. The final conditions utilised 0.40 mol eq of sodium thiosulfate in water
(1.00 rel vol), charged dropwise over 2 h, maintaining the internal temperature below 20C.
ACS Paragon Plus Environment
6

Page 7 of 25
Organic Process Research & Development
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
Figure 1. Photograph of lab-scale reaction mixture pre- (left) and during- (right) quench of
bromine.
Following the quench, the solubility of 4 in the reaction mixture, i.e. the loss to the liquors
upon filtration and isolation, was higher than expected based on the measured solubility of
isolated 4. Analysis of the pH during the sodium thiosulfate quench clearly showed a change
in pH from 4 to 2, concurrent with the overall balanced equation for reduction of bromine by
9
thiosulfate. It was postulated that a protic salt of 4 was forming during the quench. Indeed,
independent preparation and measurement of the solubility of the HBr salt of 4 revealed it had
a significantly increased solubility in methanol/water mixtures relative to the freebase form,
1
0
explaining the increased loss to liquors upon filtration and isolation. The original process
incorporated a direct filtration of the unquenched reaction mixture, followed by three
methanol washes at 20C to provide isolated material that was effective in the subsequent
stage, but typically contained both pyridine.HBr, approximately 67% w/w, and bromine,
approximately 23% (table 1 entry 1). Following the introduction of the quench procedure,
1
11
the assay of 4 was significantly reduced (80% w/w by H NMR). This batch stalled in the
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 25
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
subsequent Suzuki-Miyaura reaction but could be restarted with a fresh charge of the
palladium catalyst (table 1 entry 2). Aryl bromide 4 prepared using the original procedure was
spiked with possible inorganic contaminants from the quench process which confirmed 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
need to remove these impurities during the isolation of 4 (Table 1 entry 3/4).12
entry
Inorganic spiked (charge based on 4) Outcome of Suzuki-Miyaura
reaction
1
2
None (obtained via original process)
Complete conversion
Stalled conversion
None (obtained via quenched
process)
3
4
Sodium thiosulfate (0.20 mol eq)
Stalled conversion
Stalled conversion
Potassium tetrathionite (0.10 mol eq)
Table 1. Spiking of inorganic contaminants into Suzuki-Miyaura coupling
To isolate the desired freebase form in higher yield, the pH following the thiosulfate quench
was re-adjusted to 44.5 using an aqueous K CO solution and held at ˗10C, for at least 4 h to
2
3
further maximise isolated yield. In order to obtain 4 in sufficient purity to allow successful
ACS Paragon Plus Environment
8

Page 9 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Suzuki-Miyaura coupling, a filter cake displacement washing sequence was developed that
removed all significant impurities, residual pyridine salts, inorganics, and water (Scheme 2).13
This fully developed process performed well on 120 g lab scale, delivering 4 in acceptable yield
and purity (56% yield, 97% w/w).
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
i) Br , Pyridine,MeOH
i) K CO (aq)
2 3
2
N
N
Br
N
N
Br
N
N
5
C to 25 C,18 h
<20 C to 10 C,18 h
N
N
N
ii) Filter w ash
ii) Na S O (aq)
NH2
NH2
NH2
2
2
3
MeOH/H O
2
<20 C,18 h
H O  2
F
F
2
F
3
4HBr/H SO
MeOH  2
4 56%
2
4
Scheme 2. Developed synthetic sequence for the bromination of 3.
Manufacture of 4. The manufacture of the first of two planned batches on 100 L scale was
successfully carried out without any apparent issue using the developed process, providing 7
kg of 4 in 94% w/w purity at 56% yield. During manufacture of the second batch, a leak was
observed from the reaction vessel which was subsequently traced to the thermopocket housing
for the temperature control probe. Following removal of the thermopocket, extensive
investigation was carried out into the origin of the leak. The thermopocket consists of a hollow
bolt constructed from tantalum which seals against the glass lining of the vessel with a PTFE
gasket. Analysis of the corrosion pattern of the bolt concluded that general corrosion attack
of the bolt head had removed the sealing weld, allowing through leakage (Figure 2 (a)).
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 25
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
(
a)
(b)
Figure 2. (a) Photograph comparing a used but not corroded thermopocket (left), to the
leaking thermopocket showing signs of corrosion (right). (b) Photograph showing tantalum
coupon after immersion in the reaction mixture (right). Compared with the control (left)
there is visible corrosion below the weld line where the coupon has contacted the reaction
mixture.
Concurrently, a laboratory user trial of 4 manufactured in the first kilo lab batch in the
subsequent synthetic step showed the presence of hitherto unobserved insoluble material in
the reaction mixture after successful Suzuki-Miyaura coupling to prepare 6. Removal of this
material by filtration and subsequent analysis by ICP-MS revealed that it was comprised of
ACS Paragon Plus Environment
1
0

Page 11 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
tantalum. The level of tantalum in the input batch of 4 was quantified as 558 ppm by ICP-
OES analysis confirming the suspicion that the batch had been contaminated.14
In order to discern whether the bromination process to prepare 4 was responsible for the
observed corrosion, tantalum coupon testing of the reaction mixture was carried out on
laboratory scale by suspending a coupon in the reaction vessel, so that it contacted the reaction
mixture below the weld line. The coupon was visibly corroded where it had contacted the
reaction mixture and to a lesser extent where it had been exposed to vapours generated by the
process, providing further evidence of the incompatibility of tantalum with the reaction
mixture (Figure 2 (b)).
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
Standard materials of construction compatibility checks had not flagged any issue between
tantalum and individual components, or by-products of the reaction mixture (bromine,
pyridine, methanol, hydrobromic acid), suggesting that a combination of these components
may be responsible. Detailed literature searching found reports that tantalum is not resistant
to methanol solutions of 5% bromine containing less than 5% water, due to dissolution of the
protective oxide coating. Based on the process charges used in the manufacture of 4 (bromine
2.50 mol eq, methanol 8.00 L/kg), the bromine content in methanol is 25% w/w
(approximately 8% v/v) excluding the substrate 3 and pyridine. This suggests that the
combination of bromine and methanol at the process concentration is responsible for the
observed corrosion of tantalum, especially given the lack of added water to the process,
although water is reported to only temper the corrosive effect to a limited extent.15
ACS Paragon Plus Environment
1
1

Organic Process Research & Development
Page 12 of 25
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
Although the tantalum compatibility issue prevented further manufacture of intermediate 4,
the stock already prepared was successfully processed through to API with implementation of
an in-line filtration to remove insoluble tantalum material prior to isolation. This was
successful in reducing tantalum to an acceptable limit of less than 1 ppm in the API (see Table
2).16 Beyond component construction, tantalum is commonly used to repair damaged glass-
lined steel vessels widely prevalent in manufacturing units, therefore work is on-going to
develop alternative conditions for the bromination of 3 so as not to limit future manufacturing
options.17
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
Scale-up of C-H borylation to prepare 5. Key Suzuki-Miyaura coupling partner 5 was prepared
1
8
using a slightly modified version of the process used to deliver previous campaigns, which
has been adapted from the literature procedure (Scheme 3).19,20,21
[(1,5-COD)(OMe)Ir]2
2
DTBBP, B Pin2
C H3
C H3
n-heptane
80 C, 16 h
N
N
O
C H3
76%
Cl
B
Cl
O
C H3
C H3
7
5
H3C
Scheme 3. Preparation of boronate ester 5 by C-H borylation
Commercially available 2-chloro-6-methyl-pyridine (7) was charged to a pre-heated solution
of (1,5-cyclooctadiene)(methoxy)iridium(I) dimer (0.25 mol%), DTBBP (0.50 mol%) and
(bispinacolato)diboron (0.70 mol eq) in n-heptane (10.00 L/kg) at 50°C. The batch was then
heated further to 80°C and held for approximately 16 h to effect the conversion to 5. The
ACS Paragon Plus Environment
1
2

Page 13 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
product was isolated by concentration of the batch and cooling to initiate crystallisation. This
process was successfully run on 6 kg input scale with isolation of the product in 76% overall
yield across two batches. An essential consideration for accommodation of this process was
checking the venting capacity of the reactor to ensure safe removal of hydrogen. This
stoichiometric by-product is generated either directly, or from hydrolysis of pinacolborane
upon exposure to atmospheric moisture during isolation of 5 on an open filter.22
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
Assembly of AZD4635. During the previous manufacturing campaign, incomplete conversion
was observed during the final assembly of the API via Suzuki-Miyaura coupling, resulting in
a moderate 59% yield. The reason for this was never discerned and user trials of lab batches
1
of 4 prepared using the previous procedure (assay by H NMR >95% w/w) with the standard
coupling conditions showed no issue with the quality of 4. Having established confidence in
the input quality of 4 and the cross-coupling conditions, attention turned to the isolation and
ultimate quality of Suzuki-Miyaura product 6.
The previous delivery for clinical studies had relied on scavenging to remove palladium
residues to an appropriate level in the final isolated API. Investigation was undertaken to
assess the extent of metal removal in the isolation of AZD4635 following the Suzuki-Miyaura
reaction, achieved via solvent swap from THF to ethanol and subsequent crystallisation. This
demonstrated that the majority of both palladium and iridium (and boron) was removed in the
crystallisation liquors and subsequent displacement washes.13
ACS Paragon Plus Environment
1
3

Organic Process Research & Development
Page 14 of 25
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
The developed process transferred well to manufacture in the kilo-scale laboratory at 4 kg
input scale of 4. The conversion target was achieved and AZD4635 was isolated in 85% overall
yield across two batches. Residual iridium levels were within the range seen in laboratory
development batches, whilst palladium levels were slightly higher than expected, although
they could be adequately controlled in the subsequent crystallisation.
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
Crystallisation of AZD4635.
Despite the observation of some metal removal during the initial isolation of AZD4635 after
the Suzuki-Miyaura coupling, it was evident that an additional crystallisation step would be
required to safely control both palladium and iridium to within the specification limit of no
more than 100 ppm.23 Of equal consideration, following the Suzuki-Miyaura coupling,
AZD4635 was not isolated as a pure crystalline form in laboratory development batches.
Therefore control of the desired Form 1 must be provided by the final crystallisation.
The general approach to crystallisation development adopted by AstraZeneca comprises three
stages: (i) solubility screen and solvent selection; (ii) process outline; and (iii) process
exploration.24 Solubility screening can be carried out using a variety of small scale (i.e. <5 mL)
methods and equipment, whereas stages (ii) and (iii) are typically performed in jacketed vessels
with overhead agitation in order to provide confidence that the process will reliably transfer
to manufacturing equipment (i.e. >20 L) with control of API quality attributes (e.g. form and
purity).
ACS Paragon Plus Environment
1
4

Page 15 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The solubility screen for AZD4635 Form 1 was performed using primarily ICH Class 3 and
some Class 2 solvents by equilibrium methods.13,25 Solubility vs. temperature data was
generated for the subset of solvents with a solubility within the range 5-25 mg/mL at 25°C
using turbidity measurement.26 These single solvent systems showed a shallow response,
indicating that an anti-solvent crystallisation would be the most appropriate method to obtain
sufficient recovery. AZD4635 Form 1 had a much greater solubility in DMSO (>100 mg/mL
at 25°C) than the other studied solvents and was therefore selected to be used in the new
process in combination with an anti-solvent exhibiting the least solubility: 2-propanol
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
(1 mg/mL at 25°C).
The process outline was determined experimentally as follows: AZD4635 was dissolved in
DMSO/2-propanol (2:1, v/v) at elevated temperature to enable transfer into the crystallizer
through a screening/polishing filter, before cooling to induce supersaturation and seeding with
Form 1 to initiate crystallisation. The batch was then held for an extended period of time until
the saturation state was reached. This was identified and monitored in the laboratory using
®
PAT, specifically FBRM ; this technology indicated when the total particle counts had reached
a steady maximum, which should be related to the saturated (i.e. equilibrium) state. At this
point additional 2-propanol was charged in a controlled manner to promote further
crystallisation, subsequently followed by a ramped cool to the isolation temperature. This
process reliably delivered Form 1 in the lab in approximately 80% yield. Further process
ACS Paragon Plus Environment
1
5

Organic Process Research & Development
Page 16 of 25
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
exploration work then focused on understanding robustness of the process parameters (e.g.
seeding temperature) and consistency in the output quality, particularly metal rejection.
The crystallisation process developed transferred as expected to the kilo scale laboratory,
where two input batches were combined to deliver 6.5 kg of AZD4635 Form 1 in 83% yield.
Crucially, the crystallisation was also successful at reducing the levels of residual palladium
and iridium to within specification without the requirement for scavenging (Table 2). Control
of tantalum via in-line filtration during work-up of the Suzuki-Miyaura coupling was also
confirmed by analysis of the output from that stage prior to the final crystallisation.
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
entry
Batch
Palladium
Iridium (ppm) Tantalum (ppm)
(ppm)
1
2
3
Input C660/1
121
66
69
29
<1
<1
(3.5 kg)
Input C660/2
4.4 kg)
129
(
Output C660/1
16
not analysed
Table 2. Removal of metals via crystallisation of AZD4635
ACS Paragon Plus Environment
1
6

Page 17 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Campaign summary for AZD4635.
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
Overall, the campaign delivered 6.5 kg of AZD4635 prior to the suspension of further planned
batches, due to observed corrosion of a tantalum component of a 100 L vessel during the
bromination stage (Scheme 4). Stage 1 delivered a total of 26 kg of 3 in 6567% yield, using a
slightly modified procedure to that developed prior to in-licence. The subsequent bromination
produced 15 kg of 4 in 5663% yield, following significant re-design of the process to minimize
potential operator exposure to bromine. Suzuki-Miyaura coupling of aryl bromide 4 and
pinacol boronate ester 5, followed by a modified isolation, gave 8 kg of AZD4635 in an
improved 84-85% yield. Final stage crystallisation delivered the required form in 83% yield,
while successfully reducing the levels of residual metals to within specification without the
requirement for scavenging.
C H3
N
H2N
H2N
O
C H3
Cl
B
C H3
N

H CO
5
2
3
O
C H3
C H3
i) Br , Pyridine, MeOH
N
NH2
2
H3C
Recrystallisation
DMSO/iPrOH
HO
OH
O
2
N
N
ii) Na
2
S
2
O
3
, K
2
CO
3
Br
N
N
N
N
N
N
Cl
N
AZD4635
API
AcOH, MeOH
THF
i) Pd(OAc) , Amphos,
2
NH2
5663%
NH2
NH2
83%
K CO , THF/H O
2
3
2
6567%
ii) EtOH crystallisation
F
F
F
F
1
3
4
8485%
6
Scheme 4. Campaign summary.
CONCLUSION
ACS Paragon Plus Environment
1
7

Organic Process Research & Development
Page 18 of 25
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
AZD4635 was successfully prepared in 28% overall yield from 1. The bromination of 3 to
deliver cross-coupling partner 4 was successfully developed to introduce a quench of excess
bromine whilst ensuring successful conversion in the key Suzuki-Miyaura coupling to prepare
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
6
. Pinacol boronate ester 5 was synthesized via scale-up of the iridium-mediated borylation
of 7. Both palladium and iridium metals were controlled through well-designed
crystallisations of the API without the need for scavenging. Kilo-scale manufacture has
highlighted that the combination of bromine and methanol is incompatible with reactor
components constructed from tantalum at typical process concentrations. It is highly
recommended that all possible combinations of reaction components are checked for
compatibility with materials of construction and coupon tests are carried out where there is a
lack of any existing data.
EXPERIMENTAL METHODS
General. All reactions were carried out in dry vessels under an atmosphere of dry nitrogen
unless otherwise specified. All reagents and solvents were used as received without further
purification unless specified.
Synthesis of 5-(4-fluorophenyl)-1,2,4-triazin-3-amine (3).
1-(4-Fluorophenyl)-2,2-
dihydroxy-ethanone (1) (6.85 kg, 40.3 mol, 1.00 equiv), aminoguanidine bicarbonate (2) (6.03
kg, 44.3 mol, 1.10 equiv) and acetic acid (86.2 kg) were charged to a 100 L glass-lined reactor
which had been purged with nitrogen. After addition, the mixture was heated to 73-77°C and
then held for ≥20 h. The reaction mixture was cooled to 23-27°C over 3.5-4.5 h and then held
for ≥1 h. The mixture was filtered and washed twice with methanol (10.8 kg). The cake was
dried at 40°C under vacuum to constant weight to give 5-(4-fluorophenyl)-1,2,4-triazin-3-
1
amine (3) as a grey powder (4.94 kg, 26.0 mol, 64.5%); H NMR (400 MHz, DMSO-d6) δ 7.26
13
(
br s, 2H), 7.37–7.43 (m, 2H), 8.23–8.26 (m, 2H) and 9.24 (s, 1H); C NMR (101 MHz, DMSO-
d6) δ 164.38 (J = 249.9), 162.97, 153.84, 136.94, 130.50 (J = 2.9), 129.83 (J = 9.0), 116.15 (J =
21.9). Assay (QNMR) 95.0% w/w.
ACS Paragon Plus Environment
1
8

Page 19 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Synthesis of 6-bromo-5-(4-fluorophenyl)-1,2,4-triazin-3-amine (4). 5-(4-Fluorophenyl)-
1
,2,4-triazin-3-amine (3) (4.76 kg, 25.0 mol, 1.00 equiv), methanol (30.2 kg) and pyridine (5.94
kg, 75.1 mol, 3.00 equiv) were charged to a 100 L glass-lined reactor which had been purged
with nitrogen. After addition the mixture was cooled to 3-7°C. Bromine (10.0 kg, 62.6 mol,
2
.50 equiv) was then charged in 10 aliquots. The mixture was then warmed to to 23-27°C and
held for ≥18 h. The reaction was considered complete when conversion of 5-(4-fluorophenyl)-
,2,4-triazin-3-amine (3) was ≥99%. The reaction mixture was cooled to 13-17°C. A solution
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
1
of sodium thiosulfate (1.58 kg, 10.0 mol, 0.40 equiv) in water (15.2 kg) was added over ≥2 h
keeping the batch temperature ≤20°C. A solution of potassium carbonate (4.22 kg, 30.5 mol,
1
.22 equiv) in water (4.76 kg) was added over ≥2 h. The reaction mixture was cooled to -12 to
-
8°C over 0.5 - 1.5 h and then held for ≥4 h. The mixture was filtered and washed with pre-
mixed methanol (7.84 kg) and water (9.91 kg) cooled to -12 to -8°C. The filter cake was then
washed twice with water (19.0 kg) and twice with methanol (15.1 kg). The cake was dried at
4
0°C under vacuum to constant weight to give 6-bromo-5-(4-fluorophenyl)-1,2,4-triazin-3-
1
amine (4) as an off-white powder (4.22 kg, 15.7 mol, 62.7%); H NMR (500 MHz, DMSO-d )
δ 7.32–7.41 (m, 2H), 7.55 (s, 2H), 7.81–7.85 (m, 2H); C NMR (126 MHz, DMSO-d6) δ 115.29
6
13
(
(
J = 21.9), 131.55 (J = 3.1), 131.81 (J = 9.0), 134.56, 157.39, 162.29, 163.37 (J = 248.8). Assay
QNMR) 95.6% w/w.
Synthesis of 2-chloro-6-methyl-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine (5).
1,5-Cyclooctadiene)(methoxy)iridium(I) dimer (74.9 g, 113 mmol, 0.0025 equiv), 4,4’-di-tert-
(
butyl-2,2’-bipyridine (60.7 g, 226 mmol, 0.0050 equiv), bis(pinacolato)diboron (8.03 kg, 31.6
mol, 0.70 equiv) and n-heptane (35.5 kg) were charged to a 100 L glass-lined reactor which
had been purged with nitrogen. The reactor was purged three times with nitrogen, heated to
4
5-55°C with stirring and then held for 0.5-3 h. 2-Chloro-6-methyl-pyridine (7) (5.76 kg, 45.1
mol, 1.00 equiv) was then charged to the reactor, followed by a line wash with n-heptane (3.94
kg). The reactor was purged with nitrogen, heated to 75-85°C over 1-5 h then held for ≥16 h.
The reaction mixture was cooled to 35-45°C over 1-5 h. The reaction was considered complete
when conversion of 2-chloro-6-methyl-pyridine (7) was >95%. The batch was distilled under
vacuum at 35-45°C until the concentration of 2-chloro-6-methyl-4-(4,4,5,5-tetramethyl-1,3,2-
dioxaborolan-2-yl)pyridine (5) was >225 mg/mL. The reaction mixture was cooled to 25-35°C
over 1-3 h and then seeded if crystallisation had not already started. After holding for ≥1 h
the reaction mixture was cooled to -15 to -10°C over 2-6 h. After holding for ≥1 h the mixture
was filtered and washed twice with n-heptane (7.89 kg) cooled to -15 to -10°C. The cake was
dried at 40°C under vacuum to constant weight to give 2-chloro-6-methyl-4-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine (5) as an off-white powder (8.20 kg, 32.3 mol,
1
7
1.7%); H NMR (500 MHz, DMSO-d ) δ 1.29 (s, 12H), 2.45 (s, 3H), 7.33 (s, 1H), 7.42 (s, 1H);
6
1
3
C NMR (126 MHz, CDCl ) δ 24.6, 84.7, 125.3, 126.8, 141.0, 149.5, 159.1; Assay (QNMR)
3
95.7% w/w.
ACS Paragon Plus Environment
1
9

Organic Process Research & Development
Page 20 of 25
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
Synthesis of 6-(2-chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-triazin-3-amine (6).
6
-Bromo-5-(4-fluorophenyl)-1,2,4-triazin-3-amine (4) (4.37 kg, 16.2 mol, 1.00 equiv), 2-
chloro-6-methyl-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridine (5) (4.11 kg, 16.2
mol, 1.00 equiv), potassium carbonate (3.36 kg, 24.3 mol, 1.50 equiv), 4-(di-tert-
butylphosphino)-N,N-dimethylaniline (215 g, 810 mmol, 0.05 equiv), tetrahydrofuran (34.8
kg), water (17.5 kg) and palladium(II) acetate (72.8 g, 324 mmol, 0.02 equiv) were charged to
a 100 L glass-lined reactor which had been purged with nitrogen. The reactor was purged
three times with nitrogen, heating to 62-66°C with stirring and then held for 3-18 h. The
reaction was considered complete when conversion of 6-bromo-5-(4-fluorophenyl)-1,2,4-
triazin-3-amine (4) was >97% and conversion of 2-chloro-6-methyl-4-(4,4,5,5-tetramethyl-
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
1
,3,2-dioxaborolan-2-yl)pyridine (5) was >97%. The reaction mixture was cooled to 45-55°C
and then filtered to remove insoluble tantalum residues. The aqueous phase was separated and
then the organic phase was distilled under vacuum at 45-55°C to remove 21.0 L of solvent.
Ethanol (34.9 L) was charged at such a rate that the batch temperature remained within the
range 45-55°C. The batch was then distilled under vacuum at 45-55°C to remove 26.0 L of
solvent. The reaction mixture was cooled to -2 to 2°C over 2-4 h and then held for ≥1 h. The
mixture was filtered and washed with methyl tert-butyl ether (6.48 kg), twice with water (8.73
kg) and twice with ethanol (6.90 kg). The cake was dried at 40°C under vacuum to constant
weight to give 6-(2-chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-triazin-3-amine
1
as a white powder (4.32 kg, 13.7 mol, 84.5%; H NMR (400 MHz, DMSO-d ) δ 2.41 (s, 3H),
7
6
1
3
.16 (s, 1H), 7.24-7.31 (m, 3H), 7.4-7.51 (m, 2H), 7.71 (br s, 2H); C NMR (101 MHz, DMSO-
d6) δ 163.25 (J = 248.6), 161.87, 159.07, 155.89, 149.10, 147.86, 145.34, 129.96 - 132.28, 121.92,
20.60, 115.48 (J = 22.0), 23.56.Assay (QNMR) 98.3% w/w.
1
Recrystallisation of 6-(2-chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-triazin-3-
amine (6). 6-(2-Chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-triazin-3-amine (6)
(7.87 kg, 24.9 mol, 1.00 equiv), dimethyl sulfoxide (21.8 kg) and 2-propanol (7.80 kg) were
charged to a 100 L glass-lined reactor which had been purged with nitrogen (dissolving vessel).
The mixture was heated to 73-77°C and then held for ≥1 h to ensure dissolution. The solution
was cooled to 70°C and then screened through a 5 µm filter into a pre-heated 100 L glass-lined
reactor (crystallising vessel) which had been purged with nitrogen and then the batch
temperature was adjusted as appropriate to 68-72°C. Dimethyl sulfoxide (1.73 kg) and 2-
propanol (619 g) were charged to the dissolving vessel, heated to 70°C and then screened
through the 5 µm filter into the crystallising vessel as a line wash. The solution was cooled to
4
9-53°C over ≥30 min and then 6-(2-chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-
triazin-3-amine (6) seed was charged (31.5 g, 0.04 g/g). The mixture was held for ≥5 h to
establish crystallisation and then 2-propanol (41.3 kg) was charged over ≥5 h. The mixture
was cooled to 8-12°C over ≥2 h and then held for ≥5 h. The mixture was filtered and washed
with pre-mixed dimethyl sulfoxide (4.33 kg) and 2-propanol (9.28 kg) that had been screened
through a 5 µm filter into the crystallising vessel and then cooled to 8-12°C. The cake was
washed with 2-propanol (12.4 kg) that had been screened through a 5 µm filter into the
ACS Paragon Plus Environment
2
0

Page 21 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
crystallising vessel and then cooled to 3-7°C. The cake was dried at 40°C under vacuum to
constant weight to give 6-(2-chloro-6-methylpyridin-4-yl)-5-(4-fluorophenyl)-1,2,4-triazin-
3-amine as a white powder (6.51 kg, 20.6 mol, 82.7%; QNMR (100% w/w).
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
AUTHOR INFORMATION
Corresponding Author
*Address: F40/14 Etherow, AstraZeneca, Silk Road Business Park, Macclesfield, Cheshire,
SK10 2NA, United Kingdom. Telephone: +44(0)1625 231861, Email:
david.klauber@astrazeneca.com
Author Contributions
All authors have given approval to the final version of the manuscript.
ACKNOWLEDGMENTS
Heptares Therapeutics and Cambridge Major Laboratories are thanked for initial process
development.
ABBREVIATIONS
A R, adenosine A receptor; API, active pharmaceutical ingredient; DTBPP, 4,4’-di-tert-
2
A
2A
butyl-2,2’-bipyridine; Amphos, 4-(di-tert-butylphosphino)-N,N-dimethylaniline; PAT,
process analytical technology; PTFE, polytetrafluoroethylene; ICP-MS, inductively coupled
plasma mass spectrometry; ICP-OES, inductively coupled plasma optical emission
®
®
spectrometry; FBRM , Focused Beam Reflectance Measurement
ACS Paragon Plus Environment
2
1

Organic Process Research & Development
Page 22 of 25
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
ASSOCIATED CONTENT
Supporting Information
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
pH profile during quench of bromination; solubility data for 4 and 4.HBr; proposed mass
balance for bromination and quench; purity data for 4 during cake washing; Pd, Ir and B
1
13
content during initial isolation of 6; solubility data for AZD4635 Form 1; H and C NMR
spectra for 3, 4, 5, and 6.
REFERENCES
1
2
. Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz K-N.; Linden, J. International
Union of Pharmacology. XXV. Nomenclature and classification of adenosine receptors.
Pharmacol. Rev., 2001, 53 (4), 527–552
. Congreve, M.; Andrews, S. P.; Doré, A. S.; Hollenstein, K.; Hurrell, E.; Langmead, C. J.;
Mason J. S.; Ng, I. W.; Tehan, B.; Zhukov, A.; Weir, M.; Marshall, F. H.; Discovery of
1
,2,4-Triazine Derivatives as Adenosine A2A Antagonists using Structure Based Drug
Design. J. Med. Chem., 2012, 55 (5), 1898–1903
3
4
. Congreve, M. S.; Philippe, S.; Jonathan, A.; Mason, S.; Richardson, C. M.; Brown, G. A.;
1,2,4-triazine-4-amine derivatives. WO2011/95625 A1, 2011
. Cascioferro, S.; Parrino, B.; Spanò, V.; Carbone, A.; Montalbano, A,; Barraja, P.; Diana,
P.; Cirrincione, G.; An overview on the recent developments of 1,2,4-triazine
derivatives as anticancer compounds. Eur. J. Med. Chem., 2017, 142, 328–375
5. For a review on the chemistry of 1,2,4-triazines see: Raw, S. A.; and Taylor, R. J. K.;
Advances in Heterocyclic Chemistry, vol 100. Elsevier Press, 2010. p. 75–100
6
. We believe an excess of bromine is required due to the competitive formation of
pyridinium hydrobromide perbromide (pyridine.HBr ). Pyridine.HBr was found to be
3
3
ineffective in the bromination of 3. See Moon, M. P.; Crouch, D. R.; Pyridinium
Hydrobromide Perbromide. Encyclopedia of Reagents for Organic Synthesis, 2008
ACS Paragon Plus Environment
2
2

Page 23 of 25
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
7
. Bowman, P. T.; Ko, E. I.; Sides, P. J. A Potential Hazard in Preparing Bromine‐Methanol
Solutions. J. Electrochem. Soc., 1990, 137 (4), 13091311
8. Analysis by HPLC of the solid and filtrate samples independently, showed that the
remaining starting material 3 is predominantly out of solution at this stage of the
reaction
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
9
. The overall equation for reduction of bromine by thiosulfate is
Na S O + 4 Br + 5H O → 2H SO + 6HBr +2NaBr
2 2
3
2
2
2
4
10. We did not fully characterise the salt counterion of 4 in solution and cannot rule out
formation of the hydrogen sulfate salt of 4. See the supporting information for further
details
1
1. This material contained approximately 510% of the salt form of 4 with the remaining
mass balance unaccounted for
1
2. We did not attempt to characterize or quantify the residual inorganic content in the
2
-
low assay batch. Tetrathionate (S O ) has been reported to degrade to S under acidic
4
6
8
conditions which may have further effected the subsequent Suzuki-Miyaura reaction.
For example see Xiang, Y.; Caron, P-Y.; Lillie, B. M.; Vaidyanathan, R. Sulfur
Contamination Due to Quenching of Halogenation Reactions with Sodium Thiosulfate:
Resolution of Process Problems via Improved Quench Protocols. Org. Process Res.
Dev., 2008, 12 (1), 116–119
1
1
1
3. See the supporting information for further details
4. We did not characterize or confirm the structure of the contaminating tantalum species
5. Dechema-Werkstoff-Tabelle “Methanol” (Corrosion data sheets “Methanol”) (in
German); Dechema, D-6000: Frankfurt am Main, 1961. In addition to the effect of
water, further investigation would be required to unambiguously rule out the role of
substrate 3, pyridine, and/or the reaction solution post quench in effecting the observed
corrosion
16. An acceptable limit was assigned based on a conservative estimate of the Permitted
Daily Exposure for tantalum and the maximum clinical dose of API
1
7. During materials of construction assessment, no issues were flagged for glass-lined
vessels, however Hastelloy was flagged as incompatible due to HBr generation.
Subsequent coupon testing has confirmed incompatibility of the reaction mixture with
Hastelloy C-22 and C-276; PTFE showed no issues.
ACS Paragon Plus Environment
2
3

Organic Process Research & Development
Page 24 of 25
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
1
1
8. n-Heptane replaced hexane or THF in the literature procedures to facilitate isolation of
5
. The two-stage heating process, with addition of the substrate 7 at 45-55°C, had been
employed in previous manufacture of 5 prior to in-licence and was found to be robust
in our hands. The process was modified to provide greater control in the crystallisation
of 5.
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. Takagi, J.; Kazuaki, S.; Hartwig, F. J.; Ishiyama, T.; Miyaura, N. Iridium-catalyzed C–H
coupling reaction of heteroaromatic compounds with bis(pinacolato)diboron:
regioselective synthesis of heteroarylboronates. Tetrahedron Lett., 2002, 43 (32), 5649–
5
651
20. Ishiyama, T.; Takagi, J.; Nobuta, Y.; Miyaura., N. Iridium-catalyzed C–H borylation of
arenes and heteroarenes: 1-chloro-3-iodo-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2
-yl)benzene and 2-(4,4,5,5,-tetramethyl-1,3,2-dioxaborolan-2-yl)indole. Org. Synth.,
2005, 82, 126–132
21. Murphy, J. M.; Liao, X.; Hartwig., J. F. Meta Halogenation of 1,3-Disubstituted Arenes
via Iridium-Catalyzed Arene Borylation. J. Am. Chem. Soc., 2007, 129, 15434–15435
2
2. The risk of potential static build-up in a non-conducting solvent such as n-heptane was
mitigated by: (i) ensuring oxygen content was below the flammable range of solvent
vapour/oxygen composition (<5% oxygen); (ii) glass-lined vessels were earthed using
earthing plugs; (iii) avoiding vigorous agitation of non-conducting solvents; (iv)
reducing addition rate when charging non-conducting solvents.
23. ICH Harmonised Guideline Q3D – Guideline for Elemental Impurities, Current Step 4
version, 16 December 2014
2
4. Black, S. N.; Quigley, K.; Parker, A. A Well-Behaved Crystallisation of a
Pharmaceutical Compound. Org. Process Res. Dev., 2006, 10 (2), 241–244
25. ICH Harmonised Guideline Q3C(R6) – Impurities: Guideline for Residual Solvents,
Current Step 4 version, October 20, 2016
2
6. Muller, F. L.; Fielding, M.; Black, S. A Practical Approach for Using Solubility to Design
Cooling Crystallisations. Org. Process Res. Dev., 2009, 13 (6), 1315–1321
ACS Paragon Plus Environment
2
4

Page 25 of 25
Organic Process Research & Development
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
2
5