Development of a practical synthesis of Toll-like receptor agonist PF-4171455: 4-amino-1-benzyl-6-trifluoromethyl-1,3-dihydroimidazo [4,5-c] pyridin-2-one

Article abstract of DOI:10.1021/op200021a

The development and implementation of a scalable process for the manufacture of the Toll-like receptor (TLR7) agonist PF-4171455 (1) is described. Initial routes used to synthesise 1 in milligram quantities were unsuitable for large-scale synthesis to pro

Full text of DOI:10.1021/op200021a

ARTICLE
pubs.acs.org/OPRD
Development of a Practical Synthesis of Toll-like Receptor Agonist
PF-4171455: 4-Amino-1-benzyl-6-trifluoromethyl-1,3-dihydroimidazo
[4,5-c] pyridin-2-one
Fiona M. Adam,† Gerwyn Bish,*,† Frederick Calo,† Christopher L. Carr,‡ Nieves Castro,‡ Duncan Hay,†
Paul B. Hodgson,*,‡ Peter Jones,† Craig J. Knight,‡ Michael Paradowski,† Gemma C. Parsons,†
Katie J. W. Proctor,† David C. Pryde,† Filippo Rota,‡ Mya C. Smith,† Nicholas Smith,† Thien-Duc Tran,†
James Hitchin,§ and Rachel Dixon§
†Worldwide Medicinal Chemistry and ‡Research API, Pfizer Global Research and Development, Sandwich Laboratories, Ramsgate Road,
Sandwich, Kent CT13 9NJ, United Kingdom, and §SRG, London Office, 16 St. Helen's Place, London, EC3A 6DF, United Kingdom
ABSTRACT: The development and implementation of a scalable process for the manufacture of the Toll-like receptor (TLR7)
agonist PF-4171455 (1) is described. Initial routes used to synthesise 1 in milligram quantities were unsuitable for large-scale
synthesis to provide bulk material. As part of the transfer between Medicinal Chemistry and Research-API, collaboration provided a
fit for purpose route for the kilo-scale synthesis of 1. Key aspects of the synthesis included (i) a safe and practical synthesis of a key
nitropyridone intermediate 7 over four steps, (ii) a sequential regioselective chlorination to selectively functionalise 7 and (iii) use of
a carbamate as a tethered carbonyl group, allowing an efficient regiospecific synthesis of 1.
’ INTRODUCTION
yield. Copper mediated displacement of the chloride with
benzylamine2 and final deprotection of the benzyl fragment with
concentrated sulphuric acid3 afforded PF-4171455, 1.
In examining the route to consider both safety and practicality,
several major concerns were apparent. The acylation to prepare
enamide 4 led to a crude mixture of desired enamide and an
overacylated byproduct 144 resulting from unwanted C-acylation
followed by N-acylation (Figure 1).
The other major impurity was unreacted trifluorocrotonate
due to the overacylation pathway as only 1.2 equivalents of malonyl
chloride was used. The enamide 4 was typically purified by a very
challenging column chromatography which was impractical on large
scale. The use of extremely flammable sodium hydride as the base
for the cyclisation would be undesirable. The decarboxylation in
hydrochloric acid was high yielding, but mixing was a potential issue
on scale as the pyridone ester 5 formed a separate layer above the
reaction mixture upon addition. The use of nitric acid in acetic acid
and ethyl acetate presented a significant risk due to the potential for
formation of acetyl nitrate, solutions of which may decompose
violently above 60 °C;5 hence, alternative reagents for this trans-
formation would be required. The middle part of the sequence, i.e.
selective 4-chlorination, benzylamine displacement and 2-chlorina-
tion had good yields, and the reagents were deemed appropriate for
further scale-up although some enabling work would be required for
the workup and isolation of the chlorination steps. The iron
reduction, whilst not convenient for very large scale due to
challenges with mixing, was deemed appropriate for early bulk
campaigns, and the cyclisation was trivial.
Toll-like receptors (TLRs) are a family of conserved transmem-
brane receptors which form part of the innate immune system that
detects pathogen-associated molecular patterns and helps to
mount an immune response to challenges by specific pathogens.
Certain TLRs are known to recognise viral nucleic acids, and serve
to protect host cells from viral infection. Of these, both TLR7 and
TLR8 recognise single-stranded RNA, and have been studied for
potential applications in infectious disease and as vaccine adju-
vants. As part of a programme to identify small molecule agonists
of the TLR7 receptor for antiviral applications, PF-4171455 (1)1
emerged as a potent and selective TLR7 agonist and was advanced
through preclinical toxicology studies. In order to support the
planned studies, bulk quantities of 1 were required. Herein we
describe the development of a robust, scalable route to 1.
Initial Routes to PF-4171455 (1). The original route towards
PF-4171455 (1) is shown in Scheme 1. The synthesis commenced
with the acylation of ethyl 3-amino-4,4,4-trifluorocrotonate 3 with
ethyl malonyl chloride 2 using triethylamine in dichloromethane.
The resultant enamide 4 was then cyclised to pyridone ester 5 in
high yield by deprotonation with 60% sodium hydride dispersion in
tetrahydrofuran. Ester hydrolysis followed by decarboxylation in
refluxing concentrated hydrochloric acid furnished pyridone 6 in
excellent yield. Nitration of 6 with fuming nitric acid in a mixture of
acetic acid and ethyl acetate afforded the nitro compound 7 in
reasonable yield which was then selectively chlorinated at the
4-position in neat phenylphosphonic dichloride to give chloropyr-
idone 8. Displacement of the chloride with benzylamine in tetra-
hydrofuran afforded aminopyridone 9 which was chlorinated in the
2-position with phenylphosphonic dichloride to provide chloropyr-
idine 10. Nitro reduction with iron in aqueous acetic acid gave the
diaminopyridine 11 which was subsequently cyclised with carbonyl
diimidazole in acetonitrile to give deazapurinone 12 in excellent
The main problem in utilising this route was the copper-
mediated benzylamine displacement, which required 4 days to
Received: January 25, 2011
Published: May 09, 2011
r
2011 American Chemical Society
788
dx.doi.org/10.1021/op200021a Org. Process Res. Dev. 2011, 15, 788–796
|

Organic Process Research & Development
ARTICLE
Scheme 1. First-generation route
Scheme 2. Second-generation route
Figure 1. C-acylated impurity.
reach completion. This used stoichiometric amounts of copper
sulphate, thus making it unattractive from a “green” perspective, and
having large amounts of metal species in the penultimate step could
require additional processing to control copper levels in the API.
However, most importantly the reaction was extremely capricious in
yield, and as the scale increased to tens of grams, the yield sharply
dropped to provide extremely low amounts of product.
Due to the very low solubility of PF-4171455 1 in organic
solvents, the final step would be problematic on scale as the initial
procedure involved neutralisation of the sulfuric acid and then
extraction of the product with large amounts of ethyl acetate.
Additionally, if purification was required after initial isolation, the
low solubility could significantly reduce the options available.
The poor reactivity of the chlorodeazapurinone 12 repre-
sented a major challenge, and the most obvious solution to
address the problem was to perform the displacement before the
nitro reduction. An alternative synthesis was performed as shown
in Scheme 2.
Advanced intermediate chloropyridine 10 was reacted with
7 M ammonia in methanol, affording the displaced product 15 in
high yield. Iron reduction provided the triaminopyridine 16 in
excellent yield. Carbonyl diimidazole in acetonitrile was then
789
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
Scheme 3. New proposed route
Hence, triethylamine was replaced with pyridine which is com-
monly applied as base and catalyst for acylation reactions. A
smaller exotherm was noted when using pyridine,7 and the
reaction profile was improved in favor of desired product, but
both unreacted enamine 3 and the over acylated product 14 were
still formed. It was decided to take the crude mixture8 into the
cyclisation in the hope that the bulk could be purified at the solid
pyridone 5.
It was reasoned that sodium hydride was wholly unnecessary
as the choice of base, and instead the much safer potassium tert-
butoxide was employed.9 Gratifyingly, treatment of crude enam-
ide 4 with potassium tert-butoxide in tetrahydrofuran gave the
cyclised product 5 with complete consumption of 4 by LCꢀMS.
It was observed that having taken on the crude material, the
reaction medium became increasingly thick and difficult to stir
which had not been observed when using clean enamide 4.
Change of solvent from tetrahydrofuran to ethanol kept the
reaction sufficiently mobile, and upon concentration, the product
could be isolated cleanly by dissolving in water and then
acidifying to approximately pH 3 with citric acid. Filtration of
the resulting precipitate provided pyridone ester 5 in 63% yield
over two steps.
In the decarboxylation, the acid strength was reduced by half.
Slurrying the pyridone 5 in 6 M hydrochloric acid and then
heating with overhead stirring to effect dissolution provided the
pyridone 6 in 94% yield.
used to effect cyclisation to PF-4171455 1 but unfortunately a
mixture of regioisomers PF-4171455 1 and 17 was obtained,
which could only be separated by a very difficult chromatogra-
phy. Additionally, the reaction profile showed a number of
unidentified byproducts which may have been formed through
partial decomposition of the electron-rich pyridine 16.
Neither of these two approaches was ideal, each posing
different challenges. If the nitro group was reduced early in the
sequence, the reactivity in the chloro displacement was markedly
reduced as demonstrated in Scheme 1. Conversely, if both
chlorine substituents were displaced before nitro reduction,
significant regioselectivity issues for the cyclisation were ob-
served as demonstrated in Scheme 2.
One approach considered was to install a suitably protected
amino group at the 2-position such as dibenzyl or diallylamine,
which would be unable to participate in the cyclisation. However,
this seemed a very inelegant strategy as the deprotection would
have added one step, and isolation of pure material could be an
issue due to poor solubility of PF-4171455 1.
A more attractive approach was capping the 4-benzylamino
substituent 10 with a carbamate group. It was anticipated that the
carbamate functionality would reduce the electron density on the
pyridine ring, promoting the chloro displacement and stabilising
the diaminopyridyl product of the subsequent reduction. Most
importantly it was anticipated that upon reduction of the nitro
group in aminopyridine 19, cyclisation of the newly formed
amino group onto the tethered carbamate in situ would provide
the deazapurinone PF-4171455 1 in a regioselective fashion as
shown in Scheme 3.6 This was the strategy adopted to produce
material for exploratory toxicology studies prior to candidate
nomination, targeting initial deliveries of 50 g.
For the nitration, alternative conditions were required to avoid
the possibility of forming acetyl nitrate. It was anticipated that the
classical mixture of nitric acid and sulphuric acid would be
suitable for this transformation. Pyridone 6 was dissolved in
concentrated sulphuric acid and the exotherm observed upon
addition allowed to dissipate. One equivalent of concentrated
nitric acid was then added dropwise, with the reaction monitored
carefully by quenching of small samples and analysis by LCꢀMS
to ensure nitration was progressing, to avoid a build-up of
nitrating agent and the potential for associated thermal hazards.
Upon completion of addition, the reaction was carefully
quenched into an ice and water mixture and isolated by extrac-
tion with ethyl acetate to provide nitropyridone 7 in good yield.
For the chlorination of the 4-hydroxyl function of 7 the
number of equivalents of phenylphosphonic dichloride was
markedly reduced from 80 to just 5. Although chloropyridone
8 was isolated in high yield, the workup was slightly laborious.
After quenching into water followed by treatment with sodium
hydrogen carbonate, the pyridine 8 was extracted into ethyl
acetate as the sodium salt. Upon concentration it was dissolved in
fresh water and then precipitated by the addition of hydrochloric
acid. For the amine displacement at least 3 equivalents of
benzylamine was necessary: the first to deprotonate the pyr-
idone, the second to react, and the third to sequester the HCl
produced. Hence, reaction of chloropyridone with 3.5 equiva-
lents of benzylamine in tetrahydrofuran gave complete conver-
sion to the desired product. Evaporating the reaction mixture to
dryness and then quenching with aqueous hydrochloric acid
precipitated the benzylaminopyridone 9 which could be isolated
cleanly by filtration in good yield.
Route Development and Early Optimisation. The initial
part of the approach involved developing a scalable route toward
nitropyridine intermediate 10 to provide material for the new
end-game. A key objective was to remove the need for column
chromatography in the formation of enamide 4 (Scheme 4). It
was considered that triethylamine could deprotonate the malonyl
chloride, thereby generating a ketene as a more reactive inter-
mediate which could lead to increased levels of C-acylation.
The second chlorination took minimal optimisation to scale,
again just requiring a reduction in the number of equivalents of
phenylphosphonic dichloride from 55 to 6.5. The workup for this
second chlorination proved far more straightforward than the
first and chloropyridine 10 was isolated in high yield. Treatment
of chloropyridine 10 with potassium tert-butoxide at 0 °C in
790
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
Scheme 4. Final route to PF-4171455 1
tetrahydrofuran afforded the anion as a bright purple solution.
This was then quenched with ethyl chloroformate to provide the
ethyl carbamate 20 in excellent yield. Displacement of the
chloride of ethyl carbamate 20 with concentrated ammonia in
a sealed vessel at 80 °C gave aminopyridine 21 in good yield with
no reaction observed at the carbamate. The crude product was
then hydrogenated under 40 psi hydrogen, catalysed by 10%
palladium on charcoal to give the diaminopyridine 22 which was
recrystallised from tert-butylmethyl ether to provide clean ma-
terial for the final step. Fortunately, the diaminopyridine 22 did
not automatically cyclise to the deazapurinone upon reduction as
the low solubility of PF-4171455 1 would have led to great
difficulty in removing the catalyst. Cyclisation to the deazapur-
inone target PF-4171455 1 was easily achieved by dissolving
diaminopyridine 22 in acetic acid and warming to 80 °C whereby
the final material precipitated from solution and was isolated by
simple filtration with excellent purity and an overall yield of 12%
over 11 linear steps to provide PF-4171455 1 on 50-g scale.
The route was selected for delivery of material to support
candidate nomination and early toxicology and clinical studies.
Close collaboration between Worldwide Medicinal Chemistry
and the Research-API group directed the chemistry being
performed oneither side of the formal point of transfer. For
further scale-up, the nitropyridone 7 was selected as the in-house
starting material and successfully sourced for use in batches to
support preclinical safety studies and early clinical supply.
Development for Kilo Lab and Pilot Plant. Enabling of the
chemistry for the early campaigns was directed at rapidly
delivering processes for the developed route that could be run
on kilo-lab and pilot-plant scale to support the early program
demands. In the first of the in-house steps, the conversion of
nitropyridone 7 to chloropyridone 8, screening was performed
with alternative, more atom-efficient, chlorinating agents, but
none gave improvement over the currently used phenylpho-
sphonic dichloride.10 In development of this step, an early
challenge encountered was the reaction stalling upon scale-up
into 1-L-scale glassware. A number of hypotheses were tested
including mode and rate of agitation, but it was discovered that
nitrogen flow across the reaction was the key factor. At high
nitrogen flow rates, the reaction would stall, whereas at lower
nitrogen flow rates the reaction would proceed to completion.
This was attributed to the need for a critical concentration of
hydrogen chloride to remain in the reaction mixture. At the
reaction temperature of 95 °C, hydrogen chloride is in equilib-
rium between the reaction mixture and the head space, and high
nitrogen flow rates reduced the concentration of hydrogen
chloride in the headspace, thus reducing the concentration in
the reaction mixture.
Once understood, this factor could be easily controlled. The
reaction was monitored closely over the 19 h reaction time to
ensure the correct balance of desired product to dichloro
impurity. The workup was simplified by precipitating the crude
product from the reaction mixture by addition of water. The first
isolated material contained varying quantities of phenylpho-
sphonic acid that was readily removed by reslurrying in water.
On pilot-plant scale, this process was run, starting with 24.9 kg of
nitropyridone 7 to provide 20.2 kg of chloropyridone 8 contain-
ing 14.7% dichloro impurity.
In the formation of the benzylaminopyridone 9, the 3.5
equivalents of benzylamine employed previously gave excellent
levels of conversion, with less than 3% starting remaining. The
workup was further simplified by reducing the reaction volume
(to ∼3 mL/g) and adding aqueous hydrochloric acid, which
dissolved the benzylamine hydrochloride byproduct and
791
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
precipitated the product. The initially isolated product contained
a significant amount of the bis-benzyl impurity (12.4% on pilot-
plant scale); this was effectively reduced (to 0.62% on pilot-plant
scale) by adding tert-butylmethyl ether and reducing the volume
by distillation to approximately 50%. This process was run on
20.0 kg of chloropyridine 8 to give 22.1 kg of benzylaminopyr-
idone 9 in a yield of 64% across the first two steps.
Phenylphosphonic dichloride was retained as the reagent in
the second chlorination to give chloropyridine 10. In develop-
ment of this step, information learned from the first step-around
controlling the nitrogen flow rate and the workup was applied.
However, two water reslurries were required for this step to
remove the residual phenylphosphonic acid, possibly as a result
of the higher loading of phenylphosphonic dichloride employed.
Although the reaction and the processing on scale took longer
than those for the preparation of the chloropyridone 8, after
workup, good-quality chloropyridine 10 was obtained. In the
largest campaign, 20 kg of benzylaminopyridone 9 was converted
to 15.7 kg of chloropyridine 10 in 77% yield.
Figure 2. Impurities from the nitro reduction of aminopyridine 21.
For the conversion of chloropyridine 10 to ethyl carbamate 20,
screening demonstrated that a strong base was required and a low
temperature was critical for high levels of conversion. A number
of bases were screened at 20 °C, but none performed as well as
the previously used potassium tert-butoxide at low temperature
(0 °C and below). With potassium tert-butoxide at ꢀ10 °C,
greater than 98% conversion was achieved, whereas at 0 °C
conversion was typically 90%. Even slight deviations above this
temperature during a reaction gave a marked reduction in
conversion and an increased level of byproduct. For example,
one reaction in which the temperature rose to 4 °C after addition
gave 17% starting material remaining.
Although nitro groups frequently impart crystallinity to com-
pounds, the ethyl carbamate 20 was not readily isolated as a solid.
For the workup, the reaction was quenched with water, washed
with aqueous sodium chloride and reduced in volume to give a
tetrahydrofuran solution that was used in the next step.
The initial conditions for conversion of the ethyl carbamate 20
to aminopyridine 21 relied on the use of concentrated ammonia
and tetrahydrofuran heated to 80 °C in a sealed vessel; the
reaction generated ∼80 psi of pressure and was not amenable to
scale-up into available kilo-lab or pilot-plant facilities. Examina-
tion of the concentrated ammonia and tetrahydrofuran condi-
tions at 50 °C in a standard vessel gave excellent conversion
(<2% ethyl carbamate 20 remaining after 18 h). As with the
previous step, the product aminopyridine 21 was not readily
isolated as a solid. For the workup of this step, the reaction was
washed with aqueous sodium chloride and distilled and replaced
with ethanol, to give an ethanol solution that was used in the
hydrogenation.
In the reduction of the aminopyridine 21 to the diaminopyr-
idine 22, examination of the HPLC analysis from initial hydro-
genation reactions showed four key impurities, the triamine 23,
the des-carbamate 24, the N-hydroxy cyclised 2511 and the
nitroso 2612 (see Figure 2).
In addition to these four impurities a small amount of the API,
PF-4171455 1, was also formed. Initial catalyst screening showed
that 10% and 20% PdꢀC from Degussa (E101) gave the best
results in terms of conversion and minimising impurities. The
10% PdꢀC from Degussa (E101) was further screened at a range
of loadings (between 2% and 15%). There was very good linear
(first-order) correlation between the catalyst loading and the rate
of reaction, as judged by hydrogen uptake, and the higher
Figure 3. Ratio of product to impurities in the reduction of aminopyr-
idine 21 with varying levels of 10% PdꢀC from Degussa (E101).
loadings also gave much better ratios of product to impurities
(Figure 3).
Further experiments were performed using the 10% PdꢀC
and varying both the pressure and temperature (see results in
Table 1). One of the key aims from this series of experiments was
to minimise the amount of the nitroso compound 26 observed in
the reaction profile, eliminating the need to specifically control
this impurity in the workup. Under a number of conditions, none
of the nitroso compound was detected, and of these 100 psi and
20 °C gave the highest level of conversion and lowest overall level
of impurities. Through the development of this step, it was
observed that the other impurities could be controlled through
the crystallisation and isolation.
With diaminopyridine 22 being the first isolated solid over
three steps, crystallisation was critical to ensure high-quality PF-
4171455 1. Additionally, during the workup there was a require-
ment to minimise the temperature to prevent thermal conversion
of diaminopyridine 22 to PF-4171455 1. Early development of
the workup showed that crystallisation from a nonpolar solvent
(e.g., tert-butylmethyl ether) afforded good recovery, but no
reduction in the level of the triamine 23. Further development
showed that reducing the ethanol volume and adding an anti-
solvent (e.g., toluene, heptane or tert-butylmethyl ether) gave
good quality material, crucially maintaining ∼1% ethanol was key
to minimising the amount of triamine 23 in the isolated product.
The final workup procedure involved concentrating the ethanol
solution after filtration to 2 mL/g under vacuum below 50 °C,
addition of tert-butylmethyl ether (20 mL/g) and concentrating
792
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
Table 1. Ratio of product to impurities in the reduction of aminopyridine 21 with varying temperature and pressure
pressure (PSI) temp (°C) % product 22 % SM 21 % triamine 23 % des-carbamate 24 % N-hydroxy cyclised 25 % nitroso 26 % PF-4171455 1
50
50
20
40
20
88.2
92
0.7
3.4
5.2
3.5
4.0
1.1
1.7
0.8
ND
0.9
ND
ND
ND
1.0
1.2
1.0
0.49
0.1
100
92.9
to 10 mL/g at atmospheric pressure. The distill and replace
process was repeated twice and after cooling to 0 °C the product
was isolated by filtration.
The organic phases were combined, dried over Na2SO4, filtered
and concentrated to a dark green oil (538.7 g) which was judged
to contain ∼60% enamide 4 by 1H NMR. The crude enamide 4
was used directly in next step.
In an initial non-GMP campaign, complete removal of ethanol
during the workup gave product containing 3.5% of the triamine
23 and 6.6% of PF-4171455 1, with 85.8% of the diaminopyridine
22, in a yield of 51% from chloropyridine 10. Learnings high-
lighted above were incorporated for the subsequent GMP
campaign in which 15.7 kg of chloropyridine 10 was converted
to 11.3 kg of high quality diaminopyridine 22 in a yield of 67%
over the three steps.
The final step in the preparation of PF-4171455 1 was thermal
cyclisation in acetic acid. The key challenge with this step was to
balance complete consumption of the starting material with over-
reaction to an acylated byproduct, which although readily
purged, did result in a reduction in yield. The optimum tem-
perature was found to be 100 °C for 9 h. For a final step, the
processing was very straightforward, with the product precipitat-
ing on cooling at approximately 50 °C and after cooling to 20 °C,
the product was filtered and washed with acetic acid and ethanol
to give 8.8 kg of PF-4171455 1 of appropriate quality in 89% yield
on pilot=plant scale.
Collaboration between Worldwide Medicinal Chemistry and
Research-API during the early investment in synthetic route
identification was vital in ensuring the overall program could
progress in a rapid manner. The initial route to PF-4171455 used
in Worldwide Medicinal Chemistry consisted of 11 steps and
gave an overall yield of less than 1%. This route was found not to
be amenable to scale-up on the basis of a number of factors. Prior
to compound nomination, the key API delivery was a 50 g lot for
exploratory toxicology studies using a new route with the same
number of steps but with a 10.5% overall yield. The collaboration
ensured that the route identified at this stage was suitable for
multikilogram scale and allowed for very smooth knowledge
transfer between both partners. The familiarity with the chem-
istry allowed an 8.8 kg GMP batch of PF-4171455 1 to be
completed in similar overall yield (11%) within six months of
compound nomination.
4-Hydroxy-2-oxo-6-trifluoromethyl-2-dihydropyridine-3-
carboxylic Ethyl Ester (5). To a magnetically stirred solution of
crude enamide 4 (150 g, 60% by weight, 0.3 mol) in EtOH
(600 mL) in a water bath at ambient temperature was added
potassium tert-butoxide (62.3 g, 0.555 mol) in a portionwise
fashion over 25 min such that the temperature did not exceed
40 °C. Upon completion of addition, the reaction mixture was
heated at 70 °C for 2 h and then allowed to cool to room
temperature and stirred overnight. At this stage, the reaction was
a deep-red suspension. The solvent was then removed in vacuo to
give a solid. A saturated solution of citric acid (400 mL) was
added to the solid and the resulting suspension stirred for 30 min.
The suspension was then filtered to give a pale-pink solid which
was washed with several portions of citric acid solution. The filter
cake was triturated with pentane, filtered and dried in vacuo at
40 °C for 18 h to give pyridone ester 5 as a pale-pink solid (72 g,
95% yield, equivalent to 63% over the two steps). 1H NMR (400
MHz, DMSO-d6): δ 1.22 (t, J = 7 Hz, 3H), 4.21 (q, J = 7 Hz, 2H),
6.76 (s, 1H), 11.95 (br s, 2H). 13C NMR (400 MHz, DMSO-d6):
δ 14.06, 61.03, 102.32, 105.99, 121.03 (q, J = 274 Hz), 144.25 (q,
J = 34 Hz), 162.51, 165.00, 165.36. Anal. Calcd for C9H8F3NO4:
C, 43.04; H, 3.21; N, 5.58. Found: C, 42.66; H, 3.20; N, 5.36.
HRMS m/z Found 252.0480 [M þ H]þ C9H8F3NO4 requires
252.0478. Mp 176ꢀ177 °C. IR (KBr) 3110, 2996, 2910, 2830,
1659, 1505, 1437, 1378, 1271, 1197, 1156, 1093, 1012, 864,
677 cmꢀ1
.
4-Hydroxy-6-trifluoromethyl-1H-pyridin-2-one (6). To a
mechanically stirred 6 M HCl (aq) (1 L) at ambient temperature
was added pyridone ester 5 (165 g, 0.657 mol) in a portionwise
manner and the resulting suspension heated at reflux overnight.
After several hours at reflux a pale yellow solution had formed
followed by the formation of a pale-yellow suspension. The
reaction mixture was cooled with an ice bath to 0 °C and
neutralised via the dropwise addition of 0.88 ammonia solution
(∼500 mL). At pH 7 a thick white suspension had formed. The
suspension was filtered, the filter cake washed with cold water
and dried in vacuo at 50 °C over P2O5 for 66 h to give pyridone 6
as a white solid (110.6 g, 94%). 1H NMR (400 MHz, DMSO-d6):
δ 6.12 (s, 1H), 6.66 (s, 1H). 13C NMR (400 MHz, DMSO-d6): δ
98.01, 103.05, 121.58 (q, J = 273 Hz), 144.47 (q, J = 33 Hz),
165.90, 168.57. Anal. Calcd for C6H4F3NO2: C, 40.24; H, 2.25;
N, 7.82. Found: C, 40.33; H, 2.34; N, 7.56. HRMS m/z Found
202.0088 [M þ Na]þ C6H4F3NO2 requires 202.009183. Mp
250ꢀ251 °C. IR (KBr) 3124, 3050, 2931, 1679, 1619, 1460,
’ EXPERIMENTAL SECTION
3-(2-Ethoxycarbonyl-acetylamino)-4,4,4-trifluoro-but-2-
enoic Acid Ethyl Ester (4). To a solution of ethyl 3-amino-4,4,4-
trifluorocrotonate 2 (300 g, 1.638 mol) in pyridine (160 mL, 1.97
mol, 1.2 equiv) and DCM (1.5 L) in a water bath at ambient
temperature was added neat ethyl malonyl chloride 3 (252 mL,
1.97 mol, 1.2 equiv) dropwise over approximately 50 min,
maintaining the temperature below 30 °C. The resulting brown
solution was stirred in the water bath for 2 h (to a temperature of
22 °C) and then removed from the water bath. The resulting
dark-green solution was stirred at ambient temperature for 66 h.
The resulting solution was washed with 1 M HCl (aq) (600 mL),
and sat. aq NaHCO3 (800 mL). Each aqueous phase was
sequentially re-extracted with further DCM (800 and 600 mL).
1357, 1300, 1206, 1019, 987, 854, 724 cmꢀ1
.
4-Hydroxy-3-nitro-6-trifluoromethyl-1H-pyridin-2-one
(7). Pyridone 6 (53 g, 0.3 mol) was added portionwise to
concentrated sulphuric acid (130 mL) in 1ꢀ2 g portions, with
stirring after each addition until all solids had dissolved. During
the addition, the temperature increased to 50 °C. When all the
793
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
pyridone 6 had been added, the solution was a pale-brown
colour, and the temperature was 43 °C. Concentrated nitric acid
(∼70%, 20 mL, 0.31 mol, 1.05 equiv) was added three drops at a
time, over a period of approximately 1.5 h. The addition was
exothermic, but the rate of addition maintained the temperature
between 40 and 50 °C without external heating or cooling. When
all of the concentrated nitric acid had been added, the solution
was a bright-orange colour. Once all the nitric acid had been
added, the reaction mixture was stirred, achieving ambient
temperature over 1.5 h, and stirred at this temperature for a
further 2.5 h. The reaction mixture was then poured slowly onto
an efficiently stirred mixture of ice (800 g) and water (500 g).
The resulting fluffy white solid was filtered, dissolved in EtOAc
(200 mL) and poured into a separating funnel to remove
approximately 10 mL of water present. The organic phase was
collected, dried over Na2SO4, filtered and concentrated to give an
off-white solid (32.6 g). The main aqueous filtrate (approx 1.4 L)
was then extracted with EtOAc (3 ꢁ 500 mL). The organic
phases were combined, dried over Na2SO4, filtered and concen-
trated to give an off-white solid (23.7 g). Both solid batches were
dissolved in warm EtOAc (200 mL), and heptane (200 mL) was
added to effect precipitation. This was filtered to give a white
solid that was washed with heptane (200 mL). The solid was
dried in vacuo at 40 °C for 18 h to give nitropyridone 7 (37.5 g).
Further precipitate had emerged in the filtrate (now 600 mL) and
was filtered and dried in vacuo at 40 °C for 18 h to give additional
nitropyridone 7 (9.0 g). The combined yield of nitropyridone 7
was 46.5 g (70%). 1H NMR (400 MHz, DMSO-d6): δ 6.80 (s,
1H). 13C NMR (400 MHz, DMSO-d6): δ 102.30, 120.26 (q, J =
275 Hz), 126.66, 142.90 (m), 157.00, 159.22. Anal. Calcd for
C6H3F3N2O4: C, 32.16; H, 1.35; N, 12.50. Found: C, 31.93; H,
1.39; N, 12.29. HRMS m/z Found 225.0121 [M þ H]þ
C6H3F3N2O4 requires 225.0118. Mp 164ꢀ165 °C. IR (KBr)
3073, 2978, 2906, 2810, 1672, 1532, 1487, 1454, 1374, 1270,
IR (KBr) 3174, 3123, 2921, 2810, 2435, 1679, 1545, 1484, 1385,
1271, 1216, 1151, 999, 872, 849, 818, 711 cmꢀ1
.
4-Benzylamino-3-nitro-6-trifluoromethyl-1H-pyridin-2-one
(9). Chloropyridone 8 (20.0 kg, 83 mol) was charged to THF
(66.8 kg) at 20 °C followed by addition of a solution of
benzylamine (30.0 kg, 280 mol, 3.4 equiv) in THF (22.3 kg).
The lines were washed with THF (14.6 kg) and the mixture
heated to reflux and stirred at this temperature for 22 h. The
mixture was cooled to 20 °C and an IPC taken that confirmed the
reaction had gone to completion (pass criteria SM <3%, result of
SM 2.6%). The mixture was heated and distilled at atmospheric
pressure until the volume was approximately 60 L and then
cooled to 20 °C. Water (268 L) and 2 M HCl (aq) (83 L of water
and 20.6 L of conc. HCl) and water (33 L) were sequentially
added, and the mixture was granulated at 20 °C for 24 h. The
mixture was filtered and washed with water (2 ꢁ 40 L) and dried
in vacuo at 50 °C for 4 days to give water-wet benzylaminopyr-
idone 9 (35.48 kg).
The water wet material isolated above was charged to t-BME
(367.0 kg) and mixture heated to reflux and distilled until the
volume was approximately 248 L. The mixture was cooled to
20 °C, granulated at that temperature for 2 h, filtered and washed
with t-BME (18.4 kg ꢁ 2) dried at 45 °C in vacuo for 25 h to give
benzylaminopyridone 9 (22.1 kg, 64% over the two transforma-
tions from nitropyridone 7). 1H NMR (400 MHz, DMSO-d6): δ
4.69 (d, J = 6 Hz, 1H), 6.54 (s, 1H), 7.25ꢀ7.38 (m, 5H), 9.10 (br
s, 1H), 12.63 (br s, 1H). 13C NMR (400 MHz, DMSO-d6): δ
45.97, 95.03, 119.75 (q, J = 275 Hz), 120.88, 127.00, 127.40,
128.73, 137.77, 139.40 (m, J = 32 Hz), 151.25, 157.17. Anal.
Calcd for C13H10F3N3O3: C, 49.85; H, 3.22; N, 13.41. Found: C,
50.00; H, 3.23; N, 13.19. HRMS m/z Found 314.0759 [M þ
H]þ C13H10F3N3O3 requires 314.0747. Mp 206ꢀ207 °C. IR
(KBr) 3329, 3066, 2863, 1663, 1578, 1531, 1450, 1375, 1345,
1266, 1211, 1133, 885, 811, 735, 699 cmꢀ1
.
1211, 1156, 1080, 947, 841, 793 cmꢀ1
.
Benzyl-(2-chloro-3-nitro-6-trifluoromethylpyridin-4-yl)amine
(10). Benzylaminopyridine 9 (20.0 kg, 64 mol) was charged to a
mixture of phenylphosphonic dichloride (82.4 kg, 423 mol, 6.75
equiv) and DCM (10.0 kg line wash) at 20 °C and heated to 95 °C
and stirred at that temperature for 22 h. The mixture was cooled to
20 °C and an IPC taken that confirmed the reaction had gone to
completion (pass criteria <5% SM, result of SM < 1%). The
reaction mixture was added to water (at 20 °C) (300 L) main-
taining the temperature in the range 20ꢀ35 °C. The resulting
mixture was stirred at 20 °C for 1 h to confirm that a suspension
had formed and the suspension granulated at 20 °C for 16 h. The
resulting suspension was filtered using an agitated filter dryer and
the filter cake washed with water (2 ꢁ 25 L). Water (200 L) was
added to the agitated filter dryer and the mixture heated to 57 °C
over 6 handstirredatthis temperaturefor 3 h then cooledto 20°C
over 5 h 30 min. The mixture was stirred at this temperature for 1 h
and then filtered, and the filter cake was washed with water (2 ꢁ
20 L). Water (200 L) was added to the agitated filter dryer, and the
mixture was heated to 57 °C over 6 h and stirred at this
temperature for 3 h and cooled to 20 °C over 4 h. The mixture
was stirred at this temperature for 1 h 30 min and then filtered, and
the filter cake was washed with water (2 ꢁ 20 L). The resulting
solid was dried in vacuo at 55ꢀ65 °C for 3 days to give
4-Chloro-3-nitro-6-trifluoromethyl-1H-pyridin-2-one (8).
To a mixture of phenylphosphonic dichloride (103.0 kg, 529
mol, 4.7 equiv) and DCM (10.0 kg, added as a line wash) was
added nitropyridone 7 (24.90 kg, 111 mol) and the mixture
heated to 95 °C and stirred at that temperature for 15 h. The
mixture was cooled to 20 °C and an IPC taken that confirmed
that the reaction had gone to completion (pass criteria <5% SM,
result of SM none detected). The reaction mixture was added to
water at 20 °C (374 L) maintaining the temperature in the range
20ꢀ28 °C. The resulting mixture was stirred at 20 °C for 1 h to
confirm that a suspension had formed and the suspension
granulated at 20 °C for 5 h 30 min. The resulting suspension
was filtered and the filter cake washed with water (2 ꢁ 25 L).
Water (250 L) was added to the agitated filter dryer and the
mixture heated to 57 °C. A solution was observed during this
operation, and after cooling to 20 °C, 50 L of filtrate was removed
and the remaining mixture stirred at 57 °C for 5 h 30 min, then
cooled to 20 °C over 6 h. The mixture was filtered and the filter
cake washed with water (2 ꢁ 25 L). The resulting solid was dried
in vacuo at 50 ꢀ 55 °C for 44 h to give pyridone 8 (20.23 kg, 75%
based on 100% potency, water content <1%, dichloro impurity
15% by area normalisation). 1H NMR (400 MHz, DMSO-d6): δ
7.74 (s, 1H). 13C NMR (400 MHz, DMSO-d6): δ 113.23, 119.98
(q, J = 275 Hz), 136.28, 137.45, 144.60 (q, J = 36 Hz), 156.55.
Anal. Calcd for C6H2F3ClN2O3: C, 29.71; H, 0.83; N, 11.55.
Found: C, 30.10; H, 0.88; N, 11.13. HRMS m/z Found 242.9782
[M þ H]þ C6H2F3ClN2O3 requires 242.9779. Mp 169ꢀ170 °C.
1
chloropyridine 10 (15.7 kg, 77%, water content <0.5%). H
NMR (400 MHz, DMSO-d6): δ 4.63 (d, J = 6 Hz, 2H),
7.24ꢀ7.38 (m, 5H), 8.63 (t, J = 6 Hz, 1H). 13C NMR (400
MHz, DMSO-d6): δ 45.55, 106.07, 120.25(q, J = 275 Hz), 126.97,
127.38, 128.64, 132.66, 136.96, 142.54, 146.00 (q, J = 34 Hz),
794
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
148.33. Anal. Calcd for C13H9F3ClN3O2: C, 47.08; H, 2.74; N,
12.67. Found: C, 47.12; H, 2.70; N, 12.64. HRMS m/z Found
332.0424 [M þ H]þ C13H9F3ClN3O2 requires 332.0408. Mp
110ꢀ111 °C. IR (KBr) 3342, 1612, 1541, 1459, 1357, 1286, 1192,
4.93 (s, 2H), 6.90 (s, 1H), 7.22 ꢀ7.34 (m, 5H), 7.57 (s, 2H). 13C
NMR (400 MHz, DMSO-d6): δ 13.83, 52.47, 62.68, 106.30,
120.30 (q, J = 276 Hz), 127.63, 128.04, 128.43, 136.38, 145.93,
148.17 (q, J = 35 Hz), 153.29, 153.43.14 Anal. Calcd for
C16H15F3N4O4: C, 50.01; H, 3.93; N, 14.58. Found: C, 50.02;
H, 3.87; N, 14.55. HRMS m/z Found 385.1136 [M þ H]þ
C16H15F3N4O4 requires 385.1118. IR (neat) 3492, 3334, 3195,
2986, 1722, 1612, 1525, 1422, 1338, 1257, 1153, 1021, 944, 887,
1150, 1056, 970, 867, 756, 701 cmꢀ1
.
Benzyl-(2-chloro-3-nitro-6-trifluoromethylpyridin-4-yl)
carbamic Acid Ethyl Ester (20). A solution of potassium tert-
butoxide (5.84 kg, 52.0 mol) in THF (46.8 kg) at ꢀ10 °C was
added to a solution of chloropyridine 10 (15.7 kg, 47.3 mol) in
THF (257.2 kg) at ꢀ10 °C over 30 min maintaining the
temperature below ꢀ8 °C. The vessel containing the potassium
tert-butoxide solution and the transfer lines were washed through
with THF (27.9 kg), and the resulting mixture was stirred at
ꢀ10 °C for 1 h before a solution of ethyl chloroformate (5.90 kg,
54.4 mol) in THF (29.3 kg) at ꢀ10 °C was added over 30 min,
maintaining the temperature below ꢀ7 °C. The vessel contain-
ing the neat ethyl chloroformate, the ethyl chloroformate solu-
tion and the transfer lines were washed through with THF (14.0
kg) (cooled to ꢀ10 °C in the ethyl chloroformate solution
vessel), and the resulting mixture was stirred at ꢀ10 °C for 1 h.
The reaction was quenched by the addition of water (63 L) over
12 min, maintaining the content temperature less than 25 °C,
followed by the addition of a solution of sodium chloride (22.1
kg) in water (63 L). The resulting mixture was stirred for 15 min
and allowed to separate over 30 min. After a pH check on the
lower layer (acceptable range 5ꢀ9, result = 7.8), the lower layer
was separated and an IPC taken prior to distillation.13 The
mixture was distilled at atmospheric pressure until ∼94 L
remained, and this solution was used in the next step. Data from
a purified sample: 1H NMR (400 MHz, DMSO-d6): δ 1.13 (t, J =
7 Hz, 3H), 4.11 (q, J = 7 Hz, 2H), 5.04 (s, 2H), 7.25ꢀ7.34 (m,
5H), 8.22 (s, 1H). 13C NMR (400 MHz, DMSO-d6): δ 13.75,
52.86, 63.27, 119.65 (q, J = 275 Hz), 120.43, 127.82, 128.11,
128.44, 135.64, 143.07, 143.17, 146.13, 147.46 (q, J = 37 Hz),
152.84. Anal. Calcd for C16H13F3ClN3O4: C, 47.60; H, 3.25; N,
10.41. Found: C, 48.09; H, 3.34; N, 10.33. HRMS m/z Found
404.0635 [M þ H]þ C16H13F3ClN3O4 requires 404.0619. IR
(neat) 3087, 2986, 1730, 1553, 1425, 1310, 1253, 1154, 1011,
839, 739, 704 cmꢀ1
.
Benzyl-(2,3-diamino-6-trifluoromethylpyridin-4-yl)carba-
mic Acid Ethyl Ester (22). Palladium hydroxide on carbon
catalyst (10%, water wet, 0.89 kg) was added to a solution of
aminopyridine 21 (assumed 7.8 kg, 20.3 mol) in EtOH from the
previous step (55 kg) and EtOH (24 kg), followed by a water line
wash (2 L). The mixture was hydrogenated at 20 °C and 7.9 barg
(bar gauge) for 3 h, and a sample was taken for IPC (pass criteria
<5% SM, <2% nitrosointermediate, >80%product, resultsof SM =
1.3%, nitroso intermediate = 0.87% and product = 98%). The
catalyst was removed by recirculation over a Gauthier filter. The
filtrate was collected and the Gauthier filter washed by recircula-
tion with EtOH (50 kg) and combined with the original filtrate.
The above process was repeated at the same scale (with IPC
results of SM = 1.2%, nitroso intermediate = 0.85% and product =
98%), and the EtOH filtrates from both streams were combined
using additional EtOH (totalling 25 kg) as line washes.
The mixture from the combined streams was distilled at
reduced pressure (temperature maintained below 50 °C) until
a volume of approximately 36 L remained. t-BME (261 kg) was
added and the mixture distilled until a volume of approximately
180 L remained. t-BME (132 kg) was added and the mixture
distilled until a volume of approximately 180 L remained. t-BME
(130 kg) was added and the mixture distilled until a volume of
approximately 180 L remained. t-BME (130 kg) was added and
the mixture cooled to 0 °C at a rate of 1 °C/min, and the mixture
granulated at 0 °C for 1 h. The mixture was then filtered and
washed twice with t-BME (39 kg and 37 kg). The resulting solid
was dried at 55 ꢀ 65 °C for 16 h (result t-BME = 0.15%, EtOH =
none detected) to give diaminopyridine (11.3 kg, 31.9 mol, 67%
over the three transformations from chloropyridine 10).
949, 880, 759, 702 cmꢀ1
.
1H NMR (400 MHz, DMSO-d6): δ 1.12 (m, 3H), 4.09 (m,
2H), 4.25 (m, 1H), 5.03 (m, 1H), 5.43 (br s, 1H), 6.16 (br s, 1H),
6.33 (s, 1H), 7.22ꢀ7.31 (m, 5H). 13C NMR (400 MHz, DMSO-
d6): δ 14.47, 50.73, 61.30, 110.87, 122.39 (q, J = 272 Hz), 127.29,
128.05, 128.25, 129.54, 129.95 (q, J = 33 Hz), 137.66, 149.85,
154.78.14 Anal. Calcd for C16H17F3N4O2: C, 54.24; H, 4.84; N,
15.81. Found: C, 53.32; H, 4.71; N, 15.45. HRMS m/z Found
355.1372 [M þ H]þ C16H17F3N4O2 requires 355.1376. IR
(KBr) 3486, 3439, 3383, 3311, 3215, 2928, 1687, 1607, 1493,
(2-Amino-3-nitro-6-trifluoromethylpyridin-4-yl)-benzyl-
carbamic Acid Ethyl Ester (21). An 0.880 M ammonia solution
(31.9 kg) was added to a solution of ethyl carbamate 20
(assumed 18.1 kg, 44.9 mol) in THF (∼94 L) from the previous
step at 50 °C over 40 min maintaining the temperature in the
range 45ꢀ50 °C. The lines were washed with water (15 L) and
the resulting mixture stirred at 50 °C for 24 h. The mixture was
cooled to 20 °C and an IPC taken (pass criteria <5% SM, result of
15%). The mixture was heated to 50 °C, and 0.880 M ammonia
solution (4.0 kg) was added, maintaining the temperature in the
range 45ꢀ50 °C. The lines were washed with water (5 L), and
the resulting mixture was stirred at 50 °C for 17 h and cooled to
20 °C; a sample was taken for IPC (pass criteria <5% SM, result
of SM = 3%). A solution of sodium chloride (25.3 kg) in water
(72.4 L) was added, the mixture was stirred for 1 h 15 min and
then allowed to separate for 5 h 30 min. The aqueous layer was
removed and the organic layer distilled to a volume of ∼54 L.
Absolute EtOH (287 kg) was added and the resulting mixture
distilled to a volume of ∼54 L. Absolute EtOH (209 kg) was
added and the resulting mixture distilled to a volume of ∼163 L
to give a solution of aminopyridine 21 in EtOH (110 kg) that was
used in the next step. Data from a purified sample: 1H NMR (400
MHz, DMSO-d6): δ 1.09 (t, J = 7 Hz, 3H), 4.04 (q, J = 7 Hz, 2H),
1436, 1337, 1278, 1142, 1010, 964, 871, 767, 699, 663 cmꢀ1
.
4-Amino-1-benzyl-6-trifluoromethyl-1,3-dihydroimidazo
[4,5-c]pyridin-2-one (1). A suspension of diaminopyridine 22
(11.3kg, 31.9mol) inglacial AcOH(108 kg) was heated to100 °C
and maintained at that temperature before being cooled to 20 °C
at a rate of 2 °C/min. A sample was taken for IPC (passcriteria SM
< 1%, result of SM = 0.41%). The mixture was filtered and washed
with glacial AcOH (24.4 kg) and absolute EtOH (18.0 kg). The
resulting solid was dried at 55ꢀ65 °C for 29 h to give API PF-
4171455 1 (8.8 kg, 28.5 mol, 89%). 1H NMR (400 MHz, DMSO-
d6): δ 5.02(s, 2H), 6.22 (s, 2H), 7.00 (s, 1H), 7.21ꢀ7.32 (m, 5H),
10.77 (s, 1H). 13C NMR (400 MHz, DMSO-d6): δ 43.69, 93.62,
111.24, 122.10 (q, J = 274 Hz), 127.26, 127.54, 128.66, 135.34,
136.76, 137.47 (q, J = 33 Hz), 143.76, 153.55. Anal. Calcd for
795
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796

Organic Process Research & Development
ARTICLE
C14H11F3N4O: C, 54.55; H, 3.60; N, 18.17. Found: C, 54.48; H,
3.61; N, 17.97. HRMS m/z Found 309.0953 [M þ H]þ
C14H11F3N4O requires 309.0958. Mp 350 °C. IR (KBr) 3413,
3183, 2856, 1706, 1659, 1616, 1463, 1405, 1276, 1230, 1188, 1101,
(10) Phenylphosphonic dichloride has been reported to work well in
some cases where chlorinating agents such as POCl3 have failed and
higher temperatures are required: Robison, M. M. J. Am. Chem. Soc.
1958, 80, 5481.
(11) High yields of aromatic hydroxylamine can be obtained by
hydrogenation over platinum in lower alcohols containing 1ꢀ2%
dimethylsulfoxide: (a) Rylander, P. Aldrichimica Acta 1979, 12, 53.
Palladium catalysis has been used with hydrazine or phosphinic acid as
the reductant:(b) Rondestvedt, C. S., Jr.; Johnson, T. A. Synthesis
1977, 850. (c) Entwistle, I. D.; Gilkerson, T.; Johnstone, R. A. W.;
Telford, R. P. Tetrahedron 1978, 34, 213.
926, 812, 699, 664 cmꢀ1
.
’ AUTHOR INFORMATION
Corresponding Author
*gerwyn.bish@pfizer.com; paul.hodgson@pfizer.com.
(12) When nitro compounds are treated with most reducing agents,
either nitroso compounds are not formed, or they react further under the
reaction conditions and cannot be isolated: March, J. Advanced Organic
Chemistry 4th ed.; Wiley: New York, 1992; p1217.
’ ACKNOWLEDGMENT
We thank colleagues Dean Brick for pilot-plant support,
Wiebke Bahker and Paul Carrick for analytical support, Michael
Hawksworth for process safety support and Jinu Mathew for
hydrogenation support.
(13) An IPC was taken to ensure a low level of the starting material
chloropyridine 10 prior to starting the distillation at atmospheric
pressure to act as a control for potential thermal hazards. If the level
of chloropyridine 10 was greater than 5%, then a thermal stability test
was required before starting the distillation. The result for the process
described was 1.9% chloropyridine 10.
’ REFERENCES
(14) The 13C NMR spectra for 21 and 22 list 13 peaks, yet there are
14 nonequivalent carbons in each of these two molecules. It is likely the
missing signal overlaps with one of the other signals.
(1) (a) Jones, P.; Pryde, D. C; Tran, T. D. Patent Application WO
2007/093901 A1, 2007. (b) Pryde D. C.; Tran, T. D.; Jones, P.; Parsons,
G. C.; Bish, G.; Adam, F. M.; Smith, M. C.; Middleton, D. S.; Smith,
N. S.; Calo, F.; Hay, D.; Paradowski, M.; Proctor, K. J. W.; Parkinson, T.;
Laxton, C.; Fox, D. N. A.; Horscroft, N. J.; Ciaramella, G.; Jones, H. M.;
Duckworth, J.; Benson, N.; Harrison, A.; Webster, R. Med. Chem.
Commun., 2011, 2(3), 185.
’ NOTE ADDED AFTER ASAP PUBLICATION
This article was published on the Web on May 25, 2011. Two
authors were added. The corrected version was reposted on
June 10, 2011.
(2) Substoichiometric copper sulphate-mediated displacements of
poorly activated 2-chloropyridines with simple amines have been
sparsely reported in the literature. For example: (a) Maurel, J.; Bonnaud,
B.; Ribet, J.; Vahcer, B. Patent Application WO 2002/064585 A1, 2002.
(b) Kascheres, A.; Rodrigues, J. A. R.; Pilli, R. A. J. Org. Chem. 1983,
48, 1084.
(3) Literature examples of sulphuric acid-mediated debenzylations
of both 2- and 4-aminopyridines respectively include: (a) Wang, X.;
Spear, K. L.; Fulp, A. B.; Seconi, D.; Suzuki, T.; Ishii, T.; Moritomo, A.;
Kubota, H.; Kazami, J. Patent Application WO 2005/099711 A1, 2005.
(b) Tsuzuki, Y.; Tomita, K.; Shibamori, K.; Sato, Y.; Kashimoto, S.;
Chiba, K. J. Med. Chem. 2004, 47, 2097.
(4) Characterisation of a purified sample 14: 1H NMR (400 MHz,
CDCl3): δ 1.24ꢀ1.32 (m, 9H), 3.78 (s, 2H), 4.15ꢀ4.2 (q, 2H, J = 7 Hz),
4.20ꢀ4.26 (q, 2H, J = 7 Hz), 4.27ꢀ4.32 (q, 2H, J = 7 Hz), 6.14 (s, 1H),
11.71 (s, 1H).
(5) Bretherick’s Handbook of Reactive Chemical Hazards, 6th ed.;
Urben, P. G., Ed.; Pitt, M. J., compiler; Butterworth Heinemann: Boston,
1999; Vol. 1, entry 0765, acetyl nitrate.
(6) An acid-catalysed cyclisation to 1,3-dihydro-2H-imidazo[4,5-
c]pyridin-2-ones via a carbamate strategy had been recently reported
in the literature: Fiksdahl, A.; Bakke, J. M.; Holt, J. J. Heterocycl. Chem.
2006, 43, 787.
(7) A side-by-side comparison of the heat generated using triethy-
lamine versus pyridine was carried out by adding ethyl malonyl chloride
2 at a constant rate to 138 mmol enamine 3 over approx 3 min. The
starting temperature was noted as 18 °C. After one min the temperature
of the pyridine-mediated reaction had reached 30 °C, whereas the
triethylamine-mediated reaction was at 40 °C. At end of addition the
pyridine-mediated reaction was at a gentle reflux, whereas the triethy-
lamine-mediated reaction was vigorously refluxing and had blown the
nitrogen line free.
1
(8) Characterisation of a purified sample 4: H NMR (400 MHz,
DMSO-d6): δ 1.17ꢀ1.21 (m, 6H), 3.46 (s, 2H), 4.07ꢀ4.13 (q, 4H),
6.37 (s, 1H), 10.26 (s, 1H). 13C NMR (400 MHz, DMSO-d6): δ 14.52,
43.23, 61.17, 61.35, 116.38, 121.20 (q, J = 275 Hz), 131.28 (q, J = 34 Hz),
163.85, 164.90, 167.37. HRMS m/z Found 298.0884 [M þ H]þ
C11H14F3NO5 requires 298.0897.
(9) Potassium tert-butoxide was selected as it is an easily handled
base. No transesterification to give the tert-butyl ester was observed.
796
dx.doi.org/10.1021/op200021a |Org. Process Res. Dev. 2011, 15, 788–796