Development of a practical synthesis of a p38 kinase inhibitor via a safe and robust amination

Article abstract of DOI:10.1021/op300181r

The development of a practical synthesis for a p38 kinase inhibitor is described. The key advances include an improved route to the key intermediate, a substituted pyrrole, and a subsequent animation utilizing O-(4-nitrobenzoyl)- hydroxylamine, which provides a safe, scalable, and robust amination method. The new protocol was successfully demonstrated to generate 1.6 kg of API in seven steps and 26% overall yield.

Full text of DOI:10.1021/op300181r

Article
pubs.acs.org/OPRD
Development of a Practical Synthesis of a p38 Kinase Inhibitor via a
Safe and Robust Amination
Zhongping Shi,*,† Susanne Kiau,† Paul Lobben,† John Hynes, Jr.,‡ Hong Wu,‡ Luca Parlanti,†
Robert Discordia,† Wendel W. Doubleday,† Katerina Leftheris,‡ Alaric J. Dyckman,‡
Stephen T. Wrobleski,‡ Konstantinos Dambalas,† Srinivas Tummala,† Simon Leung,† and Ehrlic Lo†
†Chemical Development, Research & Development, Bristol-Myers Squibb Co., One Squibb Drive, P.O. Box 191, New Brunswick,
New Jersey 08903-0191, United States
‡Medicinal Chemistry, Bristol-Myers Squibb Co., Princeton, New Jersey 08543, United States
ABSTRACT: The development of a practical synthesis for a p38 kinase inhibitor is described. The key advances include an
improved route to the key intermediate, a substituted pyrrole, and a subsequent animation utilizing O-(4-nitrobenzoyl)-
hydroxylamine, which provides a safe, scalable, and robust amination method. The new protocol was successfully demonstrated
to generate 1.6 kg of API in seven steps and 26% overall yield.
ethyl ester4,5 as starting materials (Scheme 2). In the described
procedure, ethyl acetoacetate was condensed with N,N-
INTRODUCTION
Pyrrolotriazine 1 was identified as a lead candidate for efficient
■
dimethylformamide dimethylacetal in refluxing toluene to
p38α kinase inhibition and was chosen for development as a
generate a dimethyl enamine intermediate 9a. Isolation of the
potential therapy for rheumatoid arthritis and other inflamma-
tory diseases.1
dimethyl enamine and treatment with glycine ethyl ester HCl
and acetic acid at 100 °C afforded 9 in 49% yield from ethyl
In the initial Discovery Chemistry synthesis of 1 (Scheme 1),
acetoacetate.4b Our development work led to the use of tosic
condensation of ethyl isocyanoacetate with acetaldehyde under
basic conditions afforded pyrrole 2.1−3 Subsequent N-
amination of 2 with O-(2,4-dinitrophenyl)hydroxylamine
provided aminopyrrole 3, which reacted with formamide to
complete the pyrrolotriazine ring system to produce 4.
Following chlorination of 4 and coupling with aniline 6,
hydrolysis and amidation with ethylamine completed the
synthesis of the API (1).
The Discovery synthesis was successful in producing the
initial quantities of 1 in good yield. However, further
development was necessary to reduce the cost of goods,
address safety concerns, and overcome challenging scale-up
issues prior to multikilogram preparations. While finding an
alternative to the toxic and expensive starting material ethyl
isocyanoacetate was important, the key for a viable multikilo-
gram-scale process was to identify safer and more practical
reaction conditions for the N-amination of 2. We also sought to
address the elevated reaction temperature (160 °C) required
for condensation of 3 with formamide, as well as the use of neat
POCl3 as solvent for chlorination of 4. In this report, we
describe a successful development program that overcame these
challenges and provided a practical process for implementation
of the route on the multikilogram scale.
acid as catalyst for the condensation under neat conditions at
55 °C. In addition, the condensation product 9a was not
isolated and the enamine exchange with glycine ethyl ester HCl
was conducted in ethanol at 50 °C. This optimized process
reproducibly provided a 78−82% yield of enamine 9 from ethyl
acetoacetate.
Cyclization of 9 under literature conditions (NaOEt/EtOH)5
produced 2 in only 40% yield with 90% HPLC area purity. A
screen of reaction conditions indicated that the cyclization
required polar protic solvents and the presence of strong bases,
such as NaOEt potassium tert-amylate or KOBut at 60 °C. We
proposed a mechanism (Scheme 3) in which the initial
cyclization to form intermediate 12 was reversible. Protic
solvents facilitate protonation of 12 and lead to the irreversible
elimination of water to generate pyrrole 2. Under these
conditions, impurities 10, 14a, and 14b were observed at
combined levels up to 40 mol %. We hypothesized that the 1
equiv of water generated during the cyclization causes the
formation of impurities 10, 14a, and 14b via hydrolysis of 2 and
9. A variety of water scavenging reagents and additives, such as
molecular sieves, anhydrous Na2SO4, TFAA, TsCl, Tf2O,
TMSOTf, and ethyl trifluoroacetate, were investigated. Ethyl
trifluoroacetate was found to be the most effective additive and
minimized the hydrolysis impurities to 10 mol % in the
completed reaction mixture. It appears that ethyl trifluoroace-
tate acts as a sacrificial ester, being more reactive toward water
than the ester groups of 9 and 2. After the addition of 2 equiv
of base to sequester any trifluoroacetic acid generated, pyrrole 2
RESULTS AND DISCUSSIONS
■
I. Preparation of Pyrrole 2. The initial synthesis of 2 via
condensation of ethyl isocyanoacetate with acetaldehyde
afforded the desired product in an acceptable yield (66%).
However, the isocyanoacetate was both costly and toxic, and
isolation of the product required chromatography. As an
alternative, we initially investigated a slightly longer preparation
which utilized readily available ethyl acetoacetate and glycine
Received: July 6, 2012
Published: August 31, 2012
© 2012 American Chemical Society
1618
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
Scheme 1. Discovery Chemistry Synthesis of 1
Scheme 2. Alternative Synthesis of Pyrrole 2
Scheme 3. Proposed Mechanism for Formation of Pyrrole 2
was isolated as a yellow solid in 60−70% yield in good quality
(96% HPLC area purity) on up to 12 kg input scale.
under the reaction conditions. The cleanest reaction profile was
achieved when treating a solution of pyrrole 2 in DMF or
DMAc in presence of 2 equiv of KOBut with 1.4−2 equiv of an
anhydrous solution of chloramine in MTBE.9 In the presence
of >2 equiv of the base, aminopyrrole 3 reverted to pyrrole 2.
When using a higher excess of chloramine, 3 degraded to form
2-chloropyrrole 3a (Figure 1). The same degradation product
(3a) was found when treating 3 with aqueous HCl.10
II. Amination of Pyrrole 2 to Amino Pyrrole 3.
Chloramine Amination. The efficient annulation of a triazine
ring onto the pyrrole 2 required N-amination followed by
condensation with formamide (Scheme 1). Although the
Discovery Chemistry reaction of 2 with O-(2,4-dinitrophenyl)-
hydroxylamine provided 3 in good yields, concerns regarding
the detonation potential6 as well as the toxicity of the resulting
2,4-dinitrophenol byproduct required consideration. In addi-
tion, the use of NaH in DMF is hazardous.7 There are not
many N-amination conditions that lend themselves to the
safety, cost, and scalability requirements of a viable multikilo-
gram-scale process. Thus, work focused on identifying
alternative reagents to fulfill these requirements.
Employing the optimized reaction conditions, 99% HPLC
area in-process purity and 86% isolated yield of 3 could be
achieved. However, due to the limited miscibility of the DMF
Chloramine (NH2Cl) had been found earlier to be an
effective reagent for the amination of 2.8 For the initial scale-up
batches, the amination reaction was carefully optimized to
enhance safety, ensure consistent performance, and minimize
the competing degradation of the aminopyrrole product 3
Figure 1. Structure of aminopyrrole degradation product 3a.
1619
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
and MTBE phases in the reaction, the reaction was mass
transfer controlled and yields declined on scale up due to
decreasing surface/volume ratios and less efficient mixing.
In order to overcome the mass transfer issue as well as
increase volume efficiency, we subsequently developed a
process to aminate 2 using gaseous chloramine, generated
from chlorine and ammonia in a continuous fashion.11 This
new process was successfully employed to prepare amino-
pyrrole 3 in a 74% yield at up to a 2 kg scale. However, due to
the complexity of the setup and safety concerns around the use
of chlorine gas, we continued to seek more robust amination
reagents and conditions.12
hydroxylamine 15a was stable at room temperature in its
crystalline form for up to six months and at 0 °C for more than
two years. This provided a reasonable shelf life for storage and
use in manufacturing. The thermal stability analysis of 15a
displayed exothermic events with onset temperatures at 100 °C
(641 J/g) and 162 °C (657 J/g), which offered a reasonable
margin of safety if the amination was executed at 25−35 °C.
Based on these results, we concluded that 15a could be
considered for use on multikilogram-scale and demonstrated its
applicability to the N-amination of a variety of heterocyclic
substrates.19
An extensive solvent and base screen revealed that the
counterion of the base had a profound impact on the amination
of pyrrole 2 with the hydroxylamine 15a (Table 1). Generally,
the reactivity increased with less complexing counterions. For
example, while lithium bis(trimethylsilyl)amide led to only 6%
conversion in THF, sodium bis(trimethylsilyl)amide yielded
21% and potassium bis(trimethylsilyl)amide gave 76% con-
version to the desired aminopyrrole 3 (entries 1−3). The same
trend was observed in THF 3:7 NMP mixtures (entries 5−7).
A major byproduct, particularly in the presence of lithium, was
p-nitrobenzohydroxamic acid (16).20 We hypothesized that
complexation of 15a by metal cations facilitated isomerization
of the hydroxylamine to the hydroxamic acid 16. The resulting
decomposition of the aminating reagent coupled with the
protonation of the pyrrole anion 2a by 16 led to low yields in
the presence of strongly complexing cations such as lithium.
Higher yields were achieved in the more polar NMP/THF
solvent system than in 100% THF in accordance with our
hypothesis, as the metal-complexation would be stronger in
100% THF. Moisture also had a significant impact on the
reaction. Lower conversion (84% vs 92%) was observed in the
presence of more water, which was probably due to competitive
reaction of water with the aminating agent.19 The bases
KHMDS and KOBut in NMP 7:3 THF provided the best
conversion (92%).
O-(4-Nitrobenzoyl)hydroxylamine Amination. Amination
of heterocycles has been reported using a variety of
+
hydroxylamine-based electrophilic NH2 equivalents, such as
O-acyl,13 O-alkyl,14 O-sulfonyl,15 O-nitrophenyl,16 and O-
diarylphosphinyl17 derivatives. Unfortunately, most of these
reagents are not commercially available, or they are thermally
instable. Only hydroxylamine−O-sulfonic acid (HOSA) has
been used on-scale.18 However, HOSA was not a suitable
reagent, since the electron-deficient pyrrole 2 was too
unreactive to provide more than 5−10% conversion in spite
of extensive experimentation.
+
Friestad and co-workers reported the NH2 transfer to the
nitrogen of a cyclic carbamate through the use of O-(4-
nitrobenzoyl)hydroxylamine 15a (Scheme 4).13a This reagent
Scheme 4. Preparation of O-(4-Nitrobenzoyl)hydroxylamine
15a
appeared promising, and we were very pleased to observe
efficient amination of 2 with 15a. Since 15a was not readily
available in large quantities, we developed a process based on
literature methods13b,14a that provided 15a in a single step from
p-nitrobenzoyl chloride and N-Boc-hydroxylamine in 80% yield
and 99% purity. Our studies also indicated O-(4-nitrobenzoyl)-
During further development of the amination process, we
noticed the presence of impurity benzoylamide 17 that
persisted at 2−5% HPLC relative area purity (Scheme 5).
Crystallization of aminopyrrole 3 was problematic, and thus, 17
had to be carried over into the subsequent formation of
a
Table 1. Optimization of Amination Reaction
b
entry
base (1.1 equiv)
solvent syst
THF
react. time
conv
hydroxamic acid 16
1
LiHMDS
NaHMDS
KHMDS
KOBut
2 h
3 h
2 h
2 h
2 h
2 h
2 h
2 h
2 h
6%
21%
76%
75%
63%
79%
92%
92%
55%
42%
9%
2
THF
3
THF
4
THF
11%
21%
10%
4%
5
LiHMDS
NaHMDS
KHMDS
KOBut
NMP-THF (7:3)
NMP-THF (7:3)
NMP-THF (7:3)
NMP-THF (7:3)
NMP-THF (7:3)
NMP
6
7
8
2%
c
9
KOBut
84%
33%
85%
8%
d
10
11
K2CO3
70 h
1%
d
Cs2CO3
NMP
20 h
1%
a
b
c
d
Reactions were run with 1.2 equiv of 15a. By HPLC analysis. 1 equiv of water added. Slow reaction, probably due to solubility issue of base.
1620
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
Scheme 5. Conversion of 17 to 18
Scheme 6. Proposed Mechanism of the Formation of Impurity 17
pyrrolotriazine 4. It was shown that 17 transformed in the
subsequent condensation with formamide into arylated
pyrrolotriazinone 18. Due to the structural similarity of 18
with pyrrolotriazine 4, the arylated pyrrolotriazinone 18 and its
resulting impurities could not be efficiently purged from any of
the intermediates downstream. This made identification of
conditions to minimize the level of impurity 17 critical to the
successful production of high quality API.
Control experiments indicated that 17 was not generated by
the reaction of aminopyrrole 3 with O-(4-nitrobenzoyl)-
hydroxylamine 15a. Significantly, a reverse addition of the
pyrrole anion 2a to reagent 15a afforded a reduced level of 17
to 1−2%. Monitoring the impurity profile over time also
indicated that 17 was generated continuously from the
beginning of the amination reaction. Based on these
observations, we developed a mechanistic proposal (Scheme
6). We hypothesized that, under strong basic conditions, 15a
formed low levels of 3-hydroxyl-oxaziridine 19. Nucleophilic
attack by the pyrrole anion 2a and ring-opening of 19 is the
source of benzoylamide 17.
Based on the proposed mechanism, we rationalized that
benzoyl hydroxylamine reagents with electron donating
substituents should be less prone to generate the corresponding
3-hydroxyl-oxaziridines and accordingly result in less benzoy-
lamide byproducts in the amination reaction. To verify this
assumption, we prepared a series of substituted O-benzoyl
hydroxylamines 15b−f19 and determined the levels of the
corresponding benzoylamides 17b−f in the amination reaction
(Table 2). In accordance with our hypothesis, we found that
the levels of the benzoylamide byproduct decreased with more
electron-donating aromatic substituents. O-(4-Methoxy)-
hydroxylamine (15f) proved to be the most effective amination
reagent, reducing the level of byproduct 17f to 0.3 HPLC area
percent purity with excellent conversion (97%).
In parallel with these approaches, we also pursued options to
purge impurity 17a from the reaction stream by selective
reactivity. We found that 17a could be reduced to the related
aniline derivative by treating the amination reaction mixture
with aqueous sodium dithionate solution. The corresponding
aniline derivative stayed in the aqueous layer during extractive
workup and was removed from aminopyrrole 3. After
incorporating this protocol into the amination/cyclization
sequence, the levels of impurity 18 in 4 could be reduced to
<0.05% HPLC area purity.
Having identified two approaches to control the level of
benzoylamides 17, we ultimately chose to pursue the extraction
approach for scale-up due to the low melting point (35 °C) of
reagent 15f.
III. Cyclization to Pyrrolotriazine 4. Optimization work
in the subsequent condensation of 3 with formamide to form
triazinone 4 focused on lowering the reaction temperature (160
°C). Screening a series of acids22 revealed that acetic acid,
formic acid, or phosphoric acid was an effective catalyst,
affording complete conversion in 24 h at 120 °C. An added
benefit was pyrrolotriazine 4 was now being obtained as a light
yellow solid in 91% yield rather than the previously obtained
dark solid. Formic acid was not further developed due to
concerns about stability at high temperatures.23 The use of
acetic acid produced the methylated impurity 20 at 1−3%
(Figure 2).24 Since impurity 20 was difficult to remove from the
Table 2. Results of Cyclization of Substituted O-Benzoyl
Hydroxylamines
substituent const
conv
a
byproduct 17a−f
b
reagent
R-
(σ)21
(%)
(%)
15a
15b
15c
15d
15e
15f
4-NO2
4-F
0.78
0.06
92
93
95
96
95
97
4.5
1.7
1.5
1.1
0.6
0.3
Figure 2. Impurity 20 from acetic acid conditions.
3-Cl
0.37
4-Br
0.23
desired product 4, we selected phosphoric acid (0.2 equiv) as
catalyst for the cyclization reaction (20 h at 120 °C). The
optimized amination−cyclization process was smoothly scaled
up to 8 kg scale and produced 4 in 78−82% yield with >98%
HPLC area purity.
4-Cl
0.23
4-OMe
−0.27
a
b
HPLC % conversion after 0.5−3.0 h at room temperature. HPLC
area percent.
1621
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
Scheme 7. Multikilogram-Scale Synthesis of 1
IV. Chlorination of Pyrrolotriazine 4 and Coupling
with Aniline 6. For the Discovery route, the chlorination of
pyrrolotriazine 4 was performed in neat POCl3 at 100 °C. The
subsequent highly exothermic quench of the reaction with base
led to hydrolysis of ∼10% of chloride 5 back to 4. To address
this issue, we screened various reaction conditions and solvents
to reduce the amount of POCl3 required. By using toluene as
the solvent, POCl3 could be reduced to 3−4 equiv. Addition of
1 equiv of DIPEA further enabled reduction of POCl3 to only
1.2 equiv and decreased the reaction time from >30 h to ∼16 h.
The optimized reaction conditions also controlled the
hydrolysis of chloride 5 to <2%. Due to the limited stability
of isolated chloride 5 above −10 °C, we decided to telescope
the chlorination into the subsequent amide coupling. This was
done by simply performing a solvent exchange from toluene to
DMF after the workup.
Hydrolysis of chloride 5 remained a concern during the
coupling with aniline 6. Under the initial reaction conditions,
without addition of base, about 15% of 5 would hydrolyze to 4,
which was probably due to the presence of HCl generated in
the reaction to accelerate the hydrolysis. Since the process can
only tolerate 5% of 4 for isolation of 7, purging of 4 required
several successive crystallizations from DMF−water and
ethanol, followed by reslurrying in IPA or MeCN. However,
with addition of 1 equiv of DIPEA to the coupling mixture to
neutralize the released HCl from reaction, hydrolysis of 5 was
reduced to ∼3% and the product 7 could be crystallized directly
from the reaction mixture by the addition of water.
6.0−7.0, 8 was obtained as a fast-filtering granular solid with a
density of 2.0 cm3/g in 88% yield and 98.9% HPLC area purity.
Finally, the resulting acid was converted to the drug
substance (1) by coupling with ethylamine−HCl using
HOBT/EDAC-HCl in DMF.25 The API (1) could be directly
crystallized from the reaction mixture by addition of water. The
coupling process was scaled to deliver 1.6 kg of 1 in a single
batch in 88% isolated yield and >99.0% HPLC area purity.
CONCLUSIONS
■
A practical and safe synthesis of 1 has been developed (Scheme
7). Improvements to the initial process included demonstration
of a new, safe, and robust N-amination method for pyrrole 2
and optimization of the subsequent reactions for implementa-
tion in scale-up facilities. Detailed studies were conducted in
order to understand and minimize the formation of impurities
in the amination process. In addition, a more cost-effective and
less toxic alternative synthesis of pyrrole 2 from ethyl
acetoacetate was developed. The optimized processes were
executed at multikilogram scale to prepare a total of 1.6 kg of 1
in a 26% overall yield.
EXPERIMENTAL SECTION
■
All reagents purchased from vendors were used as received
unless otherwise indicated. Reported yields have not been
corrected for impurity levels or moisture content. HPLC
method: Waters XTerra RP-18, 3.5 μm, 4.6 mm × 50 mm, 10%
water/acetonitrile to 0.2% H3PO4/water, 2.5 mL/min, 8 min
gradient, 256 nm with retention time of 5.3 min for 2, 5.1 min
for 3, and 2.2 min for 16.
The preparation of intermediate 7 from 5 was successfully
executed by this telescoped process in 88% yield (>98% HPLC
area purity) on a 2.2 kg scale.
Ethyl 2-(((2-Ethoxy-2-oxoethyl)amino)methylene)-3-
oxobutanoate (9). A mixture of ethyl acetoacetate (115.5
kg), N,N-dimethylformamide dimethyl acetal (110.0 kg), and
toluenesulfonic acid monohydrate (0.231 kg) was stirred at
55−60 °C for 1 h and then cooled to 45 °C over 1 h. Glycine
ethyl ester hydrochloride (123.2 kg) and ethanol (229.3 kg)
were added to the above solution, which was then stirred at
45−53 °C for 0.5 h. The reaction mixture was then cooled over
1 h to 20−26 °C and held at 20−26 °C for 2 h. Water (231 kg)
was added over 30 min at 20−35 °C (mild exotherm).
V. Hydrolysis of 7 to Acid 8 and Subsequent Coupling
with Ethylamine. Hydrolysis of 7 was accomplished by
treatment with NaOH at 60−65 °C. Upon neutralization, 8
crystallized directly from the reaction mixture. The selection of
acid greatly impacted the ease of filtration. Adding 1 N HCl
directly to the basic reaction mixture provided 8 as very fine
particles that were slow to filter. In contrast, if the reaction
mixture was first diluted with one equal volume of water, and
then neutralized by slow addition of 1 N acetic acid to a pH
1622
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
Additional water (1155 kg) was added at 30−35 °C over 2 h to
give a thick slurry, which was then cooled to 20 °C over 1 h and
held for 2 h. The solid product was collected by filtration. The
wet cake was washed with water (1200 kg) and heptane (300
kg) and dried at 35−40 °C under vacuum to give 158.4 kg
(78%) of 9 as a slightly yellow solid (known compound in
literature). mp 70.0−72.0 °C.4b Elemental Analysis: Calcd for
C11H17NO5: C, 54.31; H, 7.04; N, 5.76. Found: C, 54.18; H,
7.13; N, 5.58.
Diethyl 3-Methyl-1H-pyrrole-2,4-dicarboxylate (2). A
solution of potassium tert-amylate (1.7 M in toluene, 61.3 L,
53.9 kg, 104 mol) was added over 30 min to a solution of 9
(12.0 kg) and CF3COOEt (6.8 kg) in toluene (13.0 kg)
containing ethanol (25.0 kg) at <27 °C (mild exotherm). The
solution was then heated to 60−63 °C over 1 h and stirred at
60−63 °C for 9 h. The reaction mixture was cooled to 15−20
°C over 1 h and quenched with HOAc (4.2 kg) at <25 °C. The
resulting mixture was treated with water (64 L). The organic
layer was separated and washed with aqueous K2HPO4 (6%, 80
kg). The organic solution was filtered through a Zeta Pad
(ZP04701-R52SP) which was then rinsed with toluene (10 kg).
The filtrate was concentrated to ∼20 L at 35−45° under
vacuum. After the reaction mixture was cooled to ∼25 °C over
1 h, heptane (68 kg) was added over 20 min for crystallization.
The resulting slurry was stirred at 20 °C for 2 h. The solid
product was collected by filtration, washed with heptane (21
kg), and dried under vacuum at 20−30 °C for one day to give
7.5 kg (67%, 95% HPLC area purity) 2 (known compound in
the literature) as a yellow solid. mp 89.0−91.0 °C.4a Elemental
Analysis: Calcd for C11H15NO4: C, 58.66; H, 6.71; N, 6.22.
Found: C, 58.43; H, 6.82; N, 6.37.
(5.5 kg) was added to a solution of 2 (10 kg) in NMP (60
L), which was stirred at 20−37 °C until all KOBut had
dissolved (mild exotherm). After the reaction mixture was
cooled to 29 °C, a solution of O-(4-nitrobenzoyl)-
hydroxylamine 15a (8.1 kg) in dry THF (30 L, KF < 0.1%)
was added over 20−30 min at 28−32 °C. The reaction mixture
was cooled to 20−25 °C over 90 min. A solution of sodium
dithionite (6.7 kg) in water (93.3 L) was added to the reaction
mixture at 20−35 °C (mild exotherm), which was then stirred
at ca. 30 °C for 30 min. Toluene (50 L) was added to the
reaction mixture, which was stirred thoroughly. The organic
phase was separated, and the aqueous layer was further
extracted with toluene (2 × 50 L). The combined organic
phases were washed with a 5% sodium hydrogen carbonate
solution (60 L) followed by water (60 L). The organic solution
was concentrated by distillation at 40−60 °C under vacuum to
10 L. Formamide (80 L) was added to the residue, which was
further concentrated at 60−80 °C under vacuum to remove
residual toluene (1−2 L). Phosphoric acid (85%, 0.85 kg) was
added to the solution, which was then heated to 120−125 °C
for 20 h. The reaction mixture was then cooled to 90 °C over 1
h. Water (100 L) was added over 30 min to give a slurry while
cooling to 20−25 °C over 2 h. After stirring for 2 h, the solid
product was collected by filtration and washed with water (2 ×
40 L), methanol (30 L), and toluene (30.0 L). The wet cake
was dried under vacuum at 30−40 °C for two days to give 8.0
kg (79%) of 4 (known compound in the literature). mp 215 °C
(dec.).1 1H NMR (DMSO-d6, 400 MHz) 11.65 (br s, 1 H),
7.87 (s, 1 H), 7.85 (s, 1 H), 4.24 (q, 2 H, J = 7.0 Hz), 2.61 (s, 3
H), 1.30 (t, 3 H, J = 7.0 Hz). 13C NMR (DMSO-d6, 100 MHz)
163.4, 155.1, 140.0, 123.2, 122.9, 117.7, 114.0, 59.6, 14.2, 11.0.
Ethyl 4-((5-(Methoxycarbamoyl)-2-methylphenyl)-
amino)-5-methylpyrrolo[2,1-f ][1,2,4]triazine-6-carboxy-
late (7). In the following order, pyrrolotriazine 4 (1.54 kg),
toluene (24.4 kg), phosphorous oxychloride (POCl3, 1.3 kg),
and N,N-diisopropylethylamine (1.1 kg) were charged to a
reactor and the reaction mixture was stirred at 110 °C for 22 h.
The resulting mixture was cooled to 2 °C over 3 h and
quenched with cold 18% potassium dibasic phosphate solution
(29.4 kg). The quenched biphasic mixture was allowed to settle
at room temperature for 1 h. The organic layer was separated
and washed with 11% aqueous potassium dibasic phosphate
solution (8.0 kg). The organic layer was polish filtered through
a 10 μm cuno filter. DMF (8.1 kg) was added to the solution,
which was then concentrated under vacuum at 49−53 °C to 9
L and then cooled to 25 °C. A solution of 6 (1.56 kg) in DMF
(2.8 kg) containing N,N-diisopropylethylamine (0.9 kg) was
added to the above solution. The mixture was heated to 58 °C.
After 25 min, two additional portions of N,N-diisopropylethyl-
amine (0.3 kg each) were added at 30 min intervals. After
stirring for an additional 2 h, the reaction mixture was cooled to
20 °C over 2 h. A 9% aqueous potassium dibasic phosphate
solution (14.7 kg) was added to give a slurry which was stirred
for 20 h. The product was collected by filtration and wash with
water (10.5 kg), acetonitrile (3.3 kg), and toluene (6.1 kg). The
wet product was dried under vacuum at 30 °C for two days to
give 2.2 kg (82%, 98.5% HPLC area purity) dry product
(known compound in the literature1) as a slightly yellow solid.
Elemental Analysis: Calcd for C19H21N5O4: C, 59.52; H, 5.52;
N, 18.27. Found: C, 59.18; H, 5.38; N, 18.23.
O-4-Nitrobenzoylhydroxylamine (15a). This is an
improved process based on a literature preparation of the
phenyl analogue.13b Triethylamine (8.0 kg, 402 mol) was added
to a solution of N-Boc-hydroxyamine (10.0 kg, 75.1 mol) in
dichloromethane (104 L) at room temperature. The solution
was cooled to −10 °C over 1 h. A solution of 4-nitrobenzoyl
chloride (13.9 kg, 75.1 mol) in dichloromethane (44 L) was
added over 2 h at −10 to 0 °C (mild exotherm). The mixture
was then stirred at −5 to 0 °C for 60 min. Water (75 L) was
added to quench the reaction at <15 °C (mild exotherm). After
the reaction mixture was stirred for 15 min, the organic layer
was separated. The aqueous layer was extracted with dichloro-
methane (20 L). The combined organic layers were washed
with aqueous K2HPO4 (4%, 40 L) and then treated with
methanesulfonic acid (10.5 kg, 109 mol) at 25 °C for 5 h. The
reaction solution was treated with aqueous K2HPO4 (20%, 124
kg) at 25 °C for 5−10 min. THF (100 L) was added. The
organic layer was separated, washed with water (50 L), and
then concentrated under vacuum at <30 °C to a final volume of
40 L. Heptane (60 L) was added over 2 h to the mixture for
crystallization. After the reaction mixture was stirred at 20 °C
for 3 h, the solid was collected by filtration, washed with THF−
heptane (1:2, 2 × 120 L), and dried under vacuum at 25 °C for
one day to give product 15a (11.4 kg, 83%) as a slightly yellow
solid. mp 85.0 °C (dec., lit,13c 111.5 °C). H NMR (CDCl3,
1
400 MHz) δ 8.32 (d, 2H, J = 8.4 Hz), 8.21 (d, 2H, J = 8.4 Hz),
6.76 (s, 2H, NH2). 13C NMR (CDCl3, 100 MHz) δ 165.5,
150.8, 133.3, 130.6 (2C), 123.7 (2C). Elemental Analysis:
Calcd for C7H6N2O4: C, 46.16; H, 3.32; N, 15.38. Found: C,
46.15; H, 3.14; N, 15.34.
Ethyl 4-Hydroxy-5-methylpyrrolo[2,1-f ][1,2,4]-
triazine-6-carboxylate (4). Solid potassium tert-butoxide
4-((5-(Methoxycarbamoyl)-2-methylphenyl)amino)-5-
methylpyrrolo[2,1-f ][1,2,4]triazine-6-carboxylic acid 8.
A mixture of ester 7 (2.06 kg), water (6.0 kg), and 50% NaOH
1623
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
(1.15 kg) was heated to 60−65 °C over 1 h and then held at
60−65 °C for 5 h. Additional water (7.2 kg) was added at 60−
65 °C. A 6 wt % aqueous solution (6.6 kg) of acetic acid was
added over 2 h to the solution, which was then held for 40 min.
Additional HOAc solution (4.9 kg) was added over 2 h. The
resulting slurry was cooled to 25 °C over 2 h and then filtered.
The wet cake was washed with water (10 kg) and then heptane
(8.0 kg). The wet cake was dried at 70−75 °C under vacuum
for two days to give 1.68 kg (88%) of 8 (known compound in
the literature1) as an off white solid. Elemental Analysis: Calcd
for C17H17N5O4: C, 57.46; H, 4.82; N, 19.71. Found: C, 57.29;
H, 4.76; N, 19.53.
N-Ethyl 4-((5-(Methoxycarbamoyl)-2-methylphenyl)-
amino)-5-methylpyrrolo[2,1-f ][1,2,4]triazine-6-carboxa-
mide (1). A mixture of acid 8 (1.66 kg), EDAC (1.076 kg), and
HOBt25 (0.76 kg, CAUTION: HOBt is potentially explosive.
Please see references for details) in DMF (7.15 kg) was stirred
at 20−25 °C for 3 h. Aqueous 70 wt % ethylamine (0.603 kg)
was added to the mixture over 1.5 h at <35 °C. The reaction
mixture was stirred at 30−35 °C for 1 h. Water (7.0 kg) was
added over 3.5 h to the mixture to give a slurry, which was then
cooled to 25 °C over 30 min. After 2 h, the product was
filtered, washed with water (8.0 kg) and acetone (6.4 kg),
respectively, and dried under vacuum at 25 °C for 2.5 days to
afford 1.57 kg (88%) of 1 (known compound in the literature1)
as an off white solid. Calcd for C17H22N6O3: C, 59.67; H, 5.80;
N, 21.98. Found: C, 59.33; H, 5.64; N, 21.67.
General Method for Preparation of Diethyl 1-(3-
Chlorobenzamido)-3-methyl-1H-pyrrole-2,4-dicarboxy-
late 17a−f from 3. A solution of m-chlorobenzoyl chloride
(0.48 g, 2.74 mmol) in toluene (5 mL) was added to a solution
of 3 (0.76 g, 80 wt %, 2.53 mmol) in toluene (15 mL) at 20−25
°C. After 15 min, a solution of N,N-diisopropylethylamine
(0.48 mL, 0.35 g, 2.7 mmol) in toluene (5 mL) was added over
10 min to the mixture at 17−25 °C. The reaction mixture was
stirred at 20−25 °C for 30 min. The reaction was quenched
with water (15 mL). The organic layer was separated, washed
with 5% aqueous sodium bicarbonate (15 mL) and then water
(15 mL), and dried over anhydrous Na2SO4 (5 g). The solids
were removed by filtration and washed with toluene (10 mL).
Concentration of the filtrate under vacuum gave the crude
product. This residue was dissolved in MTBE (20 mL), and
heptane (15 mL) was added. After the reaction mixture was
stirred at 20 °C for 2 h, the resulting crystals were collected by
filtration, washed with 1:1 heptane/MTBE (5 mL), and dried
under high vacuum at 35 °C for 15 h to give 17c (R = 3-Cl) as
a slightly yellow solid (0.85 g, 79%). mp 120−121 °C. 1H NMR
(CDCl3, 400 MHz) δ 10.18 (s, 1 H), 7.91 (t, 1 H, J = 1.7 Hz),
7.77 (ddd, 1 H, J = 7.9, 1.7, 1.5 Hz), 7.73 (s, 1 H), 7.56 (ddd, 1
H, J = 7.9, 1.7, 1.5 Hz), 7.43 (t, 1 H, J = 7.9 Hz), 4.29 (q, 2 H, J
= 7.0 Hz), 4.27 (q, 2 H, J = 7.0 Hz), 2.60 (s, 3 H), 1.34 (t, 3 H,
J = 7.0 Hz), 1.33 (t, 3 H, J = 7.0 Hz). 13C NMR (CDCl3, 100
MHz) δ 165.7, 163.9, 162.1, 135.2, 133.1, 132.9, 132.2, 130.6,
130.2, 127.9, 125.5, 118.9, 113.2, 60.8, 59.9, 14.3, 14.2, 11.8.
HR-MS: Calcd for C18H19ClN2O5 + H: 379.1061. Found
379.1062. IR (KBr, cm−1) 3310, 2984, 1721, 1682, 1525, 1419,
1264, 1233, 1079, 775.
100 MHz) 164.9, 163.9, 162.2, 150.3, 136.8, 131.9, 130.6, 128.8
(2 C), 124.1 (2 C), 118.7, 113.3, 60.9, 60.0, 14.3, 14.2, 11.8.
HR-MS: Calcd for C18H19N3O7 + NH4: 407.1567. Found
407.1560. IR (KBr, cm−1) 3308, 2991, 1719, 1682, 1529, 1420,
1349, 1268, 1240, 1081, 852, 774.
Diethyl 1-(4-Fluorobenzamido)-3-methyl-1H-pyrrole-
2,4-dicarboxylate (17b). (R = 4-F): 74%. mp 119−120 °C.
1H NMR (CDCl3, 400 MHz) 10.17 (s, 1 H), 7.95 (dd, 2 H, J =
8.7, 5.3 Hz), 7.75 (s, 1 H), 7.18 (dd, 2 H, J = 8.7, 8.5 Hz), 4.28
(q, 2 H, J = 7.0 Hz), 4.27 (q, 2 H, J = 7.0 Hz), 2.61 (s, 3 H),
1.33 (t, 3 H, J = 7.0 Hz), 1.32 (t, 3 H, J = 7.0 Hz). 13C NMR
(CDCl3, 100 MHz) 166.0, 165.6 (d, 1 C, J = 255 Hz), 164.0,
162.3, 132.2, 130.7, 130.1 (d, 2 C, J = 10 Hz), 127.5, 118.9,
116.2 (d, 2 C, J = 22 Hz), 113.1, 60.8, 59.9, 14.4, 14.2, 11.9.
HR-MS: Calcd for C18H19FN2O5 + H: 363.1356. Found
363.1359. IR (KBr, cm−1) 3329, 2985, 1716, 1686, 1605, 1504,
1421, 1271, 1230, 1078, 847, 775.
Diethyl 1-(4-Chlorobenzamido)-3-methyl-1H-pyrrole-
2,4-dicarboxylate (17e). (R = 4-Cl): 68%. mp 133−134 °C.
1H NMR (CDCl3, 400 MHz) 10.22 (s, 1 H), 7.86 (d, 2 H, J =
8.6 Hz), 7.75 (s, 1 H), 7.47 (d, 2 H, J = 8.6 Hz), 4.28 (q, 2 H, J
= 7.2 Hz), 4.27 (q, 2 H, J = 7.2 Hz), 2.61 (s, 3 H), 1.34 (t, 3 H,
J = 7.2 Hz), 1.33 (t, 3 H, J = 7.2 Hz). 13C NMR (CDCl3, 100
MHz) 166.0, 163.9, 162.2, 139.3, 132.2, 130.7, 129.7, 129.9 (2
C), 128.9 (2 C), 118.8, 113.1, 60.8, 59.9, 14.4, 14.2, 11.8.
Elemental Analysis: Calcd for C18H19ClN2O5: C, 57.07; H,
5.05; N, 7.39; Cl, 9.35. Found: C, 57.19; H, 4.88; N, 7.34; Cl,
9.21. IR (KBr, cm−1) 3307, 2983, 1717, 1689, 1677, 1555, 1418,
1264, 1236, 1078, 775.
Diethyl 1-(4-Bromobenzamido)-3-methyl-1H-pyrrole-
2,4-dicarboxylate (17d). (R = 4-Br): 50%. mp 141−142 °C.
1H NMR (CDCl3, 400 MHz) 10.22 (s, 1 H), 7.79 (d, 2 H, J =
8.6 Hz), 7.76 (s, 1 H), 7.64 (d, 2 H, J = 8.6 Hz), 4.29 (q, 2 H, J
= 7.2 Hz), 4.28 (q, 2 H, J = 7.2 Hz), 2.61 (s, 3 H), 1.34 (t, 3 H,
J = 7.2 Hz), 1.33 (t, 3 H, J = 7.2 Hz). 13C NMR (CDCl3, 100
MHz) 166.1, 164.0, 162.3, 132.3 (2 C), 131.9, 130.7, 130.2,
129.1 (2 C), 127.9, 118.8, 113.1, 60.9, 59.9, 14.4, 14.2, 11.9.
HR-MS: Calcd for C18H19BrN2O5 + H: 423.0556. Found
423.0542. IR (KBr, cm−1) 3299, 2982, 1716, 1689, 1592, 1555,
1529, 1418, 1264, 1235, 1078, 1011, 775.
Diethyl 1-(4-Methoxybenzamido)-3-methyl-1H-pyr-
role-2,4-dicarboxylate (17f). (R = 4-OMe): 57%. mp
1
112−113 °C. H NMR (CDCl3, 400 MHz) 10.08 (s, 1 H),
7.89 (d, 2 H, J = 8.8 Hz), 7.73 (s, 1 H), 6.97 (d, 2 H, J = 8.8
Hz), 4.27 (q, 4 H, J = 7.0 Hz), 3.87 (s, 3 H), 2.61 (s, 3 H), 1.33
(t, 3 H, J = 7.0 Hz), 1.31 (t, 3 H, J = 7.0 Hz). 13C NMR
(CDCl3, 100 MHz) 166.6, 164.0, 163.3, 162.1, 132.5, 130.6,
129.5 (2 C), 123.4, 119.0, 114.2 (2 C), 112.8, 60.7, 59.8, 55.5,
14.4, 14.2, 11.8. HR-MS: Calcd for C19H22N2O6 + H: 375.1556.
Found 375.1560. IR (KBr, cm−1) 3361, 2985, 1721, 1686, 1674,
1609, 1506, 1419, 1263, 1231, 1174, 1077, 843, 774.
AUTHOR INFORMATION
Corresponding Author
*E-mail: zhongping.shi@bms.com.
■
Notes
The authors declare no competing financial interest.
Diethyl 1-(4-Nitrobenzamido)-3-methyl-1H-pyrrole-
2,4-dicarboxylate (17a). (R = 4-NO2): 82%. mp 129−130
°C. 1H NMR (CDCl3, 400 MHz) 10.47 (s, 1 H), 8.35 (d, 2 H, J
= 8.8 Hz), 8.11 (d, 2 H, J = 8.8 Hz), 7.79 (s, 1 H), 4.30 (q, 2 H,
J = 7.2 Hz), 4.28 (q, 2 H, J = 7.2 Hz), 2.60 (s, 3 H), 1.35 (t, 3
H, J = 7.2 Hz), 1.34 (t, 3 H, J = 7.2 Hz). 13C NMR (CDCl3,
ACKNOWLEDGMENTS
■
The authors wish to thank the BMS pilot plant operations and
Analytical R & D staff for their valuable support during our
scale up runs. In addition, we thank Dr. Jaan A. Pesti and Dr.
1624
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625

Organic Process Research & Development
Article
Dev. 1999, 3 (2), 114. (b) Weiberth, F. J.; Hanna, R. G.; Lee, G. E.;
Polverine, Y.; Klein, J. T. Org. Process Res. Dev. 2011, 15, 704.
(19) Shi, Z.; Parlanti, L.; Discordia, R. P.; Hynes, J., Jr.; Miller, M. M.;
O’Grady, H. R. Org. Lett. 2007, 9, 3821.
(20) (a) Jencks, W. P. J. Am. Chem. Soc. 1958, 80, 4581. (b) Jencks,
W. P. J. Am. Chem. Soc. 1958, 80, 4585.
Albert J. DelMonte for useful comments and suggestions during
the preparation of this manuscript.
REFERENCES
■
(1) Hynes, J., Jr.; Dyckman, A. J.; Lin, S.; Wrobleski, S. T.; Wu, H.;
Gillooly, K. M.; Kanner, S. B.; Lonial, H.; Loo, D.; McIntyre, K. W.;
Pitt, S.; Shen, D. R.; Shuster, D. J.; Yang, X.; Zhang, R.; Behnia, K.;
Zhang, H.; Marathe, P. H.; Doweyko, A. M.; Tokarski, J. S.; Sack, J. S.;
Pokross, M.; Kiefer, S. E.; Newitt, J. A.; Barrish, J. C.; Dodd, J.;
Schieven, G. L.; Leftheris, K. J. Med. Chem. 2008, 51, 4.
(2) Fink, B. E.; Vite, G. D.; Mastalerz, H.; Kadow, J. F.; Kim, S.;
Leavitt, K. J.; Du, K.; Crews, D.; Mitt, T.; Wong, T. W.; Hunt, J. T.;
Vyas, D. M.; Tokarski, J. S. Bioorg. Med. Chem. Lett. 2005, 15, 4774.
(3) Suzuki, M.; Miyoshi, M.; Matsumoto, K. J. Org. Chem. 1974, 39,
1980.
(21) (a) Coleman, G. A.; Hauser, C. R. J. Am. Chem. Soc. 1928, 50,
1193. (b) Jaffe, H. H. Chem. Rev. 1953, 53, 191.
(22) TsOH, TFA, methanesulfonic acid, camphorsulfonic acid, citric
acid, and PPTS were examined as well and were less effective.
(23) Barham, H. N.; Clark, L. W. J. Am. Chem. Soc. 1951, 73, 4638.
(24) We propose that 20 was generated by condensation of 3 with
small amounts of acetamide that resulted from amide exchange
between acetic acid and formamide under the forcing reaction
conditions.
(25) Dry HOBt is potentially explosive. Hydrated HOBT should also
be used with caution: (a) Wehrstedt, K. D.; Wandrey, P. A.; Heitkamp,
D. J. Hazard. Mater. 2007, A126, 1. (b) Barnhart, R.; Ironside, M. D.;
Vogt, P. F. Org. Process Res. Dev. 2010, 14, 1480.
(4) (a) Stanovnik, B.; Selic, L. Synthesis 1999, 479. (b) Stanovnik, B.;
Svete, J.; Aljaz-Rozic, M. J. Heterocycl. Chem. 1997, 34, 177.
(5) Hamamoto, I. (Nippon Soda Co, Japan). Preparation of pyrrole
and thiophene derivatives. Jpn. Kokai Tokkyo Koho: 1996; 6 pp.
CODEN: JKXXAF, JP 08059611 A 19960305, Heisei. Application: JP
94-225962 19940826; CAN 125:10608.
(6) Detonation of a reaction using O-(2,4-dinitropenyl)hydroxyl-
amine has been reported: Radhakrishna, A. S.; Loudon, G. M. J. Org.
Chem. 1979, 44, 4836.
(7) (a) Magano, J.; Acciacca, A.; Akin, A.; Collman, B. M.; Conway,
B.; Waldo, M.; Chen, M. H.; Mennen, K. E. Org. Process Res. Dev.
2009, 13, 555. (b) Buckley, J.; Webb, R. L.; Laird, T.; Ward, R. J.
Chem. Eng. News 1982, 60 (28), 5. (c) De Wall, G. Chem. Eng. News
1982, 60 (37), 5.
(8) Hynes, J., Jr.; Doubleday, W. W.; Dyckman, A. J.; Godfrey, J. D.,
Jr.; Grosso, J. A.; Kiau, S.; Leftheris, K. J. Org. Chem. 2004, 69, 1368.
(9) Jaffari, G. A.; Nunn, A. J. J. Chem. Soc. C 1971, 823.
(10) For a potential degradation mechanism, see: Kos, N. J.;
Jongejan, H.; Van Der Plas, H. C. J. Chem. Soc., Perkin Trans. 1 1986,
269.
(11) Tummala, S.; Leung, S. W.; Lo, E. T.; Alvarez, M. M. Production
of gaseous chloramine. PCT Int. Appl. (2004), 27 pp. CODEN:
PIXXD2 WO 2004046280, A2 20040603, CAN 140:425737, AN
2004:453311.
(12) We would like to note that, after completion of these efforts,
Bhattacharya also reported on an alternative phase transfer catalyzed
N-amination process using chloramine: Bhattacharya, A.; Patel, N. C.;
Plata, R. E.; Peddicord, M.; Ye, Q.; Parlanti, L.; Venkatpuram, A. P.;
Grosso, J. A. Tetrahedron Lett. 2006, 46, 5341.
(13) (a) Friestad, G. K.; Shen, Y. J. Org. Chem. 2002, 67, 6236.
(b) Carpino, L. A.; Giza, C. A.; Carpino, B. A. J. Am. Chem. Soc. 1959,
81, 955. (c) Marmer, W. N.; Maerker, G. J. Org. Chem. 1972, 37, 3520.
(d) Boche, G. Encyclopedia of Reagents for Organic Synthesis; Paquette,
L. A., Ed.; Wiley: New York, NY, 1995; Vol. 5, pp 3270. (e) Carpino,
L. A. J. Org. Chem. 1965, 30, 736. (f) Carpino, L. A. J. Org. Chem.
1965, 30, 321. (g) Johnson, J. S.; Berman, A. M. J. Org. Chem. 2005,
70, 364.
(14) (a) Tamura, Y.; Minamikaw, J.; Ikeda, M. Synthesis 1977, 1.
(b) Erdik, E. Tetrahedron 2004, 60, 8747.
(15) (a) Carpino, L. J. Am. Chem. Soc. 1960, 82, 3133. (b) Taylor, E.;
Sun, J.-H. Synthesis 1980, 801.
(16) (a) Boyles, D. C.; Curran, T. T.; Partlett, R. V., IV. Org. Process
Res. Dev. 2002, 6, 230. (b) Radhakrishna, A. S.; Loudon, G. M.; Miller,
M. J. Org. Chem. 1979, 44, 4836. (c) Bellettini, J. R.; Olson, E. R.;
Teng, M.; Miller, M. J. In Encyclopedia of Reagents for Organic Synthesis;
Paquette, L. A., Ed.; Wiley: New York, 1995; Vol. 3, pp 2189.
(17) (a) Vedejs, E.; Smulik, J. Org. Lett. 2003, 22, 4187. (b) Boche,
G. In Encyclopedia of Reagents for Organic Synthesis; Paquette, L. A.,
Ed.; Wiley: New York, 1995; Vol. 4, pp 2240. (c) Boche, G.;
BernHeim, M.; Schott, W. Tetrahedron Lett. 1982, 23, 5399.
(d) Colvin, E.; Kirby, G.; Wilson, A. Tetrahedron Lett. 1982, 23, 3835.
(18) (a) Conrow, R. E.; Dean, D. W.; Zinke, P. W.; Deason, M. E.;
Sproull, S. J.; Dantanarayana, A. P.; DuPriest, M. T. Org. Process Res.
1625
dx.doi.org/10.1021/op300181r | Org. Process Res. Dev. 2012, 16, 1618−1625