Convergent, fit-for-purpose, kilogram-scale synthesis of a 5-lipoxygenase inhibitor

Article abstract of DOI:10.1021/op200299p

Process research and development of a synthetic route towards a novel 5-lipoxygenase inhibitor is described. The synthetic route provided 1 in 27% yield in nine steps (seven steps in the longest linear sequence) and was performed on kilogram scale. The synthesis began with the preparation of the coumarin core via an efficient von Pechmann condensation. The triazole fragment was obtained via a regioselective copper-catalyzed [3 + 2] cycloaddition between a chiral alkyne and the coumarin azide.

Full text of DOI:10.1021/op200299p

Article
pubs.acs.org/OPRD
Convergent, Fit-For-Purpose, Kilogram-Scale Synthesis of a
5-Lipoxygenase Inhibitor
Stephane G. Ouellet,*,† Danny Gauvreau,† Mark Cameron,‡ Sarah Dolman,† Louis-Charles Campeau,†
́
Gregory Hughes,† Paul D. O’Shea,† and Ian W. Davies‡
†Department of Process Research, Merck Frosst Canada, Kirkland, Quebec H9H 3L1, Canada
‡Department of Process Research, Merck & Co., Rahway, New Jersey 07065, United States
S
* Supporting Information
ABSTRACT: Process research and development of a synthetic route towards a novel 5-lipoxygenase inhibitor is described. The
synthetic route provided 1 in 27% yield in nine steps (seven steps in the longest linear sequence) and was performed on kilogram
scale. The synthesis began with the preparation of the coumarin core via an efficient von Pechmann condensation. The triazole
fragment was obtained via a regioselective copper-catalyzed [3 + 2] cycloaddition between a chiral alkyne and the coumarin
azide.
keto-lactone was trapped with triflic anhydride to provide the
vinyl triflate 9.
INTRODUCTION
■
The 5-lipoxygenase (5-LO) pathway is characterized by a series
of biochemical reactions in which arachidonic acid is trans-
formed into pro-inflammatory mediators called leukotrienes1
which are known to exhibit diverse biological actions and are
believed to be involved in many disease states.2 Several studies
have highlighted the role of leukotrienes in the pathogenesis of
asthma; consequently, the development of pharmaceutically
active agents with the ability to inhibit the synthesis and/or
the action of leukotrienes is being pursued for the treatment
of asthma. Our discovery efforts in this field identified 1 as a
potent 5-LO inhibitor.3 To support preclinical and clinical de-
velopment, a practical synthesis of 1 suitable for multi kilogram-
scale preparation of the active pharmaceutical ingredient was
required.
This intermediate underwent a palladium-catalyzed Suzuki
cross-coupling reaction with 3-fluorophenyl boronic acid to
provide coumarin 10. Finally, the benzylic position was acti-
vated using NBS/(BzO)2, and the resulting benzyl bromide was
displaced with NaN3 to give the coumarin azide 3 in an overall
yield of 23% (seven steps). Using this sequence, we were able
to successfully prepare this active pharmaceutical ingredient
(API) on ∼100 g scale.
While the late introduction of the functionalized aryl sub-
stituent was strategically beneficial from a SAR standpoint,
a shorter, more efficient and scalable approach was desired
in order to support further development of this compound.
Realizing the inherent inefficiency in forming the coumarin
core, followed by installation of the m-fluorophenyl side chain,
we explored methods to synthesize the coumarin which would
directly incorporate the requisite substitution pattern. To
achieve this goal, we considered a von Pechmann reaction6
between a functionalized keto-ester and m-cresol (Scheme 3).
Preparation of keto-ester 12 was achieved via a chain extension5
of the activated 3-fluorobenzoic acid in the presence of potas-
sium ethylmalonate, triethylamine, and magnesium chloride.
The unreacted starting materials could easily be rejected during
the workup, and the crude reaction mixture was used directly
in the following step. The coumarin synthesis was achieved
by successfully adapting a methodology that was initially used
for the synthesis of MK-0633.7 The keto-ester 12 reacted
smoothly with m-cresol in methanesulfonic acid at 40−45 °C to
afford the key methyl-coumarin 10. Following an aqueous
workup, the methyl-coumarin 10 was precipitated out of a mix-
ture of DCM/IPA to provide the desired product in 72%
isolated yield with an excellent purity profile (>99 LCAP). Bro-
mination using a small excess of NBS and a catalytic amount
of benzoyl peroxide was performed in acetonitrile at 80 °C.
RESULTS AND DISCUSSION
■
We focused our efforts on two synthetic strategies outlined
in Scheme 1. In both routes to 1, we envisioned that the 1,
4-substituted triazole would be constructed using a regioselec-
tive [3 + 2] cycloaddition between a chiral alkyne (2) and the
coumarin azide 3. For the construction of the coumarin core,
we considered two complementary routes. The azido-coumarin
3 could be obtained from a Suzuki cross-coupling reaction be-
tween a functionalized coumarin and a boronic acid or be pre-
pared in a more expedite way via a von Pechmann cyclization of
a ketoester with 3-methylphenol.
Early work to support medicinal chemistry efforts and early
preclinical profiling were focused on designing a synthetic se-
quence that would allow for maximum flexibility. Summarized
in Scheme 2, the synthetic route utilized for the preparation
of the azido-coumarin 3 allows for facile modification of the
4-aryl component which proved to be an important pharma-
cophore. Cresol 6 was first acylated and underwent a Fries
rearrangement4 under Lewis acidic conditions to provide aryl
ketone 8. Following a Masamune chain extension,5 the resulting
Received: October 21, 2011
Published: January 4, 2012
© 2012 American Chemical Society
214
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219

Organic Process Research & Development
Article
Scheme 1. Retrosynthetic analysis
Scheme 2. First synthetic route developed to access the
coumarin 3
straightforward manner (see Scheme 4). Addition of lithium
(trimethylsilyl)acetylene to 1,1,1-trifluoro-2-butanone had to be
Scheme 4. Preparation of the chiral acetylene ent-15
performed with careful control of the reaction temperature
to ensure the stability of the lithium-acetylide. In order
to achieve reproducible results, temperatures below −50 °C
were required during the lithiation/addition sequence. For the
electrophilic quench with p-nitrobenzoyl chloride, the reaction
mixture could be warmed up to −30 °C without any effect
on the reaction efficiency or the overall purity. We selected
p-nitrobenzoyl chloride because it could be cleaved under mild
conditions and both enantiomers could be easily separated by
chiral HPLC.
Scheme 3. Preparation of the azido-coumarin 3 via a von
Pechmann cyclization
While initial work on desilylation of the alkyne used
HF·Et3N, we elected to develop an alternative protocol to
allow for this transformation to be run in glass reactors. We
found that the TMS could be cleaved under a variety of mild
basic conditions (K3PO4 or K2CO3 in DMF/water mixture).
However, ester hydrolysis was found to be a significant side
reaction. To address this issue, an improved protocol needed to
be developed. The optimal set of conditions involved the use of
1.2 equiv of K3PO4 in DMF at 3 °C with a slow addition of
water. Upon completion of the reaction, a reverse quench into
cold MTBE/3 N HCl prevented further hydrolysis of the ester.
Under these conditions, the ester hydrolysis byproduct forma-
tion was reduced to less than 1.5 LCAP. To further upgrade the
purity of rac-15 (and to reject the undesired hydrolysis by-
product), the alkyne was recrystallized from heptane to provide
rac-15 with an overall isolated yield of 92%. The racemic alkyne
was resolved by chiral HPLC to provide the desired alkyne
in >99% ee with a 45% recovery.
The final three chemical steps in the synthesis of this API
were developed as a through-process starting from the azido-
coumarin 3 and the chiral alkyne ent-15. A copper-catalyzed
regioselective [3 + 2] cycloaddition8 allowed for the efficient
formation of the triazole 20. While these reactions were often
run with a very high catalyst loading (>25 mol % of CuI),
we successfully demonstrated that <5 mol % was sufficient to
obtain high conversion with acceptable reaction time (24 h at rt).
In order to further decrease the catalyst loading while
Under these optimized conditions, we typically observed about
85% conversion to the desired monobromide along with 5−8%
unreacted 10 and 7−10% dibromide (LCMS). Upon cooling of
the reaction mixture, the monobromide crystallized in 60%
yield with good rejection of the dibromide (<0.2 LCAP)
but is contaminated with 1−2% of the starting material (10).
Nevertheless, this purity profile was deemed suitable as these
levels of methylcoumarin 10 can easily be rejected downstream.
Finally, the displacement of the benzylic bromide with NaN3 in
ethanol afforded the azido-coumarin 3 in 55−65% yield over
both steps. This robust and straightforward approach provided
3 in four steps from 3-fluorobenzoic acid (11) in 39% overall
yield on >2.5 kg scale.
In the preparation of ent-15, we opted for a chiral resolution
of the racemic material which could be prepared in a
215
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219

Organic Process Research & Development
Article
Scheme 5. Possible pathway leading to the formation of the byproduct 16
Scheme 6. End-game through-process to access the 5-lipoxygenase inhibitor 1
maintaining acceptable reaction times, we explored the effect of
temperature on the reaction profile. The reaction rate was
found to be significantly increased upon raising the reaction
temperature to 40 °C. Under these conditions, we found that
3 mol % of catalyst was sufficient to allow the reaction to reach
complete conversion in about 1.5 h and that 1 mol % of CuI
would lead to full conversion in about 18 h. Unfortunately,
when the temperature was increased further, we observed the
formation of an impurity which tracked to the final API and was
difficult to reject by recrystallization. Somewhat surprisingly,
this impurity was identified as being alkene 16 (Scheme 5).9
Electron-rich azides typically react selectively with alkynes to
afford the 1,2,3-triazole. However, it is known that electron-
deficient azides, such as Ts-N3, react with alkynes to afford
rearranged amine products. On the basis of our observations
and recent literature precedent,10 we believe that impurity 16
could arise from such a rearrangement. Triazolyl copper species
18, the first intermediate formed in the cycloaddition, could
undergo a Dimroth rearrangement to ketenimine 17 via the
hetero-Wolff rearrangement of the α-diazoimino species 19.11
This ketenimine could then undergo a hydrolysis followed by
an elimination of the benzoyl group to provide the unsaturated
amide 16. This is consistent with the observation that amide
16 is only observed after the 4 N LiOH treatment and
reacidification with 6 N HCl. These observations warn of the
potential formation of such contaminants which may form
even when using electron-rich alkyl azides in this chemistry.
Fortunately, on the basis of our previous optimization of
temperature and catalyst loading, we were able to control the
formation of this impurity by selecting 30 °C and 3 mol % of
CuI. These conditions provided the best balance between low
catalyst loading, convenient reaction rate, and low level
(<0.6%) of the impurity 16 (Scheme 6).
Once the cycloaddition was found to be complete, the reac-
tion mixture was cooled to −10 °C, and the p-nitrobenzoate
was hydrolyzed by addition of aqueous 4 N LiOH which also
resulted in the ring-opening of the coumarin core. The lactone
was efficiently regenerated upon acidification of the reaction
mixture. Finally, on the 5 kg batch, a recrystallization (IPAc/
heptane) afforded a highly crystalline white solid (1) in 70%
isolated yield (from 3) with an acceptable purity profile (97.6
LCAP, purities >97% were typically observed).
CONCLUSION
■
In summary, we developed a practical and scalable synthesis of
a 5-LO inhibitor. The azido-coumarin fragment was prepared
via two complementary routes. On kilogram scale, this was
achieved via a von Pechmann condensation (four steps, 39%
overall yield). Careful study of the [3 + 2] cycloaddition be-
tween the chiral alkyne ent-15 and the coumarin azide 3
revealed the formation of an unexpected impurity which could
be controlled by selecting the appropriate temperature and cat-
alyst loading, resulting in the formation of triazole heterocycle
20 in high yield and selectivity. This chemistry was demon-
strated on multikilogram scale and provided the desired API in
good overall yield (27%) and high chemical purity (97.6%
LCAP, 99.5% ee).
EXPERIMENTAL SECTION
■
General Procedure. It is worth noting that the appropriate
safety evaluations were carried out and that under the reac-
tion conditions described, the risk was deemed acceptable.
216
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219

Organic Process Research & Development
Article
The azide 3, the nitro-alkyne 15, and the API (1) were found
to be not shock sensitive. The azide 3 was tested in presence
of ethanol and showed no exothermic activity below 200 °C
(maximum temperature recommended for drying the azide was
set at 40 °C). DSC testing of the API (1) slurried in THF
showed a small exotherm at 175 °C and a larger one around
250 °C. Since the triazole formation will be carried out below
66 °C (boiling point of THF), these exotherms should not be
encountered.
to 65 °C which allowed for the complete dissolution of the
acetylene-ester. The solution was cooled to 1 °C over a period
of 15 h to allow for a slow crystallization of the acetylene-ester
(seeds were added at 56 °C). The crystallized acetylene-ester
was filtered, rinsed with cold heptane (2 × 8 L), and dried
under a flow of nitrogen. The recovery of rac-15 was 6.56 kg
(92% yield). The losses to the mother liquors were estimated to
be about 4.8% (378 g). 1H NMR (400 MHz, DMSO-d6): 8.36
(d, J = 8.8 Hz, 2 H), 8.15 (d, J = 8.8 Hz, 2 H), 4.23 (s, 1 H),
2.46−2.26 (m, 2 H), 1.06 (t, J = 7.5 Hz, 3 H). 13C NMR
(101 MHz, DMSO-d6): 160.9, 150.7, 133.7, 130.9, 124.2, 122.9
(q, J = 285.9 Hz), 82.4, 77.2 (q, J = 31.2 Hz), 74.1, 27.1,
8.0. HRMS (ESI): m/z calcd for C13H11NO4F3 (M + H)
302.0635; found 302.0640. Melting point (heptane): 103.8 °C.
IR (neat, cm−1): 1738, 1523, 1349, 1278, 1196, 1089, 711.
Ethyl 3-(3-fluorophenyl)-3-oxopropanoate (12).(Flask A: 22 L).
To a suspension of 3-fluorobenzoic acid (11, 2.00 kg, 1.0 equiv)
in CH3CN (6 L) was added CDI (2.60 kg, 1.1 equiv, added
portion-wise, gas evolution was observed, and the temperature
of the reaction mixture decreased from 14 to 5 °C). This solu-
tion was stirred for 4 h (temperature of the batch was raised
to 22 °C).
(Flask B: 100 L). To a solution of potassium ethylmalonate
(3.17 kg, 5 equiv) in CH3CN (30 L) was added MgCl2 (1.47 kg,
1.08 equiv) in portions over 15 min (batch temperature rose
from 18 to 41 °C). The mixture was stirred at 35 °C for 30 min
and then cooled to 25 °C and Triethylamine (6.0 L, 3.0 equiv)
was added. After the addition of about 2 L of TEA, the reaction
mixture became very thick, and mechanical stirring was more
challenging. The slurry was stirred for 30 min. The solution
in flask A was then transferred to the slurry in flask B over
5 min. The batch temperature rose form 24 to 32 °C, and gas
evolution was observed. The reaction mixture was stirred for
1.5 h, cooled to 7 °C, and quenched with 3 N HCl (32 L),
while maintaining the batch temperature <20 °C. The resulting
by-phasic solution was distilled to remove volatiles (CH3CN)
and the resulting concentrate extracted with MTBE (35 L).
The organic phase was washed with water (8 L), 5% NaHCO3
(8 L), and 20% NaCl (8 L). The organic solution was con-
centrated under reduced pressure to give an orange-colored oil
(2.921 kg) in 97% yield.12
4-(3-Fluorophenyl)-7-methyl-2H-chromen-2-one (10). To
a solution of m-cresol (6, 2.35 kg, 1.0 equiv) in methanesulfonic
acid (16 L) was added the keto-ester 12 (4.85 kg, 1.06 equiv)
over 30 min (internal temperature rose from 19 to 39 °C). The
mixture was stirred for 3 h (39−33 °C), heated at 40−45 °C for
2 h, and allowed to cool to room temperature overnight. The
reaction mixture was partitioned between CH2Cl2 (20 L) and
water (20 L), and the aqueous layer was back-extracted with
CH2Cl2 (8 L). The combined organic layers were washed with
1 N NaOH (16 L) and water (16 L). The organic solution was
concentrated under reduced pressure to about 10 L. IPA (12 L)
was added, and volatiles were evaporated until a volume of
∼18 L of slurry was obtained. The suspension was stirred over-
night. The product was isolated by filtration (the filter cake was
washed with IPA (2 × 2 L)). The product was dried on the
filter pot under a flow of nitrogen for 8 h and then transferred
to drying trays and dried in a vacuum oven at 40 °C under a
gentle nitrogen sweep to give 10 (4.05 kg, 100 wt % 99.8
LCAP) in 72% isolated yield. 1H NMR (400 MHz, DMSO-d6):
7.63−7.55 (m, 1 H), 7.42−7.30 (m, 3 H), 7.30−7.23 (m, 2 H),
7.12 (d, J = 8.09 Hz, 1 H), 6.35 (s, 1 H), 2.38 (s, 3 H). 13C
NMR (101 MHz, DMSO-d6): 162.0 (d, J = 245.4 Hz), 159.6,
Intermediates were analyzed on an Agilent 1100 series in-
strument using the following conditions: 4.6 mm × 50 mm
Zorbax SB-C18 column, 1.8 μm, gradient elution (0.1% H3PO4/
CH3CN 90:10 to 10:90 over 8 min, then hold 4 min, flow
rate = 1.5 mL/min, detection = 215 nm, T = 35 °C, sam-
ples were diluted in MeCN, and a 10 μL sample was injected.
HPLC conditions used for the chiral separation of rac-15:
(11 cm × 25 cm DAC column packed with Chiralpak OD).
The feed was about 29 mg/mL in v/v 13/87 isopropyl alcohol/
heptane, the injection volume ranged from 1.0 to 1.2 L with
elution performed using 15% v/v isopropyl alcohol/heptane at
a flow rate of 800 mL/min, UV detection at 295 nm. The
second eluting enantiomer (S-isomer) was the desired one.
1-Ethyl-1-(trifluoromethyl)-3-(trimethylsilyl)prop-2-yn-1-yl
4-nitrobenzoate (14). A mixture of TMS−acetylene (7.06 kg,
1.5 equiv) and MTBE (24 L) was cooled to −67 °C, and
n-BuLi (2.26 M, 22.3 L, 1.05 equiv) was added over 6 h
(keeping the reaction mixture below −51 °C). The mixture was
aged 45 min, allowing the mixture to cool to −65 °C. 1,1,
1-Trifluoro-2-butanone (6.5 L, 1.0 equiv) was added over 4 h
(keeping the internal temperature below −50 °C). The mix-
ture was aged 20 min, and p-nitrobenzoyl chloride (9.33 kg,
1.05 equiv) was added over 2.5 h as a THF solution (14 L)
(keeping the reaction mixture below −33 °C). The mixture
was stirred for 4 h. The reaction was quenched by adding water
(18 L, exotherm observed) over 60 min. To the reaction
mixture was added Solka Floc (500 g, 8 wt %), and the mix-
ture was stirred for 20 min. The biphasic mixture was filtered
over Solka Floc, and the organic layer was washed with half-
saturated NaHCO3 (2 × 30 L) and brine (30 L). The crude
reaction mixture was used directly into the following step (the
crude was concentrated under vacuum and solvent switched to
DMF, temperature was kept below 35 °C, and final volume was
1
about 45 L). H NMR (400 MHz, DMSO-d6): 8.37 (d, J =
8.4 Hz, 2 H), 8.16 (d, J = 8.4 Hz, 2 H), 2.45−2.22 (m, 2 H),
1.06 (t, J = 7.6 Hz, 3 H), 0.17 (s, 9 H). 13C NMR (101 MHz,
DMSO-d6): 160.8, 150.7, 133.8, 130.9, 124.2, 122.9 (q, J =
284.9 Hz), 96.6, 95.0, 77.4 (q, J = 31.1 Hz), 27.1, 8.1, −0.8.
HRMS (ESI): m/z calcd for C16H18NO4F3SiNa (M + Na)
396.0849; found 396.0869. Melting point: 49.7 °C. IR (neat,
cm−1): 2964, 1733, 1527, 1349, 1274, 1251, 1097, 715.
1-Ethyl-1-(trifluoromethyl)prop-2-yn-1-yl 4-nitrobenzoate
(15). A solution of TMS−acetylene-ester 14 (8.85 kg, 1.0
equiv) in DMF (36 L) was cooled to 0 °C, and solid K3PO4
(6.04 kg, 1.2 equiv) was added followed by water (940 mL,
2.2 equiv, added over 60 min to control the exotherm). The
mixture was aged 30 min at 0 °C and quenched by pouring
the DMF stream onto a precooled (2 °C) mixture of MTBE
(27 L) and 3 N HCl (32 L). The residual K3PO4 was decanted
off. The mixture was filtered on Solka Floc and rinsed with
MTBE (9 L). The organic layer was washed with water (2 ×
35 L), half-saturated NaHCO3 (35 L), and brine (35 L) to give
a quantitative assay yield of rac-15 (98.5 LCAP). A suspension
of rac-15 (7.14 kg, 1.0 equiv) in heptane (32 L) was heated
217
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219

Organic Process Research & Development
Article
153.6, 153.4 (d, J = 2.1 Hz), 143.2, 136.9 (d, J = 8.2 Hz), 130.9
(d, J = 8.2 Hz), 126.2, 125.6, 124.7 (d, J = 2.7 Hz), 117.0, 116.5
(d, J = 21.1 Hz), 115.6 (d, J = 2.6 Hz), 115.4, 114.1, 21.0.
HRMS (ESI): m/z calcd for C16H12O2F (M + H) 255.0816;
found 255.0825. Melting point (IPA): 103.9 °C. IR (neat,
cm−1): 3055, 1729, 1584, 1440, 1359, 1265, 1253, 1179, 1147,
1115, 967, 716.
160 L extractor (6 L THF used to rinse the flask) and cooled
to 0 °C. To this cold solution was added 6 N HCl (15 L, small
exotherm observed but easily controlled by rate of addition),
and the reaction mixture was stirred vigorously at rt for 18 h.
MTBE (50 L) and water (10 L) were added, and the layers
were separated. The organic layer was washed with Na2CO3
(2 × 40 and 20 L, three layers were observed during the first
wash, only the lower layer was rejected) and half-saturated
brine (2 × 20 L). The reaction mixture was solvent switched to
ethanol (<0.5% THF vs ethanol was detected, final volume was
8 mL/g) and heated until complete dissolution (67 °C). The
reaction mixture was slowly cooled down to rt (seeded at 45−
47 °C) and aged for 6 h (at this point, the solubility in the
mother liquor was 25 mg/mL). This suspension was filtered
and the residual solid washed with the mother liquor (10 L),
ethanol/water (30 L of 4:6), and heptane (40 L). The beige
solid was dried under a flow of N2 overnight and transferred
into a vacuum oven at 50 °C under a gentle sweep of nitrogen.
5.15 kg of 1 was obtained (70% yield from ent-15 and 3, beige
solid, 97.6 LCAP, 33 ppm Cu, 99.5% ee, 0.5% water, 0.3%
ethanol). 1H NMR (400 MHz, acetone-d6): 8.50 (s, 1 H), 8.37
(d, J = 8.4 Hz, 2 H), 8.27 (d, J = 8.4 Hz, 2 H), 7.65−7.57 (m,
1 H), 7.48 (d, J = 8.2 Hz, 1 H), 7.40−7.22 (m, 5 H), 6.37 (s,
1 H), 5.84 (s, 2 H), 3.07−2.97 (m, 1 H), 2.88−2.77 (m, 1 H),
1.11 (t, J = 7.4 Hz, 3 H). 13C NMR (101 MHz, acetone-d6):
163.5 (d, J = 245.7 Hz), 162.6, 160.0, 155.1, 154.2 (d, J =
1.9 Hz), 151.9, 143.6, 141.1, 138.0 (d, J = 8.1 Hz), 135.7, 131.9
(d, J = 8.0 Hz), 131.8, 128.2, 125.5, 125.4 (d, J = 3.0 Hz), 125.1
(d, J = 285.3 Hz), 124.7, 124.5, 119.2, 117.3 (d, J = 21.4 Hz),
117.1, 116.6, 116.4 (d, J = 22.8 Hz), 82.3 (d, J = 29.4 Hz), 53.6,
25.9, 7.8. HRMS (ESI): m/z calcd for C29H21N4O6F4 (M + H)
597.1392; found 597.1403. Melting point (ethanol): 133.9 °C.
[α]D20: −48.1° (c = 10.0, DCM). IR (neat, cm−1): 1727, 1527,
1349, 1268, 1253, 1182, 1118, 917, 867, 715.
7-(Bromomethyl)-4-(3-fluorophenyl)-2H-chromen-2-one.
To a stirred suspension of methylcoumarin 10 (4.05 kg, 1.0
equiv) in CH3CN (10.0 L) was added NBS (3.32 kg, 1.16
equiv) followed by (BzO)2 (192 g, 0.05 equiv) and the mixture
was heated to reflux (78−82 °C) for 2 h. At this time, the
heating source was turned off and the batch stirred overnight
which allowed the product to precipitate out of solution. The
crystallized product was then isolated by filtration and the filter
cake washed with IPA (2 × 4 L) and dried under a flow of
nitrogen to give monobrominated product (3.21 kg, 97 wt %
2 mol % 10 by H/F NMR) in 60% yield. 1H NMR (400 MHz,
CDCl3): 7.58−7.49 (m, 1 H), 7.50−7.42 (m, 2 H), 7.33−7.22
(m, 3 H), 7.18 (d, J = 9.2 Hz, 1 H), 6.39 (s, 1 H), 4.55 (s, 2 H).
13C NMR (101 MHz, CDCl3): 162.7 (d, J = 249.2 Hz), 160.0,
154.0, 153.6 (d, J = 1.6 Hz), 142.3, 136.8 (d, J = 7.6 Hz), 130.7
(d, J = 8.3 Hz), 127.1, 125.0, 124.1 (d, J = 2.9 Hz), 118.3, 117.6,
116.8 (d, J = 20.7 Hz), 115.7, 115.5 (d, J = 22.3 Hz), 31.5.
HRMS (ESI): m/z calcd for C16H11O2FBr (M + H) 332.9921;
found 332.9930. Melting point (MeCN): 103.8 °C. IR (neat,
cm−1): 3032, 1723, 1579, 1373, 1254, 1221, 1148, 900, 720.
7-(Azidomethyl)-4-(3-fluorophenyl)-2H-chromen-2-one
(3). To a suspension of bromomethylcoumarin (3.22 kg, 1.0
equiv) in ethanol (32 L) was added NaN3 (0.65 kg, 1.05 equiv)
and the mixture heated at 60 °C for 3 h. At this point, the batch
was cooled to 10−15 °C. Water (30 L) was added slowly, and
the batch was stirred for an additional 60 min. The residual
slurry was transferred onto a filter pot, and the cake was washed
with water (2 × 10 L) and allowed to dry under a flow of nitro-
gen overnight. The product was then transferred onto drying
trays and dried to constant weight (vacuum oven 40 °C, nitro-
gen sweep) to provide the azidemethylcoumarin (3, 2.65 kg,
97 wt %, 98.1 LCAP) in 93% isolated yield. 1H NMR
(400 MHz, acetone-d6): 7.64 (q, J = 7.1 Hz, 1 H), 7.52 (d, J =
8.2 Hz, 1 H), 7.48−7.30 (m, 5 H), 6.39 (s, 1 H), 4.63 (s, 2 H).
13C NMR (101 MHz, acetone-d6): 163.6 (d, J = 246.7 Hz),
160.1, 155.1, 154.4 (d, J = 2.1 Hz), 141.7, 138.2 (d, J = 8.0 Hz),
131.8 (d, J = 8.5 Hz), 128.1, 125.5 (d, J = 2.9 Hz), 124.9, 119.1,
117.3 (d, J = 21.1 Hz), 117.2, 116.4 (d, J = 22.9 Hz), 116.3,
54.1. HRMS (ESI): m/z calcd for C16H11N3O2F (M + H)
296.0830; found 296.0835. Melting point (water): 90.7 °C. IR
(neat, cm−1): 2116, 2096, 1724, 1579, 1417, 1349, 1276, 1254,
1183, 1144, 890.
ASSOCIATED CONTENT
■
S
* Supporting Information
Complete characterization data and spectra for all new com-
pounds. This material is available free of charge via Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
stephane_ouellet@merck.com.
ACKNOWLEDGMENTS
■
We thank Ravi Sharma, Dean Hutchings and Wayne Mullet
for analytical support. We thank Robert A. Reamer for helpful
assistance with NMR experiments.
(1S)-1-(1-{[4-(3-Fluorophenyl)-2-oxo-2H-chromen-7-yl]-
methyl}-1H −1,2,3-triazol-4-yl)-1-(trifluoromethyl)propyl
4-nitrobenzoate (1). To a solution of azidemethylcoumarin 3
(5.0 kg, 1.0 equiv) in THF (25 L) was added alkyne ent-15
(5.44 kg, 1.1 equiv) and di-isopropylethylamine (5.74 L, 2.0
equiv). This reaction mixture was degassed by bubbling
nitrogen for 20 min. To this solution was added CuI (62.5 g,
grounded with a mortar and pestle) in one portion and the
reaction mixture was degassed for an additional 5 min before to
be heated up to 30 °C and stirred for 24 h (complete consump-
tion of the azide was observed). At this point the reaction
mixture was cooled to −10 °C and 4 N LiOH (15 L) was
added (small exotherm observed but easily controlled by rate of
addition). After 4 h, the reaction mixture was transferred into a
REFERENCES
■
(1) (a) Samuelsson, B. Science 1983, 220, 568. (b) O’Byrne, P. M.;
Israel, E.; Drazen, J. M. Ann. Intern. Med. 1997, 127, 472.
(2) (a) Peters-Golden, M.; Henderson, W. R. Jr. N. Engl. J. Med.
2007, 357, 1841. (b) Dwyer, J. H.; Allayee, H.; Dwyer, K. M.; Fan, J.;
Wu, H.; Mar, R.; Lusis, A. J.; Mehrabian, M. N. Engl. J. Med. 2004,
350, 29.
(3) Grimm, E. L.; Ducharme, Y.; Frenette, R.; Friesen, R.; Gagnon,
M.; Juteau, H.; Laliberte, S.; Mackay, B.; Gareau, Y. U.S. Pat. Appl.
Publ. U.S. 2008/0188521, 2008.
(4) For other examples of Fries rearrangement, see: (a) Paul, S.;
Gupta, M. Synthesis 2004, 1789. (b) Commarieu, A.; Hoelderich, W.;
218
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219

Organic Process Research & Development
Article
Laffitte, J. A.; Dupont, M.-P. J. Mol. Cat. A.: Chem. 2002, 182−183,
137.
(5) (a) Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem., Int.
Ed. Engl. 1979, 18, 72. (b) Masamune, S. Aldrichimica Acta 1978,
11, 23.
(6) (a) von Pechmann, H.; Duisberg, C. Ber. Dtsch. Chem. Ges. 1883,
16, 2119. (b) von Pechmann, H. Ber. Dtsch. Chem. Ges. 1884, 17, 929.
(c) For arecent example, see: Karimi, B.; Zareyee, D. Org. Lett. 2008,
10, 3989.
(7) Gosselin, F.; Britton, R. A.; Davies, I. W.; Dolman, S. J.;
Gauvreau, D.; Hoerrner, R. S.; Hughes, G.; Janey., J.; Lau, S.;
Molinaro, C.; Nadeau, C.; O’Shea, P. D.; Palucki, M.; Sidler, R. J. Org.
Chem. 2010, 75, 4154.
(8) (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596. (b) Kolb, H. C.; Finn, M. G.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40, 2004. (c) Moses, J. E.;
Moorhouse, A. D. Chem. Soc. Rev. 2007, 36, 1249. (d) Meldal, M.;
Tornoe, C. W. Chem. Rev. 2008, 108, 2952. (e) Li, L.; Zhang, G.;
Zhu, A.; Zhang, L. J. Org. Chem. 2008, 73, 3630. (f) Ackermann, L.;
Potukuchi, H. K.; Landsberg, D.; Vicente, R. Org. Lett. 2008, 10, 3081.
(9) This impurity was separated by preparative HPLC, and the struc-
ture was assigned by NMR and subsequently confirmed by indepen-
dent synthesis.
(10) (a) Cho, S. H.; Yoo, E. J.; Bae, I.; Chang, S. J. Am. Chem. Soc.
2005, 127, 16046. (b) Yoo, E. J.; Ahlquist, M.; Bae, I.; Sharpless, B. K.;
Fokin, V. V.; Chang, S. J. Org. Chem. 2008, 73, 5520.
(11) Brown, D. J.; Harper, J. S. J. Chem. Soc. 1963, 1276. (b) Perrin,
D. D. J. Chem. Soc. 1963, 1284. (c) Katritzky, A. R.; Ji, F. B.; Fan,
W.-Q.; Gallos, J. K.; Greenhill, J. V.; King, R. W.; Steel, P. J. J. Org.
Chem. 1992, 57, 190. (d) Sako, M.; Kawada, H. J. Org. Chem. 2004,
69, 8148.
(12) Huang, X.; O’Brien, E.; Thai, F.; Cooper, G. Org. Process Res.
Dev. 2010, 14, 592.
219
dx.doi.org/10.1021/op200299p | Org. ProcessRes. Dev. 2012, 16, 214−219