Development of an efficient palladium-catalyzed intramolecular carbometalation reaction for the synthesis of a dibenzoxapine containing tetra-substituted exocyclic alkene

Article abstract of DOI:10.1021/op800231b

A practical and scaleable synthesis of (Z)-3-(1-(8-bromodiben-zo[b,e] oxepin-ll(6H)-ylidene) ethyl) aniline hydrochloride (1 *HC1), a key intermediate in the synthesis of a selective nuclear hormone receptor modulator, is described. The target compound is

Full text of DOI:10.1021/op800231b

Organic Process Research & Development 2009, 13, 315–320
Development of an Efficient Palladium-Catalyzed Intramolecular Carbometalation
Reaction for the Synthesis of a Dibenzoxapine Containing Tetra-substituted
Exocyclic Alkene†
Rachel N. Richey and Hannah Yu*
Chemical Product Research and DeVelopment, Lilly Research Laboratories, Eli Lilly and Company,
Indianapolis, Indiana, 46285-4813, U.S.A.
Abstract:
and stereoselective synthesis of tetra-substituted alkenes still
poses a unique synthetic challenge.2 Although the classical
double-bond-forming methods such as the Wittig reaction or
dehydration of a carbinol could be used for the synthesis of
tetra-substituted alkenes, they are often limited in scope and
lack of stereoselectivity. The carbometalation of alkynes has
been developed as a widely used method for the formation of
tetra-substituted alkenes.3 In particular, intramolecular carbo-
palladation of alkynes (cyclocarbopalladation) has been used
for the formation of various tetrasubstituted alkenes.4 In 1988,
Grigg et al.,5 shortly followed by Negishi and Zhang,6 described
a cyclocarbopalladation process that afforded exocyclic alkenes
stereospecificly. The reaction involved an initial cyclocarbo-
palladation to form a vinylpalladium intermediate, followed by
terminating the vinylpalladium species with an organometallic
reagent. Recently, we published two stereoselective methods
for the preparation of tetra-substituted alkenes by intramolecular
cyclocarbopalladation of alkynes.7 These reports represented the
first preparations of seven-membered rings through a palladium-
catalyzed cascade reaction using organometallics as the termi-
nating trapping species. Herein, we describe the first large-scale
preparation of 2 using this strategy. The key step involved a
palladium-catalyzed intramolecular carbometalation reaction
between alkyne 3 and 3-nitrophenylboronic acid 4 to establish
the desired geometric relationship (Figure 1).
A practical and scaleable synthesis of (Z)-3-(1-(8-bromodiben-
zo[b,e]oxepin-11(6H)-ylidene)ethyl)aniline hydrochloride (1 ·HCl),
a key intermediate in the synthesis of a selective nuclear hormone
receptor modulator, is described. The target compound is prepared
in five steps from commercially available (5-bromo-2-iodophenyl)-
methanol (5) with a 47% overall yield. The key step involves a
palladium-catalyzed intramolecular carbometalation of an alkyne,
which affords the dibenzoxapine containing tetrasubstituted exo-
cyclic alkene framework stereoselectively in a single step from
readily available building blocks 4-bromo-2-(2-iodo-phenoxym-
ethyl)-1-prop-1-ynyl-benzene (3) and 3-nitrophenylboronic acid (4).
The development of each step is described. The main focus of the
paper is the description and optimization of the intramolecular
carbometalation of an alkyne. Eventually, the target compound
1·HCl was prepared in multikilogram quantities with >97%
purity.
Introduction
During our program for the development of selective nuclear
hormone receptor modulators,1 a practical and scaleable syn-
thesis of (Z)-3-(1-(8-bromodibenzo[b,e]oxepin-11(6H)-ylide-
ne)ethyl)aniline (1), was needed. This was a key intermediate
used in the late step of our convergent synthetic route. We detail
here our efforts for the preparation of 1, using a palladium-
catalyzed intramolecular carbometalation of an alkyne as the
key transformation.
Figure 1. Intramolecular carbometalation approach.
(2) For a recent review on the stereocontrolled synthesis of tetra-substituted
olefins, see: Flynn, A. B.; Ogilvie, W. W. Chem. ReV. 2007, 107, 4698.
(3) For reviews, see: (a) Normant, J. F.; Alexakis, A. Synthesis 1981,
841. (b) Doyle, M. P. In ComprehensiVe Organometallic Chemistry
II; Abel, E. W., Stone, F. G. A., Wilkinson, G., Hegedus, L., Eds.;
Pergamon Press: Oxford, 1995; Vol. 12, pp 387-420.
(4) For representative examples, see: (a) Negishi, E.-I.; Noda, Y.; Lamaty,
F.; Vawter, E. J. Tetrahedron Lett. 1990, 31, 4393. (b) Wang, R.-T.;
Chou, F.-L.; Luo, F.-T. J. Org. Chem. 1990, 55, 4846. (c) Poli, G.;
Giambastiani, G.; Heumann, A. Tetrahedron 2000, 56, 5959. (d)
Fretwell, P.; Grigg, R.; Sansano, J. M.; Sridharan, V.; Sukirthalingam,
S.; Wilson, D.; Redpath, J. Tetrahedron 2000, 56, 7525. (e) Zhou,
C.; Emrich, D. E.; Larock, R. C. Org. Lett. 2003, 5, 1579. (f) Cheung,
W. S.; Patch, R.; Player, M. R. J. Org. Chem. 2005, 70, 3741. (g)
Bour, C.; Blond, G.; Salem, B.; Suffert, J. Tetrahedron 2006, 62,
10567, and references therein.
Dibenzoxapine derivatives are an important class of biologi-
cally active compounds. The dibenzoxapine derivatives contain-
ing tetra-substituted exocyclic alkene moieties represent a novel
structure.1 Despite impressive recent progress, the efficient regio-
† We dedicate this paper to the memory of Dr. Chris Schmid, a colleague,
mentor, and friend, who passed away December 26, 2007.
* To whom correspondence should be addressed. E-mail: yu_hannah@
lilly.com.
(1) (a) Coghlan, M. J.; Green, J. E.; Grese, T. A.; Jadhav, P. K.; Matthews,
D. P.; Steinberg, M. I.; Fales, K. R.; Bell, M. G. WO 2004/052847
A2, 2004. (b) Carson, M. W.; Coghlan, M. J. WO 2008/008882, A2,
2008.
10.1021/op800231b CCC: $40.75
Published on Web 12/30/2008
2009 American Chemical Society
Vol. 13, No. 2, 2009 / Organic Process Research & Development
•
315

Figure 2. Catalytic cycle for the formation of desired product 9 and undesired byproduct 10.
Scheme 1. Synthesis of cyclization precursor 3
Scheme 2. Synthesis of 9
Results and Discussion
The cyclization precursor 3 was prepared from commercially
available (5-bromo-2-iodophenyl)methanol (5)8 in three steps
(Scheme 1). The Sonogoshira coupling between 5 and propyne
was carried out using PdCl2(PPh3)2 and CuI in diethylamine.9
Under the treatment of excess propyne, the reaction proceeded
to completion at rt after 12 h. After aqueous work up and
filtering through silica gel (3×), alkyne 6 was isolated as a pale-
yellow solid in 87% yield. Conversion of 6 to 3 could be carried
out through either a one-step Mitsunobu reaction or a two-step
alkylation procedure. To avoid chromatography purifications,
the two-step procedure was adapted. Therefore, alcohol 6 was
converted to its tosylate or corresponding halides, and the
resulting intermediates were coupled with 2-iodophenol to afford
the cyclization precursor 3. Benzyl bromide 7 was selected as
the intermediate for scale-up since it provided easy purification
when forward processed to 3. Initial bromination of 6 using
HBr and acetic acid failed to give any of the desired product
7.10 The bromination was then carried out with phosphorous
tribromide in toluene. At 60 °C, the reaction proceeded to
completion without the presence of an amine.11 After aqueous
workup, benzyl bromide 7 was subjected to the alkylation
conditions to couple with 2-iodophenol. Upon treatment with
potassium carbonate, the reaction proceeded to completion in
DMF at rt, and after the addition of water, the resulting product
3 was crystallized from the reaction mixture. Filtration followed
by drying afforded technical grade 3 with >95% HPLC area
percentage. Recrystallization of 3 in EtOH further increased
its purity to >98%. A 74% yield was obtained over the
bromination and alkylation steps.
(5) (a) Burns, B.; Grigg, R.; Sridharan, V.; Worakun, T. Tetrahedron Lett.
1988, 29, 4325. (b) Burns, B.; Grigg, R.; Ratananukul, P.; Sridharan,
V.; Stevenson, P.; Sukirthalingam, S.; Worakun, T. Tetrahedron Lett.
1988, 29, 5565.
(6) Zhang, Y.; Negishi, E. J. Am. Chem. Soc. 1989, 111, 3454.
(7) (a) Yu, H.; Richey, R. N.; Carson, M. W.; Coghlan, M. J. Org. Lett.
2006, 8, 1685. (b) Yu, H.; Richey, R. N.; Mendiola, J.; Adeva, M.;
Somoza, C.; May, S. A.; Carson, M. W.; Coghlan, M. J. Tetrahedron
Lett. 2008, 49, 1915.
(8) We also prepared compound 5 following a modified literature
procedure: Frahn, J.; Schlueter, A. D. Synthesis 1997, 11, 1301.
(9) Sashida, H.; Ohyanagi, K.; Minoura, M.; Akiba, K. J. Chem. Soc.,
Perkin Trans. 1 2002, 606.
(11) Ford-Moore, A. H.; Perry, B. J. Organic Syntheses IV; John Wiley &
Sins: New York, 1963; p 955.
(12) For a proposed mechanism on an intermolecular carbometalation
reaction, see: Zhou, C.; Larock, R. C. J. Org. Chem. 2005, 70, 3765.
(10) Kaslow, C. E.; Schlatter, J. M. J. Am. Chem. Soc. 1955, 77, 1054.
316
•
Vol. 13, No. 2, 2009 / Organic Process Research & Development

Scheme 4. Major impurities (>1%) detected by HPLC at
With cyclization precursor 3 in hand, we were ready to
attempt the key cyclization reaction. As part of method
development efforts, we first optimized reaction conditions using
compound 8 as a model substrate (Scheme 2).7a As expected,
the major byproduct was identified as the direct Suzuki coupling
product 10. A proposed catalytic cycle showing formation of
desired product 9 and undesired byproduct 10 is depicted in
Figure 2.12 Thus, the desired pathway consists of the following
key steps: (1) reduction of palladium (II) to palladium (0); (2)
oxidative addition of aryl iodide to palladium (0); (3) carbo-
palladation of the alkyne to form a vinylic palladium species;
(4) transmetalation of the vinylic palladium species with boronic
acid; (5) reductive elimination of palladium (II) to give the
desired product 9. Figure 2 also shows an undesired pathway,
in which a direct transmetalation occurs between the boronic
acid and the initial palladium (II) intermediate. As a result, the
Suzuki cross-coupling byproduct 10 forms. Our extensive ligand
screenings suggested that the desired intramolecular cascade
reaction was favored over the undesired direct coupling reaction
only when a ligandless condition was applied.7a In addition,
the results indicated that it was critical to use water as a
cosolvent to suppress the formation of polymeric byproducts,
which were likely formed as a result of slow transmetalation
between vinylic palladium and boronic acid. Thus, under our
optimized condition, 2 mol % Pd(OAc)2, 1.2 equiv of 3-ni-
troboronic acid 4, and 3 equiv of Na2CO3 at 70 °C, the desired
product 9 was isolated in 89% yield after column chroma-
tography.7a
the end of the reaction
The single largest impurity was identified as dimer 11 (∼5%),
which was formed as a result of homocoupling of 3-nitrophenyl
boronic acid 4. It had been reported that the exclusion of oxygen
should suppress the homocoupling of boronic acids in Suzuki
cross-coupling reactions.15 Since the second step of our cascade
cyclization (coupling of the vinylic palladium species with
boronic acid 4) was essentially a variant of the Suzuki reaction,
we next implemented a deoxygenation procedure to suppress
the formation of dimer 11. Thus, subsurface sparging was
implemented before and after Pd(OAc)2 addition. The results
revealed that this modification reduced the amount of dimer
11 to 1.4% at the end of the reaction.
To consume more expensive starting material 3, excess (1.2
equiv) of boronic acid 4 was used. Despite the extra boronic
acid, ∼4% of starting material 3 remained unreacted after the
reaction mixture was stirred at 70 °C for 12 h. Little conversion
was observed even if prolonged reaction time was applied.
Raising the reaction temperature pushed the reaction to comple-
tion; however, significant increase in impurity 12 (∼4% under
standard conditions) was observed. Since the remaining starting
material 3 could be efficiently rejected through crystallization
in the downstream chemistry, we made no further attempt to
drive the reaction to completion.
After fulfilling our initial goals of demonstrating proof of
concept on model substrate 8, we next investigated the carbo-
metalation reaction starting from alkyne 3 and boronic acid 4
(Scheme 3). Our challenges for the preparation of target
Scheme 3. Cyclization of 3
We next focused on monitoring and controlling the formation
of polymeric byproducts. The possibility for the formation of
polymeric byproducts during the carbometalation of alkynes
had been previously documented in the literature.3a However,
due to the fact that these byproducts could not be detected under
a variety of HPLC and NMR conditions, it was extremely
difficult to separate and characterize these byproducts. To better
monitor the yield of the reactions, we developed a quantitative
HPLC assay (measured against a pure standard) to determine
the purity of isolated products.16 Thus, water was added to the
reaction mixture at the end of the reaction to precipitate product
2. After filtration, compound 2 was collected and dried. The
potency of the isolated product was determined, and the results
are summarized in Table 1. Despite the excellent HPLC area
percentage at the end of the reaction, the potency corrected yield
with 2 mol % Pd(OAc)2 was only determined as 70% (Table
1, entry 3).
compound 1 were associated with the scale-up of the key
cyclization reaction. In particular, the challenges were associated
with defining the purification strategy to remove impurities
generated from the cyclization. We desired to establish a control
strategy that would afford target compound 1 without resorting
to chromatography.
The desired cyclization was carried out using ligandless
conditions: 2 mol % Pd(OAc)2, 1.2 equiv of 3-nitroboronic acid
and 3 equiv of Na2CO3 in dioxane/water at 70 °C. The reaction
was first run under brief nitrogen headspace sweep, and it
proceeded to give the desired product 2 with 82% in situ HPLC
area percentage. In addition to the desired product 2, we
observed several major impurities by HPLC (Scheme 4).13,14
(13) Impurities 10 and 11 were isolated by column chromatography. Their
1
structures were determined by H NMR.
(14) The corresponding direct Suzuki cross-coupling byproduct was
observed in <1%.
(15) (a) Moreno-Manas, M.; Pajuelo, F.; Plexats, R. J. Org. Chem. 1995,
60, 2346. (b) Wallow, T. I.; Novak, B. M. J. Org. Chem. 1994, 59,
5034. (c) Marck, G.; Villager, A.; Buchecker, R. Tetrahedron Lett.
1994, 35, 3277. (d) Miller, W. D.; Fray, A. H.; Quatroche, J. T.;
Sturgill, C. D. Org. Process Res. DeV. 2007, 11, 359.
We subsequently investigated the possibility of improving
the yield by suppressing the formation of polymeric byproducts.
(16) Because a heterogeneous reaction mixture was observed at the end of
the reaction, in situ quantitative HPLC assay was not used.
Vol. 13, No. 2, 2009 / Organic Process Research & Development
•
317

Table 1. Determination of potency corrected isolated yieldsa
impurity was removed to the aqueous layer during an aqueous
HCl wash. Final HCl salt formation further rejected impurity 3
(<0.2%) and afforded target compound 1 with >97% potency.
In addition, both palladium and platinum levels were determined
to be below <10 ppm in the isolated HCl salt. The strategy of
carrying out the majority of the impurity rejection at compound
1 stage controlled all of the impurities and afforded 74% overall
yield in the last two steps.
Pd loading
(mol %)
% dimer
in situ
yieldc
(%)
entry
11
HPLC areab (%)
1
2
3
4
0.1
0.5
2
0.8
1.1
89.3
87.6
87.4
85.6
83
77
70
61
1.4
10
14.6
a All reactions were run with 1.2 equiv of boronic acid 4, 3.0 equiv of Na2CO3,
and Pd(OAc)2 in 4:1 dioxane/water (0.1 M) at 70 °C with subsurface sparging.
b Remaining boronic acid 4 and dimer 11 were excluded from the calculations
since they were attributed to the extra boronic acid used. c Yield was determined
from the weight and potency of the isolated product. The product was isolated by
addition of water to the reaction mixture, followed by filtration to collect the
product. See Experimental Section for details.
Scheme 5. Synthesis of 1 ·HCl
Whereas the desired cascade reaction presumably was first order
in palladium concentration, it occurred to us that the undesired
polymer formation reactions were likely higher order in
palladium concentration. We therefore explored the options of
reducing palladium catalyst loading. When the catalyst loading
was reduced from 2 mol % to 0.5 mol %, the in situ HPLC
area percentage remained unchanged at the end of the reaction.
However, 77% potency corrected yield was obtained (Table 1,
entry 2). Further reducing the catalyst loading to 0.1 mol %
again afforded similar HPLC area percentage (entry 1). In this
case, we were delighted to observe 83% potency corrected yield.
To further demonstrate the trend, we tested the cyclization with
10 mol % of Pd(OAc)2. Although similar HPLC area percentage
was observed at the end of the reaction, significantly lower
isolated yield was obtained (Table 1, entry 4). The results from
Table 1 clearly revealed the trend that higher-level
palladium loading was correlated to higher level of
polymeric byproduct. It was also noteworthy that higher-
level of dimer impurity 11 was observed when 10 mol %
Pd(OAc)2 was employed, despite the fact that the same
subsurface sparging procedure was applied to all reaction
conditions. It was also worth mentioning that the similar
levels of impurity 12 and starting material 3 were observed
when the catalyst loading was changed from 0.1 mol %
to 10 mol %.
As described, technical grade 2 was isolated by water
addition followed by filtration at the end of the reaction. The
procedure provided an efficient rejection of remaining boronic
acid 4. However, little rejection of starting material 3, cross-
coupling impurity 12, and dimer 11 occurred during the process.
To further purify technical grade 2, attempts were made to
recrystallize the material under a variety of solvent systems.
Although starting material 3 could be easily rejected through
crystallization, the removal of impurity 12 proved to be
inefficient. To maximize the overall yield of 1, technical grade
2 was then forward processed to the next step.
Conclusion
In conclusion, we have identified and developed an efficient
and scaleable route for the synthesis of key intermediate (Z)-
3-(1-(8-bromodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)a-
niline hydrochloride (1·HCl) to support our selective nuclear
hormone receptor modulator program. The key reaction in-
volved a palladium-catalyzed intramolecular carbometalation
of an alkyne to form the desired dibenzoxapine-containing
tetrasubstituted exocyclic (Z)-alkene. Target compound 1·HCl
was prepared in 48% overall yield in five steps from com-
mercially available (5-bromo-2-iodophenyl)methanol (5). The
impurity control strategy was established in the final HCl salt
formation step. Furthermore, palladium and platinum levels
were sufficiently controlled to <10 ppm. The reactions de-
scribed were successfully scaled up at 12 L scale to provide
2.0 kg of target compound. We believe this is the first large-
scale pharmaceutical application described in the literature for
a palladium-catalyzed intramolecular carbometalation of an
alkyne.
Experimental Section
General. 1H NMR and 13C NMR spectra were recorded as
specified for each experiment with chemical shift recorded as
parts per million. Elemental analysis and high-resolution mass
spectrometry data were provided by the Physical Chemistry
group of Lilly Research Laboratories. Commercially available
reagents and solvents were used without further purification.
Compounds 1, 3, 6, and 8 were monitored using an Agilent
1100 series instrument equipped with a UV using the following
conditions: 40% acetonitrile 0.1% aqueous trifluoroacetic acid/
60% 0.1% aqueous trifluoroacetic acid to 70% acetonitrile 0.1%
aqueous trifluoroacetic acid/30% 0.1% aqueous trifluoroacetic
acid; flow rate: 1.0 mL/min; column temperature: 30 °C;
column: Zorbax SB-C8, 4.6 mm × 250 mm, 5 µm; detector:
254 nm. Compound 2 was monitored using an Agilent 1100
series instrument equipped with a UV using the following
To complete the synthesis of 1, compound 2 was reduced
under hydrogenation conditions. Sulfided platinum on carbon
(5% wt.) was selected as the catalyst to avoid the reduction of
the bromo functionality (Scheme 5). The reaction proceeded
to completion in THF at 50 psi within 2 h. The reaction mixture
was then filtered to remove residual platinum and palladium
catalysts. The solvent was then switched from THF to EtOAc
and an aqueous workup was applied. As expected, impurity
12, which was carried over from the previous step, was also
reduced under the reduction conditions. Fortunately, this
318
•
Vol. 13, No. 2, 2009 / Organic Process Research & Development

conditions: 5% acetonitrile 0.1% formic acid/95% 0.1% formic
acid in water to 100% acetonitrile 0.1% formic acid; flow rate:
0.5 mL/min; column: Zorbax SB-C8, 3.0 mm × 150 mm, 3.5
µm; detector: 254 nm.
HPLC analysis indicated that the reaction was complete. The
reaction mixture was filtered through Celite, and the residue
was rinsed with 100 mL of DMF. The filtrate was transferred
to a 2000 mL round-bottowm flask and was allowed to stir at
rt. Water (1 L) was added to the filtrate slowly to form a thick
slurry. The resulting slurry was stirred at rt for 2 h and filtered.
The solid was rinsed with water (150 mL), dried in vacuo at
45 °C to give 73.1 g of technical grade 3 as a yellow solid. To
a 1000 mL round-bottom flask equipped with a condenser,
nitrogen inlet, magnetic stirring, and a thermocouple were
charged technical grade 3 (73.1 g) and ethanol (365 mL). The
resulting slurry was heated at 80 °C, and a solution
formed. The solution was allowed to cool to room tempe-
rature. At ∼60 °C the product began to crystallize from
solution. The mixture was aged at room temperature
overnight (∼18 h) and was cooled to 5 °C for an additional
2 h. The solid was isolated by filtration and washed with
200 mL of cold ethanol. The solid was dried in vacuo at 45
°C to give the title compound 3 as an off-white solid (64.7
g, 152 mmol, 74% from 6). 1H NMR (DMSO-d6) δ
7.81-7.78 (m, 2 H), 7.51-7.49 (m, 1 H), 7.38-7.35 (m, 2
H), 7.06 (m, 1 H), 6.79-6.76 (m, 1 H), 5.21 (s, 2 H), 2.03
(s, 3 H). 13C NMR (DMSO-d6) δ 156.9, 140.7, 139.5, 134.0,
131.2, 130.5, 130.2, 123.6, 121.6, 121.5, 113.5, 94.0, 87.2,
76.4, 68.5, 4.7. Anal. Calcd for C16H12BrIO C, 45.00, H,
2.83, found C, 45.20, H, 2.79.
4-Bromo-2-((2-iodophenoxy)methyl)-1-(prop-1-ynyl)ben-
zene (2). To a 12-L three neck round-bottom flask equipped
with a condenser, nitrogen inlet, mechanical stirring, and
thermocouple were charged aryl iodide 3 (300 g, 0.702 mol),
3-nitrophenylboronic acid 4 (140.6 g, 0.842 mol), and sodium
carbonate (223 g, 2.11 mol). A solution of 5.6 L of 1,4-dioxane
and 1.4 L of water was added to the mixture. The resulting
slurry was sparged with subsurface nitrogen for 20 min.
Pd(OAc)2 (158 mg, 0.702 mmol) was added, and the reaction
was sparged with subsurface nitrogen for an additional 20 min.
The reaction mixture was warmed to 70 °C under nitrogen for
22 h. The reaction mixture was cooled to rt, and a slurry was
formed upon cooling. Water (7.5 L) was slowly added to the
reaction mixture, and the resulting thick slurry was stirred at rt
for 2 h. The reaction mixture was filtered, and the solid was
washed with water (900 mL). The solid was collected, vacuum-
dried at 40 °C to give compound 2 (302 g with 81% potency,
0.583 mol, 83% potency corrected yield). The material was
forward processed without further purification. 1H NMR (DMSO-
d6) δ 8.15 (t, 1 H, J ) 1.5 Hz), 8.02 (AB, 1 H, J ) 8.1 Hz, 2.4
Hz), 7.81 (d, 1 H, J ) 2.1 Hz), 7.74 (d, 1 H, J ) 8.1 Hz), 7.64
(AB, 1 H, J ) 7.8 Hz, 2.1 Hz), 7.51 (t, 1 H, J ) 8.1 Hz), 7.37
(d, 1 H, J ) 7.8 Hz), 6.95 (m, 1 H), 6.69 (d, 1 H, J ) 7.2 Hz),
6.47 (m, 2 H), 5.80 (d, 1 H, J ) 12.6 Hz), 5.00 (d, 1 H, J )
12.6 Hz), 2.075 (s, 3 H).
(5-Bromo-2-prop-1-ynyl-phenyl)-methanol (6). To a 12-L
three neck round-bottom flask equipped with an overhead stirrer
apparatus, a thermometer/thermocouple, and a nitrogen inlet was
charged alcohol 5 (480 g, 1.53 mol) and diethylamine (5.4 L,
52.2 mol). Copper (I) iodide (5.88 g, 30.8 mmol) was added,
followed by bis(triphenylphosphine)palladium(II) chloride (10.8
g, 15.3 mmol). Propyne gas was bubbled into the reaction
mixture for 45 min while the reaction temperature was increased
from rt to 32 °C. The reaction mixture was stirred at rt overnight.
HPLC analysis indicated that <1% of alcohol 5 remained. The
reaction mixture was filtered through water-wet Celite. The
residual catalyst on the Celite was washed with 1 L of EtOAc.
The combined filtrate was concentrated under reduced pressure
to a total volume of 500 mL. Additional EtOAc (3 L) was
added, and the mixture was washed with water (2 × 3 L)
followed by 10% brine solution (1 L). The organic layer was
dried over Na2SO4, filtered, and concentrated to give 358 g of
a dark-brown solid. The solid was purified via silica gel filtration
(975 g of silica gel, 5% EtOAc/hexanes as eluents) to afford
compound 6 as a pale-yellow solid (299 g, 1.33 mol, 87%). 1H
NMR (DMSO-d6) δ 7.60 (d, 1 H, J ) 2.0 Hz), 7.39 (m, 1 H),
7.26 (d, 1 H, J ) 8.0 Hz), 5.38 (t, 1 H, J ) 6.0 Hz), 4.57 (d,
2 H, J ) 6.0 Hz), 2.06 (s, 3 H). 13C NMR (DMSO-d6) δ 146.8,
133.7, 129.8, 129.4, 121.7, 120.1, 93.2, 76.5, 61.1, 4.5. HRMS
calcd for C10H9OBrNa [M + Na] 246.9734, found 246.9729.
4-Bromo-2-(2-iodo-phenoxy)methyl-1-(prop-1-ynyl)ben-
zene (3). To a 2000 mL round-bottom flask equipped with an
overhead stirrer apparatus, a thermometer/thermocouple, and a
nitrogen inlet was charged (5-bromo-2-prop-1-ynyl-phenyl)-
methanol (6) (46.0 g, 204 mmol) and toluene (460 mL). The
resulting mixture was cooled to 5-10 °C with an ice bath.
Phosphorus tribromide (25.0 mL, 264 mmol) was added
dropwise so the temperature did not rise above 15 °C. The
reaction mixture was then heated to 60 °C for 18 h. The reaction
mixture was then cooled to 5 °C in an ice/water bath and was
quenched slowly with ice water (200 mL) followed by saturated
aqueous sodium carbonate (300 mL). The biphasic reaction
mixture was then filtered through a thin bed of Celite, and the
residue was washed with ethyl acetate (120 mL). The filtrate
was separated, and the aqueous layer was extracted with ethyl
acetate (200 mL). The combined organic extracts were washed
with brine (400 mL), dried over magnesium sulfate, filtered,
and concentrated under reduced pressure to give benzyl bromide
7 as a yellow oil, which solidified upon standing (58.7 g, 204
mmol, 100%). The crude mixture was used in the next step
without further purification. 1H NMR (DMSO-d6) δ 7.74 (d, 1
H, J ) 2.0 Hz), 7.47 (m, 1 H), 7.32 (d, 1 H, J ) 8.5 Hz), 4.68
(s, 2 H), 2.08 (s, 3 H). 13C NMR (DMSO-d6) δ 141.8, 134.4,
133.0, 132.0, 123.1, 121.4, 94.3, 76.4, 32.1, 4.7.
(Z)-3-(1-(8-Bromodibenzo[b,e]oxepin-11(6H)-ylidene)eth-
yl)aniline hydrochloride (1 ·HCl). A hydrogenation vessel was
charged with 2 (76.5 g with 81% potency, 0.147 mol),
tetrahydrofuran (700 mL), triethylamine (20.5 mL, 14.9 g, 0.147
mmol), and platinum, 5% wt on C, sulfided (6.2 g, 1.6 mmol).
The reactor was sealed, evacuated and purged with hydrogen.
The reaction mixture was stirred under 50 psi hydrogen
To a 1000 mL round-bottom flask equipped with a nitrogen
inlet, magnetic stirrer, and a thermocouple were charged benzyl
bromide 7 (58.7 g, 204 mmol), 2-iodophenol (40.5 g, 184
mmol), powdered potassium carbonate (63.4 g, 460 mmol), and
DMF (400 mL). The resulting slurry was stirred at rt for 18 h.
Vol. 13, No. 2, 2009 / Organic Process Research & Development
•
319

atmosphere at rt for 2 h. The reaction vessel was evacuated
and flushed with nitrogen. HPLC analysis indicated that the
reaction was complete. The reaction mixture was filtered
through GFF paper, and the catalyst residue was washed with
150 mL of ethyl acetate. The filtrate was solvent exchanged to
1.2 L of ethyl acetate, and 900 mL of water was added. The
pH of the aqueous layer was adjusted to ∼1 through addition
of concentrated HCl solution. The mixture was transferred to a
separatory funnel, and the layers were separated. The organic
layer was washed with 1.2 L of saturated aqueous sodium
bicarbonate, followed by 600 mL of brine. Seventeen milliliters
of HCl (37 wt % in water) was added dropwise to the organic
layer. The mixture was heated at 50 °C, and 50% of the ethyl
acetate was removed by distillation. The resulting slurry was
then cooled to rt and stirred for 1 h. The slurry was filtered,
and the solid was washed with ethyl acetate (150 mL). Off-
white crystals were collected and dried to a constant weight to
afford 1·HCl (56.1 g, 0.131 mol, 89% yield from 2, 74% yield
from 3). 1H NMR (DMSO-d6) δ 9.77 (bs, 2 H), 7.80 (d, 1 H,
J ) 2.5 Hz), 7.62 (m, 1 H), 7.36 (d, 1 H, J ) 8.0 Hz),
7.25-7.18 (m, 3 H), 7.07 (d, 1 H, J ) 7.5 Hz), 6.96-9.92 (m,
1 H), 6.67 (m, 1 H), 6.55 (m, 1 H), 6.45 (m, 1 H), 5.73 (d, 1
H, J ) 12.0 Hz), 5.00 (d, 1 H, J ) 12.0 Hz), 2.01 (s, 3 H). 13C
NMR (DMSO-d6) δ 154.8, 144.6, 142.5, 135.8, 135.7, 134.0,
132.8, 131.88, 131.87, 129.8, 129.7, 129.3, 129.0, 125.2, 123.6,
121.7, 120.9, 120.5, 119.3, 68.2, 22.4. HRMS calcd for
C22H19BrNO [M + H] 392.0645, found 392.0642.
Acknowledgment
We thank Eric Crockett, Edward Sheldon, and Matt Thomas
for the analytical support. We are also grateful to WuXi
PharmaTech for demonstrating some of the chemistry in their
scale-up facility. We also acknowledge Drs. Scott May and
Tony Zhang for helpful discussions.
Received for review September 19, 2008.
OP800231B
320
•
Vol. 13, No. 2, 2009 / Organic Process Research & Development