Cross-coupling methods for the large-scale preparation of an imidazole - Thienopyridine: Synthesis of [2-(3-methyl-3H-imidazol-4-yl)-thieno[3,2-b]pyridin-7-yl] -(2-methyl-1H-indol-5-yl)-amine

Article abstract of DOI:10.1021/op0340457

The multihundred-gram synthesis of [2-(3-methyl-3H-imidazol-4-yl)-thieno[3,2-b]pyridin-7-yl] -(2-methyl-1H-indol-5-yl)-amine (1) is described utilizing a Stille cross-coupling of an iodothienopyridine (3) with 5-(tributylstannyl)-1-methylimidazole (11). Several cross-coupling methods were evaluated for the conversion of thienopyridine 3 to imidazole - thienopyridine 2, but only two were effective: the Stille coupling and a Negishi cross-coupling of the organozinc reagent derived from 2-(tertbutyldimethylsilyl)-1-methylimidazole and iodothienopyridine (3). The latter procedure worked well on laboratory scale (<50 g), but was capricious upon scale-up. The issues with scale-up of an organostannane reagent are discussed, including control and analysis of organotin levels.

Full text of DOI:10.1021/op0340457

Organic Process Research & Development 2003, 7, 676−683
Cross-Coupling Methods for the Large-Scale Preparation of an
Imidazole-Thienopyridine: Synthesis of [2-(3-Methyl-3H-imidazol-4-yl)-
thieno[3,2-b]pyridin-7-yl]-(2-methyl-1H-indol-5-yl)-amine
J. A. Ragan,*,† J. W. Raggon,† P. D. Hill,† B. P. Jones,† R. E. McDermott,† M. J. Munchhof,‡ M. A. Marx,‡
J. M. Casavant,‡ B. A. Cooper,‡ J. L. Doty,‡ and Y. Lu‡
Chemical Research and DeVelopment, DiscoVery Chemistry, Pfizer Gobal Research & DeVelopment,
Eastern Point Road, Groton, Connecticut 06340, U.S.A.
Scheme 1
Abstract:
The multihundred-gram synthesis of [2-(3-methyl-3H-imida-
zol-4-yl)-thieno[3,2-b]pyridin-7-yl]-(2-methyl-1H-indol-5-yl)-
amine (1) is described utilizing a Stille cross-coupling of an
iodothienopyridine (3) with 5-(tributylstannyl)-1-methylimida-
zole (11). Several cross-coupling methods were evaluated for
the conversion of thienopyridine 3 to imidazole-thienopyridine
2, but only two were effective: the Stille coupling and a Negishi
cross-coupling of the organozinc reagent derived from 2-(tert-
butyldimethylsilyl)-1-methylimidazole and iodothienopyridine
(3). The latter procedure worked well on laboratory scale (<50
g), but was capricious upon scale-up. The issues with scale-up
of an organostannane reagent are discussed, including control
and analysis of organotin levels.
of cancer. This motivated us to seek a practical synthesis of
this imidazole-thienopyridine, which could provide the
multikilogram quantities required for regulatory toxicology
studies and clinical evaluation.
Introduction
Angiogenesis is a requirement for tumor growth and
metastasis and occurs through several discrete biochemical
signaling pathways. One key pathway that initiates prolifera-
tion and migration of endothelial cells is signaling through
the vascular endothelial growth factor receptor-2 (VEGFR-
2).2 Therefore, small molecules that block this signaling
pathway through inhibition of VEGFR kinase activity could
potentially inhibit angiogenesis and tumor growth. [2-(3-
Methyl-3H-imidazol-4-yl)-thieno[3,2-b]pyridin-7-yl]-(2-methyl-
1H-indol-5-yl)-amine (1) recently emerged as a promising
Scheme 1 outlines our retrosynthesis, which identifies
chloropyridine 2 and iodothienopyridine 3 as key intermedi-
ates. The preparation of thienopyridine 4 was known,3 and
multikilogram quantities of this intermediate were available
from outside sources. Likewise, conversion of chloropyridine
2 to 1 was relatively straightforward. However, the cross-
coupling of a suitable N-methylimidazole derivative with
thienopyridine 3 or an appropriate derivative was extremely
challenging and demonstrates that, of the many transition
metal-catalyzed cross-coupling methods available, only a
portion are effective in a complex heterocyclic system such
as imidazole-thienopyridine 2.
Discussion
Scheme 2 describes our synthesis of thienopyridine 4,
which closely followed the literature synthesis.3 3-Ami-
nothiophene (6) was prepared by hydrolysis and decarboxy-
lation of 3-amino-2-carbomethoxythiophene.4 Condensation
with the adduct of triethylorthoformate and Meldrum’s acid
generated vinylogous carbamate 7, which was cyclized by
heating to 245 °C in diphenyl ether or Dowtherm. The
VEGFR kinase inhibitor (7 nM IC50 against VEGFR-2) and
was thus of interest for clinical evaluation in the treatment
† Chemical Research and Development.
‡ Discovery Chemistry.
(1) To whom correspondence should be addressed: Pfizer Global Research &
Development, Mailstop 8118D-4010, Groton, CT, 06340, U.S.A. Tele-
phone: (860) 441-6334. Fax: (860) 715-9469. E-mail: john_a_ragan@
groton.pfizer.com.
(2) (a) Fan, T. P. D.; Jaggar, R.; Bicknell, R. Trends Pharmacol. Sci. 1995, 16,
57-66. (b) Folkman, J. Nat. Med. 1995, 1, 27-31.
(3) (a) Barker, J. M.; Huddleston, P. R.; Holmes, D.; Keenan, G. J.; Wright, B.
J. Chem. Res. Miniprint 1984, 771-795. (b) Barker, J. M.; Huddleston, P.
R.; Holmes, D.; Keenan, G. J. J. Chem. Res. Miniprint 1982, 1726-1746.
(4) Barker, J. M.; Huddleston, P. R.; Wood, M. L. Synth. Commun. 1995, 25,
3729-3734.
676
•
Vol. 7, No. 5, 2003 / Organic Process Research & Development
10.1021/op0340457 CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/01/2003

Scheme 2
Scheme 3
initially generated 3-carboxy-4-pyridone underwent decar-
boxylation under these conditions, thus avoiding a separate
hydrolysis/decarboxylation (this strategy is well-precedent-
ed).5 Treatment of pyridone 8 with oxalyl chloride in
dichloroethane with DMF activation then provided 4-chlo-
ropyridine 4, which was converted to iodothienopyridine 3
via deprotonation with n-BuLi and trapping with I2.
Our first synthesis of 1 utilized a Stille cross-coupling6
with iodide 3 as shown in Scheme 3 (eq 1).7 From a scale-
up perspective, we were concerned with the issues associated
with the use of a stoichiometric organostannane reagent (e.g.,
tank contamination, contamination of drug substance), and
sought an alternative method. An organozinc coupling (eq
2, Scheme 3) was identified as an early alternative and
provided an effective, nonstannane coupling on laboratory
scale (up to 50-g batches of 3 could be coupled). Unfortu-
nately, the reaction was capricious, and further scale-up led
to reactions that would occasionally (and unpredictably) fail
completely.
Due to the toxicity concerns with the organostannane
method and the capriciousness of the organozinc method,
we examined a variety of alternative cross-coupling ap-
proaches, as summarized in Table 1. Despite these efforts,
only two methods were identified which successfully pro-
vided coupled product on laboratory scale (entries 1 and 12).
Of those, only the Stille coupling (entry 12) proved reliable
on >50-g scale.
The failure of these studies to identify non-stannane
alternatives was frustrating, particularly since the Negishi
coupling (entry 1) worked fairly reliably on 5-10-g scale.
The presence of several heteroatoms in both substrates and
the product renders them potential metal-chelators, and is a
possible culprit (although why this would be scale-dependent
is not immediately clear). Another potential issue was
(5) (a) Cassis, R.; Tapia, R.; Valderrama, J. A. Synth. Commun. 1985, 15, 125-
134. (b) Andrew, R. G.; Raphael, R. A. Tetrahedron 1987, 43, 4803-4816.
(c) Moron, J.; Huel, C.; Bisagni, E. Heterocycles 1993, 36, 2753-2764.
(d) Marcos, A.; Pedregal, C.; Avendano, C. Tetrahedron 1995, 51, 1763-
1774. (e) Toedter, C.; Lackner, H. Synthesis 1997, 567-572.
(6) For reviews, see: (a) Mitchell, T. N. In Metal-catalyzed Cross-coupling
Reactions; Diederich, F., Stang, P. J., Eds.; Wiley-VCH Verlag: Weinheim,
Germany, 1998; Chapter 4. (b) Farina, V.; Krishnamurthy, V.; Scott, W. J.
Org. React. 1997, 50, 1-652. (c) Stille, J. K. Angew. Chem., Int. Ed. Engl.
1986, 25, 508-524.
(7) Similar Stille and Negishi couplings with 4-susbstituted N-alkylimidazoles
(vs 5-substituted in the present study) have been reported: (a) Jetter, M.
C.; Reitz, A. B. Synthesis 1998, 829-831. (b) Cliff, M. D.; Pyne, S. G.
Tetrahedron 1996, 52, 13703-13712.
Vol. 7, No. 5, 2003 / Organic Process Research & Development
•
677

Table 1.
isolation of byproducts which appeared to arise from
competitive lithiation of one of the two methyl groups
attached to the TBS protecting group, suggesting that the
deprotonation of the TBS-imidazole might have been part
678
•
Vol. 7, No. 5, 2003 / Organic Process Research & Development

Table 2. Tin and Palladium Levels in Laboratory Pilotsa
Sn level
(ppm)
Pd level
(ppm)
entry
compound
purification method
1
2
3
4
2
2
1
1
aqueous pH workup, then reslurry with MTBE
(57% yield)
recrystallization of entry 1 material from ethyl acetate
(73% recovery)
aminoindole coupling of 2 (from entry 1), followed by crystallization from MeOH
(isolated HCl salt, 37% yield)
aminoindole coupling of 2 (from entry 1), followed by silica gel chromatography
(isolated free base, 50% yield)
154
19
5
52
38
7
2
<1
a For clarification, note that entries 2-4 all involve further processing of the material from entry 1: recrystalliztion (entry 2), aminoindole coupling followed by
recrystallization (entry 3), or aminoindole coupling followed by silica gel chromatography (entry 4).
of the problem for cases which required its intermediacy (e.g.,
entries 1, 5, 7, 10, and 11).
we could bypass recrystallization of 2, and proceed directly
into the final aminoindole coupling, with isolation of this
product providing further reduction in tin levels. This hope
was born out by the following two entries, which show that
utilization of the 154 ppm lot of 2 in the aminoindole
coupling provides acceptable purity by either crystallization
of the HCl salt (entry 3, 5 ppm), or silica gel chromatography
(entry 4, 2 ppm). On the basis of these results, we were
confident that we would be able to purify our bulk campaign
material to <20 ppm stannane, and this prediction was
confirmed in the GMP campaign (vide infra).
For the cGMP campaign (Scheme 4), we began with 3
kg of thienopyridine 4.3 Metalation of this material was
achieved by treatment with n-BuLi (1.6 equiv) in THF-
hexane at -70 °C for 60 min (D2O quench of an aliquot
showed >95% deuterium incorporation by 1H NMR analy-
sis), followed by addition of I2 (1.6 equiv) in THF at such a
rate that the internal temperature remained below -65 °C.
Addition of water precipitated iodide 3 as a light brown solid,
which was then washed with water and hexanes to provide
an 84% yield of 3 with >99% HPLC (area %) purity.
Running the metalation sequence at -20 °C led to increased
levels (5-10%) of a bis-iodide impurity by mass spectrom-
etry analysis. The sequential triturations were found to
With the Stille coupling as the only robust method
identified in our studies, we decided to utilize this method
for preparation of the initial cGMP bulk lot. The major issues
we faced with this approach were tank contamination and
contamination of drug substance with residual organostan-
nane byproducts. The former issue was addressed by limiting
our runs to 22 L glassware, so that we could simply dispose
of the reaction vessel when the campaign was completed.
While obviously not a long-term solution to the problem,
this strategy was acceptable for an initial campaign of <2
kg. For the latter issue, inductively coupled plasma emission
spectroscopy (ICP) analysis was used to determine total
stannane and palladium content to a lower limit of <2 ppm,
which was sufficient for releasing drug substance (the upper
limit for stannane content was set at 20 ppm by our
toxicologists, taking into consideration the proposed clinical
doses). Fortunately, both 2 and 1 are solids which crystallize
to purge residual organic impurities quite efficiently.
Table 2 summarizes tin and palladium levels from several
laboratory-scale experiments. Entry 1 shows that a simple
reslurry of the crude Stille product removes the bulk of the
residual stannane (154 ppm, vs an initial level of ca. 170 000
ppm based on the atomic weight of tin, the molecular weights
of the reactants, and the stoichiometry of the reaction).16 This
material can be further purified by recrystallization to just
below our target tin level (entry 2, 19 ppm). We hoped that
(11) For reviews, see: (a) Hiyama, T. in Metal-catalyzed Cross-coupling
Reactions; Diederich, F., Stang, P. J., Eds.; Wiley-VCH Verlag: Weinheim,
Germany, 1998; Chapter 10. (b) Hiyama, T.; Hatanaka, Y. Pure Appl. Chem.
1994, 66, 1471-1478.
(12) For reviews, see (a) Suzuki, A. J. Organomet. Chem. 1999, 576, 147-168.
(b) Miyaura, N. In AdVances in Metal-Organic Chemistry; Liebeskind, L.
S., Ed.; JAI, London, 1998; Vol. 6, pp 187-243. (c) Suzuki, A. In Metal-
catalyzed Cross-coupling Reactions; Diederich, F., Stang, P. J., Eds.; Wiley-
VCH Verlag: Weinheim, Germany, 1998; Chapter 2. (d) Miyaura, N.;
Suzuki, A. Chem. ReV. 1995, 95, 2457-2483.
(8) For reviews, see: (a) Negishi, E.; Liu, F. In Metal-catalyzed Cross-coupling
Reactions; Diederich, F., Stang, P. J., Eds.; Wiley-VCH Verlag: Weinheim,
Germany, 1998; Chapter 1. (b) Erdik, E. Tetrahedron 1992, 48, 9577-
9648.
(9) (a) Brase, S.; de Meijere, A. In Metal-catalyzed Cross-coupling Reactions;
Diederich, F., Stang, P. J., Eds.; Wiley-VCH Verlag: Weinheim, Germany,
1998; Chapter 3. (b) Pivsa-Art, S.; Satoh, T.; Kawamura, Y.; Miura, M.;
Nomura, M. Bull. Chem. Soc. Jpn. 1998, 71, 467. (c) Aoyagi, Y.; Inoue,
A.; Koizumi, I.; Hashimoto, R.; Tokunaga, K.; Gohma, K.; Komatsu, J.;
Sekine, K.; Miyafuji, A.; Kunoh, J.; Honma, R. Akita, Y., Ohta, A.
Heterocycles 1992, 33, 257.
(13) (a) Tamao, K.; Kodama, S.; Nakajima, I.; Kumada, M.; Minato, A.; Suzuki,
K. Tetrahedron 1982, 38, 3347-3354.
(14) Kobayashi, Y.; Ikeda, E. J. Chem. Soc., Chem. Commun. 1994, 1789.
(15) Mowery, M. E.; DeShong, P. J. Org. Chem. 1999, 64, 1684-1688.
(16) Based on the following calculation: (at. wt Sn) × (equiv 11)/(mol wt 3) ×
(equiv 3) + (mol wt 11) × (equiv 11) + (mol wt Pd(Ph3P)4) × (equiv cat.)
) (118 × 1.1)/(295 × 1.0) + (370 × 1.1) + (1155 × 0.05) ) 0.17 (17%)
) 170 000 ppm.
(10) Gurtler, C.; Buchwald, S. L. Chem. Eur. J. 1999, 5, 3107-3112.
Vol. 7, No. 5, 2003 / Organic Process Research & Development
•
679

Scheme 4
After our cGMP campaign, it was found that running the
aminoindole coupling in EtOH provided much cleaner and
more rapid conversion (this was not tried initially as literature
precedent suggested that ethoxide would displace the 4-
chloropyridine).3,18 Thus, combining equimolar amounts of
chloropyridine 2 and the aminoindole in refluxing EtOH for
48 h provided the desired product, which crystallized from
the reaction upon cooling in 87% yield with >99% HPLC
(area %) purity (25 g scale).
Conclusions
We have utilized a Stille coupling to prepare imidazole-
thienopyridine 2, which was then converted to clinical
candidate 1. Organostannane levels were controlled (<20
ppm) in large part due to the crystallinity of 2, 1, and salts
of 1 (HCl, camsylate). Of several methods examined, the
Stille coupling was uniquely suited to provide a robust and
scaleable cross-coupling method. This suggests that the wide
variety of cross-coupling methods demonstrated on simpler
biaryl systems is more limited when applied to complex
heterocyclic systems.
provide a more efficient isolation and purification than an
aqueous workup (e.g. partitioning between EtOAc and
aqueous Na2S2O3).
The organostannane component of the Stille coupling was
5-(tributylstannyl)-1-methylimidazole (11). This material was
prepared from N-methylimidazole, per the literature,17 as
shown in Scheme 5. The 2,5-dilithioimidazole was formed
by metalation in TMEDA-hexane at -20 °C, and quenched
with Bu3SnCl. Aqueous workup cleaves the more labile
2-stannyl moiety to provide 11. This material could be used
in crude form in lab pilots, but on scale it was deemed
prudent to remove Bu3SnCl-derived impurities via a hexane-
acetonitrile partition; the nonpolar stannane impurities were
selectively partitioned into the hexane phase, while the more
polar imidazole reagent 11 remained in the acetonitrile phase.
This method provided 1.7 kg of 11.
The Stille coupling was effected with 5 mol % Pd(Ph3P)4
in DMF at 90 °C. Unlike the other coupling methods
investigated, this reaction was robust and nondiscriminating
in the source of catalyst, and scaled from 10 to 500 g with
no significant change in isolated yield (63-67% on 530 g
scale).
Coupling of 5-amino-2-methylindole with chloropyridine
2 was accomplished in a Parr reactor in tert-butyl alcohol
and dichloroethane. High concentration (2 M in 1:1 t-BuOH/
DCE), excess indole (2 equiv), and high temperature (100
°C, 17 psi) were critical for this reaction to be driven to
reasonable (>90%) conversion. A silica gel column was
required to purify this reaction mixture (11 g of silica gel
per g of starting material 2). Following a reslurry from
2-propanol, a 65% yield of 1 was isolated (500-g scale).
Conversion to the (-)-camphorsulfonic acid salt then
provided the desired drug substance, ICP analysis of which
showed 3 ppm stannane and 3 ppm palladium, consistent
with the lab pilots described in Table 2.
Experimental Section
Palladium tetrakis(triphenylphosphine) was ordered from
Strem Chemicals (Newburyport, MA, 01950-4098). All other
chemicals were ordered from commercial suppliers and used
as received. For laboratory-scale experiments, N,N-dimeth-
ylformamide (DMF) and tetrahydrofuran (THF) were pur-
chased from Aldrich in anhydrous “Sure-Seal” glass bottles;
all other solvents were reagent grade. Laboratory-scale
reactions were run under a positive pressure of nitrogen in
glassware which was flame-dried under nitrogen. Reaction
progress was monitored by TLC, GC/MS, or HPLC. HPLC
purity refers to area %, and is uncorrected. TLC was
performed on precoated sheets of 60 F254 (Merck Art. 5719),
and visualized by UV, and/or staining with iodine, phos-
phomolybdic acid, ceric ammonium molybdate, or p-anis-
aldehyde solutions and heating. GC analyses were performed
on a Hewlett-Packard 6890 GC/MS with a 5973 mass
selective detector. Mass spectral data was collected on either
a Hewlett-Packard 6890 GC/MS (electron impact ionization),
or a Micromass (Fisons) Platform II mass spectrometer
1
(atmospheric pressure chemical ionization). H (400 MHz)
and 13C (100 MHz) NMR spectra were obtained on a Varian
Unity+400 spectrometer equipped with two RF channels,
indirect detection, and pulsed-field gradients (z-axis only).
Melting points are uncorrected. Combustion analyses were
performed by Schwarzkopf Microanalytical Laboratory
(Woodside, NY) or Quantitative Technologies, Inc. (White-
house, NJ). Tin and palladium analyses were done at
Quantitative Technologies, Inc. by inductively coupled
plasma emission spectroscopy (ICP) analysis. Samples were
Scheme 5
680
•
Vol. 7, No. 5, 2003 / Organic Process Research & Development

digested in a mixture of aqueous H2SO4-HNO3-HClO4 with
heating, then diluted with aqueous HCl in volumetric flasks.
Multiple runs of the samples as well as positive controls
demonstrated good reproducibility ((10%). The LLOQ
(lower limit of quantification) was demonstrated to be <2
ppm.
0.96 (s, 9H), 0.41 (s, 6H); MS (EI): m/z 196 (M, 5), 139
(M - C4H9, 100).
2-[2-(tert-Butyl-dimethylsilyl)-3-methyl-3H-imidazol-4-
yl]-7-chloro-thieno[3,2-b]pyridine (9). A 5 L, four-neck,
round-bottom flask equipped with two 500-mL addition
funnels was flame-dried under a positive pressure of nitrogen.
The flask was charged with 2-(tert-butyldimethylsilyl)-N-
methylimidazole 10 (66.5 g, 338 mmol). Anhydrous THF
(1.1 L) was added, and the flask was immersed in a room-
temperature water bath. A solution of sec-BuLi (1.3 M in
cyclohexane, 274 mL, 355 mmol) was then added dropwise
via addition funnel at a rate such that the temperature
remained at or below 25 °C. The reaction was stirred for an
additional 2.5 h at room temperature, then cooled to -78
°C in a dry ice-acetone bath. A solution of ZnCl2 (0.5 M in
THF, 744 mL, 372 mmol) was then added via addition funnel
at a rate such that the temperature remained at or below -65
°C. Upon complete addition, the cold bath was removed,
and the reaction was allowed to warm to room temperature
and was stirred for 60 min. Iodothienopyridine 3 (50.0 g,
169 mmol) and Pd(PPh3)4 (19.6 g, 17 mmol, 10 mol %) were
then added, and the reaction was warmed to reflux for 45
min. [Note: as discussed in the text, this coupling was
capricious; the best success was obtained with batches of
iodothienopyridine 3 that had been triturated with hot
isopropyl ether, filtered, and vacuum-dried within 24 h of
use]. HPLC analysis indicated complete conversion to
product, at which point the reaction mixture was allowed to
cool to room temperature. The reaction mixture was carefully
poured into 2 N NH4OH (4 L), and extracted with three 1.4-L
portions of CHCl3. The combined organic extracts were
washed with brine (1.2 L), dried over Na2SO4, filtered, and
concentrated to provide a solid. This material was triturated
with 400 mL of hexanes, cooling in a 0 °C ice bath. The
solids were collected and rinsed with 75 mL of cold hexanes,
and dried in a vacuum oven to provide 25-30 g (69-82
mmol, 41-49%) of the product 9 as a tan solid: 1H NMR
(CDCl3): δ 8.61 (d, J ) 5, 1H), 7.57 (s, 1H), 7.55 (s, 1H),
7.31 (d, J ) 5, 1H), 3.93 (s, 3H), 1.04 (s, 9H), 0.49 (s, 6H);
MS (EI): m/z 363 (M, 10), 306 (M - C4H9, 100).
N-Methyl-5-(tributylstannyl)-imidazole (11). A 22 L,
three-neck, round-bottom flask equipped with a mechanical
stirrer and addition funnel was charged with tetramethyleth-
ylenediamine (1.99 L, 13.2 mol). After cooling to -20 °C,
n-BuLi (2.5 M in hexane, 5.26 L, 13.2 mol) was added over
90 min, maintaining the temperature between -10 and -20
°C. After stirring 20 min, a solution of N-methylimidazole
(450 g, 5.48 mol) in THF (2.6 L) was added over 60 min,
maintaining the temperature below -10 °C. Cooling was
removed, and the resulting yellow suspension was allowed
to warm to 20 °C over 3 h. The reaction mixture was then
cooled to -20 °C, and Bu3SnCl (3.72 L, 13.7 mol) was added
over a period of 2 h. The resulting solution was allowed to
warm to 20 °C over 16 h. Water (5.7 L) was added, vigorous
stirring was maintained for 30 min, and then the layers were
allowed to settle. The top organic layer was separated and
washed with water (3 L). The combined aqueous phases were
extracted with 4 L of ethyl acetate. The organic extracts were
7-Chloro-2-iodo-thieno[3,2-b]pyridine (3). A 75-L Has-
telloy reactor was charged with 7-chloro-thieno[3,2-b]pyri-
dine 4 (3.00 kg, 17.7 mol) and THF (30 L), and the solution
was cooled to -70 °C. A solution of n-BuLi (2.5 M in
hexane, 11.3 L, 28.3 mol) was added over 60 min. The
reaction mixture was stirred for an additional 60 min and
sampled for completion (quench into D2O and analyzed by
1H NMR), which indicated >95% deprotonation. A solution
of I2 (7.18 kg, 28.3 mol) in THF (12 L) was then added to
the lithium anion solution over 80 min, maintaining the
temperature between -65 and -70 °C. The reaction mixture
was allowed to warm to 20 °C over 18 h, at which point
HPLC analysis indicated >95% conversion to iodide 3.
Water (75 L) was added over a period of 10 min, and the
resulting slurry was stirred for 3 h at 20 °C. The resulting
solids were filtered, washed with water (5 L) and hexanes
(3 portions of 2.5 L), and dried under vacuum at 40 °C,
providing 4.39 kg of iodide 3 as an off-white solid (14.9
mol, 84% yield). HPLC analysis indicated a purity of 99.5%;
mp ) 187-189 °C; IR (neat): 3089, 1564, 1531, 1353, 830
1
cm-1; H NMR (CDCl3): δ 8.55 (d, J ) 5, 1H), 7.84 (s,
1H), 7.26 (d, J ) 5, 1H); 13C NMR (CDCl3): δ 157.9, 148.4,
138.1, 136.8, 135.5, 119.0, 85.6; MS (EI): m/z 295 (M, 100),
260 (M - Cl, 15). Anal. Calcd for C7H3NSClI: C, 28.45;
H, 1.02; N, 4.74; S, 10.85; Cl, 12.00; I, 42.94. Found: C,
28.48; H, 0.86; N, 4.63; S, 11.04; Cl, 11.98; I, 43.05.
2-tert-Butyl-dimethylsilyl-N-methylimidazole (10). (Al-
though previously reported in the literature,19 experimental
details were not provided). A flame-dried, 250-mL round-
bottom flask was charged with N-methylimidazole (4.0 mL,
50 mmol) and 40 mL of THF. The reaction mixture was
cooled to -78 °C, and n-BuLi (2.5 M in hexane, 22.0 mL,
55 mmol) was added dropwise via syringe. The reaction
mixture was stirred for an additional 45 min at -78 °C, and
then tert-butyldimethylsilyl chloride (9.0 g, 60 mmol) was
added as a solution in 10 mL of THF. The solution was
stirred for 30 min at -78 °C and then allowed to warm to
room temperature overnight. The reaction mixture was
quenched by pouring into ice-cold, aqueous NH4Cl (ca. 200
mL), rinsing with isopropyl ether (IPE). The layers were
separated, and the aqueous was extracted with another 100-
mL portion of IPE. The combined organic extracts were
washed with brine, dried over MgSO4, filtered, and concen-
trated to provide a yellow oil (10.9 g, 110% of theory due
to residual solvent). This material was suitable for use in
1
the subsequent coupling without further purification. H
NMR (CDCl3): δ 7.22 (s, 1H), 6.99 (s, 1H), 3.77 (s, 3H),
(17) Gaare, K.; Repstad, T.; Benneche, T.; Undheim, K. Acta Chem. Scand. 1993,
47, 57-62.
(18) (a) Barker, J. M.; Huddleston, P. R.; Holmes, D.; Keenan, G. J.; Wright, B.
J. Chem. Res. Miniprint 1984, 771-795. (b) Cutler, R. A.; Surrey, A. R. J.
Am. Chem. Soc. 1955, 77, 2441-2444.
(19) Walters, M. A.; Lee, M. D. Tetrahedron Lett. 1994, 35, 8307-8310.
Vol. 7, No. 5, 2003 / Organic Process Research & Development
•
681

combined, dried over MgSO4, filtered, and concentrated
under vacuum to provide 4.44 kg (218% of theoretical) of
crude 11 as a yellow oil. This material was divided into two
roughly equivalent portions for the hexane-acetonitrile
extractive purification. Each portion was partitioned into
hexanes (9 L) and CH3CN (7 L). The top hexane layer was
separated and extracted with four portions of CH3CN (6, 4,
3, and 3 L). The CH3CN extracts were combined and washed
with hexanes (4 L). Concentration of the combined CH3CN
extracts under vacuum provided 1.70 kg (4.58 mol, 84%
yield) of 11 as a colorless oil. This material was used in the
Stille coupling with no further purification. 1H NMR
(CDCl3): δ 7.64 (s, 1H), 7.05 (s, 1H), 3.70 (s, 3H), 1.60-
1.50 (m, 6H), 1.42-1.30 (m, 6H), 1.15-1.09 (m, 6H), 0.92
(t, J ) 7, 9H).
7-Chloro-2-(3-methyl-3H-imidazol-4-yl)-thieno[3,2-b]-
pyridine (2). Method A (Laboratory Pilot). A 250-mL round-
bottom flask was charged with stannane 10 (15.7 g, 42.2
mmol), iodothienopyridine 3 (11.2 g, 37.9 mmol), Pd(Ph3P)4
(2.19 g, 1.89 mmol, 5 mol %), and DMF (120 mL). After
purging with nitrogen, the solution was placed in a 90 °C
oil bath for 22 h. The solution was then cooled to room
temperature, diluted with 240 mL of 1 N HCl, and extracted
with three 50-mL portions of EtOAc (the product remains
in the acidic aqueous phase). The aqueous phase was adjusted
to pH 10 with 2 N NaOH and then extracted with three 50-
mL portions of CH2Cl2. The organic extracts were combined
and dried over MgSO4. Filtration and concentration provided
an oily, tan solid. Addition of 100 mL of methyl-tert-butyl
ether (MTBE) and stirring overnight provided a tan solid,
which was filtered to provide 5.39 g of product 2, ICP
analysis of which indicated 154 ppm tin (entry 1 in Table
2). One gram of this material was recrystallized from hot
EtOAc to provide a light tan solid: 0.73 g, ICP analysis
showed 19 ppm tin (entry 2 in Table 2).
Method B (GMP Bulk Campaign). A 22-L, three-neck,
round-bottom flask equipped with a mechanical stirrer was
charged with stannane 10 (672 g, 1.81 mol), iodothieno-
pyridine 3 (535 g, 1.81 mol), Pd(Ph3P)4 (105 g, 0.091 mol,
5 mol %), and DMF (2.7 L), and heated to 95 °C under N2.
After 40 h, HPLC analysis indicated complete conversion.
The reaction mixture was cooled to 10 °C and quenched by
the addition of 1 N HCl (5.3 L). Ethyl acetate (4.2 L) was
added, and the mixture was filtered. The layers were
separated, and the aqueous phase was extracted with two
4-L portions of ethyl acetate. The combined organic layers
were washed with water (3 L). The aqueous extracts were
combined, the pH was adjusted to 10-10.5 by addition of 5
N NaOH, and extracted with three 4.5-L portions of 1:1
THF-EtOAc (v/v). The aqueous pH was monitored after
each extraction and readjusted to pH 10-10.5 as needed.
HPLC analysis indicated essentially complete extraction of
product from the aqueous phase at this point. The organic
extracts were combined, washed with water (three portions
of 4 L each) and brine (2 L), and concentrated under vacuum
to provide a tacky solid. MTBE (4 L) was added, and the
mixture was concentrated under vacuum. An additional 3 L
of MTBE was then added, and the resulting slurry was stirred
for 2 h. The solids were collected by filtration, rinsing with
MTBE. After drying at 40 °C under vacuum, the product
was obtained as an off-white solid (302 g, 1.21 mol, 67%
yield); mp ) 178-180 °C; IR (neat): 3084, 1533, 1265,
1123, 838 cm-1; 1H NMR (CDCl3): δ 8.61 (d, J ) 5, 1H),
7.63 (s, 1H), 7.57 (s, 1H), 7.45 (s, 1H), 7.31 (d, J ) 5, 1H),
3.91 (s, 3H); 13C NMR (CD3OD): δ 157.2, 148.4, 141.6,
137.9, 136.8, 132.6, 129.7, 126.3, 121.4, 119.1, 32.7; HRMS
[MH]+ (m/z) for C11H8ClN3S calcd 250.0205; obsd 250.0199;
Anal. Calcd for C11H8ClN3S: C, 52.91; H, 3.23; N, 16.83;
S, 12.84; Cl, 14.20. Found: C, 52.56; H, 2.92; N, 16.55; S,
12.75; Cl, 14.44.
[2-(3-Methyl-3H-imidazol-4-yl)-thieno[3,2-b]pyridin-7-
yl]-(2-methyl-1H-indol-5-yl)-amine (1). Method A (Labora-
tory Pilot, Isolation by Crystallization of HCl Salt: Table
2, Entry 3). A resealable tube reactor was charged with
chloropyridine 2 (1.00 g, 4.00 mmol), 5-amino-2-methylin-
dole (1.17 g, 8.00 mmol), tert-butyl alcohol (1.0 mL), and
dichloroethane (1.0 mL). The reactor was then placed in a
100 °C oil bath for 60 min, at which point HPLC analysis
indicated <5% starting material. The reaction was cooled
to room temperature, and the dark material was dissolved in
5 mL of MeOH. After stirring overnight, a green solid was
isolated by filtration to provide 1.21 g of crude product. This
material was redissolved in a minimum volume of hot MeOH
and cooled to room temperature. The resulting yellow solids
were collected to provide 0.59 g of product 1-HCl (37%
yield); ICP analysis indicated 5 ppm tin (Table 2, entry 3).
1H NMR (DMSO-d6): δ 11.3 (s, 1H), 10.7 (s, 1H), 8.26 (d,
J ) 7, 1H), 7.94 (s, 1H), 7.61 (s, 1H), 7.38-7.32 (m, 3H),
6.95 (dd, J ) 8, 2, 1H), 6.77 (d, J ) 8, 1H), 6.10 (s, 1H),
3.76 (s, 3H), 2.37 (s, 3H).
Method B (Laboratory Pilot, Isolated by Silica Gel
Chromatography of Free Base: Table 2, Entry 4). A
resealable tube reactor was charged with chloropyridine 2
(1.00 g, 4.00 mmol), 5-amino-2-methylindole (1.17 g, 8.00
mmol), tert-butyl alcohol (1.0 mL), and dichloroethane (1.0
mL). The reactor was then placed in a 100 °C oil bath for
60 min, at which point HPLC analsysis indicated <5%
starting material. The reaction was cooled to room temper-
ature, and the dark material was dissolved in 5 mL of MeOH.
After stirring overnight, a green solid was isolated by
filtration to provide 0.98 g of crude product. This material
was redissolved in MeOH and coated onto 2.5 g silica gel;
the product/silica gel mixture was placed on top of a 5-g
silica gel column and eluted with 94:5:1 CH2Cl2-MeOH-
Et3N. The product-containing fractions were combined,
concentrated, and triturated with a minimal volume of CH2-
Cl2 to provide product 1 (free base) as a light yellow solid:
0.72 g (50% yield); ICP analysis indicated 2 ppm tin (Table
2, entry 4). IR (neat): 3327-2959 (br), 1574, 1552, 1516,
1
1489, 1460, 1365, 1303, 1185, 1125, 1113 cm-1; H NMR
(CD3OD): δ 8.12 (d, J ) 6, 1H), 7.80 (s, 1H), 7.45 (s, 1H),
7.39 (s, 1H), 7.34 (d, J ) 8, 1H), 7.28 (s, 1H), 6.99 (d, J )
8, 1H), 6.70 (d, J ) 6, 1H), 6.16 (s, 1H), 3.85 (s, 3H), 2.44
(s, 3H).
Method C (Bulk Campaign, Isolated as Free Base). A
2-L Parr pressure reactor was charged with chloropyridine
682
•
Vol. 7, No. 5, 2003 / Organic Process Research & Development

2 (260 g, 1.04 mol), 5-amino-2-methylindole (304 g, 2.08
mol), tert-butyl alcohol (260 mL), and dichloroethane (260
mL). The reactor was sealed, purged with nitrogen, and
heated to 90-100 °C for 2 h (the pressure rose to 17 psi
during heating). The reactor was cooled to 60 °C and opened,
and the very thick (nearly solid) reaction slurry was
transferred out of the reactor, rinsing with methanol. This
process was repeated on the same scale, such that a total of
520 g (2.08 mol) of chloropyridine 2 was processed.
The crude product from each run was purified by silica
gel chromatography: the crude product in ca. 8 L of MeOH
was treated with 3 kg of silica gel and concentrated on a
rotary evaporator to a dark brown solid. An additional 2-L
portion of dichloroethane was added and concentrated to
solids to facilitate removal of most of the MeOH. The solids
were then charged to the top of a 20-gal glass chromatog-
raphy column previously charged with 10 gal of CH2Cl2 and
9 kg of silica gel. The column was eluted with a 98:2 (v/v)
CH2Cl2-MeOH solution to remove less polar impurities
(assayed by TLC), collecting 7-L fractions. The eluant was
then changed to a 93.5:5:1.5 CH2Cl2-MeOH-Et3N mixture
to elute the desired product. The product-containing fractions
(26-36 in the case of one column) were combined and
concentrated. This chromatography was repeated for the
second reaction, and the product-containing fractions from
both columns were combined and concentrated (distillation
under partial vacuum) to a volume of ca. 5 L. Four liters of
2-propanol was added, and distillation resumed to a volume
of ca. 3 L. Another 4-L portion of 2-propanol was added,
and distillation continued to a volume of ca. 4 L. The
resulting slurry was stirred overnight at room temperature,
then filtered and dried under vacuum to provide 483 g (1.34
mol, 65% yield) of tan solids. This material was further
purified by slurrying in 4.8 L of a 98:2 (v/v) H2O-2-
propanol mixture: the slurry was warmed to 55 °C for 60
min, then cooled to room temperature and stirred for 2 h.
Solids were collected and then recrystallized by dissolving
in 9.4 L of a 50:50 CH2Cl2-MeOH mixture with warming
to 35 °C, then concentrating under partial vacuum to ca. 4
L. 2-Propanol (8 L) was added, and distillation was resumed
to a volume of ca. 4 L. The resulting slurry was stirred at
room temperature for 2 h, and the solids were collected by
filtration. Vacuum drying provided 390 g (1.08 mol, 52%
yield for the overall reaction and purification) of off-white
solids, with an HPLC purity of 98.7%.
and 400 mL of absolute EtOH. The resulting mixture was
stirred vigorously and heated to reflux for 48 h. During the
course of the reaction, a solid precipitated from the reaction
mixture. After cooling the reaction mixture to 25 °C, the
solids were collected by filtration, washed with EtOH, and
dried in a vacuum oven. This provided the HCl salt of 1 as
a crystalline solid with 99% HPLC purity (34.5 g, mol, 87%
yield).
(1R)-(-)-10-Camphorsulfonic Acid Salt of 1. [2-(3-
Methyl-3H-imidazol-4-yl)-thieno[3,2-b]pyridin-7-yl]-(2-methyl-
1H-indol-5-yl)-amine (1) (386 g, 1.07 mol) was dissolved
in 3.9 L of CH2Cl2 and 3.9 L of MeOH with warming to 35
°C. The solution was filtered through paper, and speck-free
conditions (free of particulates) were maintained for the
duration of the operations. (1R)-(-)-10-Camphorsulfonic acid
(246 g, 1.06 mol) was dissolved in 2 L of THF, filtered into
a spec-free flask, and added to the solution of free base over
10-15 min. The resulting solution was concentrated under
partial vacuum to a volume of ca. 3 L. An additional 8 L of
THF was added, and a seed crystal was introduced. Distil-
lation was resumed to a volume of ca. 3 L. Additional THF
was added to a final volume of ca. 10 L. The resulting slurry
was stirred overnight at room temperature. The solids were
collected by filtration, rinsing with THF. Vacuum drying
provided off-white solids (589 g, 0.996 mol, 93% yield);
mp ) 269-270 °C dec. ICP analysis indicated 3 ppm for
both Sn and Pd. 1H NMR (CD3OD): δ 8.20 (d, J ) 7, 1H),
7.89 (s, 1H), 7.61 (s, 1H), 7.47 (s, 1H), 7.43 (d, J ) 8, 1H),
7.37 (s, 1H), 7.05 (dd, J ) 8, 2), 6.85 (d, J ) 7), 6.24 (s),
3.88 (s, 3H), 3.35 (d, J ) 16, 1H), 2.80 (d, J ) 16, 1H),
2.72-2.63 (m, 1H), 2.49 (s, 3H), 2.35 (dt, J ) 18, 4, 1H),
2.07-2.00 (m, 2H), 1.90 (d, J ) 18, 1H), 1.68-1.60 (m,
1H), 1.45-1.37 (m, 1H), 1.13 (s, 3H), 0.86 (s, 3H); Anal.
Calcd for C20H17N5S.C10H16SO4: C, 60.89; H, 5.62; N, 11.84;
S, 10.84. Found: C, 60.72; H, 5.62; N, 11.66; S, 10.90.
Acknowledgment
Dr. Ricardo Borjas and Mr. Eric Weiss are acknowledged
for coordinating the tin and palladium analyses. Dr. Greg
Cohee of QTI is acknowledged for timely assistance in
executing the tin and palladium analyses. Numerous helpful
discussions with Drs. Joel Hawkins, Bob Dugger, Keith
DeVries, Jennifer Rutherford, and Professor Dan Kemp
(MIT) are gratefully acknowledged.
Method D (Lab Scale in EtOH, Isolated as HCl Salt). A
2-L round-bottom flask was charged with chloropyridine 2
(25 g, 0.10 mol), 5-amino-2-methylindole (14.6 g, 0.10 mol),
Received for review April 2, 2003.
OP0340457
Vol. 7, No. 5, 2003 / Organic Process Research & Development
•
683