Pilot scale synthesis of a novel nonpeptide angiotensin II receptor antagonist

Article abstract of DOI:10.1021/op9800055

FR143187 is a novel nonpeptide angiotensin II receptor antagonist under development at Fujisawa Pharmaceutical Co. for the treatment of hypertension. Development of a process for preparation on a large scale is described. The optimized process is 10 steps in length and uses only commercially available materials for each step. Efficient methylation of the 2-position of a pyrrole derivative was achieved by reduction of a Mannich base via the quaternary ammonium salt. Selective cyanation directed by a solvent effect was also investigated. Process improvement efforts focused on optimized reaction conditions for each step, leading to a high-quality product according to a new and concise synthetic route.

Full text of DOI:10.1021/op9800055

Organic Process Research & Development 1998, 2, 230−237
Pilot Scale Synthesis of a Novel Nonpeptide Angiotensin II Receptor Antagonist
Atsuhiko Zanka,*,† Masanori Nishiwaki,† Yasuhiro Morinaga,† and Takayuki Inoue‡
Technological DeVelopment Laboratories, Fujisawa Pharmaceutical Co., Ltd.,
2-1-6 Kashima, Yodogawa-ku, Osaka 532, Japan, and Exploratory Research Laboratories,
Fujisawa Pharmaceutical Co., Ltd., 5-2-3 Tokoudai, Tsukuba, Ibaraki 300-26, Japan
Abstract:
FR143187 is a novel nonpeptide angiotensin II receptor an-
tagonist under development at Fujisawa Pharmaceutical Co.
for the treatment of hypertension. Development of a process
for preparation on a large scale is described. The optimized
process is 10 steps in length and uses only commercially
available materials for each step. Efficient methylation of the
2-position of a pyrrole derivative was achieved by reduction of
a Mannich base via the quaternary ammonium salt. Selective
cyanation directed by a solvent effect was also investigated.
Process improvement efforts focused on optimized reaction
conditions for each step, leading to a high-quality product
according to a new and concise synthetic route.
Introduction
Angiotensin II (Ang II) receptor antagonists, such as
Losartan1 (Dupont-Merck) or Candesartan2 (Takeda) which
block the receptor site of angiotensin II, have been developed
as a new generation of more effective and selective drugs
for the treatment of hypertension than the well-known
angiotensin-converting enzyme (ACE) inhibitors such as
Figure 1. Nonpeptide angiotensin II receptor antagonists.
Captopril (BMS)3 or Enalapril (Merck)4 (Figure 1). This
discovery of non-peptide Ang II receptor antagonists having
synthesis of FR143187 (1) and remove several major
exceptionally high antihypertension effects prompted our own
obstacles in the original process that prevented large-scale
search for agents with high activity. Amongst many types
synthesis. The new route uses readily available commercial
of Ang II receptor antagonists synthesized in Fujisawa,
materials, requires relatively simple equipment, and produces
several pyrrole derivatives showed highly potent activity.5
a high-quality product. Herein, we report the studies on the
Amongst these compounds, FR143187 (1) was shown to be
process development of FR143187 (1) amenable to a large-
especially effective. For the purposes of complete pharma-
scale synthesis.
cological and toxicological evaluation of FR143187 (1), we
needed an efficient synthetic route to large quantities. The
results described herein outline a new process for the
Results and Discussion
Synthetic Route Investigation. 1-(4-Ethoxycarbonylphe-
nyl)-5-methylpyrrole-2-carbonitrile (6a) is a key intermediate
for the preparation of 1. The original synthetic approach5
to 6a involved regioselective cyanation of a protected
hydroxymethylpyrrole derivative followed by reduction of
the hydroxymethyl moiety to a methyl group (Scheme 1).
However, this route had severe disadvantages for large-scale
synthesis. A major problem was that hydroxymethylpyrrole
derivatives as intermediates were too unstable to be handled
on a large scale. These compounds were initially purified
by column chromatography on silica gel, but on a scale above
100 g, most of the product decomposed during this proce-
dure. Another problem was that this 6-step route is
conceptually inefficient. For example, protection and depro-
tection of the hydroxymethyl group requires two extra steps,
† Technological Development Laboratories.
‡ Exploratory Research Laboratories.
(1) (a) Duncia, J. V.; Carini, D. J.; Chiu, A. T.; Johnson, A. L.; Price, W. A.;
Wong, P. C.; Wexler, R. R.; Timmermans, P. B. M. W. M. Med. Res. ReV.
1992, 12, 149. (b) Carini, D. J.; Duncia, J. V.; Aldrich, P. E.; Chiu, A. T.;
Johnson, A. L.; Pierce, M. E.; Price, W. A.; M. E.; Santella, , J. B., III;
Wells, G. J.; Wexler, R. R.; Wong, P. C.; Yoo, S. E.; Timmermans, P. B.
M. W. M. J. Med. Chem. 1991, 34, 2525.
(2) Kubo, K.; Kohara, Y.; Imamiya, E.; Sugiura, Y.; Inada, Y.; Furukawa, Y.;
Nishikawa, K.; Naka, T. J. Med. Chem. 1993, 36, 2182.
(3) Ondetti, M. A.; Rubin, B.; Cushman, D. W. Science 1977, 196, 441.
(4) Patchett, A. A.; Harris, E.; Tristram, E. W.; Wyvratt, M. J.; Wu, M. T.;
Taub, D.; Perterson, E. R.; Ikeler, T. J.; Broeke, J.; Payne, L. G.; Ondeyka,
D. L.; Thorsett, E. D.; Greenlee, W. J.; Lohr, N. S.; Hoffsommer, R. D.;
Joshua, H.; Ruyle, W. V.; Rothrock, J. W.; Aster, S. D.; Maycock, A. L.;
Robinson, F. M.; Hirschmann, R.; Sweet, C. S.; Ulm, E. H.; Gross, D. M.;
Vassil, T. C.; Stone, C. A. Nature 1980, 288, 280.
(5) Oku, T.; Setoi, H.; Kayakiri, H.; Satoh, S.; Inoue, T.; Saitoh, Y.; Kuroda,
A.; Tanaka, H. Eur. Pat. 0 480 204 B1.
230
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development
S1083-6160(98)00005-X CCC: $15.00 © 1998 American Chemical Society and Royal Society of Chemistry
Published on Web 04/30/1998

Scheme 1. Original route to key intermediate
Table 1. Reduction of 7a
2-cyanopyrrole (6a) (route A)a
yield
molar
ratio
temp.
(°C)
time
(h)
reagent
(%)
NaBH3CN
Me2N‚BH3
NMM‚BH3
pyridine‚BH3
pyridine‚BH3
NaBH(CH3CO2)3
73
0c
70
78
63
0d
8
8
8
8
2.2
8
70
100
100
100
100
70
18
4
4
2
2
b
2
a All reactions were carried out in N,N′-dimethylimidazolidinone. b NMM )
N-methylmorpholine. c Major product was the Mannich base (removal of MeI).
d Starting material recovery.
quaternary trimethylammonium salt. According to these
procedures, we could prepare a product of satisfactory quality
(93% chemical yield from 2; 95% chemical purity)
Step 3: Reduction with Borane-Pyridine Complex.
Itoh and co-workers7 reported methylation of aromatic
compounds by reduction of quaternary salts of Mannich bases
with sodium cyanoborohydride. It is well-known that sodium
cyanoborohydride is capable of reducing a number of
functional groups chemoselectively in the presence of a range
of sensitive functional groups.8 Despite this usefulness, it
is not acceptable for industrial synthesis because of a lack
of commercial availability on a bulk scale and the difficulty
in effecting complete decomposition of excess reagent under
mild acidic conditions. The evolution of hydrogen cyanide
gas during the reaction also necessitates elaborate safety
precautions and therefore causes many procedural problems
when employing the use of this reducing agent on a large
scale. As an alternative reagent available in bulk quantities,
borane-dimethylamine complex9 was investigated for reduc-
tion of 2, but the results were unsatisfactory. The obtained
products were mainly the Mannich base with several related
by-products. In the course of further studies, it was found
that dramatically superior results could be obtained by using
borane-pyridine complex. As shown in Table 1, use of
excess borane-pyridine complex improved the yield. How-
ever, the use of excess reagent also made the workup
procedure cumbersome. After completion of the reaction,
the organic layer was washed with water and most of the
unreacted borane-pyridine complex was thereby removed.
A small amount of remaining borane-pyridine complex in
the organic layer was completely decomposed by washing
with dilute formalin. Removal of solvent under reduced
pressure afforded an oil which contained 8 in 64% yield.
The reaction allowed complete starting material consumption
in 2 h, leading to 8 and a major by-product (33 area % by
HPLC). However, the quality of 8 was not influenced since
this unknown by-product was cleanly removed in the aqueous
layer. For the pilot plant scale synthesis, we obtained the
pre-prepared borane-pyridine complex from the Morton
International Co.; however, with the goal of gaining a more
rapid access to 8 in a manner amenable to production scale
use in the future, we endeavoured to prepare the borane-
a Reagents and conditions: 1. POCl3/DMF; 2. NaBH4/EtOH; 3. tBuPh2SiCl/
Et3N, DMAP; 4. ClSO2NCO/DMF; 5. nBu4NF; 6. Et3SiH, TFA.
compared with a direct cyanation approach. Furthermore,
protecting reagents such as tert-butylchlorodiphenylsilane are
expensive and not readily available on a bulk scale. Amongst
several postulated synthetic routes, we were particularly
interested in the route outlined in Scheme 2, since in this
route no special or expensive reagents ought to be required
and fewer potentially unstable intermediates are involved.
Moreover, according to route B, only 4 steps are required in
principle to convert 2 into 6a, compared with 6 steps in the
previous route (route A). Therefore, our interest in develop-
ing a rapid and inexpensive preparative method for 1
promoted us to explore and develop this new synthetic route
(route B).
Step 1: Pyrrole Ring Formation. In early investiga-
tions, the Wasley procedure6 (AcOH solvent) was used for
preparation of 1-(4-ethoxycarbonylphenyl)pyrrole (2). The
desired product was obtained in 47% yield; however, it
contained tar by-products which made the purification
procedure extremely tedious. We endeavoured to improve
the reaction conditions and found that the presence of water
in the reaction leads to decomposition of the desired product.
Use of a 1:1 mixture of toluene and glacial acetic acid as
solvent, and continuously removing water by azeotropic
distillation, significantly improved the quality and yield of
product. Crystallization of 2 from 2-propanol then gave the
product in good yield and satisfactory purity (78% chemical
yield; 99% chemical purity).
Step 2: Mannich Reaction and Quaternary Am-
monium Salt Formation. We applied standard Mannich
reaction conditions (37% aqueous CH2O, Me2N‚HCl-EtOH)
to prepare the dimethylamino precursor of 7, but the yield
(47%) was unacceptably low. Knowing that the product and
starting material were possibly unstable in an aqueous
system, as was observed in the previous step, we investigated
reaction in nonaqueous conditions (paraformaldehyde-
Me2N‚HCl-EtOH) and found that reaction proceeded quan-
titatively. Next, the product was converted into the stable
(7) Yamada, K.; Itoh, N.; Iwakuma, T. J. Chem. Soc., Chem. Commun. 1978,
1089.
(6) Boyer, S.; Blazier, E.; Barabi, M.; Long, G.; Zaunius, G.; Wasley, J. W.
(8) Lane, C. F. Synthesis 1975, 135.
F.; Hamdan, A. J. Heterocycl. Chem. 1988, 25, 1003.
(9) Salunkhe, A. M.; Burkhardt, E. R. Tetrahedron Lett. 1997, 38, 1519.
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
231

Scheme 2. New synthetic route to FR143187 (1) (route B)
Table 2.
pyridine complex in situ. Whilst amine borane reagents have
been proven to act as alternatives to sodium cyanoborohy-
dride10 and have remarkable stability11 and handling con-
venience, they have failed to gain widespread use in synthetic
chemistry. Furthermore, inexpensive and practical processes
for borane-pyridine complex on a large scale are not well-
known, in common with other amine borane reagents. In
an attempt to develop a viable process on a large scale, we
first improved the preparative method for borane-pyridine
complex reported by Taylor12 and explored the simple and
efficient reduction system outlined schematically in Scheme
3. The preparative method was as follows. Reaction of
pyridinium chloride, which was easily prepared from pyridine
and hydrogen chloride, with sodium borohydride afforded
borane-pyridine complex in situ and without isolation; the
quaternary ammonium species 7 could be reduced to the
corresponding methyl compound 8 by addition to the borane
solution. Removal of solvent under reduced pressure af-
forded an oil which contained 8 in 68% yield. This was the
same result as when pre-prepared borane-pyridine complex
was used.
product ratioa
solvent
8
6a
6b
CH2Cl2
PhMe
n-heptane
12
6
10
78
87
86
10
7
4
a Ratio of products was determined by HPLC.
chlorosulfonylisocyanate in methylene chloride followed by
addition of N,N-dimethylformamide (DMF) (Scheme 4).
However, when standard methods were used,13 considerable
amounts of 6b were also formed (Table 2). The similar
nature of 6a and 6b prevented efficient separation by usual
methods such as recrystallization, extraction, and column
chromatography. In addition, 6a never crystallized in the
presence of 6b (up to 10 mol %). In continuous studies, it
became clear that there was a correlation between solvent
polarity and the amount of 6b produced. Amongst several
solvents screened, n-heptane was found to be most suitable
(Table 2), but in order to avoid gummy by-products, the
addition of toluene after reaction completion served to
facilitate the isolation and purification procedure. Crude 6a,
which was triturated from isopropyl ether, along with a small
amount of 6b (∼2%) was purified by recrystallization from
aqueous isopropyl alcohol to afford 6a with satisfactory
quality (36% chemical yield; 98% chemical purity).
Step 4: Regioselective Cyanation. We next examined
the regioselective cyanation of 8. Compound 6a was
anticipated to be readily prepared via treatment of 8 with
(10) Pelter, A.; Rosser, R. M.; Mills, S. J. Chem. Soc., Perkin Trans. 1 1984,
717.
(11) Ryschkewitsch, G. J. Am. Chem. Soc. 1960, 82, 3290.
(12) Taylor, M. D.; Grant, L. R.; Sands, C. A. J. Am. Chem. Soc. 1955, 77,
1506.
(13) Lohaus, G. Org. Synth. 1970, 50, 52.
232
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development

Scheme 3. Reduction of 7 with borane-pyridine complex
Scheme 5. Inhibition of 10b
mother liquor during the filtration procedure leading to 12.
Our investigations on the inhibition of production of 10b
showed that the resulting amount of 10b depended on the
quantity and speed of addition of MsCl and the internal
reaction temperature (Scheme 5). Considering the heat
balance during the reaction on a pilot plant scale preparation,
it was estimated that 30 min would be best for introduction
of MsCl to the reaction mixture, whilst maintaining the
temperature at 5-10 °C. In this way, optimized reaction
conditions (MsCl (1 equiv)), temperature of 5-10 °C,
addition time of ca. 30 min) gave the desired product along
with 1.5% of 10b and 2.5% of the starting material.
Extraction from the reaction mixture with methylene chloride,
followed by condensation under reduced pressure, afforded
crude 10a as oil. The mesylate 10a was suitable for direct
coupling with readily prepared pyridine derivative 11 (see
the Experimental Section) according to the following pro-
cedure.
(b) Coupling 10a with 11. In our first experiment on a
lab scale, we coupled 10a with 11 to prepare 12 in the
presence of sodium hydride in DMF. However, in this
system the use of excess sodium hydride resulted in the
decomposition of the desired product, whilst in the case of
exact molar equivalents, large amounts of starting material
remained. In both experiments, the yields were poor
(∼50%). Furthermore, the known instability of NaH-DMF
mixtures precluded any development of a large-scale pro-
cess.14 We thus investigated methods to improve this
reaction system and found that by utilizing a mixture of
powdered potassium carbonate and granulated sodium hy-
droxide as the base under low-temperature conditions (3-5
°C), the desired adduct could be prepared in excellent yield
as an oil (90% from 6a).
Step 7: Reductive Cyclization. The reductive cycliza-
tion, leading to imidazole ring formation was carried out in
ethanol in the presence of acetic acid and fine iron powder.
The most difficult problem in this step was the observation
that, on both laboratory and pilot scale, the reaction would
sometimes suddenly stop, leaving variable amounts of
unchanged starting material. The reason for this problem
was not clear; however, extensive studies showed that
periodic addition of further iron powder or acetic acid could
reinitiate the reaction. In pilot manufacturing, we added
acetic acid to finish the reaction at a point 2 h from initiation.
Scheme 4. Solvent effect on cyanation of 8
Step 5: Reduction of Ethyl Ester to Alcohol. This
reaction proceeded smoothly by adding methanol dropwise
to a mixture of 6a and NaBH4 in tetrahydrofuran maintaining
high-temperature conditions (∼60 °C). However, since the
control of hydrogen evolution was difficult on a large scale,
we searched for appropriate conditions and found that
reaction proceeded even at 45-55 °C. The resulting alcohol
9 was then used directly in the next step.
Step 6: Coupling 9 with 11. (a) ActiVation of 9 with
Methanesulfonyl Chloride. Activation of 9 was easily
achieved with methanesulfonyl chloride (MsCl). However,
in our early studies, the product contained 10% of the
chloromethyl impurity 10b. This by-product did not undergo
coupling with 11 under the low-temperature conditions
investigated. By-product 10b was readily removed in the
(14) Laird, T. Chem. Ind. 1986, 17, 134.
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
233

After the removal of residual iron powder and complete
removal of solvents, the resultant oil was dissolved in ethyl
acetate and purified by column chromatography on silica in
order to improve the color of product and ensure the complete
removal of iron powder. The eluent was concentrated to an
oil and crystallized from 2-propanol to afford 13 in high yield
and quality (83% chemical yield from 6a; 99% chemical
purity).
Conclusions
We have succeeded in developing a new large-scale
synthesis of FR143187, a non-peptide angiotensin II receptor
antagonist. Key features of this work are as follows:
1. We focused on maintaining safety for the pilot scale
tetrazole reaction, one of the most dangerous and difficult
reactions for large scale, by ensuring that no heavy metals
were present and also the complete decomposition of residual
azide compounds. The final process was safely implemented
on a pilot plant scale. The described methods are useful for
reactions in which use of azide reagents is unavoidable and
especially for a large-scale synthesis.
2. A practical, efficient, and inexpensive preparative
method for the borane-pyridine complex suitable for a large-
scale synthesis was investigated. Borane-pyridine complex
is a safe reagent, is easy to handle, and can function as a
substitute for sodium cyanoborohydride.
Step 8: Tetrazole Reaction. One of the most useful
functional groups to be introduced into Ang II receptor
antagonists is the tetrazole group, functioning as a carboxylic
acid isostere. Whilst a number of methods for introducing
a tetrazole group have been reported,15 the coupling reaction
of a cyano group with sodium azide in the presence of an
ammonium salt has been the most widely used.16 This
reaction usually proceeds smoothly, but in the case of a large-
scale synthesis, there is a major safety consideration which
must be kept in mind. Hydrogen azide (HN3) is extremely
sensitive towards explosion, especially in the presence of
heavy metals, as well as being harmful.17 In pilot plant
manufacturing, we rigorously ensured that no heavy metals
were present in the starting material using an X-ray mi-
croanalyzer and also in the reaction vessel. Concerning
residual sodium azide, we planned to decompose it using
sodium nitrite under acidic conditions (pH ) <1). However,
almost all of the product decomposed under these conditions.
Therefore, we had to investigate new procedures for the
decomposition of azide. To prevent the decomposition of
the product, we envisaged that rigorous control of pH would
allow selective removal of azide. Ethyl acetate was first
added to the reaction mixture, followed by an excess amount
of sodium nitrite, and the aqueous layer was then carefully
maintained at pH ) 4.5-5.0. In our investigation, only when
the reaction was carried out at relatively high-temperature
conditions (25-30 °C) did decomposition of residual azides
completely finish. In pilot plant manufacturing, complete
decomposition of azides was analyzed by ion chromatogra-
phy (N3- was checked) as well as the conventional method18
using dilute iron(III) chloride in water. After this treatment,
the organic layer was washed with 10% brine, and replacing
the organic layer by acetonitrile gave the desired high quality
product 14 (85% chemical yield; 99% chemical purity).
Step 9: Hydrogen Chloride Salt. In early studies, we
attempted conversion of 14 into the hydrogen chloride salt
form 1 using 35% hydrogen chloride in water, ethanol, or
2-propanol. However, variable amounts (1.5-1.7%) of
residual solvent could not be removed under reduced
pressure. We investigated crystallization from a mixture of
2-propanol and water and found that, at a high ratio (up to
64%) of water, satisfactory quality product (residual 2-pro-
panol ) 0.5%, 99% purity) could be obtained in 85% yield.
The product prepared by this method has proven to be
acceptable for pharmaceutical and toxicological evaluation.
3. Selective cyanation of a methylpyrrole derivative was
achieved by modification of reaction conditions.
4. It is worth emphasizing that process improvement
efforts focused on safe and optimized reaction conditions
for each step with the aim of a large-scale synthesis giving
satisfactory quality products and of a final product suitable
for pharmaceutical and toxicological evaluation.
Experimental Section
General Methods. Ethyl 4-aminobenzoate was com-
mercially available from the Midori Chemical Co. 2,5-
Dimethoxytetrahydrofuran was commercially available from
BASF Co. N,N′-dimethylimidazolidinone (DMI) was ob-
tained from Mitsui Toatsu Chemical Co., and chlorosulfo-
nylisocyanate was from Kuraray Co. All other chemicals
were obtained from the usual commercial suppliers. HPLC
analyses were performed using a YMC GEL ODS 120 Å
S-7 column and an acetonitrile/water phase. The water
component was adjusted with KH2PO4 and Na2HPO4 to pH
) 7.8. Purity of each obtained product was determined by
comparison with purified authentic samples using quantitative
HPLC. X-ray microanalyses were performed on a HORIBA
EMAX-5770. Melting points were measured on a Thomas-
Hoover apparatus and are uncorrected. 1H NMR spectra
were obtained at 200 MHz in CDCl3 or DMSO-d6. Chemical
shifts are given in parts per million, and tetramethylsilane
was used as the internal standard. Ion chromatography
measurements were performed on a Yokogawa Model IC-
7000. IR spectra were recorded on a HITACHI IR-260-10
spectrometer. Mass spectra were measured on a Hitachi
Model M-80 mass spectrometer using EI for ionization.
Elemental analyses were carried out on a Perkin-Elmer 2400
CHN elemental analyzer.
1-(4-Ethoxycarbonylphenyl)pyrrole (2). A solution of
ethyl 4-aminobenzoate (220 kg, 1332 mol) and 2,5-
dimethoxytetrahydrofuran (176 kg, 1332 mol) in a mixture
of glacial acetic acid (660 L) and toluene (660 L) was heated
to reflux for 3 h with removal of water of reaction by
azeotropic distillation. After the reaction was complete, the
reaction mixture was washed consecutively with water (440
L, 220 L, 220 L), saturated sodium hydrogen carbonate (220
L), and then evaporated under reduced pressure. The residue
(15) Butler, R. N. AdV. Heteroat. Chem. 1977, 21, 323.
(16) (a) Finnegan, W. G.; Henry, R. A.; Lofquist, R. J. Am. Chem. Soc. 1958,
80, 3908. (b) Bernstein, P. R.; Vacek, E. P. Synthesis 1987, 1133.
(17) Lenga, R. E. Sigma-Aldrich Library of Chemical Safety Data, 2nd ed.;
Sigma-Aldrich Chemical Co., Inc.: Milwaukee, WI; Vol. 2, p 3129D.
(18) Azzam, A. A. Mikrochim. Acta 1937, 283.
234
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development

was dissolved in 2-propanol (220 L) and evaporated again
to ensure the complete removal of acetic acid. The residual
product was dissolved in 2-propanol (880 L), heated to reflux
for 1 h and cooled to 0 °C. After additional overnight stirring
at ambient temperature, the precipitate was filtered off,
washed with cooled 2-propanol (220 L) and dried under
reduced pressure to yield 2 (224 kg, 78% yield) of 99%
mL) under nitrogen atmosphere to afford borane-pyridine
complex. To this reagent was added 7 (100 g, 241 mmol),
the mixture heated to 105 °C for 2 h and then cooled to 30
°C. To this reaction mixture was added consecutively
n-heptane (500 mL) and water (300 mL). After the layers
were separated, the aqueous layer was extracted with fresh
n-heptane (300 mL) and the combined organic layer was
washed with 12% formalin (300 mL) until no borane-
pyridine complex was detected (HPLC). The organic layer
was washed with water (300 mL) and concentrated to an oil
under reduced pressure which contained 8 (37.6 g, 68% yield,
91 area % by HPLC).
1
purity; mp 74-75 °C; H NMR (200 MHz, CDCl3) δ 1.40
(t, 3H, J ) 7.1 Hz), 4.39 (q, 2H, J ) 7.1 Hz), 6.38 (t, 2H,
J ) 2.2 Hz), 7.15 (t, 2H, J ) 2.2 Hz), 7.44 (d, 2H, J ) 8.8
Hz), 8.10 (d, 2H, J ) 8.8 Hz); IR (Nujol) 1705, 1608, 1523,
1476, 1428, 1334, 1282 cm-1; MS (EI) m/z 216 (M+H)+,
75. Anal. Calcd for C13H13NO2: C, 72.54; H, 6.09; N, 6.51.
Found: C, 72.55; H, 6.10; N, 6.46.
1-(4-Ethoxycarbonylphenyl)-5-methypyrrole-2-carbo-
nitrile (6a). A mixture of 8 (46 kg, 201 mol) and n-heptane
(920 L), cooled to -10 °C, was treated dropwise with
chlorosulfonylisocyanate (39.8 kg, 281 mol) maintaining the
temperature at -10 °C, and the reaction was continued at
ambient temperature for an additional 1 h. To this solution
were added cooled toluene (0-10 °C, 460 L) and cooled
DMF (0-10 °C, 138 L), and the mixture was stirred for 1
h at -5 °C. To the reaction mixture was added 12%
hydrogen chloride in water (230 L) maintaining the temper-
ature under 0 °C. The layers were separated, and the aqueous
layer was extracted with a mixture of n-heptane (230 L) and
toluene (115 L). The combined organic layers were washed
consecutively with water (230 L) and saturated sodium
hydrogen carbonate (230 L). The organic layer was con-
centrated under reduced pressure, and to the residue were
added isopropyl ether (595 L) and n-heptane (595 L). After
the organic layer was separated by decantation, the residual
waxy products were dissolved in isopropyl ether (595 L) and
extracted with n-heptane (595 L). The combined organic
layers were concentrated to an oil under reduced pressure,
and the residual oil was dissolved in hot isopropyl ether (198
L) and cooled to 0 °C. Purified 6a (132 g) was added, and
after complete precipitation at ambient temperature, the
precipitate was filtered off and dried under reduced pressure
to afford crude 6a (33.6 kg, 66% yield) of 91% purity. The
crude 6a (33.6 kg) was recrystallized from a mixture of
2-propanol (67 L) and water (17 L) to afford purified 6a
(18.4 kg, 55% yield) of 98% purity. 1-(4-Ethoxycarbon-
ylphenyl)-5-methypyrrole-2-carbonitrile (6a): mp 56-57 °C;
1H NMR (200 MHz, CDCl3) δ 1.42 (t, 3H, J ) 7.0 Hz),
2.18 (s, 3H), 4.42 (q, 2H, J ) 7.0 Hz), 6.11 (d, 1H, J ) 4.5
Hz), 6.90 (d, 1H, J ) 4.5 Hz), 7.40 (d, 2H, J ) 9.0 Hz),
8.21 (d, 2H, J ) 9.0 Hz); IR (Nujol) 2212, 1715, 1604, 1513,
1479, 1407, 1371, 1328 cm-1; MS (EI) m/z 255 (M + H)+,
209, 141, 75. Anal. Calcd for C15H14N2O2: C, 70.85; H,
5.55; N, 11.02. Found: C, 70.68; H, 5.52; N, 10.96. 1-(4-
Ethoxycarbonylphenyl)-5-methypyrrole-2,4-dicarbonitrile (6b)
was obtained by preparative TLC: 1H NMR (200 MHz,
CDCl3) δ 1.43 (t, 3H, J ) 7.1 Hz), 2.33 (s, 3H), 4.44 (q,
2H, J ) 7.1 Hz), 7.15 (s, 1H), 7.41 (d, 2H, J ) 8.6 Hz),
8.27 (d, 2H, J ) 8.6 Hz); MS (EI) m/z 280 (M + H)+, 259,
167, 101, 75.
1-(4-Ethoxycarbonylphenyl)pyrrole-2-methyltrimethy-
lammonium Iodide (7). A solution of 2 (220 kg, 1022 mol),
dimethylammonium chloride (125 kg, 1533 mol) and
paraformaldehyde (138 kg, 4595 mol) in ethanol (660 L)
was heated to reflux (83 °C) for 3 h and cooled to 30 °C.
To this reaction mixture were added sequentially ethyl acetate
(660 L), water (220 L), and 12% sodium hydroxide in water
(440 L) whilst the temperature was maintained below 15 °C.
The aqueous layer was separated, and the organic layer was
washed with water (440 L, 220 L). Aqueous layers were
combined and extracted with ethyl acetate (220 L). The
combined ethyl acetate layers were washed with saturated
brine (220 L), and to this layer was added methyl iodide
(290 kg, 2043 mol) at 35 °C. After 2 h of stirring at ambient
temperature, the reaction mixture was cooled to 0 °C and
stirred overnight. The precipitate was filtered off, washed
with a mixture of ethyl acetate (352 L) and ethanol (88 L),
and dried under reduced pressure to yield 7 (394 kg, 93%
yield) of 95% purity: mp 287-289 °C (dec); 1H NMR (200
MHz, DMSO-d6) δ 1.35 (t, 3H, J ) 7.1 Hz), 2.78 (s, 9H),
4.37 (q, 2H, J ) 7.1 Hz), 4.37 (s, 2H), 6.42 (t, 1H, J ) 3.4
Hz), 6.74 (dd, 1H, J ) 3.6, 1.6 Hz), 7.26 (dd, 1H, J ) 2.8,
1.7 Hz), 7.62 (d, 2H, J ) 8.5 Hz), 8.12 (d, 2H, J ) 8.5 Hz);
IR (Nujol) 1698, 1605, 1470, 1379, 1333, 1280, 1174, 1130
cm-1; MS (EI) m/z 242, 228, 214, 75. Anal. Calcd for
C17H23N2O2I‚0.86 H2O: C, 47.51; H, 5.80; N, 6.52.
Found: C, 47.65; H, 5.57; N, 6.64.
1-(4-Ethoxycarbonylphenyl)-5-methylpyrrole (8). Pilot
Plant Scale Synthesis. A mixture of 7 (130 kg, 314 mol)
and borane-pyridine complex (64.2 kg, 691 mol) in DMI
(400 L) was heated to 104-113 °C for 2 h and then cooled
to 30 °C. To this reaction mixture was added consecutively
n-heptane (650 L) and water (390 L). The layers were
separated, the aqueous layer was extracted with fresh
n-heptane (390 L), and the combined organic layers were
washed with 12% formalin (390 L) until no borane-pyridine
complex was detected (HPLC). The organic layer was
washed with water (390 L) and concentrated under reduced
pressure to give 8 as an oil (46.1 kg, 64% yield, 91 area %
by HPLC).
ImproVed Synthesis on Lab Scale. Pyridinum chloride
was easily prepared from pyridine (45.75 g, 578 mmol) and
hydrogen chloride (20.20 g, 554 mmol) and was then treated
with sodium borohydride (20.05 g, 530 mmol) in DMI (300
2-(Butyrylamino)-4-methyl-3-nitropyridine (11).
A
mixture of 2-amino-4-methyl-3-nitropyridine (45.7 kg, 298
mol) and N,N-dimethylaniline (72.3 kg, 597 mol) in toluene
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
235

(229 L) was heated at 95 °C under nitrogen atmosphere. To
this solution was slowly added butyryl chloride (35.0 kg,
328 mol), and the mixture was stirred at 100 °C for 7 h.
After cooling to 25 °C, methylene chloride (451 L) and water
(229 L) were added. The organic layer (lower layer) was
separated and washed twice with water (229 L, 229 L). The
organic layer was concentrated under reduced pressure to
∼229 L and n-heptane (457 L) added. The solution was
concentrated again to ∼457 L, and n-heptane (215 L) added,
followed by concentrated to ∼457 L. To this residue was
added n-heptane (215 L), and the solution was heated at 45-
50 °C for 30 min and cooled under 10 °C. The precipitate
was filtered off, washed with n-heptane (137 L), and dried
to afford 11 (56.0 kg, 84% yield) of 100% purity: mp 109-
110 °C; 1H NMR (200 MHz, CDCl3) δ 1.01 (t, 3H, J ) 7.5
Hz), 1.64-1.85 (m, 2H), 2.43 (t, 2H, J ) 7.5 Hz), 2.48 (s,
3H), 7.10 (d, 1H, J ) 5.0 Hz), 8.26 (br s, 1H), 8.35 (d, 1H,
J ) 5.0 Hz); IR (KBr) 3296, 3020, 2962, 2929, 2875, 1674,
1603, 1556, 1533, 1512, cm-1; MS (EI) m/z 224 (M + H)+,
154, 73. Anal. Calcd for C10H13N3O3: C, 53.81; H, 5.87;
N, 18.82. Found: C, 53.94; H, 5.89; N, 18.61.
and the mixture was adjusted to 3-5 °C. To this mixture
was added the previously prepared solution of 10a in toluene
over 15 min. The reaction was then continued overnight
(12 h) at ambient temperature and then heated to 25-30 °C
for 2.5 h. After completion of the reaction, water (314 L)
was slowly added keeping the temperature below 30 °C, and
then the mixture was heated to 50 °C. The layers were
separated, and the aqueous layer was re-extracted with
toluene (125 L). The combined organic layer was washed
with water (125 L) and concentrated under reduced pressure
to afford an oil which contained 12 (72.2 kg, 88% yield from
6a). To a portion of the residual oil (62.0 kg, 149 mol) were
added ethanol (310 L), acetic acid (62 L), and iron powder
(62.0 kg), and the mixture was then heated to reflux for 2 h.
Additional acetic acid (62 L) was then added, and the reaction
was continued for a further 3 h, followed by cooling to room
temperature. The iron powder was filtered off, the filter cake
was washed with ethanol (124 L), and the filtrate was
concentrated to an oil under reduced pressure. To ensure
the complete removal of acetic acid, the residual oil was
dissolved in toluene (62 L), and the mixture was re-
concentrated. After removal of small amounts of insoluble
products contained in 13 by dissolving in ethyl acetate (186
L) and filtration, the filtrate was purified by column
chromatography on silica gel (62.0 kg) eluting with additional
ethyl acetate (248 L). The effluent was concentrated to an
oil under reduced pressure and dissolved in 2-propanol (248
L) at 65 °C, followed by cooling to 0 °C. After overnight
stirring at ambient temperature, the precipitate was filtered
off, washed with cooled 2-propanol (62 L), and dried in
vacuo overnight to afford 13 (45.5 kg, 83% yield) of 99%
purity. The product was confirmed as containing no heavy
3-[4-[1-(2-Cyano-5-methylpyrrolyl)]benzyl]-7-methyl-
2-propyl-3H-imidazo[4,5-b]pyridine (13). To a mixture of
6a (50.0 kg, 197 mol) and sodium borohydride (22.3 kg,
590 mol) in tetrahydrofuran (250 L) was added methanol
(100 L) over 1 h maintaining the temperature at 45-55 °C.
The reaction mixture was heated to reflux over 1 h followed
by cooling to 20-25 °C. After addition of water (150 L) to
the reaction mixture, the layers were separated and the
aqueous layer was re-extracted with methylene chloride (250
L, 50 L). The combined organic layers were washed with
water (150 L, 150 L) and concentrated to an oil under
reduced pressure. The residue was dissolved in methylene
chloride (418 L), treated with triethylamine (23.9 kg, 237
mol), and cooled to 5 °C. To this solution was added
methanesulfonyl chloride (22.6 kg, 197 mol) as fast as
possible (over about 30 min on this scale) keeping the
internal temperature under 10 °C. After the completion of
the reaction, water (209 L) was added over 30 min maintain-
ing the temperature in the range of 5-10 °C. The layers
were separated, and the organic layer was washed with 6%
hydrogen chloride in water (209 L) and concentrated under
reduced pressure to give crude 10a as an oil. 1-(4-
Methanesulfonyloxymethylphenyl)-5-methypyrrole-2-carbo-
1
metals by the X-ray analyzer: mp 150-151 °C; H NMR
(200 MHz, CDCl3) δ 1.00 (t, 3H, J ) 7.0 Hz), 1.69-1.91
(m, 2H), 2.10 (s, 3H), 2.72 (s, 3H), 2.87 (t, 2H, J ) 7.0
Hz), 5.57 (s, 2H), 6.06 (d, 1H, J ) 4.5 Hz), 6.86 (d, 1H, J
) 4.5 Hz), 7.08 (d, 1H, J ) 5.0 Hz), 7.28 (s, 4H), 8.23 (d,
1H, J ) 5.0 Hz); IR (Nujol) 2216, 1608, 1519, 1410, 1378,
1327, 1243 cm-1; MS (EI) m/z 370 (M + H)+. Anal. Calcd
for C23H23N5: C, 74.77; H, 6.27; N, 18.96. Found: C, 74.37;
H, 6.25; N, 18.82.
7-Methyl-3-[4-[1-(5-methyl-2-1H-tetrazol-5-yl)pyrrolyl]-
benzyl]-2-propyl-3H-imidazo[4,5-b]pyridine (14). The
reaction vessel was rigorous cleaned and confirmed as having
no residual heavy metals, holes, or cracks by measuring
electric insulation. A mixture of 13 (16.8 kg, 45.5 mol),
triethylamine hydrogen chloride (31.3 kg, 227 mol), and
sodium azide (11.8 kg, 181 mol) in DMI (118 L) was heated
to 120 °C for 8 h and then cooled to room temperature. To
this reaction mixture were added water (84 L) and sodium
nitrite (18.8 kg, 272 mol) in water (42 L), maintaining
temperature at 20-25 °C, followed by cooling to less than
10 °C. Ethyl acetate (168 L) was added, and the mixture
was treated with cooled 6% hydrogen chloride in water to
adjust the pH to 4.5-5.0 (caution: evolution of nitrogen gas).
The reaction was continued until tests using ion chromatog-
raphy measurement confirmed no remaining azide com-
pounds. To this solution was added 4% sodium hydroxide
1
nitrile (10a): H NMR (200 MHz, CDCl3) δ 2.18 (s, 3H),
3.03 (3H, s), 5.31 (2H, s), 6.10 (d, 1H, J ) 4.5 Hz), 6.89 (d,
1H, J ) 4.5 Hz), 7.37 (d, 2H, J ) 9.0 Hz), 7.59 (d, 2H, J )
9.0 Hz). 1-(4-Chloromethylphenyl)-5-methypyrrole-2-car-
1
bonitrile (10b) was obtained by preparative TLC: H NMR
(200 MHz, CDCl3) δ 2.15 (s, 3H), 4.65 (2H, s), 6.08 (d,
1H, J ) 3.9 Hz), 6.87 (d, 1H, J ) 3.9 Hz), 7.31 (d, 2H, J )
8.4 Hz), 7.54 (d, 2H, J ) 8.4 Hz); MS (EI) m/z 230 (M+),
195, 154, 127, 89. Compound 10b was also confirmed to
contain chlorine as an element. Anal. Calcd for C13H11-
ClN2: Cl, 15.37. Found: Cl, 14.62.
To a solution of 11 (44.0 kg, 197 mol) in toluene (125
L) were added granulated sodium hydroxide (15.8 kg, 395
mol) and powdered potassium carbonate (81.7 kg, 591 mol),
236
•
Vol. 2, No. 4, 1998 / Organic Process Research & Development

in water (84 L), and the layers were separated. The organic
layer was re-extracted with 4% sodium hydroxide in water
(42 L), and the aqueous layers were combined. To this
aqueous layer were added ethyl acetate (168 L) and 36%
hydrogen chloride in water to adjust the pH to 4.5-5.0 whilst
the temperature was maintained at 13-17 °C. The separated
organic layer was concentrated to ∼84 L under reduced
pressure. To this residual organic solution was added
acetonitrile (84 L), and the resulting solution was then re-
concentrated to ∼84 L under reduced pressure. After
additional overnight stirring at 3-7 °C, the precipitate was
filtered off and washed with acetonitrile (16.8 L) to afford
14 (13.3 kg, 71% yield) of 99% purity: mp 184-185 °C;
1H NMR (200 MHz, DMSO-d6) δ 0.90 (t, 3H, J ) 7.5 Hz),
1.57-1.79 (m, 2H), 1.99 (s, 3H), 2.57 (s, 3H), 2.82 (t, 2H,
J ) 7.5 Hz), 5.60 (s, 2H), 6.18 (d, 1H, J ) 4.5 Hz), 6.80(d,
1H, J ) 4.5 Hz), 7.10 (d, 1H, J ) 5.0 Hz), 7.22 (s, 4H),
8.19 (d, 1H, J ) 5.0 Hz); IR (Nujol) 1608, 1539, 1520, 1463,
1391, 1360, 1284, 1233 cm-1; MS (EI) m/z 413 (M + H)+,
385, 287, 205, 127, 105, 75. Anal. Calcd for C23H24N8:
C, 66.97; H, 5.86; N, 27.16. Found: C, 66.51; H, 5.84; N,
26.96.
hydrogen chloride in water (3.6 L) was added to this solution
above 60 °C and was followed by addition of purified water
(60 L). To this solution was added purified 1 (12.0 g) at
30-35 °C. After the precipitation of 1, stirring was
continued at the same temperature for 30 min and at 3-5
°C for 2.5 h. The precipitate was filtered off, washed with
a cooled mixture of 2-propanol (6.0 L), purified water (12
L), and 36% hydrogen chloride (0.6 L), and dried under
reduced pressure to afford purified 1 (11.1 kg, 85% yield)
of 99% purity: mp 244-246 °C (dec); 1H NMR (200 MHz,
DMSO-d6) δ 0.91 (t, 3H, J ) 7.4 Hz), 1.69 (m, 2H), 2.00
(s, 3H), 2.70 (s, 3H), 3.23 (t, 2H, J ) 8.0 Hz), 5.82 (s, 2H),
6.19 (d, 1H, J ) 3.1 Hz), 6.94 (d, 1H, J ) 3.7 Hz), 7.27 (d,
2H, J ) 8.4 Hz), 7.43 (d, 2H, J ) 8.4 Hz), 7.50 (d, 1H, J )
5.1 Hz), 8.54 (d, 1H, J ) 4.9 Hz); IR (Nujol) 1600, 1529,
1513, 1356, 1220 cm-1; MS (EI) m/z 413, 385, 287, 205,
127, 105, 75. Anal. Calcd for C23H25N8Cl: C, 61.53; H,
5.61; N, 24.96. Found: C, 61.32; H, 5.61; N, 24.89.
Acknowledgment
We especially thank Dr. David Barrett, Medicinal Chem-
istry Research Laboratories, Fujisawa Pharmaceutical Co.
Ltd., for his interest and ongoing advice in this work.
7-Methyl-3-[4-[1-(5-methyl-2-1H-tetrazol-5-yl)pyrrolyl]-
benzyl]-2-propyl-3H-imidazo[4,5-b]pyridine Hydrochlo-
ride (1) (FR143187). A solution of 14 (12 kg, 29.1 mol) in
a mixture of 2-propanol (42 L) and purified water (14 L) at
82 °C was purified by filtration through a 0.45 µm filter.
After confirming that no crystals had precipitated, 36%
Received for review January 22, 1998.
OP9800055
Vol. 2, No. 4, 1998 / Organic Process Research & Development
•
237