Development of an efficient synthesis of the pyrrolquinolone PHA-529311

Article abstract of DOI:10.1021/op050251y

An efficient synthesis of N-(4-chlorobenzyl)-2-(2-hydroxyethyl)-8- (morpholin-4-ylmethyl)-6-oxo-6H-pyrrol[3.2.1-ij]quinoline-5-carboxamide (5) was developed. The route was chosen due to its reasonable length (seven steps), solubility of intermediates, and capabilities of the pilot and production facilities. The critical transformations in this route were the selective iodination of an aniline, formation of the quinolone, and Sonogashira coupling/pyrrole formation. In addition, removal of residual palladium and copper from the penultimate and final products, which was of lower concern during the discovery phase of development, became a difficult process chemistry issue on scale-up.

Full text of DOI:10.1021/op050251y

Organic Process Research & Development 2006, 10, 493−499
Development of an Efficient Synthesis of the Pyrrolquinolone PHA-529311
R. L. Dorow,* Paul M. Herrinton,* Richard A. Hohler, Mark T. Maloney, Michael A. Mauragis, William E. McGhee,
Jeffery A. Moeslein, Joseph W. Strohbach, and Michael F. Veley
Chemical Research and DeVelopment, Pfizer Global Research and DeVelopment, Pfizer Incorporated, 7000 Portage
Road, Kalamazoo Michigan 49001, U.S.A.
Abstract:
show improved activity against herpesvirus DNA poly-
merases has been described. In vitro assays have demon-
2
An efficient synthesis of N-(4-chlorobenzyl)-2-(2-hydroxyethyl)-
8-(morpholin-4-ylmethyl)-6-oxo-6H-pyrrol[3.2.1-ij]quinoline-5-
strated potent inhibition of HCMV, HSV-1, and VZV
carboxamide (5) was developed. The route was chosen due to
its reasonable length (seven steps), solubility of intermediates,
and capabilities of the pilot and production facilities. The critical
transformations in this route were the selective iodination of
an aniline, formation of the quinolone, and Sonogashira
coupling/pyrrole formation. In addition, removal of residual
palladium and copper from the penultimate and final products,
which was of lower concern during the discovery phase of
development, became a difficult process chemistry issue on
scale-up.
polymerases, with good selectivity from human R-, δ-, or
γ-polymerases (Scheme 1).
2
We were asked to develop a synthesis of compound 5
which could be used to prepare multikilogram quantities.
After examination of several feasible synthetic routes, the
route outlined in Scheme 2 was selected for scale-up based
on the following criteria: reasonable length (seven steps),
solubility of intermediates, and capabilities of the pilot and
production plants. The critical transformations in this route
were the selective iodination to provide aniline 9, cyclization
to the quinolone 11, and Sonogashira coupling/pyrrole
formation to provide 12. Removal of residual palladium and
copper from the penultimate and final products became a
Introduction
3
difficult process chemistry issue on scale-up.
The herpesviruses comprise a large family of DNA
viruses. They are also a source of the most common viral
illnesses in humans, including the herpes simplex viruses
Results and Discussion
The preparation of nitrobenzene 7 was easily accom-
plished by amination of 4-nitrobenzyl bromide, 6, using
toluene as solvent and morpholine as the base as well as the
nucleophile. Addition of 3 equiv of morpholine to a toluene
solution of 6 results in complete reaction in less than 2 h.
The product is isolated in 93% yield after workup by
crystallization from toluene/Isopar C (available from Exxon).
On 40-kg scale the reaction proceed as expected, but isolation
required two crops to provide the expected yield. Reduction
of 7 was carried out using THF as solvent and 5% Pt/C as
catalyst. After reduction and clarification, the solvent was
exchanged from THF to Isopar C to precipitate the product
(
(
HSV), cytomegalovirus (CMV), varicella zoster virus
VZV), and Epstein-Barr virus (EBV). While the treatment
of HSV infections with acyclovir and similar nucleoside
analogues was one of the first success stories in antiviral
chemotherapy, substantial unmet medical needs still remain.
In particular, effective and safe treatments for CMV, VZV,
and EBV infections remain elusive. The herpesvirus poly-
merase was chosen as the therapeutic target because poly-
merase is known to be an essential enzyme and is a proven
antiviral target. In vitro assays were developed for CMV,
HSV, VZV, and EBV viral polymerases, and for human
alpha-, delta-, and gamma-polymerases. Previously, Phar-
macia has reported naphthalenes 1, quinolines 2, oxazino-
quinolines 3, and thienopyridines 4 to possess broad-spectrum
8
, in 91% yield in the lab. On 40-kg scale the reduction
proceeded rapidly and completely; however, the yield was
improved to 94%.
1
antiviral activity against the herpesviruses. More recently,
Iodination of 8 using either N-iodosuccinimide or iodine
resulted in none of the desired monoiodide product. We then
tried using iodine monochloride as iodinating reagent. Initial
attempts to iodinate 8 used iodine monochloride (ICl) in
methylene chloride and acetic acid. These conditions resulted
in only 56% yield of the desired iodoaniline 9 and large
amounts of black tar. NMR examination of this tarry material
a novel class of compounds, the pyrroloquinolines 5, which
*
Authors to whom correspondence should be addressed. E-mail:
paul.m.herrinton@pfizer.com; roberta.l.dorow@pfizer.com.
(
1) (a) Vaillancourt, V. A.; Cudahy, M. M.; Staley, S. A.; Brideau, R. J.; Conrad,
S. J.; Knechtel, M. L.; Oien, N. L.; Wieber, J. L.; Yagi, Y., Walthan, M.
W. Bioorg. Med. Chem. Lett. 2000, 10, 2079. (b) Knechtel, M. L.; Huang,
A.; Vaillancourt, V. A.; Brideau, R. J. J. Med. Virol. 2002, 68 (2), 234. (c)
Brideau, R. J.; K., Mary L.; Huang, A.; Vaillancourt, V. A.; Vera, E. E.;
Oie, N. L.; Hopkins, T. A.; Weiber, J. L.; Wlkinson, K. F.; Rush, B. D.;
Schwende, F. J.; Wathen, M. W. AntiViral Res. 2002, 54 (1), 19. (d) Oie,
N. L.; Brideau, R. J.; Hopkins, T. A.; Weiber, J. L.; Knechtel, M. L.; Shelly,
J. A.; Anstadt, R. A.; Wells, P. A.; Poorman, R. A.; Huang, A.; Vaillancourt,
V. A.; Clayton, T. L.; Tucker, J. A.; Wathen, M. W. Antimicrob. Agents
Chemother. 2002, 46 (3), 724. (e) Schnute, M. E.; Cudahy, M. M.; Brideau,
R. J.; Homa, F. L.; Hopkins, T. A.; Knechtel, M. L.; Oien, N. L.; Pitts, T.
W.; Poorman, R. A.; Wathen, M. W.; Wieber, J. L. J. Med. Chem. 2005,
(2) (a) Hartline, C. B.; Harden, E. A.; Williams-Aziz, S. L.; Kushner, N. L.;
Brideau, R. L.; Kern, E. R. AntiViral Res. 2005, 65, 97. (b) Vaillancourt,
V. A.; Staley, S.; Huang, A.; Nugent, R. A.; Chen, K.; Nair, S. K.; Nieman,
J. A.; Strohbach, J. W. Preparation of Pyrroloquinolones as Viral DNA
Polymerase Inhibitors for Antiviral Agents. WO 0202558.
(3) Rosso, V. W.; Lust, D. A.; Bernot, P. J.; Modi, S. P.; Rusowicz, A.;
Sedergran, T. C.; Simpson, J. H.; Srivastava, S. K.; Humora, M. J.;
Anderson, N. G. Org. Process Res. DeV. 1997, 1, 311 and references therein.
4
8, 5794.
1
0.1021/op050251y CCC: $33.50 © 2006 American Chemical Society
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
493
Published on Web 04/18/2006

Scheme 1. Reported polymerase inhibitors
Scheme 2. Synthetic scheme chosen for developmenta
a
2 2 2 2 5 2 3 2
Reagents and conditions: a) morpholine, toluene; b)H , 5%Pt/C, THF; c) ICl, HOAc, MeOH, CH Cl ; d) DEEM, toluene; e) P O , MsOH; f) CuI, Cl Pd(PPh ) ,
TEA, 3-butyn-1-ol , EtOH; g) 4-chlorobenzylamine, ethylene glycol.
showed it to be a single material, closely resembling the
desired product. The aromatic ring appeared to be iodinated,
but the morpholine ring is seen to be nonsymmetrical.
Stirring this tar in methanol and aqueous sodium bisulfite
leads to recovery of 9 along with some very polar products.
The desired monoiodide prepared by this method required
purification by column chromatography before it could be
carried into the next step. We subsequently found that
iodination with iodine monochloride in a mixture of meth-
ylene chloride/methanol/acetic acid gave good yields of the
monoiodinated product and that running at low temperature
the reaction to less than 3% starting material. Using more
than 2.0 equiv led to some overreaction byproducts, but the
yield was not significantly reduced. The major side product
isolated from this reaction was the diiodide 13, resulting from
(
<0 °C) improves the selectivity and yield in the reaction.
The stoichiometry of acetic acid was found to be
an apparent ipso attack of iodine and loss of the morpholi-
nomethyl group. The optimized conditions of adding 1.25
equiv of ICl with 2.0 equiv of acetic acid at -15 °C and
then warming to 10-15 °C gave the cleanest conversion.
Initially we purified the crude 9 by chromatography. Even
at high loading (5:1, silica/9) we achieved significant
upgrading, and the enamine 10 produced in the next step
was very clean. Although we were able to get crystalline
product from the enamine (step 4) reaction without the
important in this reaction. With less than 2 equiv of acetic
acid, the reaction stalled with more than 10% starting material
remaining and gave a dirty mixture, while with 2 equiv or
more, the reaction went well. The stoichiometry of the ICl
was more difficult to optimize. Reactions conducted with
1
.0, 1.1, 1.25, 1.5, and 2.0 equiv in the initial charge of ICl
demonstrated that at least 1.25 equiv are required to drive
494
•
Vol. 10, No. 3, 2006 / Organic Process Research & Development

Scheme 3. Cyclization Impurities
5
chromatography, the color and impurity profile were greatly
improved with chromatographed 9, and therefore we elected
to include this purification in the initial piloting.
on tertiary anilines, its use was unknown on secondary
6
anilines. We explored the use of 4.0, 3.0, 2.2, 2.0, and 1.4
equiv of phosphorus pentoxide in methanesulfonic acid. For
Iodine monochloride is available as either a solid or 1 M
solution in methylene chloride from a number of suppliers.
For our initial piloting we used the methylene chloride
solution from SAF. The iodination proceeded as expected,
and the “chromatography” was actually a filtration through
a plug of silica gel using (5:1 silica-to-crude product)
followed by washing with ethyl acetate and Isopar C until
all product was eluted. The product produced in this initial
piloting was more than 90% pure by HPLC, and a small
retention sample crystallized after standing at room temper-
ature. The yield was 74% on 40-kg scale.
these experiments we used only the commercial Eaton’s
7
reagent available from Aldrich at 7.7% w/w of P
4
O10 in
methanesulfonic acid. We found the highest yield with 4.0
equiv and observed no conversion with 1.4 equiv. The use
of 4.0 equiv of phosphorus pentoxide gave a faster, cleaner
reaction. Under these conditions there is only 2-3% of 9
formed in the reaction vs 6-7% using 2 equiv, and the
reaction is more complete. With 4 equiv the yield is 70%
on a 20-g lab scale vs 60% using 2 equiv. Pilot-scale
operations were run with 15% (w/w) P O in methanesulfonic
2
5
acid to increase volume efficiency.
Formation of the enamine 10, was a straightforward
conversion from the aniline.
Workup of the reaction was simple but volume limited.
The reaction is added to cold water, and the pH is adjusted
to 6-7 with sodium hydroxide. The product is extracted into
methylene chloride and is crystallized from methanol. We
isolated and identified only three impurities from the reaction
mixture (Scheme 3). Along with the desired 11, we found
the aniline 9, the decarboxylated analogue 14, and the
carboxylic acid 15. We prepared the authentic des-iodo
material, 16, but never observed formation of this material
during the cyclization.
Heating the aniline and diethylethoxymethylene malonate
(DEEMM) in refluxing toluene (112 °C) resulted in complete
conversion in less than 6 h. Distillation of the toluene and
addition of Isopar C led to crystallization of the desired 10
in 85% yield. On 50-kg scale the reaction proceeded as
expected, but a lower than expected yield of 74% was
recovered.
Two possible ring-closure procedures were examined for
conversion of 10 to 11. The first possibility explored was
The cyclization was run on 30-kg scale in the pilot plant,
and formation of the P O in methanesulfonic proceeded
4
the thermal closure. Heating a solution of 10 in diphenyl
2
5
ether at 240-250 °C for 1-2 h produced the desired
quinolone. The product could be crystallized by pouring the
warm diphenyl ether solution into hexanes or Isopar C. We
found that this thermal ring closure was concentration
dependent and that at less than 6 mL/g low yields and poor
product quality resulted. In addition, thermal stability data
showed that the product degraded at the reaction temperature
smoothly, requiring only about 3 h of stirring to get a clear
solution. Some problems were encountered in the extractions
which required the addition of methanol. Overall, the process
worked well, and the product was isolated in 62% yield on
this scale.
The palladium-catalyzed coupling of 11 with terminal
acetylenes was expected to provide the desired pyrrole[3.2.1-
(
∼240 °C for a neat sample), near those required for ring
8
ij]quinolines. This reaction was considered next. By gen-
closure; no reaction was observed below 225 °C. The
relatively high dilution, the very high temperature require-
ment, and the low margin for thermal stability caused us to
abandon this method for use in the pilot plant. We therefore
focused on the acid-promoted ring closure.
erating 12 first, followed by 5, metal removal could be
effected at two stages, first with the more soluble ester 12,
and then finally with amide 5. A variety of conditions were
explored to develop a consistent and high-yielding reaction.
Ethanol proved to be the best solvent for the Sonogashira
coupling even though 11 has poor ethanol solubility. To
obtain a stirrable slurry, 10 volumes of ethanol are needed,
followed by warming to effect the dissolution of 11.
To achieve complete conversion of 11 to 12, 1.5 equiv
of 3-butyn-1-ol are required. This is due to competitive
homocoupling of the alkyne to form diyne. The reaction
consistently performed well using 2 equiv of triethylamine.
Although acid-promoted ring closures for tertiary anilines
are fairly common, only a handful of reports of successful
5
closures with secondary anilines have been reported. We
chose to explore the acid-promoted reaction using Eaton’s
reagent (a solution of P
4
O10 in methane sulfonic acid). While
this reagent has been reported to affect this type of cyclization
(
(
4) Gould, R. G.; Jacobs, W. A. J. Am. Chem. Soc. 1939, 61, 2890.
5) (a) Atkins, R. J.; Breen, G. F.; Crawford, L. P.; Grinter, T. J.; Harris, M.
A.; Hayes, J. F.; Moores, C. J.; Saunders, R. N.; Share, A. C.; Walsgrove,
T. C.; Wicks, C. Org. Process Res. DeV. 1997, 1, 185. (b) Kametani, T.;
Kigasawa, K.; Hiiragi, M.; Wakisaka, K.; Kusama, O.; Sugi, H.; Kawaski,
K. J. Heterocycl. Chem. 1977, 14, 1175. (c) Willard, Alvin K.; Smith, Robert
L.; Cragoe, Edward, J., Jr. J. Org. Chem. 1981, 46, 3846. (d) Leysen, D.
C.; Zhang, M. Q.; Haemers, A.; Bollaert, W. Pharmazie 1991, 46, 557.
2 3 2
It was found that a 10:1 ratio of CuI:Cl Pd(PPh ) is
necessary to consume 11. Initially the reaction was performed
(6) Heisei. Jpn. Kokai Tokkyo Koho, 11021286, 26 Jan 1999.
(7) Eaton, P. E.; Carlson, G. R.; Lee, J. T. J. Org. Chem., 1973, 35, 4071.
(8) Blurton, P.; Brickwook, A.; Dhanak, D. Heterocycles 1997, 45, 2395.
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
495

Table 1. Metal removal from isolated 12
ethanethiol allowed the removal of copper, but it had no
effect on the palladium levels (entry 7). The use of an
alumina plug was productive as were 1,2-diphenylphosphi-
noethane and 1,3-diphenylphosphinopropane (entries 10-
sample
product
entry
treatment
EtOAc
ACN
solid TMT;
EtOAc
Pd
Cu
Pd
Cu
1
2). Unfortunately, the diphosphines were ineffective when
1
2
3
397
89
115
14
328
170
241
18
71
13
27
71
used with the crude reaction mixture.
Despite a great deal of effort, the consistent removal of
palladium and copper remained elusive. Fortunately, the
discovery that 12 is soluble in aqueous acid allowed a
convenient method for metal removal. When the reaction is
complete, the ethanol was removed by distillation followed
by the addition of 6.3 volumes of water and 1.3 equiv of
HCl. The resultant slurry was filtered through a filter aid.
The pH needed to be less than 2, otherwise the protonated
product was sequestered by the filter aid. The use of greater
amounts of HCl caused an increase in the copper levels in
the isolated 12. The aqueous filtrate was treated with 80 wt
% of Deloxan THP (available from Degussa AG) overnight
4
5
6
7
8
9
NH
acetone/MTBE
silica gel;
acetone/MTBE
2-hydroxypyridine;
EtOH/Hex
4
O
2
CH;
260
260
1
1
238
83
1
1
1551
1551
250
96
96
13
13
1113
1608
204
80
10
13
13
2 2 2
Me NCH CH SH;
MTBE
IRC-748;
MTBE
IRC-50;
MTBE
250
143
1
1
0
1
alumina
diphosphine;
MTBE
1551
151
96
50
254
27
20
10
2 3
followed by filtration and basification with Na CO . The
12
diphosphine;
MTBE
151
50
29
10
pyrrole was then isolated by methylene chloride extractions
and was crystallized from acetonitrile/tert-butyl methyl ether
to give 12 containing 20 ppm Pd and 2 ppm Cu in 60-70%
yield. This treatment consistently yields 12 with acceptable
metal levels. Compound 12 is a canary-yellow-colored solid,
but it turns purple upon exposure to light in the solid phase.
Thus, 12 must be stored in the dark.
2 3 2
with 10 mol % CuI and 1 mol % Cl Pd(Ph ) in refluxing
ethanol. Pyrrole formation was expedient under these reaction
conditions with no evidence of the uncyclized material.
Ultimately, it was determined that the optimal reaction
conditions are the use of 1.5 equiv of 3-butyn-1-ol, 2 equiv
An alternate metal-removal method was also developed.
Rather than treating the acidic filtrate with Deloxan THP,
2 3 2
of triethylamine, 5 mol % CuI and 0.5 mol % Cl Pd(Ph ) ,
in 10 volumes of ethanol as solvent with removal of half of
the ethanol by in-process atmospheric distillation. The
reaction took 12-17 h under these conditions. The reaction
was clean with no evidence of des-iodo 16 as a possible
byproduct in the Sonogashira coupling. To ease the metal
contamination issue, the Sonogashira coupling/pyrrole for-
2 3
the pH was adjusted to 9 with aqueous K CO , and 12 was
extracted into methylene chloride. The metals were removed
by stirring the methylene chloride solution over 20 wt %
Si-Thiol, a thiol end-capped silica gel (available from
SiliCycle) for 12 h. In the pilot plant on 20-kg scale 12 was
isolated by crystallization from acetonitrile/tert-butyl methyl
ether in 84% yield with 38 ppm Pd and 42 ppm Cu. Further
reduction in the metal levels was achieved by stirring a
methylene chloride solution of 12 over 7 wt % Si-Thiol to
give the desired product in 76% overall yield containing 17
ppm Pd and 1 ppm Cu.
In the final step ester, 12, is converted to amide 5. To
achieve an efficient amidation, 12 was slurried in ethylene
glycol with 3 equiv of 4-chlorobenzylamine. The slurry was
heated at 130 °C to give a homogeneous solution, which
was heated for 8 h. Upon completion of the reaction, the
mixture was cooled to 105-110 °C and diluted with toluene.
If toluene was not added, the reaction mixture solidified at
temperatures below 100 °C. The resulting solution was
slowly cooled to room temperature with added acetonitrile
to aid in ethylene glycol removal. As long as product
precipitation occurred at temperatures above 38 °C a filter-
able solid resulted. Since 4-chlorobenzylamine readily forms
an insoluble carbonate salt, the excess amine must be
removed. This was accomplished with pivaldehyde to form
imine 17. Although there are 2 equiv of 4-chlorobenzylamine
remaining, it was found that only1 equiv needs to be
converted to the imine. Crude 5 was isolated by filtration
and washed with 2:1 toluene/acetonitrile.
4
mation was attempted in the absence of copper with Bn -
9
NCl/Pd(OAc)
2
catalysis. Even after heating for 4 days, there
was no reaction.
To isolate the product, much of the ethanol was removed
by distillation followed by addition of water and vacuum
distillation to remove the remaining ethanol. Celite was added
to the aqueous solution followed by acidification to pH 1
with HCl. If Celite was added after the HCl, the filtration
was very slow. The resulting slurry was filtered with a water
rinse. When 12 was isolated from the filtrate at this stage, it
contained 242 ppm of Pd and 105 ppm of Cu. Therefore,
studies were initiated to optimize metal reduction as shown
in Table 1. The recrystallization solvent was examined, and
as shown in entries 1 and 2 of Table 1, the use of acetonitrile
offered a large benefit in terms of metal removal. Thus,
subsequent isolations of 12 employed an ACN/MTBE solvent
system. Stirring over solid trimercaptotriazine or Amberlite
IRC resins, known metal chelators, was not useful (entries
3, 8-9). Ammonium formate or 2-hydroxypyridine were
ineffective (entries 4, 6). While silica gel partially decreased
the Pd levels, it was insufficient (entry 5). Dimethylamino-
(
9) Jeffery, T. J. Chem. Soc., Chem. Commun. 1984, 1287. (b) Mori, A.;
Kawashima, J. Shimada, T.; Suguro, M.; Hirabayashi, K.; Nishihara, Y.
Org. Lett. 2000, 2, 2935.
496
•
Vol. 10, No. 3, 2006 / Organic Process Research & Development

5
3 4 2
0:50 CH CN/0.1 M NH OAc in H O; GC analyses were
carried out with a Hewlett-Packard 5890a gas chromatograph.
ICP analysis was used to determine residual palladium and
copper levels.
4-(4-Nitrobenzyl)morpholine, (7). To a 2-L, three-neck,
Purification of 5 was achieved by dissolving in hot 50%
EtOH/THF followed by clarification. The filtrate was treated
with hexanes to complete the product precipitation. After
cooling, 5 was isolated by filtration. Removal of 2% 17 was
effected by a simple re-slurry of 5 in 50% EtOH/THF to
give the desired product in 67% yield from 12. On pilot-
plant scale the desired product was isolated in 74% isolated
yield.
round-bottom flask, fitted with mechanical stirring, thermo-
couple, 250-mL addition funnel, and nitrogen inlet were
charged 4-nitrobenzylbromide (100 g, 463 mmol) (Caution.
Strong irritant and lachrymator! Weigh in hood with
door down and transport in stoppered flask!) and 500
mL of toluene. A solution of morpholine (121 g, 1389 mmol)
in 100 mL of toluene was added dropwise, keeping the
temperature less than 50 °C. The solution was rinsed in with
5
0 mL of toluene. The resulting solution was stirred at less
Conclusions
than 50 °C until the reaction was complete. TLC analysis;
add 1 mL of reaction mixture to 1 mL of water, spot upper
layer and elute with 1:1 ethyl acetate/heptane, visualize with
UV. When complete, 500 mL of water was added, and the
phases were separated. The organic phase was washed with
We were able to develop a scalable process for the
preparation of 5 in seven steps and 15.5% overall yield from
readily available commercial materials and demonstrate that
process in the pilot plant. Key features of this process are
the rugged ortho iodination of an aniline and reproducible,
acid-catalyzed cyclization of 10 to provide high-quality,
crystalline 11. Sonogashira coupling/pyrrole formation then
provides a high yield of crude 12, which due to its solubility
in aqueous acid facilitates removal of large quantities of
metals. The fact that 12 is methylene chloride soluble allows
the removal of the residual metals to an acceptable level.
Both Deloxan THP in aqueous acid and Si-Thiol in
methylene chloride effect the removal of palladium and
copper from 12.
5
00 mL of saturated sodium bicarbonate. The organic phase
was concentrated to less than 150 mL volume on the rotovap
with a 60-70 °C bath. Isopar C (500 mL) was added to
crystallize the product. The slurry was stirred at 20-25 °C
for 30 min, filtered, and washed with 100 mL of Isopar C.
The product was dried at 45 °C in a vacuum oven overnight.
The yield was 96.0 g, 93%, of 8: mp 80.3-82.1 °C. HPLC
1
purity 95%. H NMR (400 MHz, CDCl
3
) δ 8.01 (d, J ) 9
Hz, 2 H), 7.43 (2, J ) 8 Hz, 2 H), 3.62 (m, 4 H), 3.50 (s, 2
1
3
H), 2.37 (m, 4 H); C NMR (100 MHz, CDCl
1
5
3
) δ 129.9,
: C,
9.45; H, 6.35; N, 12.61; found: C, 59.52; H, 6.34; N, 12.58.
-(4-Aminobenzyl)morpholine (8). To a 1-L autoclave
were charged 5% Pt/C (3.3 g, 50% water wet) and 7 (125 g,
63 mmol). THF (600 mL) was added and the mixture
14 2 3
23.9, 67.3, 62.9, 54.0. Anal. Calcd for C11H N O
Experimental Section
4
General Procedures. All reagents were commercially
obtained and used as received unless otherwise noted. All
nonaqueous reactions were performed in dry glassware under
an atmosphere of dry nitrogen. Air- and moisture-sensitive
liquids or solutions were transferred via syringe or polypro-
pylene cannula. Organic solutions were concentrated by
rotary evaporation at ∼80 mmHg at less than 60 °C except
where noted. Chromatographic purification of products was
accomplished using forced flow chromatography on EM
silica gel 60. Thin-layer chromatography was performed on
Analtech Chromatography Products Uniplate silica gel GF
5
purged with nitrogen (3×) and then hydrogen (3×). The
resulting mixture was stirred under hydrogen pressure (50
psig) at high rpm (>300) at 60-70 °C until the uptake of
hydrogen stopped (about 2 h). The resulting mixture was
cooled to room temperature and filtered. The filter cake was
washed with THF (2 × 200 mL), and the filtrate and rinses
were charged to a 1-L, three-neck, round-bottom flask, fitted
with mechanical stirring, thermocouple, and distillation head.
The solution was vacuum distilled to a volume of 150 mL,
and 600 mL of Isopar C was added. The mixture was
concentrated to a volume of 200 mL and cooled to less than
0.25 mm plates. Visualization of the developed chromato-
gram was performed by fluorescence quenching or phos-
phomolybdic acid (PMA) stain and/or 50% sulfuric acid char.
NMR spectra were measured on a Bruker AM-400
operated at 400 and 100 MHz, for H and C, respectively,
with data reported as follows: chemical shift (ppm),
multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; m,
20 °C. The resulting slurry was cooled to 0-5 °C and stirred
1
13
for about 1 h. The slurry was filtered and washed with 100
mL of Isopar C. The product was dried in the vacuum oven
at 50 °C and provided 99.4 g, 91% yield of 8: mp 100.8-
1
1
multiplet), integration and coupling constant (J, Hz). H-
102.2 °C. HPLC purity 98%. H NMR (400 MHz, CDCl
3
)
1
3
C Multiplicities are reported on the basis of DEPT data.
δ 7.05 (d, J ) 6 Hz, 2 H), 6.43 (d, J ) 8 Hz, 2 H), 3.49 (m,
1
3
Elemental analysis, Karl Fishers (KF), melt solvates, infrared
spectra (IR), and thermogravimetric analyses (TGA) were
obtained from Pharmacia and Upjohn Physical and Analytical
Chemistry. HPLC analyses were carried out on an Agilent
4 H), 3.19 (s, 2 H), 2.22 (m, 4 H); C NMR (100 MHz,
CDCl ) δ 145.9, 130.9, 115.6, 67.3, 63.4, 53.8; Anal. Calcd
O: C, 68.72; H, 8.39; N, 14.57; found: C,
3
16 2
for C11H N
68.94; H, 8.34; N, 14.58.
1
100 System. The following HPLC method was employed:
2-Iodo-4-(4-morpholinylmethyl)benzenamine (9). To a
2-L, three-neck, round-bottom flask fitted with mechanical
stirring, thermocouple, 500-mL addition funnel, and nitrogen
stationary phase: 4.6 mm × 250 mm Zorbax RX-C8; flow
rate ) 0.7 mL/min; UV detection at 254 nm; mobile phase
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
497

inlet was charged 8 (40 g, 208 mmol) followed by methylene
chloride (400 mL), methanol (80 mL), and acetic acid (24.9
g, 416 mmol). The resulting solution was cooled to -20 to
(q, J ) 7 Hz, 2 H), 4.34 (q, J ) 7 Hz, 2 H), 3.79 (m, 4 H),
3.51 (s, 2 H), 2.51 (m, 4 H), 1.46 (t, J ) 7 Hz, 3 H), 1.40
1
3
(t, J ) 7 Hz, 3 H); C NMR (100 MHz, CDCl
166.1, 151.6, 140.6, 140.2, 130.8, 116.4, 95.6, 89.3, 67.3,
62.3, 60.9, 60.6, 53.9, 14.8, 14.7; Anal. Calcd for C19
IN : C, 49.73; H, 5.16; N, 5.78; found: C, 49.62; H, 5.14;
N, 5.68.
3
) δ 168.7,
-
15 °C. A 1 M iodine monochloride solution in methylene
chloride (370 g, 260 mL, 260 mmol) was added over about
0 min, keeping the temperature less than -15 °C. When
25
H -
3
2 5
O
the addition was completed, the cooling was removed and
the reaction warmed to 10-15 °C. After 1 h, the reaction
progress was monitored by GC: 1 mL of the reaction mixture
was quenched into 1 mL of 10% aqueous sodium sulfite,
and 1 mL of saturated bicarbonate was added; 5 drops of
the lower phase was diluted with 1 mL of methylene chloride,
and this was assayed by GC (injector 250 °C, initial oven
temp 125 °C, initial time ) 0, oven temp increased at 10
Ethyl 8-Iodo-6-(morpholin-4-ylmethyl)-4-oxo-1,4-di-
hydroquinoline-3-carboxylate (11). Preparation of Eaton’s
Reagent. To a 250-mL, three-neck, round-bottom flask fitted
with mechanical stirring thermocouple and nitrogen inlet,
was charged phosphorus pentoxide dimer (23.3 g, 82 mmol).
The stirring was started briefly to distribute the solids, and
the stirrer was raised above the solids in the flask. Meth-
anesulfonic acid (178 g, 120 mL, 1852 mmol) was added
and the mixture warmed to 90 °C until all the phosphorus
pentoxide dimer was dissolved. The resulting solution was
cooled to less than 30 °C.
Cyclization Reaction. To a 500-mL, three-neck, round-
bottom flask fitted with mechanical stirring, thermocouple,
and nitrogen inlet, was charged 10 (20 g, 41 mmol) followed
by Eaton’s reagent. There was an exotherm to about 47 °C.
The resulting dark-red solution was heated to 90 °C for at
least 4 h. The reaction was checked by TLC (about 0.5 mL
of reaction was quenched into 3-4 g of ice and the pH
adjusted to 7-8 by addition of 50% NaOH, 2 mL of
methylene chloride were added, and the lower layer was
spotted and eluted with 5% methanol/95% methylene
chloride). When completed, the reaction was cooled to less
than 30 °C and poured into 200 g of ice. A solution of 50%
NaOH (180 g, 2258 mmol) in 150 mL of water was slowly
added until the pH was 5-6. Methylene chloride (200 mL)
was added, and the addition of NaOH solution was resumed
until the pH was 7-8, keeping the temperature less than 30
°C (use an ice bath if needed). The phases were separated,
and the aqueous was washed with methylene chloride (1 ×
100 mL). The combined organic phases were filtered through
5 g of silica gel, and the cake was washed with 300 mL of
methylene chloride. The filtrates were concentrated on the
rotovap to about 150-mL volume. Methanol (200 mL) was
added and the solution concentrated to about 50-mL volume.
Methanol (100 mL) was added and the mixture again
concentrated to about 50-mL volume. The slurry was cooled
to 0 °C, filtered, and washed with 50 mL of methanol. The
product was dried in the vacuum oven at 50 °C overnight to
°C/min, final oven temp -275 °C, detector -275 °C, column
1
5
5 m DB-5SM). The reaction is complete when less than
% starting material remained. When the reaction was
complete, 250 mL of 10% aqueous sodium sulfite was added,
followed by a solution of 70 g of 50% NaOH in 250 mL of
water. The phases were separated, and the aqueous phase
was washed with 100 mL of methylene chloride. The
combined organic phases were washed with 100 mL of water
and concentrated on the rotovap to a thick red oil.
The crude oil was dissolved in 200 mL of ethyl acetate,
and 200 mL of Isopar C was added. The resulting mixture
was filtered through 200 g of silica gel, and the cake was
washed with 1500 mL of ethyl acetate. The organic phases
were concentrated to a dark red oil at 60 °C on the rotovap.
The oil was used as is in the next step; however, a purified
sample did eventually crystallize: mp 55.3-62.1 °C. HPLC
1
purity 87%. H NMR (400 MHz, CDCl
3
) δ 7.74 (s, 1 H),
7
4
.25 (d, J ) 8 Hz, 1 H), 6.82 (d, J ) 8 Hz, 1 H), 3.85 (m,
H), 3.51 (s, 2 H), 2.58 (m, 4 H); 13C NMR (100 MHz,
) δ 146.4, 139.9, 130.9, 129.3, 115.3, 84.3, 66.9, 62.5,
CDCl
3.8.
Diethyl 2-({[2-Iodo-4-(morpholin-4-ylmethyl)phenyl]-
3
5
amino}methylene)malonate (10). To a 1-L, three-neck,
round-bottom flask, fitted with mechanical stirring, thermo-
couple, distillation head, and nitrogen inlet was charged 9
(50 g, 157 mmol) followed by diethylethoxymethylene
malonate (37.4 g, 173 mmol) and toluene (115 mL). The
resulting mixture was heated under nitrogen with a heating
mantle with set point 120 °C. Toluene and ethanol (about
65 mL) were distilled during the heating period. When the
pot temperature reached 120 °C, the distillation slowed. The
reaction was assayed by TLC (elute with 1:1 ethyl acetate/
heptane) and was judged complete (starting material less than
provide 11.0 g (61% yield) of 11: mp 195-203 °C. HPLC
1
purity 96%. H NMR (400 MHz, DMSO-d
6
) δ 8.31 (s, 1
5%). The reaction mixture was cooled to 50 °C and vacuum
H), 7.99 (s, 1 H), 7.93 (s, 1 H), 4.06 (q, J ) 7 Hz, 2 H),
distilled to about 100 mL volume. Isopar C (200 mL) was
added and the mixture cooled to 20-25 °C, where the
mixture crystallized over ∼30 min. The resulting slurry was
stirred at 20-25 °C for 1 h, filtered, and washed with 75
mL of Isopar C. The product was dried in the vacuum oven
at 50 °C overnight, providing 65 g of 10. HPLC purity of
sample was 92%. An analytically pure sample was prepared
3.40 (m, 4 H), 3.37 (s, 2 H), 2.21 (m, 4 H), 1.11 (t, J ) 7
1
3
Hz, 3 H); C NMR (100 MHz, DMSO-d
146.1, 143.5, 138.9, 136.4, 128.1, 126.5, 110.2, 66.5, 61.3,
60.2, 53.3, 14.6; Anal. Calcd for C17 : C, 46.17;
6
) δ 173.2, 164.7,
2 4
H19IN O
H, 4.44; N, 6.35; found: C, 46.43; H, 4.37; N, 6.38.
Ethyl 2-(2-2-Hydroxyethyl)-8-(morpholin-4-ylmethyl)-
6-oxo-6H-pyrrol[3.2.1-ij]quinoline-5-carboxylate (12). 11
(985 g; 2.23 mol), CuI (42.4 g; 0.22 mol; 0.1 equiv), and
1
by recrystallization from toluene: mp 98-101 °C. H NMR
(400 MHz, CDCl
3
) δ 8.50 (d, J ) 13 Hz, 1 H), 7.89 (s, 1
Cl
2 3 2
Pd(PPh ) (15.7 g; 0.0226 mol; 0.01 equiv) were com-
H), 7.43 (d, J ) 8 Hz, 1 H), 7.23 (d, J ) 8 Hz, 1 H), 4.44
bined in a 22-L flask under nitrogen. Absolute ethanol (10
498
•
Vol. 10, No. 3, 2006 / Organic Process Research & Development

L) was added followed by triethylamine (630 mL; 4.52 mol;
equiv) and 3-butyn-1-ol (257 mL; 3.39 mol; 1.5 equiv).
continued cooling to 75 °C. Acetonitrile (1.2 L) was added
followed by the addition of pivaldehyde (215 mL; 2.1 mol;
1 equiv), precipitation occurred upon continued cooling to
room temperature. The slurry was stirred at room temperature
overnight and filtered on a fritted glass funnel. The poorly
filtering cake was rinsed with ethanol (2 L) and acetonitrile
(2 × 1 L), and the solids were dried in a vacuum oven at 55
°C for 2 days yielding 1318 g (132%) of the crude 5. Crude
5 (660 g; 1.38 mol) was added to a 22-L, three-necked flask
under nitrogen. Tetrahydrofuran (6 L) and ethanol (6 L) were
added, and the slurry was heated to 65 °C to effect
dissolution. The solution was transferred to a clean, 22-L
flask through a heated in-line 0.6 µm filter using nitrogen
pressure. The solution was reheated to 65 °C, and hexane
(6 L) was added followed by cooling to room temperature
and then to 5 °C. The slurry was filtered and rinsed with
hexane; the solids were dried in a vacuum oven at 60 °C for
2
The mixture was fitted with a distillation head, and 5 L of
ethanol was distilled. HPLC analysis of the pot residue
revealed that the reaction was complete. Water (7 L) was
added to the mixture, and the reaction was concentrated by
vacuum distillation to remove the remaining ethanol. The
solution was cooled to room temperature. Celite (450 g) was
then added followed by the addition of 1 N HCl (3 L; 1.3
equiv) to yield a mixture of pH 1.2. The mixture was filtered
through a Celite-coated, fritted glass funnel, and the filter
cake was rinsed with 3 L of water. The filtrate was
transferred to a 22-L flask and stirred overnight in the
presence of Deloxan THP (726 g). The slurry was filtered
on a fritted glass funnel with a 700-mL water rinse. The
filtrate was transferred to a 20-L wash tank, and solid Na
CO (287 g; 2.71 mol; 1.2 equiv) was added portionwise
Danger: foaming!) to yield a solution of pH 9. The solution
2
-
3
1
(
24 h to give 395 g of 5 (79% yield). H NMR analysis shows
was extracted with methylene chloride (3 × 11 L). The
organic layers were transferred portionwise to a 22-L flask
and concentrated by vacuum distillation to a final volume
of 6 L. Acetonitrile (1.3 L) and MTBE (7 L) were added,
and the solution was concentrated to a final volume of 6 L
by vacuum distillation. MTBE (7 L) was added, and the
solution was concentrated to 6 L. MTBE (7 L) was again
added, and the solution was concentrated to a final volume
of 4 L. The resulting yellow slurry was cooled in an ice bath
and then filtered onto a fritted glass funnel with a MTBE
rinse. The resulting yellow solids were dried in a vacuum
oven at 50 °C overnight to give 519 g of the light-sensitive
the presence of 2-4% imine, 17. High-quality material is
obtained by adding tetrahydrofuran (7.3 L) and ethanol (7.3
L) to 5 (1475 g) in a 22-L flask. The resulting slurry was
stirred for 3 h at room temperature followed by cooling to 5
°C. The slurry was filtered on a fritted glass funnel with a
2-L hexane rinse. The solids were dried in a vacuum oven
at 60 °C overnight to give 1336 g (67% from 12) of 5 as an
1
off-white solid: mp 179.5 °C. HPLC purity >98%. H NMR
(400 MHz, DMSO) δ 10.14 (t, 1H, J ) 6.1), 9.20 (s, 1H),
8.10 (s, 1H), 7.95 (s, 1H), 7.92 (s,1H), 7.38 (t, 4H, J ) 9.0),
6.89 (s, 1H), 4.98 (t,1H, J ) 4.8), 4.58 (d, 2H, J ) 6.0),
3.77 (m, 2H), 3.67 (s, 2H), 3.56 (m, 4H), 3.16 (t, 2H, J )
1
3
1
2 in 60% yield. ICP analysis showed 2 ppm Cu and 20
6.1), 2.37 (m, 4H); C NMR (100 MHz, DMSO) δ 178.5,
163.9, 141.7, 139.7, 138.5, 133.7, 131.5, 129.9, 130.0, 128.4,
127.8, 121.8, 121.5, 115.6, 109.9, 66.2, 62.4, 60.2, 53.1, 41.7,
1
ppm Pd: mp 172 °C. HPLC purity >95%. H NMR (400
MHz, CDCl ) δ 8.85 (s, 1H), 7.84 (s, 1H), 7.72 (s, 1H),
.68 (s, 1H), 4.33 (q, 2H, J ) 7.1), 4.11 (t, 2H, J ) 5.6),
3
-
1
6
3
5
28.5; IR (cm ) 3378 m, 3228 m, 1659 s, 1627 s, 1593 m,
1542 s, 1490 s; Elem. Anal. (C26 Cl) Calcd: C, 65.06;
.68 (m, 4H), 3.58 (s, 2H), 3.55 (br s, 1H), 3.18 (t, 2H, J )
26 3 4
H N O
.6), 2.43 (m, 4H), 1.38 (t, 3H, J ) 7.1); 13C NMR (100
H, 5.46; N, 8.75; Cl, 7.39; Found: C, 65.05; H, 5.41; N,
8.73; Cl, 7.43; LT 1 ppm Cu, LT 1 ppm Pd.
MHz, CDCl
3
) δ 174.4, 163.1, 137.9, 137.4, 134.1, 132.5,
1
5
27.8, 125.0, 120.9, 120.7, 114.2, 108.3, 65.0, 61.4, 59.1,
-
1
Acknowledgment
1.6, 27.1, 12.4; IR (cm ) 3250 m, 2956 m, 1650 s, 1590
+
We gratefully acknowledge the help of Allen Scott, Mark
Lyster, Chris Lockhart, Brian Conway, and Dar Ruimveld.
Without their help we would have not been able to complete
this work in a timely manner. Thanks to Jeff Weber for the
many Pd and Cu assays that he performed, to Doug
Grevenstuk for his efforts in the pilot-plant campaign, and
to Rajappa Vaidyanathan for his suggestions on the amidation
chemistry.
s, 1551 s, 1493 s, 1221 s; MS (70 eV) 384 (M ); Elem.
Anal. (C21H N O ) Calcd: C, 65.61; H, 6.29; N, 7.29;
24 2 5
Found: C, 65.23; H, 6.39; N, 7.24; 20 ppm Pd; 2 ppm Cu.
N-(4-Chlorobenzyl)-2-(2-hydroxyethyl)-8-(morpholin-
-ylmethyl)-6-oxo-6H-pyrrol[3.2.1-ij]quinoline-5-
4
carboxamide (5). 12 (800 g; 2.1 mol) was added to a 22-L,
three-necked flask under nitrogen. Ethylene glycol (4.8 L)
and 4-chlorobenzylamine (755 mL; 6.2 mol; 3 equiv) were
added, and the suspension was heated at 132 °C for 8 h.
HPLC analysis showed 98% conversion, so the reaction was
cooled to 102 °C; toluene (2.4 L) was added followed by
Received for review December 20, 2005.
OP050251Y
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
499