Process development of the synthetic route to sulamserod hydrochloride

Article abstract of DOI:10.1021/op0002040

Sulamserod hydrochloride is a potent S-HT₄ receptor antagonist and was a clinical candidate for the treatment of gastrointestinal disorders. Process development of the fairly long synthetic route (12 linear, 14 overall steps) was undertaken. Process improvements were highlighted by aromatic chlorination choices in making dichlorobenzodioxan 2 and acetylaminochloroketone 7, a transfer hydrogenation reducing a nitro group and simultaneous aromatic dechlorination without ketone reduction to give aminoketone 5, and use of the potential mutagenic iodosulfonamide 14 to make quaternary salt 11. The chemistry was scaled-up into pilot plant reactor vessels to produce multikilogram amounts of Sulamserod hydrochloride suitable for drug development.

Full text of DOI:10.1021/op0002040

Organic Process Research & Development 2001, 5, 116−121
Process Development of the Synthetic Route to Sulamserod Hydrochloride
Bruce A. Kowalczyk,* James Robinson, III, and John O. Gardner
Chemical DeVelopment, Roche Bioscience, 3401 HillView AVenue, Palo Alto, California 94304, U.S.A.
Abstract:
multikilogram lots of Sulamserod hydrochloride was required
for the development needs from toxicology, formulation, the
clinic, and pharmacology. To meet production requirements,
process development of a synthesis resulting in a scalable
synthetic route to Sulamserod hydrochloride was needed.
The synthetic route to Sulamserod hydrochloride is
outlined in Schemes 1 and 2. The route was used by
medicinal chemistry but had several shortcomings as a route
4
Sulamserod hydrochloride is a potent 5-HT receptor antagonist
and was a clinical candidate for the treatment of gastrointestinal
disorders. Process development of the fairly long synthetic route
(12 linear, 14 overall steps) was undertaken. Process improve-
ments were highlighted by aromatic chlorination choices in
making dichlorobenzodioxan 2 and acetylaminochloroketone 7,
a transfer hydrogenation reducing a nitro group and simulta-
neous aromatic dechlorination without ketone reduction to
give aminoketone 5, and use of the potential mutagenic iodo-
sulfonamide 14 to make quaternary salt 11. The chemistry
was scaled-up into pilot plant reactor vessels to produce multi-
kilogram amounts of Sulamserod hydrochloride suitable for
drug development.
1
for supplying bulk quantities of drug substance. The route
was very long (12 linear steps, 14 overall) and required many
chromatographic purifications, and the chemistry at several
steps needed major modifications. Despite these problems,
this synthetic route was chosen and used in process develop-
ment to supply drug substance. The reasons for this decision
were lack of a significantly better alternative synthetic route,
high potency of the drug (therefore, low projected demand
for material), and belief that the existing route could be
turned into a suitable process with adequate development
time. We report the process development of the synthetic
route to Sulamserod hydrochloride in this paper.
Introduction
Sulamserod hydrochloride is a potent selective 5-HT
receptor antagonist. The pK for Sulamserod hydrochloride
4
i
Results and Discussion
The dichlorination of benzodioxan 1 was investigated
using the initial procedure of chlorine gas as the chlorination
1
,4
source. The procedure was optimized to yield 72-76%
of pure dichlorobenzodioxan 2 by using CH Cl as the
2
2
reaction solvent and a MeOH drown to force out product.
The procedure worked well and could have been used in a
pilot plant or plant environment, but there were several
problems associated with using chlorine gas. Chlorine gas
is a regulated toxic gas and in our pilot plant facilities
required specialized detection equipment which would have
necessitated additional cost and time. The amount of chlorine
gas that we could have at any one time was also strictly
limited. A procedure generating chlorine in situ by adding
hydrogen peroxide to aqueous HCl did work but was plagued
by producing colored dichlorobenzodioxan 2.
has been determined to be 9.9-10.7 in human, guinea pig,
and rat tissues using in vitro radioligand binding assays. It
has been proposed that an antagonist for the 5-HT receptor
4
1
would be useful for the treatment of gastrointestinal dis-
orders.2 A program was launched with Sulamserod hydro-
chloride as the lead clinical candidate. The production of
,3
*
To whom correspondence should be addressed. Present address: Applied
The procedure used for dichlorination of 1 was based on
N-chlorosuccinimide (NCS) as the chlorinating source. The
process was optimized using HOAc as the reaction solvent
and a water drown yielding 75% of pure dichlorobenzo-
dioxan 2. The reaction was run between 70 and 100 °C which
consumed the NCS quickly, so that the reaction exotherm
was easily controlled by the addition rate of the NCS. The
water drown was optimized to a fairly narrow range because
if there was too little water, succinimide crystallized out,
and if there was too much water, other impurities were forced
Biosystems, 850 Lincoln Centre Drive, M/S 405, Foster City, CA 94404.
E-mail: kowalcba@pebio.com
(
1) Clark, R. D.; Jahangir, A.; Flippin, L. A.; Langston, J. A.; Leung, E.;
Bonhaus, D. W.; Wong, E. H. F.; Johnson, L. G.; Eglen, R. M. Bioorg.
Med. Chem. Lett. 1995, 5(18), 2119.
(
2) Review of 5-HT
Res. ReV. 1997, 17(2), 163. Review of 5-HT
4
receptor antagonists: Gaster, L. M.; King, R. D. Med.
receptor: Bockaert, J.; Fagni,
4
L.; Dumuis, A. Handbook of Experimental Pharmacology; Baumgarten, H.
G., Gothert, M., Eds.; Serotoninergic Neurons and 5-HT Receptors in the
CNS, Vol. 129; Springer-Verlag: Berlin, Heidelberg, 1997; p 439.
4
(3) Recent medicinal chemistry on 5-HT receptor: Tapia, I.; Alonso-Cires,
L.; Lopez-Tudanca, P. L.; Mosquera, R.; Labeaga, L.; Innerarity, A.; Orjales,
A. J. Med. Chem. 1999, 42, 2870. Clark, R. D.; Jahangir, A.; Langston, J.
A.; Weinhardt, K. K.; Miller, A. B.; Leung, E.; Bonhaus, D. W.; Wong, E.
H. F.; Eglen, R. M. Bioorg. Med. Chem. Lett. 1994, 4(20), 2481. Clark, R.
D.; Jahangir, A.; Langston, J. A.; Weinhardt, K. K.; Miller, A. B.; Leung,
E.; Eglen, R. M. Bioorg. Med. Chem. Lett. 1994, 4(20), 2477.
(4) Heertjes, P. M.; Knape, A. A.; Talsma, H.; Andriesse, P. J. Chem. Soc.
1954, 18.
1
16
•
Vol. 5, No. 2, 2001 / Organic Process Research & Development
10.1021/op0002040 CCC: $20.00 © 2001 American Chemical Society and The Royal Society of Chemistry
Published on Web 12/13/2000

Scheme 1. Synthetic route to Sulamserod hydrochloride
Scheme 2. Synthetic route to iodosulfonamide 14
pletion, and desired aminoketone 5 was contaminated by
various amounts of under-reduced and over-reduced impuri-
ties 15-18. The process was most significantly improved
out. The NCS was added as a slurry in HOAc, avoiding
difficult direct solids addition to the reaction vessel.
The process optimization of the conversion of dichloro-
benzodioxan 2 to methyl ketone 3 focused on the order of
the addition of reagents and the number of equivalents
necessary for complete conversion in a reasonable time
frame. It was convenient to charge the solid components first
to the reaction vessel (dichlorobenzodioxan 2, aluminum
chloride) before sealing the porthole on the reaction vessel
by eliminating hydrogen gas and switching to a transfer
hydrogenation. The source of the hydrogen was chosen as
6
potassium formate because it produced KOH needed in the
neutralization of HCl produced from the dechlorination, and
unlike ammonium formate does not produce a sublimate
which tends to clog condenser units. By switching to the
transfer hydrogenation, the hydrogenation could be run to
completion, eliminating the under-reduced by-products, and
remarkably not causing over-reduction of the ketone func-
tionality. Typically, the transfer hydrogenation in the pilot
plant did stall out near completion but was easily finished
off by addition of a small charge of additional catalyst. THF
was added during the work-up to help fully solubilize
aminoketone 5 and force out inorganic salts before filtration.
The direct chlorination of aminoketone 5 would have been
an expedient route to aminochloroketone 8, but under all
conditions attempted, it produced unacceptably impure
product. The acetylation of aminoketone 5 was essential for
clean introduction of the chloro group. The initial procedure
2 2
and then charging the liquid components (CH Cl , acetyl
chloride). This avoided any difficult solids addition. The
crystallization procedure used MeOH as the crystallization
solvent and included a seeding during cool-down which
ensured large, easily filtered crystals.
The safety analysis and process optimization of the
nitration of methyl ketone 3 to nitroketone 4 took an
5
enormous effort and has been reported previously.
The reduction of nitroketone 4 was done in medicinal
chemistry by reduction of the nitro functionality first by
hydrogenation using Pd/C catalyst under neutral conditions
with hydrogen gas followed by dechlorination either as a
separate step or in the same pot by base addition. The
streamlining of the process by carrying out both the nitro
reduction and the dechlorination in the same reaction vessel
was very attractive but was plagued by several shortcomings.
The dechlorination was susceptible to stalling before com-
(
6) Marques, C. A.; Selva, M.; Tundo, P. J. Org. Chem. 1995, 60, 2430.
Rajagopal, S.; Spatola, A. F. J. Org. Chem. 1995, 60, 1347 Wiener, H.;
Blum, J.; Sasson, Y. J. Org. Chem. 1991, 56, 6145. Anwer, M. K.; Sherman,
D. B.; Roney, J. G.; Spatola, A. F. J. Org. Chem. 1989, 54, 1284.
(5) Kowalczyk, B. A.; Roberts, P. N.; McEwen, G. K.; Robinson, J., III. Org.
Process Res. DeV. 1997, 1, 355.
Vol. 5, No. 2, 2001 / Organic Process Research & Development
•
117

of two separate steps for the conversion of aminoketone 5
to acetylaminochloroketone 7 was replaced with a process
that in one pot did the same conversion. The one-pot process
used HOAc as the solvent and acetic anhydride for acetyl-
ation followed by NCS as the chlorinating agent.
Scheme 3. Attempted aromatic chlorine introduction
The conversion of acetylaminochloroketone 7 to enone 9
was best done in one pot but as two separate steps. The
all-at-once approach did not work nearly as well. The in-
efficiency of having to put on the acetyl group on the amino
functionality for clean chlorine introduction was of little
overall consequence, because it was put on and taken off in
the same reaction vessel as that for other steps, causing little
additional labor costs.
The optimization of the hydrogenation of enone 9 to
pyridine 10 was straightforward. The best solvent for the
hydrogenation was found to be THF. THF provided good
solubility compared to other typical hydrogenation solvents,
thus, keeping the reaction volume at a reasonable level.
The medicinal chemistry effort used bromosulfonamide
The manufactured Sulamserod hydrochloride was closely
examined at a quantification limit of 1 ppm for the presence
of the potential mutagen iodosulfonamide 14 and none was
found.
During process development several general strategies
were used. The isolation of many intermediates, which is
labor intensive, was deemed worthwhile. The loss of a single
campaign would have meant a serious delay in the program.
Therefore, the close analysis of many intermediates provided
quick feedback on how the synthesis was progressing and
left the opportunity for additional purifications before a
serious contaminant threatened a batch. Solids addition to a
reaction was avoided throughout process development. The
difficulty and safety issues associated with solids addition
was of too great a concern. A general strategy of running a
reaction followed by addition of an antisolvent and temper-
ature control to crystallize out intermediates was used
repeatedly throughout the synthesis. This strategy simplified
work-ups and made isolation by filtration convenient. The
annual demand for Sulamserod hydrochloride was not
expected to be more than 5 metric tons per year due to low
clinical dose (1-2.5 mg, q.d.). The low demand made a
relatively long synthetic route to Sulamserod hydrochloride
viable as the long-term route with expected improvements
coming from additional process development in manufactur-
ing.
19 for the conversion of pyridine 10 to quaternary salt 11,
because it was accessible in one step from 2-bromoethyl-
amine. Upon process optimization, it was found that iodo-
sulfonamide 14 was superior to bromosulfonamide 19,
because the reaction time between pyridine 10 and iodo-
sulfonamide 14 was shorter and produced purer quaternary
salt 11. The additional step required to make iodosulfonamide
14 compared to bromosulfonamide 19 was not in the long
linear route to make Sulamserod hydrochloride, and there-
fore, was not a serious cost concern. Early in the process
development effort, iodosulfonamide 14 was recognized as
a potential mutagen; thus, an Ames assay was conducted.
Iodosulfonamide 14 did prove mutagenic in the Ames
salmonella gene mutation assay. A decision was made to
outsource production of iodosulfonamide 14 to a group better
able to handle containment issues during manufacturing of
this material in reactor vessels. To expedite the outsourcing
of iodosulfonamide 14, the process development work was
done in-house, and only the manufacturing was outsourced.
The process optimization of the conversion of quaternary
salt 11 to Sulamserod hydroiodide focused on solvent and
Attempted Alternative Aromatic Chlorine Introduction
The development of a shorter route to Sulamserod
hydrochloride was desired. One strategy that was particularly
attractive was an alternate selective introduction of the
aromatic chlorine of Sulamserod. Three different routes
for the selective introduction of the aromatic chlorine of
Sulamserod were tried, see Schemes 3-5.
With the introduction of bromine into the 6- and 7-
positions of benzodioxan 1, it was planned to displace one
bromine with a chlorine atom selectively at the correct
position for Sulamserod, and reduce the other bromine, see
Scheme 3. The bromination of benzodioxan 1 proceeded
well, but the acetylation of dibromobenzodioxan 20 produced
desired 21 along with major “halogen dance” by-products
22 and 23. The nitration of 21 was fine. Unfortunately, the
attempted selective displacement of one bromine atom from
dibromide 24 only yielded dichloride 4.
2
catalyst choice. The catalyst initially used was PtO and could
not be replaced with anything less expensive despite
significant experimentation. The complete solubility of
quaternary salt 11 was determined not to be a critical factor
for the hydrogenation to proceed. The solvent for hydrogena-
tion was optimized to maximize the solubility of Sulamserod
hydroiodide produced; thus, upon completion the reaction
solution could be filtered to remove only the catalyst. One
interesting finding was the hydrogenation could be done in
DMF, but the catalyst was very difficult to filter off. An
extractive work-up procedure converted Sulamserod hydro-
iodide to Sulamserod free base. Sulamserod free base was
converted to Sulamserod hydrochloride in ethanol with HCl,
consistently producing the only known anhydrous polymorph.
118
•
Vol. 5, No. 2, 2001 / Organic Process Research & Development

Scheme 4. Second attempted aromatic chlorine
introduction
nitrogen unless noted otherwise. All reactors were glass-lined
1
steel vessels. H NMR spectroscopy was performed at 300
1
3
MHz and C NMR at 75 MHz. Reactions were monitored
for completion by removing a small sample from the reaction
mixture and analyzing the sample by TLC and HPLC.
6
,7-Dichloro-2,3-dihydro-1,4-benzodioxin (2). In a 30-
gal reactor was created a slurry of N-chlorosuccinimide
53.25 kg, 399 mol) in acetic acid (63.6 kg). Into a 50-gal
(
reactor, the N-chlorosuccinimide slurry was added over 2 h
to 2,3-dihydro-1,4-benzodioxin (1) (24.95 kg, 183 mol),
keeping the temperature between 70 and 100 °C. A rinse of
the slurry reactor and lines into the reaction with acetic acid
(
2 × 10 kg) was done. After complete addition, the reaction
was aged 30 min at 90-100 °C. To the reaction mixture
Scheme 5. Third attempted aromatic chlorine introduction
was added water (15.7 kg). The mixture was cooled to 20
°
C and aged overnight. The resulting crystals were filtered
off, washed with methanol (17.8 kg) and water (13 kg), and
dried at 60-65 °C to give 2 (28.11 kg, 75%) as a white
1
solid: mp 151.2-151.6 °C; H NMR (CDCl
3
) δ 4.24 (s, 4),
), 118.50 (CH),
Cl : C,
1
3
6
1
4
.96 (s, 2); C NMR (CDCl
3
) δ 64.24 (CH
2
23.97 (C), 142.78 (C). Anal. Calcd for C
6.86; H, 2.95. Found: C, 46.82; H, 2.91.
8
H
6
2 2
O
5-Acetyl-6,7-dichloro-2,3-dihydro-1,4-benzodioxin (3).
Into a 100-gal reactor was charged aluminum chloride (29.2
kg, 219 mol), 2 (27.86 kg, 135.9 mol), and dichloromethane
Trying to circumvent the problem of selective displace-
ment of one bromine from dibromide 24, a route incorporat-
ing a chlorine and bromine initially was attempted, see
Scheme 4. The bromination and chlorination of benzodioxan
(207 kg) followed by acetyl chloride (16.2 kg, 206 mol).
The mixture was stirred at 18-24 °C for 18 h. The reaction
mixture was added to cooled water (187 kg) in a 100-gal
reactor, keeping the temperature between 1 and 25 °C. The
resulting mixture was stirred at 20-25 °C for 1 h and then
allowed to settle, and the two layers were separated. The
organic phase was washed with 2 N HCl (99 kg), aqueous
sodium bicarbonate (50 kg), and water (99 kg). In a 50-gal
reactor, the organic phase was concentrated by distillation,
removing most of the dichloromethane. The concentrate was
dissolved in methanol (131 kg), and a portion of the solvent
was distilled out (108 kg). The solution was cooled to 45
1
yielded bromochlorobenzodioxan 25. Unfortunately, the
nitration of bromochlorobenzodioxan 25 was unselective,
yielding both nitro compounds 26 and 27. The acetylation
of bromochlorobenzodioxan 25 gave an unacceptable mixture
of 22, 23, 28-30.
A third approach to selective introduction of the aromatic
chlorine of Sulamserod planned to use the amino functional-
ity of aniline 31 to direct introduction of the other groups
onto the aromatic ring and eventually reduce off the amino
functionality, see Scheme 5. From commercially available
aniline 31 acetylation and chlorination yielded desired 33.
However, the introduction of the methyl ketone functionality
proved troublesome, requiring large excess acetyl chloride
and aluminum chloride. Methyl ketone 34 produced was also
unacceptably impure.
°
C and seeded with authentic product (100 mg) which
immediately initiated crystallization. The crystallization
mixture was further cooled to 20-25 °C and aged overnight.
Water (67 kg) was added to the crystallized solution, and
the mixture was further cooled to 10-15 °C for 1 h. The
mixture was filtered, and the cake was washed with
methanol/water (10 kg/15 kg) and dried in a forced air oven
at 40-50 °C to give 3 (32.86 kg, 98%) as a white solid:
Summary
1
mp 89.5-90.1 °C; H NMR (CDCl
3
) δ 2.52 (s, 3), 4.27 (s,
) δ 31.59 (CH ), 64.16
), 118.75 (CH), 119.18 (C), 124.95 (C),
131.46 (C), 139.25 (C), 142.86 (C), 199.46 (C). Anal. Calcd
The synthetic route to Sulamserod hydrochloride under-
went process development to yield a route which produced
multikilogram amounts of drug and was capable of transi-
tioning to a manufacturing environment. The chief process
improvements included eliminating all chromatographic
purifications, developing reactor friendly processes, and
improving the chemistry at all steps to give a reliable and
scalable process.
1
3
4
), 7.00 (s, 1); C NMR (CDCl
3
3
(CH ), 64.48 (CH
2
2
8 2 3
for C10H Cl O : C, 48.61; H, 3.26. Found: C, 48.52; H,
3.26.
5-Acetyl-6,7-dichloro-8-nitro-2,3-dihydro-1,4-benzo-
dioxin (4). See ref 5.
5-Acetyl-8-amino-2,3-dihydro-1,4-benzodioxin (5). Into
Experimental Section
a 100-gal reactor was charged potassium formate (77.7 kg,
924 mol) and water (81 kg), creating a solution. To the
solution was added methanol (51 kg). Into a 200-gal reactor
was charged 10% palladium on carbon (11.6 kg, 50% water
General Procedures. All materials were purchased from
commercial suppliers and used without further purification.
All reactions were conducted under an atmosphere of
Vol. 5, No. 2, 2001 / Organic Process Research & Development
•
119

wet), 4 (30.85 kg, 105.6 mol), water (31 kg), and methanol
226 kg). The mixture in the 200-gal reactor was brought to
gentle reflux, and the contents of the 100-gal reactor were
added over 2 h. Care was taken to avoid any ingress of air,
to 20 °C, and 4-pyridinecarboxaldehyde (9.57 kg, 89.3 mol)
and methanol (12 kg) were added. After 18 h, to the solution
was added water (172 kg). The mixture was cooled for 3 h
at 5-10 °C. The crystals were filtered off, washed with water
(
and generated CO
h at reflux an additional charge of 10% palladium on carbon
0.5 kg) slurried in methanol (7.9 kg) was added to the
reaction. After 43 h at reflux, solvent (183 kg) was distilled
out of the reaction mixture. The mixture was cooled to 30-
2
was allowed to rapidly escape. After 25
(91 and 59 kg), and dried in a N
2
/vacuum oven at 50 °C to
give 9 (20.27 kg, 80%) as a bright orange solid: mp 207.5-
1
(
208.4 °C; H NMR (DMSO-d
(br), 2), 7.35 (s, 1), 7.51 (d, 1, J ) 15.8), 7.68 (d (br), 2, J
) 4.5), 7.86 (d, 1, J ) 15.8), 8.64 (d (br), 2, J ) 4.5); 13
NMR (DMSO-d ) δ 66.34 (CH ), 67.04 (CH ), 112.03 (C),
6
) δ 4.34-4.43 (m, 4), 5.90 (s
C
35 °C, and tetrahydrofuran (165 kg) was added. The solution
6
2
2
(30-35 °C) was filtered through a bed of Celite (5 kg)
118.52 (C), 124.82 (CH), 125.70 (CH), 132.57 (C), 133.60
packed into a filter. The cake was washed with 40-45 °C
tetrahydrofuran (2 × 54 kg). In a 100-gal reactor, the filtrate
was concentrated under vacuum until the volume of solution
had been reduced to approximately 90 L, and crystallization
of product had begun. To the mixture was added water (79
kg). The solution was cooled to 20-25 °C and aged for 1 h.
The crystals were collected by filtration, washed with water
(CH), 140.56 (CH), 142.18 (C), 144.91 (C), 145.77 (C),
153.00 (CH), 188.53 (C). Anal. Calcd for C16
0.15 mol H O: C, 60.16; H, 4.20; N, 8.77. Found: C, 60.10;
H, 4.11; N, 8.75.
13 2 3
H N ClO ‚
2
8-Acetyl-7-chloro-5-(3-pyridin-4-yl-1-oxoprop-1-yl)-
2,3-dihydro-1,4-benzodioxin (10). Into a 200-gal reactor
was charged 9 (19.7 kg, 62.2 mol), 10% palladium on carbon
(2.5 kg, 50% water wet), and tetrahydrofuran (331 kg). The
mixture was stirred under a hydrogen atmosphere for 3 h.
The reaction mixture was filtered through a bed of Celite (3
kg), and the cake was washed with tetrahydrofuran (2 × 39
kg). In a 100-gal reactor the filtrate was concentrated by
vacuum distillation to a volume of 20 L. To the concentrate
was added 2-propanol (155 kg). The solution was concen-
trated by distillation (78 kg removed). The solution was
cooled to 20 °C, aged overnight, and cooled to 5-10 °C for
2 h. The crystals were filtered off, washed with 2-propanol
(
2 × 21 kg), and dried using a flow of N
2
. Into a 100-gal
reactor was charged the crude product and toluene (181 kg).
The mixture was brought to reflux and concentrated by
distillation (100 kg distillate). The solution was cooled to
5-10 °C and aged overnight. The crystals were filtered off
and washed with hexanes (2 × 22 kg) to give 5 (18.24 kg,
1
8
9%) as an off white solid: mp 142.7-143.7 °C; H NMR
(
8
3
CDCl ) δ 2.53 (s, 3), 4.30-4.38 (m, 4), 6.30 (d, 1, J )
13
.6), 7.36 (d, 1, J ) 8.6); C NMR (CDCl
3
) δ 31.36 (CH
3
),
6
3.73 (CH ), 64.43 (CH ), 106.75 (CH), 118.05 (C), 123.97
2
2
(
CH), 130.06 (C), 141.05 (C), 144.83 (C), 196.66 (C). Anal.
Calcd for C10 : C,62.17; H, 5.74; N, 7.25. Found:
C, 62.38; H, 5.71; N, 7.40.
-Acetyl-8-acetylamino-7-chloro-2,3-dihydro-1,4-ben-
zodioxin (7). Into a 50-gal reactor was charged 5 (17.9 kg,
2.6 mol) and acetic acid (140 kg) followed by acetic
(30 kg) and dried in a N
(16.01 kg, 81%) as a white solid: mp 152.1-153.5 °C; H
NMR (DMSO-d ) δ 2.88 (t, 2, J ) 7.4), 3.20 (t, 2, J ) 7.4),
4.31-4.36 (m, 4), 5.76 (s (br), 2), 7.26 (d (br), 2, J ) 4.5),
2
/vacuum oven at 50 °C to give 10
1
H11NO
3
6
5
1
3
7.28 (s, 1), 8.44 (d (br), 2, J ) 4.5); C NMR (DMSO-d
6
)
9
δ 29.05 (CH ), 42.94 (CH ), 63.46 (CH ), 64.16 (CH ),
2
2
2
2
anhydride (10.6 kg, 104 mol). After 18 h, to the reaction
mixture was added to N-chlorosuccinimide (15.1 kg, 113
mol) slurried in acetic acid (46 kg). The mixture was warmed
to 50 °C. After 4 h, the solution was concentrated by vacuum
distillation to a volume of 55 L. To the solution was added
methanol (19.2 kg). The mixture was brought to reflux and
then cooled to 35 °C which initiated crystallization. To the
crystallization mixture was added water (301 kg). The
mixture was aged at 20 °C for overnight. The crystals were
filtered off, washed with water (2 × 71 kg), and dried in a
109.14 (C), 114.88 (C), 122.12 (CH), 123.87 (CH), 129.76
(C), 139.02 (C), 143.11 (C), 149.31 (CH), 150.61 (C), 194.88
(C). Anal. Calcd for C16H N ClO : C, 60.29; H, 4.74; N,
15 2 3
8.79. Found: C, 60.22; H, 4.71; N, 8.76.
4-[3-(8-Amino-7-chloro-2,3-dihydro-1,4-benzodioxin-
5-yl)-3-oxopropyl]-1-(1-methanesulfonylaminoeth-2-yl)-
pyridinium Iodide (11). Into a 50-gal reactor was charged
10 (4.0 kg, 12.5 mol), 14 (5.3 kg, 21.3 mol), and acetonitrile
(19 kg). The mixture was refluxed for 6 h. To the solution
was added 2-propanol (48 kg). The solution was cooled to
25 °C, aged overnight, and cooled to 5-10 °C for 2 h. The
crystals were filtered off, washed with 2-propanol (8 kg),
N
2
/vacuum oven at 50 °C to give 7 (22.06 kg, 88%) as a
white solid, analytic sample crystallized from toluene: mp
1
1
(
85.3-186.8 °C; H NMR (CDCl
s, 3), 4.32-4.39 (m, 4), 7.03 (s (br), 1), 7.38 (s, 1);
NMR δ 23.0 (CH , br), 31.61 (CH ), 63.89 (CH ), 64.08
CH ), 121.69 (CH), 123.49 (C), 126.58 (C), 126.84 (C),
40.54 (C), 142.93 (C), 168.6 (C, br), 196.77 (C). Anal.
Calcd for C12 : C, 53.44; H, 4.48; N, 5.19. Found:
C, 53.15; H, 4.41; N, 5.28.
-Acetyl-7-chloro-5-(3-pyridin-4-yl-1-oxoprop-2,3-en-
-yl)-2,3-dihydro-1,4-benzodioxin (9). Into a 200-gal reactor
was charged 7 (21.5 kg, 79.7 mol), methanol (170 kg), water
57 kg), and 45% potassium hydroxide (37.6 kg, 302 mol).
The solution was refluxed for 4.5 h. The mixture was cooled
3
) δ 2.18 (s (br), 3), 2.57
and dried in a N
kg, 85%) as a white solid: mp 122.0-124.9 °C; H NMR
(DMSO-d ) δ 2.91 (s, 3), 3.18 (t, 2, J ) 6.9), 3.39 (t, 2, J )
2
/vacuum oven at 50 °C to give 11 (6.03
1
3
1
C
3
3
2
6
(
1
2
6.9), 3.53-3.57 (m, 2), 4.33-4.37 (m, 4), 4.62 (t, 2, J )
5.3), 5.80 (s (br), 2), 7.29 (s, 1), 7.35 (t, 1, J ) 6.0), 8.09 (d,
1
3
H
12NClO
4
2, J ) 6.6), 8.86 (d, 2, J ) 6.6); C NMR (DMSO-d
25.48 (CH ), 29.70 (CH ), 41.86 (CH ), 42.63 (CH ), 59.89
(CH ), 63.57 (CH ), 64.34 (CH ), 109.24 (C), 114.51 (C),
6
) δ
3
2
2
2
8
2
2
2
1
122.27 (CH), 127.48 (CH), 129.81 (C), 139.32 (C), 143.37
(C), 144.29 (CH), 162.84 (C), 194.07 (C). Anal. Calcd for
(
C
19
H
23
N
3
ClIO
5
S: C, 40.19; H, 4.08; N, 7.40. Found: C,
40.24; H, 4.06; N, 7.47.
120
•
Vol. 5, No. 2, 2001 / Organic Process Research & Development

Sulamserod hydrochloride. In a 100-gal reactor 11 (6.0
kg, 10.6 mol), platinum oxide (792 g), water (30 kg), and
methanol (48 kg) were stirred at 45-50 °C under a hydrogen
atmosphere for 18 h. The warm reaction mixture was filtered
through a filter coated with Celite (0.6 kg), and the cake
was washed with 55-65 °C methanol/water (2 × 24/12 kg).
The filtrate was concentrated in a 100-gal reactor by vacuum
distillation to a volume of 60 L. The concentrate was cooled
to 10 °C. To the concentrate was added 50% NaOH (1.0
kg), water (3 kg), and tetrahydrofuran (21 kg). The mixture
was extracted three times with ethyl acetate (60 kg, 41 kg,
8.6 mol) and methylene chloride (8 L). To the solution
was added N-methylmorpholine (1.82 kg, 18.0 mol), while
maintaining the temperature between 0 and 15 °C. Methane-
sulfonyl chloride (1.1 kg, 10.2 mol) was added to the reaction
mixture, maintaining the temperature between -25 and 20
°C. After 1.5 h, the mixture was washed with water (0.5 L),
2 N HCl (0.5 L), and water (0.5 L). The organic layer was
dried over sodium sulfate, filtered, and concentrated to yield
crude N-(2-chloroethyl)methanesulfonamide (1.24 kg, 92%)
which was used without further purification. Into a 22-L flask
was placed N-(2-chloroethyl)methanesulfonamide (1.20 kg,
7.6 mol), methyl ethyl ketone (8.0 L), and sodium iodide
(2.0 kg, 13.3 mol). The mixture was refluxed for 2 h. The
solution was cooled to 50-55 °C, filtered, and concentrated.
The residue was dissolved in ethyl acetate (4.0 L)/hexane
(1.0 L) and washed with water (0.6 L), sodium thiosulfate
(0.1 kg)/water (0.5 L), and brine (0.5 L). The organic layer
was dried over sodium sulfate, filtered, and concentrated to
a solid. The residue was slurried in hexane, filtered, and dried
41 kg). The organic extracts were combined, washed with
aqueous NaCl (50 kg), dried over sodium sulfate (20 kg),
filtered, and concentrated in a 50-gal reactor by distillation
to a volume of 13 L. To the concentrate was added ethanol
(43 kg). The solution was warmed to 70 °C. To the solution
was added concentrated hydrochloric acid (1.175 kg). The
solution was cooled to 20 °C, aged overnight, and further
cooled to 5 °C for 2 h. The crystals were filtered off, washed
with ethanol (14 kg), and dried in a N
2
/vacuum oven at 50
to give 14 (1.61 kg, 85%) as a white solid: mp 70.9-72.2
1
°
C to give Sulamserod hydrochloride (4.5 kg, 88%) as a
°C; H NMR (CDCl
3.50 (dt, 2, J ) 6.5, 6.5), 5.03 (s (br), 1); C NMR (CDCl
3
) δ 3.03 (s, 3), 3.31 (t, 2, J ) 6.5),
1
13
white solid: mp 198.9-199.9 °C; H NMR (DMSO-d
.48-1.70 (m, 5), 1.80-1.84 (m, 2), 2.83-2.95 (m, 4), 2.98
s, 3), 3.13-3.25 (m, 3), 3.40-3.52 (m, 3), 4.33-4.37 (m,
), 5.72 (s (br), 2), 7.25 (s, 1), 7.52 (t, 1, J ) 5.8), 10.59 (s
br), 1); 13C NMR (DMSO-d
) δ 28.61 (CH ), 30.10 (CH ),
), 38.95 (CH ), 52.04 (CH ), 55.55
), 64.21 (CH ), 109.16 (C), 115.31 (C),
22.18 (CH), 129.89 (C), 138. 89 (C), 142.99 (C), 196.47
Cl S: C, 47.30; H, 6.06;
6
) δ
3
)
+
1
(
4
δ 4.59 (CH
2
), 41.25 (CH
3
), 45.30 (CH
2
); ms m/z (M ) 249.
Anal. Calcd for C
3
H
8
NIO
2
S: C, 14.47; H, 3.24; N, 5.62.
Found: C, 14.51; H, 3.26; N, 5.57. Caution: This material
was mutagenic in the Ames salmonella gene mutation assay
in several tester strains. LD50 (oral-rat): 354 mg/kg.
(
6
2
2
3
2.55 (CH), 37.04 (CH
2
3
2
(CH
2
), 63.57 (CH
2
2
1
(C). Anal. Calcd for C19
H
29
N
3
2 5
O
Received for review March 31, 2000.
OP0002040
N, 8.71. Found: C, 47.20; H, 6.01; N, 8.73.
N-(2-iodoethyl)methanesulfonamide (14). Into a 22-L
flask was placed 2-chloroethylamine hydrochloride (1.0 kg,
Vol. 5, No. 2, 2001 / Organic Process Research & Development
•
121