Organic Process Research & Development 1999, 3, 114−120

Enantioselective Synthesis of Brinzolamide (AL-4862), a New Topical Carbonic

Anhydrase Inhibitor. The “DCAT Route” to Thiophenesulfonamides
Raymond E. Conrow,* W. Dennis Dean, Paul W. Zinke, Michael E. Deason,† Steven J. Sproull,§
Anura P. Dantanarayana, and Mark T. DuPriest
Alcon Laboratories, Inc., 6201 South Freeway, Fort Worth, Texas 76134

Abstract: 1998 as Azopt ophthalmic suspension (brinzolamide) for the

A large scale synthesis of the topical carbonic anhydrase treatment of elevated intraocular pressure associated with
inhibitors AL-4623A (13a‚HCl) and AL-4862 (13b) from glaucoma.3
3-acetyl-2,5-dichlorothiophene (“DCAT”, 1) is described. Reac-
tion of 1 with NaSBn gave thioether 2, which was converted
via sulfenyl chloride 3 and sulfenamide 5 to sulfonamide 6.
Bromination of 6 gave bromo ketone 7, which upon reduction
with (+)-B-chlorodiisopinocampheylborane and cyclization of
the resulting bromohydrin produced S thieno[3,2-e]-1,2-thiazine
8a (96% ee) after chromatography. Treatment of 8a in THF
with n-BuLi at -70 °C resulted in Li-Cl exchange. Reaction
of the thienyllithium with SO2 and hydroxylamine O-sulfonic
acid afforded bis-sulfonamide 11a. Protection of 11a as the
acetimidate 12a, followed by tosylation and amination, gave R
amine 13a. The synthesis of 13b proceeded via primary
sulfonamide 16, which was brominated, reduced, and cyclized
to give S thieno[3,2-e]-1,2-thiazine 18 (>98% ee). By virtue of
3-Acetyl-2,5-dichlorothiophene (“DCAT”, 1), available
the ionizable NH, 18 was separable from reduction byproducts
in bulk, was selected as the starting material (Scheme 1).
by base extraction. Alkylation of 18 with 3-bromopropyl methyl
The sulfonamides were envisioned to arise from thioethers
ether afforded 8b, which was converted as above, via 11b, to
via oxidative chlorination4 to the sulfonyl chloride (2 f 4
AL-4862 (13b). These procedures provided multihundred gram
f 6). This would be an operationally simpler route to
lots of 13a and 13b.
secondary sulfonamides than the lithium sulfinate method,5
facilitating the option of installing the eventual primary
We previously reported a large-scale synthesis of the sulfonamide in protected form, e.g., 9 f 10. A key strategic
water-soluble thieno[3,2-e]-1,2-thiazine-6-sulfonamide car- feature was the sequential regioselective introduction of the
bonic anhydrase inhibitor (CAI) AL-4414A starting with thioethers by nucleophilic addition-elimination, controlled
3-acetylthiophene.1 The N-methyl sulfonamide, destined to by the keto group (1 f 2) and later by the first-formed
form the thiazine ring, was formed by ketal-directed lithia- sulfonamide (8a f 9).
tion, addition of SO2 to give the lithium sulfinate, chlorination Thioether 2 formed cleanly upon reaction of 1 with benzyl
to the sulfonyl chloride, and amination with methylamine. mercaptan and NaOH in refluxing aqueous ethanol or THF.
The primary sulfonamide arose by reaction of its lithium For scale-up we used S-benzylisothiuronium chloride, pre-
sulfinate precursor with hydroxylamine O-sulfonic acid pared in situ from benzyl chloride and thiourea, as a
(HOSA).2 Early lots of a second clinical candidate, the convenient precursor of NaSBn.4a,6 The product 2 precipitated
N-methoxyethyl compound AL-4623A, were also prepared
using this chemistry. (3) (a) Clinical studies: Silver, L. H. and the Brinzolamide Primary Therapy
Study Group. Am. J. Ophthalmol. 1998, 126, 400. (b) Crystallographic study
Herein we describe an improved synthesis of CAI’s of of carbonic anhydrase isozymes complexed with brinzolamide: Stams, T.;
this structural class. One goal of these studies was to evaluate Chen, Y.; Boriack-Sjodin, P. A.; Hurt, J. D.; Liao, J.; May, J. A.; Dean, T.;
Laipis, P.; Silverman, D. N.; Christianson, D. W. Protein Sci. 1998, 7, 556.
alternatives to low-temperature lithiation for elaborating the (c) The synthetic sequence proceeding via 18 is disclosed in Alcon
sulfonamide groups. We have employed this new sequence Laboratories, Inc. U.S. patents 5,344,929 (6 Sept 1994), 5,424,448 (13 Jun
to produce multihundred gram lots of both AL-4623A and 1995), and 5,473,067 (5 Dec 1995).
(4) (a) Baker, R. H.; Dodson, R. M.; Riegel, B. J. Am. Chem. Soc. 1946, 68,
the N-methoxypropyl homologue AL-4862. The latter com- 2636. (b) Kwart, H.; Miller, R. K. J. Am. Chem. Soc. 1956, 78, 5008. (c)
pound, as the free base, was introduced by Alcon in April Langler, R. F. Can. J. Chem. 1976, 54, 498. (d) Barnette, W. E.; Dean, T.
R.; Petersen, W. C.; Wexler, B. A. US Patent 4,789,465. (e) Kim, D.-W.;
Ko, Y. K.; Kim, S. H. Synthesis 1992, 1203.
† Present address: Agouron Pharmaceuticals, San Diego, CA. (5) The method of ref 2 was not adaptable to direct synthesis of secondary
§ Present address: Zeneca LifeScience Molecules, Wilmington, DE. sulfonamides: use of N-methylhydroxylamine O-sulfonic acid (Schmitz,
(1) DuPriest, M. T.; Zinke, P. W.; Conrow, R. E.; Kuzmich, D.; Dantanarayana, E.; Ohme, R.; Murawski, D. Chem. Ber. 1965, 98, 2516) to quench lithium
A. P.; Sproull, S. J. J. Org. Chem. 1997, 62, 9372. thiophenesulfinates gave the N-methylsulfonamide in poor yield (D.
(2) Graham, S. L.; Scholz, T. H. Synthesis 1986, 1031. Kuzmich, unpublished observations).

114 • Vol. 3, No. 2, 1999 / Organic Process Research & Development 10.1021/op9802125 CCC: $18.00 © 1999 American Chemical Society and Royal Society of Chemistry
Published on Web 02/18/1999
Scheme 1 oxidative degradation en route to 4. Another pathway was
suggested by our observation that the N-methylsulfonamide
analogue of 6, upon reaction with NaSBn (DMSO, room
temperature), regenerated thioether 2 (87%) by elimination
of MeNHSO2-. Given the greater electronegativity of the
chlorosulfonyl group, an analogous addition-elimination
might compete during the amination of 4. Conversion of the
carbonyl group to the acid-soluble imine was detected in
small-scale trials; the ketone was regenerated on standing
in the acidic aqueous extract.
In anhydrous media, chlorination of thioethers affords
sulfenyl chlorides.9 Thus 3 was obtained in high yield by
reaction of 2 with Cl2 in toluene, which served well as a
replacement for the hazardous classical solvent CCl4. Toluene
also proved to be the solvent of choice for the subsequent
amination and oxidation steps, enabling a one-flask process
that proved reliable and easy to scale up. Addition of
2-methoxyethylamine to the toluene solution of 3, removal
of excess amine by acid extraction, and treatment of the
resulting solution of sulfenamide 5 with m-CPBA and
aqueous NaHCO310 gave sulfonamide 6. Trituration of the
crude product with hexane dissolved the benzylic byproducts
and precipitated 6 in 72-83% yield.
upon quenching into water. Oxidation of excess thiol in The bromination-reduction-cyclization sequence 6 f
solution with NaOCl was followed by filtration to afford 2 7 f 8a was accomplished as described for the N-methyl
in 95-97% yield on a 15-mol scale. series.1 Treatment of ketone 6 in THF with pyridinium
Conversion of 2 to sulfonamide 6 via sulfonyl chloride 4 bromide perbromide (PBP) and catalytic HBr produced
proved erratic. Addition of gaseous chlorine to 2 in acetic bromo ketone 7, typically accompanied by 10-15% of the
acid-water,4a-c followed by removal of excess Cl2 by air dibromo ketone.11 Crude 7 was reduced with (+)-B-chloro-
stream, resulted in clean formation of sulfonyl chloride 4 diisopinocampheylborane ((+)-Ipc2BCl)12 in THF at ca. -30
(TLC). Sulfenyl chloride 3 was the only intermediate °C, and the resulting bromohydrin was cyclized in situ to
observed. Quenching the reaction mixture into water invari- the thienothiazine 8a. The dibromo ketone proved to be
ably produced an oil (isolated 4 is a low-melting solid). relatively innocuous, as it was found to undergo slow
Alternatively, removal of HOAc by rotary evaporation with reductive debromination to 7 prior to carbonyl reduction.13
moderate warming caused decomposition of 4. Therefore, Consistent with our earlier observations,1 and those of
extractive workup (CH2Cl2-ice cold 2.5% NaOH)4c was used others,12d workup using the recommended12b,c diethanolamine
to provide an acid-free solution of 4. Addition of 2-meth- or acetaldehyde quenches did not adequately segregate the
oxyethylamine produced sulfonamide 6 in 75-78% yield on product 8a from pinene, isopinocampheol, and other organic-
a 2-10-g scale, but on a 50-g scale the yield dropped to soluble byproducts. A tedious chromatographic purification
30-50%. Oxidative chlorination conditions were varieds was necessary. Thienothiazine 8a of 96% ee was thereby
temperature, solvent (CH2Cl2,4d TFA, EtOAc, THF,7 mix- obtained in 55-60% yield from 6. The S configuration
tures), aqueous component (water, HCl at different concen- follows from our previous work in the N-methyl series1 and
trations,4d buffers), chlorine source (NaOCl, NCS), and (9) (a) Drabowicz, J.; Kielbasinski, P.; Mikolajczyk, M. In The Chemistry of
thioether substitution (n-Bu, i-Bu, p-methoxybenzyl)8s Sulfenic Acids and Their DeriVatiVes; Patai, S., Ed.; Wiley: Chichester,
without improvement. The benzyl sulfoxide derived from 2 UK, 1990; p 221. (b) Capozzi, G.; Modena, G.; Pasquato, L. ref 9a, p 403.
(c) Kühle, E. Synthesis 1970, 561; 1971, 563, 617. (d) Kurzer, F.; Powell,
also afforded 4 upon oxidative chlorination; amination then J. R. Organic Syntheses; Wiley: New York, 1963; Collect Vol. IV, p 934.
delivered 6 in yields comparable to those observed above. (10) The presence of aqueous bicarbonate was required for good yields in this
m-CPBA oxidation. Chlorine in aqueous HOAc, while effective for oxidizing
The crude amination mixtures contained numerous com- 3 to 4, was unsatisfactory for oxidizing 5 to 6. Oxone or H2O2 in aqueous
ponents, complicating the interpretation of these results. In HOAc converted 5 to 6 but too slowly to be preparatively useful.
contrast, a ketalized analogue of 4 had undergone amination (11) (a) Use of methanol as solvent decreased dibromination, but concomitant
ketalization proved troublesome: Gaudry, M.; Marquet, A. Organic
with little difficulty.1 The thiophene ring could have suffered Syntheses; Wiley: New York, 1988; Collect Vol. VI, p 193. (b) Phenylt-
rimethylammonium tribromide in THF also converted 6 to 7, plus dibromo
(6) (a) Speziale, A. J. Organic Syntheses; Wiley: New York, 1963; Collect ketone, at a slower rate than PBP: Jacques, J.; Marquet, A. Organic
Vol. IV, p 401. (b) Kofod, H. Organic Syntheses; Wiley: New York, 1963; Syntheses; Wiley: New York, 1988; Collect Vol. VI, p 175.
Collect Vol. IV, p 491. (12) (a) Chandrasekharan, J.; Ramachandran, P. V.; Brown, H. C. J. Org. Chem.
(7) (a) CAUTION: THF and DME undergo delayed exothermic polymerization 1985, 50, 5446. (b) Brown, H. C.; Ramachandran, P. V. Acc. Chem. Res.
upon exposure to Cl2 and are therefore not recommended as solvents for 1992, 25, 16. (c) Dhar, R. K. Aldrichim. Acta 1994, 27(2), 43. (d) Simpson,
oxidative chlorinations. (b) Acetonitrile and DMF are useful water-miscible P.; Tschaen, D.; Verhoeven, T. R. Synth. Commun. 1991, 21, 1705.
chlorination solvents; their utility was not assessed in the present studies. (13) This pathway was confirmed using isolated dibromo ketone, best prepared
(8) The tert-butyl thioether did not form efficiently from DCAT and t-BuSH by condensing 6 with DMF dimethyl acetal (1.2 equiv, MeCN, 60 °C, 70-
in THF-aq NaOH. 80%) and reacting the resulting enaminone with 2 equiv of Br2 in wet THF.

Vol. 3, No. 2, 1999 / Organic Process Research & Development • 115

Scheme 2 temperature also favored protonation of the lithiated inter-
mediates, presumably by THF: 2.3 equiv of BuLi was added
to a solution of 8b in THF at -10 °C, followed by stirring
at this temperature for 4 h, simulating a process delay.
Quenching with SO2 and HOSA as above gave a 1:1 mixture
of 8b and dechloro-8b in >90% yield, and little sulfonamide.
This “hybrid” approach to bis-sulfonamides exploits the
best features of both sulfonamide-forming methods. The
thioether route was advantageous early in the synthesis where
the scale was greatest and the properties of the keto group
could be utilized, in contrast with the cumbersome ketaliza-
tion required previously.1 The organolithium chemistry was
deployed at a more advanced stage, saved one step compared
to the route via 9, and was the preferred route to primary
sulfonamides 11 which turned out to be the intermediates
from the results obtained by Brown for phenacyl halide of choice. The low temperature required for efficient Li-Cl
reductions.12a-c exchange was not viewed as a serious drawback, as -80 °C
Reaction of 8a with NaSBn in DMF at 40-50 °C gave capability is becoming increasingly common in manufactur-
thioether 9 in 64-72% yield. Oxidative chlorination of 9 ing facilities.
(Cl2 or NaOCl, HOAc, H2O) followed by extractive workup4c There remained a tosylation-amination sequence to
and amination with t-BuNH2 provided bis-sulfonamide 10 convert 11a to 13a. In the N-methyl series, we had observed
in about 50% yield on a gram scale. Use of two-phase tosylation of the primary sulfonamide concurrent with
mixtures4d for oxidative chlorination gave inferior results.7b O-tosylation.1 A tert-butyl group, as in compound 10,
Upon scale-up, the yield of 10 declined markedly, and much effectively suppressed this side reaction and could be cleaved
water-soluble material was formed. When the sulfonyl with TFA (45 °C, 18 h) after amination. Access to 10 via
chloride was quenched with ammonia, primary sulfonamide the thienyllithium, however, required a chlorination step and
11a was obtained in poor yield. The sulfenyl chloride to was low-yielding and inconvenient. Better results were
sulfenamide alternative gave no useful results. Substitution realized via acetimidate 12a, readily prepared17 by condens-
of NHEt for OH (see below) in 9, not surprisingly, led to ing 11a with trimethyl orthoacetate. This protecting group
decomposition upon subsequent chlorination. Hydroxyl survived tosylation, then cleaved during amination to furnish
protection was not considered an attractive tactic due to 13a in about 55% yield from 11a.18 Acetonitrile proved
increasing step count. superior to THF as a solvent for the protection step, while
These results were not taken as an abrogation of the the converse was true of the tosylation and amination; hence
original plan. However, the necessity of delivering material a solvent exchange was required. Treatment of 13a with
in quantity within a practical time frame occasioned a gaseous HCl gave the target compound AL-4623A. Single-
speculative return to organolithium chemistry (Scheme 2). crystal X-ray crystallographic analysis of AL-4623A con-
To our delight, 8a underwent efficient lithium-chlorine firmed the R configuration, consistent with inversion of the
exchange,14 ascertained by quenching with MeSSMe, upon assigned S configuration of alcohol 8a.19
reaction with n-butyllithium (2 equiv) in THF at -65 to -70 The synthesis of the homologue AL-4862 was first
°C. Addition of SO2 to this dianion solution, followed by accomplished essentially as described above. As in the case
reaction of the resulting lithium sulfinate with an aqueous of 8a, chromatographic purification of the reduction-
solution of HOSA and NaOAc2 produced sulfonamide 11a cyclization product 8b constituted a troublesome bottleneck.
in 65-67% yield on a kilo scale. Conversion of the alkoxy We therefore developed a modified sequence proceeding via
group to the sulfite or sulfamate could account for some of the unsubstituted thienothiazine 18 (Scheme 3), which by
the material loss, although this was not confirmed. Proton virtue of the ionizable NH was isolable from neutral organic
transfer from OH to the thienyllithium rather than to BuLi, materials by extraction into aqueous hydroxide.20 Oxidative
a likely side reaction dependent upon mixing rate,15 would chlorination of 2 to 4 followed by introduction of ammonia
be of no practical consequence since further reaction of the
resulting dechloro compound with BuLi is regioselective for (17) (a) Loev, B.; Kormendy, M. F. Can. J. Chem. 1964, 42, 176. (b) This
condensation was later found to be markedly accelerated, and to proceed
C-6 (R) thienyllithium formation.1 Butyllithium treatment of more cleanly, in the presence of powdered NaHCO3.
8a at higher temperature (-40 to -20 °C) led to the (18) (a) The formamidine Me2NCHdNSO2R, formed by condensation of 11a
appearance of the 6-chloro-5-sulfonamide, arising from with DMF dimethyl acetal, was also examined as a primary sulfonamide
masking group. Tosylation/amination (EtNH2) then gave the expected amine,
deprotonation16 in competition with Li-Cl exchange. Higher plus an isomer having a substituted protecting group EtNdCHNHSO2R (13C
and 1H NMR, MS). The former product was cleaved to 13a by dilute acid,
(14) Li-Cl exchange is a rarely reported event, in comparison with Li-Br and but the latter material resisted 2 M HCl treatment for several days at room
Li-I exchange: (a) Slocum, D. W. FMC Lithium Link, No. 5, Winter 1993, temperature (D. Pierce, unpublished observations). (b) Variations tried on
pp 1-12. (b) March, J. AdVanced Organic Chemistry, 4th ed.; John Wiley the tosylation-amination reaction included the use of DMAP; MsCl, Ms2O,
& Sons: New York, 1992; pp 625-6. and Ts2O in place of TsCl; gaseous vs aqueous EtNH2; and Mitsunobu
(15) Gallagher, D. J.; Beak, P. J. Am. Chem. Soc. 1991, 113, 7984. conditions: Edwards, M. L.; Stemerick, D. M.; McCarthy, J. R. Tetrahedron
(16) (a) Snieckus, V. Chem. ReV. 1990, 90, 879. (b) FMC Lithium Link, No. 3, Lett. 1990, 31, 3417.
Spring 1992, pp 1-14; No. 4, Spring 1993, pp 1-17. (19) May, J. A. J. Med. Chem. Manuscript in preparation for submission.

116 • Vol. 3, No. 2, 1999 / Organic Process Research & Development

Scheme 3 For the (+)-Ipc2BCl reduction of 17, the water-im-
miscible, relatively nonhygroscopic solvent methyl tert-butyl
ether (MTBE) proved superior to THF.23 After complete
reduction (3.5 h at -25 to -20 °C), aqueous NaOH was
added to the crude mixture to cyclize the bromohydrin and
partition the product 18 into the aqueous phase. Acidification
and extraction into EtOAc was followed by dilution with
toluene to effect crystallization, affording 18 in 70-77%
yield and assaying at >98% for both HPLC chemical purity
and ee.
Reactions of 18 with 3-bromopropyl methyl ether in the
presence of a variety of bases and solvents were discouraging
at first. Slow or incomplete conversion to multiple products,
including both N- and O-alkylated materials, was observed.
gave primary sulfonamide 16 in up to 80% yield on a 10- The combination of solid potassium carbonate and DMSO
20-g scale. Once again, scale-up (>50 g) resulted in a dropoff was then found to promote N-alkylation with excellent
in yield. The preferred route to 16 proceeded via sulfenyl selectivity, affording the product 8b in 94% yield. The
remaining steps of the synthesis of 13b (AL-4862) were
chloride 3 and primary sulfenamide 14. Oxidation of 14 with
performed according to Scheme 2 and proceeded in com-
m-CPBA/NaHCO3 produced sulfinamide 15 rapidly, but
parable yields to those observed in the a series.
further oxidation of 15 to 16 was sluggish unless a large
Compound 18 also proved to be a versatile intermediate
excess of reagents was used. This contrasted with the
for preparation of diversely substituted experimental carbonic
oxidation of 5 to 6, in which the sulfinamide was observed
anhydrase inhibitors. Notably, alkylation of 18 with 1-bromo-
by TLC as a minor, transitory component. Sulfinamide 15
3-chloropropane (K2CO3, DMSO) provided the N-(3-chlo-
was poorly soluble in most common solvents, and once
ropropyl) derivative in 94% yield. Subsequent Li-Cl ex-
precipitated could not be oxidized to 16 effectively.
change selectively engaged the thienyl chloride and gave after
The combination of H2O2 (20 equiv) and sodium tungstate
SO2/HOSA treatment the 3-chloropropyl analogue of 11b
(0.5 equiv) in ethyl acetate-water21 gave superior results
(65%, mp 135-137 °C). The primary chloride could then
for the oxidation of 14 to 16. This reaction was characterized
be displaced with various nucleophiles, for example cyanide
by an accelerating exotherm from ambient to reflux tem-
which provided an entry to carboxylic ester-bearing com-
perature, requiring careful monitoring. A workable modifica- pounds.
tion involved adding the H2O2 to a preheated reaction mixture These studies resulted in practical preparative routes to
to allow better control over the oxidation rate.22 The prior AL-4623A and AL-4862. The stepwise thioetherification
chlorination and amination steps also proceeded well in ethyl strategy proved sound, but difficulties were encountered in
acetate, so a one-flask procedure for the conversion of subsequent processing to form the sulfonamide groups.
thioether 2 to sulfonamide 16 was possible. A series of eleven Consequently, the original goal of replacing both low-
1-kg runs delivered 16 in 64-79% yield and 98-99% purity. temperature lithiations was realized only in part. Lithium-
Reaction of 16 with PBP and H2SO4 in ethyl acetate gave chlorine exchange provided an efficient alternative entry to
more reliable initiation and cleaner conversion than the earlier the primary sulfonamide functionality. Use of the base-
(HBr, THF) method. The use of freshly opened PBP in slight soluble thienothiazine 18, in concert with a highly selective
deficiency (0.9 equiv), plus a generous amount of acid, method for N-alkylation, proved to be a critical advance for
minimized dibromination; dibromo ketone in small amounts high material throughput. The procedures described have
(<5%) was acceptable (see above), but in larger amounts been successfully adapted, with further refinements, to
was detrimental to the subsequent reduction. Bromo ketone manufacturing scale.
17 of >90% purity, containing e3% each of 16 and dibromo
ketone, was obtained in 72-80% yield after trituration of Experimental Section
the crude product with CH2Cl2.
General Methods. 3-Acetyl-2,5-dichlorothiophene
(20) The base solubility of secondary sulfonamides was also used to advantage
(“DCAT”) was obtained from Lancaster Synthesis. Anhy-
for isolating the waxy methoxypropyl homologue of keto sulfonamide 6, drous THF and (+)-Ipc2BCl were used as received from
which could not be separated from the crude reaction mixture by trituration Aldrich Chemical Co. Pyridinium bromide perbromide (PBP,
as done for 6.
(21) Blacklock, T. J.; Butcher, J. W.; Sohar, P.; Lamanec, T. R.; Grabowski, E. manufacturer’s assay 91-98%) was used as received from
J. J. J. Org. Chem. 1989, 54, 3907. Aldrich or Lancaster. Temperatures recorded are those of
(22) CAUTION: Addition of brine to the oxidation mixture resulted in an
exothermic reaction and eruption of the flask contents. Blank experiments
the reaction mixture. Concentration refers to removal of
verified that brine reacts exothermically with a combination of hydrogen volatile components by rotary evaporation in Vacuo. Melting
peroxide and sodium tungstate. Separation of the aqueous oxidant phase, points are uncorrected. Coupling constants (J) are reported
followed by thorough washing of the organic phase with bisulfite until a
peroxide test was negative, remedied this problem. Sulfite is more effective
than bisulfite for reducing organic peroxy compounds: Sharpless, K. B.; (23) Solutions of (+)-Ipc2BCl in MTBE could be prepared in advance and stored
Verhoeven, T. R. Aldrichim. Acta 1979, 12(4), 63-74: see p 64 and refrigerated for several days; however, upon longer storage (>1 month),
footnotes 17 and 18 therein. several bottles burst from pressure buildup.

Vol. 3, No. 2, 1999 / Organic Process Research & Development • 117

in hertz. Elemental analyses were performed by Atlantic 11.6 mol) in THF (28 L). PBP (“90%”, 3.69 kg, 10.4 mol)
Microlab (Norcross, GA). Reactions were monitored by TLC was added in 0.5-kg portions over 10 min, and the mixture
on E. Merck Silica Gel 60 F254 plates, with visualization by was stirred at 0 °C until TLC showed consumption of 6.
UV light and phosphomolybdic acid staining. Purity of Water (15 L) was added and the mixture was extracted with
reaction products was assayed by reverse-phase HPLC on EtOAc (2 × 4L). The combined organic solution was washed
C18 columns with UV detection at 255 nm. Enantiomeric with brine, dried (MgSO4), filtered, and concentrated. 2-Pro-
excesses (ee) were determined using a Daicel OF analytical panol (4 L) was added to the residue to initiate crystallization.
HPLC column with a mobile phase consisting of hexane/i- The mixture was chilled in an ice bath and the solid was
PrOH mixtures, containing 0.0-0.1% Et3N. collected by filtration, washed with 2-propanol (3 × 1.3 L)
3-Acetyl-5-chloro-2-(benzylthio)thiophene (2). A mix- and hexane (2 × 2 L), and dried in air at room temperature
ture of thiourea (1.29 kg, 16.9 mol), benzyl chloride (1.86 to give 3.02 kg (70%) of 7. This material contained 9.6% of
L, 2.04 kg, 16.1 mol), EtOH (13.5 L), and water (4.5 L) the dibromo ketone by HPLC. A sample was recrystallized
was heated to reflux over 2 h, then allowed to cool to 74 °C from EtOH: mp 84-86 °C; 1H NMR (CDCl3) δ 3.20-3.28
over 20 min. 3-Acetyl-2,5-dichlorothiophene (1) (3.00 kg, (m, 2H), 3.26 (s, 3H), 3.39-3.44 (m, 2H), 4.28 (s, 2H), 6.27
15.4 mol) was added, followed by 4 M NaOH (10 L). The (br t, 1H), 7.40 (s, 1H).
mixture was refluxed for 3 h, cooled to room temperature, (S)-3,4-Dihydro-6-chloro-4-hydroxy-2-(2-methoxyethyl)-
diluted with water (10 L), stirred for 30 min, and treated 4H-thieno[3,2-e]-1,2-thiazine 1,1-Dioxide (8a). Under N2,
with bleach (3 L of 5.25% NaOCl). After the mixture was a precooled solution of (+)-Ipc2BCl (1.96 kg, 6.09 mol) in
stirred for 30 min, the solid was collected by filtration, anhydrous THF (8 L) was added to a stirred solution of 7
washed with water (4 × 2.5 L) and 2-propanol (3 × 2 L), (1.51 kg, 4.02 mol) in anhydrous THF (28 L) keeping the
and dried in air at room temperature to give 4.22 kg (97%) temperature between -40 and -30 °C. The solution was
of 2. A sample was recrystallized from EtOH: mp 86-88 stirred for 6 h at -35 to -25 °C, then allowed to warm to
°C; 1H NMR (CDCl3) δ 2.42 (s, 3H), 4.15 (s, 2H), 7.17 (s, 20 °C over 15 h. TLC showed incomplete reaction. The
1H), 7.25-7.34 (m, 5H); 13C NMR (CDCl3) δ 28.7, 40.5, solution was cooled to -30 °C and (+)-Ipc2BCl (0.50 kg,
125.9, 127.3, 127.9, 128.7, 129.1, 134.9, 135.1, 147.2, 191.6. 1.56 mol) was added. After 1 h the solution was concentrated
Anal. Calcd for C13H11OS2Cl: C, 55.21; H, 3.92. Found: and the residual oil was dissolved in Et2O (15 L) and cooled
C, 55.34; H, 3.96. to 10 °C. Diethanolamine (1.5 kg) was added over 10 min,
N-(2-Methoxyethyl)-3-acetyl-5-chloro-2-thiophene- forming a yellow precipitate. After 30 min the mixture was
sulfonamide (6). Chlorine was bubbled into a stirred, ice- filtered and the precipitate was washed with Et2O (3 × 1
cooled solution of 2 (1.24 kg, 4.38 mol) in toluene (14 L) L). 1 M NaOH (16 L) was added and the mixture was stirred
for 25 min, keeping the temperature below 8 °C, to form at 20 °C for 15 h. The aqueous phase was separated and
sulfenyl chloride 3. Dry air was then bubbled through the extracted with EtOAc (2 × 4L). The combined organic
solution for 1.2 h. 2-Methoxyethylamine (1.13 L, 13.2 mol) solutions were washed with 1 M HCl (10 L) and brine (10
was added dropwise over 40 min, keeping the temperature L), dried (MgSO4), filtered, and concentrated. The residual
below 12 °C. The resulting solution of sulfenamide 5 was oil (2.77 kg) was combined with that from another run on
washed with 1 M HCl (2 × 3L), and the combined washes the same scale and applied as a slurry in 10% EtOAc-hexane
were back-extracted with CH2Cl2 (2 × 2L). The combined to a 30 × 50 cm silica gel column. The column was eluted
organic solutions were washed with brine. Water (15 L) and at 250 mL/min with a gradient of 10% (60 L), 20% (20 L),
NaHCO3 (1.48 kg, 17.6 mol) were added to the organic 30% (20 L), 40% (20 L), 50% (20 L), and 70% EtOAc-
solution, and the two-phase mixture was stirred vigorously hexane (40 L), followed by EtOAc (20 L), collecting 3-L
and cooled to 8 °C. m-CPBA (84%, 2.85 kg, 13.8 mol) was fractions. The appropriate fractions were combined and
added in portions over 1 h, taking care to avoid excessive concentrated to give 1.98 kg (83%) of 8a as an oil. A sample
foaming. The temperature was allowed to rise to 20 °C over was rechromatographed: [R]25D +4.0° (c ) 1, MeOH); 1H
15.5 h. NaHSO3 (1.5 kg) was added over 30 min and the NMR (CDCl3) δ 3.31 (s, 3H), 3.38-3.91 (m, 5H), 4.16 (br
phases were separated. The organic solution was washed with s, 1H), 4.33 (dd, 1H, J ) 16, 4), 4.57 (br s, 1H), 6.98 (s,
aqueous NaHCO3 and brine, dried (MgSO4), filtered, and 1H). Anal. Calcd for C9H12NO4S2Cl: C, 36.30; H, 4.06; N,
concentrated. The residual solid was triturated with hexane 4.70. Found: C, 36.23; H, 4.05; N, 4.66.
(4 L), collected by filtration, washed with hexane, and dried 3-Acetyl-5-chloro-2-thiophenesulfonamide (16). Chlo-
in air at room temperature to give 1.01 kg (77.5%) of 6. A rine was bubbled into a stirred solution of 2 (1.00 kg, 3.53
sample was recrystallized from MTBE: mp 92-93 °C; 1H mol) in EtOAc (20 L) at 2-10 °C to form sulfenyl chloride
NMR (CDCl3) δ 2.57 (s, 3H), 3.17-3.22 (m, 2H), 3.26 (s, 3. The solution was purged with a stream of air for 1 h.
3H), 3.41-3.46 (m, 2H), 6.46 (br t, 1H), 7.33 (s, 1H); 13C Ammonia was bubbled in, keeping the temperature between
NMR (CDCl3) δ 29.4, 43.6, 58.7, 70.4, 128.9, 134.9, 138.4, 5 and 15 °C, to form sulfenamide 14. The mixture was
143.9, 192.5. Anal. Calcd for C9H11NO3S2Cl: C, 36.30; H, purged with air for 1 h. Water (5 L) was added and the
4.06; N, 4.70. Found: C, 36.38; H, 4.02; N, 4.79. solution was cooled to 15 °C. Sodium tungstate dihydrate
N-(2-Methoxyethyl)-3-bromoacetyl-5-chloro-2-thiophe- (0.5 equiv, 583 g, 1.77 mol) was added, followed by 30%
nesulfonamide (7). A 30% solution of HBr in HOAc (231 H2O2 (8 L) over 5 min. The mixture was heated at 35 °C for
mL) was added to a stirred, 0 °C solution of 6 (3.43 kg, 2 h, then stirred at room temperature for 16 h. Water (5 L)
118 • Vol. 3, No. 2, 1999 / Organic Process Research & Development
was added and the aqueous phase was separated. Water (5 d6) δ 49.1, 61.7, 127.1, 132.7, 133.3, 146.6. Anal. Calcd for
L) was added to the organic phase, followed by solid C6H6NO3S2Cl: C, 30.06; H, 2.52; N, 5.84. Found: C, 30.14;
NaHSO3 with agitation, until peroxide test paper gave a H, 2.56; N, 5.80.
negative test. The phases were separated and the organic (S)-3,4-Dihydro-6-chloro-4-hydroxy-2-(3-methoxypro-
solution was washed with saturated NaHCO3 to pH 8, then pyl)-4H-thieno[3,2-e]-1,2-thiazine 1,1-Dioxide (8b). To a
with brine, dried (Na2SO4), filtered, and concentrated. The stirred mixture of 18 (350 g, 1.46 mol), DMSO (1.75 L),
residual semisolid was triturated with MTBE to give a solid and K2CO3 (605 g, 4.38 mol) was added 3-bromopropyl
that was collected by filtration, washed with MTBE, and methyl ether (268 g, 1.75 mol) in eight equal portions spaced
dried in air to give 597 g (71%) of 16: mp 178-179 °C; 1H 1 h apart. After a further 1.5 h the mixture was poured into
NMR (DMSO-d6) δ 2.55 (s, 3H), 7.70 (s, 1H), 7.72 (s, 2H, saturated NaCl (18 L), extracted with MTBE (2 × 4 L), and
exchanges); 13C NMR (DMSO-d6) δ 30.0, 130.0, 131.5, the combined extracts were washed with 1 M NaOH (2 L),
138.8, 145.9, 193.2. Anal. Calcd for C6H6ClNO3S2: C, 30.06; 1:1 bleach/water (2 L), and brine (2 L), dried over Na2SO4
H, 2.52; N, 5.84; S, 26.75. Found: C, 30.19; H, 2.51; N, (500 g), filtered, and concentrated to provide 427 g (94%)
5.80; S, 26.70. of 8b as a light-yellow syrup. A sample was purified by
3-Bromoacetyl-5-chloro-2-thiophenesulfonamide (17). chromatography on silica (5% acetone-CH2Cl2): [R]25D
To a stirred suspension of 16 (1.09 kg, 4.55 mol) and EtOAc +11.4° (c ) 1, MeOH); 1H NMR (CDCl3) δ 1.83-2.04 (m,
(22 L) at 1 °C was added PBP (mfr. assay 98%, 1.30 kg, 2H), 3.25 (s, 3H), 3.28-3.81 (m, 6H), 4.08 (dd, 1H, J )
4.00 mol) in one portion. H2SO4 (concentrated, 544 mL) was 15, 4), 4.64 (br s, 1H), 6.96 (s, 1H); 13C NMR (CDCl3) δ
added over 10 min, causing the temperature to rise to 5 °C. 28.9, 46.9, 53.5, 58.5, 61.3, 69.8, 126.0, 132.4, 135.9, 143.4.
The mixture was stirred for 1.5 h, water (5 L) was added, Anal. Calcd for C10H14NO4S2Cl: C, 38.52; H, 4.53; N, 4.49.
the mixture was stirred for 5 min, and the phases were Found: C, 38.65; H, 4.54; N, 4.47.
separated. The organic solution was washed with brine to (S)-3,4-Dihydro-4-hydroxy-2-(3-methoxypropyl)-4H-
pH 3 (4 × 5 L required), dried over Na2SO4 (1 kg), filtered, thieno[3,2-e]-1,2-thiazine-6-sulfonamide 1,1-Dioxide (11b).
and concentrated. The residue was triturated with CH2Cl2 Under N2, a solution of 8b (1.06 kg, 3.42 mol) in anhydrous
(2 L) and chilled for 15 min. The solid was collected by THF (27 L) was cooled to -70 °C using a dry ice/2-propanol
filtration, washed with cold CH2Cl2 (2 L), and dried in air bath. n-BuLi (7.7 mol, 2.3 equiv, 3.08 L of a 2.5 M hexane
at room temperature to give 1.04 kg (72%) of 17. A sample solution) was added dropwise over 2.5 h while the temper-
was recrystallized from CH2Cl2-EtOAc-hexane: mp 147- ature was kept below -66 °C. After 1 h, SO2 was introduced
148 °C; 1H NMR (acetone-d6) δ 4.76 (s, 2H), 7.11 (br s, into the mixture until an aliquot quenched into water showed
2H, exchanges), 7.76 (s, 1H). Anal. Calcd for C6H5NO3S2- pH 4. The mixture was allowed to warm to room temperature
ClBr: C, 22.62; H, 1.58; N, 4.40; S, 20.13. Found: C, 22.66; overnight and then concentrated. The residue was dissolved
H, 1.60; N, 4.35; S, 20.12. in water (5 L), and this solution was added in one portion to
(S)-3,4-Dihydro-6-chloro-4-hydroxy-4H-thieno[3,2-e]- a 0 °C solution of NaOAc‚3H2O (2.80 kg, 20.5 mol) and
1,2-thiazine-1,1-dioxide (18). Under N2, a stirred suspension hydroxylamine-O-sulfonic acid (1.55 kg, 13.7 mol) in water
of 17 (855 g, 2.68 mol) and MTBE (12.5 L) was cooled to (6 L), causing the temperature to rise to 25 °C. After being
-40 °C using a dry ice/2-propanol bath. (+)-Ipc2BCl (4.5 stirred for 15 h at room temperature the solution was
L of a 1.2 M solution in MTBE, 5.4 mol, 2 equiv) was added extracted with EtOAc (3 × 4 L). The combined extracts were
via a cannula over 30 min, causing the temperature to rise washed with saturated NaHCO3 until basic, then with brine,
to -32 °C. After 3.5 h at -25 to -20 °C, reduction was dried (Na2SO4), filtered, and concentrated. CH2Cl2 (6 L) was
complete. The bromohydrin solution was warmed to 0 °C added to the residual oil along with with 5 g of seed crystals,
and 1 M aqueous NaOH (11 L) was added over 10 min, and the mixture was chilled and agitated to induce crystal-
causing the temperature to rise to 22 °C. The biphasic lization. The solid was collected by filtration, washed with
mixture was stirred vigorously at room temperature for 2 h. CH2Cl2, and dried in air at room temperature to give 748 g
The phases were separated and the aqueous solution was (61%) of 11b of 98.5% purity by HPLC. Another run starting
extracted with MTBE (3 L), acidified to pH 1 using 12 M with 700 g of 8b yielded 552 g (69%) of 11b of comparable
HCl, and extracted with EtOAc (2 × 4 L). The combined purity. A sample was recrystallized from MeOH-H2O: mp
EtOAc extracts were washed with brine (3 L), dried (Na2- 111-113 °C; [R]23405 +21.9° (c ) 0.45, MeOH); 1H NMR
SO4, 1 kg), filtered, and concentrated to about 1 L. Toluene (DMSO-d6) δ 1.83 (pent, 2H, J ) 7), 3.21 (s, 3H), 3.3-3.4
(2 L) was added. Upon further concentration, removing (m, 4H), 3.73 (dd, 1H, J ) 15, 5.6), 3.92 (dd, 1H, J ) 15,
EtOAc, the product crystallized and was collected by 4.5), 4.82 (br q, 1H, J ) 5), 6.15 (d, 1H, J ) 6, exchanges),
filtration, washed with toluene (2 L) and CH2Cl2 (2 L), and 7.60 (s, 1H), 8.05 (s, 2H, exchanges); 13C NMR (DMSO-
dried in air at room temperature to give 498 g (77%) of 18. d6) δ 28.5, 46.6, 52.6, 57.9, 60.2, 68.8, 129.7, 135.4, 145.9,
HPLC analysis of this material showed >98% chemical 148.8. Anal. Calcd for C10H16N2O6S3: C, 33.69; H, 4.53;
purity and >98% ee. A sample was recrystallized from N, 7.86. Found: C, 33.61; H, 4.55; N, 7.77.
CHCl3: mp 126-127 °C; [R]25D -5.9° (c ) 1, MeOH); 1H (R)-3,4-Dihydro-4-(ethylamino)-2-(3-methoxypropyl)-
NMR (DMSO-d6) δ 3.35-3.50 (m, 1H), 3.55-3.68 (m, 1H), 4H-thieno[3,2-e]-1,2-thiazine-6-sulfonamide 1,1-Dioxide
4.58 (q, 1H, J ) 7), 5.90 (d, 1H, J ) 7, exchanges), 7.20 (s, (13b, AL-4862). A solution of 11b (438 g, 1.23 mol),
1H), 8.15 (br t, 1H, J ) 6, exchanges); 13C NMR (DMSO- acetonitrile (4.0 L), and trimethyl orthoacetate (630 mL, 2.83
Vol. 3, No. 2, 1999 / Organic Process Research & Development • 119
mol) was heated at reflux (85 °C) for 15 h. After cooling to an additional 40 g of 13b, for a total yield of 286 g (61%)
30 °C (4 h), the solution was concentrated and the imidate of 13b. HPLC analysis of this material showed 97.4% purity
12b (563 g) was dissolved in anhydrous THF (2.2 L) and and >99% ee. A sample was recrystallized from 2-pro-
cooled to -10 °C under N2. Et3N (377 mL, 2.70 mol) and panol: mp 125-127 °C; [R]25312.6 -26.1° (c ) 1, pH 3 citric
TsCl (469 g, 2.46 mol) were added. A precipitate formed acid buffer); 1H NMR (DMSO-d6) δ 1.04 (t, 3H, J ) 7),
and the temperature rose to -2 °C over 10 min. After 2 h, 1.83 (pent, 2H, J ) 7), 2.5 (br s, 1H, exchanges), 2.60 (q,
TLC indicated complete tosylation. 70% aqueous EtNH2 (3.0 2H, J ) 7), 3.1-3.2 (m, 1H), 3.22 (s, 3H), 3.38 (t, 2H, J )
L, 37 mol) was added dropwise (0.4 L over 1 h, exothermic 7), 3.3-3.5 (m, 1H), 3.80 (m, 2H), 4.12 (t, 1H, J ) 7), 7.68
reaction with excess TsCl; then 2.6 L over 0.25 h), keeping (s, 1H), 8.02 (br s, 2H, exchanges); 13C NMR (DMSO-d6) δ
the temperature below 11 °C. The mixture was stirred at 15.6, 28.6, 40.0, 45.6, 49.0, 50.3, 57.9, 68.8, 129.8, 135.6,
room temperature for 15 h and cooled to -5 °C, and 12 M 146.6, 148.4. Anal. Calcd for C12H21N3O5S3: C, 37.58; H,
HCl (3.0 L) was added dropwise over 1 h causing a 5.48; N, 10.96. Found: C, 37.66; H, 5.56; N, 10.98.
temperature rise to 50 °C. The cooled (25 °C) solution was
washed with MTBE (2 × 2 L) and the combined washes Supporting Information Available
were back-extracted with 1 M HCl (200 mL). The pH of Infrared and mass spectra; experimental procedures and
the aqueous solution was adjusted to 8 over 1 h using solid analytical data for 4, 9, 10, 11a and 13a; analytical data for
NaHCO3, causing a white solid to precipitate. After the 3, 5, 12a, 12b, 14 and 15. This material is available free of
solution was stirred overnight, the solid was collected by charge via the Internet at http://pubs.acs.org.
filtration and washed with water and dried to give 246 g of
13b. The filtrates were extracted with EtOAc (2 × 4 L), the
Received for review August 14, 1998.
extract was dried (MgSO4), filtered, and concentrated, and
the residue was recrystallized from 2-propanol to provide OP9802125

120 • Vol. 3, No. 2, 1999 / Organic Process Research & Development