Delineating an alternate convergent synthesis of brexpiprazole: a novel use of commercial 6,7-dihydrobenzo[b]thiophen-4(5H)-one as precursor to an efficacious Buchwald–Hartwig amination step

Article abstract of DOI:10.1007/s12039-018-1470-z

Abstract: Brexpiprazole – an anti-psychotic drug approved for the treatment of schizophrenia – has been synthesized via an extremely concise and convergent route, which involves in essence two key C–N bond formation (amination) steps that serve to link the piperazine core between the constituent benzo[b]thiophene and 7-butoxyquinolin-2(1H)-one fragments. The highlight of this synthesis is the first amination step, which was effected quite efficaciously by a novel palladium mediated Buchwald–Hartwig coupling between N-Boc-piperazine and the triflate ester of benzo[b]thiophen-4-ol (conveniently prepared in three steps from commercially available 6,7-dihydrobenzo[b]thiophen-4(5H)-one). Indeed, even without an extensive screening of catalysts, ligands and reaction conditions, this amination step could be performed quite efficiently with merely 1?mol% catalyst loading, which cleanly afforded in 87% overall yield 1-(benzo[b]thiophen-4-yl)piperazine—the starting material for the second C–N bond formation and the final step in the synthesis of Brexpiprazole. Graphical Abstract: Synopsis An alternate convergent synthesis of the anti-psychotic drug Brexpiprazole has been described. In particular, a novel palladium catalyzed Buchwald–Hartwig coupling of N-Boc-piperazine with benzo[b]thiophen-4-yl trifluoromethanesulfonate (prepared from commercially available 6,7-dihydrobenzo[b]thiophen-4(5H)-one) has been shown to furnish 1-(benzo[b]thiophen-4-yl)piperazine (a key intermediate in the synthesis of the API) in excellent yield even with 1?mol% catalyst loading. [Figure not available: see fulltext.].

Full text of DOI:10.1007/s12039-018-1470-z

J. Chem. Sci. (2018) 130:72
© Indian Academy of Sciences
https://doi.org/10.1007/s12039-018-1470-z
REGULAR ARTICLE
Special Section on Transition Metal Catalyzed Synthesis of Medicinally Relevant Molecules
Delineating an alternate convergent synthesis of brexpiprazole: a
novel use of commercial 6,7-dihydrobenzo[b]thiophen-4(5H)-one
as precursor to an efficacious Buchwald–Hartwig amination step
A SRAVANTH KUMAR, SAI GIRIDHAR SARMA KANDANUR, SAIKAT SEN and
SRINIVAS ORUGANTI∗
Centre for Process Research and Innovation, Dr. Reddy’s Institute of Life Sciences, University of Hyderabad
Campus, Gachibowli, Hyderabad, Telangana 500 046, India
E-mail: soruganti@drils.org
MS received 26 March 2018; revised 26 April 2018; accepted 26 April 2018
Abstract. Brexpiprazole – an anti-psychotic drug approved for the treatment of schizophrenia – has been
synthesized via an extremely concise and convergent route, which involves in essence two key C–N bond
formation (amination) steps that serve to link the piperazine core between the constituent benzo[b]thiophene
and 7-butoxyquinolin-2(1H)-one fragments. The highlight of this synthesis is the first amination step, which
was effected quite efficaciously by a novel palladium mediated Buchwald–Hartwig coupling between N-
Boc-piperazine and the triflate ester of benzo[b]thiophen-4-ol (conveniently prepared in three steps from
commercially available 6,7-dihydrobenzo[b]thiophen-4(5H)-one). Indeed, even without an extensive screening
of catalysts, ligands and reaction conditions, this amination step could be performed quite efficiently with merely
1 mol% catalyst loading, which cleanly afforded in 87% overall yield 1-(benzo[b]thiophen-4-yl)piperazine—the
starting material for the second C–N bond formation and the final step in the synthesis of Brexpiprazole.
Keywords. Schizophrenia; anti-psychotic drugs; brexpiprazole; Buchwald–Hartwig amination; heterocycles;
benzo[b]thiophen-4-ol.
1. Introduction
restlessness and nausea.1,3 In fact, the Thomson Reuters
2015 annual report named Rexulti® as one of the pre-
dicted blockbuster entrants and slated its annual sales to
reach $1.353 billion in 2019.4
1.1 Background
The USFDA approval of the atypical (or, ‘second
generation’) anti-psychotic drug brexpiprazole 1(Rex-
ulti®, Otsuka Pharmaceutical Company Ltd.) in 2015
strengthened a new way forward in the treatment of
schizophrenia, a severe mental disorder that as of 2016,
affects more than 21 million people worldwide.1,2 Most
notably, 1 is a partial agonist of the dopamine receptor
D2 and prior to its approval, the only D2 partial agonist
that had reached the market with the approved indi-
cation of schizophrenia was its top-selling predecessor
aripiprazole 2(Abilify®).1a Projected as a better toler-
ated variant, brexpiprazole 1 has been reported to have
a lesser tendency than 2 to bring about D2 partial ago-
nist induced adverse effects such as akathisia, insomnia,
From a structural perspective (and a viewpoint of
retrosynthetic analysis as well), brexpiprazole 1 and
aripiprazole 2 are broadly similar, and consist essen-
tially of an N-arylpiperazine moiety joined by a n-butyl
linker to a subunit derived from 7-hydroxyquinolinone
3. Unsurprisingly, the foregoing structural analysis also
defines the two key retrosynthetic disconnections which
are common to best known syntheses of 1 and 2
(Scheme 1). Thus, the most widely employed method of
aripiprazole synthesis involves N-alkylation of 1-(2,3-
dichlorophenyl)piperazine4with7-(4-halobutoxy)-3,4-
dihydroquinolinone 5; the latter in turn is obtained
via etherification of 7-hydroxy-3,4-dihydroquinolinone
6 (Scheme 1).5 Likewise, a majority of the synthetic
*
For correspondence
Electronic supplementary material: The online version of this article (https://doi.org/10.1007/s12039-018-1470-z) contains
supplementary material, which is available to authorized users.

72 Page 2 of 10
J. Chem. Sci. (2018) 130:72
Scheme 1. Commonalities in the retrosynthetic bond disconnection strategies most widely adopted for
brexpiprazole 1 and aripiprazole 2.
Scheme 2. Synthesis of the key intermediate 7 and its elaboration to 1 (Otsuka; WO 2006/112464 A1).
routes to 1 invariably involve reacting 1-(benzo[b] [b]-thiophene 9b and piperazine was utilized as means
thiophen-4-yl)piperazine 7 with 7-(4-halobutoxy)- to obtain the key intermediate 7 (Scheme 3). Prepa-
quinolinone 8 (prepared by etherification of 3).6 Given ration of 9b involved two key steps: (a) elaboration
the well demonstrated (and predicted) efficiency of the of 2-carboxy-4-chlorobenzo[b]thiophene 11 from either
N-alkylation step, advances in commercial manufactur- 2,6-dichlorobenzaldehyde 10a or 2-chloro-6-fluoroben-
ing of 1 have, needless to say, primarily targeted an zaldehyde 10b by employing either rhodanine or thio-
efficacious acquisition of the tricyclic framework of 7.7 glycolic acid as a one carbon synthon and an even-
For example, the first synthetic route to 1, as disclosed tual source of a thiol functionality and (b) proto-
by the innovator in 2006,8 employed a Pd2(dba)3/(R)- decarboxylation of 11 to 9b at temperatures in excess
◦
BINAPcatalyzedBuchwald–Hartwigcouplingbetween of 140 C.9,10 While application of specialized lig-
4-bromobenzo[b]-thiophene9aandexcesspiperazineto ands (such as RuPhos and tri-tert-butylphosphonium
obtain 7 as a yellow oil in 64% yield after purification tetraphenylborate) in the Pd(OAc)2 catalyzed N-
by silica gel column chromatography. The free base 7, arylation of piperazine with 9b minimized unwanted
thus obtained, was characterized after conversion into its reactions, it did not completely eliminate formation of
recrystallized hydrochloride salt; the latter was reacted such hard-to-separate diarylated impurities as 12 and
with 7-(4-chlorobutoxy)-1H-quinolin-2-one 8a in pres- 13.
ence of K2CO3 and NaI to furnish 1 (Scheme 2). As
A workaround solution to deal with the impurities
elaborated upon as a prelude to an improved strategy in encountered in Otsuka’s preparation of 7 was provided
a subsequent 2013 Otsuka patent,9 this original process in a recent communication by Xiangrui and co-workers
to1isinherentlyinefficientattheN-arylationstep;itsuf- (Scheme 4).11 Herein, the key C–N bond formation en
fers from the formation of a large number of by-products route to 7 was achieved at the outset of the synthesis
and inevitably requires careful chromatographic separa- by a controlled SN Ar reaction between 10b and 1-Boc-
tion to obtain 7 in high purity. In Otsuka’s second gener- piperazine to furnish 14. Thereafter, the aldehyde 14 was
ation approach to brexpiprazole, a Pd(OAc)2 mediated converted into benzo[b]thiophene 17 in a down-stream
Buchwald–Hartwig coupling between 4-chlorobenzo process that closely followed Otsuka’s synthesis of 9b

J. Chem. Sci. (2018) 130:72
Page 3 of 10 72
Scheme 3. An example of Otsuka’s synthesis of 9b and its use in Buchwald–Hartwig coupling with
piperazine to obtain 7 (Otsuka; WO 2013/015456 A1).
Scheme 4. Synthesis of 7·2HCl from 2-chloro-6-fluorobenzaldehyde 10b as reported by Xiangrui and co-workers.
from 10b. Finally, removal of the N-Boc protection in 17 1-Boc-piperazine and the aripiprazole precursor 7-(4-
with concentrated HCl afforded 7 as a dihydrochloride bromobutoxy)-3,4-dihydroquinolinone 5b as the only
salt. Despite providing an alternate amination strategy, building blocks.
Xiangrui’s synthesis of 7 is nonetheless a linear one,
and suffers from formation of impurities such as 18
and 19 [in the step 10b → 14] and 20 [in the step 16
→ 17, on account of the instability of the tert-butyl-
carboxylate moiety at the high temperatures (∼140 ◦C)
employed to effect decarboxylation in 16]. Against this
background and in keeping with our on-going interest
in investigating applications of transition metal catalysis
in the synthesis of contemporary active pharmaceutical
ingredients,12 we decided to re-design Otsuka’s conver-
gent synthesis of brexpiprazole 1, and delineate a highly
concise route to the API by employing commercially
available 6,7-dihydrobenzo[b]thiophen-4(5H)-one 21,
1.2 Retrosynthetic analysis
As shown in Scheme 5, our retrosynthetic analysis of 1
relied on two key C–N bond formation steps, namely: (a)
N-alkylation of 7 with 7-(4-bromobutoxy)quinolinone
8b, the latter being obtained by DDQ oxidation of
5b; and (b) Buchwald–Hartwig coupling of 1-Boc-
piperazine with the triflate ester 23 of benzo[b]thiophen-
4-ol 22, which in turn can be prepared via aromatization
of 21. While commercially available, the bicyclic ketone
21 can also be easily prepared from thiophene in
three steps, viz., Friedel-Crafts acylation with succinic

72 Page 4 of 10
J. Chem. Sci. (2018) 130:72
Scheme 5. Retrosynthetic analysis of 1 from 6,7-dihydrobenzo[b]thiophen-4(5H)-one 21,
1-Boc-piperazine and the aripiprazole precursor 7-(4-bromobutoxy)-3,4-dihydroquinolinone 5b.
anhydride, Wolff-Kishner reduction of the γ -ketoacid were calculated on the basis of product assay determined by
quantitative NMR (qNMR) with 1,3,5-trimethoxybenzene as
an internal standard.
intermediate and annulation via intramolecular Friedel-
Crafts acylation in the 4-(thiophen-2-yl)butanoic acid
obtained.13
2.2 Experimental procedures and spectral
characterization
2.2a Benzo[b]thiophen-4-ol (22): To a 100 mL two-
neck round-bottom flask fitted with a magnetic stir bar
and nitrogen inlet was charged 6,7-dihydrobenzo[b]thiophen-
4(5H)-one 21 (1.29−g, 8.48 mmol) in 20 mL dry THF and
2. Experimental
2.1 General information
Reactions were carried out in oven-dried glassware under added PhN+Me3Br3 (3.18 g, 8.48 mmol) portion wise at
◦
◦
a positive pressure of nitrogen unless otherwise mentioned. 0 C. The resulting solution was stirred at 0 C for 90 min,
Air-sensitive reagents and solutions were transferred via warmed slowly to rt, and further stirred at rt for 30 min. Upon
syringe and were introduced to the apparatus via rubber septa. consumption of starting material as indicated by TLC (as the
1H and 13C NMR spectra were recorded on a Varian 400 ratio of product increases in the reaction mixture, the color
MHz spectrometer. Deuterated solvents for NMR spectro- of the reaction mixture changed from pale orange to light
scopic analyses were used as received. Coupling constants yellow), the reaction mixture was quenched with water and
are reported in Hz. All chemical shifts are quoted in ppm, rel- extracted with ethyl acetate (3 × 70 mL). The organic layer
ative to TMS, using the residual solvent peak as a reference was dried over anhydrous Na2SO4 and concentrated under
standard. The following abbreviations are used to explain the reduced pressure to afford a musk colored solid (2.97 g).
multiplicities: s = singlet, bs = broad singlet, d = doublet, dd
The musk color solid obtained in the previous step was
= doublet of doublet t = triplet, q = quartet, m = multiplet, taken in DMF (20 mL) without any further purification into
br = broad. Mass spectra were recorded on an Agilent 6430 a 100 mL round bottom flask. Li2CO3 (2.16 g, 29.23 mmol)
triple quad LC/MS system. The purity of the compounds were and LiBr (3.30 g, 38.00 mmol) were charged consecutively
determined by HPLC with a Waters Alliance e2695 separa- and resulting reaction mixture was heated to 150 ◦C for
tion module, equipped with a Waters 2998 photodiode array 5 h. Upon consumption of starting material (as indicated by
detector and analysed by Empower2 software. Reactions were TLC), the reaction mixture was quenched with water and
monitored by thin-layer chromatography (TLC) performed on extracted with ethyl acetate (3 × 50 mL). The organic layer
Merck TLC silica gel 60 F254 aluminum plates. Visualization was washed with brine solution, water, dried over anhydrous
was accomplished with either UV light, or by immersion in an Na2SO4 and concentrated under reduced pressure to get the
ethanolic solution of phosphomolybdic acid (PMA), ninhy- crude compound as a brownish solid. The crude compound
drin, or KMnO4, followed by heating with a heat gun for 15 s. was purified by column chromatography over 60–120 mesh
Dry toluene and THF were either distilled over sodium ben- (5% ethyl acetate/ hexane) to afford the title compound 22
1
zophenone ketyl, or procured commercially. Yields of 7·2HCl as a white solid. Yield: 88% (1.13 g); H NMR (400 MHz,

J. Chem. Sci. (2018) 130:72
Page 5 of 10 72
CDCl3): δ 7.50–7.43 (m, 2H), 7.39–7.35 (m, 1H), 7.20 (t, The residue obtained was passed through a short pad of silica
J = 7.9 Hz, 1H), 6.72 (dd, J = 7.7, 0.6 Hz, 1H), 5.10 gel (60–120 mesh) using 4% ethyl acetate/hexane to remove
(s, 1H); 13C NMR (100 MHz, CDCl3): δ 150.71, 141.86, metal salts, inorganic matter and excess 1-Boc piperazine.
129.23, 125.28, 125.07, 119.69, 115.24, 108.80; ESI (MS):
The partially purified product 17, (NOTE: Analytical data
151 [M+H]+; HPLC purity: 99.18% [RT: 11.930 min; UV obtained for a sample of 17 purified by column chromatog-
detection at 220 nm; Column: X-Terra RP 18, 150 × 4.6 mm, raphy: 1H NMR (400 MHz, CDCl3): δ 7.57 (d, J = 7.9 Hz,
5 µm particle size; Mobile phase: A) 0.1% TFA in water B) 1H), 7.41 (brs, 2H), 7.28 (t, J = 7.7 Hz, 1H), 6.88 (d,
Acetonitrile; T/%B: 0/20, 3/20, 12/95, 23/95, 25/20, 30/20;
J = 7.2 Hz, 1H), 3.66 (brs, 4H), 3.10 (brs, 4H), 1.50 (s,
Flow rate: 1.0 mL/min; Diluent: Acetonitrile:Water (80:20)]. 9H); 13C NMR (100 MHz, CDCl3): δ 154.83, 148.22, 141.17,
1H NMR spectral data of 22 was found to be consistent with 134.17, 125.29, 124.96, 121.61, 117.40, 112.40, 79.84, 52.11,
the values reported in Ref.14
.
28.46; ESI (MS): 319.1 [M+H]+, H and 13C NMR spec-
tral data of 17 were found to be consistent with the values
reported in Ref.11), thus obtained was dissolved in 25 mL
THF and taken in a 100 mL round bottom flask. 25 mL of
1
2.2b Benzo[b]thiophen-4-yl trifluoromethanesulfon-
ate (23): Benzo[b]thiophen-4-ol (22) (500 mg, 3.33 mmol)
was charged in a 100 mL two-neck round-bottom flask
fitted with a magnetic stir bar, nitrogen inlet and dry
CH2Cl2 (10 ml). To this reaction mixture was added triethy-
lamine (1.16 mL, 8.32 mmol) following which the contents
were cooled to 0 ◦C and trifluoromethanesulfonic anhydride
(0.61 mL, 3.66 mmol) was added carefully drop wise. The
reaction was allowed to stir at 0 ◦C for 1 h. Upon consump-
tion of starting material (as indicated by TLC), the reaction
mixture was quenched with saturated NaHCO3 solution. The
organic layer was washed thoroughly with brine solution,
water, dried over anhydrous Na2SO4 and concentrated under
reduced pressure to afford the crude product. The crude com-
pound was purified by column chromatography over 60–120
mesh (2% ethyl acetate/ hexane) to afford the 23 as color-
less oil. Yield: 90% (846 mg); 1H NMR (400 MHz, CDCl3):
δ 7.88 (d, J = 7.9 Hz, 1H), 7.58 (d, J = 5.5 Hz, 1H),
7.47 (d, J = 5.5 Hz, 1H), 7.39 (t, J = 8.0 Hz, 1H),
7.32 (d, J = 7.9 Hz, 1H); 13C NMR (100 MHz, CDCl3):
δ 144.04, 142.53, 132.63, 128.91, 124.74, 122.69, 119.25,
118.73 (q, JC F = 319 Hz, CF3), 116.44; ESI (MS): 281
[M−H]−; HPLC purity: 98.94% [RT: 14.212 min, UV detec-
tion at 225 nm; Column: X-Terra RP 18, 150×4.6 mm, 5 µm
particle size; Mobile phase: A) 0.1% TFA in water B) Ace-
tonitrile; T/%B: 0/20, 3/20, 12/95, 23/95, 25/20, 30/20; Flow
◦
6N HCl was charged at 0 C and the reaction mixture was
stirred overnight at rt. Upon completion of the starting mate-
rial(asindicatedbyTLC), THFwasevaporatedunderreduced
pressure. The reaction mixture was diluted with 30 mL 6N
HCl, and washed with dichloromethane (3 × 20 mL). The
washed aqueous layer was evaporated under reduced pressure
at 55 ◦C to furnish the required product 1-(benzo[b]thiophen-
4-yl) piperazine dihydrochloride (7·2HCl) as an off-white
1
solid. 61% over two steps by q-NMR (1.37 g); H NMR
(400 MHz, DMSO-d6): δ 9.60 (brs, 2H), 8.68 (brs, 1H),
7.71 (brs, 1H), 7.65 (d, J = 7.2 Hz, 1H), 7.49 (brs, 1H), 7.27
(brs, 1H), 6.92 (d, J = 5.9 Hz, 1H), 3.27 (s, 8H); 13C NMR
(100 MHz, DMSO-d6): δ 147.41, 141.04, 133.84, 126.96,
125.52, 122.36, 118.15, 112.98, 49.00, 43.53; ESI (MS): 219
[M+H]+; HPLC purity: 99.20% [RT: 8.623 min; UV detec-
tion at 220 nm; Column: X Bridge C-18, 150 × 4.6 mm,
5 µm particle size; Mobile phase: A) 0.1% TFA in water, B)
Acetonitrile; T/%B: 0/10, 3/10, 12/95, 23/95, 25/10, 30/10;
Flow rate: 1.0 mL/min; Diluent: Acetonitrile:Water (80:20)].
1H and 13C NMR spectral data of 7·2HCl were found to be
consistent with the values reported in Ref.11
.
2.2d 1-(Benzo[b]thiophen-4-yl)piperazinedihydroch-
loride (7·2HCl) [Table 1, Entry 13]: To a 50 mL two-
neck round-bottom flask fitted with a magnetic stir bar and
argon inlet, benzo[b]thiophen-4-yl trifluoromethanesulfonate
23 (1.0 g, 3.54 mmol) was charged in dry toluene (10 mL).
1-Boc-piperazine (0.98 g, 5.31 mmol), Pd2(dba)3·CHCl3
(0.036 g, 0.035 mmol; 1 mol%), XPhos (0.05 g, 0.106 mmol;
1
rate: 1.0 mL/min; Diluent: Acetonitrile:Water (80:20)]. H
and 13C NMR spectral data of 23 were found to be consistent
with the values reported in Ref.15
.
2.2c 1-(Benzo[b]thiophen-4-yl)piperazinedihydroch-
loride (7·2HCl) [Table 1, Entry 10]: Benzo[b]thiophen- 3 mol%) and Cs2CO3(3.45 g, 10.62 mmol) were added con-
4-yl trifluoromethanesulfonate 23 (2.0 g, 7.08 mmol) was secutively to the reaction mixture under argon atmosphere.
charged in dry toluene (20 mL) into a 100 mL two- The reaction mixture was degassed with vacuum pump and
neck round-bottom flask fitted with a magnetic stir bar then heated to 100 ◦C for 8 h under argon atmosphere. Upon
and argon inlet. To this reaction mixture, 1-Boc-piperazine completion of the starting material (as indicated by mass and
(1.97 g, 10.62 mmol), Pd2(dba)3·CHCl3 (0.73 g, 0.708 mmol; TLC analysis) the reaction mixture was filtered over a short
10 mol%), XPhos (0.50 g, 1.062 mmol; 15 mol%) and pad of Celite and the filtrate was evaporated under reduced
Cs2CO3(6.92 g, 21.24mmol) wereaddedconsecutivelyunder pressure. The residue obtained was passed through a short pad
argon atmosphere. The reddish brown solution obtained was of silica gel (60–120 mesh) using 4% ethyl acetate/hexane as
degasse◦d with vacuum pump. The reaction was then heated solvent system to remove metal salts, inorganic matter and
to 100 C for 8 h under argon atmosphere. Upon comple- excess 1-Boc-piperazine.
tion of the starting material (as indicated by mass and TLC
The partially purified product 17 thus obtained was dis-
analysis) the reaction mixture was filtered over a short pad of solved in 15 mL of THF and taken in a 100 mL round
◦
Celite and the filtrate was evaporated under reduced pressure. bottom flask. 15 mL of 6N HCl was charged at 0 C and

72 Page 6 of 10
J. Chem. Sci. (2018) 130:72
Table 1. Buchwald–Hartwig coupling of the triflate 23 and 1-Boc-piperazine.
Boc
H
N
N
•
2HCl
S
Tf
O
N
N
B
oc
HN
N
6N HCl, 1:1 THF-H2O
°
0 C RT, overnight
S
S
Pd-catalyst, ligand, base,
PhMe, 100 °C, 8-12 h
23
7•2HCl
17
Entry
Catalyst (loading)
Ligand (loading)
Base Yield (%) of 7·2HCl (over two steps from 23)a
1
2
3
4
5
6
7
8
9
Pd(OAc)2(20 mol%)
Pd(OAc)2(10 mol%)
Pd(OAc)2(10 mol%)
CyJohnPhos (20 mol%)
NaOt Bu
–
Trace
–
tri(o-tolyl)phosphine (15 mol%) NaOt Bu
tri(o-tolyl)phosphine (15 mol%) Cs2CO3
Pd(OAc)2(10 mol%)
XPhos (15 mol%)
XPhos (15 mol%)
XPhos (15 mol%)
XPhos (15 mol%)
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
70
10
64
78
68
25
61
66
96
87
Pd(PPh3)4(10 mol%)
Pd(PPh3)2Cl2(10 mol%)
Pd2(dba)3·CHCl3(10 mol%)
Pd2(dba)3·CHCl3 (10 mol%) tri(o-tolyl)phosphine (15 mol%) Cs2CO3
Pd2(dba)3·CHCl3 (10 mol%)
CyJohnPhos (15 mol%)
XPhos (15 mol%)
XPhos (6 mol%)
XPhos (3 mol%)
XPhos (3 mol%)
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
10 Pd2(dba)3·CHCl3 (10 mol%)
11
12
13
Pd2(dba)3·CHCl3 (3 mol%)
Pd2(dba)3·CHCl3 (1 mol%)
Pd2(dba)3·CHCl3 (1 mol%)
Reaction Conditions: All reactions were performed at 100 ◦C in dry toluene (4 mL for Entries 1–9, 11 and 12) with 1.0 mole
equiv. of triflate 23, 1.5 mole equiv. of 1-Boc-piperazine and 3.0 mole equiv. of the base. Entries 1–9, 11 and 12 were performed
with ∼200 mg of 23. Entries 10 and 13 were performed with 2 g and 1 g of 23, respectively (cf. Sections 2.2c and 2.2d for
details).
aFor Entries 1–9, yields were calculated based simply on the weight of the product isolated, assuming the sample to be entirely
7·2HCl. For Entries 10–13, yields of 7·2HCl were calculated on basis of the assay of the isolated product as determined by
quantitative NMR (qNMR) with 1,3,5-trimethoxybenzene as an internal standard (cf. Sections 2.2c and 2.2d for details).
the contents were stirred overnight at rt. Upon completion of the contents were stirred at rt until the completion of the reac-
the starting material (indicated by TLC), THF was evaporated tion. Upon the consumption of starting material (as indicated
underreducedpressure. Thereactionmixturewasdilutedwith by TLC), THF was evaporated under reduced pressure. The
15mL6NHClandwashedwithdichloromethane(3×20 mL). residual product obtained was washed with water (200 mL)
The washed aqueous layer was evaporated under reduced and extracted with ethyl acetate (3 × 100 mL). The organic
◦
pressure at 55 C to furnish an off-white solid which was layer was washed with brine, dried over anhydrous Na2SO4,
subsequently washed with diethyl ether to afford required and concentrated under reduced pressure to furnish the crude
product 1-(benzo[b]thiophen-4-yl) piperazine dihydrochlo- product. The crude product was purified by column chro-
ride (7·2HCl) as an off white solid. Yield: 87% over two steps matography using Ethyl acetate and n-Hexane as solvent
by q-NMR (919 mg,); HPLC purity: 98.73% [RT: 8.645 min; system to afford the title compound 8b. Yield: 98.99% (9.8 g);
UV detection at 220 nm; matches with the RT of Table 1, 1H NMR (400 MHz, DMSO-d6) δ 11.57 (s, 1H), 7.77 (d,
Entry 10 obtained under the given HPLC conditions; Col-
J = 9.4 Hz, 1H), 7.53 (d, J = 9.4 Hz, 1H), 6.83–6.69 (m,
umn: X-Terra RP 18, 150 × 4.6 mm, 5 µm particle size; 2H), 6.27 (d, J = 9.6 Hz, 1H), 4.02 (t, J = 6.1 Hz, 2H), 3.59
Mobile phase: A) 0.1% TFA in water B) Acetonitrile; T/%B: (t, J = 6.6 Hz, 2H), 2.03–1.89 (m, 2H), 1.89–1.74 (m, 2H);
0/20, 3/20, 12/95, 23/95, 25/20, 30/20; Flow rate: 1.0 mL/min; 13C NMR (100 MHz, DMSO-d6): δ 162.67, 160.72, 141.05,
Diluent: Acetonitrile:Water (80:20)].
140.44, 129.69, 118.97, 113.77, 111.19, 99.05, 67.28, 35.26,
29.46, 27.74; ESI (MS): 296 [M+H]+; HPLC purity: 99.26%
[RT: 11.545 min; UV detection at 210 nm; Column: X Bridge
C-18, 150 × 4.6 mm, 5 µm particle size; Mobile phase: A)
0.1% TFA in water, B) Acetonitrile; T/%B: 0/10, 3/10, 15/95,
23/95, 25/10, 30/10; Flow rate: 1.0 mL/min; Diluent: Acetoni-
trile:Water (80:20)]. 1H NMR spectral data of 8b was found
2.2e 7-(4-Bromobutoxy)quinolin-2-(1H)-one (8b):
To a 500 mL two-neck round-bottom flask fitted with a mag-
netic stir bar and argon inlet was charged 7-[4-bromobutoxy]-
3,4-dihydro-2(1H)-quinolinone 5b (10 g, 0.033 mol) in THF
(200 mL). To the aforementioned mixture 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (30.4 g, 0.13 mol) was added and
to be consistent with the values reported in Ref.16
.

J. Chem. Sci. (2018) 130:72
Page 7 of 10 72
Scheme 6. Synthesis of the Buchwald–Hartwig arylation partner 23.
2.2f Brexpiprazole (1):
1-(Benzo[b]thiophen-4-yl) obtained was invariably contaminated with a significant
piperazinedihydrochloride (7·2HCl) (282 mg, 0.97 mmol) in
dry DMF was taken in a 100 mL two-neck round-bottom flask
fitted with a magnetic stir bar and argon inlet. To this reaction
mixture, K2CO3 (488mg, 3.53mmol)wasaddedandthereac-
tion was allowed to stir at rt for 30 min. One mole equivalent
of KI (167 mg, 1.01 mmol) and 7-(4-bromobutoxy)quinolin-
2(1H)-one 8b (300 mg, 1.01 mmol) were added consecutively.
The contents were heated at 100 ◦C for 4 h. Upon complete
consumption of starting material (a◦s indicated by TLC), the
reaction mixture was cooled to 0 C, then 5 mL of water
was added and stirred for 30 min. The resultant precipitate
(off-white solid) observed was filtered using sintered funnel,
amount of 2,5-dibromo-6,7-dihydrobenzo[b]thiophen-
4(5H)-one 25.18 Indeed, even on a 500 mg scale,
bromination of 21 with Br2 would at times result in the
formation of nearly an equal amount of 25 as the desired
product 24.
◦
When heated at 150 C in DMF with a mixture of
Li2CO3 and LiBr, the crude α-bromoketone 24 under-
went a tandem dehydrobromination-aromatization and
cleanly furnished the phenol 22 in nearly 88% yield.14
Esterification of 22 with triflic anhydride under basic
conditions15 afforded the triflate 23 that was then sub-
redissolved in a mixture of MeOH and CH2Cl2 and subjected jected to a palladium catalyzed Buchwald–Hartwig
to column chromatography (100–200 mesh) using 3% MeOH
in CH2Cl2 as solvent system to furnish the pure product Brex-
piprazole 1 as a white solid. Yield: 85% (357 mg); 1H NMR
(400 MHz, DMSO-d6): δ 11.60 (s, 1H), 7.79 (d, J = 9.4 Hz,
1H), 7.67 (d, J = 5.5 Hz, 1H), 7.59 (d, J = 8.1 Hz, 1H),
7.54 (d, J = 9.0 Hz, 1H), 7.37 (d, J = 5.7 Hz, 1H), 7.25 (t,
J = 7.9 Hz, 1H), 6.85 (d, J = 7.7 Hz, 1H), 6.82–6.73 (m,
2H), 6.28 (d, J = 9.4 Hz, 1H), 4.03 (t, J = 6.4 Hz, 2H), 3.04
(brs, 4H), 2.62 (brs, 4H), 2.44 (t, J = 6.9 Hz, 2H), 1.84–1.70
(m, 2H), 1.70–1.53 (m, 2H); 13C NMR (100 MHz, DMSO-
d6):δ 162.73, 160.89, 148.62, 141.09, 140.85, 140.49, 133.80,
129.69, 126.28, 125.53, 122.33, 118.89, 117.11, 113.72,
amination with 1-Boc-piperazine. Since the use of 23 as
an N-arylation partner was found to be unprecedented in
literature, a series of catalyst, ligand and base screening
experiments were performed to arrive at the best pos-
sible reaction conditions for effecting the coupling of
1-Boc-piperazine and 23 (Table 1).
Thus, based on reaction conditions reported for
Buchwald–Hartwig amination of various aryl triflates
with N-protected piperazines, our first attempt at ary-
lating 1-Boc-piperazine with 23 was carried out by
employing Pd(OAc)2 as the catalyst, CyJohnPhos as the
112.47, 111.32, 99.03, 68.04, 57.74, 53.36, 52.05, 26.98, ligand and NaOt Bu as the base. However, even upon
23.06; ESI (MS): 434 [M+H]+; HPLC purity: 97.97% [RT:
increasing the catalyst and ligand loading to 20 mol%,
10.504 min; UV detection at 215 nm; Column: X-Terra RP
no evidence for formation of the desired product could
18, 150 × 4.6 mm, 5 µm particle size; Mobile phase: A)
be discerned (Table 1, Entry 1); rather, the only observ-
able reaction in most of these attempts was a gradual
hydrolysis of 23 over time. By replacing CyJohnPhos
0.1% TFA in water B) Acetonitrile; T/%B: 0/20, 3/20, 12/95,
23/95, 25/20, 30/20; Flow rate: 1.0 mL/min; Diluent: Acetoni-
trile:Water (80:20)]. 1H NMR spectral data of 1 was found to
with tri(o-tolyl)phosphine, we were able to detect traces
be consistent with the values reported in Ref.9.
ofthedesiredproduct17inthecrudereactionmixtureby
mass analysis (Table 1, Entry 2). However, hydrolysis of
23 still prevailed, prompting us to replace with Nat OBu
with Cs2CO3as a base for subsequent screening exper-
3. Results and Discussion
iments. Indeed, while a milieu of Pd(OAc)2, (o-tol)3P
and Cs2CO3failed to promote the desired Buchwald–
Hartwig coupling, no evidence for triflate hydrolysis
was observed even after allowing the reaction mixture to
be heated at 100 ◦C for 12 h (Table 1, Entry 3). Much to
our delight, we were eventually successful in realizing
the coveted N-arylation by employing XPhos as the lig-
and, and the key intermediate 7·2HCl could be isolated
in 70% yield (over two steps) after N-Boc deprotection
in the Buchwald product 17 (Table 1, Entry 4).
As already alluded to, the Buchwald–Hartwig arylation
partner 23 was conveniently prepared from 21 in three
steps via the intermediacy of the phenol 22 (Scheme 6).
Thus, following a known procedure,14 the ketone 21
was treated with trimethylphenylammonium tribromide
in order to mono-brominate selectively the C-5 posi-
tion and obtain 24, which was carried forward after
work-up to the next step without any further purifica-
tion. While the α-bromo derivative 24 could also be
obtained with molecular bromine in CCl4,17 the product

72 Page 8 of 10
J. Chem. Sci. (2018) 130:72
Scheme 7. Preparation of 8b and the end-game C–N bond formation to afford 1.
With XPhos as the ligand, other well-known palla- of Pd2(dba)3·CHCl3 and XPhos. After a few trials,
dium catalysts, namely Pd(PPh3)4, Pd(PPh3)2Cl2 and the desired conversion of 23 into 17 could indeed be
Pd2(dba)3·CHCl3,were screened (Table 1, Entries 5–7) achieved with not only 3 mol% Pd2(dba)3·CHCl3 and
to evaluate their efficiency vis-à-vis Pd(OAc)2 (Table 1, 6 mol% XPhos (Table 1, Entry 11), but also 1 mol%
Entry 4) in promoting the Buchwald–Hartwig coupling Pd2(dba)3·CHCl3 and 3 mol% XPhos (Table 1, Entry
between 1-Boc-piperazine and 23. Thus, a combina- 12). It was also possible to employ the latter cata-
tion of Pd2(dba)3·CHCl3 and X-Phos (Table 1, Entry lyst/ligand loading to effect a complete consumption
7) was found to perform best and furnished 7·2HCl in of 23 even on a 1 g scale, and obtain 7·2HCl in 87%
78% isolated yield. Indeed, when employed with other yield and ∼99% purity (Table 1, Entry 13). Thus, hav-
phosphine ligands [viz., (o-tol)3P (Table 1, Entry 8) ing accessed one of the two building blocks necessary
and CyJohnPhos (Table 1, Entry 9)], Pd2(dba)3·CHCl3 for the construction of 1, the stage was set for us to
fared worse in comparison so that Pd2(dba)3·CHCl3 and usher in the second fragment 8b and converge on the
X-Phos were eventually chosen as the catalyst/ligand well-known end-game C–N bond formation step.
combination in our first attempt to effect the desired
As already alluded to in Section 1.2, it is known that
Buchwald–Hartwig amination of 23 on a gram scale. 7-(4-bromobutoxy)quinolinone 8b can be prepared by
Gratifyingly, coupling of the triflate 23 with 1-Boc- DDQ oxidation of the commercially available Aripipra-
piperazine in presence of 10 mol% Pd2(dba)3·CHCl3 zole precursor 5b.19 With minor modifications to the
and 15 mol% XPhos was uneventful even on a 2 g reported procedure, it was possible for us to carry out
scale (Table 1, Entry 10), and the reaction went to this transformation on a 10 g scale, and conveniently
near completion within 8 h at 100 ◦C to cleanly afford obtain 8b in near quantitative yield and with HPLC
17. In fact, absence of any major side reactions in purity >99%. Thus, with both 7·2HCl and 8b in hand,
this coupling step rendered a complete purification of thefinalN-alkylationstepwaseffectedinthepresenceof
the N-Boc derivative 17 prior to its conversion into K2CO3 and KI to furnish brexpiprazole 1 (purity ∼98%)
7·2HCl entirely unnecessary. The crude reaction mix- in 85% yield (Scheme 7).
ture, obtained from the Buchwald–Hartwig amination
of 23, needed to be passed through a short silica gel col-
umn merely to remove inorganic matter and unreacted
1-Boc-piperazine. The partially purified product 17 was
thereafter subjected to N-Boc deprotection with dilute
HCl, and the desired intermediate 7·2HCl isolated with
>99% purity simply by extracting out the non-basic
impurities from the acidic milieu with dichloromethane.
Having established the proof-of-concept for a novel
synthesis of 7·2HCl from 23, we were goaded to inves-
tigate if the Buchwald–Hartwig coupling could be made
more economical by lowering the required loading
4. Conclusions
To summarize, we have demonstrated a convergent and
highly concise synthesis of brexpiprazole 1, wherein
the entire framework of the API has been assembled in
a few simple steps from three readily available build-
ing blocks, viz., 6,7-dihydrobenzo[b]thiophen-4(5H)-
one 21, 1-Boc-piperazine and the aripiprazole precursor
7-(4-bromobutoxy)-3,4-dihydroquinolinone 5b. Of the
two efficacious C–N bond formations that constitute

J. Chem. Sci. (2018) 130:72
Page 9 of 10 72
the crux of our synthetic strategy, the one eventu-
ated by Buchwald–Hartwig coupling between 1-Boc-
piperazine and 23 is particularly noteworthy. Indeed,
while the chemistry of benzo[b]thiophen-4-ol 22 is well
documented,20 any report on the use of its triflate ester
23 to arylate amines is hitherto unknown. Our present
endeavor should therefore stimulate further investiga-
tions into potential application of 23 as a source of
benzo[b]thiophen-4-yl fragment in Buchwald–Hartwig
aminations.
8. Yamashita H, Matsubara J, Oshima K, Kuroda H, Ito
N, Miyamura S, Shimizu S, Tanaka T, Oshiro Y, Shi-
mada J, Maeda K, Tadori Y, Amada N, Akazawa H,
Yamashita J, Mori A, Uwahodo Y, Masumoto T, Kikuchi
T, Hashimoto K and Takahashi H 2006 Piperazine-
substituted benzothiophenes for treatment of mental
disorders PCT Application WO 2006/112464 A1
9. Shinhama K, Utsumi N, Sota M, Fujieda S and Oga-
sawara S 2013 Method for producing benzo[b]thiophene
compound PCT Application WO 2013/015456 A1
10. Miyake M, Shimizu M, Tsuji K and Ikeda K 2016 Safe
and Efficient Decarboxylation Process: A Practical Syn-
thetic Route to 4-Chlorobenzo[b]thiophene Org. Process
Res. Dev. 20 86
Supplementary Information (SI)
11. Wu C, Chen W, Jiang D, Jiang X and Shen
1
Scanned copies of H and 13C NMR spectra of 1, 7·2HCl,
J
2015 An Improved Synthesis of 4-(1-
8b, 17, 22 and 23 are available at www.ias.ac.in/chemsci.
Piperazinyl)benzo[b]thiophene Dihydrochloride Org.
Process Res. Dev. 19 555
Acknowledgements
12. (a) Bollikonda S, Mohanarangam S, Jinna R R,
Kandirelli V K K, Makthala L, Sen S, Chaplin D A,
Lloyd R C, Mahoney T, Dahanukar V H, Oruganti S and
Fox M E 2015 An Enantioselective Formal Synthesis of
Montelukast Sodium J. Org. Chem. 80 3891; (b) Nevu-
luri N R, Rapolu R K, Iqbal J, Kandagatla B, Sen S,
Dahanukar V H and Oruganti S 2017 A morpholine-
free process amenable convergent synthesis of apixa-
ban: a potent factor Xa inhibitor Monatsh. Chem. 148
1477
13. (a) Harel D, Schepmann D and Wünsch B 2013 New
combination of pharmacophoric elements of potent σ1
ligands: Design, synthesis and σ receptor affinity of
aminoethyl substituted tetrahydrobenzothiophenes Eur.
J. Med. Chem. 69 490; (b) Acheson R M and Cooper M
W 1980 The synthesis of thiophenium and of oxazolium
salts from diazoketones J. Chem. Soc., Perkin Trans. 1
1185
14. Kawakita Y, Banno H, Ohashi T, Tamura T, Yusa
T, Nakayama A, Miki H, Iwata H, Kamiguchi
H, Tanaka T, Habuka N, Sogabe S, Ohta Y and
Ishikawa T 2012 Design and Synthesis of Pyrrolo[3,2-
d]pyrimidineHumanEpidermalGrowthFactorReceptor
2 (HER2)/Epidermal Growth Factor Receptor (EGFR)
Dual Inhibitors: Exploration of Novel Back-Pocket
Binders J. Med. Chem. 55 3975
15. Clark P D and Irvine N M 1996 Preparation of 3,4-
disubstituted benzo[b]thiophenes for use in thia-ergoline
synthesis Phosphorus, Sulfur Silicon Relat. Elem. 118
61
16. Oshiro Y, Sato S, Kurahashi N, Tanaka T, Kikuchi
T, Tottori K, Uwahodo Y and Nishi T 1998 Novel
Antipsychotic Agents with Dopamine Autoreceptor
Agonist Properties: Synthesis and Pharmacology of
7-[4-(4-Phenyl-1-piperazinyl)butoxy]-3,4-dihydro-
2(1H)-quinolinone Derivatives J. Med. Chem. 41
658
17. Kinnick M D, Lin H-S, Martinelli M J, Morin J M and
Richett M E 2004 sPLA2 inhibitors U.S. Patent 6713505
18. Drewry D T and Scrowston R M 1969 Bromination and
Vilsmeier–Haack formylation of 6,7-dihydrobenzo[b]-
thiophen-4(5H)-one J. Chem. Soc. C 2750
ASK and SGSK thank Dr. Reddy’s Laboratories (DRL) for
support through DRILS Postdoctoral Fellowship Program.
We thank Dr. S. Vittal and Ms. Shweta Sawner for their pre-
liminary investigations into the synthesis of 7·2HCl and 8b.
References
1. (a) Das S, Barnwal P, Winston A B, Mondal S and Saha
I 2016 Brexpiprazole: so far so good Ther. Adv. Psy-
chopharmacol. 6 39;(b)BruijnzeelDandTandonR2016
Spotlight on brexpiprazole and its potential in the treat-
ment of schizophrenia and as adjunctive therapy for the
treatment of major depression Drug Des. Devel. Ther. 10
1641
2. World Health Organization (WHO). Schizophrenia.
Fact Sheet No. 397. http://www.who.int/mediacentre/
factsheets/fs397/en/ (reviewed in April 2016; accessed
on 19 April 2018)
3. (a) Goff D C 2015 Brexpiprazole: A New Antipsychotic
Following in the Footsteps of Aripiprazole Am. J. Psy-
chiatry 172 820; (b) Stahl S M 2016 Mechanism of
action of brexpiprazole: comparison with aripiprazole
CNS Spectr. 21 1
4. Jago C In 2015 Drugs to Watch Thomson Reuters Market
Insight Report. http://assets.fiercemarkets.net/public/
lifesciences/LS.2015.DrugstoWatch.MarketInsight_
Report%20FINAL%203.23.15.pdf (Accessed on 19
April 2018)
5. Kowalski P, Jas´kowska J and Majka Z 2012 Recent
Approaches to the Synthesis of Aripiprazole – A New
Generation Antypsychotic Drug Mini-Rev. Org. Chem. 9
374
6. Flick A C, Ding H X, Leverett C A, Kyne R E, Liu
K K C, Fink S J and O’Donnell C J 2017 Synthetic
Approaches to the New Drugs Approved During 2015
J. Med. Chem. 60 6480
7. Yamashita H, Sato T, Minowa T, Hoshika Y, Toyofuku
H, Yamaguchi T, Sota M, Kawano S, Nakamura T, Eto
R, Ikebuchi T, Moriyama K and Ito N 2013 Dihydrate
of benzothiophene compound or of a salt thereof, and
process for producing the same PCT Application WO
2013/162046 A1
19. Poorna Chander T, Satyanarayana B, Ramesh Kumar N,
Pratap Reddy P and Anjaneyulu Y 2007 Synthesis and

72 Page 10 of 10
J. Chem. Sci. (2018) 130:72
Characterization of Related Compounds of Aripiprazole, 20. Mukherjee C and De A 2001 Certain aspects of the chem-
an Antipsychotic Drug Substance Synth. Commun. 37
4337
istry of hydroxybenzo[b]thiophene J. Indian Inst. Sci. 81
417