A novel intramolecular palladium-mediated cyclization for the synthesis of substituted 2-(aryl- or benzylamino)-4H-1,3-benzothiazines

Article abstract of DOI:10.1055/s-2008-1078206

Synthesis of 2-aryl/benzylamino 4H-1,3-benzothiazine derivatives was achieved via an unprecedented intramolecular palladium cyclization for creating a sulfur-aryl bond. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2008-1078206

LETTER
2433
A Novel Intramolecular Palladium-Mediated Cyclization for the Synthesis of
Substituted 2-(Aryl- or Benzylamino)-4H-1,3-Benzothiazines
S
ynthesis of Subst
a
itute
d
2-(A
r
v
yl- or Benzy
i
lamino
d
)-4
H
-1,3-Benzothi
O
Global Discovery Chemistry-Neuroscience, Novartis Institute for Biomedical Research, WKL-122.2.43, 4002 Basel, Switzerland
Fax +41(61)6962455; E-mail: david.orain@novartis.com
Received 21 April 2008
R2
S
R2
R1
R1
Abstract: Synthesis of 2-aryl/benzylamino 4H-1,3-benzothiazine
derivatives was achieved via an unprecedented intramolecular pal-
ladium cyclization for creating a sulfur–aryl bond.
R3
R3
i
N
N
N
H
H
R
X
X
Key words: palladium, sulfur, heterocycles, ring closure, catalysis
1a–g
R1, R2, R3 = H, alkyl
R = aryl
X = Br, I
ii
R2
R2
R1
R1
For one of our research programs, an efficient synthesis of
unknown 2-(a,a¢-arylamino)-4H-1,3-benzothiazine deriv-
atives was needed. A literature survey of 2-substitued
amino-4H-1,3-benzothiazine systems showed only very
limited reported examples and none of them with 2-(aryl-
amino) substituents. Recently, synthetic access of 2-di-
alkylamino-4H-1,3-benzothiazines, via a benzyne inter-
mediate, was reported by Sathunuru and co-workers.2
Nevertheless, this approach did not appear suitable for our
synthesis of 2-(arylamino) derivatives, especially in terms
of the potential scaffold’s functionalization/substitution.
In this communication, we would like to report a new and
efficient synthesis of 2-(aryl- and benzylamino)-4H-1,3-
benzothiazine derivatives allowing general substitution
patterns. For this synthesis, an unprecedented intramolec-
ular palladium cyclization creating a sulfur–aryl bond was
developed.
R3
N
N
N
or
R
R
N
S
S
H
R3 = H
R3 ≠ H
2a–g
Scheme 1 2-(Arylamino)benzothiazine synthesis. Reagents and
conditions: i) RNCS, Et3N, dioxane, r.t., 8 h; ii) Pd(PPh3)4 (10% mol),
Ph3P (10% mol), Et3N or DBU (2 equiv), dioxane, reflux.
To test the cyclization process, 1-(2-bromobenzyl)-3-phe-
nyl-thiourea (1a) and 1-(2-iodobenzyl)-3-phenyl-thiourea
(1a¢) were prepared by condensation of the corresponding
o-halobenzylamine and phenylisothiocyanate. These in-
termediates were then submitted to a standard palladium
protocol using 10% mol of Pd(PPh3)4 in the presence of
triethylamine in refluxing dioxane, and the reaction was
monitored by LC-MS. (Scheme 1)
The key step of our new synthetic approach was based on
a palladium-mediated intramolecular cyclization between
an o-iodo/bromo aryl and a thiourea. The intermediate
thiourea is obtained from the condensation of an o-halo-
substituted benzylic amine and an aryl/benzylisothiocya-
nate.
When 1a was used, the benzothiazine adduct 2a was ob-
tained in a modest 16% isolated yield after 16 hours of re-
action whereas with 1a¢, the yield was increased to 55%
and the reaction was complete in only 40 minutes. By add-
ing, 10 mol% of triphenylphosphine as a ligand, the yield
of benzothiazine was increased up to 67%.6 From this
small pilot screen, several conclusions could be made:
a) standard palladium chemistry is able to afford 2-(ami-
nophenyl)-4H-1,3-benzothiazine, b) iodine substitution
proved essential to ensure a fast and efficient conversion,
and c) addition of a co-ligand is improving the chemical
yield. It is noteworthy to mention that when applying the
nickel catalysis reported by Takagi, no adduct formation
was detected.
Our hypothesis was that under basic conditions the
isothiourea tautomer would be present and be able to
achieve intramolecular cyclization under palladium catal-
ysis.
Formation of S–aryl bonds using standard palladium cou-
pling procedures between various thiols and iodo aryl de-
rivatives has been described.3 In addition, Takagi and co-
workers showed that alkylthioamides were able to react
under Pd(0) conditions with o-iodoanilines to form 2-
alkylbenzothiazoles via the formation of a sulfur–aryl
bond before cyclization.4 The same author showed as well
that sulfur–aryl bond could be formed between thiourea
and iodaryls but this time under nickel catalysis.5
To establish the scope and limitations of this new reaction
with regards to steric hindrance, a selection of 2-(arylami-
no)-4H-1,3-benzothiazine was prepared by combination
of o-iodobenzylamines and a,a¢-arylisocyanates as report-
ed in Table 1.
When this new intramolecular cyclization was applied to
a,a¢-arylthiourea (1b and 1c), steric hindrance was not
detrimental to the reaction, and the benzothiazine deriva-
SYNLETT 2008, No. 16, pp 2433–2436
Advanced online publication: 22.08.2008
DOI: 10.1055/s-2008-1078206; Art ID: G13008ST
x
x
.x
x
.2
0
0
8
© Georg Thieme Verlag Stuttgart · New York

2434
D. Orain et al.
LETTER
Table 1 Synthesis of 2-(Arylamino)-4H-1,3-benzothiazines
Entry
1
Starting
thiourea
R
R1, R2
H, H
R3
H
X
Benzothiazine adducta
Yield (%)b
N
N
N
1a
Ph
Br
2a
16
S
S
S
N
H
2
3
1a¢
Ph
H, H
H, H
H
H
I
I
2a
2b
67
75
N
H
1b
1c
2,6-MeC6H3
N
H
Cl
Cl
N
4
5
2,6-ClC6H3
2,6-MeC6H3
H, H
H, Et
H
H
I
I
2c
63
86
S
N
H
N
1d
2d
S
N
H
Cl
Cl
N
6
7
1e
1f
2,6-ClC6H3
2,6-MeC6H3
Me, Me
H
I
I
2e
2f
32
83
S
N
H
N
H, H
Me
S
N
Cl
Cl
N
8
1g
2,6-Cl-4-F-C6H2
H, Et
Me
I
2g
63
S
N
F
a Reaction carried out analogously to the example synthesis reported in ref. 5 except that Et3N was replaced by DBU for entries 3–8.
b Isolated yield.
tives 2b and 2c were isolated in good yields. Compared to To further extend the possible scope of this new reaction,
our initial test conditions, the base was switched from tri- cyclization to a seven-membered ring and an intermolec-
ethylamine to DBU which led to better yields. Monosub- ular S–aryl bond formation between a thiourea and iodo-
stitution of the benzylic position (1d) was still tolerated benzene were examined under our standard protocol
but for the gem-dimethyl derivative (1e) a decreased yield (Scheme 2). In both cases, adducts 4 or 6 were not ob-
was observed. Interestingly, when a secondary amine (1f) served.
was used instead of a primary benzylamine, the cycliza-
tion was still very efficient. By combining benzylic and
After showing that N-benzyl-N¢-arylthioureas were good
substrates for intramolecular cyclization, we turned our
nitrogen susbsituents (1g), a highly substituted benzothi-
attention to N-benzyl-N¢-benzyl derivatives (Scheme 3).
azine 2g was synthesized in a 63% yield.
The first attempt to cyclize o-iodo-N-benzylthiourea 7 un-
Synlett 2008, No. 16, 2433–2436 © Thieme Stuttgart · New York

LETTER
Synthesis of Substituted 2-(Aryl- or Benzylamino)-4H-1,3-Benzothiazines
Table 2 Screening Conditions for the Synthesis of 8
2435
Entry Base Catalyst (mol%)
Ligand
(mol%)
Conversion at
1 h (%)a 16 h (%)a
Cl
Cl
Cl
N
H
N
NH
i
S
1
–
94
N
Et3N Pd(PPh3)4 (25)
S
H
I
3
4
Cl
2
3
DBU Pd(PPh3)4 (25)
DBU Pd(PPh3)4 (10)
DBU Pd(PPh3)4 (5)
DBU Pd(PPh3)2Cl2 (10)
DBU Pd(dppf)2Cl2 (10)
DBU Pd2(dba)3 (10)
DBU Pd(OAc)2 (10)
DBU Pd(OAc)2 (10)
DBU Pd(OAc)2 (10)
DBU Pd(OAc)2 (5)
DBU Pd(OAc)2 (5)
DBU Pd(OAc)2 (10)
DBU Pd(OAc)2 (10)
–
–
–
–
–
–
–
100
95
5
4
i
HN
N
H
5
5
N
NH
I
S
6
45
12
35
S
5
6
7
Scheme
2
Unsuccessful cyclization/addition when using
8
i) Pd(PPh3)4 (10 mol%), Ph3P (10 mol%), DBU (2.5 equiv) in dioxane
(90 °C).
9
PPh3 (25) 13
dppf (25) 100
10
11
12
13
14
dppf (2.5)
dppf (5)
dppf (5)
0
S
N
12
20
13
N
N
H
H
S
N
40
I
H
8
7
dppf (10) 46
100
Scheme 3 Synthesis of N-benzylbenzothiazine 8. Reagents and
conditions: palladium catalyst, ligand, DBU (2 equiv), dioxane, reflux
(cf. Table 2 for catalyst/ligand charge).
a Percentage of conversion of 7 to 8 by LC-MS at 254 nm based on
peak areas.
der the previously described conditions [Pd(PPh3)4 (0.1 cessible by increasing the pressure. Finally, by running
equiv), Et3N (2 equiv), dioxane, reflux) afforded 87 in a the reaction at 0.5 M (still in screw-capped vessels), the
low isolated yield (10%).
conversion was complete in less than two hours with an
isolated yield of 72%.9 When the optimized conditions for
the synthesis of 2-benzylamino benzothiazines was ap-
plied back to the synthesis of 2-(arylamino) derivatives,
the compounds were obtained only in low yields.
So, we decided to investigate this reaction by looking at
the catalyst/ligand source and the base. The reactions
were run in screw-capped tubes8 at 80 °C in dioxane and
analyzed by LC-MS after one hour and 16 hours
(Table 2).
In conclusion, a new, simple, and efficient synthetic ac-
cess to aryl- and benzylamino-substituted 4H-1,3-ben-
zothiazines has been developed, allowing a wide range of
possible diversification points. This new synthesis was re-
alized through an unprecedented palladium coupling reac-
tion for the generation of an S–aryl bond.
The lowest amount of catalyst needed was determined to
be 10 mol% (entry 3). With 5 mol% no reaction took
place. The best reaction conditions were found to be 10
mol% Pd(OAc)2 and 25 mol% dppf [1-1¢-bis(diphe-
nylphosphinino)ferrocene)] (entry 10). The ligand was
crucial for the reaction rate. When Ph3P was used instead
of dppf (entry 9), only 13% of conversion was observed
after one hour. Finally, after refinement of catalyst/ligand
charge, the optimal conditions were found when using 10
mol% of Pd(OAc)2 with 10 mol% of dppf.
Acknowledgment
The authors would like to thanks Mr. Alexandre Luneau for NMR
analysis and Dr. Henri Mattes and Dr. Bernard Roy for fruitful dis-
cussions.
The reaction was run on a gram scale at 0.25 M in an open
vessel with the above-mentioned conditions found during
the screening. After 30 hours, only 40% conversion was
observed. Concentration turned out to be a crucial factor
for this type of cyclization as well as high energy required.
Indeed, when we switched from an open vessel to the
screw-capped vessel system we used during the screening,
the cyclization run (at 0.25 M) was completed after 24
hours, and the cyclized material was obtained in an im-
proved isolated yield of 45%. This result showed that this
particular type of cyclization needs some high energy ac-
References and Notes
(1) Current address: Actelion Pharmaceuticals Ltd,
Gewerbestrasse 16, 4123 Allschwil, Switzerland.
(2) Sathunuru, R.; Zhang, H.; Rees, C. W.; Biehl, E.
Heterocycles 2005, 65, 1615.
(3) (a) Reaction on thiopurine: He, H.; Llauger, L.; Rosen, N.;
Chiosis, G. J. Org. Chem. 2004, 69, 3230. (b) Reaction on
thioimidazole and thiobenzothiazole: Schopfer, U.;
Schlapbach, A. Tetrahedron 2001, 57, 3069.
(4) Takagi, K.; Iwachido, T.; Hayam, N. Chem. Lett. 1987, 839.
Synlett 2008, No. 16, 2433–2436 © Thieme Stuttgart · New York

2436
D. Orain et al.
LETTER
(5) (a) Takagi, K. Chem. Lett. 1986, 265. (b) Takagi, K. Chem.
Lett. 1990, 2205.
(6) Synthesis of 2a
(8) A 40 mL clear vial, screw top hole cap with PTFE/silicone
septa form SUPELCO (ref 27180).
(9) Synthesis of 8
To a mixture of 1-(2-iodobenzyl)-3-phenylthiourea (1a¢, 500
mg, 1.4 mmol) and Et3N (0.38 mL, 2 equiv) in dioxane (20
mL), Ph3P (158 mg, 0.1 equiv), and Pd(PPh3)4 (36 mg, 0.1
equiv) were added under argon. The resulting mixture was
refluxed for 40 min. The solvent was removed in vacuo, and
the crude was partitioned between EtOAc and sat. NaHCO3
soln. The organic phase was separated, washed with sat.
NH4Cl, dried over Na2SO4, and concentrated in vacuo to
afford a crude brown oil (520 mg). The crude material was
purified by flash chromatography on SiO2 using hexanes–
EtOAc (100:0 to 60:40) as solvent system. From the
purification, 2 (220 mg, 67.4% yield) was isolated as a light
yellow oil. 1H NMR (400 MHz, CDCl3): d = 7.00–7.50 (m,
9 H), 4.52 (s, 2 H). 13C NMR (100 MHz, CDCl3): d = 152.6,
149.2, 134.6, 131.7, 129.7, 128.1, 127.7, 127.6, 127.4,
123.7, 120.6, 53.1. IR (solution in CH2Cl2): 3418, 3054,
1638, 1590, 1518, 1498, 1437, 1311, 1252, 1031, 1025 cm–1.
HRMS: m/z calcd for C14H12N2S: 241.07940 [M + H+];
found : 241.07935 [M + H+].
In a screw-capped vial, to a suspension of 1-benzyl-3-(2-
iodobenzyl)-thiourea (7, 945 mg, 2.5 mmol) and DBU (0.76
mL, 2 equiv) in dioxane (5 mL), dppf (141 mg, 0.1 equiv),
and Pd(OAc)2 (57 mg, 0.1 equiv) were added under argon.
The resulting mixture was stirred at 80 °C for 2 h. The
solvent was removed in vacuo, and the crude material was
purified by flash chromatography on SiO2 using hexanes–
EtOAc (70:30) as solvent system. The fractions containing
the product were concentrated in vacuo to afford 570 mg of
a light yellow solid. This solid was sonicated in hexanes and
the resulting precipitate was filtered off, washed with
hexanes, and after high-vacuum drying, 8 (452 mg, 71.9%
yield) was isolated as a white powder. 1H NMR (400 MHz,
DMSO-d6): d = 7.41 (br m, 1 H, NH), 7.15–7.36 (m, 9 H),
4.36 (d, J = 4.8, 2 H), 4.27 (s, 2 H). 13C NMR (100 MHz,
CDCl3): d = 154.9, 139.0, 135.5, 131.8, 129.3, 128.6, 128.2,
127.8, 127.6, 127.5, 127.4, 54.3, 48.3. IR (solution in
CH2Cl2): 3427, 3034, 1633, 1483, 1468, 1251, 1205, 1192,
1129, 1066 cm–1. HRMS: m/z calcd for C15H14N2S:
255.09505 [M + H+]; found: 255.09503 [M + H+].
(7) Klatsmanyi-Gabor, P.; Meisel, T.; Erdey, L. Acta Chim.
Acad. Sci. Hung. 1964, 40, 99.
Synlett 2008, No. 16, 2433–2436 © Thieme Stuttgart · New York