A new approach for the ortho-substitution of anilines and for the synthesis of indolines

Article abstract of DOI:10.1016/j.tetlet.2004.04.111

Intermolecular radical addition of a xanthate to a vinyl sulfanilide is followed by ring closure to the aromatic ring to give a dihydrobenzoisothiazole dioxide structure, which upon heating loses sulfur dioxide to give a 2-substituted aniline; in some exa

Full text of DOI:10.1016/j.tetlet.2004.04.111

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 4631–4634
A new approach for the ortho-substitution of anilines
and for the synthesis of indolines
*
ꢀ
Cecile Moutrille and Samir Z. Zard
Laboratoire de Synthꢁese Organique associꢀe au CNRS, Ecole Polytechnique, F-91128Palaiseau, France
Received 10 March 2004; revised 13 April 2004; accepted 21 April 2004
Abstract—Intermolecular radical addition of a xanthate to a vinyl sulfanilide is followed by ring closure to the aromatic ring to give
a dihydrobenzoisothiazole dioxide structure, which upon heating loses sulfur dioxide to give a 2-substituted aniline; in some
examples, the presence ofDBU during heating induces the of rmation ofan indoline.
Ó 2004 Elsevier Ltd. All rights reserved.
R
The functionalisation of anilines in the ortho position is
often a key transformation in the synthesis of medici-
nally important compounds. For this task, the powerful
ortho-directed metallation is complemented by a number
ofless general but nevertheless quite useufl reactions.
Some indirect methods have also been occasionally
applied to this problem. These include the addition of
organometallic reagents to nitroarenes or the vicarious
substitution reactions explored extensively by Makosza
et al.1 and his collaborators, both followed by reduction
ofthe nitro group. Radical based approaches have
received in contrast scant attention. One reason is the
relative sluggishness ofthe radical addition to the aro-
matic ring, in comparison with other reaction pathways
open to the radical. Furthermore, the intermolecular
O
O
O
1) Et3N
2) Et3N
N
S
S
O
Cl
Cl
R
Y
NH
2
S
1
R'
S
O
Y
3
(lauroyl peroxide)
R
O
N
R
lauroyl peroxide
(stoichiometric)
S
O
N
O
S
Y
S
O
Y
R'
4
5
S
O
R'
variant suffers from generally low yields and an essen-
2
Scheme 1.
tially complete lack ofregioselectivity.
We have
attempted to remedy this situation by combining an
intermolecular radical addition with the use ofa tem-
porary sulfonyl tether to force a second, intermolecular
addition to occur at the ortho position on the aromatic
ring.
ing the formation of 5 by ring-closure onto the aromatic
ring. The required starting sulfanilides 2 are easily
obtained by reaction ofthe commercially available
2-chloroethylsulfonyl chloride with a base to induce
b-elimination ofthe chloride anion followed by reaction
with an aniline.4
Our conception is outlined in Scheme 1. A peroxide
initiated xanthate transfer radical addition3 to vinylsul-
fanilide 2 would lead to adduct 4, which could then be
subjected to a stoichiometric amount ofperoxide caus-
In the event, we found that the radical addition to the
vinyl sulfanilide 2a derived from N-methylaniline could
indeed be accomplished using a number ofdifferent
xanthates. The reaction is carried out by simply reflux-
ing a solution ofthe xanthate and 2a in 1,2-dichloro-
methane in the presence oflauroyl peroxide as initiator. 5
Although it was possible to isolate the expected adduct 4
Keywords: Radical addition; Anilines; Indolines; Cheletropic reaction.
* Corresponding author. Tel.: +33-(0)169334872; fax: +33-(0)169333-
851; e-mail: zard@poly.polytechnique.fr
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.04.111

4632
C. Moutrille, S. Z. Zard / Tetrahedron Letters 45 (2004) 4631–4634
Table 1
2
3
4
5 (%)
R
Y
R0
a
CH3
CH3
CH3
CH3
CH3
H
H
H
H
H
H
H
H
F
CH2CN
70
40
34
63
45
45
44
24
44
50
b
CH2COCH2CO2Et
(CH3)2CH2COCH3
CH2(3,4,5-OCH3)C6H2
CH2COt-Bu
CH2COPh
c
d
48%
e
fCH
3
g
h
i
H
CH2CN
CH2CN
50 (66)%
52 (62)%
38 (69)%
COCH3
CH3
CH2–Ph
CH2CN
CH2CN
j
H
Ph
N
N
O
O
k
CH2CO(4-OCH3)C6H4
20
S
in some cases, it was better to add directly a stoichi-
ometric amount ofthe peroxide to induce cyclisation to
the aromatic ring. This gave the corresponding di-
hydrobenzoisothiazole dioxide 5 in moderate to good,
but still unoptimised, yields. The results are compiled in
Table 1 (the yields in parentheses are based on recovered
starting material).
R
N
Y
∆
6
R
R'
-SO2
R
N
O
N
S
O
Y
R'
Y
Various types ofradicals could be added: benzylic (entry
d), tertiary (entry c), or radicals substituted with an
electrophilic group such as a nitrile (entry a), a ketone
(entries e and f) or ketoester (entry b). In the case of the
addition ofa tertiary radical, a quaternary centre is
generated and this is worth underlining. Replacing the
methyl group on the nitrogen with hydrogen or with an
acetyl group caused a lowering in the yield (entries g and
h). In one example, a 2-aminopyridine derivative could
be used instead ofan aniline (entry k).
5
R'
∆
R
-SO2
N
H
R'
Y
Y
∆
6'
R
σ [1,5]
NH
R'
7
Heating the cyclised derivatives in refluxing o-dichloro-
benzene caused a retro-cheletropic loss ofsulfur dioxide
to give a highly reactive intermediate diene 6 followed
by a 1,5-sigmatropic shift of a hydrogen to give the
2-substituted aniline 7,6 as outlined in Scheme 2. This
transformation is illustrated by the conversion of 5a, 5c
and 5d into anilines 7a, 7c and 7d in 68%, 60% and 77%
yields, respectively7 (Scheme 3). Although such extru-
Scheme 2.
NH
N
S
O
O
o-C6H4Cl2
CN
5a
68%
7a
CN
6
sions ofsulfur dioxide have been previously described,
they have seldom found use in synthesis because of a
lack ofconvenient routes to the cyclic precursors 5.
NH
N
S
O
O
∆
o-C6H4Cl2
5c
60%
7c
O
O
In the case of 5a, heating in the presence ofDBU caused
a further transformation of the initial product 7a into
indoline 9a through base induced migration ofthe ole-
finic bond followed by intramolecular Michael addition
(Scheme 4). The yield ofthis sequence was 70%, a quite
N
S
O
O
O
NH
O
O
∆
o-C6H4Cl2
5d
8
good yield given the nature ofthe intermediates. This
77%
7d
modification could be applied to the adducts possessing
an electron withdrawing group, which facilitates the
migration ofthe olefin and subsequent conjugate addi-
tion ofthe aniline nitrogen. In this way, indolines 9e,f
and 9i–k were prepared from compounds 5e,f and 5i–k
in 65%, 66%, 59%, 50% and 45% yields, respectively
O
O
O
Scheme 3.

C. Moutrille, S. Z. Zard / Tetrahedron Letters 45 (2004) 4631–4634
4633
Ph
Ph
DBU
o-C6H4Cl2
∆
DBU
NH
NH
N
N
N
N
O
S
CN
CN
CN
O
O
7a
8a
6k
5k
∆
-SO2
O
O
O
N
N
O
O
O
S
Ph
Ph
9a
70%
N
5a
N
N
O
CN
O
25%
10f
Scheme 4.
trace
10k
Scheme 6.
(Scheme 5). Under the same conditions, for compounds
5g and 5h, with no substituent on the nitrogen or with
an N-acetyl group, only a complex mixture was
obtained. The reason for this behaviour is still not clear.
intramolecular Diels–Alder reaction (Scheme 6). With
this knowledge in hand, a more careful re-examination
ofthe crude reaction mixture in the case ofthe aniline
analogue 5f indicated the presence oftrace amounts of
the corresponding tricyclic derivative 10f.
One final twist observed with compound 5k is the con-
comitant formation of tricyclic derivative 10k in 25%
yield, in addition to the ‘normal’ product 9k (45%). This
substance arises by capture ofthe especially reactive
intermediate diene 6k by the carbonyl group through an
In summary, we have established a flexible, versatile
route to anilines functionalised in the ortho position.
The starting materials and reagents are cheap and
readily available. The approach can further be modified
to access indolines and in principle the corresponding
indoles through dehydrogenation. These compounds are
ofspecial interest in medicinal chemistry and many of
those described in the present study would be difficult to
obtain by more classical approaches. Further studies
aimed at exploring the scope and improving the yield of
the addition–cyclisation sequence are under way.
O
DBU
o-C6H4Cl2
N
N
O
O
N
S
S
∆
65%
9e
5e
O
O
DBU
o-C6H4Cl2
Ph
O
O
N
∆
9f
66%
5f
Acknowledgements
Ph
O
One ofus (C.M.) grateuflly acknowledges generous
financial support from Rhodia. We also thank Drs.
Franßcois Metz and Jean-Marc Paris ofRhodia ofr
friendly discussions.
DBU
o-C6H4Cl2
CN
N
N
O
S
∆
O
F
F
59%
9i
5i
CN
Ph
Ph
References and notes
DBU
o-C6H4Cl2
N
N
CN
O
O
N
S
S
1. (a) Makosza, M.; Surowiec, M. J. Organomet. Chem. 2001,
624, 167–171, and references cited therein; (b) Paradisi, C.
In Comprehensive Organic Synthesis; Trost, B. M., Fleming,
I., Eds.; Pergamon: Oxford, 1991; Vol. 4, pp 423–450.
2. Studer, A.; Bossart, M. In Radicals in Organic Synthesis;
Renaud, P., Sibi, M., Eds.; Wiley-VCH: Weiheim, 2002;
Vol. 2, pp 63–80.
3. (a) Zard, S. Z. Angew. Chem., Int. Ed. Engl. 1997, 36, 672–
685; (b) Quiclet-Sire, B.; Zard, S. Z. Phosphorus, Sulfur, and
Silicon 1999, 153, 137–154; (c) Zard, S. Z. In Radicals in
Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-
VCH: Weinheim, 2001; Vol. 1, pp 90–108.
∆
50%
9j
5j
CN
Ph
Ph
O
DBU
N
N
O
O
O
N
o-C6H4Cl2
∆
5k
9k
45%
O
O
4. Marchand-Brynaert, J.; Bouchet, M.; Touillaux, R.;
Beauve, C.; Fastrez, J. Tetrahedron 1996, 52, 5591–
5606.
Scheme 5.

4634
C. Moutrille, S. Z. Zard / Tetrahedron Letters 45 (2004) 4631–4634
5. Typical procedure is as follows: to a boiling solution of 2a
(200 mg, 1.0 mmol) and 3a (330 mg, 2.0 mmol) in 1,2-
dichloroethane (2 mL), 10% oflauroyl peroxide was added
every hour until disappearance ofstarting materials (usu-
ally 120% was necessary). The reaction mixture was purified
by flash chromatography on silica gel to give an orange oil
(165 mg, 0.7 mmol, 70% yield). dH (CDCl3, 400 MHz): 2.33–
2.48 (m, 2H, CH2–CH2–CN), 2.62 (ddd, 1H, J ¼ 5:9, 8.2,
17.3 Hz, CH2–CH2–CN), 2.75 (ddd, 1H, J ¼ 7:9, 7.9,
17.3 Hz, CH2–CH2–CN), 3.13 (s, 3H, N–CH3), 4.34 (dd,
1H, J ¼ 5:3, 8.2 Hz, CH–SO2), 6.76 (d, 1H, J ¼ 8:2 Hz,
CHAr), 7.07 (dt, 1H, J ¼ 0:9, 7.6 Hz, CHAr), 7.25 (d, 1H,
J ¼ 7:6 Hz, CHAr), 7.38 (t, 1H, J ¼ 7:9 Hz, CHAr); dC
(CDCl3, 100 MHz): 14.3 (N–CH3), 26.2 (CH2), 26.5 (CH2),
58.6 (CH–SO2), 109.5 (CHAr), 118.2 (Cq), 120.9 (Cq), 122.4
(CHAr), 124.9 (CHAr), 130.2 (CHAr), 140.9 (CqAr). Anal.
Calcd for C11H12N2O2S: C, 55.91; H, 5.12. Found: C, 55.71;
H, 5.18.
CH3–N), 3.32 (dd, 2H, J ¼ 1:8, 5.3 Hz, CH2–CN), 4.05 (br
s, 1H, NH), 5.93 (dt, 1H, J ¼ 5:3, 15.6 Hz, CH@CH–CH2–
CN), 6.69 (d, 1H, J ¼ 8:2 Hz, CHAr), 6.73–6.77 (m, 1H,
CHAr), 6.80 (dt, 1H, J ¼ 1:8, 15.6 Hz, CH@CH–CH2–CN),
7.20–7.28 (m, 2H, CHAr); dC (CDCl3, 100 MHz): 21.1
(CH2), 30.9 (CH3), 110.5 (CH), 111.5 (Cq), 117.5 (CH),
118.9 (CH), 122.3 (Cq), 127.5 (CH), 129.6 (CH), 130.4
(CH), 146.1 (Cq). m (cmÀ1) 3449, 3046, 2985, 2918, 2873,
2816, 2253 (w), 1604 (s), 1579, 1511 (s), 1463, 1418, 1340,
1310, 1266, 1166, 1067, 970. Mass (ICP NH3) m=z
173 ¼ MHþ.
8. Typical procedure is as follows: a solution of 5a and DBU
was heated to reflux in o-dichlorobenzene for 30 min. The
reaction mixture was purified by flash chromatography on
silica gel to give an orange oil. dH (CDCl3, 400 MHz): 2.67
(dd, 1H, J ¼ 6:8, 16.7 Hz, Ph–CH2–CH), 2.76 (dd, 1H,
J ¼ 4:4, 16.7 Hz, Ph–CH2–CH), 2.81 (s, 3H, CH3–N), 2.92
(dd, 1H, J ¼ 9:1, 15.6 Hz, CN–CH2–CH), 3.30 (dd, 1H,
J ¼ 8:8, 15.6 Hz, CN–CH2–CH), 3.69 (m, 1H, CH2–CH–
CH2), 6.52 (d, 1H, J ¼ 7:9 Hz, CHAr), 6.76 (dt, 1H, J ¼ 0:9,
7.6 Hz, CHAr), 7.11–7.17 (m, 2H, CHAr); dC (CDCl3,
100 MHz): 21.9 (CH2), 34.4 (CH3), 35.1 (CH2), 62.9 (CH),
107.7 (CH), 117.5 (Cq), 118.8 (CH), 124.3 (CH), 127.3 (Cq),
127.9 (CH), 152.3 (Cq). Anal. Calcd for C11H12N2: C, 76.71;
H, 7.02. Found: C, 76.53; H, 7.03.
6. (a) Cava, M. P.; Deano, A. A. J. Am. Chem. Soc. 1959, 81,
4266–4267; (b) Wojciechowski, K. Synlett 1991, 571–572;
(c) Wojciechowski, K. Eur. J. Org. Chem. 2001, 3587–3605.
7. Typical procedure is as follows: a solution of 5a was heated
to reflux in o-dichlorobenzene for 30 min. The reaction
mixture was purified by flash chromatography on silica gel
to give an orange oil. dH (CDCl3, 400 MHz): 2.89 (s, 3H,