Cyclic guanidines. Part 6.1 synthesis of benzimidazoles by intramolecular vicarious nucleophilic substitution of hydrogen

Article abstract of DOI:10.1039/a900553f

m-Nitrophenylguanidines, which carry a vicarious leaving group, cyclize under basic conditions yielding nitrobenzimidazoles. Preliminary observations of the regioselectivity of the ring closure reaction are reported.

Full text of DOI:10.1039/a900553f

View Article Online / Journal Homepage / Table of Contents for this issue
Cyclic guanidines. Part 6.1 Synthesis of benzimidazoles by
intramolecular vicarious nucleophilic substitution of hydrogen
Franz Esser,* Peter Ehrengart and Hans Peter Ignatow
Department of Medicinal Chemistry, Boehringer Ingelheim, 55216 Ingelheim, Germany.
E-mail: ESSERF@ing.boehringer-ingelheim.com; Fax: 06132-773418
Received (in Cambridge) 20th January 1999, Accepted 2nd March 1999
H
N
H
N
m-Nitrophenylguanidines, which carry a vicarious leaving
O2N
NCS
O2N
R1
H2N R1
group, cyclize under basic conditions yielding nitro-
benzimidazoles. Preliminary observations of the regio-
selectivity of the ring closure reaction are reported.
CH3CN
rt, 1.3 h
S
2 - 4
1
The principle of vicarious nucleophilic substitution of hydro-
gen (VNS) was first described by Golinski and Makosza.2 This
type of reaction has been used for the amination of nitro-
arenes3 and also examples with O-alkyl as a vicarious leaving
group are known.4 Occasionally intramolecular VNS has been
applied in the preparation of heterocycles.5 Among the less
frequently employed procedures for preparing benzimidazoles
are those which start from N-substituted NЈ-arylamidines or
-guanidines.6 The intramolecular attack of the nitrogen
atom on the aromatic moiety is accompanied by loss of the
N-substituent. Although related to the VNS reaction, these
processes are not clearly defined nucleophilic reactions since the
arenes are not activated by suitable functional groups.
Previously we described the preparation of benzimidazoles
by ring closure of (3-nitrophenyl)guanidines.7 This base
induced redox process suffers from the concomitant formation
of azoxy compounds which limits the yield to ≤60%. In order to
transform this reaction to a VNS reaction, we tried to intro-
duce a potential vicarious leaving group into the starting
guanidine. By this means we should be able to circumvent the
formation of azoxy side products and raise the theoretical
yields to 100%. As a matter of fact, in none of the described
cyclization reactions have azoxy compounds been detected as
by-products.
The preparation of the adducts for the VNS is depicted in
Scheme 1. The addition of an amine to 1 proceeds almost quanti-
tatively. For the synthesis of 5–11 from 2–4 the procedure of
Moynihan8 was followed which is performed under nitrogen.
When the guanidines 5–11 are treated with 1 to 3.8 equiv-
alents of KOtBu in DMSO9 or DMF,10 cyclization occurs
resulting in the formation of 12–17 as yellow to orange
coloured substances (Scheme 2). In some instances side prod-
ucts of type 18–20 could be identified, the structures of which
correspond to 15–17 with the NO2-group being replaced by
OR. Analogous ethers corresponding to 12–14 were not
observed. However, 18–20 are not formed by direct interconver-
sion from 15–17 since treatment of 15 with methoxide under
cyclization conditions did not yield any 18. Therefore, 18–20
must result from the reaction of some intermediate species with
in situ generated alkoxide.
Et2O
Et3N
H2NOR
HgO
rt,
1-3 d
H
N
H
N
O2N
R1
N
OR
5 - 11
Compound
R
R1
Total yield (%)a
5
6
Me
Et
Bn
Me
Et
Bn
Me
Bu
Bu
Bu
Pri
Pri
Pri
Bn
65
63
89
79
80
68
75
7
8
9
10
11
a Isolated yield.
Scheme 1
decrease of side products. All these observations taken together
do not allow clearcut regioselectivity rules to be deduced, but
they only show rough trends. Generally the ortho cyclization
seems to be preferred, which is in accord with mechanistic
ideas,11 as well as with observations from related cyclizations.7
As far as the mechanism of the reaction is concerned, we
believe, that cyclization is initiated by deprotonation of the
–NH–OR group. In general the –NORϪ nucleophile attacks
predominantly ortho to the nitro group leading to an anionic
σ-complex 21. Elimination of ROH followed by reprotonation
would give 12–14. At present we regard the para cyclization as
the exception.
– O
O –
H
+
N
O
R
N
H
N
R1
N
Various observations about the regioselectivity of the ring
closure reaction can be made. The nucleophile attacking the
nitroaromatic moiety is always represented by the guanidino-N
bearing the alkoxy-group and never by the guanidino-N carry-
ing the alkyl (or arylalkyl) group. The ratio of the ortho cycliz-
ation products 12–14 to para products 15–17 seems to depend
on several parameters, for example, the nature (size) of R,
reaction temperature and solvent. Clearly OMe as a vicarious
leaving group seems to favour the ortho cyclization irrespective
of the nature of R1, solvent and temperature. Also when R1 is
Pri (8–10) then the ortho cyclization is predominantly triggered.
The two experiments in DMF reveal a tendency towards a
21
The significance of this reaction is as a route to specifically
substituted benzimidazoles. The regioselectivity seems to be
influenced by the nature of the vicarious leaving group OR as
well as by the reaction conditions. Future experiments with this
type of ring closure will allow rules for the expected regio-
selectivity to be established. There are many possibilities for
chemically modifying the nitro group which offer the potential
of preparing benzimidazoles as starting materials for multiple
purposes.
J. Chem. Soc., Perkin Trans. 1, 1999, 1153–1154
1153

View Article Online
NO2
H
N
H
N
R1
H
N
O2N
N
KOtBu
R1
N
HN
12 - 14
OR
5 - 11
+
H
N
O2N
H
N
R1
N
15 - 17
+
H
N
R
O
H
N
R1
N
18 - 20
Starting material
R
R1
KOtBu (equiv.)
Solvent
Reaction temp. (ЊC)
Products (yield [%])a
5
6
7
8
Me
Et
Bn
Me
Bu
Bu
Bu
Pri
2.5
1.0
1.1
3.0
3.0
3.0
3.0
3.8
2.5
DMSO
DMSO
DMSO
DMSO
DMF
DMSO
DMF
20
70
70
20
20
50
50
20
60
12 (57)
12 (4)
15 (1.6)
15 (16)
15 (16)
16 (1b)
—
18 (4.7)
—
—
12 (3.8)
13 (64)
13 (65)
13 (24)
13 (36)
13 (47)
14 (42)
—
—
9
Et
Pri
16 (24b)
16 (9b)
—
19 (12b)
—
10
11
Bn
Me
Pri
Bn
DMSO
DMSO
19 (7b)
20 (13)
17 (7.1)
a Isolated yield, unless stated otherwise. b Yield calculated from mixture.
Scheme 2
20%), were obtained 12 (1.42 g, 57%), mp 197–200 ЊC; δH (250 MHz,
CDCl3): 10.14 (1H, s, NH), 7.81 (1H, d, J 7.6, ArH5), 7.65 (1H, d,
J 7.6, ArH7), 7.14 (1H, dd, J 7.6, 7.6, ArH6), 5.39 (1H, t, J 4.5,
NHCH2), 3.54 (2H, m, NHCH2), 1.70 (2H, m, CH2), 1.44 (2H, m,
CH2CH3), 0.95 (3H, t, J 7.0, CH3); MS (ion spray) m/z 235
(M ϩ Hϩ, 100%); 15 (0.04 g, 1.6%), mp 182–183 ЊC; δH (250 MHz,
CDCl3–DMSO-d6 = 3:1): 8.06 (1H, d, J 2.4, ArH4), 7.98 (1H, dd,
J 8.9, 2.4, ArH6), 7.24 (1H, d, J 8.9, ArH7), 5.83 (1H, t, J 4.3,
NHCH2), 3.47 (2H, m, NHCH2), 1.65 (2H, m, CH2), 1.44 (2H, m,
CH2CH3), 0.96 (3H, t, J 7.5, CH3); MS (ion spray) m/z 235
(M ϩ Hϩ, 100%) and 18 (0.11 g, 4.7%) as a highly viscous oil, δH (250
MHz, CDCl3): 8.37 (1H, s, NH), 6.99–6.86 (2H, m, ArH4 ϩ ArH7),
6.58 (1H, dd, J 6.4, 2.4, ArH6), 5.56 (1H, s, NHCH2), 3.86 (3H, s,
OCH3), 3.38 (2H, t, J 7.0, NCH2), 1.50 (2H, m, CH2), 1.23 (2H, m,
CH2CH3), 0.78 (3H, t, J 7.5, CH2CH3); MS (ion spray) m/z 220
(M ϩ Hϩ, 93%), 164 (57), 149 (100).
Notes and references
1 Part 5: F. Esser, K.-H. Pook, A. Carpy and J.-M. Léger, Synthesis,
1994, 77.
2 J. Golinski and M. Makosza, Tetrahedron Lett., 1978, 3495.
3 (a) A. R. Katritzky and K. S. Laurenzo, J. Org. Chem., 1988, 53,
3978; (b) M. Makosza and M. Bialecki, J. Org. Chem., 1992, 57,
4784.
4 S. Seko and N. Kawamura, J. Org. Chem., 1996, 61, 442.
5 (a) M. Makosza and K. Wojciechowski, Tetrahedron Lett., 1984, 25,
4791; (b) K. Wojciechowski and M. Makosza, Synthesis, 1992,
571; (c) M. Makosza and H. Hoser, Heterocycles, 1994, 37, 1701;
(d) S. V. Litvinenko, Y. M. Volovenko and F. S. Babichev, Khim.
Geterotsikl. Soedin., 1993, 367; (e) M. Makosza and J. V. Krylova,
Eur. J. Org. Chem., 1998, 2229.
6 (a) P. N. Preston, in Heterocyclic Compounds, Vol 40, Part 1, ed. P. N.
Preston, John Wiley & Sons, New York 1981, p. 42; (b) J. Garapon
and B. Sillion, Bull. Soc. Chim. Fr., 1975, 2677; (c) P. Held, M. Gross
and H. Schubert, Z. Chem., 1973, 13, 292.
10 Representative experimental procedure with DMF as solvent: 8
(5.05 g, 20 mmol) was dissolved in DMF (90 ml) and KOtBu (6.72 g,
60 mmol) was added under nitrogen. The solution was stirred at
room temperature for 17 h. It was poured into 1 M Na2CO3 solution
(1 l), which was extracted with ethyl acetate (2 × 0.5 l). The com-
bined extracts were filtered and concentrated under reduced pres-
sure. The crude product was chromatographed over silica gel using
ethyl acetate–cyclohexane (1:1) as eluent. From the main fraction 13
(2.84 g, 65% yield) was obtained as orange coloured crystals, mp
246–249 ЊC; δH (250 MHz, CDCl3): 9.80 (1H, s, NH), 7.81 (1H, dd,
J 8.5, 0.9, ArH5), 7.65 (1H, d, J 7.9, ArH7), 7.14 (1H, dd, J 8.5, 7.9,
ArH6), 4.94 (1H, d, J 7.3, NH-CH), 4.15 (1H, m, NH-CH), 1.36
(6H, d, J 6.4, (CH3)2CH); MS (ion spray) m/z 221 (M ϩ Hϩ, 100%).
11 M. Makosza and J. Winiarski, Acc. Chem. Res., 1987, 20, 282.
7 F. Esser and K.-H. Pook, Synthesis, 1992, 596.
8 H. A. Moynihan, S. M. Roberts, H. Weldon, G. H. Allcock, E. E.
Ångga˚rd and T. D. Warner, J. Chem. Soc., Perkin Trans. 1, 1994,
769.
9 Representative experimental procedure with DMSO as solvent: 5
(2.82 g, 10.6 mmol) was dissolved in DMSO (50 ml) and KOtBu (3 g,
26.4 mmol) was added under nitrogen. The solution, which was
stirred at room temperature for 9.5 h changed colour from light
brown via dark blue to red–brown. It was poured into ice-cold water
(700 ml), extracted with a 1:1 mixture of ether–ethyl acetate and
with ether. The combined organic phases were washed with water,
filtered and the solvent evaporated. The crude product was flash
chromatographed over silica gel using ethyl acetate–cyclohexane
(1:1) and ethyl acetate as eluents. Apart from unchanged 5 (0.55 g,
Communication 9/00553F
1154
J. Chem. Soc., Perkin Trans. 1, 1999, 1153–1154