Electroreduction of 1-methyl 5-nitroindole, 5-nitrobenzofurane, and 5-nitrobenzothiophene in acidic and basic hydroorganic media: Generation and trapping of iminoquinone-type intermediates and electrosynthesis of ring-substituted amino derivatives

Article abstract of DOI:10.1139/v03-149

Preparative electrolysis of 1-methyl-5-nitroindole (1b, X = NCH ₃), 5-nitrobenzofurane (1c, X = O), and 5-nitrobenzothiophene (1d, X = S) at Hg, in acidic hydromethanolic media, leads to the formation of the corresponding 4-substituted amino derivatives 5, which result from the 100% regioselective addition to iminoquinone-type intermediate 4 of methanol or of any other good nucleophile present in the electrolytic solution. In acidic medium, the iminoquinonium intermediates 4b and 4c were trapped in a cycloaddition reaction with cyclopentadiene added to the electrolysis medium. The regiochemistry of the nucleophilic addition is discussed in light of AM1 calculations.

Full text of DOI:10.1139/v03-149

1
108
Ele ctrore duction of 1-me thyl 5-nitroindole , 5-
nitrobe nzofurane , and 5-nitrobe nzothiophe ne in
acidic and bas ic hydroorganic me dia: Ge ne ration
and trapping of iminoquinone -type inte rme diate s
and e le ctros ynthe s is of ring-s ubs titute d amino
de rivative s
Luc Bouc ha rd, Ia n Ma rc otte , J e a n Ma rc Cha puze t, a nd J e a n Le s s a rd
Abstract: Preparative electrolysis of 1-methyl-5-nitroindole (1b, X = NCH ), 5-nitrobenzofurane (1c, X = O), and 5-
3
nitrobenzothiophene (1d, X = S) at Hg, in acidic hydromethanolic media, leads to the formation of the corresponding
4-substituted amino derivatives 5, which result from the 100% regioselective addition to iminoquinone-type intermediate 4
of methanol or of any other good nucleophile present in the electrolytic solution. In acidic medium, the iminoquin-
onium intermediates 4b and 4c were trapped in a cycloaddition reaction with cyclopentadiene added to the electrolysis
medium. The regiochemistry of the nucleophilic addition is discussed in light of AM1 calculations.
Key words: 1-methyl-5-nitroindole, 5-nitrobenzofurane, 5-nitrobenzothiophene, iminoquinone, electrosynthesis.
Résumé : L’électrolyse préparative du 1-methyl-5-nitroindole (1b, X = NCH ), du 5-nitrobenzofurane (1c, X = O) et
3
du 5-nitrobenzothiophene (1d, X = S), sur Hg et en milieu hydrométhanolique, mène à la formation des amines corres-
pondantes 5 substituées en position 4, lesquelles proviennent de l’addition, aux intermédiaires de type iminoquinonique
4
, du méthanol ou d’autres bons nucléophiles présents au sein de la solution électrolytique. L’addition du nucléophile
est 100% régiosélective. En milieu acide, les intermédiaires diiminoquinonium 4b et 4c ont été piégés par une réaction
péricyclique avec le cyclopentadiène ajouté à la solution électrolytique. La régiochimie de l’addition du nucléophile est
discutée à la lumière des résultats de calculs AM1.
Mots clés : 1-méthyl-5-nitroindole, 5-nitrobenzofurane, 5-nitrobenzothiophène, iminoquinone, électrosynthèse.
Bouchard et al. 1118
Introduc tion
and 6 are potential intermediates, and one could consider the
possibility of a diprotonated intermediate 4aH , depending
+
Recently, we reported the electroreduction at a mercury
cathode of 5-nitroindole (1a) in acidic, neutral, and basic
hydroorganic media (1, 2). As shown in Scheme 1, prepara-
tive electrolysis of 5-nitroindole (1a) in acidic or basic meth-
anolic solutions gives 4-substituted-5-aminoindoles 5a. In
acidic medium, the latter are formed through dehydration of
the protonated hydroxylamine 2a to the diiminoquinonium
intermediate 4a followed by the regioselective addition of
methanol and (or) of any other good nucleophile present in
on the basicity of the imino nitrogen of 4a and of the ring
nitrogen of 6. In basic medium, the hydroxylamine and the
diiminoquinone are not protonated (see 7a and 8, respec-
tively). In the presence of a good nucleophile, such as bro-
mide ion in acidic medium or thiophenolate anion in basic
medium, substituted aminoindole 5a is the sole compound
isolated in yields of 71%–86% and 59%, respectively (2).
–
1
the electrolytic solution. In such acidic medium (0.15 mol L
in acid, pH = 0.3), two monoprotonated diiminoquinones 4a
Received 25 June 2003. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 7 October 2003.
L. Bouchard, I. Marcotte, J.M. Chapuzet, and J. Lessard.1
Laboratoire de chimie et électrochimie organiques,
Département de chimie, Université de Sherbrooke,
Sherbrooke, QC J1K 2R1, Canada.
We studied the electroreduction of 1-methyl-5-nitroindole
1b), 5-nitrobenzofurane (1c), and 5-nitrobenzothiophene
1d) at a mercury cathode, in the same acidic (pH = 0.3) and
(
(
1Corresponding author (e-mail:
Jean.Lessard@USherbrooke.ca).
basic (pH = 13.5) aqueous methanolic solutions (MeOH–
H O 95:5 (volume fraction)) as those used for the electro-
2
Can. J. Chem. 81: 1108–1118 (2003)
doi: 10.1139/V03-149
© 2003 NRC Canada

Bouchard et al.
1109
Scheme 1.
reduction of 1a (2). The aim was to get a better understand-
the voltammetric behaviour of these nitro derivatives in
basic medium is similar to the well-known voltammetric be-
haviour of nitroaryl groups (3). The decreasing order of ease
of reduction in basic medium (1c > 1b > 1a > 1d) differs
from that in acidic medium (1a > 1b > 1d > 1c), so there
seems to be no straightforward relationship between the
structure (for instance, the electronegativity of X and its
electron-withdrawing or electron-donating ability) and the
reduction potential.
ing of the nature and reactivity towards nucleophiles of
+
cationic iminoquinone intermediates 4 and (or) 4H (neutral
diiminoquinones are not possible intermediates in these cases)
and to see if substituted 5-amino derivatives 5 could also be
obtained in good yield and high regioselectivity from this
family of nitro compounds.
Volta m m e tric m e a s ure m e nts
Cyclic voltammetry was used to determine the reduction
potential of 1-methyl-5-nitroindole (1b), 5-nitrobenzofurane
1c), and 5-nitrobenzothiophene (1d) in acidic and basic aque-
Pre pa ra tive e le c trolys e s of 1-m e thyl-5-
nitroindole (1b)
(
ous methanolic solutions. The data are summarized in Ta-
ble 1. An increase of the scan rate from 100 to 500 mV s
caused the expected increase of current for the diffusion
controlled processes. In acidic medium (entries 2–4 and 6–
The results of preparative electroreduction of 1-methyl-5-
nitroindole (1b) at Hg, in methanol–water (95:5 volume
fraction) solutions and under controlled potential conditions
(EW ~ (E ) – 100 mV in acidic medium, E ~ (E ) –
–1
pc 1
W
pc 1
8
), the cyclic voltammograms of the three nitro compounds
300 mV in basic medium), are presented in Table 2.
are similar to that of 5-nitroindole (1a) (entries 1 and 5) (2)
and show only one peak, corresponding to a six-electron re-
duction to the amine. 5-Nitrobenzofurane (1c) has the most
negative reduction potential, –0.58 and –0.51 V vs. ECS
In acidic medium (entries 1 to 3), the results are very sim-
ilar to those reported previously for the electroreduction of
the parent 5-nitroindole (1a) in the same media (2). With
–
1
HBr as acid and supporting electrolyte (0.15 mol L , pH =
(
entries 3 and 7, respectively), followed by 5-nitrobenzothio-
0.3) (entry 1), the electroreduction of 1b consumed
–
1
phene (1d), –0.51 and –0.47 V (entries 4 and 8), then by 1-
methyl-5-nitroindole (1b), –0.47 and –0.44 V (entries 2 and
4.6 F mol (1 F = 96 488 C) and led to 1-methyl-4-bromo-
5-aminoindole (5b, Nu = Br) as the unique product in 57%
yield. With H SO as acid and supporting electrolyte
6
–
), and finally by 5-nitroindole (1a), –0.34 V (entry 1) and
0.35 V (entry 5). In basic medium, the cyclic voltammo-
2
4
–
1
(0.15 mol L ) (entry 2), the electrolysis of 1b gave a 0.7 ra-
–1
grams of 1-methyl-5-nitroindole (1b) (entry 10: (E ) =
tio of 3b and 5b (Nu = MeO) and consumed also 4.6 F mol .
The mechanistic hypothesis for the formation of 3b and 5b
(Scheme 1) is the same as that proposed previously for the
electroreduction of 5-nitroindole (1a) in an acidic medium
(2). First, 1b is reduced to the protonated hydroxylamine 2b
through four successive electron and proton transfers. Then,
there could be a competition between path A, the two-
electron reduction of 2b to aminoindole 3b (protonated in
the medium), and path B, acid-catalyzed elimination of wa-
ter from 2b to give a diiminoquinonium intermediate (4b
pc 1
–
1.03 V and (E ) = –1.25 V) and of 5-nitrobenzofurane
pc 2
(
1c) (entry 11: (E ) = –1.03 V and (E ) = –1.25 V) are
pc 1 pc 2
also similar to that of 5-nitroindole (1a) (entry 9: (E ) =
pc 1
–
1.06 V and (E ) = –1.28 V) (2) in showing two reduction
pc 2
peaks (i + i /i of about 2), corresponding, respectively, to
1
2 1
the reversible monoelectronic reduction to the radical anion
and to the irreversible three-electron reduction of the radical
anion to the hydroxylamine. However, the cyclic voltammo-
gram of 5-nitrobenzothiophene (1d) shows only one four-
+
electron reduction peak (entry 12: (E ) = –1.09 V). The
and (or) 4bH ). Since 3b was not detected in the products of
pc 1
cyclic voltammograms of all four nitro derivatives show the
presence, on the second and third half-cycles at a less nega-
tive potential, of the hydroxylamine–nitroso reversible redox
couple (see peaks (E ) and E in entries 9 to 12). Thus,
the electrolysis in the HBr medium (Table 2, entry 1), it is
reasonable to assume that reaction A is slower than reaction
B and thus may be neglected as proposed for the electro-
reduction of 1a (2). 1-Methyl-5-aminoindole (3b) would
pc 3
pa
©
2003 NRC Canada

1
110
Can. J. Chem. Vol. 81, 2003
Table 1. Cyclic voltammetric data of 1-methyl-5-nitroindole (1b), 5-nitrobenzofurane (1c), and 5-nitrobenzothiophene (1d) at Hg in
methanol–water (95:5 volume fraction).
Entry Electrolytea Substrateb (E ) (V) (i ) (µA)
(E ) (V) (i ) (µA) Epa (V) ipa (µA) (E ) (V) (i ) (µA)
c
pc 1
–0.34
–0.47
–0.58
–0.51
–0.35
–0.44
–0.51
–0.47
–1.06
–1.03
–0.95
–1.09
pc 1
pc 2
pc t
pc 3
pc 3
1
2
3
4
5
6
7
8
9
0
1
2
HBr
1a
1b
1c
1d
1a
1b
1c
1d
1a
1b
1c
1d
–7.05
–5.36
–2.55
–4.54
–6.95
–4.91
–3.18
–4.05
–2.70
–1.56
–0.91
–3.48
d
d
H SO4
2
KOH
–1.28
–1.25
–1.42
–5.07
–2.24
–2.10
–0.58
–0.56
–0.50
–0.46
0.41
0.56
0.46
0.63
–1
–0.61
–0.68
–0.57
–0.52
–0.17
–0.48
–0.20
–0.30
1
1
1
–
Note: T = 25 ± 2 °C; glassy carbon as auxiliary electrode; Ag/AgCl/(Cl ) as reference electrode; v = 100 mV s . Variations in peak current are due
dil.
in part to variations of the surface of the mercury hanging drop electrode.
a
b
c
–1
–1
–1
Electrolytes: 0.15 mol L HBr; 0.15 mol L H SO ; 0.15 mol L KOH.
2
4
–4
–1
Substrate concentration: 2.5 × 10 mol L .
i represents the sum i + i where i is the current of the first peak and i the current of the second peak.
t
1
2
1
2
d
Taken from ref. 2.
Table 2. Controlled potential electroreduction of 1-methyl-5-nitroindole (1b) (6 × 10–3 mol L–1 solution) at Hg in MeOH–H O
2
(
95:5 volume fraction).
Yield of products (%)a
Entry
Supporting electrolyteb
HBr
Added nucleophilec
EWd (V)
Charge (F mol–1)
3b
5b
1
2
3
—
—
PhSH
–0.54
–0.56
–0.56
4.6
4.6
3.8
—
13
5e
57 (Nu = Br)
19 (Nu = MeO)
9 (Nu = MeO)
e
e
e
H SO4
2
H SO4
2
e,f
2
0 (Nu = PhS)
2e,g
7 (Nu = MeO)
e
4
KOH
—
–1.31
8.5
a
Yield of product(s) isolated by flash chromatography on silica gel (unless stated otherwise) and based on two to three electrolyses (range given when
variations were larger than ±4%).
b
Supporting electrolyte concentration; see footnote b of Table 1.
c
–2
–1
Added nucleophile concentration: 1.2 × 10 mol L .
d
Working potential vs. SCE.
e
Determined by vapor-phase chromatography.
f
A third substitution product, 5-amino-1-methyl-3-thiophenoxyindole (9b), was isolated in 20% yield.
g
The 5,5′-azo derivative 11b was isolated in 27% yield.
then be formed by a two-electron reduction of the diimino-
quinonium intermediates 4b and (or) 4bH (path C), and 4-
are again similar to those obtained with 5-nitroindole (1a)
under the same conditions (formation of 9a in 14% yield)
(2). It is reasonable to assume that both isomers 5b (Nu =
SPh) and 9b come from the addition of thiophenol to 4b and
+
bromo-5-aminoindole (5b, Nu = Br) would result from the
addition of a bromide ion to these intermediates (path D).
The fact that 4-bromo-5-amino-1-methylindole (5b, Nu =
Br) was the sole product isolated shows that the addition of
+
(or) 4bH . The larger proportion of thiophenoxyindoles (5b
(Nu = SPh) + 9b: 40%) over 4-methoxy-5-aminoindole (5b,
Nu = OMe: 9%) shows that thiophenol is a better nucleo-
phile than methanol in the electrolysis medium (protic me-
dium), which is in agreement with the fact that hydrogen
bonding to sulphur is weaker than hydrogen bonding to oxy-
gen. That thiophenol is a better nucleophile than methanol in
such protic medium is seen also in the relative proportion
of substituted-1-methyl-5-aminoindoles (5b + 9b) and 1-
methyl-5-aminoindole (3b) (proportional to the ratio of the
rate of reaction D over the rate of reaction C, Scheme 1),
–
+
Br to 4b and (or) 4bH (path D) is faster than their reduc-
tion to 3b (path C). In the absence of bromide ions — meth-
–
anol being a weaker nucleophile than Br in such protic
–
medium (hydrogen bonds to Br are weak) — the reduction
+
of 4b and (or) 4bH to 5b (path C) now competes effec-
tively with the addition of methanol to these intermediates
(
path D), the rate of the two reactions being similar (rate of
reaction D : rate of reaction C ~ 1.5, entry 2). In the pres-
ence of thiophenol (entry 3), the electroreduction of 1-
methyl-5-nitroindole (1b) gave 1-methyl-5-aminoindole (3b)
which
9b(%))/3b(%) ~
is larger in the presence of thiophenol ((5b(%) +
10, entry 3) than in its absence
(
(
5%), 1-methyl-4-methoxy-5-aminoindole (5b, Nu = MeO)
9%), and two substituted 5-amino indoles in equimolar
(5b(%)/3b(%) ~ 1.5, entry 2). The material balance based on
the weight of crude product after work-up was nearly quanti-
tative for all electrolyses in acidic medium (entries 1 to 3),
amounts: the 4-thiophenoxy derivative 5b (Nu = PhS) (20%)
and the 3-thiophenoxy derivative 9b (20%). These results
©
2003 NRC Canada

Bouchard et al.
1111
Scheme 2.
and both 3b and 5b were shown to be stable in the electroly-
sis medium, as well as during the work-up and chromato-
graphic separation. So (i) the low material balances (32% in
entry 2, 54% in entry 3, and 57% in entry 1) after chromato-
graphic separation and (ii) the fact that the higher the mate-
rial balance, the better the nucleophile suggests processes
giving molecules that are retained on silica gel must be com-
peting with path D and, to a lesser extent, path C
and (ii) the dehydration of unprotonated hydroxylamine 7b
to the diiminoquinonium cation 4b occurs by elimination of
OH . Such elimination would be expected to be slower than
–
in acidic medium (acid-catalyzed elimination of H O from
2
2b (see path B in Scheme 1)). Assuming that the low mate-
rial balance after chromatographic separation (36%) is due
to polymerization and (or) oligomerization of the diimino-
quinonium cation 4b, which occurs also in basic medium
(the crude material balance is again close to 100%), the
(
(
(
Scheme 1). One probable hypothesis is polymerization and
or) oligomerization of the diiminoquinonic intermediate 4b
–
elimination of OH from hydroxylamine 7b would be faster
polymerization and (or) oligomerization of the diprotonated
than the condensation between 7b and its nitroso precursor
(structure not shown), a well-documented reaction in the
electroreduction of arylnitro groups (3) in basic medium;
rate of elimination : rate of condensation ~ 73:27. The
hydroxylamine–nitroso coupling gives the azoxy derivative
first, which is reduced to the azo compound 11b at the elec-
trolysis potential (3). Azo compounds are more easily reduc-
ible than azoxy compounds (3), and 11b is immediately
reduced to the hydrazo derivative during the electrolysis.
The latter is rapidly oxidized back to the azo compound 11b
in the presence of oxygen during the basic work-up (3), and
the azo derivative 11b was isolated in 27% yield. The
hydroxylamine–nitroso condensation is much slower in
acidic medium (3) and did not occur during the electro-
reduction of 1b under acidic conditions discussed above. It
was not observed either during the electroreduction of 5-
nitroindole (1a) in basic medium (2) because the dehydra-
tion of the unprotonated hydroxylamine 7a, being base cata-
+
form 4bH appears less likely). The polymers and (or) oli-
gomers formed would be strongly adsorbed on silica gel,
and this would explain why all attempts to isolate any poly-
meric and (or) oligomeric material failed. In the presence of
–
a good nucleophile (Br or PhSH, entries 1 and 2), polymer-
ization and (or) oligomerization would account for about
4
0% of the products, and in methanol only (entry 2), it
would be the main reaction (about 70% of the products).
+
The formation of 4b and (or) 4bH was confirmed beyond
any doubt by the isolation of the tetracyclic compound 10b
(
R = CH ) in 31% yield (Scheme 2) when electrolysis of 1b
3
–
1
in acidic medium (H SO 0.15 mol L ) was carried out in
2
4
the presence of cyclopentadiene in excess (8 equiv). The for-
mation of 10a has been previously reported for the electroly-
sis of 1a under the same conditions (2). Adducts 10 most
6
4
probably result from a [π s + π s] cycloaddition reaction be-
tween the fulvene part of the diminoquinonic intermediate(s)
and cyclopentadiene, as has already been suggested (2).
–
lyzed (abstraction of the proton at position 1 by OH ), is
6
4
Fulvenes are known to undergo [π s + π s] cycloadditions
much faster than the nitroso–hydroxylamine coupling reac-
tion. Finally, the low yield of 1-methyl-5-aminoindole (3b)
and of 1-methyl-4-methoxy-5-aminoindole (5b) shows that
polymerization and (or) oligomerization of the diiminoquin-
onium cation 4b (about 64% of the reaction products) is
faster than its reaction with nucleophiles (path D of
Scheme 1) and its reduction (path C) (9% of reaction prod-
ucts).
with cyclic 1,3-dienes (4, 5).
Pre pa ra tive e le c trolys e s of 5-
nitrobe nzofura ne (1c )
In basic medium (Table 2, entry 4), the electroreduction of
1
-methyl-5-nitroindole (1b) gave 2% of 1-methyl-5-
The results of preparative electroreduction of 5-
aminoindole 3b, 7% of 4-methoxy-5-aminoindole (5b, Nu =
MeO), and 27% of the 5,5′-azo derivative 11b. The forma-
tion of 3b and 5b (Nu = MeO) is again explained by the
mechanistic hypothesis of Scheme 1 with the following dif-
ference: the intermediate hydroxylamine is not protonated
nitrobenzofurane (1c) at Hg, in MeOH–H O (95:5 volume
2
fraction) solutions, are presented in Table 3. In acidic me-
dium (entries 1 to 3), the electrochemical reduction of 5-
nitrobenzofurane (1c) gives 4-substituted-5-aminobenzo-
furanes (5c: Nu = Br, OCH , SPh) as the major products
3
(
see structure 7b). Therefore, (i) path A does not occur at all
(Scheme 1). They represent 70%–100% of the isolated prod-
ucts. The other products isolated are 5-aminobenzofurane
since unprotonated arylhydroxylamines are not reducible (3)
©
2003 NRC Canada

1
112
Can. J. Chem. Vol. 81, 2003
Table 3. Electroreduction of 5-nitrobenzofurane (1c) (6 × 10–3 mol L–1 solution) at Hg in MeOH–H O (95:5 volume fraction).
2
Yield of products (%)a
Entry
1
Supporting electrolyteb
HBr
Added nucleophilec
—
EW (V)d
Charge (F mol–1)
4.1
3c
5c
–0.67
6
63–69 (Nu = Br)
e
1
1 (Nu = MeO)
f
2
3
H SO4
—
PhSH
–0.61
–0.61
4.0
3.8
—
18–27
66–79 (Nu = MeO)
46–55 (Nu = MeO)
2
H SO4
2
4
—
(Nu = PhS)
g
4
KOH
—
–1.40
8.4
10
a
Yield of product isolated by flash chromatography on silica gel after N-acetylation (see the Experimental section) unless stated otherwise and
based on two to three electrolyses (range given when variations were larger than ±4%); see Scheme 1 for structures of 3c and 5c.
b
Supporting electrolyte concentration; see footnote b of Table 1.
c
–2
–1
Added nucleophile concentration: 1.2 × 10 mol L .
d
e
Working potential vs. SCE.
Trace amounts of trans-5-amino-2,3-dihydro-2,3-dimethoxybenzofurane (12).
f
5
-Amino-2,3-dihydro-2,3-dimethoxybenzofurane (12) was also isolated in an 11%–18% yield (ratio trans:cis = 1.5).
Azo derivative 11c was isolated in a 69% yield.
g
5
7%, 32%, and 53% in entries 1, 2, and 3, respectively) (the
Scheme 3.
material balance before separation was 100%, as in the
electroreduction of 1b). This difference in material balance
strongly suggests that the cationic intermediate 4c has a
lesser tendency to polymerize and (or) oligomerize and (or)
is more reactive towards nucleophiles (paths D of Scheme 1)
than the cationic intermediate 4b. The diprotonated interme-
diate 4cH+ would also be more reactive towards nucleo-
+
philes than the diprotonated intermediate 4bH . Moreover,
when a nucleophile other than methanol is also present in
the reaction medium (is competing with methanol), the rela-
tive amount of methoxy derivative (Nu = OCH ) in the 4-
3
substituted-5-amino derivatives 5 is higher in the benzo-
–
furane series c (17% (MeOH vs. Br ) in entry 1 of Table 3,
9
6% (MeOH vs. PhSH) in entry 3) than in the indole series
–
b (0% (MeOH vs. Br ) in entry 1 of Table 2, 23% (MeOH
vs. PhSH) in entry 3). This can be readily explained by the
+
fact that the cation 4c and dication 4cH would be harder
electrophiles than cation 4b and dication 4bH , oxygen be-
(
3c) (6% in entry 1, 18%–27% in entry 3) and cis- and
trans-2,3-dihydro-2,3-dimethoxy-5-aminobenzofurane (see
2 in Scheme 3) (trace amounts of trans isomer in entry 2
and 11%–18% of a mixture of cis and trans isomers
cis:trans = 0.66) in entry 3). The aminofuranes 3c and 5c
must be formed through the same mechanistic pathways
Scheme 1, X = O) as those discussed above for the electro-
+
ing more electronegative than nitrogen and MeOH being a
1
–
harder nucleophile than Br and PhSH for electronegativity
reasons as well. When the electrolysis of 5-nitrobenzofurane
(
(
1c) was carried out in acidic medium in the presence of
cyclopentadiene in excess (8 equiv), the cycloadduct 15 was
isolated in a low yield of 7% (Scheme 4). As seen above, the
yield of cycloadduct 10a was 32% for the electrolysis of 1-
methyl-5-nitroindole (1b) carried out under the same condi-
tions (Scheme 2). This difference in yield of cycloaddition is
readily explained again by the greater reactivity of cationic
(
reduction of 5-nitroindole (1a) (2) and 1-methyl-5-
nitroindole (1b) (Scheme 1, X = NH and NCH , respec-
tively). The dimethoxy derivatives 12 could have been
formed as described in Scheme 3 — addition of a first mole-
cule of methanol to the dication 4cH , followed by the addi-
tion of a second methanol molecule on the resulting dication
1
3
+
+
electrophiles 4c and (or) 4cH towards the hard nucleophile
methanol when compared with the reactivity of the softer
3 — which would be faster than deprotonation of 13 since
-methoxy-5-aminobenzofurane (14) was not detected. The
+
cationic electrophiles 4b and (or) 4bH . Cyclopentadiene is
3
–
1
a much softer nucleophile than methanol. It must be noted
charge of about 4 F mol in the first three entries of Table 2
correlates very well with the fact that the main products iso-
lated are 4-substituted-5-aminobenzofuranes (5c), the forma-
tion of which involves four electrons (Scheme 1, paths B
and D). These results also show that the acid-catalyzed de-
hydration of the protonated hydroxylamine 2c (path B) is
faster than its reduction to 3c (path A), as for the electro-
reduction of 1a and 1b. The material balance, after chro-
matographic separation, is higher with 5-nitrobenzofurane
+
that cations 4c and (or) 4cH act as dienophiles (2-electron
4
partner) at the 5,6 positions in a Diels–Alder reaction ([π s +
2
π s] cycloaddition) with cyclopentadiene (Scheme 4),
+
whereas cations 4b and (or) 4bH act as the 6-electron com-
6
4
ponent (fulvene moiety) in a [π s + π s] cycloaddition, as al-
ready seen (Scheme 2).
In basic medium (Table 3, entry 4), the electroreduction of
5-nitrobenzofurane (1c) gave the azo derivative 11c as the
main product (69%) and some 5-aminobenzofurane (3c)
(10%). Since unprotonated arylhydroxylamines are not re-
(
1c) (Table 3: 83%, 79%, and 78% in entries 1, 2, and 3, re-
spectively) than with 1-methyl-5-nitroindole (1b) (Table 2:
©
2003 NRC Canada

Bouchard et al.
1113
Scheme 4.
ducible in protic medium, as already mentioned (3), 5-
aminobenzofurane (3c) must have been formed by reduction
of the cationic intermediate 4c (path C of Scheme 1). No
(1c) and 5-nitrobenzothiophene (1d) is the high
regioselectivity of the addition of nucleophiles (i) to cationic
+
+
+
intermediates 4a–4aH (2), 4b–4bH , 4c–4cH , and 4d–
+
methoxy derivative such as 5c (Nu = OCH ), resulting from
4dH (this work) in acidic medium and (ii) to neutral inter-
3
the addition of a methoxide anion to 4c (path D), was ob-
served. This shows that, in basic medium, the reduction of
mediate 8 (2), as well as to cationic intermediates 4b, 4c,
and 4d in basic medium (this work). The addition occurs ex-
clusively at carbon 4 and not at the other electrophilic posi-
tions (positions 2, 3, 6, and 7) except (i) for the addition of
4
c is faster than its reaction with methoxide ions and (or)
–
methanol. Moreover, elimination of OH from hydroxyl-
amine 7c to give the cationic intermediate 4c is slower than
the hydroxylamine–nitroso coupling reaction, leading ulti-
mately to the isolation of azo derivative 11c. In the case of
+
methanol to 4c and (or) 4cH (Table 3, entry 2) in acidic
medium with a regioselectivity of about 83%, some addition
+
taking place at carbon 2 of 4cH (see Scheme 3) and (ii) for
–
hydroxylamine 7b, the elimination of OH to form 4b was
the addition of PhSH to cationic intermediates 4a and (or)
+
faster than the coupling reaction, as seen above. This differ-
ence in reactivity between 7c and 7b can be ascribed to the
fact that oxygen is more electronegative than nitrogen, the
cationic intermediate 4c being more difficult to form (having
a higher enthalpy) than intermediate 4b.
4aH with a regioselectivity of 62% (38% of 3-thiophenoxy
derivative 5a) (2) and to intermediates 4b and (or) 4bH+
with a regioselectivity of 50% (50% of 3-thiophenoxy deriv-
ative 5b) (Table 2, entry 3).
According to frontier molecular orbital theory, the most
important frontier orbital interaction in a reaction between
an electrophile and a nucleophile (5, 6) is that between the
LUMO of the electrophile and the HOMO of the nucleo-
phile. In the case of the addition of a nucleophile to a multi-
valent electrophile, the reaction would be the fastest at the
carbon where the coefficient of the LUMO is the largest (5,
6). The results of theoretical calculations using the AM1
method (7, 8) on intermediates 4 (Scheme 1) and their con-
Pre pa ra tive e le c trolys e s of 5-
nitrobe nzothiophe ne (1d)
The results of preparative electroreduction of 5-
nitrobenzothiophene (1d) under the same conditions as those
used for the electroreduction of the two other nitro com-
pounds, 1b and 1c, are presented in Table 4. Those results
are similar to those reported in Table 3 for the electro-
reduction of 5-nitrobenzofurane (1c) for both acidic and ba-
sic media (charge passed, yields, and relative proportion of
+
jugate acids 4H are gathered in Table 5. They show that the
LUMO of unprotonated intermediates 4 and protonated 4H+
have the largest coefficient on carbon 4, which agrees
entirely with the experimental results: the regioselective for-
mation of 4-substitued-5-amino derivatives 5 in acidic me-
dia. However, the largest coefficient of the LUMO of the
protonated intermediate 6 (reduction of 5-nitroindole (1a)) is
on carbon 6 (2). Although calculations give a slightly higher
gas-phase enthalpy of formation for cationic intermediate 4a
5
5
-amino product 3 and of 4-substituted-5-amino derivative
), except that no dimethoxy derivative analogous to 12 was
detected in the products of electroreduction of 5-
nitrobenzothiophene (1d) in acidic medium (compare entries
2
and 3 in Table 4 with entries 2 and 3, respectively, in Ta-
ble 3). Another slight difference in reactivity was noted
when the electroreduction of 5-nitrobenzothiophene (1d) in
acidic medium was carried out in the presence of cyclopen-
tadiene under the same conditions as those used with 5-
nitrobenzofurane (1c). No cycloadduct analogous to 15 (or
any other possible adduct) was detected (15 was isolated in a
low 7% yield from the electroreduction of 1c). So the mech-
anisms must be the same for the three nitro compounds (see
–
1
(257.6 kcal mol ) than for cationic intermediate 6
–
1
(255.4 kcal mol ) (2), the similarity of the experimental re-
sults obtained with 5-nitroindole (1a) (2) and 1-methyl-5-
nitroindole (1b) (the cationic intermediate 4b is the sole pos-
sible) and the calculated largest coefficient on carbon 4
strongly suggest the participation of intermediate 4a and (or)
+
4aH instead of intermediate 6 in the electroreduction of 1a.
Scheme 1), and the reactivities of the various intermediates
As for the low regioselectivity of substitution observed in
the electroreduction of 5-nitroindole (1a) (62%) and 1-
methyl-5-nitroindole (1b) (50%) in acidic medium in the
presence of thiophenol, it would be surprising that it would
be because of the fact that the addition of thiophenol to 4a–
+
(
2d, 4d and 4dH , 7d) must be comparable to those of the
related intermediates involved in the reduction of 5-
nitrobenzofurane (1c). This close similarity seems surprising
since sulphur is less electronegative and is a third-row ele-
ment.
+
+
4
aH and 4b–4bH would be much less regioselective than
the addition of the other nucleophiles (bromide ion and
methanol). One possibility is that the addition of thiophenol
Re gios e le c tivity of ring s ubs titution
+
would occur regioselectively at position 4 of 4a–4aH and
+
A very interesting aspect of the electroreduction of 5-
4b–4bH . Then there would be a competition between
nitroindoles 1a and 1b and of the related 5-nitrobenzofurane
deprotonation of the initial adduct 16 to give the 4-thio-
©
2003 NRC Canada

1
114
Can. J. Chem. Vol. 81, 2003
Table 4. Electroreduction of 5-nitrobenzothiophene (1d) (6 × 10–3 mol L ) at Hg in MeOH–H O (95:5 volume
–1
2
fraction).
Yield of products (%)a
Entry Supporting electrolyteb Added nucleophilec E (V)d
Charge (F mol–1) 3d
5d
1
HBr
—
–0.61
4.2
—
47 (Nu = Br)
1
2–22 (Nu = MeO)
2
3
H SO4
—
PhSH
–0.57
–0.57
4.1
4.9
—
70–80 (Nu = MeO)
2
H SO4
34–40 46–52 (Nu = MeO)
2
7
—
(Nu = PhS)
e
4
KOH
—
–1.19
8.5
—
a
Yield of product isolated by flash chromatography on silica gel (unless stated otherwise) and based on two to three electrolyses
(
range given when variations were larger than ±4%).
b
Supporting electrolyte concentration; see footnote b of Table 1.
Added nucleophile concentration: 1.2 × 10 mol L for PhSH.
Working potential vs. SCE.
c
–2
–1
d
e
Azo derivative 11d was isolated in a 78% yield.
Table 5. Heats of formation, coefficient of the LUMO, and Mulliken charges of intermediates 4 and 4H+.
Coefficient of the LUMO (charge)a
–
1
Intermediate
∆H (kCal mol )
C2
0.01 (–0.09)
–0.13 (–0.06)
0.10 (–0.08)
–0.11 (–0.04)
0.08 (–0.48)
0.03 (–0.55)
C3
C4
0.49 (–0.06)
0.34 (–0.05)
0.49 (–0.02)
–0.36 (–0.08)
0.50 (0.15)
C6
0.35 (–0.06)
–0.23 (–0.15)
0.41 (–0.00)
–0.30 (–0.16)
0.32 (–0.08)
0.18 (–0.18)
C7
f
4
4
4
4
4
4
b (X = NCH )
bH
259.5
496.9
237.8
491.3
252.3
490.6
–0.24 (–0.09)
0.21 (–0.07)
–0.16 (–0.20)
0.13 (–0.19)
–0.25 (–0.02)
–0.30 (0.06)
–0.13 (–0.14)
0.27 (–0.01)
–0.05 (–0.24)
0.22 (–0.11)
–0.07 (–0.11)
–0.22 (0.04)
3
+
c (X = O)
+
cH
d (X = S)
+
dH
0.39 (0.01)
a
C , C , etc. correspond to the numbering of carbons of indole, benzofurane, and benzothiophene (see 1, Scheme 1).
2
3
Scheme 5.
phenoxy derivative 5 (Nu = SPh) and rearrangement of 16
into adduct 17 as shown in Scheme 5 for the monocationic
intermediates 4. Deprotonation of 17 would give the 3-
thiophenoxy derivatives 9. Interestingly, the electroreduction
of 5-nitroindole (1a) in the presence of thiophenol in basic
medium was reported to give 59% of 4-thiophenoxy-5-
aminoindole (9a) as the sole substitution product (2), which
then means that the addition of a thiophenolate anion to
diiminoquinone 8 is 100% regioselective, as with the other
nucleophiles, and occurs solely at position 4. The AM1 cal-
culations once more agree with this result; the LUMO of the
neutral diiminoquinone 8 has the largest coefficient on car-
bon 4 (2).
Conc lus ion
In this paper, we have shown that the electroreduction of
-methyl-5-nitroindole (1b), 5-nitrobenzofurane (1c), and 5-
1
nitrobenzothiophene (1d) at Hg, in acidic hydromethanolic
media, gives the corresponding 4-substituted amino deriva-
tives 5b, 5c, and 5d in good yields and high regioselectivity.
The proposed mechanism (Scheme 1) involves first the
electroreduction of the nitro substrate 1 to the corresponding
protonated hydroxylamine 2, which is quickly dehydrated to
pericyclic reaction with cyclopentadiene in the electrolysis
of 1-methyl-5-nitroindole (1b) (Scheme 2), and 5-
nitrobenzofurane (1c) (Scheme 4) in acidic medium. The
material balance was always lower than 100% after chro-
matographic separation. This has been attributed to a compe-
tition between, on the one hand, reduction of the
intermediate quinonium species 4 and its reaction with
nucleophiles (steps C and D) and, on the other hand, their
polymerization and (or) oligomerization. The electro-
+
an electrophilic quinonic intermediate 4–4H (step B). The
latter reacts with nucleophiles present in the medium
(
step D) faster than it is reduced to the corresponding amino
compound 3 (step C). The mechanism was confirmed by un-
equivocal trapping of the intermediates diiminoquinonium
+
+
cation 4b–4bH and iminoquinonium cation 4c–4cH , via a
©
2003 NRC Canada

Bouchard et al.
1115
reduction of all of the nitro compounds 1 in basic medium
gives the corresponding azo derivative 11 as the major
product. In the case of 5-nitrobenzofurane (1c) and 5-
nitrobenzothiophene (1d), the hydroxylamine–nitroso
coupling reaction is faster than dehydration of the hydroxyl-
amine in basic medium. It is slower in the case of 1-methyl-
was considered more adequate than the PM3 method for cal-
culations on cationic species.
Substrate and authentic sample synthesis
All prepared compounds had spectral and physical charac-
teristics identical to those reported in the literature.
5
-nitroindole (1b), and the main process in basic medium is
polymerization and (or) oligomerization of 4b.
1-Methyl-5-nitroindole (1b) was prepared by alkylation of
commercially available 5-nitroindole (1a) with iodomethane
(
(
9). 5-Nitrobenzofurane (1c) was synthesized by nitration
10) and decarboxylation (11) of coumarillic acid, previ-
Expe rim e nta l
ously prepared by the method of Suzuki et al. (12). 5-
Nitrobenzothiophene (1d) was prepared by the method of
Zambias and Hammond (13).
General information
All commercially available chemicals were purchased
from Aldrich and BDH, and all the solvents were distilled
before use. Infrared spectra were recorded in chloroform on
a PerkinElmer 1600 FT-IR spectrometer, and nuclear mag-
netic resonance spectra were recorded in deuterochloroform
or deuteroacetone on a Bruker AC-300 spectrometer using
the signal from chloroform or acetone as reference (s = sin-
glet, d = doublet, t = triplet, q = quartet, dd = doublet of
doublets, m = multiplet). Vapor phase chromatography was
performed on a Hewlett-Packard (HP) 5710A chromato-
graph with an FID detector, an HP 3390A integrator, and a
DB-5 fused silica capillary column (30 m). High resolution
mass spectra were recorded on a ZAB-2F model VG appara-
tus. Flash chromatography was carried out on Silicycle Inc.
silica gel (70–230 mesh). Solutions for all electrochemical
experiments were deoxygenated by bubbling dry argon
through them for 15 min, and an argon atmosphere was
maintained throughout the experiments. The cyclic volt-
ammetry (CV) experiments were performed using a Prince-
1
-Methyl-5-aminoindole (3b) was synthesized by reduc-
tion of 1-methyl-5-nitroindole (1b) by NaBH S (14) and
was used as an authentic sample. To a suspension of freshly
prepared NaBH2S3 (15) (300 mg, 2.28 mmol) in THF
70 mL), a solution of 1-methyl-5-nitroindole (1b) (200 mg,
.14 mmol) in anhydrous THF was slowly added. The mix-
ture was refluxed for 24 h. The solvent was removed at room
temperature under reduced pressure, and a solid residue was
obtained. The crude material was hydrolyzed with HCl
2
3
(
1
–
1
(
1 mol L ) in the presence of diethyl ether. The solids were
filtered and rinsed with diethyl ether. The filtrate was
–
1
basified to pH 11 with NaOH (1 mol L ). The aqueous
phase was extracted with diethyl ether. The organic phase
was washed with brine, and the organic phases were com-
bined, dried (Na SO ), and filtrated. The solvent was evapo-
rated under vacuum, and a dark oil was obtained. After flash
chromatography on silica gel, 67 mg of 1-methyl-5-
aminoindole (3b) were obtained (40% yield) as a dark oil
lit. (15) mp = 100–101 °C). IR (CHCl ) (cm ) : 3005 (CH
aro), 2930 (CH ali), 1630 (C=C), 1497 (C=C), 1455 (CH ),
250 (C-N). H NMR (300 MHz, CDCl ) (ppm) δ: 7.17 (1H,
d, J = 8.6, H7), 7.14 (1H, d, J = 2.2, H4), 7.00 (1H, d, J =
.0, H2), 6.86 (1H, dd, J = 2.2, J = 8.6, H6), 6.34 (1H, d,
2
4
–
1
(
3
ton Applied Research PAR 273 potentiostat in a single
compartment cell with 7 mL of 0.25 mmol L–1 solutions of
3
1
1
3
the substrate in methanol–water (95:5 volume fraction) con-
taining the supporting electrolyte (see Table 1). No correc-
tion was applied to compensate for the solution resistance.
The voltammograms were recorded and processed using
EG&G software M270. The working electrode was a mer-
3
1
3
J = 3.0, H3), 3.73 (3H, s, CH3). C NMR (75 MHz, CDCl3)
ppm) δ: 139.1, 131.8, 129.1, 128.0, 112.5, 109.7, 105.7,
9.5, 32.8. HR-MS, m/z: calcd. for C H N : 146.0844;
(
9
9 10 2
cury hanging drop Metrohm, model E410. A glassy carbon
–
found: 146.0849.
rod was used as the counter electrode and an Ag/AgCl/(Cl )
dil
electrode, the potential of which was adjusted to that of an
SCE, served as the reference. The controlled potential elec-
trolyses were carried out on a mercury pool, in a vertical
glass cell with two concentric compartments separated by a
fritted glass, using an Electrosynthesis Company (ESC) 410
potentiostat with an ESC 640 coulometer and an ammeter
connected in series within the auxiliary circuit. The volume
of the cathodic compartment was 45 mL and that of the an-
odic compartment 10 mL. The same reference electrode as
the one described above was used, and a platinum grid was
used as the counter electrode. The other conditions are given
in Tables 2–4.
Typical electrolysis procedure
Electroreduction of 1-methyl-5-nitroindole (1b) in acidic
HBr medium (entry 1, Table 2). Argon was bubbled through
the cathodic compartment containing the electrolyte for 1 h,
and an argon atmosphere was maintained throughout the
electrolysis. A pre-electrolysis was performed at –0.54 V vs.
SCE until the current was 0.1 mA or less. The circuit was
opened, then 1b (48.9 mg, 0.27 mmol) was added to the
catholyte. The potential was applied, and the electrolysis
was stopped when the current had dropped to 0.1 mA or
less. The electrolysis solution was transferred to a round-
bottomed flask, the cell was washed with methanol and wa-
ter, and methanol was removed on a rotatory evaporator. The
pH of the resulting solution was brought to 12 with aqueous
Computational procedure
All calculations were performed at the RHF level using
the AM1 method (7, 8). The first input files for MOPAC
–
1
NaOH (1 mol L ); the solution was saturated with NaCl and
extracted with ether (4 × 50 mL). The organic phase was
washed with brine, dried (Na SO ), and the solvent removed
6
.00 were created by means of SYBIL 6.01 (Tripos Associ-
ates, Inc., St-Louis, Mo.) for IBM RS/6000 computers. The
gradients of the norm of these draft structures were then
fully optimized using the EF subroutine. The AM1 method
2
4
under vacuum. The crude materiel was purified by flash
chromatography on silica gel (elution with 20% volume
©
2003 NRC Canada

1
116
Can. J. Chem. Vol. 81, 2003
fraction AcOEt–hexanes) to give 35.4 mg of 5b (Nu = Br)
4-Bromo-5-acetamidobenzofurane (acetylated 5c, Nu =
Br)
(
57% yield).
–
1
The crude product from the electroreduction of 5-
Beige solid. mp = 160–161 °C. IR (CHCl ) (cm ): 3416
3
nitrobenzofurane (1c) was acetylated by acetic anhydride in
ethyl acetate before chromatographic separation, and the
amino compounds were isolated as their N-acetyl derivative.
The acetylation was quantitative.
(NH), 3010 (CH aro.), 1693 (C=O), 1518 (NH), 1432 (CH ),
3
1
1252 (C-O), 1211 (C-N). H NMR (300 MHz, acetone-d )
6
(ppm) δ: 8.66 (1H, m, NH), 7.94 (1H, d, J = 2.2, H2), 7.81
(1H, d, J = 8.8, H4), 7.52 (1H, d, J = 8.8, H6), 6.87 (1H, d,
1
3
J = 2.2, H3), 2.08 (3H, s, CH3). C NMR (acetone-d6)
ppm) δ: 168.4 (C=O), 146.1, 119.4, 110.8, 107.2, 151.2,
31.1, 129.0, 105.6, 24.6. HR-MS, m/z: calcd. for
C H NO Br: 252.9738; found: 252.9743.
(
1
Characterization of electrolysis products
5
-Acetamidobenzofurane (acetylated 3c)
10
8
2
–
1
Dark oil. IR (CHCl ) (cm ): 3438 (NH), 3011 (CH aro.),
3
1
4-Methoxy-5-acetamidobenzofurane (acetylated 5c, Nu =
MeO)
1
684 (C=O), 1524 (C=C), 1215 (C-N). H NMR (300 MHz,
CDCl ) (ppm) δ: 7.89 (1H, d, J = 2.1, H4), 7.63 (1H, m,
3
–
1
Dark oil. IR (CHCl ) (cm ): 3426 (NH), 3320 (NH),
NH), 7.59 (1H, d, J = 1.9, H2), 7.40 (1H, d, J = 8.8, H7),
3
3
006 (CH aro.), 2947 (CH ali.), 2838 (CH ), 1677 (C=O),
7
.21 (1H, dd, J = 2.3, J = 6.6, H6), 6.70 (1H, d, J = 2.1,
2
1
1
3
1597 (C=C), 1544 (NH), 1254 (C-O). H NMR (300 MHz,
H3), 2.03 (3H, s, CH3). C NMR (75 MHz, CDCl ) (ppm)
3
CDCl ) (ppm) δ: 8.20 (1H, d, J = 8.9, H7), 7.75 (1H, m,
δ: 168.6, 145.8, 117.7, 113.2, 111.3, 106.8, 152.0, 133.0,
3
NH), 7.52 (1H, d, J = 2.3, H2), 7.15 (1H, d, J = 8.9, H6),
1
27.8, 24.4. HR-MS, m/z: calcd. for C H NO : 175.0633;
10 9 2
6
.87 (1H, d, J = 2.3, H3), 4.08 (3H, s, O-CH ), 2.08 (3H, s,
found: 175.0636.
3
1
3
CH3). C NMR (75 MHz, CDCl ) (ppm) δ: 168.2, 144.4,
3
1
18.0, 115.3, 103.4, 153.0, 141.8, 123.7, 117.1, 59.9, 24.6.
5
-Aminobenzothiophene (3d)
Dark oil. IR (acetone) (cm ): 3369 (NH), 3075 (CH aro.),
667 (C=C), 1608 (C=C), 1491 (C=C). H NMR (300 MHz,
–
1
HR-MS, m/z: calcd. for C11H11NO3: 205.0739; found:
205.0742.
1
1
3
4
-Thiophenoxy-5-acetamidobenzofurane (acetylated 5c,
Nu = PhS)
Dark oil. IR (CHCl ) (cm ): 3476 (R N-H), 3367 (R N-
–
1
3
2
2
H), 3009 (C-H arom.), 1679 (C=O amide), 1613, 1582
1
(
C=C), 1215 (C-N). H NMR (300 MHz, CDCl ) (ppm) δ:
3
8
.56 (1H, m, NH), 8.24 (1H, d, J = 8.6 Hz, H ), 7.45 (1H, d,
7
J = 8.8 Hz, H ), 7.20 (1H, M, H ), 7.16 (2H, d, J = 7.7 Hz,
6
2
1
-Methyl-4-bromo-5-aminoindole (5b, Nu = Br)
H phenyl), 7.05 (3H, m, H phenyl), 6.58 (1H, M, H ), 2.09
–
1
3
Yellow oil. IR (CHCl ) (cm ): 3007 (CH aro.), 2947 (CH
13
3
(
1
1
3H, s, CH -). C NMR (75 MHz, CDCl ) (ppm) δ:
1
3
3
ali.), 1495 (C=C), 1440 (CH ), 1254 (C-N). H NMR
3
68.4 136.0, 133.9, 132.7, 131.6, 129.1, 126.6, 126.0, 125.6,
16.2, 113.2, 102.7, 24.7. HR-MS, m/z: calcd. for
(
(
300 MHz, CDCl ) (ppm) δ: 7.10 (1H, d, J = 8.5, H7), 7.01
1H, d, J = 3.0, H2), 6.75 (1H, d, J = 8.5, H6), 6.37 (1H, d,
3
C H NO S: 282.0827; found: 282.0833.
1
3
16 13
2
J = 3.0, H3), 3.73 (3H, s, CH3). C NMR (75 MHz, CDCl3)
ppm) δ: 136.8, 131.3, 129.6, 129.3, 112.2, 109.1, 100.2,
(
4
-Bromo-5-aminobenzothiophene (5d, Nu = Br)
9
9.9, 33.0. HR-MS, m/z: calcd. for C H N Br: 223.9949;
–1
9
9
2
Dark oil. IR (acetone) (cm ): 3464 (NH), 3001 (CH aro.),
found: 223.9945.
1
1
623 (C=C), 1502 (C=C). H NMR (300 MHz, CD OD)
3
(
ppm) δ: 7.57 (1H, d, J = 8.7, H7), 7.54 (1H, d, J = 5.7, H2),
1
-Methyl-4-methoxy-5-aminoindole (5b, Nu = MeO)
13
7
.27 (1H, d, J = 5.5, H3), 6.92 (1H, d, J = 8.6, H6).
C
–
1
Dark oil. IR (CHCl ) (cm ): 2930 (CH ali.), 1501 (C=C),
3
NMR (75 MHz, CD OD) (ppm) δ: 143.5, 141.5, 131.0,
3
1
1
446 (CH ), 1265 (C-N). H NMR (300 MHz, CDCl3)
3
128.8, 124.4, 122.8, 116.2, 102.6. HR-MS, m/z: calcd. for
C H NSBr: 226.9404; found: 226.9409.
(
ppm) δ: 6.94 (1H, d, J = 3.1, H2), 6.90 (1H, d, J = 8.4, H7
8
6
or H6), 6.76 (1H, d, J = 8.4, H6 or H7), 6.45 (1H, d, J = 3.1,
H3), 4.02 (3H, s, O-CH ), 3.72 (3H, s, N-CH ). C NMR
1
3
3
3
4-Methoxy-5-aminobenzothiophene (5d, Nu = MeO)
–
1
(
75 MHz, CDCl ) (ppm) δ: 135.7, 133.3, 129.9, 128.4,
3
Dark oil. IR (acetone) (cm ): 3459 (NH), 3005 (CH aro.),
1
1
13.3, 112.1 104.8, 96.8, 58.5, 33.0. HR-MS, m/z: calcd. for
2
937 (CH ali.), 1620 (C=C), 1509 (C=C), 1454 (CH ). H
3
C H N O: 176.0950; found: 176.0953.
1
0
12
2
NMR (300 MHz, CD OD) (ppm) δ: 7.44 (1H, d, J = 5.5,
3
H2), 7.39 (1H, d, J = 8.5, H7), 7.31 (1H, d, J = 5.5, H3),
6.92 (1H, d, J = 8.6, H6), 4.91 (3H, s, CH3). 13C NMR
1
-Methyl-4-thiophenoxy-5-aminoindole (5b, Nu = PhS)
–
1
Dark oil. IR (CHCl ) (cm ): 3001 (CH aro.), 2946 (CH
(75 MHz, CD OD) (ppm) δ: 143.0, 137.0, 135.5, 132.9,
3
3
1
ali.), 1613 (C=C), 1505 (C=C), 1253 (C-N). H NMR
300 MHz, CDCl ) (ppm) δ: 7.25 (1H, d, J = 8.6, H6), 7.18–
127.5, 120.4, 119.2, 117.4, 60.8. HR-MS, m/z: calcd. for
(
C H NOS: 179.0405; found: 179.0399.
3
9
9
7
.14 (2H, m, H phenyl), 7.07–7.03 (3H, m, H phenyl), 6.98
(
1H, d, J = 3.0, H2), 6.82 (1H, d, J = 8.6, H7), 6.44 (1H, d,
4-Thiophenoxy-5-aminobenzothiophene (5d, Nu = PhS)
1
3
–1
J = 3.0, H3), 3.76 (3H, s, CH3). C NMR (75 MHz, CDCl3)
ppm) δ: 142.8, 137.1, 129.6, 128.8, 128.5, 127.9, 127.3,
26.0, 124.8, 112.2, 111.9, 100.0, 33.0 (CH ). HR-MS, m/z:
Dark oil. IR (acetone) (cm ): 3494 (NH), 3004 (CH aro.),
1
(
1
1616 (C=C). H NMR (300 MHz, CD OD) (ppm) δ: 7.71
3
(1H, d, J = 8.6, H7), 7.46 (1H, d, J = 5.5, H2), 7.34 (1H, d,
J = 5.5, H3), 7.17–7.12 (2H, m, H phenyl), 7.07–6.95
3
calcd. for C H N S: 254.0878; found: 254.0884.
1
5
14
2
©
2003 NRC Canada

Bouchard et al.
1117
(
(
4H, m, H6 and H phenyl). 13C NMR (75 MHz, CD OD)
ppm) δ: 149.4, 145.9, 138.2, 130.8, 129.9, 128.8, 127.2,
26.2, 125.8, 124.3, 115.8, 106.9. HR-MS, m/z: calcd. for
trans-2,3-Dimethoxy-2,3-dihydro-5-acetamidobenzofurane
(trans 12)
3
–
1
1
Dark oil. IR (CHCl ) (cm ): 3437 (NH), 3011 (CH aro.),
3
1
C H NS : 257.0333; found: 257.0339.
1684 (C=O), 1490 (C=C), 1230 (C-N), 1194 (C-O).
H
1
4
11
2
NMR (300 MHz, CDCl ) (ppm) δ: 7.67 (1H, d, J = 2.3, H4),
3
7
.25 (1H, dd, J = 2.3, J = 8.6, H6), 7.08 (1H, broad peak,
1
-Methyl-3-thiophenoxy-5-aminoindole (9b)
NH), 6.83 (1H, d, J = 8.6, H7), 5.41 (1H, s, H2), 4.64
–
1
Dark oil. IR (CHCl ) (cm ): 3004 (CH aro.), 2974 (CH
3
(
(
1H, s, H3), 3.54 (3H, s, OCH ), 3.43 (3H, s, OCH ), 2.15
3 3
3H, s, CH3). C NMR (75 MHz, CDCl ) (ppm) δ: 168.3,
1
ali.), 1627 (C=C), 1493 (CH ), 1241 (C-N). H NMR
13
3
3
(
300 MHz, CDCl ) (ppm) δ: 7.24 (1H, s, H2), 7.19 (1H, d,
3
1
5
56.5, 144.8, 131.5, 123.3, 119.2, 111.2, 110.8, 85.1, 56.4,
J = 8.6, H7), 7.15–7.09 (2H, m,, H phenyl), 7.07–7.03
3H, m, H phenyl), 6.93 (1H, d, J = 2.2, H4), 6.78 (1H, dd,
6.2, 24.4. HR-MS, m/z: calcd. for C H NO : 237.1001;
1
2
15
4
(
found: 237.1005.
1
3
J = 2.2, J = 8.6, H6), 3.78 (3H, s, CH3). C NMR (75 MHz,
3
1
Compound 10b (cyclopentadiene cycloadduct)
NMR (300 MHz, CDCl ) (ppm) δ: 7.55 (1H, d, J = 2.2, H4),
3
–
1
Dark oil. IR (CHCl ) (cm ): 3431 (NH), 3359 (NH),
7.00 (1H, dd, J = 2.2, J = 8.5, H6), 7.12 (1H, broad peak,
NH), 6.77 (1H, d, J = 8.5, H7), 5.48 (1H, d, J = 5.2, H2),
4.87 (1H, d, J = 5.2, H3), 3.63 (3H, s, OCH3), 3.52 (3H, s,
3
3
1
001 (CH aro.), 2946 (CH ali.), 1506 (C=C), 1456 (CH ),
3
1
256 (C-N). H NMR (300 MHz, CDCl ) (ppm) δ: 6.92 (1H,
3
1
3
d, J = 8.4, H7), 6.66 (1H, s, H2), 6.63 (1H, d, J = 8.4, H6),
.95–5.92 (2H, m, H9 and H10), 3.81–3.77 (1H, m, H11 or
H8), 3.71–3.69 (1H, m, H8 or H11), 3.65 (3H, s, CH ),
OCH3), 2.14 (3H, s, CH3). C NMR (75 MHz, CDCl3)
(ppm) δ: 168.2, 154.1, 131.5, 126.0, 123.1, 118.9, 110.5,
106.8, 79.9, 58.2, 57.2, 29.7. HR-MS, m/z: calcd. for
C12H15NO4: 237.1001; found: 237.1003.
5
3
1
3
2
.48–2.41 (1H, m, H12b), 1.99 (1H, d, J = 9.8, H12a).
C
NMR (75 MHz, CDCl ) (ppm) δ: 130.8, 132.3, 130.5, 124.8,
3
Acetylated compound 15 (acetylated cycloaddition
product)
1
21.1, 117.1, 113.9, 112.7, 108.2, 39.4, 38.7, 44.1, 32.7.
HR-MS, m/z: calcd. for C H N : 210.1157; found:
1
4
14
2
–
1
Dark oil. IR (CHCl ) (cm ): 3001 (CH aro.), 1654
2
10.1299.
3
1
(
C=O), 1468 (CH ), 1437 (CH ). H NMR (300 MHz,
2
3
CDCl ) (ppm) δ: 8.25 (1H, m, NH), 7.59 (1H, d, J = 2.3,
H2), 7.30 (1H, d, J = 9.3, H4), 6.82 (1H, d, J = 2.3, H3),
5.82 (1H, dd, J = 2.4, J = 5.5, H6), 5.16 (1H, m, H7), 3.79
3
Azo compound 11b
–
1
Yellow solid. mp = 196–198 °C. IR (CHCl ) (cm ): 3007
CH aro), 2946 (CH ali), 1613 (C=C), 1490 (C=C), 1449
3
(
(
2H, m, H5 and H8), 2.45 (3H, s, CH ), 2.30 (2H, m, H9).
C NMR (75 MHz, CDCl ) (ppm) δ: 170.3, 150.7, 130.5,
3
3
1
(
8
1
CH ), 1237 (C-N). H NMR (300 MHz, CDCl ) (ppm) δ:
.25 (2H, d, J = 1.7, H4 and H4′), 7.94 (2H, dd, J = 8.3, J =
.7, H6 and H6′), 7.40 (2H, d, J = 8.8, H7 and H7′), 7.11
13
3
3
1
4
2
25.2, 124.3, 116.1, 145.2, 139.3, 120.6, 108.9, 104.8, 60.2,
1.0, 37.3, 25.3. HR-MS, m/z: calcd. for C H NO :
1
5
13
2
(
2H, d, J = 3.1, H2 and H2′), 6.64 (2H, d, J = 3.1, H3 and
39.0946; found: 239.0951.
1
3
H3′), 3.85 (6H, s, CH3). C NMR (75 MHz, CDCl ) (ppm)
3
δ: 130.0, 118.3, 115.3, 109.5, 102.9, 147.1, 137.9, 128.5,
3
2
3.2. HR-MS, m/z: calcd. for C H N : 288.1375; found:
88.1371.
Ac knowle dge m e nts
1
8
16
4
Financial support from the Natural Sciences and Engi-
neering Research Council of Canada (NSERC) and the
Fonds FCAR du Québec is gratefully acknowledged.
Azo compound 11c
–
1
Orange solid. mp = 160–162 °C. IR (CHCl ) (cm ): 3020
3
1
(
(
CH aro), 1454 (C=C), 1269 (C-O), 1095 (C-O). H NMR
Re fe re nc e s
300 MHz, CDCl ) (ppm) δ: 8.09–7.93 (6H, m, H2, H2′, H4,
3
H4′, H7 and H7′), 7.71 (2H, d, J = 8.8, H6 and H6′), 7.09
1
. J.M. Chapuzet, R. Labrecque, M. Lavoie, E. Martel, and J.
Lessard. J. Chim. Phys. 93, 601 (1996).
(
(
2H, d, J = 2.2, H3 and H3′). 13C NMR (75 MHz, CDCl3)
ppm) δ: 146.9, 119.3, 117.2, 112.1, 107.8, 156.7, 149.4,
2. I. Marcotte, M. Lavoie, Y. Dory, J.M. Chapuzet, and J.
Lessard. Acta Chem. Scand. 53, 849 (1999).
1
28.4. HR-MS, m/z: calcd. for C H N O : 262.0742;
16 10 2 2
found: 262.0747.
3. H. Lund and M.M. Baizer. In Organic electrochemistry: An in-
troduction and a guide. Edited by M.M. Baizer and H. Lund.
Marcel Dekker, NewYork. 1990. pp. 412–419.
Azo compound 11d
4
. K.N. Houk and L.J. Luskus. J. Org. Chem. 38, 3836 (1972).
–
1
Orange solid. mp = 165–166 °C. IR (CHCl ) (cm ): 3018
CH aro), 1463 (C=C). H NMR (300 MHz, CDCl ) (ppm)
3
5. I. Flemming. Frontier orbitals and organic chemical reactions.
Wiley, London. 1976.
1
(
3
δ: 8.42 (2H, s, H4 and H4′), 8.10–7.95 (4H, m, H2, H2′, H7
6
. N. Trong Anh. Orbitales frontières: Manuel pratique.
InterEditions and CNRS Éditions, Paris. 1995.
1
3
and H7′), 7.58–7.49 (4H, m, H3, H3′, H6 and H6′). C NMR
75 MHz, CDCl ) (ppm) δ: 150.2, 140.2, 131.8, 129.7,
(
1
7. M.J.S. Dewar, E.G. Zoebish, E.F. Healy, and J.J.P. Stewart. J.
Am. Chem. Soc. 107, 3902 (1985).
3
24.9, 123.0, 120.3, 117.4. HR-MS, m/z: calcd. for
C H N S : 294.0285; found: 294.0290.
8. J.J.P. Stewart. J. Comput. Chem. 10, 209 (1989); 10, 221 (1989).
1
6
10
2 2
©
2003 NRC Canada

1
118
Can. J. Chem. Vol. 81, 2003
9
. G. Bartoli, M. Bosco, R. Dalpozzo, and P.E. Todesco. J. Org.
Chem. 51, 3694 (1986).
13. R.A. Zambias and M.L. Hammond. Synth. Comm. 21, 959
(1991).
1
1
1
0. G. Lamotte, P. Demerseman, and R. Royer. J. Heterocyclic
Chem. 15, 1373 (1978).
1. F.G. Bordwell and C.J. Albisetti. J. Am. Chem. Soc. 70, 1948
14. J.M. Lalancette and J.R Brindle. Can. J. Chem. 49, 2990 (1971).
15. J.M. Lalancette, A. Frêche, and R. Monteux. Can. J. Chem.
46, 2754 (1968).
(
1955).
2. T. Suzuki, T. Horayuchi, and T. Shimizu. Bull. Chem. Soc.
Jpn. 56, 2726 (1983).
©
2003 NRC Canada