A tandem highly stereoselective FeCl3-promoted synthesis of a bisindoline: Synthetic utility of radical cations in heterocyclic construction

Article abstract of DOI:10.1016/j.tet.2004.10.002

A conceptually distinctive stereoselective construction of the novel dimer, N-[N′-acetyl-7,7′-bis-(3,4-dimethoxy-phenyl)-7,8,7′,8′- tetrahydro-N′H-[8,8′]biindolyl-N-yl]-ethanone 25 (bisindoline) is described below. These structures, which include 7-(3,4-dimethoxyphenyl)- indoline 24 (veratryl indoline), were obtained by the tactical combination of palladium-catalysed coupling which produced 10-acetamido-3,4-dimethoxystilbene 9, followed by FeCl₃ induced oxidative cyclization/dimerization. All new structures were fully characterized by 1- and 2D NMR spectroscopy, (proton, carbon-13, COSY, HMBC, HMQC) and mass spectrometry. Configurational assignments were further supported by semi-empirical AM1 calculations. Mechanistic interpretations, consistent with our results, are discussed. Graphical Abstract

Full text of DOI:10.1016/j.tet.2004.10.002

Tetrahedron 60 (2004) 11733–11742
A tandem highly stereoselective FeCl -promoted synthesis of a
3
bisindoline: synthetic utility of radical cations in
heterocyclic construction
a,*
a
Noel F. Thomas, Saraswati S. Velu, Jean-Fr e´ d e´ ric F. Weber, K. C. Lee,
b
c,†
a
A. Hamid A. Hadi, Pascal Richomme, David Rondeau, Ibrahim Noorbatcha
d
d
e
a
and Khalijah Awang
a
Department of Chemistry, Faculty of Science, University Malaya, 59100 Kuala Lumpur, Malaysia
iKUS, Faculty of Pharmacy, Universiti Teknologi MARA, 40450 Shah Alam, Selangor D.E., Malaysia
b
c
Department of Pharmacy, Faculty of Allied Health Sciences, Universiti Kebangsaan Malaysia, 50300 Kuala Lumpur, Malaysia
d
Facult e´ des Sciences, Universit e´ d’Angers, 2, Boulevard Lavoisier, 49045 Angers, France
Department of Biotechnology, Faculty of Science, International Islamic University, Jalan Gombak, 53100 Kuala Lumpur, Malaysia
e
Received 25 June 2004; revised 10 September 2004; accepted 1 October 2004
Available online 18 October 2004
0
0
Abstract—A conceptually distinctive stereoselective construction of the novel dimer, N-[N -acetyl-7,7 -bis-(3,4-dimethoxy-phenyl)-
7
0
0
0
0
,8,7 ,8 -tetrahydro-N H-[8,8 ]biindolyl-N-yl]-ethanone 25 (bisindoline) is described below. These structures, which include 7-(3,4-
dimethoxyphenyl)-indoline 24 (veratryl indoline), were obtained by the tactical combination of palladium-catalysed coupling which
produced 10-acetamido-3,4-dimethoxystilbene 9, followed by FeCl induced oxidative cyclization/dimerization. All new structures were
3
fully characterized by 1- and 2D NMR spectroscopy, (proton, carbon-13, COSY, HMBC, HMQC) and mass spectrometry. Configurational
assignments were further supported by semi-empirical AM1 calculations. Mechanistic interpretations, consistent with our results, are
discussed.
q 2004 Elsevier Ltd. All rights reserved.
1
. Introduction
construction by means of Heck coupling has recently been
7
the subject of some fascinating reviews. The compre-
,8
8
The importance of the indole ring has been demonstrated by
the fact that this skeleton, may be found in a variety of
biologically active natural products, for example the
hensive text by Li and Gribble is recommended reading.
From the vast literature relating to indole construction, a
few of these make use of ortho-iodoaniline as a starting
material. The palladium acetate-catalysed synthesis of
1
1,2
essential amino acid tryptophan, serotonin (a neuro-
transmitter in the brain), the hallucinogenic indole, psilocin
1
9
10
3-spiro-2-oxindoles and tryptamine are good examples.
With regard to bi-indole syntheses, by palladium-catalysed
3
and the alkaloids reserpine and strychnine. Of particular
4
5
significance is the ‘dimeric’ indole alkaloid vincristine
coupling, the synthesis of N-methylareyriacyanine by
1
5
leukaemia) and indomethacin (rheumatoid arthritis). With
1
(
Steglich has been reported. Syntheses of bis(indolyl)-
maleimides in 55% yield which exploited a phosphine-free
regards to the indolines, quite apart from oxido-reductive
connections with the indoles, other synthetic developments
have been reported.
12
palladium catalyst and caesium fluoride, have been
13
5
,6
0
described. In Fukuyama’s method a C-2 and C-2 linked
bisindole (biindole) was prepared and used in the synthesis
4
In 1954 Woodward et al. reported the total synthesis of
strychnine from veratryl indole, the latter being constructed
by Fischer indole methodology. In contrast, indole
14
of indolocarbazoles. This approach illustrates the syn-
thetic utility of isonitriles. However, Grigg’s synthesis of
0
the unsymmetrical 3,3 -biindole (Scheme 1), constructed by
1
5
the use of Pd(OAc) , PPh , (Me Sn) is noteworthy.
2
3
3
2,
Keywords: Ferric chloride; Radical cation; Bisindole; Stereospecific;
Stilbene.
This type of structural entity is relatively rare. We have
highlighted the Grigg dimer because (i) it contains a C-3/C-
3 link; (ii) one stereogenic centre is generated; and (iii) the
molecule lacks C-2 substituents.
*
Corresponding author. Tel.: C60 3 7967 4065; fax: C60 3 7967 4193;
e-mail: noelfthomas@um.edu.my
0
†
Present address: Department of Chemistry, Faculty of Sciences, National
University of Singapore, Singapore 117543.
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.10.002

11734
N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
1
9
cations) generated from stilbenes. Previously we reported
the manganese triacetate-mediated oxidative lactonization
for the regio-controlled syntheses of g-butyrolactones from
stilbenes, in which benzylic radicals were invoked. In a
2
0
subsequent paper, we described the FeCl -promoted
3
tandem pericyclic synthesis of catechol analogues of
restrytisol (Scheme 2). In that paper, we suggested that
the crucial reactive intermediate was the radical cation 2.
0
Scheme 1. The Grigg synthesis of the unsymmetrical 3 ,3-biindole.
The relevance of the above to our synthetic studies in
relation to indoles, indolines and the corresponding dimers
possessing C-2 substituents, will become clear during the
course of our discussion. We have found that such
substituents can exert profound stereo-directing effects.
We reasoned that, in a departure from previously reported
approaches, indole construction could be realized from
suitably functionalized stilbenes, in which the benzylic
cation was trapped intramolecularly by a nucleophile. Our
hypothesis was that cyclization would produce the indole,
provided the stilbene possessed an amino group at the ortho
position (Scheme 3).
1
6
With regard to C-2 substituted indoles, Larock’s elegant
one-step synthesis (oxidative addition), starting from
protected ortho-iodoaniline, is a substantial achievement.
Many syntheses along these lines have been described using
In order to translate the above plan into reality, the reactions
described in the next section were performed.
16–18
internal alkynes and dienes as acceptor molecules. The
pyrrole-construction step’ in these syntheses is often
‘
mediated by palladium reagents, for example, Pd(OAc)2.
2. Results and discussion
0
10 in dry DMF with sodium hydride and acetic anhydride to
In contrast to the above, we have discovered a conceptually
different approach exploiting benzylic radicals (or radical
2 -Iodoacetanilide 7 was prepared by reacting iodoaniline
3
Scheme 2. The FeCl -promoted tandem pericyclic synthesis of catechol analogues of restrytisol.
3
Scheme 3. Proposed indole construction by FeCl induced oxidative cyclization of the amino stilbene.
Scheme 4. Synthesis of amino- and acetamidostilbenes from iodoaniline and iodoacetamide.

N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
11735
Scheme 5. The synthesis of 2-amino-5-hydroxystilbene via Heck coupling.
produce 7 as brown crystals in 79%. The styrene was
prepared by reacting methyltriphenylphosphonium iodide in
THF with potassium tert-butoxide, to generate the ylide
followed by addition of 3,4-dimethoxybenzaldehyde 11,
which produced 8 in 60% yield as a yellowish oil. Stilbene 9
was prepared by treatment of ortho-iodoacetanilide 7 with
mixture, and our previous experience of stilbene-Mn(OAc)
3
reactions (Scheme 7), suggested to us the mechanistic
hypothesis described in Scheme 9.
1
9
In contrast to Scheme 3, we have found that treatment of the
acetamidostilbene 9 with ferric chloride in dichloromethane
produced the indoline 24 in 38% yield and the bisindole 25
in 15% yield (Scheme 9). In the case of 25 only one
diastereoisomer was obtained. The problem of the relative
configuration at the four contiguous stereogenic centres,
3
,4-dimethoxystyrene 8 in the presence of Pd(OAc) , Et N
2 3
in DMF. This Heck reaction produced the acetamidostilbene
in 24% yield. The aminostilbene 12 was generated in an
9
analogous manner. In accordance with the plan (Scheme 4)
we prepared stilbenes 9 and 12 in three and two steps,
respectively.
0
0
H-7, H-8, H-8 and H-7 and how the assignments were
made will be discussed later. In this transformation, the
formation of the bisindoline is highly stereoselective. These
compounds were isolated by preparative TLC, solvent
2
1
Ziegler and co-workers have reported that highly activated
-amino-5-hydroxyiodobenzene and styrene react via the
system hexane:ethyl acetate (7:3, v/v); R for the indoline 24
f
2
and for bisindoline 25 were 0.4 and 0.27, respectively. With
regard to the brownish yellow base line material, all
attempts to move it up the plate by addition of methanol
to the solvent system were unsuccessful. We think it
unlikely that these materials contain any bisindoline
diastereoisomers. These baseline products are much more
likely to be the result of polymerisation and/or degradation.
On this basis we say that this transformation leading to the
formation of 25c, is apparently stereospecific since no other
diastereomers are formed. However, we believe highly
stereoselective is a better description because the stereo-
chemical outcome (on the product side) cannot be directly
related to configurational/stereochemical features in the
Heck coupling reaction to form 2-amino-5-hydroxystilbene
in 50% yield (Scheme 5).
Treatment of aminostilbene 12 with manganese triacetate in
acetic anhydride/acetic acid produced the indole 6, which
was isolated in 13% yield as the major component from the
complex mixture (Scheme 6). The pyrrole NH is acetylated
under the reaction conditions. The polar baseline materials,
as shown on TLC, were not investigated further at this stage.
(All indole structures described in this manuscript, for
example, 6, are numbered according to the stilbene
numbering system; see Scheme 6).
2
3
starting stilbene. The origin of this stereoselectivity lies in
radical cation dimerization mechanistic pathways (and the
stereo-directing influence of C-2 aryl substituents). This will
be more fully explained in Schemes 11(a)–(d) and 12 and
Table 1. This reaction may actually be stereospecific in spite
2
3
of the above (Scheme 8).
2.1. Spectroscopic evidence for synthesised compounds
Scheme 6. Indole synthesis by manganese triacetate oxidative cyclization.
We have previously reported that under the same reaction
conditions (Scheme 7) the stilbene 1 was converted to the
g-butyrolactone 12.
2.1.1. Indole 6. The UV spectrum of indole 6 showed
absorption maxima at 293 and 213 nm (log 3 3.90 and 4.14,
respectively) typical of an indole chromophore. High
resolution EI mass spectrum gave a molecular ion peak
with accurate mass 295.1209 compatible with the molecular
formula C H O N (calculated mass 295.1208, D
1
9
1
8
17
3
K1
0.34 ppm). The IR spectrum showed a band at 1702 cm
representing the C]O stretch. No N–H bands were
1
observed, suggesting that this N was tertiary. The
H
NMR spectrum of 6 integrated for 17 protons. At d2.07 a
singlet corresponding to the methyl protons of the acetyl
was observed. The singlets of the methyls at C-3 and C-4
appeared at d 3.87 and d 3.90, respectively. H-8 resonated as
a singlet at d6.56. The signals of the seven aromatic ring
protons were observed between d 6.89 and d 8.32. The
chemical shift of each signal in the aromatic region is
similar to corresponding protons in the starting amino-
stilbene, except for H-11, which was deshielded due to its
Scheme 7. Synthesis of the g-butyrolactone.
We are aware of a report of the construction of a C-2 phenyl
substituted indole from an ortho-nitrostilbene by Akazome
et al.
22
The low yield of the indole 6, the complexity of the reaction

11736
N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
Table 1. The four possible stereoisomers of bisindoline 25a–25d
Stereoisomer
Symmetry element
Heat of formation (kcal/mol)
Dihedral angle H-7/H-8
2
2
2
2
5a
5b
5c
5d
C axis
2
K70.44
15.88
1
0.78
C axis
2
K81.07
K82.85
K66.65
1218
G18
1218
G18
31.08
Plane (MESO)
Plane (MESO)
2
7.28
Scheme 8. Synthesis of the indoline and bisindoline from acetamido-stilbene 9.
proximity to the electron-withdrawing amide group. The
absence of the trans coupling (16 Hz) present in the
aminostilbene 11, coupled with the disappearance of
the C-7 proton (originally present in the stilbene 11) in 6
proved that cyclization had occurred resulting in the
quaternary C-7 in 6. The complete assignment of all carbon
resonances was obtained from the HMQC and HMBC
spectra. Crucial HMBC correlations are shown in
Figure 1(a) and (b).
1
3
formation of the indole ring. In addition, the C spectrum
showed that the methyne C-7 in 11 had changed to a
2.1.2. N-[7-(3,4-Dimethoxy-phenyl)-7,8-dihydro-indol-
N-yl]-ethanone 24. The UV spectrum of the dihydroindole

N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
11737
2
4
3
Figure 1. (a) Key J and J HMBC correlations in 6. (b) Key J HMBC
correlations in 6.
2
(
4 showed absorption maxima at 282, 253, 239 and 213 nm
log 3 3.84, 3.10, 3.12 and 3.34) typical of a dihydroindole
chromophore. In addition the HREI mass spectrum gave a
molecular ion peak with an additional two mass unit
compared to 6; accurate mass measured as 297.1377 is
compatible with the molecular formula C H O N (calcu-
3
Figure 3. Key J HMBC correlations in 25.
1
8 19 3
that the m/z 297 peak originated from the m/z 592 ion
derived from a symmetrical dimer. The IR spectrum showed
lated mass 297.1365, D 4.0 ppm). The IR spectrum
K1
indicated the presence of a C]O stretch at 1660 cm
K1
the presence of C]O stretch at 1666 cm due to the C]O
1
3
due to the tertiary amide carbonyl. The C NMR spectrum
showed two methoxyl carbons, one acetyl, one methylene,
one methine (stereogenic centre), seven tertiary aromatic
1
of the tertiary amide. The H NMR spectrum of the
bisdihydroindole 25 was similar to the monomer dihydro-
indole 24 except for the 5-membered ring system. This
revealed two methyne singlets (d3.51 and d4.66)
1
carbons and six quaternary carbons. H NMR spectrum
showed the chemical shift of each signal in the aromatic
region to be shifted slightly upfield compared to the indole
6
. This is due to the greater availability of the lone electron
pair on the indoline nitrogen (mesomeric delocalization).
The COSY spectrum showed correlation signals between
0
H-7 and H-8/8 thus confirming the existence of the indoline
skeleton. In addition the olefinic protons (H-7 and H-8) of
the acetamidostilbene (starting material) were replaced by
0
the signals of H-7 at d 5.29 and H-8/8 at d 2.94 and d 3.78,
therefore indicating that a cyclization had occurred resulting
in the indoline 24. Important HMBC correlations are shown
in Figure 2(a) and (b).
2
3
Figure 2. (a) Key J HMBC correlations in 24. (b) Key J HMBC
correlations in 24.
0 0
.1.3. N-[N -Acetyl-7,7 -bis-(3,4-dimethoxyphenyl)-
,8,7 ,8 -tetrahydro-N H-[8,8 ]biindolyl-N-yl]-ethanone
5. The UV spectrum showed absorption band at 282 and
14 nm (log 3 3.84 and 4.31) typical of a dihydroindole
2
7
2
2
0
0
0
0
chromophore. High resolution mass spectrum gave a
molecular ion peak at 592.2574 (calculated mass
5
formula C H O N corresponding to a dimeric species.
92.2573, D C0.7 ppm) consistent with the molecular
3
6 36 6 2
The huge (64.7%) peak at m/z 297 corresponds to the
monomeric 24, formed by cleavage of the dimer with the
loss of one H. The possibility that dimerization might have
occurred within the ionisation chamber was ruled out by the
result of a MIKES experiment, which clearly demonstrated
Scheme 9. Proposed mechanism for the formation of 2-phenyl indole via
manganese acetate-mediated oxidative cyclization.

11738
N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
Scheme 10. Proposed mechanism of formation of 24 and 25.
0
0
corresponding to H-7 and H-8, and H-7 and H-8 . This also
proves that 25 is a symmetrical dimer since the methylene
signal of H-8 in 24 was replaced by the methyne signal. The
to the lactam 18. Ring opening and proton transfer 18–20
will generate a second manganese triacetate carbon-centred
radical 20. This will now attack the double bond in the
manner depicted (see 21) resulting in 22 (after further
oxidation). Rapid attack by the nucleophilic acetamide NH
on the suitably disposed empty p orbital leads to 23. The
conformation of 23 allows efficient antiperiplanar elimin-
ation of the mangano-acetate complex leading to indole 6.
The complex reaction mixture is better explained by this
pathway rather than that originally proposed in Scheme 3.
1
3
C spectrum provides additional support for the dimeric
structure of by revealing two carbon methyne peaks at d64.2
and d56.8 of C-7 and C-8, respectively. The complete
assignments of the protons and carbons were confirmed by
the COSY, HMQC and HMBC spectra (Fig. 3). The
bisdihydroindole 25 possesses 4 stereogenic centres.
2
.2. Mechanistic considerations
2
.2.1. Mechanism for formation of the indoline 24 in the
presence of FeCl . As we have previously observed
2
0
Although the indole synthesis has not been fully investi-
gated, a mechanistic interpretation must account for the
complex mixture as revealed by TLC (see Schemes 6 and 9).
Oxomanganate carbon centred radical complex 15 attacks
the double bond in 11 to give benzylic radical 17.
Intramolecular nucleophilic acyl substitution will give rise
3
3
C
(Scheme 2), olefin (stilbene) oxidation by means of Fe
gives rise to the radical cation 2. However, the dimerization
of the reactive intermediate followed by 6p electrocyclic
ring closure that we have observed and previously
2
0
reported with stilbenes such as 1 (Scheme 10) is not
Scheme 11. The Nicolaou radical cation-mediated synthesis of hybocarpone.

N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
11739
Scheme 12. (a) Dimerisation pathways leading to stereoisomer 25a. (b) Dimerisation pathways leading to stereoisomer 25b. (c) Dimerisation pathways leading
to stereoisomer 25c. (d) Dimerisation pathways leading to stereoisomer 25d.
2
4
observed with the acetamidostilbene 9. Intramolecular
nucleophilic attack by the ortho-acetamido group would
appear to be the favoured pathway leading to 24. The fact
that indole 6 is not obtained under these condition may be
attributed to two factors. Firstly, oxidation of the indolyl
radical 27 to the corresponding benzylic cation followed by
the deprotonation here proceeds more slowly. This may be
because chelation of the acetamido group to Fe has the
effect of withdrawing electrons from the benzene ring,
thereby destabilising benzylic carbocations. Secondly,
radical promoted dimerisation may be the kinetically more
favourable process. We suspect the indolyl radical mono-
mers are held in close proximity to each other by FeCl3
Nicolaou has reported the total synthesis of hybocarpone,
which involved a single-electron transfer process that leads
to the highly reactive radical cation intermediate
(Scheme 11). Kam has reported in the same year the
electrochemical one-electron oxidation of the hexacyclic
indole alkaloid kopsamine that produced a symmetrical
2
5
bisindole dimer (via radical cations). Oxidative generation
leading to radical cation intermediates has been proposed
for the synthesis of bisnaphthol dimers in the presence of
copper (II) salt (Smrcina et al. ). The construction of aryl–
aryl bonds (dibromotetraalkylbiphenyls) by means of ferric
chloride reported by Bushby exerted a profound influence
on our work. Evidence for the intermediacy of radical
cations in the carbon–carbon bond forming step has come
from the work of Creasson et al. Oxidative dimerisation
involving radical cations is discussed in an interesting
review entitled ‘Enantioselective radical processes’ by Sibi
3C
2
6
2
7
(
chelation effect). Mechanistic aspects of radical dimeriza-
2
8
tion pathways leading to 25 will be discussed later. With
regard to our synthesis of the dimer 25c (Scheme 8) for
which we have proposed a radical cation pathway, the
following observations are worth noting. We have pre-
viously reported the first synthesis of restrytisol analogues
from stilbenes and proposed a radical cation dimerization/
pericyclic pathway. One might propose an alternative
pathway involving attack by the indolyl radical 27 on the
stilbene 9 (Scheme 10) followed by cyclization which
would install the second indole ring. Although this
suggestion is worthy of consideration, it lacks the
explanatory power of the proposal depicted in
Schemes 12(a)–(d) and 13 that effectively addresses the
stereochemical issues.
2
9
et al. The mechanistic and stereochemical implications of
our investigations will now be discussed.
2.2.2. Symmetry and the stereochemistry problem.
Although there are fourteen theoretically possible stereo-
isomers, our conclusion that 25c represents the correct
stereostructure is based on the following considerations We
have already described key features of the NMR spectrum
of the bisindoline dimer 25 that strongly suggests a
symmetrical environment in the vicinity of the H-7 and
0
0
H-8 (or H-7 and H-8 protons). The various possibilities are

11740
N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
thus narrowed down to the four structures (25a, 25b, 25c
and 25d) indicated in Table 1. The precise description of
symmetry shown in the table applies strictly to the
planar structural representations and not necessarily to the
3-dimensional energy minimized structures (for which other
terminology could be applied). For 25a, 25b, 25c and 25d,
It is possible to discriminate even more between (G)-25b
and 25c by examining physical molecular models. As
indolyl radicals approach each other even more closely,
formation of (G)-25b (Scheme 13) is prohibited by
0
repulsive interactions between C(2)–H, C(3)–H and C(2 )–
H. These interactions are reduced to a degree in the case of
the formation of 25c (Scheme 13) 25c is thus the favoured
stereostructure on both kinetic and thermodynamic grounds.
The elucidation of the structure of the bisindoline dimer
energy minimized structures were obtained by means of the
annealing method and semi-empirical AM1 calculations.
3
0,31
Heats of formation and dihedral angles obtained from these
calculations are shown in the above-mentioned Table 1. The
relative lack of coupling, (H-7 and H-8 are singlets) reflect a
dihedral angle close to 908. That there is a weak coupling
(
25c) is now complete (Fig. 5).
between H-7 and H-8 is shown in the COSY spectrum
0
(
500 MHz). The proximity of the H-7/H-7 to the electron-
withdrawing acetamide nitrogens would further decrease
the coupling constant between the H-7 and H-8 (as well as
0
0
between H-7 and H-8 ) below the theoretical value obtained
from the Karplus equation. Therefore, coupling constants
3
2
for structures 25a and 25d would be inconsistent with the
experimental data. The calculated heats of formation of
33
2
5a and 25d are also much higher than those for 25b and
5c. On the basis of these arguments structures 25a and 25d
2
must be ruled out.
Figure 4. Structural analogy between 25c and amurensin A.
3
. Conclusion
In both (G)-25b and 25c, semi-empirical calculations
indicate that the H-7-H-8 dihedral angle is large enough to
account for the extremely weak coupling (for (G)-25b and
The following inferences may be drawn from this study: (1)
the use of the FeCl reaction to install two C–N bonds, one
3
2
5c, 1218). Stereostructure 25c is, however, 1.8 kcal/mol
C–C bond as well as four new stereogenic centres in one
step; (2) the synthetic utility of amino stilbenes prepared by
Heck coupling in indole construction. These are to the best
of our knowledge very few examples of this strategy; (3) the
more stable than (G)-25b. Examination of indolyl radical
dimerization pathways leading to (G)-25a and 25d
provides further convincing reasons for the fact that these
stereoisomers are not obtained. It is clear that the approach
of the two indolyl radicals (25a or 25a ). (Scheme 12(a)) or
use of FeCl in the oxidative cyclization of the above
3
1
3
stilbene to produce indoles and indolines (our catechol
indoline is a new compound). Our attention has been drawn
2
5d or 25d (Scheme 12(d)) would result in a destabilizing
1 3
interaction either between the C-2 substituents (catechol
rings) or alternatively between the catechol ring of one
component and the indole ring system of the other.
Pathways leading to 25a and 25d would therefore have
higher energy of activation barriers.
3
to an intriguing 2002 paper by Sayre et al., in which he
4
revisited the FeCl dimerisation of 3-methylindole, first
3
reported in 1957. From a clinical standpoint, Van Vranken’s
3
5
report of the first total synthesis of the antitumour
antibiotic AT2433-A1, a bisindolylmaleimide, is note-
worthy; (4) the elaboration of the novel bisindoline
containing four contiguous stereogenic centres by a
mechanistic pathway that involves radical cations. This is
certainly a unique feature of this transformation; (5) in the
case of bisindole formation, and in contrast to the
dimerization of 1 (Scheme 2), the pericyclic pathway is
completely suppressed; (6) the solution to the elucidation of
the relative configuration of the bisindole by spectroscopic
and computational methods.
With respect to the formation of (G)-25b and 25c,
examination of pathways (Scheme 12(b) and (c), respect-
ively) indicates that both indolyl radicals can approach each
other without interference from the C-2 phenyl substituents.
Scheme 13. Dimerisation pathways (within bonding distance) leading to
25b and 25c.
Figure 5. Two views of the computerized model of 25c.

N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
11741
It has not escaped our attention that the bisindole dimer,
0
9 in 24% yield after chromatography (ethyl acetate/hexane
8:2).
with its C8–C8 bond, bears some relation to the known
3
6
oligostilbenoid dimer amurensin A (Fig. 4). In the light of
our interest in oligostilbenoid dimerization by one-electron
HRMS-EI 297.1353 (calculated mass 297.1365, D C4 ppm);
UV (MeOH) lmax nm (log 3): 323 (4.48), 209 (4.64); IR
1
9,20
oxidants,
this bisindole can be considered, with respect
1
nmax: 2837, 1665, 1515; H NMR (400 MHz, CDCl ) d: 7.71
to the carbon skeleton, as a novel analogue of a natural
oligostilbenoid dimer. We are currently investigating the
relationship between the substitution pattern of the starting
stilbene, the mechanism and the dimerization products from
3
(d; JZ8.08 Hz; 1H) (H-11), 7.61 (s; 1H) (N–H), 7.51 (d;
JZ7.56 Hz; 1H) (H-14), 7.24 (t; JZ7.56 Hz; 1H) (H-12),
7.16 (t; JZ7.56 Hz; 1H) (H-13), 7.06–6.98 (m; 3H) (H-6,
H-2, H-7), 6.89 (d; JZ16.12 Hz; 1H) (H-8), 6.85 (d; JZ
the FeCl reaction.
3
8
.08 Hz; 1H) (H-5), 3.90 (s; 3H) (OMe), 3.88 (s; 3H)
1
3
(OMe), 2.19 (s; 3H) (COCH3); C NMR (100.6 MHz,
CDCl ) d: 168.9, 149.1, 148.9, 134.4, 131.4, 130.8, 130.2,
4. Experimental
3
127.7, 126.3, 125.5, 124.6, 121.7, 119.8, 111.2, 109.3, 55.8,
23.8.
4
4
.1. General
.1.1. N-[7-(3,4-Dimethoxy-phenyl)-indol-N-yl]-etha-
4.1.5. Amino stilbene 11. The starting material, iodoaniline
10 (8 g, 0.01826 mol) was dissolved in dry DMF (80 ml) in
a dry, clean two-necked round bottom flask. The solution
was heated to 120 8C and allowed to stir under nitrogen for
5 min. Palladium acetate (0.154 g, 0.3445 mmol) was then
added, followed by addition of triethylamine (18.2 ml,
none 6. The Heck product, amino stilbene 11 (0.5 g,
.96 mmol), manganese triacetate (1.052 g, 3.92 mmol),
1
potassium acetate (1.92 g, 0.0196 mol) and 13% acetic acid
in acetic anhydride as the solvent (26 ml) were refluxed
under nitrogen for 6 h. The mixture was then cooled, diluted
with saturated sodium chloride solution and extracted with
ethyl acetate. The combined organic layers were dried over
anhydrous sodium sulphate. The crude product was purified
by column chromatography and preparative thin layer
chromatography to yield indole 6 in 13%.
0.0657 mol) and 3,4-dimethoxy styrene
8 (7.5 g,
0.0228 mol). The mixture was refluxed under nitrogen for
48 h. The reaction mixture was then filtered and saturated
sodium chloride added. The aqueous mixture was extracted
with ethyl acetate. The combined ethyl acetate extracts were
dried over anhydrous sodium sulphate. The crude was
purified by column chromatography (ethyl acetate/hexane
8:2) to give 43% of orange crystals of the aminostilbene 11.
HRMS-EI 295.1209 (calculated mass 295.1208, D
0
.3 ppm); UV (MeOH) lmax nm (log 3): 282 (5.84), 253
6.10), 239 (6.12), 213 (6.34); IR nmax: 1605, 1586, 1702; H
1
(
NMR (400 MHz, CDCl ) d: 8.32 (d; JZ8.32 Hz; 1H)
HRMS-EI 255.1267 (calculated mass 255.1259,
C0.8 ppm); UV (MeOH) lmax nm (log 3): 340 (4.18), 294
D
3
(
H-11), 7.5 (d; JZ7.32 Hz; 1H) (H-14), 7.31 (t; JZ7.08 Hz;
1
(4.16), 222 (4.2824); IR nmax: 3373, 1514, 1266; H NMR
1
8
6
3
H) (H-12), 7.24 (t; JZ6.84 Hz; 1H) (H-13), 6.99 (dd; JZ
.32, 1.96 Hz; 1H) (H-6), 6.92 (d; JZ2.44 Hz; 1H) (H-2),
.89 (d; 1H) (H-5), 6.56 (s; 1H) (H-8), 3.90 (s; 3H) (OMe),
(400 MHz, CDCl ) d: 7.37 (d; JZ7.8 Hz; 1H) (H-11), 7.12–
3
7.00 (m; 4H) (H-12, H-13, H-2, H-8), 6.91 (d; JZ16.12 Hz;
1H) (H-7), 6.84 (d; JZ8.04 Hz; 1H) (H-5) 6.79 (d; JZ
1
3
.87 (s; 3H) (OMe), 2.07 (s; 3H) (COCH3); C NMR
1
3
(
100.6 MHz, CDCl ) d: 171.6 (COCH ), 149.5 (C-12),
3
7.56 Hz; 1H) (H-14), 6.71 (d; JZ7.8 Hz; 1H) (H-6);
C
NMR (100.6 MHz, CDCl ) d: 149.0, 148.7, 143.7, 130.6,
3
1
49.1 (C-11), 139.5 (C-6), 137.5 (C-9), 128.9 (C-5), 126.6
3
(
1
(
C-2), 125.1 (C-4), 124.9 (C-3), 123.6 (C-14),121.7 (C-1),
20.1 (C-13), 115.9 (C-10), 112.0 (C-8), 56.2 (OMe), 111.2
OMe), 110.9 (C-7), 56.0, 55.1 (OMe), 27.6 (COCH3).
129.9, 128.2, 126.9, 123.9, 122.2, 119.6, 119.6, 119.0,
116.1, 111.1, 108.7, 55.8, 55.7.
4
.1.6. N-[7-(3,4-Dimethoxy-phenyl)-7,8-dihydro-indol-1-
0
0 0
yl]-ethanone 24 and N-[N -acetyl-7,7 -bis-(3,4-
4
1
0
0
.1.2. 2 -Iodoacetanilide 7. Iodoacetanilide 7 (CAS No
9591-17-4) was prepared by reacting iodoaniline 10 (5 g,
.023 mol) in dry DMF (40 ml) with sodium hydride (1.096,
.046 mol) and acetic anhydride (10.87 ml, 0.115 mol) to
0
0
0
0
dimethoxy-phenyl)-7,8,7 ,8 -tetrahydro-N H-[8,8 ]bi-
indolyl-N-yl]-ethanone 25. Protected amino stilbene 9
(0.079 g, 0.266 mmol) was dissolved in 6 ml of dichloro-
methane. FeCl $6H O (0.72 g, 2.66 mmol) was added to the
mixture. The mixture was allowed to stir overnight at room
temperature and was monitored by TLC. After the
consumption of the starting material, the reaction mixture
was diluted with saturated sodium chloride and extracted
with ethyl acetate. The crude product obtained after
evaporation under reduced pressure was subjected to
preparative thin layer chromatography. Two pure com-
pounds, 24 (38% yield) and 25 (15% yield) were isolated.
produce N-(2-iodo-phenyl)-acetamide 7 as brown crystals in
7
3
2
9% yield after chromatography (ethyl acetate/hexane 8:2).
4
2
.1.3. 3,4-Dimethoxystyrene 8. Styrene 8 (CAS No 6380-
3-0) was prepared by reacting methyltriphenylphos-
phonium iodide (4.86 g, 0.012 mol) in THF (40 ml), 0 8C
with potassium tert-butoxide (12 ml, 0.012 mol) and 3,4-
dimethoxybenzaldehyde (2 g, 0.012 mol) via Wittig
methodology to give yellowish oil 8 in 60% yield after
chromatography (ethyl acetate/hexane 8:2).
Spectroscopic data for 24. HRMS-EI 297.1377 (calculated
mass 297.1365, D C1.2 mmu); UV (MeOH) lmax nm
(log 3): 282 (3.84), 253 (4.10), 239 (4.12), 213 (4.34); IR
4
.1.4. Acetamido stilbene 9. Replacement of the iodo-
aniline 10 by 7 (0.3 g, 1.15 mmol) with 3,4-dimethoxy-
styrene 8 (0.236 g, 1.44 mmol), palladium acetate (4.87 mg,
1
nmax: 1660, 1481, 1462, 1027; H NMR (400 MHz, CDCl )
3
0.0115 mmol) and triethylamine (0.57 ml, 4.14 mmol) in
dry DMF (10 ml) produced the protected acetamidostilbene
d: 8.28 (d; JZ6.8 Hz; 1H) (H-11), 7.23 (t; JZ7.8 Hz; 1H)
(H-12), 7.09 (d; 1H) (H-14), 7.02 (t; JZ7.32 Hz; 1H)

11742
N. F. Thomas et al. / Tetrahedron 60 (2004) 11733–11742
(
H-13), 6.75 (d; JZ8.04 Hz; 1H) (H-5), 6.68 (dd; JZ6.8,
a Guide for the Synthetic Chemist; Pergamon: Amsterdam,
2000; (indole, indoline and bisindole syntheses are covered in
Chapter 3) .
1
1
1
(
(
(
(
.48 Hz; 1H) (H-6), 6.62 (s; 1H) (H-2), 5.29 (d; JZ9.04 Hz;
H) (H-7), 3.81 (s; 3H) (OMe), 3.75 (s; 3H) (OMe), 3.78 (d;
0
H) (H-8 ), 2.94 (d; JZ15.88 Hz; 1H) (H-8), 2.03 (s; 1H)
9. Grigg, R.; Rutnikovic, B.; Urch, C. J. Tetrahedron Lett. 1996,
37, 695.
1
3
COCH );
C NMR (100.6 MHz, CDCl ) d: 169.9
3
3
COCH ), 149.8 (C-4), 148.9 (C-3), 143.6 (C-10), 136.0
3
C-1), 129.5 (C-9), 127.9 (C-12), 125.1 (C-14), 124.3
C-13), 117.4 (C-6), 117.2 (C-11), 111.8 (C-5), 108.3 (C-2),
10. Dong, Y.; Busacca, C. A. J. Org. Chem. 1997, 62, 6464.
11. Brenner, M.; Mayer, G.; Terpin, A.; Steglich, W. Chem. Eur. J.
1997, 3, 70.
6
(
3.6 (C-7), 56.2 (OMe), 56.1 (OMe), 39.4 (C-8), 24.4
COCH3).
12. Wallow, T. I.; Novak, B. M. J. Org. Chem. 1994, 59, 5034.
13. Wright, S. W.; Hageman, D. L.; McClure, L. D. J. Org. Chem.
1
994, 59, 6095.
Spectroscopic data for 25. HRMS-EI 592.2574 (calculated
mass 592.2573, D C0.1 mmu); UV (MeOH) l nm
14. (a) Kobayashi, Y.; Fukuyama, T. J. Heterocycl. Chem. 1998,
35, 1043. (b) Kobayashi, Y.; Peng, G.; Fukuyama, T.
Tetrahedron Lett. 1999, 40, 1519.
max
1
log 3): 282 (3.84) 214 (4.31); IR nmax: 1666, 1026; H NMR
(
(
(
400 MHz, CDCl ) d: 8.47 (d; JZ8.08 Hz; 1H) (H-11), 7.45
t; JZ7.84 Hz; 1H) (H-12), 7.31 (d; JZ7.32 Hz; 1H) (H-14)
15. Grigg, R.; Teasdale, A.; Sridharan, V. Tetrahedron Lett. 1991,
32, 3859.
3
7
.21 (t; JZ7.08 Hz; 1H), (H-13), 6.61 d; JZ8.32 Hz; 1H)
16. Larock, R. C.; Yum, E. K. J. Am. Chem. Soc. 1991, 113, 6689.
17. Larock, R. C.; Yum, E. K.; Refvik, M. D. J. Org. Chem. 1998,
63, 7652.
(
H-5), 6.22 (d; JZ8.04 Hz; 1H) (H-6), 5.80 (s; 1H) (H-2),
4
3
.66 (s; 1H) (H-7), 3.75 (s; 3H) (OMe), 3.59 (s; 3H) (OMe),
.51 (s; 1H) (H-8), 1.88 (s; 3H) (COCH3); C NMR
1
3
18. Larock, R. C. J. Organomet. Chem. 1999, 576, 111.
19. Thomas, N. F.; Lee, K. C.; Paraidathathu, T.; Weber, J. F. F.;
Awang, K. Tetrahedron Lett. 2002, 43, 3151.
20. Thomas, N. F.; Lee, K. C.; Paraidathathu, T.; Weber, J. F. F.;
Awang, K.; Rondeau, D.; Richomme, P. Tetrahedron 2002,
58, 7201.
(
(
100.6 MHz, CDCl ) d: 169.8 (COCH ) 149.4 (C-3), 148.4
3 3
C-4), 143.9 C-10), 134.1 (C-1), 130.1 (C-9), 129.3 C-12),
1
1
25.2 (C-14), 124.7 (C-13), 117.7 (C-11), 115.9 (C-6),
11.4 (C-5), 107.5 (C-2), 64.2 (C-7), 56.8 (C-8), 23.6
(
COCH3).
2
1. Ziegler, C. B.; Heck, R. F. J. Org. Chem. 1978, 43, 2941.
2. Akazome, M.; Kondo, T.; Watanabe, Y. J. Org. Chem. 1994,
59, 3375.
2
Acknowledgements
This work was supported by a grant from the Malaysian
Government (IRPA No. 09-02-03-0802).
2
3. We thank the referee for his suggestion.
References and notes
24. Nicolaou, K. C.; Gray, D. Angew. Chem., Int. Ed. 2001, 40,
61.
7
1
2
3
. Mann, J. Secondary Metabolism; Oxford University Press:
Oxford, 1987.
25. Kam, T. S.; Lim, T. M.; Tan, G. H. J. Chem. Soc., Perkin
Trans. 1 2001, 1594.
. (a) Mann, J. Chem. Ber., 478. (b) Mann, J. Murder, Magic and
Medicine; Oxford University Press: Oxford, 1992.
26. Smrcina, M.; Polakova, J.; Vyskocil, S.; Kocovsky, P. J. Org.
Chem. 1993, 58, 4534.
. (a) Woodward, R. B.; Bader, F. E.; Bickel, H.; Frey, A. J.;
Kierstead, R. W. J. Am. Chem. Soc. 1956, 78, 2023. (b)
Woodward, R. B.; Bader, F. E.; Bickel, H.; Frey, A. J.;
Kierstead, R. W. J. Am. Chem. Soc. 1956, 78, 2657. (c)
Woodward, R. B.; Bader, F. E.; Bickel, H.; Frey, A. J.;
Kierstead, R. W. Tetrahedron 1958, 2, 1.
27. Boden, N.; Bushby, R. J.; Lu, Z.; Headdock, G. Tetrahedron
Lett. 2000, 41, 10117.
28. Creason, S. C.; Wheeler, J.; Nelson, R. J. Org. Chem. 1972, 37,
4440.
29. Sibi, M. P.; Manyem, S.; Zimmerman, J. Chem. Rev. 2003,
103.
4
. (a) Woodward, R. B.; Cava, M. P.; Ollis, W. D.; Hunger, A.;
Daeniker, H. U.; Schenker, K. J. Am. Chem. Soc. 1954, 76,
30. Dewar, M. J. S.; Zocbish, E. G.; Healy, E. F.; Stuart, J. J. P.
J. Am. Chem. Soc. 1985, 107, 3902.
4
749. (b) Woodward, R. B.; Cava, M. P.; Ollis, W. D.; Hunger,
31. AMPAC version 6.0, 7204 Mullen, SemiChem, Shawne, KS,
USA, 1997.
A.; Daeniker, H. U.; Schenker, K. Tetrahedron 1963, 19, 247.
. Joule, J. A.; Mills, A. K. In Heterocyclic Chemistry, 4th ed.;
Blackwell: Oxford, 2000.
5
6
32. Friebolin, H. Basic One- and Two-dimensional NMR Spec-
troscopy; Wiley: Weinheim, 1998; p 93.
. (a) Solid phase synthesis: Nicolaou, K. C.; Roecker, A. J.;
Pfefferkorn, J. A.; Cao, G. Q. J. Am. Chem. Soc. 2000, 122,
33. In the light of the dihedral angles for (G)-25a and 25d, there is
some doubt about the ‘symmetry’ of these compounds. The
results of further computational chemistry will be reported in
due course.
2
966. (b) Catalytic (Gribb’s): Witulski, B.; Stengel, T.;
Fernandez-Hernandez, J. M. Chem. Commun. 2000, 1672.
c) Photolytic: Benali, O.; Miranda, M. A.; Tormas, R.; Gill, S.
(
34. Ling, K. Q.; Ren, T.; Protasiewicz, J. D.; Sayre, L. M.
Tetrahedron Lett. 2002, 43, 6903.
J. Org. Chem. 2002, 67, 7915.
. Frank, W. C.; Kim, Y. C.; Heck, R. F. J. Org. Chem. 1978, 43,
7
8
35. Chisholm, J. D.; Van Vranken, D. L. J. Org. Chem. 2000, 65,
7541.
36. Huang, K. S.; Lin, M. J. Asian Nat. Prod. Res. 1999, 2, 21.
2947.
. Li, J. J.; Gribble, G. W. Palladium in Heterocyclic Chemistry: