Ferrocenyl induced one-pot synthesis of 3,3′-ferrocenylbiindoles

Article abstract of DOI:10.1080/00958972.2020.1770235

When we used 2-ferrocenethynylaniline (1a) as reactant and NaAuCl₄·2H₂O (5%) as catalyst to prepare 2-ferrocenylindole (2a), we were surprised to find that 3,3′-ferrocenylbiindole (3a) was obtained simultaneously. 3a was characterized by elemental analysis, FT-IR, MS, NMR, and X-ray single crystal diffraction. The reactant substitution reaction, in-situ HNMR, XPS, EPR characterization, and calculations were carried out to interpret why 3a was obtained in one pot. The results showed that the hydroperoxyl radical played a key role in the reaction, which induced synthesis of 3a.

Full text of DOI:10.1080/00958972.2020.1770235

Journal of Coordination Chemistry
ISSN: 0095-8972 (Print) 1029-0389 (Online) Journal homepage: https://www.tandfonline.com/loi/gcoo20
Ferrocenyl induced one-pot synthesis of 3,3′-
ferrocenylbiindoles
Ligang Yan, Limin Han & Ruijun Xie
To cite this article: Ligang Yan, Limin Han & Ruijun Xie (2020): Ferrocenyl induced
one-pot synthesis of 3,3′-ferrocenylbiindoles, Journal of Coordination Chemistry, DOI:
10.1080/00958972.2020.1770235
To link to this article: https://doi.org/10.1080/00958972.2020.1770235
View supplementary material
Published online: 02 Jun 2020.
Submit your article to this journal
Article views: 15
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=gcoo20

JOURNAL OF COORDINATION CHEMISTRY
https://doi.org/10.1080/00958972.2020.1770235
Ferrocenyl induced one-pot synthesis of
0
3
,3 -ferrocenylbiindoles
Ligang Yan, Limin Han and Ruijun Xie
Chemical Engineering College, Inner Mongolia University of Technology, Hohhot, P. R. China
ABSTRACT
ARTICLE HISTORY
When we used 2-ferrocenethynylaniline (1a) as reactant and
Received 22 January 2020
Accepted 29 April 2020
NaAuCl
4
ꢀ2H
2
O (5%) as catalyst to prepare 2-ferrocenylindole (2a),
0
we were surprised to find that 3,3 -ferrocenylbiindole (3a) was
obtained simultaneously. 3a was characterized by elemental ana-
lysis, FT-IR, MS, NMR, and X-ray single crystal diffraction. The react-
ant substitution reaction, in-situ HNMR, XPS, EPR characterization,
and calculations were carried out to interpret why 3a was obtained
in one pot. The results showed that the hydroperoxyl radical played
a key role in the reaction, which induced synthesis of 3a.
KEYWORDS
Ferrocenyl; biindoles;
one-pot synthesis; reaction
mechanism; homo-coupling
CONTACT Limin Han
hanlimin_442@hotmail.com
Chemical Engineering College, Inner Mongolia University of
Technology, Hohhot 010051, P. R. China
Supplemental data for this article can be accessed https://doi.org/10.1080/00958972.2020.1770235.
ß 2020 Informa UK Limited, trading as Taylor & Francis Group

2
L. YAN ET AL.
1. Introduction
The indole nucleus is a common substructure of many natural products and pharma-
ceutical compounds, found in calycanthidine, chimonanthine, psychotriasine, mela-
tonin, and tryptophan [1–4]. Synthesis methods of indole and its derivatives, such as
Fischer indole synthesis, cyclization of 2-alkynylaniline, and reductive cyclizations of
nitro compounds [5], have been studied extensively. Transition metal-catalyzed C–H
activation of indoles also played important roles for the synthesis of indole derivatives,
for example, palladium-catalyzed electrophilic substitution reaction of indole and
umpolung of 2-indolylmethanols can be used to construct a large number of indole
derivatives [6–12].
0
0
Recently, the synthesis of biindole derivatives, such as 6,6 -biindole,3,3 -biindole,
biindole-methane [13–15], has attracted much attention. Up to now, the construction
of biindolyl compounds can be classified into three main methods. First, the biindoles
can be formed through oxidative dimerization of indoles (Scheme 1(a)) [16–19]; sec-
ondly, the biindoles can be constructed by transitional metal-catalyzed coupling reac-
tion or substitution reaction of pre-functionalized indole [20–25] (Scheme 1(b) and (c));
shirdly, the cyclization reaction and C–H oxidative homo-coupling reaction of the
2
2
-alkynylanilines (Scheme 1(d)) [26, 27]. Overall, one-pot synthesis of diindole from
-alkynylanilines is an economical method, but the synthesis mechanism of diindole
still needs to be further studied.
Ferrocene has a stereoscopic sandwich structure and redox properties. Therefore, if
a heterocyclic compound involves ferrocene units, it may exhibit some unique chem-
ical properties [28, 29]. Despite our research group having synthesized many ferrocenyl
heterocyclic compounds [30], we want to synthesize the rare ferrocenyl
indole compounds.
First, we want to use Au-catalyzed cyclization of 2-ferrocenethynylaniline (1a) to
synthesize ferrocenyl indoles (2a) [31, 32]. Interestingly, the target product was not
0
the desired product 2a, but 3,3 -ferrocenylbiindole (3a) was obtained. Hence, we inter-
pret why 3a was obtained in one pot. Through reactant–substitution reaction, in-situ
HNMR characterization, XPS, EPR characterization, and theory calculations, we propose
a ferrocenyl-induced hydroperoxyl radical catalytic mechanism in this article.
2
. Experimental
.1. General procedures
All the reactions were monitored with thin-layer chromatography (TLC). NaAuCl ꢀ2H O,
2
4
2
2-iodoaniline, 2-phenylethynylaniline (1b), 2-phenylindole (2b), ferrocenylacetylene,
NH Cl, and K CO were commercially available. Column chromatographic separations
4
2
3
1
13
and purifications were performed on 100–200 mesh silica gel. H and C spectra were
recorded on an Agilent Technologies-500 MHz spectrometer. FT-IR spectra were meas-
ured on a Nexus-870 FT-IR spectrometer. Elemental analyses were carried out on an
Elementarvar III-type analyzer. Mass spectra were determined using a Shimadzu LCMS-
2020 instrument. X-ray photoelectron spectra (XPS) and valence bond spectra (VBS)

JOURNAL OF COORDINATION CHEMISTRY
3
0
Scheme 1. The synthesis of 3,3 -biindoles.
were measured on a Thermo-Fisher ESCALAB 250Xi spectrometer equipped with a
monochromatic Al X-ray source (1486.6 eV). Electron paramagnetic resonance (EPR)
was measured using a Bruker EPR spectrometer (A300, X band) with the microwave
frequency 9.5 GHz and modulation frequency (100 kHz). Theory calculations were per-
formed using Gaussian09 [33]. The geometry optimizations, frequency analysis, and
thermodynamic corrections were performed at the B3LYP (D3BJ), MDF10 (for Fe), and
Def2SVP (for others) level of theory [34–36].
2.2. Synthesis
2.2.1. Synthesis of 2-ferrocenethynylaniline (1a)
An ethanol solution (10 mL) containing 2-iodoaniline (88 mg, 0.4 mmol), ferrocenylace-
5
5
tylene (42 mg, 0.2 mmol), PdClf[(g -C H )]Fe[(g -C H )]-2-Pyg(Py)] (3 mg, 0.005 mmol),
5
5
5 3
ꢁ
and K CO (57 mg, 0.4 mmol) was stirred at 40 C for 4 h under an argon atmosphere
2
3
(Scheme 2). The reaction mixture was washed with degassed water and extracted with

4
L. YAN ET AL.
Scheme 2. The synthesis of 2-ferrocenethynylaniline.
0
Scheme 3. The synthesis of 3,3 -ferrocenylbiindole.
dichloromethane. The organic layer was filtered and the solvent was removed under
vacuum. The residue was subjected to chromatographic separation on a silica gel
column (2.0 ꢂ 30 cm). Elution with a mixture of dichloromethane and hexane (1:2/V:V)
gave the second red band as 1a. Single crystals of 1a were obtained from a mixture
of hexane:dichloromethane (3:1/V:V).
ꢁ
2-Ferrocenethynylaniline (1a) Yield: 39 mg, 65%. M.p. 123–125 C (decomp). Anal.
Calcd for C H FeN: C, 71.76%; H, 4.98%; N, 4.65%. Found: C, 72.06%; H, 5.53%; N,
1
8 15
ꢃ1
4
.98%. FT-IR (cm ): 3472, 3374, 3090, 2921, 2201, 1605, 1446, 1308, 1161, 815, 752.
1
H NMR (500 MHz, CDCl ): d: 7.35 (d, 1H, J ¼ 7.5 Hz, Ph), 7.13 (t, 1H, J ¼ 8 Hz, Ph), 6.72
3
13
(m, 2H, Ph), 4.52 (s, 2H, Cp), 4.25 (s, 7H, Cp), 3.99 (s, 1H, NH ) ppm. C NMR (125 MHz,
2
CDCl ): d: 131.85, 129.12, 118.05, 114.33, 93.55, 81.99, 71.34, 69.95, 68.87, 65.91. ESI-MS
3
þ
(
CH Cl , m/z): 300.95 (M ).
2
2
0
.2.2. Synthesis of 2-ferrocenylindole (2a) and 3,3 -ferrocenylbiindole (3a)
2
2
0
-Ferrocenethynylaniline (1a) (60 mg, 0.2 mmol) and NaAuCl ꢀ2H O (5%) (4 mg,
4
2
ꢁ
.01 mmol) were dissolved in ethanol (5 mL) and stirred at 40 C for 8 h in air (Scheme 3).
Then, the reaction mixture was quenched with saturated NH Cl solution and extracted
4
with dichloromethane. The organic solvent was removed in vacuo and the residues were
subjected to chromatographic separation on a silica gel column (2.0 ꢂ 30 cm). Elution
with a mixture of dichloromethane and hexane (1:2/V:V) gave the first yellow band as 2a.
The second orange band was 3a; single crystals of 3a were obtained from a mixture of
hexane:dichloromethane (3:1/V:V).
ꢁ
2-Ferrocenylindole (2a), Yield: 3 mg, 5%. M.p. 168–170 C (decomp). Anal. Calcd for
C H FeN: C, 71.76%; H, 4.98%; N, 4.65%. Found: C, 72.01%; H, 5.15%; N, 5.01%. FT-IR
1
8
15
ꢃ1
1
(
(
cm , KBr): 3435, 1443, 1417, 1344, 1293, 1105, 1029, 819, 791, 734. H NMR
500 MHz, CDCl ): d: 8.05 (bs, 1H, N–H), 7.55 (d, 1H, J ¼ 7.5 Hz, Ph), 7.36 (d, 1H,
3

JOURNAL OF COORDINATION CHEMISTRY
5
J ¼ 8.0 Hz, Ph), 7.15 (t, 1H, J ¼ 7 Hz, Ph), 7.09 (t, 1H, J ¼ 7.5 Hz, Ph), 6.52 (s, 1H, pyrrole),
1
3
4
.64 (s, 2H, C H ), 4.34 (s, 2H, C H ), 4.11 (s, 5H, C H ) ppm. C NMR (125 MHz, CDCl ):
5 4 5 4 5 5 3
d: 137.06, 136.23, 129.34, 121.39, 119.97, 119.78, 110.33, 98.88, 69.55, 68.82, 65.91. ESI-
þ
MS (CH Cl , m/z): 300.95 (M ).
2
2
0
ꢁ
3
,3 -Ferrocenylbiindole (3a), Yield: 36 mg, 60%. M.p. 115–118 C (decomp). Anal.
Calcd for C H Fe N : C, 71.76%; H, 4.65%; N, 4.65%. Found: C, 72.21%; H, 4.88%; N,
3
6
28
2 2
ꢃ1
5
.05%. FT-IR (cm ): 3473, 3091, 1608, 1592, 1494, 1461, 1447, 1309, 1164, 818, 754.
1
H NMR (500 MHz, CDCl ): d: 7.36 (d, 2H, J ¼ 8 Hz, Ph), 7.13 (t, 2H, J ¼ 7.5 Hz, Ph), 6.86
3
(
d, 2H, J ¼ 7.5 Hz, Ph), 6.77 (d, 2H, J ¼ 7.5 Hz, Ph), 4.52 (s, 4H, C H ), 4.23 (s, 14H, C H ,
5
4
5 4
1
3
C H ) ppm. C NMR (125 MHz, CDCl ): d: 147.30, 131.75, 129.07, 117.76, 114.56, 93.56,
5
5
3
þ
8
1.76, 71.27, 69.92, 68.88, 65.09. ESI-MS (CH Cl , m/z): 599.95 (M ).
2 2
2.2.3. Synthesis of 2-phenylindole (2b) and diphenylbiindole (3b)
2
-Phenylethynylaniline (1 b) (39 mg, 0.2 mmol), ferrocene (19 mg, 0.1 mmol), and
NaAuCl ꢀ2H O (5%) (4 mg, 0.01 mmol) were dissolved in ethanol (5 mL) and stirred at
4
2
ꢁ
ꢁ
4
0 C for 8 h under air or 70 C for 24 h under oxygen. After the reaction was com-
pleted, the reaction mixture was extracted by dichloromethane. The organic solvent
was removed in vacuo and the residues were subjected to chromatographic separation
on a silica gel column (2.0 ꢂ 30 cm). Elution with a mixture of dichloromethane and
hexane (1:3/V:V) gave the corresponding pure product 2b (in air) and 3b (in oxygen).
The spectroscopic data were consistent with those reported previously [26].
1
2
-Phenylindole (2b): H NMR (500 MHz, CDCl ): d: 8.31 (s, 1H, NH), 7.67–7.62 (m, 3H,
3
ArH), 7.45–7.39 (m, 3H, ArH), 7.32 (t, 1H, J ¼ 6.5 Hz, ArH), 7.20 (t, 1H, J ¼ 2.5 Hz, ArH),
1
3
7
1
.12 (t, 1H, J ¼ 9 Hz, ArH), 6.83 (s, 1H, C-3H) ppm. C NMR (125 MHz, CDCl ): d: 137.9,
3
32.4, 129.2, 127.7, 125.2, 122.4, 120.7, 120.3, 110.1, 100.0.
Diphenylbiindole (3b): H NMR (500 MHz, DLM): d: 10.65 (s, 2H, NH), 7.55 (d, 4H,
1
J ¼ 7.5 Hz, ArH), 7.49 (d, 2H, J ¼ 8 Hz, ArH), 7.15–7.03 (m, 10H, ArH), 6.87 (t, 2H,
13
J ¼ 7.5 Hz, ArH) ppm. C NMR (125 MHz, DLM): d: 137.6, 135.5, 133.9, 130.7, 129.2,
27.7, 127.4, 122.5, 120.3, 119.6, 111.9, 107.5.
1
2.3. X-ray crystallography
X-ray single crystal data for 1a and 3a were recorded on a Bruker APEX II CCD diffract-
ometer with graphite-monochromated Mo Ka (k ¼ 0.71073 Å) radiation [37]. Empirical
absorption corrections by semi-empirical methods from equivalents were carried out.
The structures were solved by direct methods using SHELXS-97 and refined with
2
SHELXL-97 by full-matrix least-squares techniques on F [38]. All non-hydrogen atoms
were refined anisotropically, while the hydrogens were located geometrically and
refined isotropically.
3. Results and discussion
3.1. Synthesis and characterizations of 1a, 2a, and 3a
Compounds 1a and 3a were synthesized and their molecular structures were charac-
terized by NMR, FT-IR, and X-ray single crystal diffraction. According to the gold-

6
L. YAN ET AL.
Table 1. Crystal data and relevant structural parameters of 1a and 3a.
1
a
3a
Fe
Formula
C
18
H
15FeN
C
37
H
30Cl
2
2 2
N
Formula weight
T/ C
301.16
20(2)
685.23
20(2)
ꢁ
k/Å
0.71073
Monoclinic
0.71073
Crystal system
Space group
Monoclinic
P2 /c
P2
1
/c
1
a/Å
b/Å
c/Å
9.2634(4)
10.6602(5)
14.6444(7)
90
106.272(2)
90
1388.20(11)
4
1.441
12.8235(9)
15.5818(11)
9.1541(7)
90
99.475(2)
90
3044.2(2)
4
1.495
ꢁ
a/
ꢁ
b/
ꢁ
c/
3
V/Å
Z
q
ꢃ3
Calcd/g m
ꢃ1
Abs. coeff./mm
1.074
1.159
F(0 0 0)
624
5.796–56.808
1408.0
5.77–52.836
ꢁ
2
h range/
Reflections collected
27,514
49,806
Unique
3471
6231
Completeness to h
99.5%
99.9%
T
max. and Tmin.
0.746 and 0.600
1.065
0.578 and 0.573
1.041
2
Goodness-of-fit on F
Final R indices [I > 2r(I)]
R indices (all data)
R1 ¼ 0.0549
wR2 ¼ 0.1263,
R1 ¼ 0.0732,
wR2 ¼ 0.1369
R1 ¼ 0.1221,
wR2 ¼ 0.2692,
R1 ¼ 0.1547,
wR2 ¼ 0.2922
catalyzed protocol in the literature [39], for the NaAuCl ꢀ2H O-catalyzed cyclization of
4
2
1
a, the desired 2a should be obtained. But in the present preparation, we were sur-
1
13
prised to find that 3a was obtained in good yield (Scheme 3). The H and C NMR
spectra of 1a and 3a were similar in CDCl . The chemical shift of the two protons of
3
the substituted Cp-rings appeared at d ¼ 4.50–4.53 ppm, whereas the other seven pro-
tons of Cp rings appeared at d ¼ 4.23–4.25 ppm. Protons of the benzene groups
appeared at d ¼ 7.35, 7.13, and 6.72 ppm, respectively. But the protons of amino
groups (1a) were at d ¼ 3.99 ppm, which could be used to distinguish 1a and 3a in
NMR spectra [26]. The molecular structures of 1a and 3a were further confirmed by X-
ray single crystal diffraction analysis. Crystal data and relevant structural parameters are
listed in Table 1. Selected bond lengths and angles are listed in Table 2. ORTEP diagrams
of 1a and 3a are shown in Figures 1 and 2. Owing to the formation of a pyrrole ring via
cyclization, the bond lengths of the C–N and C–C in 1a are shorter than the correspond-
ing bond lengths in 3a. For instance, the N(1)–C(18) bond length is 1.331(5) Å and the
C(11)–C(12) bond length is 1.197(4) Å in 1a. The N(2)–C(25) bond length is 1.372(11) Å,
the N(1)–C(12) bond length is 1.366(13) Å and the C(11)–C(18) bond length is 1.355(12) Å
in 3a. In the crystal structure of 3a, because the bond length of the two pyrrole rings
were different, the dihedral angles between the one pyrrole ring (C18–C17–C12–N1) and
ꢁ
the Fc ring is 19.688 while the dihedral angles between another pyrrole ring
ꢁ
(
C20–C19–C26–N2) and the Fc ring is 4.224 . The dihedral angle of the two pyrrole rings
ꢁ
0
was 81.586 , which was the axially chiral 3,3 -bisindole frameworks due to introducing
the Fc group. The molecular structure of 2a can be further determined by the 2D COSY,
2
D HSQC, and 2D HMQC (see Supporting Information). d ¼ 8.05 ppm and d ¼ 6.52
showed no spatial correlation with other protons. The d ¼ 8.05 ppm is attributed to the

JOURNAL OF COORDINATION CHEMISTRY
7
ꢁ
Table 2. Selected bond lengths (Å) and angles ( ) of 1a and 3a.
1
a
3a
Fe(1)–C(10)
Fe(1)–C(6)
C(11)–C(10)
C(11)–C(12)
N(1)–C(18)
C(12)–C(14)
C(11)–C(10)–C(6)
C(18)–C(14)–C(12)
N(1)–C(18)–C(14)
C(11)–C(12)–C(14)
C(9)–C(8)–C(7)
C(14)–C(13)–C(15)
2.045(3)
2.034(3)
1.432(4)
1.197(4)
1.331(5)
1.455(5)
125.4(3)
118.9(3)
120.7(4)
179.6(4)
108.3(3)
118.7(4)
N(1)–C(11)
N(1)–C(12)
C(11)–C(18)
N(2)–C(26)
C(19)–C(18)
N(2)–C(25)
C(18)–C(19)–C(26)
C(17)–C(18)–C(19)
C(10)–C(11)–C(18)
C(18)–C(19)–C(20)
C(27)–C(26)–C(19)
C(19)–C(20)–C(25)
1.392(10)
1.366(13)
1.355(12)
1.379(10)
1.469(11)
1.372(11)
128.3(7)
125.9(8)
130.5(7)
125.8(7)
130.5(7)
108.2(7)
Figure 1. The molecular structure of 1a. Hydrogens have been omitted for clarity.
Figure 2. The molecular structure of 3a. Hydrogens have been omitted for clarity.

8
L. YAN ET AL.
Scheme 4. Primary mechanism studies.
H-1 protons. The d ¼ 6.52 ppm corresponded to the H-3 of pyrrole. In order to distinguish
the H-1 and H-3 protons accurately, the chemical shift of 8.05 ppm disappeared in 2D
HSQC, illustrating that the chemical shifts of H-3 of pyrrole was 6.52 ppm. Through ana-
lysis of 2D COSY, 2D HSQC, and 2D HMQC, the chemical shifts of H-5, H-8, H-7, and H-6
protons were at d ¼ 7.53, 7.35, 7.14, and 7.09 ppm, respectively. The protons of the ferro-
cenyl units were at d ¼ 4.11 ppm (C H ), 4.34 ppm, and 4.64 ppm (C H ), respectively [32].
5
5
5 4
It is concluded that 2a was synthesized.
3.2. Mechanism studies
0
In order to answer why 3,3 -ferrocenylbiindole (3a) can be obtained in one pot, several
experiments were carried out (Scheme 4). First, we investigated the catalytic

JOURNAL OF COORDINATION CHEMISTRY
9
Figure 3. XPS Fe2p3/2 spectra for the residues of reaction.
Reaction residues
1
1
50000
00000
5
0000
0
-
50000
-
-
100000
150000
0
2
4
6
8
10 12 14 16
Value (g-Factor)
Figure 4. The EPR spectra for iron(III).
performance of Au for synthesizing the biindole. When 2-ferrocenethynylaniline (1a)
was substituted by 2-phenylethynylaniline (1b) under the standard reaction conditions,
only 2-phenylindole (2b) was obtained in 80% yield (Scheme 4, eq. 1). This demon-
strates that NaAuCl ꢀ2H O only catalyzes intramolecular cyclization reaction of 1b to
4
2
0
2
b; NaAuCl ꢀ2H O could not catalyze oxidative coupling of 2b to form 3,3 -diphenyl-
4
2
biindole (3b). With 1a as reactant under the same reaction conditions, why was 3a
obtained? With this question, we investigated the synthesis of 3,3 -biindole derivatives
extensively from the literature [16–27]. Fe can catalyze oxidative homo-coupling of
-arylindole to form the homo-coupling product 3,3 -biindole with molecular oxygen
0
3
þ
0
2
as the oxidant [18].

1
0
L. YAN ET AL.
1
Figure 5. The proceeding of the reaction was determined by in situ H NMR analysis of 2a under
1
1
1
air and acidity. H NMR(CD O). (A) The H NMR of 2a when reaction starts in air. (B) The H NMR
4
1
1
of 2a in acidic solution after 1 h. (C) The H NMR of 2a in acidic solution after 1 day. (D) The H
NMR of 3a.
Ferrocene and ferrocenyl derivatives are classical acid-sensitive organometallics;
3
þ
under acidic media of NaAuCl ꢀ2H O–ethanol, the ferrocene will be oxidized to Fe in
4
2
oxygen or in air [40]. To test this, we repeated the reaction (Scheme 4, eq. 2) with the
addition of ferrocene (0.5 eq.) under air and 3,3 -diphenylbiindole (3b) was obtained
0
with yield of 15%. When the same reaction was performed under oxygen, the yield of
3
b increased to 50%. This phenomenon demonstrated that ferrocene could be the
3
þ
source of Fe which was generated easily under oxygen atmosphere. Therefore, the
Fe and oxygen had key roles in oxidative homo-coupling reaction [18].
Is Fe involved in the oxidative homo-coupling of 2a to form 3a? In order to
obtain the valence of iron in the system, we used XPS to measure the surface chem-
ical state of Fe species of the reaction residues (Scheme 3). There are two Fe 2p
3
þ
3
þ
3/2
2
þ
3þ
peaks, which could be described as mixed valence of Fe and Fe (Figure 3). The
XPS binding energy for the 2p orbital of Fe is about 707.5 eV relative to C 1s at
2þ
3þ
2þ
284.8 eV, while the binding energy for Fe is about 710.7 eV. The Fe signal may ori-
3
þ
ginate from the ferrocene moiety, while the Fe can be attributed to the oxidation
product of ferrocene [41]. The Fe has also been confirmed by EPR spectroscopy of
3
þ
the reaction residues (Figure 4), where g ¼ 4.3 (1500 G) is a characteristic value of
3
þ
Fe [42].
3
þ
Does Fe come from the ferrocenyl of 1a or 2a? When 1a is the reactant (Scheme 4,
eq. 4), only 2a was obtained under Ar atmosphere. This revealed that 2a was synthesized
3
þ
through Au-catalyzed intramolecular cyclization of 1a, and Fe was not generated in
argon. Under air atmosphere, 2a (5%) and 3a (60%) were obtained together (Scheme 3).
3
þ
These comparative experiments proved that air is necessary for formation of Fe , which
comes from ferrocene of 1a or 2a.
To investigate the effect of acid and air on the transformation of 2a to 3a, the in
1
situ H NMR experiment of 2a was carried out under simulated reaction conditions

JOURNAL OF COORDINATION CHEMISTRY
11
Scheme 5. The redox reaction of ferrocene and its derivatives in air and acid condition.
1
1
5000
0000
5
000
0
-
5000
-
-
10000
15000
3
460 3480 3500 3520 3540 3560
Magnetic field (G)
Figure 6. EPR spectra for DMPO adduct with 2a (0.2 g/L) and HCl (25 mM) under atmos-
pheric conditions.
with air and acidity (Scheme 4, eq. 5). The 6.43 ppm could be attributed to the chem-
ical shift of 2a. Initially, the 6.43 ppm was found in NMR spectra, which could be
attributed to the chemical shift of the H-3 of pyrrole in 2a (Figure 5(A)). With the add-
ition of acid, the peaks disappeared at 6.43 ppm after 1 h (Figure 5(B)). As the reaction
proceeded, the peaks at 7.5–7.2 ppm and 7.0–6.8 ppm gradually split and moved after
24 h (Figure 5(C)), in accord with the chemical shift of 3a (Figure 5(D)). This result indi-
cates that C–H bond cleavage at the C-3 position in 2a could be key in the homo-cou-
pling reaction to synthesize the desired 3a through a single-electron-transfer pathway
3
þ
[
18]. When Fe was formed from the ferrocene of 1a or 2a under acid and air condi-
tions, homo-coupling reaction was initiated.
3
þ
Calculations were used to determine if Fe was a real catalyst in the homo-coupling
reaction. The redox reaction of ferrocene and its derivatives with air (or oxygen gas) in
acidic solution have been done by several workers (Scheme 5) [43–46]. In C–H oxidative
homo-coupling reaction, oxidants may be the radicals produced by oxidation of ferro-
cene and its derivatives. To verify this hypothesis, DMPO adduct with 2a (0.2 g/L) and
HCl (25 mM) were recorded under atmospheric conditions by EPR spectra. The result
showed a strong four-line EPR signal with the relative intensity ratio of 1:2:2:1 (Figure 6)
[
45]. This illustrated that the addition of DMPO led to the EPR peak of DMPO-OH and the
formation of hydroxyl radical (HOꢀ) in the system. If the HOꢀ was used as oxidant
Scheme 5, eqs. 1, 2, and 4), it would consume at least two equivalents of 2a to produce
(
one equivalent of 3a. The yield of 3a would be much lower than 50%. If the hydroper-
oxyl radical (HOOꢀ) was utilized as oxidant (Scheme 5, eqs. 1, 2, and 3), one equivalent of

1
2
L. YAN ET AL.
Figure 7. Energy profile for the reaction between 2a and 3a. Relative energies are given in
kcal/mol.
0
Scheme 6. Prospective catalytic cycle for the formation of 3,3 -ferrocenylbiindole.
2a would generate one equivalent of 3a. Then, the yield of 3a would be higher than
50%, but lower than 75%; the yield of 3a was always 60–70% in this experiment.
Comparing eq. 3 and eq. 4 in Scheme 5, the pH of the reaction system dominated the
kinds of radicals. Strong acid promoted HOꢀ in eq. 4; weak acid produced HOOꢀ in eq. 3
[
47]. In the EPR experiment, excessive acid led to the production of HOꢀ.
To assess whether HOOꢀ was involved in the C–H oxidative homo-coupling reaction,
we complemented the mechanistic experimental with theoretical calculations (Figure 7).
The relative free energy profile showed that there are three different intermediates in
the process (INT1, INT2, and INT3). The highest energy intermediate (INT3) was
176.1 kcal/mol and the energy intermediate (INT1) was 139.2 kcal/mol above the reac-
tants. The reason of the unusual high barrier was that the precursor of the intermediate

JOURNAL OF COORDINATION CHEMISTRY
13
(
INT3 or INT1) provided one electron to HOOꢀ which would react with protons simultan-
eously (Scheme 5, eq. 2). For chemical calculation, this reaction was artificially divided
into two stages, resulting in presenting the energy of half-reaction intermediates INT3
and INT1. When protons were involved in the reaction, the energy declined to
ꢃ35.9 kcal/mol at the last step. The present experiments and calculations confirmed that
HOOꢀ was an oxidant in C–H oxidative homo-coupling reaction.
Based on the experiments, theory calculations, and the literature, a possible mech-
anism was proposed. The formation of 3a may proceed via two steps (Scheme 6). The
first step is the cyclization reaction initiated by the formation of the adduct between
Au(III) and 1a forming intermediate A. Then, intermediate A undergoes protonolysis to
afford 2a in the presence of strong acid [48]. The second step begins with generating
of HOOꢀ (Scheme 5, eq. 1); in eq. 1, the ferrocenyl of 1a or 2a was partially oxidized
3
þ
into Fe by air and acid, HOOꢀ was generated simultaneously. Then, HOOꢀ captured
an electron from another part of 2a to form the radical cation intermediate INT1
which is then coupled to neutral 2a to form intermediate INT2. Intermediate INT2
loses an electron to convert into intermediate INT3. Ultimately, deprotonation of INT3
leads to the corresponding 3a [19].
4. Conclusion
0
We have synthesized a diindole compound, 3,3 -ferrocenylbiindoles (3a), and charac-
terized it by elemental analysis, FT-IR, MS, NMR, and X-ray single crystal diffraction. We
proposed a possible reaction mechanism through experiments and calculations. The
research provides a new synthesis of biindole derivatives.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
We are grateful to the Program for New Century Excellent Talents in University [NCET-08-858]
and the Natural Science Foundation of China [NSFC-21462029].
References
[
[
[
1] E.C. Gentry, L.J. Rono, M.E. Hale, R. Matsuura, R.R. Knowles. J. Am. Chem. Soc., 140, 3394
2018).
2] C.R. Jamison, J.J. Badillo, J.M. Lipshultz, R.J. Comito, D.W.C. MacMillan. Nat. Chem., 9, 1165
2017).
3] A.G. Turjanski, F. Leonik, D.A. Estrin, R.E. Rosenstein, F. Doctorovich. J. Am. Chem. Soc.,
22, 10468 (2000).
(
(
1
[4] A. Crespo, A.G. Turjanski, D.A. Estrin. Chem. Phys. Lett., 365, 15 (2002).
[5] G.R. Humphrey, J.T. Kuethe. Chem. Rev., 106, 2875 (2006).
[6] S. Cacchi, G. Fabrizi. Chem. Rev., 111, PR215 (2011).
[7] S.W. Youn, T.Y. Ko. Asian J. Org. Chem., 7, 1467 (2018).
[8] V. Pirovano. Eur. J. Org. Chem., 2018, 1925 (2018).

1
4
L. YAN ET AL.
[9] M.M. Xu, H.Q. Wang, Y.J. Mao, G.J. Mei, S.L. Wang, F. Shi. J. Org. Chem., 83, 5027 (2018).
[
[
10] Z.Q. Zhu, Y. Shen, J.X. Liu, J.Y. Tao, F. Shi. Org. Lett., 19, 1542 (2017).
11] B. Lu, Y.D. Luo, L.Z. Liu, L.W. Ye, Y.Z. Wang, L.M. Zhang. Angew. Chem. Int. Ed. Engl., 50,
8
358 (2011).
12] M. Sun, C. Ma, S.J. Zhou, S.F. Lou, J. Xiao, Y.C. Jiao, F. Shi. Angew. Chem. Int. Ed. Engl., 58,
703 (2019).
[
8
[
[
[
[
13] A.B. Gamble, P.A. Keller. Chem. Commun. (Camb.), 46, 4076 (2010).
14] D.S. Black, N. Kumar. Adv. Heterocycl. Chem., 101, 97 (2010).
15] Y. Wan, H.Q. Wang, M.M. Xu, G.J. Mei, F. Shi. Org. Biomol. Chem., 16, 1536 (2018).
16] P.P. Huang, X.J. Peng, D. Hu, H.W. Liao, S.B. Tang, L.X. Liu. Org. Biomol. Chem., 15, 9622
(
2017).
17] Y. Li, W.H. Wang, S.D. Yang, B.J. Li, C. Feng, Z.J. Shi. Chem. Commun. (Camb.), 46, 4553
2010).
[
(
[
[
18] T.M. Niu, Y.H. Zhang. Tetrahedron Lett., 51, 6847 (2010).
19] T. Wirtanen, M.K. M €a kel €a , J. Sarfraz, P. Ihalainen, S. Hietala, M. Melchionna, J. Helaja. Adv.
Synth. Catal., 357, 3718 (2015).
[20] U. Berens, J.M. Brown, J. Long, R. Selke. Tetrahedron: Asymm., 7, 285 (1996).
[21] T.R. Li, M.M. Zhang, B.C. Wang, L.Q. Lu, W.J. Xiao. Org. Lett., 20, 3237 (2018).
[22] S. Naskar, P. Paira, R. Paira, S. Mondal, A. Maity, A. Hazra, K.B. Sahu, P. Saha, S. Banerjee,
P. Luger, M. Webe, N.B. Mondal. Tetrahedron, 66, 5196 (2010).
[23] J. Guo, J. Weng, G.B. Huang, L.J. Huang, A.S.C. Chan, G. Lu. Tetrahedron Lett., 57, 5493
(2016).
[
[
[
24] H.A. Duong, S. Chua, P.B. Huleatt, C.L.L. Chai. J. Org. Chem., 73, 9177 (2008).
25] B.O.A. Tasch, D. Antovic, E. Merkul, T.J.J. M u€ ller. Eur. J. Org. Chem., 2013, 4564 (2013).
26] J.E. Perea-Buceta, T. Wirtanen, O.V. Laukkanen, M.K. M €a kel €a , M. Nieger, M. Melchionna, N.
Huittinen, J.A.L. Sanchez, J. Helaja. Angew. Chem. Int. Ed. Engl., 52, 11835 (2013).
27] M. Yamashita, T. Noro, A. Iida. Tetrahedron Lett., 54, 6848 (2013).
[
[
28] B. Lal, A. Kanwal, A.A. Altaf, A. Badshah, F. Asghar, S. Akhter, S. Ullah, S.I. Khan, M.N. Tahir.
J. Coord. Chem., 72, 2376 (2019).
[
[
[
[
29] C. D ꢀı az, G. Ferraudi, A.G. Lappin, A. Oliver, M. Isaacs. J. Coord. Chem., 71, 2584 (2018).
30] Y.-Q. Hu, L.-M. Han, N. Zhu, H.-L. Hong, R.-J. Xie. J. Coord. Chem., 66, 3481 (2013).
31] L.G. Yan, L.M. Han, R.J. Xie. J. Coord. Chem., 72, 2705 (2019).
32] J.C. Torres, R.A. Pilli, M.D. Vargas, F.A. Violante, S.J. Garden, A.C. Pinto. Tetrahedron, 58,
4
487 (2002).
33] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman. Gaussian
9, Revision E.01, Gaussian, Inc., Wallingford, CT (2013).
[
0
[
[
[
[
[
[
[
[
34] A.D. Becke. J. Chem. Phys., 98, 5648 (1993).
35] M. Dolg, U. Wedig, H. Stoll, H. Preuss. J. Chem. Phys., 86, 866 (1987).
36] A. Sch €a fer, H. Horn, R. Ahlrichs. J. Chem. Phys., 97, 2571 (1992).
37] Bruker, APEX2 and SAINT-Plus, Bruker AXS Inc., Madison, Wisconsin (2004).
38] G.M. Sheldrick. Acta Crystallogr. C Struct. Chem., 71, 3 (2015).
39] A. Arcadi, G. Bianchi, F. Marinelli. Synthesis, 4, 610 (2004).
40] J.P. Hurvois, C. Moinet. J. Organomet. Chem., 690, 1829 (2005).
41] L.Z. Zhang, H. Chen, X.F. Zhao, Q. Zhai, D.J. Yin, Y.F. Sun, J.H. Li. Appl. Catal. B: Environ.,
1
93, 47 (2016).
[42] T.G. Costa, B. Szpoganicz, G.F. Caramori, V.R. de Almeida, A.S. Mangrich, A.P. Mangoni.
J. Coord. Chem., 67, 986 (2014).
[43] Y. Okada, T. Nakagawa, T. Hayashi. Inorg. Chim. Acta, 312, 197 (2001).
[44] Q. Wang, S.L. Tian, P. Ning. Ind. Eng. Chem. Res., 53, 643 (2014).
[45] Y.J. Li, B.J. Zhang, X.L. Liu, Q. Zhao, H. Zhang, Y.C. Zhang, P. Ning, S.L. Tian. J. Hazard.
Mater., 353, 26 (2018).
[46] C.K. Duesterberg, S.E. Mylon, T.D. Waite. Environ. Sci. Technol., 42, 8522 (2008).
[47] Q. Wang, S.L. Tian, J. Cun, P. Ning. Desal. Water Treat., 51, 5821 (2013).
[48] G. Abbiati, F. Marinelli, E. Rossi, A. Arcadi. Isr. J. Chem., 53, 856(2013).