Synthesis of Indoles by Reductive Cyclization of Nitro Compounds Using Formate Esters as CO Surrogates

Article abstract of DOI:10.1002/ejoc.202100789

Alkyl and aryl formate esters were evaluated as CO sources in the Pd- and Pd/Ru-catalyzed reductive cyclization of 2-nitrostyrenes to give indoles. Whereas the use of alkyl formates requires the presence of a ruthenium catalyst such as Ru₃(CO)₁₂, the reaction with phenyl formate can be performed by using a Pd/phenanthroline complex alone. Phenyl formate was found to be the most effective CO source and the desired products were obtained in excellent yields, often higher than those previously reported using pressurized CO. The reaction tolerates many functional groups, including sensitive ones like a free aldehydic group or a pendant pyrrole. Detailed experiments and kinetic studies allow to conclude that the activation of phenyl formate is base-catalyzed and that the metal doesn't play a role in the decarbonylation step. The reactions can be performed in a single thick-walled glass tube with as little as 0.2 mol-% palladium catalyst and even on a 2 g scale. The same protocol can be extended to other nitro compounds, affording different heterocycles.

Full text of DOI:10.1002/ejoc.202100789

European Journal of Organic Chemistry
Accepted Article
Accepted Article
Title: Synthesis of Indoles by Reductive Cyclization of Nitro
Compounds Using Formate Esters as CO Surrogates
Authors: Manar Ahmed Fouad, Francesco Ferretti, Dario Formenti,
Fabio Milani, and Fabio Ragaini
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.202100789
Link to VoR: https://doi.org/10.1002/ejoc.202100789
01/2020

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
Synthesis of Indoles by Reductive Cyclization of Nitro
Compounds Using Formate Esters as CO Surrogates
Manar Ahmed Fouad,[a],[b] Francesco Ferretti,[a] Dario Formenti,[a],[c] Fabio Milani,[a] Fabio Ragaini*[a]
[a]
Manar A. Fouad, Dr. Dario Formenti, Dr. Francesco Ferretti, Fabio Milani, Prof. Dr. Fabio Ragaini
Dipartimento di Chimica, Università degli Studi di Milano
Via C. Golgi 19, 20133 Milano - Italy
E-mail: fabio.ragaini@unimi.it:
[b]
[c]
Chemistry Department, Faculty of Science, Alexandria University, P.O. Box 426, Alexandria, 21321, Egypt
Current address: Institut für Anorganische Chemie - RWTH Aachen, Landoltweg 1a, 52074 Aachen - Germany
Supporting information for this article is given via a link at the end of the document.
Abstract: Alkyl and aryl formate esters were evaluated as CO
sources in the Pd- and Pd/Ru-catalyzed reductive cyclization of 2-
nitrostyrenes to give indoles. Whereas the use of alkyl formates
requires the presence of a ruthenium catalyst such as Ru3(CO)12, the
We started searching for a suitable CO surrogate to perform
nitroarene reductive cyclization reactions many years ago and
selected formate esters because of their commercial availability,
low cost and low toxicity. We also chose to investigate the Pd-
catalyzed synthesis of indoles from o-nitrostyrenes (Scheme 1,
path a) as the first reaction to test and decided to operate in a
single glass pressure tube because this equipment is cheap and
available in different sizes. A steel autoclave can also be
employed, but without using pressurized CO. Note that some of
other known CO-releasing reagents require the use of two-
chamber reactors to separate the CO-releasing reaction from
the carbonylation one, a less convenient and versatile apparatus.
When we started this study, no alternative had been reported to
the use of gaseous CO in nitroarene cyclization reactions. More
reaction with phenyl formate can be performed by using
a
Pd/phenanthroline complex alone. Phenyl formate was found to be
the most effective CO source and the desired products were
obtained in excellent yields, often higher than those previously
reported by the use of pressurized CO. The reaction tolerates many
functional groups, including sensitive ones like a free aldehydic
group or a pendant pyrrole. Detailed experiments and kinetic studies
allow to conclude that the activation of phenyl formate is base-
catalyzed and that the metal play no role in the decarbonylation step.
The reactions can be performed in a single thick-walled glass tube
with as little as 0.2 mol-% palladium catalyst and even on a 2 g scale. recently, the use of Mo(CO)6 was reported to this aim.[5]
The same protocol can be extended to other nitro compounds,
affording different heterocycles.
However, this CO source is highly toxic and is not suitable for
large scale applications. Very recently, Wu also employed aryl
formates to carbonylate nitroarenes.[6]
Introduction
The catalytic reductive carbonylation of aromatic nitro
compounds is an efficient method for the preparation of a large
variety of bulk[1] and fine[2] chemicals. Specifically, many nitrogen
containing heterocycles have been previously synthesized from
suitably ortho-substituted nitroarenes using CO as
a
stoichiometric reductant and examples of inter-molecular
reactions and of reactions employing nitroalkenes as reagents
are also known (Scheme 1).[3] Despite of the high efficiency of
many of these reactions, they have not become of widespread
use. The most likely reason for this is that they involve the use of
pressurized CO, requiring safety measures that are not available
in most synthetic organic laboratories. This problem is shared
with other transformations employing gaseous CO. Indeed,
despite the fact that most of the largest scale homogeneously
catalyzed industrial processes are carbonylation reactions, the
use of the same reactions at the laboratory scale is much less
common. To allow standard organic synthesis laboratories to
employ carbonylation reactions, a number of solid or liquid
molecules able to liberate CO in situ have been developed,
especially in the last decade.[4] These allow performing the
reactions in thick-walled borosilicate glass vessels (also called
pressure tubes), thus avoiding the need for high-pressure
equipment and CO lines.
Scheme 1. Selected transition-metal catalyzed syntheses of N-heterocycles
from nitro compounds using CO as the reductant.
1
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
In this paper, we report a full account of our studies on the use
conversion and an increased selectivity in both indoles 2a and
of formate esters in the synthesis of indoles from 2-nitrostyrenes. 3a. Under the same reaction conditions, benzyl and methyl
A preliminary account of some of the results here reported has
been previously published.[7]
formate gave similar results (entries 9 and 10, respectively). A
few attempts were also made to use solvent mixtures in which
only 1 mL of n-butyl formate was employed with 9 mL of another
solvent (DMF, CH3CN, toluene. Entries 11-13), but in all cases
conversion and selectivity in indole decreased.
Results and Discussion
The formation of 3a when Et3N is employed as base was not
expected. Small amounts of the ethyl ester of the starting
nitrostyrene were also observed. It is known that some
mononuclear ruthenium complexes catalyze the decomposition
of Et3N to afford, among others, ethanol[14] and the latter may be
involved in a transesterification reaction. However, this is
unlikely to be the correct explanation because larger amounts of
n-butanol are surely formed as the reaction proceeds and no or
at most trace amounts of the butyl ester of 1H-indole-2-
carboxylic acid could be detected by GC-MS among the reaction
products. On the other hand, it has also been reported that
Ru3(CO)12 catalyzes the exchange of alkyl groups among Et3N
and a different tertiary amine.[15] The reaction proceeds by the
formation of different trinuclear cluster species. Although, to the
best of our knowledge, it has not been reported that esters can
play the role of the second amine, the formation of 3a can be
easily rationalized if we suppose that the same kind of
intermediates that are able to break and reform C-N bonds are
also able to do the same with C-O bonds. Note that even several
palladium complexes are able to activate the C-N bond of
Use of alkyl formates
Owing to their very low cost, our initial investigation was devoted
to prove if alkyl formates could be successful CO surrogates in
the reductive cyclization of methyl (E)-2-nitrocinnamate (1a).
Among formate esters, methyl formate is not only the cheapest,
but also the one that has been more extensively investigated. It
has long been known that its decomposition into CO and
methanol can be promoted by strong bases, but the latter would
not be compatible with most catalytic processes. On the other
hand, Ru3(CO)12 has also been found to be an active catalyst for
this decomposition[8] and this complex is surely compatible with
the present reaction because it has long been known to catalyze
reductive carbonylation and cyclization reactions of nitroarenes,
including the synthesis of indoles.[9] The activity of Ru3(CO)12 for
the decarbonylation of methyl formate is best promoted by
phosphine ligands, but the latter are not compatible with a stable
catalytic system for nitroarene reductive carbonylation because
they react directly with the substrate to give phosphine oxides
under typical reaction conditions.[10] Fortunately, 1,10-
phenanthroline (Phen) was also reported to be effective as a
Ru3(CO)12 activator.[11] Besides this, Phen and its substituted
derivatives are also the ligands of choice to stabilize palladium
to give the most active to date catalytic systems for a number of
Table 1. [Pd]/[Ru]-catalyzed reductive cyclization of 1a using n-butyl formate
as CO source: influence of various reaction parameters.
carbonylation and reductive carbonylation reactions of
3o, 3q, 3u, 12]
nitroarenes.[3d,
Moreover, the combination of a
palladium complex, Ru3(CO)12 and phenanthroline has also
been already tested and the two metals found to cooperate
rather than to inhibit each other.[11, 13]
1a
2a
3a
Select.
[%][c]
2a+3a
Select.
[%][c]
<1
47
48
23
86
29
24
Basic
promoter
Entry
Solvent
Conv.
[%][b]
15
72
85
53
96
>99
97
71
Select.
[%][c]
traces
45
40
23
65
29
24
24
With this knowledge in mind, we started investigating the
decomposition of methyl formate in the presence of Ru3(CO)12
and Phen in an autoclave. However, it turned out that in order to
decompose methyl formate at a convenient rate, the reaction
has to be performed at 180 °C and at this temperature methyl
formate (b.p. 36 °C) generates by itself a vapor pressure of
about 11 bar. The latter is at the limits of what can be withstood
by glass pressure tubes, making their use unsafe. We thus
moved to n-butyl formate (b.p. 107 °C) as a viable alternative.
Results are summarized in Table 1. The reaction conducted with
palladium alone only gave a small conversion of the starting
nitrostyrene (entry 1). In contrast, the use of Ru3(CO)12 alone
was found to be effective A fair selectivity in indole was achieved,
but only a moderate conversion of the nitrostyrene was
observed (entry 2). The combination of the two metals allowed
for an increased yield (entry 3). Organic bases accelerate the
reaction (compare entry 4 with entries 3, 6-8), but only in the
presence of Et3N was an increase in selectivity towards the
desired indoles 2a observed (entry 3). Note that the addition of
the weak base Et3N to the reaction mixture has been shown to
be beneficial for most, although not all, reductive cyclization
reactions of nitroarenes by CO.[2d] However, in the present case,
when Et3N was used, ethyl 1H-indole-2-carboxylate (3a), was
also produced. Increasing the reaction time allowed for a higher
1[d]
2[e]
3
4
5[f]
6
7
8
9
10
n-Butyl formate
n-Butyl formate
n-Butyl formate
n-Butyl formate
n-Butyl formate
n-Butyl formate
n-Butyl formate
n-Butyl formate
Methyl formate[j]
Benzyl formate
Et3N
Et3N
Et3N
-
-
2
8
-
Et3N
21
DBU[g]
DABCO[h]
DIPEA[i]
Et3N
-
-
-
traces
-
24
52
53
48
>99
52
53
Et3N
Et3N
n-Butyl
formate/CH3CN
53
70
28
18
40
21
5
23
40
21
11[k]
12[k]
13[k]
Et3N
Et3N
n-Butyl
formate/DMF
<1
n-Butyl
formate/Toluene
traces
[a] Reaction conditions: 1a
= 0.27 mmol, mol ratio [Pd(Phen)2][BF4]2 :
Ru3(CO)12 : Phen : 1a = 1:1:20:100, mol ratio basic promoter: 1a = 2 : 1,
solvent (10 mL), at 180 °C for 6 hours in a sealed pressure tube. Results
based on GC analysis using biphenyl as the internal standard. [b] With respect
to starting 1a. [c] With respect to converted 1a. [d] Only [Pd(Phen)2][BF4]2 was
used. [e] Only Ru3(CO)12 was used; [f] Reaction time was 10 h; [g] 1,8-
diazabicycloundec-7-ene.
[h]
1,4-diazabicyclo[2.2.2]octane;
[i]
N,N-
diisopropylethylamine. [j] The reaction was performed in a Teflon-coated 200
mL stainless steel autoclave. [k] 1 mL n-butyl formate and 9 mL of the other
solvent.
2
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
tertiary amines, although, to the best of our knowledge, an
exchange reaction such as that described for ruthenium has not
been reported.[16]
Table 2. [Pd]/[Ru]-catalyzed reductive cyclization of 1b using n-butyl formate
as CO source.
The main side product was methyl 2-aminocinnamate (4a,
Scheme 2). The latter was accompanied by its cyclization
product, namely a lactam (5a). Traces of 3,4-dihydroquinolin-
2(1H)-one (6a) were also detected by GC-MS, which can arise
from the reduction of the conjugated C=C double bond in 5a.
Z:E mol
ratio
4b
[%][b]
Select.
27
Entry
1b Conv. [%][b]
2b Select. [%][b]
2:1
2:1
65
>99
>99
>99
41
69
52
45
32
23
74
63
1
2[c]
3[d]
4[e]
5[f]
6
29
22
31
32
20
19
2:1
2:1
2:1
Only Z
Only E
61
81
7
[a]
Reaction
conditions:
0.27
mmol
1b,
mol.
ratio
Scheme 2. Formation of byproducts during the reductive cyclization of 1a into
2a in the presence of n-butyl formate.
[Pd(Phen)2][BF4]2:Ru3(CO)12:Phen:1b = 1:1:20:100; mol. ratio Et3N:1b = 2:1; n-
butyl formate 10 mL. Reactions were performed into heavy-wall glass tubes.
[b] Conversion and selectivity were determined by GC analysis using biphenyl
as the internal standard. [c] Reaction time was 10 h. [d] DBU was used instead
Et3N. [e] DABCO was used instead Et3N. [f] DIPEA was used instead Et3N.
The formation of 4a can be explained by a Ru-catalyzed
reduction of nitro group. Indeed, it is known that Ru3(CO)12 in the
presence of Phen[17] or even just Et3N[18] is a very active catalyst
for the reduction of nitroarenes to anilines by CO/H2O. In fact,
despite the use of distilled solvents and dried glassware,
adventitious traces of water may be present both in the solvents
and in the gaseous nitrogen. However, the over reduction of the
nitro group might also be due to a ruthenium catalyzed transfer
hydrogenation reaction in which butanol formed by formate
decomposition act as the H-donor.
Using the reaction conditions previously employed in the case of
methyl (E)-2-nitrocynnamate (1a), we extended the optimization
study to 2-nitrostylbene (1b) for the synthesis of 2-phenylindole
(2b) (Table 2). Since we obtained the starting material as a
diasteroisomeric E and Z mixture, we examined the reactivity of
both the pure diasteroisomers and their mixture. At variance with
what observed with 1a, in the current case the corresponding
amine 4b was detected as by-product to a much higher extent,
which may simply be due to the fact that in this case it cannot
evolve to other cyclic products (vide supra). As in the previous
case, an elongation of the reaction time allowed for better
conversions although a decrease in the selectivity in the desired
indole 2b was observed (Table 2, entry 2). Even in this case,
bases other than Et3N gave a less selective reaction (entries 3-
5). It is worth noting that Et3N is one of the most convenient
organic bases both in terms of cost and easiness of removal
from the reaction mixture, owing to its relatively low boiling point
(89 °C).
nitrostyrenes react more quickly) and with the Hammet

constant of the substituent on the nitroaryl ring when the same
kind of cyclization reaction is performed with a Pd/Phen catalyst
and under CO pressure.[3d] The activation of the nitroarene was
consequently considered to be the r.d.s. of the reaction. This is
in accord with the fact that the initial activation of nitroarenes by
late-transition metal complexes has been shown to involve a
single electron transfer from the metal to the nitroarene in all the
cases in which it has been investigated.[20] The electronic effect
of styryl group is expected to be essentially unvaried if in the E
or Z configuration, but steric effects indirectly affect the reduction
potential of the nitroarene. Indeed, it has long been known that
o-nitrotoluene has a more negative reduction potential than p-
nitrotoluene, despite electronic effects should be roughly the
same.[21] The effect is due to the fact that the presence of an
ortho substituent causes a tilting of the plane of the nitro group
with respect to that of the phenyl ring and this tilting decreases
the efficiency of the delocalization of the negative charge in the
reduced species.
In order to check if the same effect of steric hindrance on the
reduction potential operates in our case, a cyclic voltammetric
study was performed. Only data for the E isomer have been
previously reported in the literature.[22] As depicted in Figure 1,
three cathodic peaks with peak potentials around -1.12
(A), -1.80 (B) and -2.30 V (C) were detected.
According to the literature, the first peak (A in Figure 1) is
attributed to the quasi-reversible one-electron reduction of the
nitro group to its radical anion.[22] At more negative potentials,
the second peak (B) is attributed to the reversible one-electron
reduction of the nitro radical-anion to the nitrodianion (Scheme
3). Finally, the third cathodic peak (C) at around -2.30 V arises
from the irreversible reduction of the stilbene moiety. The
A close look at the unreacted styrene in those cases in which
the conversion was not complete evidenced that the starting
material was much enriched into the Z isomer with respect to the
initial 2:1 Z/E ratio. To better understand this point, the two
isomers were separated and independently reacted. The
outcome of the two separate reactions (Table 2, entries 6 and 7)
evidences that the reaction is indeed slower in the case of the Z
than in the case of E one. However, the selectivity followed an
opposite trend. A similar effect has been previously observed in
the case of the related cyclization of dimethylamino-substituted
nitrostilbenes.[19]
involved species are depicted in Scheme
corresponding Epc values are reported in Table 3.
3
and the
Analyzing the available data as a whole, it is evident that the Z
isomer is more difficult to reduce than the E one, in accord with
our expectations. That Z-2b feels a strong steric repulsion is
confirmed by its crystal structure clearly showing not only a
twisting of the nitro group with respect to the aryl ring but even a
Davies reported that the reactivity of a series 2-nitrostilbenes
correlated with their reduction potential (more easily reducible
3
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
20
complete conversion and a moderate selectivity was achieved
(Table 4, entry 1). Moreover, when the reaction temperature was
lowered from 180 °C to 140 °C, the selectivity increased to 92%
(entry 2). At lower temperatures, the selectivity decreased,
although the conversion was still virtually complete. Thus,
140 °C was chosen for a further optimization.
Z
E
0
-20
-40
A
In the first place, several Pd sources were tested (Table 4). In
B
-60
-80
Table 4. Reductive cyclization of 1a with HCOOPh: screening of Pd and Ru
precatalysts.[a]
C
-3,5
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
Potential (V vs Cp2Fe)
Figure 1. Cyclic voltammograms recorded of Z-1b and E-1b. The CV tests
were conducted in anhydrous DMF with a substrate concentration of 6·10-4
M
+ 0.1 M tetraethylammonium tetrafluoborate as supporting electrolyte, with a
potential scan rate of 0.2 V·s-1.
Entry
T [°C]
t [h]
Catalyst
1a Conv.
[%][b],[c]
2a Select.
[%][b],[c]
1[d]
2[d]
3[d]
4[d]
180
140
120
90
6
6
6
6
[Pd(Phen)2][BF4]2/
Ru3(CO)12
> 99
> 99
> 99
> 99
65
92
86
76
Scheme 3. Electrochemical reduction steps of 1b.
Table 3. Electrochemical data for reduction of 1b.[a]
[Pd(Phen)2][BF4]2/
Ru3(CO)12
[Pd(Phen)2][BF4]2/
Ru3(CO)12
Epc for Z
isomer
Epc for E
isomer
ΔEpc
Peak
[mV]
[Pd(Phen)2][BF4]2/
Ru3(CO)12
[V]
[V]
A
B
-1.15
-1.83
-1.12
-1.81
30
20
5
140
140
140
140
140
140
3
3
3
3
3
3
[Pd(Phen)2][BF4]2
Pd(CH3CN)2Cl2
Pd(OAc)2
>99 (32)
>99 (19)
>99 (4)
86 (4)
23
94 (34)
98 (17)
92 (traces)
93 (traces)
79
[a] See footnote to Figure 1 for experimental conditions.
6
remarkable twisting of the two aryl moieties.[23] Unfortunately, no
crystal structures of the E isomer has been reported in the
literature. However, unsubstituted trans-stilbene shows
completely flat structure.[24]
7
a
8
Pd(dba)2
It is also worth noting that at the end of the reaction employing
either pure Z-2b or pure E-2b, a small amount of the other
isomer was detected by GC. However, a control experiment in
which pure Z-2b was reacted under the same experimental
conditions (see caption to Table 2), but in the absence of any
metal showed that 13% of the starting olefin had been converted
into the E isomer. Thus, this effect may be ascribed to the high
temperature employed only, without the necessity to claim for an
effect of the catalysts. On the other hand, the byproduct 4b was
exclusively obtained as the E isomer even when pure Z-2b was
used as reagent. This indicates that the catalytic system has the
ability to isomerize the double bond and that the isomerization
occurs more easily when the reduction of the nitro group
proceeds.
9
Ru3(CO)12
10[e]
Pd(CH3CN)2Cl2
58
95
[a] Reaction conditions: 0.27 mmol 1a,
1 mol% of Pd catalyst and/or
Ru3(CO)12, 20 mol% Phen, 500 μL of HCOOPh (corresponding to 17 equiv.
with respect to 1a or 8.5 fold the CO required by the reaction), 2 equiv. Et3N
(80 μL), in CH3CN (10 mL). Reactions were performed into heavy-wall glass
tubes. [b] Conversion and selectivity were determined by GC analysis using
biphenyl as the internal standard. [c] Values in parentheses refer to the
reactions performed omitting Phen. [d] Reaction performed in 1 mL HCOOPh
+ 9 mL CH3CN. [e] Reaction performed in an autoclave under CO pressure
(30 bar). No HCOOPh was added.
order to better differentiate the catalysts, both the reaction time
and the formate amounts were halved.
The n-butyl formate protocol was applied to the synthesis of a
few indoles (vide infra Table 9, conditions A), but isolated yields
were lower than those achievable by the use of gaseous CO.
All the tested palladium compounds were very effective in the
transformation, leading to complete conversion. Significantly, an
almost complete selectivity towards 2a was observed with
Pd(CH3CN)2Cl2 (entry 6). In contrast to the n-butyl formate
based protocol, the addition of Ru3(CO)12 is not necessary, but
the presence of the ligand is still essential and poor results were
obtained in its absence (values in parentheses in Table 4).
Note that by working at this temperature and in the absence of
any ruthenium species, 3a was not formed even though Et3N
Use of phenyl formate
Since we started working in this field many years ago, the use of
phenyl formate as a more effective CO surrogate than alkyl
formates has been reported by several groups.[25] Gratifyingly,
when phenyl formate was employed in a 1:9 mixture with CH3CN,
4
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
was still present. This supports the idea that a ruthenium
species is responsible for the apparent transesterification
reaction.
The amount of phenyl formate was optimized. As shown in
Figure S1 (Supporting Information), the into a pressure tube
formate amount can be lowered up to 3.4-fold the stoichiometric
amount with respect to the required CO (2 equivalents of CO are
needed) without altering the reaction outcome significantly.
Further diminishing the HCOOPh amount led to a slight
decrease of the selectivity, but to a drop in the conversion. The
base amount was also optimized and just 1 equivalent of Et3N
(40 µL) is needed to achieve complete conversion and selectivity
(Figure S2). Finally, the Phen amount could be lowered to 5 mol-
% (Figure S3).
To handle suitable amounts of both starting materials and
reaction products for a substrate scope study, the reaction was
scaled-up by a factor of two. In order not to shift the coordination
equilibria in the system, which are concentration-dependent: a)
the catalyst and nitroarene amounts were doubled; b) the Phen,
Et3N and CH3CN amounts were not changed; c) the amount of
HCOOPh was such that its excess with respect to the
stoichiometric amount required was unvaried. Under these
conditions, complete conversion of 1a and selectivity (GC
analysis) into 2a were achieved.
With the best conditions in our hands, the reaction scope was
explored (vide infra Table 9, conditions B). A discussion of the
obtained results is given later in this paper. Here it suffices to
say that excellent results (isolated yield > 90%) were obtained in
several cases and the reaction also gave good results with
several more substrates. However, only moderate yields were
obtained in the presence of sensitive groups such as an
aldehyde or a pyrrole ring and a few substrates gave mixtures of
products.
For this reason, we decided to make a further effort to optimize a
new set of experimental conditions that could allow the reaction
to be performed at a lower temperature, in the hope that this
feature would have allowed to deal also with very reactive
functional groups on the molecule. A further goal was to
decrease the amount of catalyst in order to make the procedure
more interesting even from an industrial point of view.
The identity of the best pre-catalyst, a chloride-containing
complex, was somehow unexpected. It has long been known
that when Pd/Phen complexes are employed as catalysts for the
related carbonylation of nitroarenes to methyl arylcarbamates,
non-coordinating anions must be used.[2c] Specifically, even if
less than one equivalent of chloride with respect to palladium
26]
can have a promoting effect under certain conditions,[12b,
larger amounts strongly inhibit the catalytic activity.[12a] This is at
least in part due to the very low solubility of Pd(Phen)Cl2 in most
solvents. Apparently, CH3CN solubilizes the complex enough at
the reaction temperature that the low solubility is no longer a
problem.
Control experiments (Table S1, Supporting Information) show
that removing any of the components of the catalytic system,
palladium complex, phenyl formate, Et3N, from the reaction
mixture completely quench the catalytic activity.
Especially important is the comparison with the result obtained
by working under CO pressure (in an autoclave) and in the
absence of any formate (Table 4, entry 10). Both the selectivity
and, more markedly, the conversion are lower for the reaction
run under CO, showing that the use of a CO surrogate can be a
first-choice strategy and not only a second option when an
autoclave and a source of pressurized CO are not available.
While an autoclave is not necessary, a sealed vessel is still
required. A reaction was performed in a standard Schlenk flask
connected to a dinitrogen line. The temperature had to be
lowered at the reflux temperature of the mixture (82 °C). The
reaction was very slow (Figure 2): after 9 h conversion was just
8%. However, when further 200 μL of HCOOPh were added, the
reaction accelerated and a 19% conversion was reached. As a
reference, performing the reaction at the same temperature into
a pressure tube, 65% conversion and 97% selectivity were
achieved after 9 h (Figure 2). These experiments prove that the
low reactivity in an open vessel system is not due to a
deactivation of the Pd catalysts but to a very low CO
concentration in the liquid phase. They also show that pressure
tubes are crucial to maximize the amount of CO in solution.
We initially tested the effect of lowering the catalyst amount
when working at 120 °C (Table 5).
Table 5. Effect of catalyst loading in the Pd(CH3CN)2Cl2 catalyzed reductive
cyclization of 1a using phenyl formate as a CO source.[a]
Entry
Catalyst loading [mol-%]
1
1a Conv. [%][b]
2a Select. [%][b]
98
>99
1
2
3
4
5
6
0.33
0.2
99
99
97
27
3
>99
98
0.1
87
0.02
0.01
21
48[c]
[a] Reaction conditions: 0.27 mmol 1a, 5 mol-% Phen, 200 µL HCOOPh, 40 µL
Et3N, CH3CN 10 mL, at 120 °C for 3 h. Reactions were performed into heavy-
wall glass tubes. [b] Conversion and selectivity were determined by GC
analysis using biphenyl as the internal standard. [c] Value affected by a large
experimental error because of the low conversion value.
Gratifyingly, the catalyst amount can be decreased to 0.2 mol-%
essentially without affecting the results (Table 5, entry 3) and the
catalyst is active even at lower loadings, although with a
decreased performance (Table 5, entries 4-6).
Figure 2. Reaction performed in a Schlenk flask under dinitrogen atmosphere
and comparison reaction run in a sealed pressure tube. Reaction conditions:
0.27 mmol 1a, 1 mol-% Pd(CH3CN)2Cl2, 20 mol-% Phen, 1 equiv. Et3N (40 µL),
500 μL of HCOOPh, in 10 mL CH3CN. Reaction was conducted at reflux
temperature (82 °C). Conversion and selectivity were determined using GC
(biphenyl as internal standard).
5
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
When we attempted to further lower the temperature to 100 °C
Neat DMF is clearly inferior to neat CH3CN as a reaction
we encountered a problem with the reproducibility of the reaction. medium (Table 7, entries 1 and 6). However, lower amounts of
Considering that this may have been due to an insufficient rate
of CO generation (see also later), we tested the effect of
increasing the amount of base (Table S2) and found that use of
2.5 eq. of Et3N with respect to the substrate (100 µL absolute
amount) not only improved conversion and selectivity, but also
solved the reproducibility problem.
The ligand identity plays a very important role in most catalytic
systems and phenanthrolines substituted with electrondonating
substituents affords better results than unsubstituted
phenanthroline in several cases in the field of nitro compounds
reduction by CO.[3p, 3u, 5b, 27] Moreover, the ligand-to-metal ratio is
also very important and may vary between differently substituted
phenanthrolines.[28] Thus, several experiments were run
employing four different phenanthrolines. The most significant
data from the point of view of evidencing the differences (the
reaction time was lowered to 2 h to enhance them) are reported
in Table 6. More data is reported in the Supporting Information
(Tables S3 and S4).
DMF in CH3CN can be highly beneficial on the reaction rate, with
1 mL DMF allowing an increase in conversion from 60% in neat
CH3CN to 92% in the 9:1 CH3CN/DMF mixture. The selectivity in
indole was less sensitive to the reaction medium, but the best
result was again obtained at the 9:1 ratio. It may be noted that
DMF was the ideal solvent under the reaction conditions
reported by Davies for the same cyclization, but with gaseous
CO.[3d]
Yet, such a large effect on conversion of just 10 % DMF was
unexpected and will be worth to be investigated even in other
reactions.
Further experiments were also run at 80 °C and 60 °C (Tables
S6 and S7). A full conversion and virtually quantitative selectivity
could still be obtained at 80 °C, but the catalyst loading had to
be increased to 1 mol-% and 5 eq. of Et3N were necessary to
decompose phenyl formate at an acceptable rate. Thus, we
decided not to develop these reaction conditions any further.
However, working at less than 100 °C is clearly possible if this is
advisable due to stability problems of the substrate or product.
One point that is virtually never analyzed in the literature on CO
surrogates is the effect of the level of filling of the tube or flask
employed to perform the reaction, which influence the
headspace volume. As a matter of fact, neither for traditional
reactions nor for reactions run in an autoclave, where the
gaseous reagent is usually present in a large molar excess or is
continuously refilled, this parameter is important. However, in
the case of reactions in which CO is generated during the
reaction, the headspace volume can have a very relevant role.
Indeed, one must consider that, contrary to a reaction run under
pressure, in which CO is initially present only in the gas phase
and gradually dissolves in the liquid phase, here CO is
generated directly in the liquid phase, but can then both escape
into the gas phase and reenter the liquid phase from the latter.
We have already shown above (Figure 2 and associated text)
that if the reaction is run in a traditional Schlenk flask connected
to a dinitrogen line, the reaction is much slower and stops well
before the starting nitro compound is fully consumed. This
observation evidences that some of the liberated CO can
probably be intercepted by the palladium catalyst when still in
the liquid phase, but under these conditions most of it escapes
into the gas phase and is recaptured only to a small extent, if
any. Likely, performing the same reaction in a closed vessel of
large volume would produce a similar outcome. Overall, one
must consider that different and competing effects arise when
decreasing the solvent amount keeping all of the rest constant.
1) The decreasing of the solvent will result in a more
concentrated solution and this should have a positive effect at
least on the reaction rate.
Table 6. Effect of ligand identity in the Pd(CH3CN)2Cl2 catalyzed reductive
cyclization of 1a using phenyl formate as CO source.[a]
1a
[%][b]
Conv.
2a Select.
[%][b]
Entry
Ligand
Phenanthroline
84
97
1
2
3
4
4,7-Dimethoxyphenanthroline
3,4,7,8-Tetramethylphenanthroline
4,7-Dimethylphenanthroline
76
67
74
97
93
93
[a] Reaction conditions: 0.27 mmol 1a, mol. 1 mol-% Pd(CH3CN)2Cl2, 2.5 mol-
% ligand; 200 µL HCOOPh, 100 µL Et3N, CH3CN 10 mL, at 100 °C for 2 h.
Reactions were performed into heavy-wall glass tubes. [b] Conversion and
selectivity were determined by GC analysis using biphenyl as the internal
standard.
Over a range of ligand ratios and reaction times, phenanthroline
was confirmed to be the best ligand.
Even the use of different phosphates as inorganic bases in place
of Et3N gave worse results (Table S5), but during attempts to
better solubilize them in the reaction mixture by using DMF as a
solvent or co-solvent we found an unexpected solvent effect
(Table 7).
Table 7. Effect of DMF addition in the Pd(CH3CN)2Cl2 catalyzed reductive
cyclization of 1a using phenyl formate as CO source.[a]
Entry
Volume DMF [mL]
10
1a Conv. [%][b]
2a Select. [%][b]
25
86
1
2
1
48
92
86
60
60
15
93
97
97
95
94
92
2
3
2) The corresponding increasing of the headspace volume will
result in a lower average CO pressure during the reaction and a
less effective CO utilization. This may slow down the reaction
and possibly also cause the deactivation of the catalyst or the
onset of competing reactions that do not use CO (e.g. a Heck
reaction in the present case when substrates containing aryl
bromide moieties are employed).
3) Increasing the solvent amount should have the opposite effect,
but much caution should be exerted in not leaving a too small
headspace volume because this may cause the onset of a too
high pressure and the explosion of the glass vessel. The
0.5
0.1
0
4
5
6
7[c]
1
[a] Reaction conditions: 0.27 mmol 1a, 0.2 mol% Pd(CH3CN)2Cl2, 5 mol%
Phen, 200 µL HCOOPh, 100 µL of Et3N, at 100 °C in CH3CN + DMF (total
volume 10 mL), for 2 h. Reactions were performed into heavy-wall glass tubes.
Conversion and selectivity were determined using GC (biphenyl as the internal
standard). [c] Na3PO4 was employed as a base in place of Et3N.
6
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
maximum amount of phenyl formate we employed was chosen
so that even in the worst scenario, i.e. complete decomposition
of the formate with no consumption of the produced CO, the
pressure inside the tube would not exceed 10 bar even at the
highest temperature.
Obviously, a strong dependency of the reaction outcome on the
solvent and vessel volumes would detract from the general
applicability of our protocol. Thus, we tested the effect of
decreasing the solvent amount (Table 8)
Table 9. Substrate scope under different experimental conditions.
Yield[a]
B
Entry
1
Substrate
Product
A
C
95
62[b,c]
99[b]
(99[b]
)
Table 8. Effect of the solvent volume in the Pd(CH3CN)2Cl2 catalyzed
reductive cyclization of 1a using phenyl formate as CO source.[a]
92
2
3
4
5
Entry
Volume CH3CN [mL]
10
1a Conv. [%][b]
2a Select. [%][b]
60
94
1
2
3
87
96
5
67
67
96
90
2.5
[a] Reaction conditions: 0.27 mmol 1a, 0.2 mol% Pd(CH3CN)2Cl2, 5 mol%
Phen, 200 µL HCOOPh, 100 µL of Et3N, at 100 °C for 2 h. Reactions were
performed into heavy-wall glass tubes. Conversion and selectivity were
determined using GC (biphenyl as the internal standard).
90[d]
Reducing the solvent from 10 to 5 and even 2.5 mL (resulting in
an increase of the headspace from 13 to 18 and 20.5 mL
respectively) had a small, whether not negligible, effect, showing
that the optimized protocol tolerates quite large variations in the
solvent and headspace volumes without compromising the
outcome of the reaction.
On the other hand, a strict exclusion of air while assembling the
reactor (see Experimental for a description of the employed
protocol) and use of dry solvents is essential to obtain good
results. When a reaction was run under the conditions or entry 1
in Table 5, but assembling the reactor in the air, the conversion
dropped from 98 to 10 %. Clearly, the catalytically active species
is irreversibly destroyed by even low amounts of dioxygen.
Conversely, the addition of 50 L of water, in the absence of air,
had little effect on the conversion, but the selectivity in indole
dropped to 34% and larger amounts of 4a were formed. Thus,
use of dry solvents and strict exclusion of air are key points for
obtaining good and reproducible results.
79[e]
6
7
8
50[f] 65
73 91
45 64
9
94
10
11
Substrate scope investigation
95[g]
With this new set of optimized reaction conditions in our hands,
we reinvestigated the substrate scope of the reaction, focusing
on those substrates that either had not given high yields or had
failed at all to give an isolable product under Conditions B.
Several new substrates were also explored. To ensure that even
less reactive substrates would give complete conversion under
the same set of experimental conditions, a 1 mol-% catalyst
loading was employed and the reaction time increased to 6 h.
Results are shown in Table 9, Conditions C. Pure E or Z isomer
of the starting nitrostyrenes or a mixture of the two was
employed as a starting material depending on the synthetic
procedure employed to prepare them.
A first glance to the results reported in the table immediately
evidences that despite good results can be obtained in some
cases under Conditions A, the obtained yields are always
markedly lower than those achievable by the use of phenyl
formate under both Conditions B and C. The results obtained
under Conditions A will not be further discussed. As already
95[g]
12
13
51 77
52[b] 90[d]
14
15
95[b]
7
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
96[b]
[k]
16
-
34
[a] Isolated yields unless otherwise noted. Reaction conditions: Conditions A:
0.27 mmol 1, mol. 1 mol-% [Pd(Phen)2][BF4]2, 1 mol-% Ru3(CO)12, 20 mol-%
Phen; mol. 40 µL (0.29 mmol) Et3N; in butyl formate (10 mL), at 180 °C for
10h. Conditions B: 0.54 mmol 1, 1 mol-% Pd(CH3CN)2Cl2, 2.5 mol-% Phen,
240 µL (2.2 mmol) HCOOPh, 40 µL (0.29 mmol) Et3N, in CH3CN (10 mL),
140 °C for 3 h (unless otherwise noted). Conditions C: 0.54 mmol 1, 1 mol%
Pd(CH3CN)2Cl2, 5 mol% Phen, 260 µL (2.38 mmol) HCOOPh, 100 µL (0.72
mmol) Et3N, in CH3CN + DMF (9+1 mL), at 100 °C for 6 h. Reactions were
performed into heavy-wall glass tubes. [b] Determined by GC analysis using
biphenyl as the internal standard. [c] 3a, 20 % yield, was also obtained. [d]
Reaction time 4h. [e] Reaction time 8h. [f] Reaction time 12h. [g] Reaction time
5h. [h] Reaction time 6h. [j] Yield measured by 1H NMR with 2,4-dinitrotoluene
as an internal standard. [k] Complex mixture of products. [l] 0.27 mmol of 1y or
1z was employed as substrate.
71 96
17
18
37 98[g]
78
19
20
21
45[j] 60
mentioned, Conditions B gave excellent results in many cases,
but results are not fully satisfactory for some substrates and in a
few cases the reaction yielded mixtures of products in which the
target indole was only a minor component. Conditions C finally
constitute a marked improvement and allowed to get higher
yields of the products in all cases in which a comparison with
Conditions B was made.
70 92[h]
69 94
22
Coming to the specific functional groups present on the reagent,
esters (1a,c-f) derived from 2-nitrocinnamic acid were
successfully cyclized providing the corresponding indoles in
good to excellent yields. Remarkably, a bromine substituent on
the aryl ring (1f) was well tolerated despite the fact that the
presence of the nitro group on the same aryl ring should activate
it towards oxidative addition to a low-valent palladium complex.
Very likely, the presence of CO inhibits the C-Br oxidative
addition.
[k]
-
71
23
24
25
8[j] 72
91
Amide 1i afforded somewhat lower yields under Conditions B,
but a > 90% yield could be reached under Conditions C. Moving
the ester group to the alpha position of the starting nitrostyrene
(1j) lead to a less satisfactory yield. This is the only substrate
among those tested for which markedly better results have been
52 98
26
previously obtained by using gaseous CO, albeit with much
27[l]
29]
higher catalyst loadings (10-12 mol-%).[3f,
Prolonging the
78
reaction time did not improve the yield, indicating that the
catalyst deactivates before the end of the reaction. On the other
hand, a keto group (1g), and even a free aldehydic group (1h)
were tolerated. Cyclization of 1h to 2h by any means has been
previously reported only once, using gaseous CO as reductant,
using 10 mol-% of a rhodium complex as catalyst and affording
a lower (52%) yield with respect to that achieved by us under
Conditions C.[3f] Worthy of note, the synthesis of 2h would not
be possible by any alternative reaction employing an arylamine
as starting material or in which the arylamine is formed as an
intermediate.
28[l]
61
90
29[l]
30
[k]
[k]
-
-
Indoles having a cyano group in the position 3 and an aryl group
in the position 2 (2k-2m) were obtained in excellent yield
independent of the presence of electronwithdrawing or donating
groups on the aryl ring. When a sensitive pyrrole ring was
present in the place of the aryl ring (2n), the yield was quite
lower under Conditions B. However, moving to Conditions C
markedly improved the results.
The Z/E nitrostilbene (Z/E-1b) mixture gave much better yields
under Conditions B than under Conditions A and even better
results could be obtained under Conditions C employing the
[k]
-
31
[k]
[k]
-
-
32
33
[k]
[k]
-
-
separate isomers. Under these conditions,
a
complete
8
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
conversion was observed for both isomers. However, if the
reactions in entries 16 and 17 were run for just 3 h instead of 6,
a 64 % conversion of E-1b was observed, but the conversion of
Z-1b was just 29%. This parallels the lower reactivity of Z isomer
with respect to the E one already observed and discussed above
for the [Pd(Phen)2][BF4]2:Ru3(CO)12:Phen catalytic system with
n-butyl formate. However, at variance with the results obtained
at 180 °C, no isomerized nitrostilbene was detected in the
unreacted substrate, confirming that the previously observed
nitrostilbene isomerization was due to the high temperature only.
This observation has also a mechanistic implication. If any
isomerization of nitrostilbene had been observed even at a
from the reaction mixture. A few double cyclizations of non-
symmetrically substituted bis-(2-nitrophenyl)dienes to give 2-2'
indoles have been reported in the literature,[3h] but the reactions
were always run to complete conversion of the starting dinitro
compound. That the reaction is faster on the side of the cyano-
substituted olefin is not surprising, given what discussed earlier
in this paper, but the observation that the reactivity difference is
large enough to allow for a selective monocyclization is, to the
best of our knowledge, an unprecedented observation. This
possibility paves the way to following reactions in which the
unreacted nitro group is converted into different functionalities,
such as an amino or imino group.
temperature low enough that a thermal process can be excluded, The only substrates for which our reaction failed to give clean
this would have implied that the initial nitroarene activation were
reversible. That no isomerization of the starting material is
observed whereas isomerization clearly occurs during the
cyclization is a clear proof that the initial substrate activation by
the catalyst is irreversible. This fact may not be surprising, but,
to the best of our knowledge, has never been stated before.
Nitrostilbenes substituted with electronwithdrawing groups on
either or both aryl rings (1p, 1q, 1s, 1t) gave excellent results
and the same apply to the 2-pyridyl substituted compound 1x.
The case of 1q is especially interesting. Two nitro groups are
present on the substrate and the one in the para position with
respect to the vinyl substituent should be the most reactive for
steric reasons. Indeed, when 2,4-dinitrotoluene is carbonylated
to the corresponding methyl dicarbamate by Pd/Phen catalysts
and the reaction is stopped before completion, the nitro group in
the position 4 is carbonylated to a quite larger extent than that in
the position 2 over a range of experimental conditions.[12c] The
fact that in the present case the reactions proceeds selectively
on the ortho nitro group supports the previous proposal that the
olefin groups may coordinate to palladium before the reduction
occurs, facilitating and addressing the reaction.[2d, 19]
Electrondonating groups on the aryl ring are known to deactivate
the nitro group in these reactions (vide supra). Yet, good to
excellent results could be obtained not only in the presence of
the mildly electrondonating methyl group (1o), but even in the
presence of dimethylamino (1r) or two alkoxy (1u, 1v) groups. In
this case, use of Conditions C is required or anyway highly
preferable.
Noteworthy, 2-phenyl-6-azaindole 2w was synthetized in very
high yield. Azaindoles are a class of nitrogen rich molecules that
found broad applications in many fields such as pharmaceuticals
and agrochemicals and also exhibits peculiar photophysical
properties.[30]
isolable products under all conditions are those containing a free
amino (1aa) or phenolic (1ab) groups and those in which the
double bond is part of an allylic acetate (1ac) or allylic alcohol
(1ad) moiety. The reason for the formation of a mixture of
compounds in the first case may be due to the easy reaction of
anilines with nitroarenes in the presence of Pd/Phen catalysts
and CO to give diarylureas.[26] Indole 2aa may however be easily
obtained by reduction of 2q[32] and this indirect route is
preferable in any case. Indeed, preparation of the starting 1aa
following the only published procedure[33] proceeded in our
hands in poor yields, whereas synthesis of 1q can be easily
performed by reacting dinitrotoluene with benzaldehyde in the
presence of a base and the reagents are very cheap and
available on a multi-ton scale if desired. Allylic acetates are
typical substrates for the generation of -allyl complexes and this
is the probable entry to different reactions for 1ac. Activation of
allylic alcohols is usually more difficult, but, on the other hand,
alcohols can enter a carbonylation reaction of nitroarene to give
carbamates, a reaction that is again very efficiently catalyzed by
Pd/Phen complexes.[12a, 34] Note anyway that these compounds
are very difficult substrates for this kind of reaction. To the best
of our knowledge, cyclization of 1aa to 2aa or 1ab to 2ab has
never been reported by any means, whereas that of 1ab to 2ab
and 1ac to 2ac has been reported in just one case to occur with
CO and a rhodium catalyst in low yields (11 and 26%
respectively) despite the use of a high rhodium loading (10
mol-%).[3f]
Additionally, the presence of a substituent in the second position
ortho to the nitro group also blocked the reaction. The sensitivity
of this kind of cyclization to steric hindrance in this position is a
known problem[3f] and our protocol did not solve it.
It worth noting that in several cases the achieved yields are
higher than those previously reported in the literature working
under CO pressure in an autoclave. For example, the best yields
obtained in the case of compounds 2a,[3f] 2h,[3f] 2x, [9b] 2w[35] and
2p[5b] were 83 %, 52 %, 63 %, 87 % and 55 %, respectively. This
fact clearly demonstrates that our protocol is an effective
alternative to those based on the use of pressurized CO and not
simply a second-choice strategy when use of the latter is not
possible.
Furthermore,
double
cyclization
of
2,6-bis((E)-2-
nitrostyryl)pyridine (1y) led to the formation of 2,6-bis(2’-
indoyl)pyridine in good yields (2y). Compounds of this type have
been successfully used as ligands in metal-based
phosphorescent and electroluminescent compounds.[31]
When the double cyclization of a non-symmetrically substituted
conjugated diene was attempted under Conditions C (0.27 mmol
of substrate were used instead of 0.54 to keep the same number
of nitro groups as for the other reactions), after the standard 6 h
reaction time the main product, isolated in a 61 % yield was that
in which cyclization had only occurred on the cyano-substituted
side (2z'). The recovered unreacted diene accounted for the rest
of the mass balance. The doubly-cyclized bis-indole (2z") could
be obtained in high yield by increasing the reaction time to 16 h.
Interestingly, the product was isolated by simple precipitation
To further assess the synthetic value of our protocols, the
scalability of our reaction under both Conditions
Conditions C was investigated.
B and
A 14-fold scaled-up reaction (7.5 mmol of the starting material)
of 1a was performed under Conditions B to give 2a in 84 %
isolated yield.
More interestingly, an even larger scale cyclization of 1b (2.0
g, .8.9 mmol, 16.5-fold scale-up) was performed under
9
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
conditions similar to Conditions C, but with half the amount of
catalyst (0.5 mol-%) and less solvent (45+5 mL CH3CN/DMF, 5-
fold increase). At the end of the reaction, metallic palladium had
been precipitated and the product could be isolated as an
analytically and spectroscopically pure compound (1.63 g, 96 %
yield) by a simple precipitation and extraction procedure, without
the need for any chromatographic purification.
Application to other reductive cyclization reactions of nitro
compounds
In order to test the general applicability of our protocol, it was
applied to other related reactions. Four reactions were chosen,
featuring respectively the functionalization of an aromatic C-H
bond,[3r] the formation of a six-membered ring,[3s] the cyclization
of
a
β-nitrostyrene,[3q] and an inter-molecular cyclization
reaction[3m, 3t] (Scheme 1, paths f, b, h, and k respectively and
Scheme 4). The protocol worked in all cases even if
experimental conditions were not optimized for these specific
reactions. Since publication of our preliminary communication on
this topic, the synthesis of oxazines has already been the object
of a specific study and > 90% yields could be obtained in many
cases.[36] Moreover, work is in progress on the synthesis of
carbazoles and conditions that allow obtaining these products in
excellent yields have been identified.[37] Thus, the use of phenyl
formate is not limited to the synthesis of indoles.
Scheme 4. Synthesis of other N-heterocycles (isolated yields are
reported).
literature, several papers claim that phenyl formate can be
decarbonylated by bases,[25a, 25f, 38] However, one work reported
that Pd(OAc)2 in combination with P-based ligands is also able
to perform this transformation.[25e]
Kinetic study of phenyl formate decarbonylation and
reaction mechanism proposal
In the aim of gaining insight into the reaction mechanism, the
decarbonylation step was separately studied. In the previous
Figure 3. Kinetic analysis of the Et3N-mediated decarbonylation of phenyl formate in refluxing CH3CN (for the experimental procedure see Supporting
Information).
10
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
In order to discriminate between this two kind of activations,
HCOOPh was reacted with either 1 equiv. of Et3N or a catalytic
amount of Pd(CH3CN)2Cl2 at 140 °C in a pressure tube (Figures
S5 and S6). In the first case, a complete conversion to CO and
phenol was achieved in 3 h, but no reaction occurred in the
presence of Pd(CH3CN)2Cl2. A control experiment carried with
HCOOPh alone in CH3CN, ruled out thermal decarbonylation. In
addition, phenanthroline, which may also act as a weak base
was found to be inactive in the same transformation. All these
tests are in accord with the control experiments previously
mentioned (Table S3). To track the course of the transformation
over time, the reaction of HCOOPh with Et3N was performed in
refluxing CH3CN and monitored by FT-IR spectroscopy. A
decreasing of the two peaks corresponding to the C=O bond
was observed over time, accompanied by a parallel increasing
of the absorption at 3400 cm-1 due to OH stretching in phenol.
Noteworthy, no other C=O stretching bands were detected at all
stages of the reactions, ruling out the formation of stable
intermediates in the reaction, which may affect the kinetic
measurement. The reaction rate was found to be first order in
both HCOOPh and Et3N (Figure 3), with a second order constant
k = 5.1410-2 M-1min-1 (in CH3CN at 82 °C). Under the same
reaction conditions, n-butyl formate was unreactive, indicating a
different activation mode.[39]
When the reaction is complete, no oxidant is any longer present
and palladium is reduced to palladium black. The appearance of
a black ring on the glass tube is a typical indication that the
reaction is complete or close to completion.
Finally, the mechanism of the reductive cyclization was
considered. Nitrosoarenes are generally considered active
intermediates in the reduction of nitro compounds by CO. The
intermediate formation of nitrosoarenes in the synthesis of
indoles is strongly supported by the very efficient synthesis of
oxazines under the same experimental conditions (Scheme 4).
We were not able to intercept the nitrosoarene derived from 1a
by addition of a conjugated diene to the reaction mixture.
However, this is not surprising because of the high reactivity of
nitrosoarenes with olefins[40] and the higher rate of intra-
molecular reactions with respect to inter-molecular ones. As a
matter of fact, all previous attempt to obtain o-nitrosostyrenes
met with failure.[3d, 41]
In the light of the previous experiments and of the results here
reported, the following mechanism can be proposed for the
present process (Scheme 5). Two independent reactions occur:
the phenyl formate decarbonylation and the reductive cyclization.
After CO is formed by the Et3N-mediated decarbonylation
reaction, it reduces the starting Pd(II) complex to give a Pd(0)
phenanthroline complex, likely coordinated to one or two CO
molecules. As discussed above, the starting nitrostyrene
appears to coordinate to the zerovalent palladium complex and
stabilize it before a reaction with CO results in the formation of
o-nitrosostyrene A. The latter undergoes an intra-molecular
cyclization affording nitronate B, which is converted by a 1,5-H-
shift to nitronate C. The latter isomerizes to the N-hydroxyindole
D that is finally deoxygenated by a second CO molecule leading
to the formation of the indole and the regeneration of the
palladium active species. This proposal is supported by
experimental[3e] and theoretical evidence.[42]
Scheme 5. Proposed reaction mechanism.
Conclusion
In conclusion, we have presented here an efficient, convenient,
and general synthetic procedure for the synthesis of nitrogen
heterocycles from nitro compounds, based on the use of CO
surrogates. The protocol affords the desired products with
selectivities and yields that are most often higher than those
previously reported using pressurized CO. Additionally, the
catalyst (just 1 mol-% or even lesser), the ligand and the CO
surrogate are commercially available., thus increasing the user
friendliness and the economics of this protocol both on a
laboratory and on a larger scale. Worth of note, in at least one
case, the indole could be isolated in a pure form and with
minimal losses by a simple precipitation/extraction procedure,
avoiding the need for a chromatographic purification that would
limit the scale on which the reaction can be performed and
significantly increase the overall cost of the process.
The decomposition of formates has been investigated. Whereas
the presence of a ruthenium catalyst is required to decompose
alkyl formates, at least in the absence of very strong bases,
organic bases of medium strength are sufficient to catalyze the
decomposition of phenyl formate. The kinetics of the reaction
was investigated and
a first order dependence of the
decomposition rate on the base concentration was observed,
stressing the need for its presence.[43] [5c, 44]
The apparatus employed and tricks to avoid problems with the
O-rings are carefully described in the Experimental Section. This
should render it easier to other groups to reproduce our protocol
or to extend it to other reactions.
Note that during the reaction the oxidation state of palladium
shuttles between 0 and II. Pd(0) complexes with nitrogen ligands
are not stable in the presence of CO, but until the substrate is
present, coordination of the double bond and prompt oxidation to
Pd(II) prevent the decomposition of the catalyst. Excess
phenanthroline also plays an important role in this respect.
11
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
1,10-Phenanthroline (Phen) was purchased as hydrate (Sigma-Aldrich).
Before use, it was dissolved in CH2Cl2, dried over Na2SO4 followed by
filtration under a dinitrogen atmosphere and evaporation of the solvent in
vacuo. Phen was weighed in the air but stored under dinitrogen to avoid
water uptake. All the Pd precursors employed in this work were prepared
starting from commercially available PdCl2 following procedures reported
in the literature (see Supporting Information). Ru3(CO)12 and the starting
nitrostyrenes were synthesized as described in the Supporting
Information. All other reagents were purchased from Sigma-Aldrich or
Alfa-Aesar and used without further purifications. 1H-NMR and 13C-NMR
spectra were recorded at room temperature on a Bruker Avance DRX
300, Bruker Avance DRX 400 or Bruker Avance 600. Chemical shifts are
reported in ppm relative to TMS. Gas-chromatographic analyses were
performed using a Shimadzu 2010 gas chromatograph equipped with a
Supelco SLBTM-5ms capillary column. A standard analysis involves the
preparation of a sample solution in CH2Cl2 (conc. 0.1 mg/mL-1 calculated
with respect to biphenyl used as the internal standard). GC-MS analyses
were recorded using a Thermo Fisher ISQTMQD Single Quadrupole GC-
MS equipped with a ZB-1MS 60m column. Elemental analyses were
performed on a Perkin Elmer 2400 CHN elemental analyzer. IR spectra
were registered on a Varian Scimitar FTS-1000.
Experimental Section
Apparatus:
In our early experiments, we employed a 18 mL ACE Pressure tube with
"front" (inner) seal bushing. This tube has a female-threatened glass end
onto which a male-threatened PTFE stopper with a rubber O-ring is
placed (Figure S1 left). A weak point of this setting is the rubber O-ring.
Those enclosed with the tube or others made of Viton rapidly degraded
even during a single run at the temperatures and with the solvents
employed in this work, resulting in a pressure drop and possible
contamination of the solution by rubber oligomers. Moving to very
expensive Karlez O-rings did not solve the problem, neither did
employing the tube type with "back" (outer) seal. However, it turned out
that the O-ring is not required and simply removing it and hand tightening
the stopper provides a good seal. By doing this, the tube can be
employed many times. However, upon prolonged (> 50 reactions) use,
the PTFE stopper slowly deformed and the tube started to leak. Since the
stoppers are not sold separately from the tubes, in the second part of this
work we moved to Fisher Scientific heavy walls (2.5 mm) borosilicate
glass tubes with Duran PTB screw caps completed with PTFE protected
seals (Figure S1 left, Supporting Information). In this case, the seals can
be changed independently from the rest of the apparatus. The
disadvantage of this apparatus for the occasional user is that the tubes
are sold with an open bottom and need to be sealed at the desired length
by an experienced glass blower not to compromise the glass strength
(the volume of our tubes is 23 mL). Both setups worked well and the
choice of which to employ should be based on the type of use foreseen.
However, it is important to stress that normal glass should not be
employed for these reactions and that care should always be taken in
opening the (cold) tubes because residual carbon monoxide pressure is
usually present in the tube at the end of the reaction. Heating was
performed either by means of an oil bath or by inserting the tubes in a
preheated aluminum block in which eight 22.5 mm wide holes had been
made to fit the pressure tubes (Figure S1 right). All reactions must be
conducted in a well-ventilated hood and a safety screen or other
protecting shield should be present during the reaction. We never had
problems while performing our reactions, but care must be taken that
scratches are not formed on the glass surface that may compromise its
resistance. NOTE: ACE 18 mL pressure tube were employed for all
reactions involving the use of alkyl formates and for reactions run under
Conditions B or during the optimization performed to identify them
(results reported in Tables 1, 2, 4, S1, Figures S2-S4, and results for
Conditions A and B in Table 9). The following optimization (results in
Tables 5-8, S2-S8, and results for Conditions C in Table 9) were
performed in Fisher Scientific heavy walls borosilicate 23 mL glass tubes.
General methods for catalytic reactions
In order to avoid weighing very small amounts of the catalysts, stock
solutions (under dinitrogen) of all ruthenium and palladium catalysts were
prepared in the appropriate alkyl formate or CH3CN (in the case of phenyl
formate). For catalytic runs that required an amount of Phen less than 10
mg, a stock solution was prepared. For a typical reaction, the pressure
tube (see Note in the apparatus section) containing a magnetic stirring
bar was charged with the substrate and Phen (when the amount of the
latter is larger than 10 mg) and the tube was then immediately placed in
a Schlenk tube with a wide mouth. The tube was evacuated and filled
with dinitrogen three times to remove any air. The appropriate volume of
stock solutions of the catalysts, Phen (if not added as a solid), HCOOPh
(if required), the solvent (total solvent amount was 10 mL) and the base
were added in this order. It is important that the base is added at the end
because in some cases CO evolution may start immediately after its
addition even at RT. Then, the reaction vessel was immediately plugged
with a PTFE or PTFE-lined screwed-cap and the mixture let to stir for 10
min. This time is required to allow for the coordination of Phen to
palladium. Formation of a precipitate of Pd(Phen)Cl2 may be observed
depending on the catalyst identity and concentration. This complex is
little soluble at RT, but dissolves at higher temperatures and when it
reacts to give the catalytically active species. Heating the mixture without
waiting for this time results in a lower activity apparently because
reduction of uncoordinated palladium by liberated CO causes the
formation of inactive aggregates. The tube was then either immersed in a
pre-heated oil bath or placed in a preheated custom-made aluminum
block with holes fitting the tube width (Figure S1). Reagents amounts and
other experimental conditions are reported in the captions to the tables.
At the end of the reaction, the pressure tube was lifted and let to cool to
room temperature. Then, the screw cap was carefully removed, the
excess of CO was vented and the content analyzed by GC (biphenyl as
an internal standard), GC-MS or the solvent and excess formate were
evaporated and the residue subjected to column chromatography (silica
gel) using hexane/CH2Cl2 or hexane/AcOEt as the eluent (in both cases,
from 1 to 2 % of Et3N was added to the eluent to partially deactivate
acidic sites of silica gel. Failure to do this causes extensive
decomposition of most indoles).
General procedure:
Unless otherwise stated, all reactions and manipulations were performed
under a dinitrogen atmosphere using standard Schlenk apparatus. All
glassware and magnetic stirring bars were kept in an oven at 120 °C for
at least two hours and let to cool under vacuum before use. CH3CN was
dried by distillation from CaH2. DMF was dried by distillation over CaH2,
under reduced pressure at 60 °C. Dried solvents were stored under a
dinitrogen atmosphere. Butyl and methyl formate were dried over MgSO4,
filtered by cannula techniques and then distilled under a dinitrogen
atmosphere. Benzyl formate was degassed by three freeze-pump-thaw
cycles but not distilled. Et3N, DIPEA (N,N-diisopropylethylamine), and
DBU (1.8-diazabicyclo[5.4.0]undecen-5-ene) were distilled from CaH2
under reduced pressure. Phenyl formate was purchased form Sigma
Aldrich or prepared following a procedure reported in the literature (see
Supporting Information). Deuterated solvents were purchased from
Sigma-Aldrich: DMSO-d6 (commercially available in 0.75 mL vials under
dinitrogen atmosphere) was used as purchased while CDCl3 was filtered
on basic alumina and stored under dinitrogen over 4 Å molecular sieves.
The same reaction protocol was employed for running the catalytic test in
standard Schlenk glassware.
The large-scale reactions were performed in a 250 mL Fisher-Porter
pressure bottle. The reaction for the synthesis of 2a under conditions B
was scaled up by a factor of 14. All the components of the reaction were
scaled-up and the product was purified by column chromatography as
12
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
described above. The reaction for the synthesis of 2b under Conditions C
was scaled-up 16.5-fold with respect to the substrate (2.0 g, 8.9 mmol)
and Et3N (1.17 g, 1.6 mL, 11.5 mmol), but half the amount of catalyst (0.5
mol-%) and less phenyl formate (3.7 g, 3.3 mL, 30.3 mmol, 3.4 eq
instead of 4.4) and solvent (45+5 mL CH3CN/DMF, 5-fold increase) were
employed. To ensure completion of the reaction, the reaction time was
increased from 6 to 8 h, although the reaction may had reached
completion before. At the end of the reaction, metallic palladium had
precipitated on the walls of the bottle. The solution was then filtered
through a pad of Celite in a Pasteur pipette by cannula technique to
remove even colloidal palladium particles possibly present. The product
was then precipitated with water, dissolved in CH2Cl2 (50 mL) and
washed with a saturated NaHCO3 aqueous solution (3 × 30 mL), brine (1
× 50 mL) and water (1 × 50 mL). The organic phase was then dried over
Na2SO4, filtered and the solvent was evaporated under vacuum to yield
the final product as a white, analytically and spectroscopically pure
52.4 ppm. Elemental analysis calcd for C10H9NO2.: C 68.56; H 5.18; N
8.00, found: C 68.50; H 5.23; N 8.01.
2-Phenyl-1H-indole (2b): Obtained as a white solid (94 mg, 0.49 mmol,
90 % yield) after column chromatography (hexane:CH2Cl2 = 7:3 + 1 %
Et3N). 1H NMR (300 MHz, CDCl3) δ 8.33 (br, 1H), 7.73 – 7.64 (m, 3H),
7.53 – 7.40 (m, 4H), 7.36 (t, J = 7.3 Hz, 1H), 7.25 (dd, J = 13.4, 5.3 Hz,
1H), 7.18 (t, J = 7.4 Hz, 1H), 6.87 ppm (s, 1H). 13C NMR (75 MHz,
CDCl3) δ 138.3, 137.3, 132.8, 129.7, 129.4, 128.1, 125.6, 122.8, 121.1,
120.7, 111.3, 100.4 ppm. Elemental analysis calcd for C14H11N: C 87.01;
H 5.74; N 7.25, found: C 86.85; H 5.80; N 7.30.
Ethyl 1H-indole-2-carboxylate (2c): Obtained as a white solid (94 mg,
0.50 mmol, 92 % yield) after column chromatography (hexane:CH2Cl2 =
3:7 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 8.90 (br, 1H), 7.70 (dd, J =
8.1, 0.9 Hz, 1H), 7.43 (dd, J = 7.9, 1.4 Hz, 1H), 7.33 (ddd, J = 8.3, 7.0,
1.1 Hz, 1H), 7.24 (dd, J = 2.1, 1.0 Hz, 1H), 7.16 (ddd, J = 8.0, 7.0, 1.0 Hz,
1H), 4.42 (q, J = 7.1 Hz, 1H), 1.42 ppm (t, J = 7.1 Hz, 1H). 13C NMR (101
MHz, CDCl3) δ 162.0, 136.8, 127.5, 125.3, 122.6, 120.8, 111.8, 108.6,
61.0, 14.4 ppm. Elemental analysis calcd for C11H11NO2: C 69.83; H
5.86; N 7.40, found: C 69.67; H 5.80; N 7.40.
crystalline solid (1.63 g, 96
% yield), without the need for any
chromatographic purification. Phenanthroline and any other by-product
present in small amounts remained in solution by this procedure.
Elemental analysis calcd for C14H11N: C, 87.01; H, 5.74; N, 7.25, found:
C, 86.69; H, 5.85; N, 7.44.
Concerning the control reactions conducted in the autoclave using
pressurized CO, Phen and the substrate were weighed in the air in a
glass liner that was then placed inside a Schlenk tube with a wide mouth
under a dinitrogen atmosphere. The tube was evacuated and filled with
nitrogen three times to remove any air. The catalyst stock solution
(Pd(CH3CN)2Cl2 in CH3CN) and Et3N were added by volume and the liner
was closed with a screw cap having a glass wool-filled open mouth which
allows gaseous reagents to exchange. The Schlenk tube was immersed
in liquid nitrogen until the solvent froze, evacuated, and filled with
dinitrogen three times. The liner was rapidly transferred to a 200 mL
stainless steel autoclave equipped with magnetic stirring. The autoclave
was then evacuated and filled with dinitrogen three times, charged with
CO at RT and subsequently immersed in a pre-heated oil bath. At the
end of the reaction, the autoclave was quickly cooled with an ice bath
and vented.
t-Butyl-1H-indole-2-carboxylate (2d): Obtained as a white solid (102
mg, 0.46 mmol, 87 % yield) after column chromatography (hexane:AcOEt
= 1:1 + 1 % Et3N). 1H NMR (300 MHz, CDCl3) δ 8.95 (s, 1H), 7.68 (d, J =
8.9 Hz, 1H), 7.42 (d, J = 8.3 Hz, 1H), 7.31 (t, J = 7.6 Hz, 1H), 7.19 – 7.11
ppm (m, 2H). 13C NMR (75 MHz, CDCl3) δ 161.7, 137.0, 129.4, 128.0,
125.4, 122.9, 121.0, 112.2, 108.5, 82.2, 28.7 ppm. Elemental analysis
calcd for C13H15NO2: C 71.87; H 6.96; N 6.45, found: C 71.71; H 6.95; N
6.70.
Methyl 6-methyl-1H-indole-2-carboxylate (2e): Obtained as a white
solid (98 mg, 0.52 mmol, 96 % yield) after column chromatography
(hexane:CH2Cl2 = 3:7 + 1 % Et3N). 1H NMR (600 MHz, CDCl3) δ 8.79 (br,
1H), 7.57 (d, J = 8.2 Hz, 1H), 7.19 (d, J = 11.6 Hz, 2H), 7.00 (d, J = 8.2
Hz, 1H), 3.94 (s, 3H), 2.47 ppm (s, 3H). 13C NMR (151 MHz, CDCl3) δ
162.5, 137.4, 135.7, 126.6, 125.4, 123.0, 122.2, 111.5, 108.8, 51.9, 21.9
ppm. Elemental analysis calcd for C11H11NO2: C 69.83; H 5.86; N 7.40,
found: C 69.67; H 5.80; N 7.40.
Procedure for the FT-IR investigations of the HCOOPh
decarbonylation. Procedure followed under typical catalytic reaction
conditions: HCOOPh (109 μL, 1 mmol) was dissolved in 5 mL of CH3CN
inside a ~ 18 mL ACE pressure tube. Subsequently, the desired reagent
(Et3N or Pd(CH3CN)2Cl2) was added. The reaction vessel was plugged
with a PTFE screw cap and immersed in a preheated oil bath (140 °C) for
3h. At the end of the reaction, the pressure tube was lifted and let to cool
to room temperature. Then, the screw cap was carefully removed and the
content analyzed at FT-IR using a 0.1 mm CaF2 cell for liquid samples.
Methyl 6-bromo-1H-indole-2-carboxylate (2f): Obtained as a white
solid (124 mg, 0.49 mmol, 90 % yield) after column chromatography
(hexane:CH2Cl2 = 4:6 + 1 % Et3N). 1H NMR (600 MHz, CDCl3) δ 8.86 (br
s, 1H), 7.59 (s, 1H), 7.55 (d, J = 8.6 Hz, 1H), 7.27 (d J = 8.6 Hz, 1H),
7.18 (s, 1H), 3.95 ppm (s, 3H). 13C NMR (151 MHz, CDCl3) δ 162.0,
137.4, 127.8, 126.3, 124.5, 123.8, 119.2, 114.7, 108.8, 52.1 ppm.
Elemental analysis calcd for C10H8BrNO: C 47.27; H 3.17; N 5.51, found:
C 47.64; H 3.32; N 5.44.
Procedure followed under refluxing CH3CN conditions (kinetic analysis):
HCOOPh (65.4 μL, 0.6 mmol) was dissolved in CH3CN (3 mL) inside a
Schlenk tube. Subsequently, the desired amount of Et3N was added. The
reaction vessel was immersed into a pre-heated oil bath reaching
refluxing conditions. At the desired time, the Schlenk tube was lifted-up
and after 5 minutes a very small aliquot was analyzed by FT-IR. The
disappearing of HCOOPh was evaluated monitoring the decreasing of
the two C=O stretching bands at 1741 cm-1 and 1761 cm-1.
(1H-Indol-2-yl)(phenyl)methanone (2g): Obtained as a white solid (95
mg, 0.43 mmol, 79 % yield) after column chromatography (hexane:AcOEt
= 7:3 + 1 % Et3N). 1H NMR (300 MHz, CDCl3) δ 9.40 (s, 1H), 8.01 (d, J =
7.0 Hz, 2H), 7.73 (d, J = 8.1 Hz, 1H), 7.63 (t, J = 7.3 Hz, 1H), 7.59 – 7.45
(m, 3H), 7.39 (t, J = 7.6 Hz, 1H), 7.22 – 7.13 ppm (m, 2H). 13C NMR (75
MHz, CDCl3) δ 187.6, 138.4, 137.9, 134.8, 132.7, 129.6, 128.9, 128.2,
126.9, 123.6, 121.4, 113.1, 112.6 ppm. Elemental analysis calcd for
C15H11NO: C 81.43; H 5.01; N 6.33, found: C 81.60; H 5.11; N 6.30.
Characterization of the products of the catalytic reactions Yields and
amounts of the product reported (mg and mmol) refer to the method
affording the highest yield (see Table 9 for yields under different
experimental conditions).
1H-Indole-2-carbaldehyde (2h): Obtained as a white solid (51 mg, 0.35
mmol) after column chromatography (hexane:CH2Cl2 = 4:6 + 1 % Et3N).
1H NMR (400 MHz, CDCl3) δ 9.86 (s, 1H), 9.24 (br, 1H), 7.76 (d, J = 8.1
Hz, 1H), 7.47 (d, J = 8.3 Hz, 1H), 7.40 (t, J = 7.6 Hz, 1H), 7.29 (d, 1.8 Hz,
1H), 7.19 ppm (t, J = 7.5 Hz, 1H). 13C NMR (100 MHz, CDCl3) δ 182.0,
138.3, 136.3, 127.7, 127.5, 123.6, 121.5, 114.7, 112.6 ppm. Elemental
analysis calcd for C9H7NO: C 74.47; H 4.86; N 9.65, found (Conditions
B): C 74.30; H 4.55; N 9.60, found (Conditions C): C 74.07; H 4.85; N
9.28.
Methyl 1H-indole-2-carboxylate (2a): Obtained as a white solid (90 mg,
0.51 mmol, 95 % yield) after column chromatography (hexane:CH2Cl2 =
7:3 + 1 % Et3N). 1H NMR (300 MHz, CDCl3) δ 8.98 (br, 1H), 7.70 (d, J =
8.1 Hz, 1H), 7.43 (d, J = 8.3 Hz, 1H), 7.33 (t, J = 7.6 Hz, 1H), 7.25 (d, J =
8.2 Hz, 1H), 7.16 (t, J = 7.5 Hz, 1H), 3.96 ppm (s, 3H). 13C NMR (75 MHz,
CDCl3) δ 162.85, 137.27, 127.9, 127.5, 125.8, 123.0, 121.2, 112.3, 109.2,
13
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
(1H-Indol-2-yl)(piperidin-1-yl)methanone (2i): Obtained as a white
solid (112 mg, 0.49 mmol, 91 % yield) after column chromatography
(from hexane:CH2Cl2 = 1:1 + 1 % Et3N to hexane:CH2Cl2 = 2:8 + 1 %
Et3N). 1H NMR (400 MHz, CDCl3) δ 1H NMR (400 MHz, CDCl3) δ 9.22 (br,
1H), 7.65 (d, J = 8.0 Hz, 1H), 7.44 (d, J = 8.3 Hz, 1H), 7.32 – 7.23 (m, 1H,
overlapped with CDCl3), 7.14 (t, J = 7.5 Hz, 1H), 6.78 (s, 1H), 3.86 (br,
4H), 1.81 – 1.64 ppm (m, 6H). 13C NMR (100 MHz, CDCl3) δ 162.4,
135.7, 129.9, 127.7, 124.3, 121.9, 120.6, 111.8, 104.9, 26.3, 26.2, 24.8
ppm. Elemental analysis calcd for C14H16N2O: C 73.66; H 7.06; N 12.27,
found (Conditions B): C 73.76; H 7.20; N 12.01, found (Conditions C):
C,73.40; H,7.04; N,11.94.
2.48 ppm (s, 3H). 13C NMR (75 MHz, CDCl3) δ 137.7, 137.6, 132.9,
132.7, 129.4, 127.9, 127.5, 125.4, 122.5, 120.7, 111. 3, 100.2, 22.2 ppm.
Elemental analysis calcd for C15H13N: C 86.92; H 6.32; N 6.76, Found: C
86.85; H 6.40; N 6.51.
2-Phenyl-6-(trifluoromethyl)-1H-indole (2p): Obtained as a white solid
(139 mg, 0.53 mmol, 98
% yield) after column chromatography
(hexane:CH2Cl2 =8:2 + 1 % Et3N). 1H NMR (600 MHz, CDCl3) δ 8.55 (br,
1H), 7.71-7.68 (m, 4H), 7.48 (t, J = 7.7 Hz, 2H), 7.40-7.36 (m, 2H), 6.88
ppm (d, J = 1.7 Hz, 1H). 13C NMR (151 MHz, CDCl3) δ 140.6, 135.6,
131.60, 131.57 129.2, 128.5, 125.4, 125.2 (q, 1JCF = 271.5 Hz), 124.2 (q,
2JCF = 32.2 Hz), 120.9, 117.02 (q, 3JCF = 3.4 Hz), 108.4 (q, 3JCF = 4.3 Hz),
100.1 ppm. 19F NMR (282 MHz, CDCl3) δ -60.96 ppm (s). Elemental
analysis calcd for C15H10F3N: C 68.96; H 3.86; N 5.36, found: C 68.70; H
3.66; N 5.15.
Methyl 1H-indole-3-carboxylate (2j): Obtained as a white solid (62 mg,
0.35 mmol, 64 % yield) after column chromatography (hexane:CH2Cl2
=8:2 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 8.69 (br, 1H), 8.26 – 8.10
(m, 1H), 7.92 (d, J = 3.0 Hz, 1H), 7.46 – 7.36 (m, 1H), 7.35 – 7.21 (m,
2H), 3.93 ppm (s, 3H). 13C NMR (101 MHz, CDCl3) δ 165.6, 136.1, 131.0,
125.8, 123.2, 122.1, 121.5, 111.5, 108.8, 51.1 ppm. Elemental analysis
calcd for C21H15N3: C 68.56; H 5.18; N 8.00, found (Conditions B): C
68.75; H 5.03; N 8.24, found (Conditions C):C, 68.48; H, 5.26; N, 7.76.
6-Nitro-2-phenyl-1H-indole (2q): Obtained as a white solid (100 mg,
0.42 mmol, 78 % yield) after column chromatography (hexane:AcOEt =
1
70:30 + 1 % Et3N). H NMR (400 MHz, DMSO-d6) δ 12.34 (br, 1H), 8.29
(d, J = 1.8 Hz, 1H), 7.94 (d, J = 7.4 Hz, 2H), 7.91 (dd, J = 8.8, 2.1 Hz, 1H),
7.72 (d, J = 8.8 Hz, 1H), 7.54 (t, J = 7.7 Hz, 2H), 7.43 (t, J = 7.4 Hz, 1H),
7.15 ppm (s, 1H). 13C NMR (100 MHz, DMSO-d6) δ 144.0, 141.7, 135.2,
133.5, 130.6, 129.0, 128.8, 125.6, 119.9, 114.6, 107.6, 99.7 ppm.
Elemental analysis calcd for C14H10N2O2: C, 70.58; H, 4.23; N, 11.76,
found: C,70.78; H,4.27; N,11.46.
2-(4-Chlorophenyl)-1H-indole-3-carbonitrile (2k): Obtained as a white
solid (129 mg, 0.51 mmol, 94 % yield) after column chromatography
(hexane:CH2Cl2 = 7:3 + 1 % Et3N). 1H NMR (600 MHz, DMSO-d6) δ
12.65 (br, 1H), 8.00 (d, J = 8.6 Hz, 2H), 7.72 (d, J = 8.6 Hz, 2H), 7.66 (d,
J = 7.9 Hz, 1H), 7.57 (d, J = 8.1 Hz, 1H), 7.34 (t, J = 8.1 Hz, 1H), 7.28
ppm (t, J = 7.4 Hz, 1H). 13C NMR (151 MHz, DMSO-d6) δ 143.8, 136.1,
135.1, 129.9, 129.2, 128.7, 124.6, 122.6, 118.9, 117.2, 113.2, 82.3 ppm.
Elemental analysis calcd for C15H9ClN2: C 71.29; H 3.59; N 11.09, found:
C 71.29; H 3.62; N 10.97.
N,N-Dimethyl-2-phenyl-1H-indol-6-amine (2r): Obtained as a white
solid (77 mg, 0.33 mmol, 60 % yield) after column chromatography
(hexane:AcOEt = 80:20 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 8.15
(br, 1H), 7.61 (d, J = 7.9 Hz, 2H), 7.49 (d, J = 8.6 Hz, 1H), 7.41 (t, J = 7.7
Hz, 2H), 7.27 (t, J = 7.3 Hz, 1H, overlapped with CDCl3), 6.82 – 6.70 (m,
3H), 2.98 ppm (s, 6H). 13C NMR (100 MHz, CDCl3) δ 148.1, 138.6, 136.0,
132.9, 129.1, 127.0, 124.6, 121.8, 121.1, 110.1, 99.9, 94.9, 42.1 ppm.
Elemental analysis calcd for C16H16N2: C, 81.32; H, 6.82; N, 11.85, found:
C,81.23; H,7.06; N,11.49.
2-(p-Tolyl)-1H-indole-3-carbonitrile (2l): Obtained as a white solid (120
mg, 0.52 mmol, 95
%
yield) after column chromatography
(hexane:CH2Cl2 = 7:3 + 1 % Et3N). 1H NMR (400 MHz, DMSO-d6) δ
12.52 (br, 1H), 7.89 (d, J = 8.1 Hz, 2H), 7.63 (d, J = 7.6 Hz, 1H), 7.55 (d,
J = 8.0 Hz, 1H), 7.44 (d, J = 8.0 Hz, 2H), 7.28 (dt, J = 22.4, 7.1 Hz, 2H),
2.41 ppm (s, 3H). 13C NMR (101 MHz, DMSO-d6) δ 145.4, 140.3, 135.9,
130.3, 128.8, 127.3, 127.0, 124.2, 122.4, 118.7, 117.6, 113.0, 81.4, 21.4
ppm. Elemental analysis calcd for C16H12N2: C 82.73; H 5.21; N 12.06,
found: C 82.38; H 5.23; N 11.72.
6-Chloro-2-(4-chlorophenyl)-1H-indole (2s): Obtained as a white solid
(131 mg, 0.50 mmol, 92
% yield) after column chromatography
(hexane:CH2Cl2 =7:3 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 8.28 (br,
1H), 7.63 – 7.52 (m, 3H), 7.48 – 7.36 (m, 3H), 7.12 (d, J = 8.4 Hz, 1H),
6.79 ppm (s, 1H). 13C NMR (101 MHz, CDCl3) δ 137.43, 137.19, 133.78,
130.42, 129.31, 128.37, 127.73, 126.31, 121.56, 121.25, 110.88, 100.42
ppm. Elemental analysis calcd for C14H9Cl2N: C 64.15; H 3.46; N 5.34,
found: C 64.50; H 3.66; N 5.15.
2-(4-(Diethylamino)phenyl)-1H-indole-3-carbonitrile (2m): Obtained
as a white solid (149 mg, 0.51 mmol, 95 % yield) after column
1
chromatography (hexane:CH2Cl2 = 8:2 + 1 % Et3N). H NMR (300 MHz,
DMSO-d6) δ 12.19 (br, 1H), 7.84 (d, J = 9.0 Hz, 2H), 7. 53 (dd, J = 6.6,
1.6 Hz, 1H), 7.47 (dd, J = 6.8, 1.6 Hz, 1H), 7.28 – 7.15 (m, 2H), 6.85 (d, J
2-(2-Chlorophenyl)-1H-indole (2t): Obtained as a white solid (116 mg,
0.51 mmol, 64 % yield) after column chromatography (hexane:CH2Cl2
=7:3 + 1 % Et3N); 94 % yield. 1H NMR (300 MHz, CDCl3) δ 8.78 (br, 1H),
7.78 – 7.66 (m, 1H), 7.56 – 7.51 (m, 1H), 7.46 (d, J = 8.1 Hz, 1H), 7.39
(dd, J = 7.5, 1.4 Hz, 1H), 7.34 (dd, J = 5.3, 1.7 Hz, 1H), 7.32 – 7.27 (m,
1H), 7.24 – 7.15 (m, 1H), 6.92 ppm (d, J = 1.3 Hz, 1H). 13C NMR (75
MHz, CDCl3) δ 136.8, 135.5, 131.8, 131.6, 131.2, 131.1, 129.2, 128.6,
127.7, 123.1, 121.2, 120.6, 111.5, 104.0 ppm. Elemental analysis calcd
for C14H10ClN: C 73.85; H 4.43; N 6.15, found: C 73.50; H 4.66; N 6.01.
= 9.0 Hz, 2H), 3.43 (q, J = 7.0 Hz, 4H), 1.14 ppm (t, J = 7.0 Hz, 6H). 13
C
NMR (101 MHz, DMSO-d6) δ 148.9, 146.6, 135.7, 129.12, 128.5, 123.4
122.0, 118.2, 118.1, 115.7, 112.5, 111.8, 78.8, 44.1, 12.9 ppm.
Elemental analysis calcd for C19H19N3: C 78.86; H 6.62; N 14.52, found:
C 78.87; H 6.60; N 14.50.
2-(1-Methyl-1H-pyrrol-2-yl)-1H-indole-3-carbonitrile (2n): Obtained as
a
white solid (92 mg, 0.42 mmol, 77
%
yield) after column
1
chromatography (hexane:CH2Cl2 = 4:6 + 1 % Et3N). H NMR (400 MHz,
CDCl3) δ 8.59 (br, 1H), 7.75 (m, 1H), 7.44 (m, 1H), 7.36 – 7.26 (m, 2H),
6.89 – 6.82 (m, 1H), 6.58 (dd, J = 3.7, 1.7 Hz, 1H), 6.28 (dd, J = 3.7, 2.8
Hz, 1H), 3.85 ppm (s, 3H). 13C NMR (100 MHz, CDCl3) δ 138.2, 134.8,
128.6, 126.8, 124.3, 122.9, 122.5, 119.5, 116.9, 112.7, 111.7, 109.3,
85.3, 35.8 ppm. Elemental analysis calcd for C14H11N3: C, 76.00; H, 5.01;
N, 18.99, found (Conditions B): C 76.25; H 5.13; N 18.70, found
(Conditions C): C, 76.38; H, 5.24; N, 18.69
5,6-Dimethoxy-2-phenyl-1H-indole (2u): Obtained as a white solid (97
mg, 0.38 mmol, 71 % yield) after column chromatography (hexane:AcOEt
1
= 75:25 + 1 % Et3N). H NMR (400 MHz, CDCl3) δ 8.21 (br, 1H), 7.64 –
7.58 (m, 2H), 7.44 – 7.38 (m, 2H), 7.29 (m, 1H overlapped with residual
solvent signal), 7.08 (s, 1H), 6.91 (s, 1H), 6.73 (dd, J = 2.2, 0.8 Hz, 1H),
3.93 ppm (d, J = 5.6 Hz, 6H). 13C NMR (100 MHz, CDCl3) δ 147.5, 145.6,
136.7, 132.8, 131.5, 129.1, 127.2, 124.7, 122.3, 102.4, 100.0, 94.6, 56.5,
56.3 ppm. Elemental analysis calcd for C16H15NO2: C, 75.87; H, 5.97; N,
5.53, found: C, 75.49; H, 6.09; N, 5.19.
6-Methyl-2-phenyl-1H-indole (2o): Obtained as a white solid (107 mg,
0.52 mmol, 96 % yield) after column chromatography (hexane:AcOEt =
85:15 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 8.19 (br, 1H), 7.65 (d, J
= 7.7 Hz, 2H), 7.52 (d, J = 8.0 Hz, 1H), 7.44 (t, J = 7.7Hz, 2H), 7.31 (t, J
= 7.4 Hz, 1H), 7.19 (s, 1H), 6.97 (d, J = 8.0 Hz, 1H), 6.79 (d, 1.0 Hz, 1H),
6-Phenyl-5H-[1,3]dioxolo[4,5-f]indole (2v): Obtained as a yellow solid
(92 mg, 0.39 mmol, 72
% yield) after column chromatography
(hexane:AcOEt = 70:30 + 1 % Et3N). 1H NMR (400 MHz, DMSO-d6) δ
14
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
11.37 (s, 1H), 7.75 (d, J = 7.4 Hz, 2H), 7.41 (t, J = 7.7 Hz, 2H), 7.24 (t, J
= 7.4 Hz, 1H), 6.99 (s, 1H), 6.90 (s, 1H), 6.75 (d, J = 1.8 Hz, 1H), 5.94
ppm (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 144.1, 142.2, 136.0, 132.3,
131.9, 128.7, 126.5, 124.1, 122.3, 99.98, 98.95, 98.3, 91.8 ppm.
Elemental analysis calcd for C15H11NO2: C, 75.94; H, 4.67; N, 5.90,
found: C, 75.69; H, 4.71; N, 5.55.
white solid precipitated from the mixture . It was filtered through Büchner
funnel, washed with cold acetonitrile and dried under vacuum (91 mg,
0.36 mmol, 67 % yield). H NMR (600 MHz, DMSO-d6) δ 11.57 (br, 1H),
8.09 (d, J = 8.0 Hz, 1H), 7.81 (d, J = 8.7 Hz, 2H), 7.77 (d, J = 8.3 Hz, 1H),
7.66 (t, J = 7.3 Hz, 1H), 7.32 (t, J = 7.5 Hz, 1H), 7.15 (d, J = 8.7 Hz, 2H),
6.31 (s, 1H), 3.86 ppm (s, 3H). 13C NMR (151 MHz, DMSO-d6) δ 177.3,
161.5, 150.1, 141.0, 132.1, 129.3, 126.7, 125.3, 125.2, 123.5, 119.1,
114.9, 107.0, 55.9 ppm. Elemental analysis calcd for C16H13NO2: C
76.48; H 5.21; N 5.57, found: C 76.28; H 4.96; N 5.51.
1
2-Phenyl-6-azaindole (2w): Obtained as a white solid (95 mg, 0.49
mmol, 91 % yield) after column chromatography (AcOEt + 2 % Et3N +
1 % MeOH). 1H NMR (400 MHz, CDCl3) δ 10.27 (br, 1H), 8.96 (s, 1H),
8.27 (d, J = 5.5 Hz, 1H), 7.82 (d, J = 7.3 Hz, 2H), 7.63 – 7.36 (m, 4H),
6.85 ppm (s, 1H). 13C NMR (101 MHz, CDCl3) δ 142.4, 138.8, 134.3,
134.1, 133.8, 129.2, 128.8, 126.0, 115.1, 98.9 ppm. Elemental analysis
calcd for C13H10N2: C 80.39; H 5.19; N 14.42, found: C 80.29; H 4.98; N
14.23.
Carbazole: Obtained as a white solid (42 mg, 0.25 mmol, 46 % yield)
after column chromatography (hexane:AcOEt=8:2 + 1 % Et3N). 1H NMR
(300 MHz, CDCl3) δ 8.10 (d, J = 7.8 Hz, 1H), 7.99 (br, 1H), 7.48 – 7.40
(m, 2H), 7.32 – 7.17 ppm (m, 1H). 13C NMR (75 MHz, CDCl3) δ 139.9,
126.2, 123.7, 120.7, 119.8, 111.0 ppm. Elemental analysis calcd for
C12H9N: C 86.20; H 5.43; N 8.38, found: C 86.30; H 5.35; N 8.35.
2-(Pyridin-2-yl)-1H-indole (2x): Obtained as a white solid (103 mg, 0.53
mmol, 98 % yield) after column chromatography (hexane:CH2Cl2 = 7:3 +
1 % Et3N). 1H NMR (300 MHz, CDCl3) δ 9.75 (br, 1H), 8.58 (d, J = 4.6 Hz,
1H), 7.82 (d, J = 8.0 Hz, 1H), 7.73 (td, J = 7.8, 1.7 Hz, 1H), 7.66 (d, J =
7.9 Hz, 1H), 7.42 (d, J = 8.2 Hz, 1H), 7.26-7.16 (m, 2H, overlapping with
CDCl3), 7.12 (t, J = 7.5 Hz, 1H), 7.04 ppm (d, J = 1.1 Hz, 1H). 13C NMR
(75 MHz, CDCl3) δ 150.5, 149.04, 137.4, 137.1, 136.6, 129.4, 123.8,
122.4, 121.3, 120.6, 120.4, 111.9, 101.4 ppm. Elemental analysis calcd
for C13H10N: C 80.39; H 5.19; N 14.42, found: C 80.19; H 5.23; N 14.35.
2-Methylindole: Obtained from β-nitrostyrene (29 mg, 0.22 mmol, 41 %
yield) after column chromatography (hexane:CH2Cl2=7:3 + 1 % Et3N). 1H
NMR (300 MHz, CDCl3) δ 7.82 (br, 1H), 7.54 (d, J = 7.5 Hz, 1H), 7.29 (d,
J = 7.2 Hz, 1H), 7.18 – 7.05 (m, 2H), 6.22 (s, 1H), 2.45 ppm (s, 3H). 13C
NMR (75 MHz, CDCl3) δ 136.5, 135.5, 129.5, 121.4, 120.1, 110.7, 100.7,
14.1 ppm. Elemental analysis calcd for C9H9N: C 82.41; H 6.92; N 10.68,
found: C 82.38; H 6.87; N 10.65.
4,5-Dimethyl-2-phenyl-3,6-dihydro-2H-1,2-oxazine: Obtained as
a
2,6-Di(1H-indol-2-yl)pyridine (2y): Obtained as a white solid (65 mg,
0.21 mmol, 78 % yield) after column chromatography (hexane:CH2Cl2
=7:3 + 1 % Et3N). 1H NMR (300 MHz, DMSO-d6) δ 11.71 (s, 1H), 7.88 (t,
2H), 7.61 (dd, J = 17.9, 8.0 Hz, 4H), 7.28 (d, J = 1.2 Hz, 2H), 7.25 – 7.18
(m, 2H), 7.07 ppm (t, J = 7.1 Hz, 2H). 13C NMR (75 MHz, DMSO-d6) δ
150.3, 138.7, 137.9, 137.6, 129.5, 123.6, 121.8, 120.5, 118.7, 112.4,
101.7 ppm. Elemental analysis calcd for C21H15N3: C 81.53; H 4.89; N
13.58, found: C 81.75; H 4.83; N 13.34.
white solid (98 mg, 0.52 mmol, 96 % yield) after column chromatography
(hexane:CH2Cl2 =7:3 + 1 % Et3N). 1H NMR (400 MHz, CDCl3) δ 7.33 (t, J
= 10.6 Hz, 2H), 7.16 (dd, J = 11.4, 1.1 Hz, 2H), 7.02 (t, J = 9.7 Hz, 1H),
4.35 (s, 2H), 3.69 (s, 2H), 1.76 (s, 3H), 1.67 ppm (s, 3H). 13C NMR (101
MHz, CDCl3) δ 150.8, 129.2, 125.4, 122.8, 122.7, 116.3, 72.4, 56.8, 16.3,
14.06 ppm. Elemental analysis calcd for C12H15NO: C 76.16; H 7.99; N
7.40, found: C 76.36; H 8.10; N 7.40.
(E)-2-(2-Nitrostyryl)-1H-indole-3-carbonitrile (2z’): Obtained as
a
Acknowledgements
yellow solid (48 mg, 0.17 mmol, 61 yield) after column
%
chromatography (hexane:CH2Cl2 = 40:60 + 1 % Et3N). The yield reported
in table 9 was calculated considering the molecular weight of the
hemihydrate molecule. 1H NMR (300 MHz, DMSO-d6) δ 12.62 (br, 1H),
8.11 (d, J = 7.6 Hz, 1H), 8.05 (dd, J = 8.1, 0.9 Hz, 1H), 7.83 (d, J = 16.3
Hz, 1H, olefinic CH overlapped with multiplet at 7.84 – 7.76), 7.84 – 7.76
(m, 1H), 7.67 – 7.58 (m, 2H), 7.54 (d, J = 8.1 Hz, 1H), 7.320(d, J = 16.3
Hz, 1H, olefinic CH overlapped with multiplet at 7.39 – 7.31), 7.39 – 7.31
(m, 1H), 7.28 – 7.21 ppm (m, 1H). 13C NMR (75 MHz, DMSO-d6) δ 148.1,
142.4, 136.1, 133.7, 130.2, 129.8, 128.3, 127.4, 127.1, 124.9, 124.6,
122.1, 119.8, 118.6, 115.7, 112.51, 85.25 ppm. MS (ESI+), m/z [M+Na]+
calcd for C17H11N3NaO2+ : 312.07, found 312.32. MS (ESI-), m/z [M-H]-
calcd for C17H10N3O2- : 289.08, found 288.57. Elemental analysis calcd
for C17H11N3O2∙½ H2O: C, 68.45; H, 4.05; N, 14.09, found: C, 68.58; H,
4.61; N, 14.09.
We thank Dr. M. Hagar and Mrs. E. Storer for preliminary work,
Prof. P. Mussini for recording and discussing the cyclic
voltammograms, Dr. M. El-Atawy for helpful discussions. F.F.
thanks the Dipartimento di Chimica, Università degli Studi di
Milano, for financial support (Piano di Sostegno alla Ricerca
2019).
Keywords: C-H Amination • CO Source • Nitrogen heterocycles
• Palladium • Phenyl Formate
[1]
[2]
a) F. Ragaini, in Reference Module in Chemistry, Molecular Sciences
and Chemical Engineering, Elsevier, 2016. doi: 10.1016/B978-0-12-
409547-2.12210-3; b) F. Ragaini, Dalton Trans. 2009, 6251-6266; c) F.
Paul, Coord. Chem. Rev. 2000, 203, 269-323; d) A. M. Tafesh, J.
Weiguny, Chem. Rev. 1996, 96, 2035-2052.
1H,1'H-[2,2'-Biindole]-3-carbonitrile (2z’’): Obtained as a white solid
(62 mg, 0.24 mmol, 90
% yield) and recovered without column
chromatography. The product was precipitated from the reaction mixture
by addition of water (15 mL), collected by filtration on a Buchner funnel,
let to dry on the filter paper and then washed several times with mixture
of hexane and CH2Cl2 (3:1). 1H NMR (400 MHz, DMSO-d6) δ 12.42 (br,
1H), 11.54 (br, 1H), 7.69 (d, J = 7.9 Hz, 1H), 7.65 (d, J = 7.7 Hz, 1H),
7.59 (d, J = 7.9 Hz, 1H), 7.55 (d, J = 8.0 Hz, 1H), 7.33 (dd, 7.2, 1.2 Hz,
1H), 7.30 – 7.21 (m, 2H, partially overlapped with singlet at 7.21), 7.21 (s,
1H), 7.11 ppm (t, J = 7.9 Hz, 1H). 13C NMR (100 MHz, DMSO-d6) δ 138.2,
137.1, 135.3, 128.0, 127.9, 127.3, 123.9, 123.3, 122.2, 120. 9, 120.3,
118.2, 116.6, 112.6, 112.0, 103.3, 81.1 ppm. MS (ESI-), m/z [M-H]- calcd
a) F. Ragaini, S. Cenini, E. Gallo, A. Caselli, S. Fantauzzi, Curr. Org.
Chem. 2006, 10, 1479-1510; b) B. C. G. Soderberg, Curr. Org. Chem.
2000, 4, 727-764; c) S. Cenini, F. Ragaini, Catalytic Reductive
Carbonylation of Organic Nitro Compounds, Kluwer Academic
Publishers, Dordrecht, The Netherlands, 1996; d) F. Ferretti, D. R.
Ramadan, F. Ragaini, ChemCatChem 2019, 11, 4450-4488; e) A. A.
Tsygankov, M. Makarova, D. Chusov, Mendeleev Commun. 2018, 28,
113-122; f) F. Ferretti, D. Formenti, F. Ragaini, Rend. Fis. Acc. Lincei
2017, 28, 97-115.
-
for C17H10N3 : 256.09, found 256.43. Elemental analysis calcd for
C17H11N3: C, 79.36; H, 4.31; N, 16.33, found: C, 78.98; H, 4.65; N, 15.97.
[3]
a) F. Ragaini, A. Rapetti, E. Visentin, M. Monzani, A. Caselli, S. Cenini,
J. Org. Chem. 2006, 71, 3748-3753; b) F. Ragaini, F. Ventriglia, M.
Hagar, S. Fantauzzi, S. Cenini, Eur. J. Org. Chem. 2009, 2009, 2185-
2-(4-Methoxyphenyl)quinolin-4(1H)-one: After the catalytic reaction, a
15
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
2189; c) A. Penoni, K. M. Nicholas, Chem. Commun. 2002, 484-485; d)
I. W. Davies, J. H. Smitrovich, R. Sidler, C. Qu, V. Gresham, C. Bazaral,
Tetrahedron 2005, 61, 6425-6437; e) F. Ragaini, P. Sportiello, S.
Cenini, J. Organomet. Chem. 1999, 577, 283-291; f) K. Okuro, J.
Gurnham, H. Alper, J. Org. Chem. 2011, 76, 4715-4720; g) K. Okuro, J.
Gurnham, H. Alper, Tetrahedron Lett. 2012, 53, 620-622; h) N. H.
Ansari, C. A. Dacko, N. G. Akhmedov, B. C. G. Söderberg, J. Org.
Chem. 2016, 81, 9337-9349; i) Y. L. Zhang, J. W. Hubbard, N. G.
Akhmedov, J. L. Petersen, B. C. G. Soderberg, J. Org. Chem. 2015, 80,
4783-4790; j) M. Akazome, T. Kondo, Y. Watanabe, J. Org. Chem.
1993, 58, 310-312; k) D. K. O'Del, K. M. Nicholas, J. Org. Chem. 2003,
68, 6427-6430; l) B. C. G. Söderberg, J. M. Wallace, J. Tamariz, Org.
Lett. 2002, 4, 1339-1342; m) F. Ragaini, S. Cenini, D. Brignoli, M.
Gasperini, E. Gallo, J. Org. Chem. 2003, 68, 460-466; n) R. S.
Srivastava, K. M. Nicholas, Organometallics 2005, 24, 1563-1568; o) T.
H. H. Hsieh, V. M. Dong, Tetrahedron 2009, 65, 3062-3068; p) M. A. El-
Atawy, F. Ferretti, F. Ragaini, Eur. J. Org. Chem. 2017, 1902-1910; q)
F. Ferretti, M. A. El-Atawy, S. Muto, M. Hagar, E. Gallo, F. Ragaini, Eur.
J. Org. Chem. 2015, 5712-5715; r) J. H. Smitrovich, I. W. Davies, Org.
Lett. 2004, 6, 533-535; s) S. Tollari, S. Cenini, F. Ragaini, L. Cassar, J.
Chem. Soc., Chem. Comm. 1994, 1741-1742; t) F. Ragaini, S. Cenini,
E. Borsani, M. Dompe, E. Gallo, M. Moret, Organometallics 2001, 20,
3390-3398; u) M. A. EL-Atawy, F. Ferretti, F. Ragaini, Eur. J. Org.
Chem. 2018, 4818-4825.
[15] R. B. Wilson, Jr., R. M. Laine, J. Am. Chem. Soc. 1985, 107, 361-369.
[16] K. Ouyang, W. Hao, W.-X. Zhang, Z. Xi, Chem. Rev. 2015, 115, 12045-
12090.
[17] a) E. Alessio, G. Clauti, G. Mestroni, J. Mol. Catal. 1985, 29, 77-98; b)
F. Ragaini, S. Cenini, M. Gasperini, J. Mol. Catal. A: Chem. 2001, 174,
51-57.
[18] K. Nomura, J. Mol. Catal. 1992, 73, L1-L4.
[19] G. Glotz, B. Gutmann, P. Hanselmann, A. Kulesza, D. Roberge, C. O.
Kappe, RSC Adv. 2017, 7, 10469-10478.
[20] a) R. S. Berman, J. K. Kochi, Inorg. Chem. 1980, 19, 248-254; b) S. J.
Skoog, W. L. Gladfelter, J. Am. Chem. Soc. 1997, 119, 11049-11060;
c) A. J. Kunin, M. D. Noirot, W. L. Gladfelter, J. Am. Chem. Soc. 1989,
111, 2739-2741; d) Y. A. Belousov, Russ. Chem. Rev. 2007, 76, 41-58;
e) Y. A. Belousov, T. A. Kolosova, Polyhedron 1987, 6, 1959-1970; f) F.
Ragaini, J. S. Song, D. L. Ramage, G. L. Geoffroy, G. A. P. Yap, A. L.
Rheingold, Organometallics 1995, 14, 387-400; g) F. Ragaini,
Organometallics 1996, 15, 3572-3578; h) P. H. Liu, H. Y. Liao, C. H.
Cheng, J. Chem. Soc., Chem. Comm. 1995, 2441-2442; i) F. Ragaini,
S. Cenini, F. Demartin, J. Chem. Soc., Chem. Comm. 1992, 1467-
1468; j) F. Ragaini, S. Cenini, F. Demartin, Organometallics 1994, 13,
1178-1189.
[21] O. H. Wheeler, Can. J. Chem. 1963, 41, 192-194.
[22] P. Du, J. L. Brosmer, D. G. Peters, Org. Lett. 2011, 13, 4072-4075.
[23] Z. V. Todres, G. V. Gridunova, K. I. Dyusengaliev, Y. T. Struchkov, Zh.
Org. Khim. 1987, 23, 1805-1807.
[4]
a) L. Wu, Q. Liu, R. Jackstell, M. Beller, Angew. Chem., Int. Ed. 2014,
53, 6310-6320; b) P. Gautam, B. M. Bhanage, Catal. Sci. Technol.
2015, 5, 4663-4702; c) S. D. Friis, A. T. Lindhardt, T. Skrydstrup, Acc.
Chem. Res. 2016, 49, 594-605; d) H. Konishi, K. Manabe, Synlett 2014,
25, 1971-1986; e) J. B. Peng, X. X. Qi, X. F. Wu, Synlett 2017, 28, 175-
194; f) J. Cao, Z.-J. Zheng, Z. Xu, L.-W. Xu, Coord. Chem. Rev. 2017,
336, 43-53; g) V. V. Grushin, H. Alper, Organometallics 1993, 12, 3846-
3850; h) S. N. Gockel, K. L. Hull, Org. Lett. 2015, 17, 3236-3239; i) Z.
Chen, L.-C. Wang, X.-F. Wu, Chem. Commun. 2020, 56, 6016-6030; j)
O. I. Afanasyev, I. Cherkashchenko, A. Kuznetsov, F. Kliuev, S.
Semenov, O. Chusova, G. Denisov, D. Chusov, Mendeleev Commun.
2020, 30, 112-113; k) K. Mondal, P. Halder, G. Gopalan, P. Sasikumar,
K. V. Radhakrishnan, P. Das, Org. Biomol. Chem. 2019, 17, 5212-5222.
a) N. Jana, F. Zhou, T. G. Driver, J. Am. Chem. Soc. 2015, 137, 6738-
6741; b) F. Zhou, D.-S. Wang, T. G. Driver, Adv. Synth. Catal. 2015,
357, 3463-3468; c) Z. Su, B. Liu, H. Liao, H.-W. Lin, Eur. J. Org. Chem.
2020, 4059-4066; d) J.-B. Peng, H.-Q. Geng, F.-P. Wu, D. Li, X.-F. Wu,
J. Catal. 2019, 375, 519-523; e) H.-Q. Geng, C.-Y. Hou, L.-C. Wang, J.-
B. Peng, X.-F. Wu, J. Catal. 2020, 381, 271-274.
[24] C. J. Finder, M. G. Newton, N. L. Allinger, Acta Crystallogr. Sect. B-
Struct. Sci. 1974, 30, 411-415.
[25] a) T. Ueda, H. Konishi, K. Manabe, Org. Lett. 2012, 14, 3100-3103; b)
H. Konishi, H. Nagase, K. Manabe, Chem. Commun. 2015, 51, 1854-
1857; c) W. J. Chang, J. F. Li, W. L. Ren, Y. Shi, Org. Biomol. Chem.
2016, 14, 3047-3052; d) S. P. Chavan, B. M. Bhanage, Eur. J. Org.
Chem. 2015, 2015, 2405-2410; e) Y. Katafuchi, T. Fujihara, T. Iwai, J.
Terao, Y. Tsuji, Adv. Synth. Catal. 2011, 353, 475-482; f) T. Fujihara, T.
Hosoki, Y. Katafuchi, T. Iwai, J. Terao, Y. Tsuji, Chem. Commun. 2012,
48, 8012-8014; g) H. Li, H. Neumann, M. Beller, X.-F. Wu, Angew.
Chem., Int. Ed. 2014, 53, 3183-3186.
[26] M. Gasperini, F. Ragaini, C. Remondini, A. Caselli, S. Cenini, J.
Organomet. Chem. 2005, 690, 4517-4529.
[5]
[27] a) S. Tollari, S. Cenini, C. Crotti, E. Gianella, J. Mol. Catal. 1994, 87,
203-214; b) M. A. EL-Atawy, F. Ferretti, F. Ragaini, Eur. J. Org. Chem.
2019, 4968-4968.
[28] a) F. Ferretti, F. Ragaini, R. Lariccia, E. Gallo, S. Cenini,
Organometallics 2010, 29, 1465-1471; b) F. Ferretti, E. Gallo, F.
Ragaini, J. Organomet. Chem. 2014, 771, 59-67.
[6]
[7]
a) R. Zhou, X. Qi, X.-F. Wu, ACS Comb. Sci. 2019, 21, 573-577; b) X.
Qi, R. Zhou, J.-B. Peng, J. Ying, X.-F. Wu, Eur. J. Org. Chem. 2019,
2019, 5161-5164.
[29] B. C. G. Soderberg, S. R. Banini, M. R. Turner, A. R. Minter, A. K.
Arrington, Synthesis 2008, 903-912.
a) D. Formenti, F. Ferretti, F. Ragaini, ChemCatChem 2018, 10, 148-
152. b) See also the PhD thesis of Dr. D. Formenti (DOI:
10.13130/formenti-dario_phd2017-03-24).
[30] a) J.-Y. Mérour, F. Buron, K. Plé, P. Bonnet, S. Routier, Molecules
2014, 19, 19935; b) J. J. Song, J. T. Reeves, F. Gallou, Z. Tan, N. K.
Yee, C. H. Senanayake, Chem. Soc. Rev. 2007, 36, 1120-1132.
[31] a) W.-L. Jia, Q.-D. Liu, R. Wang, S. Wang, Organometallics 2003, 22,
4070-4078; b) Q. Liu, L. Thorne, I. Kozin, D. Song, C. Seward, M.
D'Iorio, Y. Tao, S. Wang, J. Chem. Soc., Dalton Trans. 2002, 3234-
3240.
[8]
[9]
G. Jenner, E. M. Nahmed, H. Leismann, Tetrahedron Lett. 1989, 30,
6501-6502.
a) S. Cenini, M. Pizzotti, C. Crotti, F. Porta, G. La Monica, J. Chem.
Soc., Chem. Commun. 1984, 1286-1287; b) C. Crotti, S. Cenini, B.
Rindone, S. Tollari, F. Demartin, J. Chem. Soc., Chem. Comm. 1986,
784-786; c) F. Ragaini, S. Cenini, J. Mol. Catal. A: Chem. 2000, 161,
31-38.
[32] W.-M. Dai, L.-P. Sun, D.-S. Guo, Tetrahedron Lett. 2002, 43, 7699-
7702.
[33] A. McGookin, J. Soc. Chem. Ind., London 1949, 68, 195-196.
[34] a) E. Alessio, G. Mestroni, J. Mol. Catal. 1984, 26, 337-340; b) F.
Ragaini, C. Cognolato, M. Gasperini, S. Cenini, Angew. Chem., Int. Ed.
2003, 42, 2886-2889.
[10] P. Wehman, H. M. A. vanDonge, A. Hagos, P. C. J. Kamer, P. W. N. M.
van Leeuwen, J. Organomet. Chem. 1997, 535, 183-193.
[11] A. Ben Taleb, G. Jenner, J. Organomet. Chem. 1993, 456, 263-269.
[12] a) P. Wehman, V. E. Kaasjager, W. G. J. Delange, F. Hartl, P. C. J.
Kamer, P. W. N. M. van Leeuwen, J. Fraanje, K. Goubitz,
Organometallics 1995, 14, 3751-3761; b) A. Bontempi, E. Alessio, G.
Chanos, G. Mestroni, J. Mol. Catal. 1987, 42, 67-80; c) M. Gasperini, F.
Ragaini, C. Cazzaniga, S. Cenini, Adv. Synth. Catal. 2005, 347, 105-
120.
[35] S. P. Gorugantula, G. M. Carrero-Martinez, S. W. Dantale, B. C. G.
Soderberg, Tetrahedron 2010, 66, 1800-1805.
[36] M. A. EL-Atawy, D. Formenti, F. Ferretti, F. Ragaini, ChemCatChem
2018, 10, 4707-4717.
[37] D. R. Ramadan, F. Ferretti, F. Ragaini, work in progress.
[38] H. Konishi, M. Matsubara, K. Mori, T. Tokiwa, S. Arulmozhiraja, Y.
Yamamoto, Y. Ishikawa, H. Hashimoto, Y. Shigeta, H. Tokiwa, K.
Manabe, Adv. Synth. Catal. 2017, 359, 3592-3601.
[13] F. Ragaini, F. Ventriglia, M. Hagar, S. Fantauzzi, S. Cenini, Eur. J. Org.
Chem. 2009, 2185-2189.
[14] Y. Shvo, M. Abed, Y. Blum, R. M. Laine, Israel J. Chem. 1986, 27, 267-
275.
[39] Since the submission of our preliminary communication on this topic,
Manabe and coworkers have published an in depth study of aryl
16
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
formate decomposition by bases.[38] Their results are in full agreement
with ours. Moreover, the fact that pyridine was found to be the least
effective of all the bases tested supports our result that the small
amount of phenanthroline present in our system has no significant role
in catalyzing phenyl formate decomposition.
[40] L. M. Kogan, Russ. Chem. Rev. 1986, 55, 1164-1174.
[41] D. Kuzmich, C. Mulrooney, Synthesis 2003, 2003, 1671-1678.
[42] a) I. W. Davies, V. A. Guner, K. N. Houk, Org. Lett. 2004, 6, 743-746; b)
A. G. Leach, K. N. Houk, I. W. Davies, Synthesis 2005, 2005, 3463-
3467.
[5c, 44]
[43] In at least two cases
since the publishing of our preliminary
communication on this topic, the use of phenyl formate was compared
with that of Mo(CO)6 in reductive cyclization reactions of nitro
compounds and in both cases the latter afforded better results.
However, in neither case was a base added to the reaction mixture. In
our view, this makes the comparison completely meaningless.
[44] Y. Gao, F. Yang, F. Sun, L. Liu, B. Liu, S.-P. Wang, C.-W. Cheng, H.
Liao, H.-W. Lin, Org. Lett. 2019, 21, 1430-1433.
17
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100789
European Journal of Organic Chemistry
FULL PAPER
Entry for the Table of Contents
We here describe a very efficient, general and scalable protocol for the preparation of indoles and other N-heterocycles from suitably
substituted nitroarenes using alkyl and phenyl formates as CO surrogates. Using phenyl formate, the products were isolated in yields
often higher than those previously achieved by using gaseous CO. The mechanism of both the decarbonylation reaction of phenyl
formate and the cyclization reaction were clarified by kinetic and mechanistic studies.
18
This article is protected by copyright. All rights reserved.