A bifunctional palladium/acid solid catalyst performs the direct synthesis of cyclohexylanilines and dicyclohexylamines from nitrobenzenes

Article abstract of DOI:10.1039/c3cc44064h

Nitroderivatives are transformed to cyclohexylanilines at room temperature in good yields and selectivity via a hydrogenation-amine coupling cascade reaction using Pd nanoparticles on carbon as a catalyst and a Broensted acid.

Full text of DOI:10.1039/c3cc44064h

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
A bifunctional palladium/acid solid catalyst performs
the direct synthesis of cyclohexylanilines and
dicyclohexylamines from nitrobenzenes†
Cite this: Chem. Commun., 2013,
49, 8160
Received 30th May 2013,
Accepted 17th July 2013
´
´
Paula Rubio-Marques, Antonio Leyva-Perez* and Avelino Corma*
DOI: 10.1039/c3cc44064h
www.rsc.org/chemcomm
Nitroderivatives are transformed to cyclohexylanilines at room and co-workers.7,8 This excellent homogeneous process and, to
temperature in good yields and selectivity via a hydrogenation- our knowledge, all the other catalytic systems reported to date
amine coupling cascade reaction using Pd nanoparticles on carbon operate at temperatures >120 1C.6 In contrast, the reduction of
as a catalyst and a Bro¨nsted acid.
nitrobenzene to aniline 2 - 3 and the subsequent reduction of
aniline to cyclohexylamine 3 - 4 operate under milder condi-
Secondary amines are ubiquitous in nature and fundamental tions.9 Thus, it would be desirable to have a suitable catalyst to
building blocks in chemical synthesis. Production is based on couple aniline and cyclohexylamine and, ideally, to be able to
two different processes: dehydroxylative amination of alcohols engage the two hydrogenations in one-pot to obtain cyclohexyl-
for manufacturing alkyl amines,1,2 and reductive amination of aniline 5 from nitrobenzene 2.10 This process would avoid the
ketones.3,4 The latter includes the multiton/year production cyclohexane–cyclohexanone11–cyclohexanol
7
-
8
+
9
of aniline 3 from nitrobenzene 2 as shown in Scheme 1. route,2,7,12–15 with the corresponding low yields obtained during
Cyclohexylaniline 5 and dicyclohexylamine 6 are not produced the oxidation of cyclohexane to produce cyclohexanone.
by the downstream hydrogenation route 2 -- 5 - 6 because
Fig. 1 shows the kinetics for the hydrogenation of nitro-
of the lack of efficient and selective catalysts for these trans- benzene 2 with a catalytic amount of commercially-available
formations. Instead, the longer cyclohexanone route 1 -- palladium on carbon (Pd/C), wherein each point corresponds to
8 -- 6 must be followed.5
an independent batch reaction. The results show that aniline 3
To date, the cross amination of aniline and cyclohexylamine, is rapidly formed as a primary and unstable product that, by
3 + 4 - 5, remains a challenge.6 The most active and selective partial hydrogenation, gives cyclohexylaniline 5 as a secondary
catalyst for this transformation is an organoruthenium(II) product. This, by further hydrogenation, produces minor amounts
hydride complex (known as Shvo’s catalyst) reported by Beller of dicyclohexylamine 6, as a tertiary product. No intermediate
diazo products are detected (GC).16–18 The reaction sequence
was confirmed by reacting cyclohexylaniline 5 in the presence
of Pd/C and obtaining dicyclohexylamine 6 with high selectivity,
as a primary product.
Cyclohexylamine 4 was not observed among the reaction
products, and since the hydrogenation of the primary product
aniline 3 is very plausible, a possible way to produce cyclo-
hexylaniline 5 would be the coupling of 3 with 4. To check this
hypothesis, aniline 3 and cyclohexylamine 4 were reacted under
Scheme 1 Main industrial routes for primary and secondary benzene-derived
amines. Dashed lines represent the processes studied here.
´
´
´
Instituto de Tecnologıa Quımica, Universidad Politecnica de Valencia-Consejo
´
Superior de Investigaciones Cientıficas (UPV-CSIC.) Avda. De los Naranjos S/N,
Valencia, Spain. E-mail: anleyva@itq.upv.es, acorma@itq.upv.es;
Fax: +34 96 387 7809; Tel: +34 96 387 7800
Fig. 1 Plot-time yield for the hydrogenation of nitrobenzene 2 with the Pd/C
catalyst at 30 1C, without (left) or with (right) 1 eq. of methanesulfonic acid.
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3cc44064h
c
8160 Chem. Commun., 2013, 49, 8160--8162
This journal is The Royal Society of Chemistry 2013

View Article Online
Communication
ChemComm
Table 1 Scope for the Pd/C-catalyzed one-pot synthesis of secondary amines by
hydrogenation of nitrocompounds as single starting materials
Yield
(%)
Fig. 2 A proposed mechanism for the Pd/C catalyzed hydrogenation of nitro-
benzene 2 to cyclohexylaniline 5 and dicyclohexylamine 6.
Run Substrate (s)
1
Product
82
the above reaction conditions in the presence of Pd/C and the
coupled products 5 and 6 were observed. This result strongly
suggests that Pd/C catalyzes the formation of 5 and 6 from
nitrobenzene 2 in the presence of H2 by in situ coupling of
aniline 3 and cyclohexylamine 4. A hydrogen-borrowing mecha-
nism would explain the formation of cyclohexylaniline 5, as
shown in Fig. 2, which is consistent with the kinetic experi-
mental results shown in Fig. 1 since aniline 3 is smoothly
consumed while the coupled products are formed.
Following the reaction scheme presented in Fig. 2, where the
cyclohexanimine formed is considered the key intermediate for
the aminocoupling reaction, it appears that one could increase 6a
the rate of formation of secondary amines by activating this
2
71
55
3a
4
5
75
79
88
90
90
57
50
50
72
60
¨
imine. To do so, different Bronsted acids, which are able to
7
activate the imine, were added to the reaction mixture (Table S1,
ESI†) and the yield (>80%) and selectivity (>90%) to cyclo-
hexylaniline 5 were highly increased. Fig. 1 shows that good
yields of 5 are obtained in o3 h reaction time and that the
formation of 5 starts exactly when the concentration of aniline
3 decreases, in accordance with the mechanism proposed in
Fig. 2. Cyclohexane was not detected (GC) under the optimized
reaction conditions.‡
To check the generality of the transformation, different
nitroaromatics were tested, and the results in Table 1 show that
good yields of different cyclohexylanilines and phenylethylamines
can be obtained by homocoupling (entries 1–10) and also by
heterocoupling (entries 11–15) of nitroderivatives, asymmetric
cyclohexylanilines being obtained in reasonable yields. Note
that the hydrogen-borrowing amination coupling reaction has
been reported in the literature with different metal catalysts at
reaction temperatures >140 1C6 and that other methods to
obtain substituted secondary amines involve substituted arenes,
ketones and/or amines. Here the reaction temperature is
drastically reduced and the starting nitrocompounds are early
products in the manufacturing chain. In addition, the process is
also valid for obtaining dicyclohexylamines (entries 16 and 17)
by just permitting the reaction to proceed for longer times or
increasing the reaction temperature to 60 1C. Indoline can also
be obtained (entry 18).
8
9a
10a
11a
12a
13
14
55
78
72
15
16a
17a
18
96
64
Two additional Pd/C catalysts with different particle size
distributions were then tested in the reaction and the results
are shown in Fig. 3.19,20 High angle annular dark field scanning
transmission electron microscopy (HAADF-STEM) together with
a
GC yields. Average of three experiments. 60 1C.
HR-TEM and EDX measurements (Fig. 3 and Fig. S1–S3, ESI†) and this order fits the catalytic activity of the material towards
show that the relative population of o3 nm palladium particles cyclohexylaniline 5. The corresponding high-resolution images
on the catalysts decreases in the order Pd/C > Pd/C (b) > Pd/C (c) show small microfaceted nanoparticles and these spectroscopic
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 8160--8162 8161

View Article Online
ChemComm
Communication
here operates at room temperature (B100 1C lower than homo-
geneous catalysts previously reported) directly from nitro-
derivatives. This strategy constitutes, therefore, a cost-effective
and environmentally-friendly methodology to prepare secondary
amines. In addition, this process exemplifies the concept of
modern catalysis based on the identification and optimization
of the active sites on reusable solids for atom-economical
domino reactions.
Financial support by Consolider-Ingenio MULTICAT 2010,
subprograma de Apoyo a Centros y Universidades de Excelencia
Severo Ochoa (SEV 2012 0267), and MAT2009-00889 projects
from MICINN is acknowledged. P. R.-M. thanks the Ministry of
Education for a concession of a FPU contract. A. L. P thanks ITQ
a
Fig. 3 Dependence on the catalytic activity of Pd/C on the particle size. The
Pd/C (c) catalyst presents a bi-modal particle size distribution with a minor number
of small (o3 nm) Pd particles that account for the catalytic activity observed. HR-TEM
images of each palladium/carbon: (A) Pd/C; (B) Pd/C (b) (C) Pd/C (c).
´
for a contract. We thank Dr J. C. Hernandez-Garrido for the
microscopic images.
results suggest that the reaction proceeds on the surface of very
small nanoparticles with a significant number of exposed
palladium atoms.
The process was scaled-up in the laboratory and very good
yields and selectivity were still found at the gram scale (Table S2,
ESI†). The possible leaching of metals during the reaction was
studied by the filtration test and the results show that no active
catalyst species in solution were released during the process
(Fig. S4, ESI†), and recycling of the catalyst was also successful
at the gram scale (Table S3, ESI†).
Notes and references
‡ A typical reaction procedure for the synthesis of cyclohexylaniline 5
from nitrobenzene 2: Pd/C (5 wt%, 21.2 mg, 0.01 mmol of Pd) and
hexane (0.5 mL) were placed into the reactor (2 mL capacity) equipped
with a magnetic stirrer. Nitrobenzene 2 (21 mL, 0.2 mmol) and methane-
sulfonic acid (10 mL, 0.15 mmol) were added, and after the micro-
reactor was sealed, air was purged by flushing out four times with
hydrogen and then pressurized with 6 equivalents of H2 (B10 bar). The
resulting mixture was magnetically stirred overnight under static pressure
at room temperature. During the experiment, the hydrogen pressure
was decreased as the reaction proceeds. The product composition was
determined by means of GC, once the catalyst particles were removed
from the solution by filtration. The products were identified using
GC-MS and also by 1H NMR and 13C NMR.
Proton Nuclear Magnetic Resonance (1H-NMR) analyses of
the liquid phase after the reaction showed the presence of
ammonia as ammonium salt (Fig. S5, ESI†), which supports the
hydrogen-borrowing mechanism shown in Fig. 2, but opens the
question if the role of the acid is activating the imine or
quenching ammonia, or both. To differentiate those possibilities,
a solid substitute of methanesulfonic acid such as Amberlyst A15,
used in separated particles to those of Pd/C, was introduced as
a solid acid. Since catalysis operates on the solid surface, the
use of a separated solid acid should hamper the reaction yield
by the restricted contact between solids. Indeed, it was found
that the yield of 5 (Table S4, ESI†) strongly decreased in the
presence of Amberlyst A15, supporting the hypothesis that
1 J. He, J. W. Kim, K. Yamaguchi and N. Mizuno, Angew. Chem.,
Int. Ed., 2009, 48, 9888–9891.
2 M. H. S. A. Hamid, P. A. Slatford and J. M. J. Williams, Adv. Synth.
Catal., 2007, 349, 1555–1575.
3 E. Merino, Chem. Soc. Rev., 2011, 40, 3835–3853.
4 H.-U. Blaser, H. Steiner and M. Studer, ChemCatChem, 2009, 1,
210–221.
5 K. S. Hayes, Appl. Catal., A, 2001, 221, 187–195.
´
6 G. Guillena, D. J. Ramon and M. Yus, Chem. Rev., 2009, 110,
1611–1641.
7 D. Hollmann, S. Bahn, A. Tillack and M. Beller, Chem. Commun.,
2008, 3199–3201.
¨
8 D. Hollmann, S. Bahn, A. Tillack and M. Beller, Angew. Chem.,
Int. Ed., 2007, 46, 8291–8294.
9 T. Kiyuma and I. Naramoto, Method of production of cyclohexylamines,
New Japan Chemical Co., Ltd., Kyoto, JP, 1990.
10 G. Deng, W. Chen and C.-J. Li, Adv. Synth. Catal., 2009, 351,
353–356.
11 S. Han and M. Movassaghi, J. Am. Chem. Soc., 2011, 133, 10768–10771.
12 D. Pingen, C. Mu¨ller and D. Vogt, Angew. Chem., Int. Ed., 2010, 49,
8130–8133.
¨
the Bronsted acid probably activates the imine. In a second
experiment, H2 was continuously flowed in order to desorb the
released ammonia and avoid the poisoning of the metal
catalyst. When the reaction was executed with this experimental
set-up and without adding any external acid, cyclohexylaniline 5
´
was produced in good yield and selectivity. This indicates that 13 A. Corma, T. Rodenas and M. J. Sabater, Chem.–Eur. J., 2010, 16,
254–260.
the presence of ammonia is detrimental for the reaction and
that the role of the acid is not only to activate the imine
14 G. E. Dobereiner and R. H. Crabtree, Chem. Rev., 2009, 110,
681–703.
´
intermediate but also to quench the 0.5 equivalents of released 15 A. Prades, R. Corberan, M. Poyatos and E. Peris, Chem.–Eur. J., 2008,
14, 11474–11479.
ammonia.
´
16 A. Corma, P. Concepcion and P. Serna, Angew. Chem., Int. Ed., 2007,
In conclusion, an efficient and selective one-pot formation of
46, 7266–7269.
cyclohexylanilines and dicyclohexylamines from nitroderivatives 17 A. Corma and P. Serna, Science, 2006, 313, 332–334.
´
18 A. Grirrane, A. Corma and H. Garcıa, Science, 2008, 322, 1661–1664.
has been accomplished with commercially-available small
(o3 nm) Pd nanoparticles on carbon as catalysts. The solid-
19 S.-Y. Huang, S.-M. Chang and C.-T. Yeh, J. Phys. Chem. B, 2005, 110,
234–239.
catalyzed hydrogen-borrowing amination coupling presented 20 M. T. Reetz and M. Maase, Adv. Mater., 1999, 11, 773–777.
c
8162 Chem. Commun., 2013, 49, 8160--8162
This journal is The Royal Society of Chemistry 2013