Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics

Article abstract of DOI:10.1016/j.cattod.2016.04.020

We report herein on the development of Ni-based catalyst using activated carbon fibres (ACFs) as a structured support and its application for the three‐phase hydrogenation of nitroarenes (T?=?353?K; P?=?10?bar). It was shown that metallic Ni⁰ n

Full text of DOI:10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Ni-based structured catalyst for selective 3-phase hydrogenation of
nitroaromatics
Oliver Beswicka, Daniel Lameya, Félicien Muriseta, Thomas LaGrangeb, Loïc Obersonb,
Songhak Yoonc, Esther Sulmand, Paul J. Dysona, Lioubov Kiwi-Minskera,d,∗
a Group of Catalytic Reaction Engineering, Ecole Polytechnique Fédérale de Lausanne (EPFL), Station 6, Lausanne CH-1015, Switzerland
b Interdisciplinary Centre for Electron Microscopy (CIME), Ecole Polytechnique Fédérale de Lausanne (EPFL), Station 12, Lausanne CH-1015, Switzerland
c Laboratory for Solid State Chemistry and Catalysis, Swiss Federal Laboratories for Materials Science and Technology (EMPA), 8600 Dübendorf, Switzerland
d Regional Technological Centre, Tver State University, 170100 Tver, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
We report herein on the development of Ni-based catalyst using activated carbon fibres (ACFs) as a struc-
tured support and its application for the three-phase hydrogenation of nitroarenes (T = 353 K; P = 10 bar).
It was shown that metallic Ni0 nanoparticles (NPs) with a mean diameter of ∼2.0 nm stabilized by the
ACF microporous network were responsible for the catalytic transformation. To obtain optimum catalytic
activity, the Ni/ACF catalyst must be freshly prepared and activated in situ by H2 at T > 353 K. Pre-treatment
of the ACFs by nitric acid boosted the activity of the Ni/ACF catalyst, which exhibits high performance in
hydrogenation of nitrobenzene to aniline (yield, Y ∼100%). The catalyst was tested for the reuse attaining
a quasi-steady-state after the sixth reaction thereafter remaining relatively stable over seven consec-
utive runs. Near-quantitative transformation (Y > 99%) of p-chloronitrobenzene to p-chloroaniline was
achieved under mild conditions over the Ni/ACF catalyst with a ca. 20-fold higher activity than conven-
tional Raney Ni. Thus, new catalyst reported here represents a significant step forward towards a simple,
heterogeneous catalytic selective hydrogenation of nitroarenes that employs H2 as the hydrogen source.
© 2016 Elsevier B.V. All rights reserved.
Received 1 December 2015
Received in revised form 19 April 2016
Accepted 25 April 2016
Available online xxx
Keywords:
Selective hydrogenation
Nitrobenzene
p-Chloroaniline
Ni nanoparticles
Activated carbon fibres
Structured catalyst
1. Introduction
gas-phases [1,18] while Raney nickel is the only commercial cat-
alyst used on the industrial scale for Cl-nitroarenes hydrogenation
Functionalized anilines are valuable intermediates used in the
synthesis of agrochemicals, pharmaceuticals, pigments and dyes
[1]. They are generally prepared by hydrogenation of the cor-
responding nitroarenes. Traditional non-catalytic processes have
been largely abandoned, being replaced by catalytic hydrogena-
tion processes [2]. The catalytic route allows high yields without
the generation of toxic wastes and smaller amount of by-products
with cost efficient separation and treatment processes. For substi-
tuted anilines the preservation of functional groups is one of the
main challenges, with the exclusive reduction of the nitro-group
while not affecting other groups being crucial [1,3].
[1]. The influence of metal nanoparticles (NPs) size on catalytic
response including the catalyst selectivity and stability has been
widely reported [20–25], although the control of the NP size dur-
ing the preparation of supported catalysts is difficult to ensure. In
general, the highest dispersion of active metal is targeted in order
to have a sufficient catalytic performance per mole of active metal.
The majority of hydrogenation catalysts are used as powders
to avoid transport limitations, but these powders lead to prob-
lems during downstream separation. Possible solutions lie in the
use of structured catalysts that allow simple handling and con-
trolled fluid dynamics [14,26]. ACFs in the form of tissues are ideal
catalytic structured supports, characterized by a highly porous net-
work, high specific surface areas (SSAs) (up to 3000 m2 g−1) and low
resistance to fluid flow [10,27–30]. ACFs have been used for the
deposition of noble metal NPs and exhibit good performances in
the multi-phase hydrogenation [31–34]. The carbonaceous surface
can be pre-treated in acidic media generating different oxygen-
containing groups [35,36]. In this way, the morphology of the
supported Me NPs and their catalytic response can be controlled.
The majority of heterogeneous catalytic systems used to hydro-
genate nitroarenes are based on precious metals such as Au, Pd,
Pt, Rh or Ru [4–14]. Nevertheless, catalysts based on Fe, Co, Cu
or Ni [1,9,11,15–19] show promising activity in both liquid and
∗
Corresponding author at: Group of Catalytic Reaction Engineering, Ecole Poly-
technique Fédérale de Lausanne (EPFL), Station 6, Lausanne CH-1015, Switzerland.
E-mail address: lioubov.kiwi-minsker@epfl.ch (L. Kiwi-Minsker).
http://dx.doi.org/10.1016/j.cattod.2016.04.020
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
2
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
Herein, we present a facile preparation of ACF-based structured
catalysts with supported Ni-NPs of high dispersion (1–2 nm) as an
active phase. Different pre-treatments of ACF were applied and
the influence of preparation conditions on catalytic response has
been assessed in nitroarenes hydrogenation. A battery of physico-
chemical techniques, such as N2 physisorption, XRD, XPS, AAS, SEM,
STEM-HAADF imaging, temperature programmed decomposition
(TPD) and reduction by hydrogen (TPR) was applied for catalysts
characterization. The activity/selectivity of the various Ni/ACF cat-
alysts was compared to industrial Raney nickel which serves as a
benchmark catalyst.
BET surface areas were measured using a Micromeritics 3Flex.
Prior to analysis, the samples were degassed at 393 K for 3 h under
vacuum (<1 × 10−5 bar). N2 adsorption/desorption isotherms were
recorded over the range 0.000002 ≤ P/P0 ≤ 0.99. The specific surface
area and the total pore volume were obtained using the BET method
[38].
Temperature Programmed Decomposition (TPD) was conducted
in a He flow (50 cm3 min−1) from RT to 835 K at 5 K min−1 in a
tubular reactor inside an oven (Carbolite MTF 10/25/130) and the
outlet flow was continuously monitored using a Pfeiffer Vacuum
ThermoStarTM GSD 300 T2. The NO2 signal was mainly detected as
a NO fragment (m/z = 30).
Powder X-ray diffractograms (XRD) were recorded on
a
2. Experimental
Bruker/Siemens D500 incident X-ray diffractometer using Cu K␣
radiation. The samples were scanned at 0.004◦ s−1 over the range
20◦ ≤ 2␪ ≤ 90◦ (scan time = 5 s step−1). Diffractograms were identi-
fied using the JCPDS-ICDD reference standard, i.e. Ni (89-7129), NiO
(89-5881) and NiO2 (89-8397). In situ XRD patterns were obtained
using a PANalytical X’Pert PRO ␪–␪ scan system with CuK␣ radi-
ation (␭ = 1.5418 Å). The diffraction patterns were recorded at a
scanning rate of 0.015◦ s−1. The Ni(NO3)2/ACFHNO3-373 sample was
placed in the heating chamber (XRK 900, Anton Paar) and gas was
applied through the mass flow controller (5850 TR, Brooks instru-
ment). About 0.03 g of sample was heated under a 60 cm3 min−1
flow of N2 up to 723 K with a heating rate of 2 K min−1. Long scans
(30◦ ≤ 2␪ ≤ 85◦) were acquired on the fresh sample directly after
heating to 723 K. In a separate experiment on the same starting
material, the procedure was modified by replacing the N2 by 10%
v/v H2/N2. A long scan was performed at the end of the procedure.
Peak fitting was performed using the EVA (DIFFRAC.Suite) software.
X-ray photoelectron spectroscopy (XPS) analysis was conducted
on PHI VersaProbe II (Physical Instruments AG). The monochro-
matic Al K␣ X-ray source power was maintained at 24.8 W and the
emitted photoelectrons were obtained from a 100 ␮m × 100 ␮m
samples. The spherical capacitor analyser was set at a 45◦ take-
off angle with respect to the sample surface. The analyser pass
energy was 188 eV for survey spectra (0–1300 eV) and 47 eV for
high resolution spectra (Ni 2p3/2 and Ni 2p1/2). For the latter res-
olution a full width at half maximum of 0.91 eV for the Ag 3d 5/2
peak was obtained. The adventitious C (284.8 eV) 1 s peak was used
as an internal standard to compensate for any charging effects.
2.1. Materials
Nickel(II) nitrate hexahydrate (Fluka, ≥98.0%), nitric acid (VWR
chemicals, 65%), hydrogen peroxide (reactolab SA, 30%), nitroben-
zene (99%, Acros Organics), 1-chloro-4-nitrobenzene (99%, Acros
Organics), 1,3-dinitrobenzene (Tokyo Chemical Industry, ≥99.0%),
dodecane (99%, Acros Organics), methanol (MeOH, ≥99.8%, Sigma-
Aldrich), ethanol (EtOH, 99.8%, Sigma-Aldrich), toluene (>99%,
AppliChem) were used as received. The solvent used for the reac-
tion was technical EtOH (95% + 5% MeOH, Brenntag). All gases (H2,
N2, and Ar) were of high purity (Carbagas Switzerland, >99.9%). The
ACF K-20 (∼2000 m2 g−1) was purchased from Kynol Europa GmbH.
2.2. Catalyst preparation
2.2.1. Structured Ni/ACF
The Ni/ACF catalysts were prepared as follows. Commercial ACF
was pre-treated in 15 wt.% HNO3 aq. solution for 15 min at 373 K
(ACFHNO3-373) and for 3 min at 298 K (ACFHNO3-298) to increase the
concentration of oxygen-containing groups on the carbon surface
[35,37]. The supports denoted as ACF were not pre-treated and used
as received. The ACF was impregnated with an ethanolic solution
of the Ni precursor (Ni(NO3)2·6H2O) ensuring complete filling of
the pores. The Ni loading was adjusted by varying the precursor
concentration. The impregnated ACF samples were dried at room
temperature (RT) overnight. The precursor decomposition was per-
formed via thermal treatment in a flow reactor (50 cm × 3 cm i.d.)
from RT to 673 K, 6 K min−1 under Ar flow (280 cm3 min−1), main-
+
Sputtering was realized using a 20 kV Ar2500 cluster source on a
tained at 673 K (1 h) under 17% v/v H2/Ar flow (340 cm3 min−1
)
2 nm × 2 nm sample surface at a rate of 0.45 nm/min (referenced to
SiO2). Curve fitting was performed using the CasaXPS software.
The Ni NPs size distribution was measured from “Z-contrast”
or high angle annular dark-field scanning transmission elec-
tron microscope (HAADF STEM) images acquired on a FEI Talos
F200S instrument operating at 200 keV. The Ni/ACF catalysts were
infiltrated-embedded in an EPON 812 epoxy resin and polymer-
ized at 333 K for 24 h, cut (20 nm) by ultramicrotomy to analyse
the fibre cross-section. The TEM samples were subjected to mild
(10 eV) plasma cleaning for 1 min using a Fischone 8070 plasma
cleaner operated with a forward power of 9 W.
Up to 1400 individual metal particles were counted for the
investigated catalyst and the circular diameter (di) was determined
from the area measured using imageJ software from images of
different magnifications. The mean circular diameter (d) was cal-
culated using Eq. (1):
and cooled to RT under Ar. To prevent oxidation of the pyrophoric
Ni/ACF, the samples were passivated (1 h) at RT in 2.8% v/v air/Ar
flow (145 cm3 min−1). Gas flows were controlled using an Agilent
Technologies ADM1000 Universal Gas Flowmeter, values are given
in IUPAC STP [40].
2.2.2. Raney nickel catalyst
1 g of Ni-Al (50:50 wt.% alloy, Alfa Aesar) was treated with an
aq. solution of KOH (10 wt.%) and stirred at RT until bubbling (H2)
stopped (∼1 h), then at 333 K for 0.5 h. The resulting solid (Raney
Ni) was decanted, washed with distilled water and ethanol.
2.3. Catalyst characterization
The Ni content was determined by absorption atomic spec-
troscopy (AAS) using a Shimadzu AA-6650 spectrometer with
an air-acetylene flame. Temperature-programmed reduction in
hydrogen (H2-TPR) was carried out on a Micromeritics Autochem
II 2920 by heating the sample in 17 cm3 min−1 5% v/v H2/N2 from
RT to 973 K at 2 K min−1. The exit gas was passed through a liquid
N2 trap and changes in H2 consumption/release were monitored
by TCD with data acquisition/manipulation using the TPR WinTM
software.
ꢀ
nidi
i
d =
(1)
ꢀ
ni
i
where ni is the number of particles of diameter di.
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
3
2.4. Hydrogenation of nitroaromatic compounds: set-up and
analytics
Liquid phase hydrogenation reactions (T = 353–423 K;
PH = 10 bar) were carried out in
a commercial semi-batch
sti2rred reactor (150 cm3 stainless steel autoclave, Büchi AG, Uster,
Switzerland) equipped with baffles. The reactor temperature was
controlled using a water heating bath (Colora K4, Lorch, Germany)
connected to the reactor jacket. Mixing was achieved with a stirrer
with a 6-blade disk turbine impeller and four wire-mesh blades
onto which the structured ACF catalyst was fixed. The pressure and
temperature inside the reactor as well as the stirring speed were
monitored via a control unit (bpc 6002/bds mc, Büchi AG, Uster,
Switzerland). Prior to each experiment, the reactor was charged
with the catalyst and 100 cm3 of an ethanolic solution of dodecane
1 g L−1 (internal standard) and flushed with N2. The reactor was
then heated to 353 K under stirring (2000 rpm) and stabilized for
5 min. H2 was introduced into the reactor to a total pressure of
10 bar (sum of the partial pressures of EtOH and H2), which was
maintained throughout the reaction. The catalyst was reduced
in-situ for 30 min. The substrate was subsequently injected (time
t = 0) via a gas-tight syringe.
Fig. 1. TPD profile of Ni(NO3)2/ACFHNO3-373
(
CO; CO2; NO).
An initial NO2/Ni molar ratio of 50 was targeted. A non-
invasive liquid sampling system allowed the controlled removal
of aliquots (≤0.4 cm3) which were analysed using a Perkin-Elmer
Clarus 500 GC equipped with programmed split/splitless injector,
FID and Stabilwax (Cross-bond Carbowax-PEG, Restek, USA) capil-
lary column (i.d. = 0.32 mm, length 30 m, film thickness 0.25 ␮m).
Data acquisition/manipulations were performed using TotalChrom
Workstation v.6.3.2. The concentrations of organic species and the
conversion of the substrate, A, is defined using Eq. (2):
3.1.1. BET
BET analyses were performed with different ACF samples and
the associated SSAs and pore volumes are summarized in Table 1.
A very high SSA (>2000 m2 g−1) was found for the as received, non-
treated ACF (entry 1). Nitric acid treatment at 373 K decreased
both the SSA and the pore volume. This effect has been already
reported by Shim et al. [39] and is due to: i) the increased amount
of functional O-groups on the surface and ii) a change in the carbon
morphology. By heating under an inert gas the HNO3 pre-treated
ACF recovered the initial SSA and the total pore volume. It is plau-
sible that the O-containing groups rendered part of the porous
structure inaccessible for nitrogen adsorption. The deposition of
5 wt.% Ni (entries 4 and 5) changed only slightly the SSA (by 5–8%)
and the pore volume of the ACF compared to the initial value, indi-
cating that the deposition of the metal salt does not lead to pore
blocking [40].
CA,0 − CA,i
XA(%) =
× 100
(2)
CA,0
where CA,0 and CA,i represent the substrate molar concentrations
at the beginning and time t, respectively. The selectivity towards
product B is calculated using Eq. (3):
CB
ꢀA
ꢀB
SB(%) =
×
× 100
(3)
CA,0 − CA,i
where CB0 and CB represent the molar concentrations of product B
at the beginning and time t, respectively, and ␯A and ␯B correspond
to the stoichiometric coefficients.
The yield towards the product B with respect to substrate A is
defined from Eq. (4):
3.1.2. Temperature programmed decomposition (TPD) and
reduction (TPR)
The ACF freshly impregnated by the Ni precursor, after dry-
ing, was studied by TPD to determine the decomposition of the
precursor. The NO signal at the reactor outlet was monitored and
the results are presented in Fig. 1. The NO peak at 432 K corre-
sponds to the decomposition (323–473 K) of the surface nitrates
leading to the formation of NO2 [41]. The second NO peak at 521 K
results exclusively from the decomposition of the HNO3 vapour
(473 K–573 K) into NO2 [42]. At the end of the calcination step NiO
NPs are formed. Therefore, during catalyst preparation, the calci-
nation step was at 673 K to ensure complete decomposition of the
nitrates. The surface O-groups are decomposed giving CO and CO2
during TPD runs [37].
CB,0 − CB,i
ꢁA
YB % =
( )
·
× 100
(4)
CA,0
ꢁB
Repeated reaction runs with the same batch of the catalyst
delivered values of the reaction mixture compositions that were
reproducible to within 4%. The catalytic activity was quantified in
terms of the initial substrate consumption rate (−RA,0) determined
from a linear regression of the temporal substrate concentration
profile, Eq. (5):
In order to choose a suitable reduction temperature, after the
calcination step the catalyst was studied by TPR. The H2-TPR pro-
file of the NiOx/ACFHNO3-373 material exhibits two peaks at 602 and
778 K (Fig. 2A). The first one (602 K) is attributed to the reduction
of NiOx NPs and the second one is due to methanation of ACF. Per-
forming the same procedure with MS instead of TCD confirmed the
peak attribution to CH4, H2 and H2O signals. Therefore, the reduc-
tion step during the catalyst preparation was set at 673 K for 1 h to
ensure formation of metallic Ni NPs while minimizing detrimental
methanation of the ACF support.
(CA,0 − CA) · V
−RA,0(molAmol-N1imin-1) =
(5)
nNi · tR
where V is the reaction volume, nNi is the number of moles of Ni,
and tR the reaction time.
3. Results and discussion
3.1. Catalyst characterization
In order to rationalize the observed catalytic results, all samples
were characterised by different physico-chemical methods.
Catalytic testing has demonstrated that in-situ activation under
H2 of freshly prepared catalysts was necessary to obtain significant
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
4
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
Table 1
BET analysis for ACF support with and without acid treatment and Ni/ACF catalysts.
N◦
Sample
Ni content (wt%)
Comment
SSA 50 (m2 g−1
)
Specific pore volume 0.02 (cm3 g−1
)
1
2
3
4
5
ACF
0
0
0
4.8
3.8
as received
2120
1750
2000
1960
2000
0.88
0.73
0.84
0.83
0.79
ACFHNO3-373
ACFHNO3-373-ꢂ
Ni/ACF
HNO3 treateda
HNO3a + thermal treatment in Ar at 673 K
Calcination 673 K − reduction 673 K − passivation at RT
Calcination 673 K − reduction 673 K − passivation at RT
Ni/ACFHNO3-373
a
Treated with nitric acid: 15% HNO3, 373 K, 15 min.
Fig. 3. In-situ XRD profile of ACF (A), Ni(NO3)2/ACFHNO3-373 before (B) and after treat-
ments in N2 (C) and N2/H2 (D) at 723 K with JCPDS-ICDD reference diffractogram for
Ni0 (standard card 04-0850).
the temperature and the larger the hydrogen consumption. The H2-
TPR profile of the sample exposed for to air 24 h exhibited a peak
maximum at 486 K and for the sample exposed for 6 days a peak
maximum was at 501 K. After 47 days of air exposure the profile
has two peaks at 521 and 564 K, suggesting that to activate the
catalyst in-situ a temperature of 353 K is not enough.
3.1.3. In-situ XRD analysis
The nature of the Ni phase was studied in-situ by XRD analysis
following each step of the precursor decomposition/reduction until
no more changes were observed to the diffraction pattern. Pattern
A shown in Fig. 3 represents the initial ACF support and pattern
B corresponds to the ACF after deposition of the Ni(NO3)2·6H2O
precursor. A sample of impregnated ACF was subjected to a ther-
mal treatment (298 K–723 K, 2 K min−1) under a 60 cm3 min−1 flow
of N2 and left during one hour at the final temperature (Fig. 3C).
Finally, pattern D represents the impregnated sample subjected to a
thermal treatment (298 K–723 K, 2 K min−1) under a 60 cm3 min−1
flow of 10% v/v H2/N2. Diffraction patterns C and D show addi-
tional components compared to patterns A and B, suggesting that
both N2 and H2/N2 treatments lead to the formation of a Ni(0) phase
by comparison with the JCPDS-ICDD reference standard, i.e. fcc Ni
(04-0850).
Fig. 2. TPR-H2 profile of (A) NiOx/ACFHNO3-373 (after nitrates decomposition) and
(B) Ni/ACFHNO3-373 after reduction, passivation and exposition in air during different
time.
catalytic activity. Since the catalyst was passivated after reduction,
this activation seems to be responsible for the removal of thin oxide
layer from the Ni NPs surface. To confirm this assumption and to
get a better insight on the dynamics of oxidation of the Ni NPs,
the passivated samples were exposed to air for different periods
of time and analysed by H2-TPR. Without exposure to air (0 day,
Fig. 2B) the hydrogen consumption was lower with the maximum at
426 K. Therefore, the Ni-oxides species are readily reducible during
in-situ activation. The longer was the air exposure time, the higher
The XRD analysis of the passivated Ni/ACFHNO3-373 catalyst
exposed to air over 6 days generated the pattern shown in Fig. 4. The
profile exhibits four main peaks over the 2ꢃ range 20◦–90◦, corre-
sponding to the (222), (400), (440) and (440) planes of face-centred
cubic NiO (JCPDS-ICDD 89-5881), and confirming the presence of
nickel oxide after the standard treatment and air exposure.
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
5
2p1/2 peak satellite at 878.1 eV, half-way from NiO (879 eV) and Ni0
(874 eV), could indicate some contributions from Ni0.
In summary, XRD and XPS studies confirm that the conditions of
Ni/ACF preparation lead to the formation of metallic Ni0 NPs during
the reduction under hydrogen. These particles were covered by a
thin Ni-oxide/hydroxide layer formed during the passivation step.
3.1.5. STEM
HAADF STEM measurements were performed to observe the
size and morphology of the Ni NPs. In the images they appear as
bright contrast within the ACF support composed of a lower density
carbon material (Z-contrast).
Fig. 6 shows the image of Ni/ACFHNO3-373 (A) and outlines of the
Ni NPs of the same sample (B). The STEM derived histogram for this
sample (C) is characterized by a distribution of the NP size between
1.6–2.6 nm with 53% within the range. The mean particle diameter
was found 2.1 1 nm. Analysis of NiOx/ACF (with non-treated by
HNO3 support) reveals NPs in the range of 1.4–2.3 nm for 52% of the
NPs with a mean particle size of 1.7 0.8 nm. The mean diameter of
NiOx NPs on acid pre-treated support, ACFHNO3-373, formed before
the reduction in H2-flow (images not shown) is 1.8 0.9 nm, being
similar to those without acid treatment.
Fig. 4. XRD pattern of Ni/ACFHNO3-373 after 6 days of exposure in air with JCPDS-ICDD
reference diffractogram for NiO (standard card 89-5881).
Thus, the thermal treatment in the presence of hydrogen leads
to a slight increase of mean NPs size. Note that the presence
of particles with sizes below the detection threshold, not visible
with the high magnification (1.3 Mx), cannot be excluded. Imag-
ing with the aberration corrected microscope (FEI Titan Themis) of
the Ni/ACFHNO3-373 specimen suggested that a larger proportion of
NPs in the sub-nm range (0.5–1 nm) exist than detected with the
FEI Talos F200S. The enhanced catalytic properties of these sub-
nm range NPs could arise from their irregular shape and dangling
chains of atoms on the surface, increasing the number active sites.
Also, the distribution of the Ni/ACFHNO3-373 specimen (Fig. 6C) sug-
gests that the Ni-NPs are attached rather than stacked. Therefore,
the effective NPs surface available for catalysis might be greater for
this sample, despite its larger mean diameter.
In summary, the oxidative pre-treatment of ACF support by
HNO3 weakly affects the NiOx NPs size. In contrast, the reduc-
tive treatment at 673 K slightly diminishes the Ni-NPs dispersion
with a concomitant increase of average diameter from 1.8 0.9 to
2.1 1.0 nm.
3.2. Catalytic results
3.2.1. Effect of the ACF acidic pre-treatment on catalytic
performance
Fig. 5. XPS profiles of freshly prepared Ni/ACFHNO3-373 (A) before and (B) after 5 min
of Ar sputtering.
It is well-known that the concentration of oxygen-containing
(O-) groups on an ACF surface can be increased by acidic treatments
[37]. In this study we applied a HNO3 treatment at 298 K or 373 K
to modify the ACF (see Section 2). Nickel (5 wt.%) was deposited on
pre-treated and non-treated ACFs and the obtained catalysts were
tested in p-chloronitrobenzene (p-CNB) to p-chloroaniline (p-CAN)
hydrogenation. Fig. 7 shows that the selectivity of all the catalysts
was close to 100% while the activity drastically increases (3-fold) for
the ACF treated at 373 K. It is worth to noting that a yield of p-CAN
close to 100% was obtained for all catalysts before dechlorination
(cleavage C Cl) starts. Catalytic tests also demonstrated that in-
situ activation under H2 of freshly prepared/passivated catalysts
was necessary to obtain catalytic activity.
The effect of the ACF oxidative treatment can be related to the
Ni NPs size and to the concentration and the nature of functional
O-groups on the carbon surface. The HNO3-treated carbon was pre-
viously shown to display a high density of weak and strong acid
sites [44]. During the calcination step in the catalyst preparation it
is likely that the strong O-containing groups (rendering CO2 dur-
3.1.4. XPS analysis
Freshly prepared passivated Ni/ACFHNO3-373 was studied by XPS
before and after argon sputtering to gain information on the Ni
NPs composition. Before sputtering the profile exhibits a symmet-
ric Ni 2p3/2 peak at 855.7 eV and its satellite at 861.3 eV, respectively
(Fig. 5A). The contributions of the main peak of NiO produce two
visible maxima (at 853.7 and 855.4 eV) whereas a single symmetric
peak is detected corresponding to Ni(OH)2 [43]. The higher inten-
sity of the Ni 2p1/2 peak measured at 873.7 eV compared to the
satellite at 880.1, indicates that the NPs surface consists mainly of
Ni(OH)2. A contribution of NiO cannot be completely ruled out since
the carbonaceous nature of catalyst results in low signal-to-signal.
After 5 min of Ar sputtering (2 nm) the position of the Ni 2p3/2
peak maxima was found to be shifted by 1.5 eV to 854.2 eV (Fig. 5B).
The shape presents an asymmetry visible on the left side of the peak
that is characteristic of a NiO contribution. The position of the Ni
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
6
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 6. Representative STEM images (A), outlines of the particles isolated after ImageJ processing (B) and related particles size distribution (C) for 5% Ni/ACFHNO3-373 after
reduction in H2 followed by passivation.
Fig. 7. Effect of ACF pre-treatment by HNO3 on the selectivity to p-CAN (hatched
bars) and activity (open bars) for the hydrogenation of p-CNB. Reaction conditions:
T = 353 K; PH2 = 10 bar; Cp-CNB,0 = 1.2 × 10−1 mol L−1, molar p-CNB:Ni ratio = 50.
Fig. 8. Concentrations vs reaction time during NB hydrogenation over
Ni/ACFHNO3-373 (ᮀ NB; AN: azobenzene; azoxybenzene; nitrosobenzene).
ing TPD runs) were removed whereas majority of weak groups
(rendering CO in TPD) remain. In fact, during TPD performed over
the catalysts after the calcination step, a 60% higher amount of
CO was detected from Ni/ACFHNO3-373 as compared to Ni/ACF (4.3
vs 2.7 mmol g−1). These weak O-groups were reported by Toebes
et al. [45] to maintain the dispersion of Pt NPs on ACF while the
activity towards cinnamaldehyde hydrogenation was attributed to
electronic effects. Ramaker et al. demonstrated that electronega-
tive support oxygen atoms leads to the higher ionization potential
of the supported Pt NPs, increasing the TOF during hydrogenation
of a tetralin ring [46]. Therefore, we suggest that the larger num-
ber of weak O-groups of Ni/ACFHNO3 as compared to Ni/ACF favor
substrate adsorption on Ni NPs via the NO2 group and thereby
increasing the catalytic activity [47].
Reaction conditions: T = 353 K; PH2 = 10 bar; CNB,0 = 1.2 × 10−1 mol L−1, molar ratio
NB:Ni = 50.
the selectivity at close to full conversion) is the most important
parameter to assess the catalytic performance.
The 5 wt.% Ni/ACFHNO3-373 catalyst was also evaluated in other
challenging hydrogenations (Table 2). For the nitro-aromatics con-
taining functional groups which are sensitive to the reduction
conditions (e.g. alkenes, halogens, etc.), the maximum yields of the
desired anilines are the main issue. These yields were compared
to ones obtained using Raney Ni under the same reaction condi-
tions. The developed 5 wt.% Ni/ACFHNO3-373 catalyst outperformed,
by ca. 20-fold, the Raney Ni (industrial non-noble metal catalyst
used for this reaction) in terms of activity in p-CNB hydrogena-
tion and the maximum yield was also slightly higher (99.0 0.5
vs 98.0 0.5%). It is worth to note that dechlorination after reac-
tion completion (100% conversion of p-CNB) was much higher over
Raney Ni resulting in its rapid deactivation. At reaction completion
(X = 100%), the selectivity towards aniline of 0.25% and 1.50% were
obtained for the Ni/ACFHNO3-373 and the Raney nickel catalysts,
respectively. The 5 wt.% Ni/ACFHNO3-373 catalyst was also highly
efficient in m-dinitro-benzene hydrogenation affording a high yield
of m-nitro-aniline (99.0 0.5%).
3.2.2. Reaction scope: nitrobenzene hydrogenation
The Ni/ACFHNO3-373 catalyst is the most active and also has
been tested in the hydrogenation of nitrobenzene (NB). Represen-
tative concentration vs time profiles are shown in Fig. 8. Linear
dependency of NB consumption was observed until close to 100%
conversion. The aniline formation passes via azo- and azoxy-
intermediates indicative of the condensation route established by
Haber [1]. It is important to note that the selectivity and yield of
the desired aniline product continues to increase even after the full
conversion of NB due to the completion of consecutive transfor-
mations of intermediates. Therefore, the maximum yield (and not
Catalyst stability was assessed via 13 consecutive reaction runs
using the same 5 wt.% Ni/ACFHNO3-373 catalyst. Deactivation was
more pronounced during the first run with a tendency towards
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
7
Table 2
Activity and maximum yield of the amine for the hydrogenation of nitro-aromatics (NC) over 5 wt.% Ni/ACFHNO3 and Raney Ni. Reaction conditions: T = 353 K; PH2 = 10 bar;
CNC,0 = 1.2 × 10−1 mol L−1, molar NC:Ni ratio = 50.
−1
N◦
Catalyst
Substrate (S)
Activity, r0·102 (molNC mol−1
s
)
Maximum Yield 0.5 (%)
Time to reach X = 98% (min)a
171
15
409
23
Ni
1b
2
NB
NB
m-DNB
p-CNB
p-CNB
0.24
3.2
0.15
2.8
100.0
100.0
97.0
99.0
98.0
Ni/ACFHNO3-373
3b,c
4
5
Raney Ni
0.13
672
a
Normalized according to molar NC:Ni ratio towards the NB reduction.
Without in-situ activation.
Catalyst was tested at 423 K.
b
c
stabilization after the sixth run. The total activity drop (about 45%)
may be due to leaching of loosely attached Ni NPs during the first
run, some mechanical losses and slow accumulation of intermedi-
ates on the catalyst surface during the following reaction runs. The
post-reaction mixture was analysed using AAS and no traces of Ni
were detected. Therefore, the deactivation is probably due to con-
tamination from some residuals or due to reconstructions of active
sites in the course of the hydrogenation reactions, leading to the
quasi-steady-state attained after 6–7 runs.
[3] A. Corma, P. Serna, Science 313 (2006) 332–334.
[4] A. Corma, P. Concepcion, P. Serna, Angew. Chem. Int. Ed. 46 (2007) 7266–7269.
[5] F. Cardenas-Lizana, X.D. Wang, D. Lamey, M.S. Li, M.A. Keane, L. Kiwi-Minsker,
Chem. Eng. J. 255 (2014) 695–704.
[6] F. Cardenas-Lizana, D. Lamey, N. Perret, S. Gomez-Quero, L. Kiwi-Minsker,
M.A. Keane, Catal. Commun. 21 (2012) 46–51.
[7] F. Cardenas-Lizana, S. Gomez-Quero, M.A. Keane, ChemSusChem 1 (2008)
215–221.
[8] F. Cardenas-Lizana, S. Gomez-Quero, M.A. Keane, Catal. Commun. 9 (2008)
475–481.
[9] A. Corma, P. Serna, P. Concepcion, J.J. Calvino, J. Am. Chem. Soc. 130 (2008)
8748–8753.
In summary, the structured 5 wt.% Ni/ACF catalyst is highly effi-
cient for the 3-phase hydrogenation of nitro-aromatics tolerating
different functionalities. In addition to obtaining near-quantitative
yields of the desired anilines, the catalyst is reusable and simple to
handle.
[10] E. Joannet, C. Horny, L. Kiwi-Minsker, A. Renken, Chem. Eng. Sci. 57 (2002)
3453–3460.
[11] R. Raja, V.B. Golovko, J.M. Thomas, A. Berenguer-Murcia, W.Z. Zhou, S.H. Xie,
B.F.G. Johnson, Chem. Commun. (2005) 2026–2028.
[12] F. Cardenas-Lizana, C. Berguerand, I. Yuranov, L. Kiwi-Minsker, J. Catal. 301
(2013) 103–111.
[13] B. Coq, A. Tijani, F. Figueras, J. Mol. Catal. 68 (1991) 331–345.
[14] V. Holler, D. Wegricht, I. Yuranov, L. Kiwi-Minsker, A. Renken, Chem. Eng.
Technol. 23 (2000) 251–255.
[15] R.V. Jagadeesh, A.E. Surkus, H. Junge, M.M. Pohl, J. Radnik, J. Rabeah, H.M.
Huan, V. Schunemann, A. Bruckner, M. Beller, Science 342 (2013) 1073–1076.
[16] F.A. Westerhaus, R.V. Jagadeesh, G. Wienhofer, M.M. Pohl, J. Radnik, A.E.
Surkus, J. Rabeah, K. Junge, H. Junge, M. Nielsen, A. Bruckner, M. Beller, Nat.
Chem. 5 (2013) 537–543.
4. Conclusions
Structured catalysts consisting of small Ni NPs (∼2.0 nm) sup-
ported on activated carbon fibers (ACF) have been prepared which
efficiently catalyze the chemoselective hydrogenation of nitro-
aromatics under relatively mild conditions (T = 353 K; P = 10 bar).
The method used to prepare the catalyst was optimised, with pre-
treatment of the ACF by nitric acid (prior to the impregnation by
Ni-precursor) not affecting the Ni0 dispersion, but creating a high
concentration of functional O-groups leading to the enhanced (3-
fold) activity of the resulting catalyst in the hydrogenation of p-CNB.
In-situ XRD and XPS analysis showed that the Ni NPs (generated
during the calcination-reduction of the nickel salt) are metallic. The
TPD and H2-TPR data allowed the optimal reduction temperature
in H2-flow of 673 K to be established and indicated the necessity
for the passivated catalyst to undergo in-situ activation in the reac-
tion mixture at 353 K. Although deactivation of the catalyst was
observed following storage in ambient air, catalytic activity could
be regenerated by reduction at temperatures >560 K.
[17] K. Bhaumik, K. Akamanchi, Can. J. Chem. 81 (2003) 197–198.
[18] F. Cardenas-Lizana, S. Gomez-Quero, M.A. Keane, Appl. Catal. A Gen. 334
(2008) 199–206.
[19] M.V. Twigg, M.S. Spencer, Appl. Catal. A Gen. 212 (2001) 161–174.
[20] L.Z. Nikoshvili, A.S. Makarova, N.A. Lyubimova, A.V. Bykov, A.I. Sidorov, I.Y.
Tyamina, V.G. Matveeva, E.M. Sulman, Catal. Today 256 (2015) 231–240.
[21] G.C. Bond, Chem. Soc. Rev. 20 (1991) 441–475.
[22] N. Semagina, L. Kiwi-Minsker, Catal. Rev. 51 (2009) 147–217.
[23] A. Borodzinski, Catal. Lett. 71 (2001) 169–175.
[24] N. Semagina, L. Kiwi-Minsker, Catal. Lett. 127 (2009) 334–338.
[25] M. Crespo-Quesada, A. Yarulin, M.S. Jin, Y.N. Xia, L. Kiwi-Minsker, J. Am. Chem.
Soc. 133 (2011) 12787–12794.
[26] L. Kiwi-Minsker, I. Yuranov, V. Holler, A. Renken, Chem. Eng. Sci. 54 (1999)
4785–4790.
[27] S.R. de Miguel, J.I. Villella, E.L. Jablonski, O.A. Scelza, C.S.M. de Lecea, A.
Linares-Solano, Appl. Catal. A Gen. 232 (2002) 237–246.
[28] M.C.M. Perez, C.S.M. deLecea, A.L. Solano, Appl. Catal. A Gen. 151 (1997)
461–475.
[29] V. Gaur, A. Sharma, N. Verma, Carbon 43 (2005) 3041–3053.
[30] J. Economy, M. Daley, C. Mangun, Abstr. Pap. Am. Chem. Soc. 211 (1996) 84
(Fuel).
[31] C. Wang, J.S. Qiu, C.H. Liang, L. Xing, X.M. Yang, Catal. Commun. 9 (2008)
1749–1753.
[32] N. Mahata, A.F. Cunha, J.J.M. Orfao, J.L. Figueiredo, Appl. Catal. A Gen. 351
(2008) 204–209.
[33] N. Mahata, A.F. Cunha, J.J.M. Orfao, J.L. Figueiredo, Catal. Commun. 10 (2009)
1203–1206.
The best catalytic performance was observed for 5 wt.%
Ni/ACFHNO3-373, which is ca. 20-fold more active than the bench-
mark Raney Ni catalyst in the hydrogenation of p-CNB. Thus, the
new catalyst reported here represents a significant step forward
towards a simple, heterogeneous catalytic selective hydrogenation
of nitroarenes that employs H2 as the hydrogen source.
[34] V. Holler, D. Wegricht, I. Yuranov, L. Kiwi-Minsker, Chem. Eng. Sci. Technol. 23
(2000) 251–255.
[35] D. Rosenthal, M. Ruta, R. Schlogl, L. Kiwi-Minsker, Carbon 48 (2010)
1835–1843.
Acknowledgments
[36] C. Subrahmanyam, D.A. Bulushev, L. Kiwi-Minsker, Appl. Catal. B Environ. 61
(2005) 98–106.
[37] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Orfao, Carbon 37 (1999)
1379–1389.
[38] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[39] J.W. Shim, S.J. Park, S.K. Ryu, Carbon 39 (2001) 1635–1642.
[40] G.B. Baur, I. Yuranov, L. Kiwi-Minsker, Catal. Today 249 (2015) 252–258.
[41] W. Brockner, C. Ehrhardt, M. Gjikaj, Thermochim. Acta 456 (2007) 64–68.
[42] H.S. Johnston, L. Foering, Y.S. Tao, G.H. Messerly, J. Am. Chem. Soc. 73 (1951)
2319–2321.
Danièle Laub and Colette Vallotton of CIME are acknowledged
for the samples preparation for STEM analysis. The financial support
from the Swiss National Science Foundation (Grant 200020-
149869) and the Russian Science Foundation (Grant 15-19-20023)
is highly appreciated.
References
[43] M.C. Biesinger, B.P. Payne, L.W.M. Lau, A. Gerson, R.S.C. Smart, Surf. Interface
Anal. 41 (2009) 324–332.
[1] H.U. Blaser, H. Steiner, M. Studer, ChemCatChem 1 (2009) 210–221.
[2] R.S. Downing, P.J. Kunkeler, H. vanBekkum, Catal. Today 37 (1997) 121–136.
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020

G Model
CATTOD-10168; No. of Pages8
ARTICLE IN PRESS
8
O. Beswick et al. / Catalysis Today xxx (2016) xxx–xxx
[44] G.C. Torres, E.L. Jablonski, G.T. Baronetti, A.A. Castro, S.R. deMiguel, O.A.
Scelza, M.D. Blanco, M.A.P. Jimenez, J.L.G. Fierro, Appl. Catal. A Gen. 161
(1997) 213–226.
[46] D.E. Ramaker, J. de Graaf, J.A.R. van Veen, D.C. Koningsberger, J. Catal. 203
(2001) 7–17.
[47] B. Coq, A. Tijani, R. Dutartre, F. Figueras, J. Mol. Catal. 79 (1993) 253–264.
[45] M.L. Toebes, Y.H. Zhang, J. Hajek, T.A. Nijhuis, J.H. Bitter, A.J. van Dillen, D.Y.
Murzin, D.C. Koningsberger, K.P. de Jong, J. Catal. 226 (2004) 215–225.
Please cite this article in press as: O. Beswick, et al., Ni-based structured catalyst for selective 3-phase hydrogenation of nitroaromatics,
Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.04.020