Performance of carbon nanofiber and activated carbon supported nickel catalysts for liquid-phase hydrogenation of cinnamaldehyde into hydrocinnamaldehyde

Article abstract of DOI:10.1007/s10562-013-1146-8

The herringbone and platelet carbon nanofibers and the activated carbon were used as the supports of nickel catalyst for the selective hydrogenation of cinnamaldehyde to hydrocinnamaldehyde (HALD). The hydrogenation of HALD and cinnamyl alcohol under the same experimental conditions was also performed to determine the susceptibility of the isolated and conjugated C=O and C=C bonds to hydrogenation. The catalytic activity and selectivity of the nickel supported on carbons of different structure were evaluated based on the determination of the rate constants of the hydrogenation reaction. The nickel particle size effect was found to be crucial for the catalyst performance.

Full text of DOI:10.1007/s10562-013-1146-8

Catal Lett (2014) 144:62–69
DOI 10.1007/s10562-013-1146-8
Performance of Carbon Nanofiber and Activated Carbon
Supported Nickel Catalysts for Liquid-Phase Hydrogenation
of Cinnamaldehyde into Hydrocinnamaldehyde
´
Stanisław Gryglewicz Agata Sliwak
•
•
´
Joanna Cwikła Grazyna Gryglewicz
•
_
Received: 7 August 2013 / Accepted: 14 October 2013 / Published online: 5 November 2013
Ó Springer Science+Business Media New York 2013
Abstract The herringbone and platelet carbon nanofibers
and the activated carbon were used as the supports of nickel
catalyst for the selective hydrogenation of cinnamaldehyde
to hydrocinnamaldehyde (HALD). The hydrogenation of
HALD and cinnamyl alcohol under the same experimental
conditions was also performed to determine the suscepti-
bility of the isolated and conjugated C=O and C=C bonds to
hydrogenation. The catalytic activity and selectivity of the
nickel supported on carbons of different structure were
evaluated based on the determination of the rate constants of
the hydrogenation reaction. The nickel particle size effect
was found to be crucial for the catalyst performance.
favored [1]. The partially hydrogenated intermediates of
CALD hydrogenation are the most desirable chemicals in the
market. The selective hydrogenation of the carbonyl group is
preferred because the unsaturated alcohols are of high
importance for the production of perfumers and pharma-
ceuticals. However, HALD is also valuable, for example, as
an intermediate in the preparation of pharmaceuticals used in
the treatment of HIV [2]. Generally, the most important task
is to achieve the high selectivity of the hydrogenation reac-
tion to partially hydrogenated CALD derivatives, CALC or
HALD, for high conversion level.
Direct hydrogenation processes based on heterogeneous
catalysis are being developed to produce unsaturated
alcohols and hydrogenated aldehydes. Porous carbon
materials such as activated carbons (ACs), carbon nano-
tubes (CNTs) and carbon nanofibers (CNFs) have been
mainly used as support of mono- and bi-metallic Pt, Ru and
Ir catalysts for the selective hydrogenation of CALD to
CALC [3–7]. Some studies addressed the metal particle
size effect on the hydrogenation of CALD in terms of the
selectivity towards CALC. It was shown for the Pt and Ru-
based catalysts that the selectivity to unsaturated alcohol
increases with increasing metal particle size [8–10].
The carbon support has an advantage over mineral one
because organic compounds, especially aromatics, show
the affinity for hydrophobic carbon surface in contrast to
the mineral support. CNFs seem to be very promising
carbon support due to their mesoporous nature and devel-
oped surface area, reaching up to 300 m2/g, and the open
edges of graphene layers, which constitute sites that are
readily available for chemical and physical interactions
[11, 12]. It was reported the superiority of the Pd/CNF
catalyst compared to the high surface area activated carbon
supported Pd for the hydrogenation of CALD in liquid-
phase reactions [13]. A 90 % selectivity towards HALD
Keywords Heterogeneous catalysis ꢀ XRD ꢀ
Selective hydrogenation ꢀ Nickel catalyst ꢀ Carbon
nanofibers ꢀ Cinnamaldehyde hydrogenation ꢀ
Kinetics ꢀ Particle size effect
1 Introduction
Hydrogenation of a,b-unsaturated aldehydes such acrolein,
crotonaldehyde, citral and cinnamaldehyde (CALD) pro-
vides the products which can be considered as precious
starting materials for synthesis of fine chemicals. The final
product of CALD hydrogenation is hydrocinnamyl alcohol
(HALC, 3-phenyl propanol). The process can proceed via the
formation of hydrocinnamaldehyde (HALD) or cinnamyl
alcohol (CALC). The first route is thermodynamically more
´
´
S. Gryglewicz ꢀ A. Sliwak ꢀ J. Cwikła ꢀ G. Gryglewicz (&)
Department of Polymer and Carbonaceous Materials, Faculty of
´
Chemistry, Wrocław University of Technology, Gdanska 7/9,
50-344 Wrocław, Poland
e-mail: grazyna.gryglewicz@pwr.wroc.pl
123

Performance of Carbon Nanofiber
63
was achieved over Pd/CNF. The higher activity of Pd/CNF
compared to activated carbon supported palladium catalyst
can be related to the reduced mass transfer limitation due to
the absence of microporosity in the CNF support. Ruthe-
nium supported on CNFs showed also higher selectivity
towards HALD (70–90 %) than that obtained with Ru on
activated carbon (30–40 %) [9]. Recently, nitrogen-doped
CNTs have been revealed as promising support for Pd in
the liquid-phase hydrogenation of CALD. Nitrogen incor-
poration in the CNT structure improved not only the cat-
alytic activity of Pd/N-CNTs but also the selectivity
towards the C=C bond hydrogenation [14, 15].
ULTRA CAT (AC) was supplied by Norit Company. The
gases used in this work, hydrogen (99.999 %), nitrogen
(99.999 %), methane (99.995 %) and propane (99.95 %)
were provided by Messer.
2.2 CNF Synthesis
Herringbone-type CNFs (hCNF) were synthesized by
decomposition of methane at 550 °C over Ni/Al2O3 with a
yield of 5.3 gCNF/gcat. Platelet-type CNFs (pCNF) were
obtained by catalytic decomposition of propane at 500 °C
using the same catalyst. The yield of pCNF was slightly
lower, 4.9 gCNF/gcat. For each synthesis, 0.2 g of catalyst
was spread in the bottom of a quartz boat which was placed
into a horizontal type furnace. After catalyst reduction
under hydrogen flow, a mixture of carbon precursor and
hydrogen (1:1 v/v) was introduced into the reactor for 1 h.
The as-received CNFs were purified with hydrofluoric acid
at room temperature for 1 h in sonication bath to remove
the nickel and alumina. Finally, the samples were thor-
oughly washed with hot distilled water (until the pH
reached 7) and then dried overnight at 120 °C.
Since the nickel promotes the hydrogenation of C=C
bond rather than C=O bond when both double bonds are
present in the molecule of reactant, HALD is preferentially
formed during hydrogenation of CALD. Chin et al. [16]
obtained a high value of conversion (74 %) and HALD
selectivity (80 %) during vapour phase hydrogenation of
CALD at 300 °C over Ni/c-Al2O3 catalyst. A significant
improvement of the catalytic activity and selectivity
towards HALD was obtained in the liquid phase hydro-
genation of CALD over Pt–Ni/CNTs catalyst reduced by
KBH4 due to the electronic synergetic effect of Pt–Ni–B
[17]. In our work, the catalytic CALD hydrogenation was
carried out in the presence of CNF and activated carbon
supported nickel catalysts in a liquid phase medium. Until
know little work has been carried out on the hydrogenation
of a,b-unsaturated aldehydes over CNF supported nickel
catalysts [18]. To our knowledge, the catalytic performance
of the nickel supported on different types of CNF for the
hydrogenation of CALD has been not reported yet. This is
important in terms of searching for a highly active and
selective catalyst towards HALD.
2.3 Catalysts Preparation
hCNF, pCNF and AC were used as the support of Ni cat-
alyst. CNF and AC supported Ni catalysts (Ni/hCNF, Ni/
pCNF, Ni/AC) were prepared by incipient wetness method.
Prior to impregnation the support was wetted with ethanol
to overcome hydrophobicity of CNF and subsequently was
impregnated with an appropriate amount of aqueous nickel
nitrate solution to yield a Ni loading of 10 wt% in the
catalyst. AC before Ni loading was demineralized with HCl
and HF to remove the mineral constituents. Another AC-
based Ni catalyst was prepared by spray-drying method
(Ni/AC-SD) [19]. An aqueous suspension of AC and nickel
nitrate was mixed for 1 h to receive a homogeneous mix-
ture and subsequently subjected to a Bu¨chi mini spray
dryer 190 with a 0.5 mm nozzle at an inlet temperature of
230 °C. After drying, the Ni-loaded CNFs and ACs were
heat treated at 400 °C for 3 h under argon atmosphere.
ASA analysis showed that the Ni content in all prepared
catalysts is comparable, between 9.8 and 10.1 wt%.
In this paper, the results of the study on the hydroge-
nation of the C=C bond and the C=O group in CALD over
carbon nanofiber and activated carbon supported nickel
catalysts have been presented. The susceptibility of the
isolated and conjugated of C=C and C=O bonds to
hydrogenation was evaluated based on the constant rates of
hydrogenation reactions. The effects of the structure of
carbon support, the nickel particle size and the catalyst
method preparation on the catalytic activity and selectivity
of the prepared catalysts in the hydrogenation of CALD
were determined.
2.4 Catalyst Characterization
2 Materials and Methods
The porous structure parameters of the supports and cata-
lysts were determined from sorption of N2 at 77 K with a
Nova 2200 gas sorption analyzer (Quantachrome). Prior to
the measurements, the sample had been outgassed over-
night at 300 °C. The specific surface area was calculated
from the BET equation. The Gurvitch rule and Kelvin
condensation theory [20] were applied to establish the
2.1 Materials
Cinnamaldehyde, HALD, HALC, CALC, nickel nitrate
Ni(NO3)2ꢀ6H2O, and alumina were purchased from Aldrich
(Steinheim, Germany). The activated carbon Norit SX
123

64
S. Gryglewicz et al.
extent of microporosity and the mesopore volume distri-
bution, respectively. The amount of nitrogen adsorbed at
the relative pressure of p p-0 1 = 0.96 was applied to
determine the total pore volume VT which corresponds to
the sum of the micropore Vmic (\2 nm) and mesopore Vmes
(2–50 nm). The mesopore fraction is expressed as the ratio
of mesopore volume to the total pore volume (Vmes/VT).
The structural arrangement of graphene layers in the CNF
and the dispersion of NiO nanoparticles on the carbon
surface support was investigated by high resolution trans-
mission electron microscopy (HRTEM) performed on FEI
Tecnai G2 20X-TWIN microscope, operating at an accel-
eration voltage of 200 kV. The sample was prepared by
ultrasonic dispersion in ethanol and a drop of the resultant
suspension was placed on a holey carbon support grid. The
X-ray diffraction (XRD) was applied to determine the size
of NiO crystallites, using a Rigaku Ultima IV X-ray dif-
fractometer and CuKa radiation.
The analysis of reaction products was performed using
an Agilent GC 6850 gas chromatograph equipped with a
FID detector and a HP-1 column (30 m 9 0.25 mm 9
0.25 lm film thickness). The injector temperature was
250 °C, split 5, and a constant pressure of 100 kPa. The
temperature of GC oven was kept constant at 100 °C. The
identification of compounds in the reaction products was
done by comparison of retention time of standard. Two
independent chromatographic analyses were performed for
each sample.
3 Results and Discussion
3.1 Carbon Supported Nickel Catalysts
The catalysts containing 10 wt% of Ni were prepared using
two types of CNFs, i.e., platelet and herringbone, and the
activated carbon as the supports. Both types of CNFs are
constituted of graphene layers with the exposed edges but
different orientation within the structure. The platelet CNFs
have graphene layers orientated perpendicular to the fiber
axis (Fig. 1a). In the herringbone CNFs they are stacked at
an angle to the fiber axis (Fig. 1b). The different orienta-
tion of graphene layers determines the catalytic behavior of
metal particles supported on the CNF. Park and Baker [21]
documented well how strong influence has a change in
orientation of the graphene layers within the nanofiber
structure on the selectivity of the supported nickel particles
for hydrogenation of alkenes and dienes. In our work, the
performance of nickel supported on the traditional carbon
material such activated carbon has been also studied to
compare with that of supported on the CNF.
2.5 Hydrogenation
The hydrogenation reactions were carried out in a micro-
autoclave under hydrogen pressure of 3 MPa at 160 °C.
20 mL of substrate solution (1 % in toluene) and 50 mg of
catalyst were placed into the reactor for each experiment.
The reaction mixture was kept homogeneous with a mag-
netic stirrer at stirring frequency of 120 min-1. Before
starting the reaction, the catalyst was reduced at 360 °C for
1 h under a 3 MPa hydrogen pressure. Afterwards the
reactor was cooled to the temperature required for the
process, and the reaction substrate was injected into the
reactor. The reactor was equipped with the sampling line of
reagents. Toluene was used as a solvent because cinnamyl
alcohol is poorly dissolved in aliphatic hydrocarbons. It
should be added that the catalytic hydrogenation of toluene
was not observed in the applied process conditions.
Table 1 shows the textural parameters of the supports
and the prepared nickel catalysts. The CNFs are meso-
porous in their nature. The mesopore ratio for the
Fig. 1 TEM images of
as-grown platelet-type
CNFs (a) and herringbone-type
CNFs (b)
a
b
10 nm
10 nm
123

Performance of Carbon Nanofiber
65
Table 1 Characteristics of porous texture of the supports and the catalysts, and the size of NiO particles determined by XRD
Support/catalyst SBET (m2/g) VT (cm3/g) Vmes (cm3/g) Vmic (cm3/g) Mesopore volume distribution (cm3/g) Average NiO size (nm)
2–5 nm
5–10 nm
10–50 nm
hCNF
150
127
251
229
1020
959
682
0.417
0.337
0.486
0.367
0.809
0.613
0.460
0.323
0.282
0.386
0.277
0.437
0.254
0.201
0.094
0.055
0.100
0.090
0.372
0.359
0.259
0.036
0.036
0.060
0.053
0.117
0.090
0.076
0.062
0.060
0.115
0.064
0.076
0.058
0.480
0.225
0.186
0.211
0.160
0.244
0.106
0.077
Ni/hCNF
pCNF
5.5
4.7
Ni/pCNF
AC
Ni/AC
5.3
4.2
Ni/AC-SD
synthesized CNFs is almost 0.8. Comparing the porous
texture of both CNFs samples, pCNFs are characterized by
better developed porosity. The BET surface area of pCNFs
is high, 250 m2/g, whereas 150 m2/g is obtained for
hCNFs. As expected the porosity development of the cat-
alysts is lower compared to respective support due to pore
blockage by NiO particles.
conjugated with the carbonyl group, and aromatic ring.
However, the aromatic ring does not undergo hydrogena-
tion in the presence of Ni catalyst under mild process
conditions. Therefore, our work is focused on the hydro-
genation of C=C and C=O bonds in CALD (Fig. 4). It has
been assumed that the conversion of CALD to HALC
proceeds directly without desorption of intermediate
product, i.e. HALD, from the catalyst surface. This implies
that the hydrogenation of CALD would proceed on the
active sites through three parallel transformation pathways.
Further, k3 = 0 since CALC was not detected in the
reaction products. k2 corresponds to the rate constant kCO for
the C=O hydrogenation and the sum of k1 ? k2 is attributed
to the rate constant k= for the C=C hydrogenation.
However, the mesoporous character of the CNFs is
preserved after Ni loading. AC shows a much lower con-
tribution of mesopores to the total pore volume than CNFs
(0.54), and its Ni loading leads to the predominance of
micropores in the porous texture of the obtained catalysts. It
is interesting to note that the use of spray-drying for metal
loading results in a decrease of mesopore volume only,
whereas incipient wetness impregnation reduces signifi-
cantly both the micropores and mesopores volumes. In
contrast to CNF supported Ni catalysts, AC-based catalysts
are distinguished by very high surface area (Table 1).
Figure 2 shows HRTEM images of NiO particle distri-
bution in the prepared catalysts. It can be clearly seen that
the Ni/AC-SD catalyst is characterized by the largest dis-
persion of NiO nanoparticles on the surface.
The kinetics of hydrogenation reaction can be described
by the pseudo-first order kinetics model. The rate of hy-
drogenolysis of the C=C bond or the C=O group is a
function of one substrate only due to the hydrogen excess
(3 MPa) used under the conditions examined in our work.
This implies that the concentration of hydrogen is nearly
constant during hydrogenation reaction. Therefore, the
reaction rate can be expressed by a classical equation:
XRD examination was applied to determine the average
size of NiO crystallites in the prepared catalysts. The dif-
fractogramms are shown in Fig. 3 and the NiO particle size
determined based on XRD is given in Table 1. As can be
observed the average NiO size increases in the direction:
Ni/AC-SD \ Ni/pCNF \ Ni/AC \ Ni/hCNF which con-
firms that the Ni/AC-SD catalyst has the smallest size of
NiO particles amongst the studied catalysts. Comparing
CNF-based catalysts, the size of NiO supported on pCNFs
is considerable smaller than that on hCNFs (4.7 vs.
5.3 nm).
C0
ln
¼ k ꢁ t
C
where C denotes the concentration of a given reactant, C0
is the initial concentration and k is the rate constant. The
plot of ln C as a function of time gives the linear rela-
0
C
tionship where the slope of the line corresponds to the
reaction rate constant.
Figure 5 shows the catalytic activity of different carbon
supported Ni catalysts as a function of reaction time for the
CALD hydrogenation. Prior to the hydrogenation runs, we
performed optimization of stirring frequency to assure that
the reactions are carried out within the intrinsic kinetic
region. The catalysts contain the same amount of Ni but
differ in the support which determines the porous texture of
the catalyst and the size of NiO particles (Table 1).
It can be seen from Fig. 5 that the CALD hydrogenation
over carbon supported Ni catalyst leads mainly to HALD,
3.2 Hydrogenation of CALD Over Carbon-Supported
Ni Catalysts
Cinnamaldehyde has three potential hydrogenation sites in
its molecule: the carbonyl group C=O, the C=C bond
123

66
S. Gryglewicz et al.
Fig. 2 TEM images of NiO
distribution for different carbon
supports, Ni/AC-SD (a), Ni/AC
(b), Ni/hCNF (c) and
a
b
Ni/pCNF (d)
50 nm
50 nm
d
c
50 nm
50 nm
Fig. 3 XRD spectra of carbon-
supported nickel catalysts. Ni/
hCNF (a), Ni/pCNF (b), Ni/AC
(c) and Ni/AC-SD (d). White
circle, carbon; black square,
NiO
o
a
o
b
o
o
o
o
10 20 30 40 50 60 70 80 90
2 theta
10 20 30 40 50 60 70 80 90
2 theta
c
d
o
o
o
o
10 20 30 40 50 60 70 80 90
2 theta
10 20 30 40 50 60 70 80 90
2 theta
indicating the preferential hydrogenation of C=C bond. The
calculated values of rate constants for the catalytic
hydrogenation of CALD are presented in Table 2. The
correlation coefficient (R2) of the linear plots for the
determination of k= and kCO were in the range of
0.996–0.998 and 0.983–0.998, respectively. The high
123

Performance of Carbon Nanofiber
67
Generally, an increase in the catalytic activity is followed
by a decrease in the selectivity to HALD. The Ni/pCNF
catalyst is the most promising amongst all studied catalysts,
taking into account its highest catalytic activity and a rel-
atively high selectivity for the hydrogenation of C=C bond.
Amongst platelet, ribbon and spiral-like CNFs, the best
performance of the nickel dispersed on the platelet CNFs
was also reported for the partial hydrogenation of 1,3-
butadiene to 1-butene [21]. The changes in the catalytic
activity and selectivity of the nickel supported on the CNFs
of different structure were explained in terms of the vari-
ations in crystallographic orientations of the metal particles
associated with the structure of the support. It is believed
that orientation of graphene layers determines the faces of a
metal particle which are required to catalyze certain reac-
tions [22].
O
HALD
k1
OH
k2
O
HALC
CALD
k3
OH
CALC
Fig. 4 Possible reaction pathways for the hydrogenation of
cinnamaldehyde
values of R2 demonstrate that the hydrogenation of CALD
under experimental conditions used in the work can be
described by a pseudo first-order rate equation for two
parallel reactions.
Ni/AC-SD shows only slightly worse catalytic activity
than Ni/pCNF, however its selectivity is poor. It seems that
the high reaction rate for the hydrogenation of C=C bond
As expected the hydrogenation of double bond over Ni
catalyst is much more favorite than carbonyl group. The
rate for the C=C hydrogenation compared to the C=O
hydrogenation is 18–44-fold higher, depending on the
examined catalyst. The ratio k=/kCO is considered as a
measure of the catalyst selectivity toward the formation of
HALD (Table 2). The k=/kCO value is the largest for the
Ni/AC catalyst, indicating the highest selectivity for the
C=C hydrogenation, however, at the lowest reaction rate.
Table 2 Reaction rate constants for hydrogenation of CALD over
different carbon supported Ni catalysts (160 °C, pH = 3 MPa) min-1
2
Catalyst
k=
kCO
k=/kCO
Ni/hCNF
Ni/pCNF
Ni/AC
0.0261
0.0383
0.0174
0.0341
0.0008
0.0015
0.0004
0.0019
32.6
25.5
43.5
17.9
Ni/AC-SD
Fig. 5 Cinnamaldehyde
100
100
Ni/hCNF
Ni/pCNF
conversion and product
distribution as a function of
reaction time at 160 °C and
under 3 MPa over different
carbon supported nickel
catalysts. Black circle, CALD;
white circle, HALD; cross,
HALC
80
60
40
20
0
80
60
40
20
0
0
30
Time, min
60
90
0
30
60
Time, min
90
100
80
60
40
20
0
100
80
60
40
20
0
Ni/AC
Ni/AC-SD
0
30
Time, min
60
90
0
30
Time, min
60
90
123

68
S. Gryglewicz et al.
obtained for Ni/AC-SD is attributed to the smallest NiO
particles in size which makes it very active. The very fine
NiO distribution in the Ni/AC-SD catalyst was obtained
due to the introduction of the nickel on the carbon surface
by spray drying method which does not lead to the
blockage of the porosity. The Ni/AC-SD catalyst has only
slightly lower BET surface area compared to the initial AC
(960 vs. 1020 m2/g). The same catalyst (Ni/AC) but pre-
pared by classical incipient wetness impregnation shows
the worst catalytic activity probably due to the larger NiO
particles. This is in line with the finding reported by Giroir-
Fendler et al. [8]. In case of large particles, the C=C bond
remains far from the metal surface due to steric hindrance
related to the phenyl group. However, there is no steric
constraint for approach and adsorption of the C=C bond on
the surface of smaller particles because the phenyl group
lies aside the metal surface. The lower catalytic activity of
Ni/hCNFs than that of Ni/pCNF can be also partially
attributed to the difference in the NiO crystallite size
(Table 1).
'
k=
OH
OH
Fig. 7 Hydrogenation pathway of cinnamyl alcohol
CALC, compared to Ni/hCNF and Ni/AC, could be
attributed to the size of Ni particles. Ni/pCNF and Ni/AC-
SD are characterized by distincly smaller NiO particles (4.7
and 4.2 nm, respectively) in size that those of Ni/hCNF and
Ni/AC (5.5 and 5.3 nm, respectively). Moreover, the small
size of NiO particles ensures better dispersion of the metal,
resulting in the higher concentration of the active sites on
the carbon surface. The obtained results indicate that the
size of nickel crystallite are crucial for the hydrogenation
of C=O bond in CALD over different carbon supported
catalysts in the conditions applied in this work. The lower
size of metal particles the smaller steric hindrances to reach
the active site by the reagent molecule.
The comparable values of the rate constants for the
hydrogenation of C=O group in CALD (Table 2) to those
obtained for HALD (Table 3) indicate that the hydroge-
nation rate of C=O group is not affected by the presence of
conjugated unsaturated bond. On the contrary, the conju-
gation of the C=C bond with C=O group enhances the C=C
The conjugation of the C=C bond with C=O bond in the
CALD molecule has an influence on their reactivity, in
particular in case of contribution of the ionic reactions.
Such bonds combination determines the molecular struc-
ture, affecting the configuration and rigidity of the mole-
cule. This suggests that the hydrogenation reaction of C=C
bond and C=O bond should proceed with different rate
depending on the system, conjugated (CALD) or isolated
(HALD and CALC).
0
hydrogenation over all catalysts to a high extent (k= vs. k= ,
Tables 2, 3), i.e., even up to fourfold, as observed for Ni/
hCNF. Presumably, the impact of the adjacent C=O group
on the hydrogenation rate of the C=C bond is related to the
enhancement of the polarity of CALD molecule. This leads
to the higher affinity of the substrate to the support surface
that increases the probability of the reaction occurrence.
The nickel supported on the hCNFs shows the lower
catalytic activity compared to the Ni/pCNF and Ni/AC-SD.
One of the reasons of this phenomenon could be the larger
size of supported NiO particles. The lowest catalytic
activity among studied catalysts for the hydrogenation of
CALD, HALD and CALC was obtained for the Ni/AC
catalyst prepared by incipient wetness method. While the
nickel dispersed on the same support by spray drying
exhibits the superior behavior. The difference between both
catalysts (Ni/AC vs. Ni/AC-SD) lies in different metal
3.3 Hydrogenation of HALD and Cinnamyl Alcohol
Over Carbon Supported Ni Catalysts
HALD and CALC were subjected to the hydrogenation in
the same conditions as applied for CALD, i.e., at 160 °C
under a hydrogen pressure of 3 MPa for 30, 60 and 90 min,
to determine the reaction rate constants for different cata-
lysts. The hydrogenation of C=O bond in HALD and C=C
group in CALC leads to the formation of HALC (Figs. 6,
7).
The reaction rate constants for the hydrogenation of
HALD and CALC are given in Table 3. The R2 of the
0
0
linear plots for the determination of kCO and k= were in
the range of 0.951–0.990 and 0.966–0.999, respectively.
The considerable higher catalytic activity of Ni/pCNF and
Ni/AC-SD for the hydrogenation of both HALD and
Table 3 Reaction rate constants for hydrogenation of HALD and
CALC over different carbon supported Ni catalysts (160 °C,
pH = 3 MPa) min-1
2
0
0
Catalyst
HALD kCO
CALC k=
'
kCO
Ni/hCNF
Ni/pCNF
Ni/AC
0.0008
0.0016
0.0004
0.0022
0.0062
0.0193
0.0067
0.0160
OH
O
Ni/AC-SD
Fig. 6 Hydrogenation pathway of hydrocinnamaldehyde
123

Performance of Carbon Nanofiber
69
References
particle size (5.3 vs. 4.2 nm) and textural parameters (SBET
682 vs. 959 m2/g, VT 0.460 vs. 0.613 cm3/g). Based on the
comparison of Ni/AC-SD and Ni/pCNF in terms of
porosity development (Table 1) and the reaction rate con-
stants (Table 2) it can be concluded that the nickel particle
size has a crucial influence on their catalytic activity. Both
catalysts differ in their porosity to a high extent, whereas
the metal particle size is not so much different.
1. Gallezot P, Richard D (1998) Catal Rev Sci Eng 40:81
2. Muller A, Bowers J, WO Patent application 99/08989
3. Planeix JM, Coustel N, Coq B, Brotons V, Kumbhar PS, Dutartre
R, Geneste P, Bernier P, Ajayan PM (1994) J Am Chem Soc
116:7935
4. Toebes ML, Zhang Y, Hajek J, Nujhuis TA, Bitter JH, van Dillen
AJ, Murzin DYu, Koningsberger DC, de Jong KP (2004) J Catal
226:215
5. Toebes ML, Nujhuis TA, Bitter JH, Hajek J, Bitter JH, van Dillen
AJ, Murzin DYu, de Jong KP (2005) Chem Eng Sci 60:5682
6. Vu H, Goncalves F, Philippe R, Lamouroux E, Corrias M, Kihn
Y, Plee D, Kalck P, Serp P (2006) J Catal 240:18
7. Jung A, Jess A, Schubert T, Schu¨tz W (2009) Appl Catal A
362:95
8. Giroir-Fendler A, Richard D, Gallezot P (1990) Catal Lett 5:175
9. Galvagno S, Capanelli G, Neri G, Donato A, Pietropaolo R
(1991) J Mol Catal 64:237
10. Plomp AJ, Vuori H, Krause AOI, Jong KP, Bitter JH (2008) Appl
Catal A 351:9
11. Serp P, Corrias M, Kalck P (2003) Appl Catal A 253:337
12. Chinthaginjala JK, Seshan K, Lefferts L (2007) Ind Eng Chem
Res 46:3968
13. Pham-Huu C, Keller N, Ehret G, Charbonniere LJ, Ziessel R,
Ledoux MJ (2001) J Mol Catal A 170:155
14. Amadou J, Chizari K, Houlle M, Janowska I, Ersen O, Begin D,
Pham-Huu C (2008) Catal Today 138:62
15. Chizari K, Janowska I, Houlle M, Florea I, Ersen O, Romero T,
Bernhardt P, Ledoux MJ, Pham-Huu C (2010) Appl Catal A
380:72
16. Chin S-Y, Lin F-J, Ko A-N (2009) Catal Lett 132:389
17. Li Y, Ge CH, Zhao J, Zhou RX (2008) Catal Lett 126:280
18. Salman F, Park C, Baker RTK (1999) Catal Today 53:385
19. Gryglewicz G, Stolarski M, Gryglewicz S, Klijanienko A, Pie-
chocki W, Hoste S, Van Driessche I, Carleer R, Yperman J
(2006) Chemosphere 62:135
4 Conclusions
The hydrogenation of CALD and its partially hydrogenated
products HALD and CALC over carbon supported nickel
catalyst was found to be strongly dependent on the metal
particle size. Regardless of the type of carbon support, the
smaller nickel particles, the better performance of the
catalyst in the liquid-phase hydrogenation of CALD to
HALD at 160 °C under hydrogen pressure of 3 MPa. For a
given category of carbon support, i.e. CNFs or ACs, the
higher surface area enhances the dispersion of the nickel
particles. It was found that Ni/pCNF and Ni/AC-SD cata-
lysts show the highest catalytic activity in the hydrogena-
tion reactions. The comparison of the rate constants for
CALD and HALD hydrogenation under the same experi-
mental conditions revealed that the presence of the C=O
group increases strongly the hydrogen addition to the C=C
bond. While the hydrogenation of the C=O group is not
affected by the presence of conjugated C=C bond as was
found comparing the constant rates for the hydrogenation
of CALD and CALC.
20. Gregg SJ, Sing KSW (1982) Adsorption, surface area and
porosity. Academic Press, London
21. Park C, Baker RTK (1998) J Phys Chem B 102:5168
22. Park C, Baker RTK (1999) J Phys Chem B 103:2453
Acknowledgments This work was financed by a statutory activity
subsidy from the Polish Ministry of Science and Higher Education for
the Faculty of Chemistry of Wrocław University of Technology.
123