i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Selective hydrogenation of 4-carboxybenzaldehyde

over palladium catalysts supported with different
structural organization

Nosrat Izadi a,*, Ali Morad Rashidi a, Ahad Izadi b, Saeed Emami c,
Vahid Samimi d, Heshmatollah Varmazyar d
a
Nano Technology Center, Research Institute of Petroleum Industry, Tehran, Iran
b
Department of Chemical and Petroleum Engineering, Shahid Bahonar University of Technology, Kerman, Iran
c
Shahid Tondgooyan Petrochemical Company, Mahshahr Bandar, Iran
d
Department of Instrumental Analysis, Research Institute of Petroleum Industry, Tehran, Iran

article info abstract

Article history: The palladium-incorporated catalyst, (0.5% Pd/C) (nanospace engineering KOH-activated
Received 23 August 2016 carbon material) catalyst with macro-structured carbon nanofibers aggregates and
Received in revised form micro-nano-porous activated carbons, has been studied as a candidate for hydro-
13 November 2016 purification of crude terephthalic acid containing of 4- carboxybenzaldehyde (4-CBA) as
Accepted 29 November 2016 an impurity. The efficiency of different carbon structures (macro/micro) was investigated
Available online 6 February 2017 over selectivity catalyst. The reaction products were analyzed by HPLC to determine the
amounts of 4-CBA, benzoic acid (BA), 4-Hydroxymethyl benzoic acid (4-HMBA) and para-
Keywords: toluic acid (p-tol). It has been confirmed that 0.5% Pd/microporous catalyst gave an
KOH activation excellent performance to catalyze the hydrogenation of 4-CBA to para-toluic acid due to
Carbon nanofiber high micro-porous surface area and the most desirable selectivity to 4-HMBA was obtained
Pd-catalyst with macro-structured carbon nanofibers. Pd/AC catalyst with more micro surface area
4-CBA (85.26%) achieved a maximum yield to the intermediate product of 4-HMBA only 30%
4-HMBA whereas the macro-structured CNFs achieved typically 66%. The desired selectivity to para-
Para-toluic acid toluic acid was found to be deep dependent on the micro-porous structure.
© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

homogeneous catalyst Co(OAc)2, HBr, and a co catalyst

Introduction Mn(OAc)2. The obtained crude terephthalic acid (CTA) usually
contains approximately 2000e3000 ppm of 4-
Purified terephthalic acid (PTA) is the important industrial carboxybenzaldehyde (4-CBA) as the main impurity, which
chemical [1], which is used as a raw material for producing would decrease the polymerization rate during the production
saturated polyester e mainly polyethylene terephthalate as of PET, PBT and PTT [4]. The crude terephthalic acid (CTA)
PET [2]. In the process of company, i.e. Shahid Tondgooyan produced from PX oxidation contains a by-product, 4-
Petrochemical Co [3], theterephthalic acid (TA) production is carboxybenzaldehyde (4-CBA), which causes discoloration of
based on the liquid-phase oxidation of p-xylene (PX) using a

* Corresponding author.
E-mail address: Izadin@ripi.ir (N. Izadi).
http://dx.doi.org/10.1016/j.ijhydene.2016.11.189
0360-3199/© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2971

the CTA. The main upgrading step in refining purified ter-

Table 2 e The derived experimental parameters for
ephthalic acid (PTA) is the catalytic hydrogenation of 4-CBA to
optimal modified activation carbons during KOH agent
4-hydroxymethylbenzoic acid (4-HMBA) that is subsequently egroup 2.
converted to p-toluic acid. Simultaneously, 4-CBA is decar-
Sample PH/char KOH/PH-AC Activation
bonylated to benzoic acid (BA) producing an equimolar
ID ratio (w/w) ratio (w/w) temperature (
C)
amount of carbon monoxide, which is a well-known poison
KOH-PH-AC1 0.90 2.50 700
for palladium hydrogenation catalysts (shown in Eq. (1)) [4,5].
KOH-PH-AC2 0.90 2.50 800
KOH-PH-AC3 0.90 2.50 900
KOH-PH-AC4 0.90 3.50 700
KOH-PH-AC5 0.90 3.50 800
KOH-PH-AC6 0.90 3.50 900
KOH-PH-AC7 1.20 2.50 700
KOH-PH-AC8 1.90 2.50 700
(1) KOH-PH-AC9 2.50 2.50 700

The unfavorable hydrogenation reaction, however, can

occur over Pd catalysts [6e9]. The catalyst selectivity toward
each hydrogenation path is affected by many factors,
including the interaction between active phase and support,
the cluster size of active sites, and the type of support. Acti-
vated carbon supported palladium is the catalyst commonly
used for 4-CBA hydrogenation [10]. However, palladium par-
ticles on the carbon carrier have a sintering problem which
results in catalyst deactivation. It has been reported that
bimetallic PdeRu catalysts could increase the sintering resis-
tance of Pd metal particles [11]. Recently, Jei Zhu et al. [12]
prepared Palladium over carbon nanofiber coated monolith
for purification of CTA. They investigated the effect of mass
transfer on the catalyst performance. In this study, we
investigated the efficiency of different carbon structures
(macro/micro) over selectivity of Pd-catalyst in hydrogenation
of 4-CBA. The present work compares the performance and
characterization of Pd/KOH-modified carbon nanofiber as
macro-pore structures and Pd/KOH-activated carbon as
micro-pore structure for selectivity of catalysts. The majority
of this paper has been focused on the KOH-treatment of car-
bon nanofiber optimization, factors affecting activation car-
bon production and performance of catalyst. Carbon
nanofibers (CNFs) are cylindrical nanostructures with gra-
phene layers arranged as stacked cones, cups or plates.
Chemical vapor deposition is the dominant commercial
technique for the fabrication of CNFs. Thermal catalytic
decomposition (TCD) of methane for simultaneous production
of COx-free hydrogen and carbon nano-fibers over NieCoeAl,
NieFe/Al2O3 and NiePd/Al2O3 catalysts has been studied in
many researchers [13e16]. In previous paper [17], we synthe-
sized carbon nanofibers with average 25 nm diameters by
decomposition of methane over NieMgO catalyst. Thus in this
paper, the role of KOH as an agent was analyzed in the macro-
high surface area evolution and the properties of the perfor-
mance of catalysts. Seong-Ho Yoon [18] reported that the in-
crease of surface area in KOH activation of carbon nanofiber is
first ascribed to local broadening of graghene interstices or
local burn-off of graghenes, which limit the surface area in-
crease of CNF with the fiber structure maintained. J Romanos
et al. [19] in their paper demonstrate that nanospace engi-
neering of KOH activated carbon is possible by controlling the
degree of carbon consumption and metallic potassium inter-
calation into the carbon lattice during the activation process.
High specific surface areas, porosities, sub-nanometer (<1 nm)
and supra-nanometer (1e5 nm) pore volumes are quantita-
tively controlled by a combination of KOH concentration and
activation temperature. Controlling the pore network is

Table 1 e The derived experimental parameters for optimal activation carbons during phosphoric acid egroup 1.
Sample ID PH/char ratio (w/w) Carbonization time (h) Carbonization Activation
temperature (
C) temperature (
C)
PH-C1 0.75 12 150 550
PH-C2 0.90 12 150 550
PH-C3 1.05 12 150 550
PH-C4 1.20 12 150 550
PH-C5 0.90 6 150 550
PH-C6 0.90 12 150 550
PH-C7 0.90 24 150 550
PH-C8 0.90 48 150 550
PH-C9 0.90 24 50 550
PH-C10 0.90 24 100 550
PH-C11 0.90 24 150 550
PH-C12 0.90 24 200 550
PH-C13 0.90 24 150 500
PH-C14 0.90 24 150 550
PH-C15 0.90 24 150 600
PH-C16 0.90 24 150 650
2972 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3
crucial to optimizing activated carbon in numerous other
applications as well [20e25].
Experimental
Chemicals
In all catalyst samples, transition metal salts including re-
agent grade nickel nitrate, magnesium nitrate, palladium
chloride, sodium bicarbonate, potassium hydroxide and the
analytical grade of chemical solvents such as hydrochloric
acid 37% and phosphoric acid 85% were obtained from Merck
sources.
Preparation and activation carbon nanofibers
Carbon nanofibers with the average diameter of 25 nm were
synthesized by decomposition of methane over NieMgO
catalyst. The preparation of CNF has been described in details
in our previous publication [17]. In summary, Ni catalyst
powder (20 mg) was placed in a 1.5 m vertical two pass-fixed
bed tubular reactor. The catalyst was reduced in a 5% (v/v)
H2/N2 mixture at 850
C for 1 h, followed by CNFs synthesis in
methane ðPCH4 ¼ 0:33 atmÞ at temperature of 570
C. When the
reaction was completed, the CNFs products were let to cool
down under a nitrogen atmosphere. Activation of resulting
CNFs with KOH solution was performed at selected tempera-
tures and KOH concentration for 1 h under nitrogen flow of
100 ml/min. The ratio of KOH and CNFs was fixed in 4:1 (g/g).
The resulting material was washed with DI water (PH ¼ 7) and
dried at 100
C to form the final product.
Initial activated carbon and textural modification technique
by chemical agents
Active carbon samples were prepared in a multi step activa-
tion process from walnut shell. The PH-C samples (Group 1)
were synthesized during phosphoric acid activation. Walnut
Fig. 1 e The adsorption isotherms of N2 on CNFs and modified structures at different temperatures e (700
C, 800
C, and
900
C).
shell was soaked with phosphoric acid, followed by different
successive textural modification treatments leading to opti-
mization of factors affecting activation carbon production.
The ratios of phosphoric acid and carbon, time and tempera-
ture carbonization and activation temperature parameters
were optimized in carbon samples. KOH activated carbons
(KOH-AC) of group 2 were prepared by KOH chemical agent.
The char produced from group 1 was mixed with KOH pallets,
resulting in certain weight ratio of the agent (IR ¼ wKOH/wchar).
After dehydrating in an oven at 100
C for 2 h, the mixture was
held in a ceramic boat and the activation was achieved in a
tubular electric furnace on a nitrogen flow rate of 200 ml/min.
The heat treatment composed of a first heating ramp from the
ambient temperature to the final heat treatment at the heat-
ing rate of 5
C min1 followed by a 1 h before cooling back to
the initial temperature. Then the sample was rinsed with hot
distilled water until neutrality of the solution was obtained.
The obtained material was dried for 24 h at 120
C under air to
remove water. The experimental parameters were derived for
the given assumptions, resulting in 4 conditions that were
categorized into two groups in Tables 1 and 2.
Pd deposition
0.5 wt% Pd catalysts were prepared by the adsorption and
depositioneprecipitation methods for CNFs and activated
carbon structures, respectively. Specially, pre treatment CNFs
and KOH activated nanofibers agglomerates with the average
size of 80 mm were dispersed in a HCl solution (0.2 M) of
palladium (II) chloride followed by drying in a rotary vacuum
and reduced in a 20% (v/v) H2/N2 mixture at 250
C for 2 h. The
AC carriers with the same size of carbon nanofibers were
preliminarily purified from dust by boiling in distilled water
and dried at 120
C. Aqueous solutions of Na2CO3 and H2PdCl4
in the molar ratio of 2:1 were sprayed into carriers and dried
under vacuum at 100
C. Subsequent operation of reducing
was carried out in a tubular reactor with a flow of hydrogen
and nitrogen (20% v/v) at the temperature of 250
C for 2 h [26].
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2973

Table 3 e Properties of CNFs and modified structures with KOH agent at temperatures of (700
C, 800
C, and 900
C).
Samples Stotal (m2/g) Vmicro (cm3/g) Vtotal (cm3/g) Fmicro (%) Smicro (m2/g) Smicro (%) Da (nm)
CNF-pure 177.5 0.013 0.780 2.240 31.54 17.76 17.58
KOH-CNF-700 618.3 0.028 0.930 3.011 68.60 11.09 6.02
KOH-CNF-800 714.1 0.030 0.950 3.158 62.30 8.72 5.32
KOH-CNF-900 1242.6 0.0001 2.92 0.0034 9.74 0.780 9.40
a
D ¼ 4*Vtotal/Stotal.

Fig. 2 e The adsorptive properties of activated carbon during phosphoric acid: (a) the efficacy of (a) concentration of
phosphoric acid to carbon (PH/C ratio); (b) carbonization time; (c) carbonization temperature and (d) activation temperature.

Fig. 3 e The adsorptive properties of activated carbon during KOH treatment: the efficacy of (a) Activation temperature for
concentration of KOH/PH-AC ¼ 2.5 (b) Activation temperature for concentration of KOH/PH-AC ¼ 3.5 (c) concentration of
phosphoric acid to char (PH/char ratio) for concentration of KOH/PH-AC ¼ 2.5 and T ¼ 700
C.
2974 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3
Apparatus
Studies of the characterization and the morphological
appearance and the crystalline structure of samples were
presented by ASAP, XRD, SEM, TEM and ICP-AES analyses.
Accelerated surface area and porosimetry analysis was car-
ried out with sub critical nitrogen at 77.3 K on a Micromeritics
American Instrument (model 2010), which is calibrated by
using a surface area reference material Al2O3. Before the ASAP
measurement, the samples were heated at 30
C and 200
C for
30 and 180 min (heating rate 10
C/min) under N2 gas flow of
40 ml/min in order to remove any adsorbed species. After the
samples were cooled down to room temperature under N2
flowing. Specific of surface area were determined from sub
critical nitrogen isotherms using BrunauereEmmetteTeller
(BET) theory in the range of P/P0 (0.06e0.2). The total pore
volume was determined at P/P0 ¼ 0.995. The t-plot method
was applied to calculate the micropore volume. The X-ray
diffraction (XRD) pattern of samples were recorded on a PW-
Fig. 4 e SEM micrographs for (a) CNF-pure (b) KOH-CNF-800
(c) KOH-CNF-900.
1840 X-ray diffract meter (Philips), using Cu-Ka (1.5
A) radia-
tion operated at 40 kV and 30 mA. The SEM images were ob-
tained with a Cambridge S-360 operated at 16 kV and 2.5 mA.
The TEM photographs were obtained with a LEO-912-AB
operated at 85 kV. The Pd loading was determined by (ICP-
AES) using a PerkinElmer Optima 3000DV instrument.
Results and discussions of activated carbons
Chemical activation of carbon structures and adsorptive
properties
Among the chemical activation agents currently tested in the
literature, KOH is known to be the most efficient but its in-
dustrial use is severely restricted by induced liquid effluent
treatments and large amount of chemicals needed [27e29]. It
is well known that preparation of carbon materials by KOH
activation generally leads to a very high surface area of even
more than 2000 m2/g. Such a surface area is ascribed to partial
gasification and expansion of the interlayer spacing between
graghenes through simultaneous intercalation and dein-
tercalation [19]. For each activating agent, various activating
agent to carbon ratio and activation temperature, acknowl-
edged to be the most important activation parameters. In this
study, the weight ratio of KOH to CNFs was kept constant in
4:1 (g/g), followed by activation at 700e900
C for 1 h. The ef-
fect on BET surface area and pore size distribution was
measured. The adsorption isotherms of N2 on all samples
have been presented in Fig. 1. The isotherms of CNFs activated
with KOH (Fig. 1) exhibited hysteresis loops, suggesting
macro-porous features. The total surface area increased with
KOH activation and temperature. The results of the BET spe-
cific surface area have been presented in Table 3. At 800
C,
KOH activation allowed the tenability of the surface area up to
714 m2 g1. Ultimately, surface area was increased by a factor
of 4 when comparing activated nano fibers to the raw CNFs.
CNFs activated at 900
C with a high surface area
(1242.6 m2 g1) exhibited a high macro-pore structure which
was 7 times higher than as prepared CNFs. From the results
presented in Table 3, modification of CNFs produced a high
macro-porous structure. Apparent micro-pore area calcu-
lated, based on N2 adsorptionedesorption data, to be insig-
nificant, and the average pore size diameter for CNFs-pure
and KOH-CNF-800 obtained 17.58 and 5.32 nm (Table 3), which
inferred the macro-porous structure of the material. Activated
carbon from walnut shell has been successfully produced
through chemical activation process using phosphoric acid
Table 4 e Reaction conditions for 4-CBA hydrogenation.
Reaction condition Value


Reaction temperature, T ( C) 250
Total pressure, P (bar) 50
Agitation, A (rpm) 400
Mass of H2O, mH2 O (g) 130
Mass of CTA, mCTA (g) 120
Mass of the catalyst, mcat (g) 1.30
Initial concentration of 4-CBA, C04-CBA (mmol) 1.83
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2975

and KOH as activating agents. In the present study, the acti-
vated carbons were prepared using chemical activation
method by varying the preparation variables. The variables
studied were (i) activation temperature (ii) activation time (iii)
carbonization time (iv) impregnation ratio of activating agent
to carbon on BET surface area and volume microporosity. For
phosphoric acid activation, the results of the adsorptive
properties have been presented in Fig. 2. From Fig. 2(a),
Phosphoric acid:char impregnation ratio played a decisive role
in the formation of total high surface area. At high PH
impregnation ratio (1.2 g/g), the high BET development was
mostly due to the intercalation of phosphoric acid in the
carbon structure but the highest total BET internal surface
area (Smicro) obtained by using 0.9 (g/g) ratio of phosphoric acid
to char was 71.71% (Fig. 2(a)). Hence, the optimum tempera-
tures of carbonization and activation for production of acti-
vated carbon based on highest total BET internal surface area
were obtained at 150
C and 550
C, respectively and optimal
carbonization time was selected at 24 h (sample PH-C14)
(Fig. 2(b)e(d)). Further activation of the resulting carbon (PH-
C14) with KOH solution was performed at selected tempera-
ture and KOH concentration for 1 h. In this study, the efficacy
of preliminary phosphoric acid concentration to char was
studied also. The results of the volume microporosity and BET
specific surface area were presented in Fig. 3. Fig. 3(a)e(b)
shows the effects of activation temperature on surface area
and porosity of carbons at different KOH concentration. It can
be observed that the KOH increased the surface of carbons
more than other agents (almost 2.2 times the activated
phosphoric acid area). However, KOH activation doubles the
sub-nanometer pore volume and generates a very high BET
internal surface area [19]. From Fig. 3, the highest BET internal
surface area of the activated carbon obtained by KOH: PH-AC
impregnation ratio of 2.5 at 700
C that was 85.42% (KOH-PH-
AC1). The lowest BET internal surface area (69.88%) was ob-
tained at KOH-PH-AC8 sample. From this, it can be concluded
that the KOH/PH-AC molar ratio of 2.5 displays agent for
improved micropore surface area at a reaction temperature of
700
C. The morphology of deposited carbon nanofibers in
purified and activated samples were studied by SEM analysis
and presented in Fig. 4. In Fig. 4(a) and (b), CNFs as prepared
and activated with KOH at 800
C maintained the fiber form
and it was difficult to find any change of overall diameter and
structure of CNFs before (Fig. 4(a)) and after KOH activation
(Fig. 4(b)) under SEM. However, the fibrous shape of the KOH-
CNF-900 sample with a high surface area of 1242.6 m2 g1 was
holly collapsed to become other type bulk structure, although
a minority of fibers remained to be aggregated as shown in
Fig. 4(c).
Selection to a series of carbon structures for Pd
deposition
Macro-porous structures including pre-treatment CNFs (CNF-
pure, SBET ¼ 177.5 m2/g, %Smicro ¼ 17.76) and KOH activated
nanofibers (KOH-CNF-800, SBET ¼ 714.1 m2/g, %Smicro ¼ 8.72)
and micro-porous structures activated carbons including (PH-
C14, SBET ¼ 1329 m2/g, %Smicro ¼ 71.71, KOH-PH-AC1,
SBET ¼ 1706 m2/g, %Smicro ¼ 85.42, KOH-PH-AC8,
SBET ¼ 2107 m2/g, %Smicro ¼ 69.88) were selected for Pd
deposition.
Performance of Pd/C catalysts on 4-CBA
hydrogenation
The catalytic performance of Pd/C catalysts ( macro and micro
structures) was evaluated for hydro-purification of crude ter-
ephthalic acid (CTA) to convert 4-CBA to 4-HMBA, BA and p-tol
and the results were compared with each other. The experi-
ments were conducted in a 1000 ml pressurized stirred batch
autoclave made of stainless steel (supplied by Parr In-
struments Co type 4848). The detailed reaction conditions
were summarized in Table 4. In a typical run, 120 g of
unpurified CTA, 1.30 g of Pd/C catalyst and 280 g of de-ionized
water were mixed together and charged into the reactor. The
agitator speed was set at 400 rpm and then the autoclave was
flushed with nitrogen before pressurizing with hydrogen, in
order to remove dissolved oxygen and the reaction mixture
was then heated to the desired temperature (250
C). At the
reaction temperature, a mixture of hydrogen and nitrogen
ðPH2 =PN2 ¼ 0:5Þ was introduced into the reactor at a desired
pressure. The total pressure was kept constant at 735 psig
during reaction. Samples (1 ml) were withdrawn from the
reactor at different time during reaction, filtered immediately
to separate the catalyst particles and analyzed (solid and
liquid) by high-performance liquid chromatography (HPLC)

Table 5 e The adsorptive properties of Pd/carbon structures.

Samples SBET (m2/g) Vmicro (cm3/g) Vtotal (cm3/g) Fmicro (%) Smicro (m2/g) Smicro (%) Da (nm)
CNF-pure 177.5 0.013 0.780 2.240 31.54 17.76 17.58
Pd/CNF-pure 170.0 0.011 0.740 1.960 29.56 17.38 17.41
KOH-CNF-800 714.1 0.030 0.950 3.157 62.30 8.72 5.32
Pd/KOH-CNF-800 710.3 0.026 0.900 2.880 60.29 8.48 5.07
PH-C14 1329.0 0.470 0.670 70.15 953.00 71.71 2.02
Pd/PH-C14 1320.0 0.410 0.610 67.21 940.30 71.23 1.85
KOH-PH-AC1 1706.0 0.580 0.700 82.85 1457.28 85.42 1.64
Pd/KOH-PH-AC1 1701.0 0.560 0.680 82.35 1450.28 85.26 1.60
KOH-PH-AC8 2107.0 0.670 0.780 67.00 1472.31 69.88 1.48
Pd/KOH-PH-AC8 2102.0 0.660 0.720 68.75 1468.11 69.84 1.37
a
D ¼ 4*Vtotal/Stotal.
2976 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3

equipment. HPLC was performed with a Waters model 600E

instrument equipped with 150 mm long C18 column and a UV
detector. 4-CBA conversion (X), the selectivity toward the 4-
HMBA (S4-HMBA), BA (SBA) and p-tol (Sp-tol) were defined as
Eqs. (2)e(5).

C04CBA  C4CBA
X¼  100 (2)
C04CBA

C4HMBA
S4HMBA ¼  100 (3)
C04CBA  C4CBA

CBA
SBA ¼  100 (4)
C04CBA  C4CBA

Cptol
Sptol ¼  100 (5)
C04CBA  C4CBA

where C4-CBA, C4-HMBA, CBA and Cp-tol are the instant molar
concentration of 4-CBA, 4-HMBA, BA and p-tol during the re-
action; C04CBA is the initial molar concentration of 4-CBA in
CTA (Table 4).

Results and discussions of Pd/C catalysts

Catalyst characterization

The total BET surface area, micro-pore area and average pore
size diameter were summarized in Table 5. The adsorbed
analysis revealed that there was a slight reduction of the total
surface area and micro pore volume for the catalyst materials
in comparison with those of not loaded structures. In all
samples, there were no significant changes in surface area on
loading. For example, compared with CNF-pure, the BET sur-
face area of Pd/CNF-pure decreased with Pd deposition, where
it resulted in 4.22% and 6.28% reduction for the total and micro
BET surface area, respectively. However, The BET results ob-
tained in this investigation are in good agreement with those
reported in the literature [12]. According to Jie zhu' work, it is
well known that After Pd loading, the catalyst Pd/CNF/TiO2
held the similar textural properties as CNF/TiO2 did which
indicated that as-prepared CNF/TiO2 possessed a high mate-
rial stability [12]. The XRD patterns of the solid carbons before
and after Pd treatment have been shown in Fig. 5. The width of
XRD patterns of graphitic sheets carbon (002, 101) located at 2q
(26
C, 44
C). Quantitative data obtained from further treat-
ment of the powder XRD patterns, as shown in Table 6. The

Table 6 e Quantitative data obtained from of the powder

XRD patterns, TEM and ICP analysis for carbon structures.
Sample d-value (d002) Pd particle size Pd loading
from XRD (nm) from TEM (wt%) from
ICP
CNF-pure/Pd- 3.46/3.47 ~25 0.46
KOH-CNF-800/Pd- 3.48/3.49 ~16 0.48
KOH-CNF-900 4.04/- e e
Fig. 5 e XRD patterns of (a) CNFs pre treatment (b) Pd/CNFs PH-C14/Pd- 3.73/4.05 ~10 0.5
catalysts (c) Activated carbon pre treatment (d) Pd/activated KOH-PH-AC1/Pd- 3.93/4.59 ~5 0.5
carbon catalyst. KOH-PH-AC8/Pd- 4.08/4.77 ~2 0.5
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3
Fig. 6 e TEM images for (a) Pd/CNF-pure (b) Pd/KOH-CNF-
800 (c) Pd/PH-C14 (d) Pd/KOH-PH-AC8 (e) Pd/KOH-PH-AC1.
2977
results indicated that the CNFs materials (CNF-pure, KOH-
CNF-800) were composed of both crystalline and amorphous
carbon. The former interlayer spacing d (002) of the deposited
carbon went from approximately 3.46
A in sample CNF-pure
to 3.47
A in sample KOH-CNF-800 and 4.05
A in sample
KOH-CNF-900. Since the interlayer distance of graphite carbon
is 3.354
A, these d (002) values indicated that the crystalline
structure contains some imperfections specify in KOH-CNF-
900 sample. In carbon nanofibers with higher surface area
than 400 m2/g, the d002 increased as the surface area increased
[22]. Fig. 4(b) shows the XRD pattern of CNFs with decoration
of Pd nanoparticle. As shown in Fig. 4(b) X-ray diffraction
analysis of Pd/CNFs catalysts revealed the graphitized nature
of carbon in the CNT support that was confirmed by the
presence of C (002) and C (101) reflections at 2q close to 26 and
44, respectively. Deposition of palladium species on CNF
support and reduction of catalyst precursors led to the for-
mation of Pd0 phases that could not be detected by means of
the XRD technique in 0.5 wt% loading. The broadening of
diffraction peaks was too large to identify Pd in the Pd/CNF
catalysts. From the data in Table 6, adding of Palladium spe-
cies on CNFs, the former interlayer spacing d (002) of the
carbons was shifted to the larger amount, compared with that
of CNFs before treatment. The same effect was found with the
samples as named PH-C14, KOH-PH-AC1 and KOH-PH-AC8. In
macro and micro structures the d-value increased by a factor
of 0.1 and 0.7, respectively, when comparing before loading.
The size and dispersion of palladium particles were obtained
from TEM images (Fig. 6) and summarized in Table 6. The Pd
size nanoparticle increased from 2 nm on the Pd/KOH-PH-AC8
to 20 nm on Pd/CNF-pure with decreasing BET surface area of
the support. From Fig. 6, the larger size of palladium nano-
particles resulted in the smaller surface area. Based on Pd/AC
catalysts, Pd/PH-C14 with a total surface area of 1320 m2/g,
that is about 0.7 times the total surface area of Pd/KOH-PH-
AC8 and Pd/KOH-PH-AC, contributing to increasing particle
size and decreasing dispersion. Furthermore, Pd was distrib-
uted through the KOH-PH-AC8 support, as observed with TEM
in Fig. 6(d). This study is in good agreement with those re-
ported in the literature [12]. A typical ICP-AES results for all
catalysts presented in Table 6. The Pd loading was signifi-
cantly smaller on the Pd/CNF-pure (0.46 wt%, Table 6),
contributing to increasing particle size and decreasing
dispersion [12].
Catalytic performance of Pd on the carbon structures as
support
To investigate the influence of catalyst media on hydrogena-
tion of 4-CBA as main impurity in CTA and products at oper-
ating conditions in the high pressure reactor, it was
experimentally checked that main products contained ben-
zoic acid (BA), para-toluic acid (p-tol) and intermediate prod-
uct 4-HMBA. The conversion of 4-CBA and product yields
versus reaction time and the values as product selectivity
against the 4-CBA conversion on Pd/CNF-pure, Pd/KOH-CNF-
800, Pd/PH-C14, Pd/KOH-PH-AC1 and Pd/KOH-PH-AC8 cata-
lysts has been presented in Figs. 7 and 8. Considering for
macro-structured carbons (Pd/CNF-pure and Pd/KOH-CNF-
800), the conversion of 4-CBA was completed in around 150
2978 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3

Fig. 7 e The 4-CBA conversion and product yields as a function time over (a) Pd/CNF-pure (b) Pd/KOH-CNF-800 (c) Pd/PH-C14
(d) Pd/KOH-PH-AC8 (e) Pd/KOH-PH-AC1.

and 165 min, respectively. For Pd/CNF-pure catalyst
(Smicro ¼ 17.76%), at completed reaction, the total yield prod-
ucts obtained 31.52% p-tol, 20.68% BA and the 46.29% inter-
mediate product of 4-HMBA. On the main product (99%
conversion) was almost the hydrogenated 4-CBA to 4-HMBA.
Using Pd/KOH-CNF-800 catalyst (Smicro ¼ 8.72%), the p-tol yield
decreased, whatever the more total surface area for catalyst
preparation used, down to around 60%. However, the inter-
mediate product 4-HMBA was dominant finally (52.46%). In
fact, by increasing the total surface area, the amount of
metallic palladium dispersed into the carbon matrix
increased. KOH activation of carbon nanofibers resulted with
a reduced microporous region which was not suitable for p-tol
selectivity application. However, it was observed that for
carbon nanofibers catalysts exhibited much higher selectivity
to 4-HMBA over 47e53%. From this, it can be concluded that
macro-structured carbons displays a good catalytic activity for
improved quality and high yield 4-HMBA at a reaction tem-
perature of 250
C. Also, the evolution of the products in the
activity test for micro-structured carbons with different sur-
face area (Pd/PH-C14, Pd/KOH-PH-AC1 and Pd/KOH-PH-AC8)
has been shown in Figs. 7 and 8. The results on p-tol yield in
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2979

Fig. 8 e The product selectivity as a function 4-CBA conversion over (a) Pd/CNF-pure (b) Pd/KOH-CNF-800 (c) Pd/PH-C14 (d)
Pd/KOH-PH-AC8 (e) Pd/KOH-PH-AC1.

Fig. 7(d) showed that the Pd/KOH-PH-AC1 catalyst prepared by
higher internal surface area (85.26%) had a higher p-tol con-
centration over 76.98% than the other ones. For this catalyst,
the conversion of 4-CBA completed in 186 min and the main
products (76.98% and 19.82%) were the fully hydrogenated p-
tol and BA, respectively. The inter-mediate product 4-HMBA
was dominant initially, but was decreasing down to 3.20%
when approaching full conversion (Fig. 7(d)). For both of cat-
alysts (Pd/KOH-PH-AC8 and Pd/PH-C14) with almost the same
internal surface area (69.84% & 71.71%), p-tol yield was not
almost kept constant along the test. However, for the Pd/PH-
KOH-AC8 catalyst prepared by higher total surface area
(2102 m2/g) after an catalytic activity producing a p-tol yield
(54.20%) about 175 min. Compared with Pd/KOH-PH-AC1, Pd/
PH-C14 and Pd/KOH-PH-AC8 catalysts, Pd/KOH-PH-AC8 (Fig. 8)
exhibited a lower observed p-tol selectivity for the two sam-
ples with identical Pd amounts; which was probably due to the
lower diffusion phenomena on Pd/KOH-PH-AC8 catalyst.
While the time for full conversion of 4-CBA of catalyst Pd/
KOH-PH-AC8 was estimated at about 175 min. Thus
2980 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3

Table 7 e Comparing of the products selectivity and maximum yields of 4-HMBA for all catalysts.
Samples Max yield of 4-CBA conversion at p-tol selectivity (%) at BA selectivity (%) at 4-HMBA selectivity (%) at
4-HMBA (%) max 4-HMBA yield (%) completed reaction completed reaction completed reaction
Pd/CNF-pure 55.22 89.67 32.00 21.00 47.00
Pd/KOH-CNF-800 66.44 90.29 12.60 34.40 53.00
Pd/PH-C14 42.89 87.71 60.90 21.70 17.40
Pd/KOH-PH-AC1 30.46 75.78 76.98 19.82 3.20
Pd/KOH-PH-AC8 50.98 88.70 54.47 22.25 23.27

Table 8 e Effect of any or significant diffusion limitations, R-square of Eqs. (10) and (11) ratio of k1 to k2 for 4-CBA
hydrogenations on Pd catalysts.
Samples Eq. (10) Eq. (11) H2 4-CBA
R-square k1/k2 ratio R-square k1/k2 ratio NWeP NWeP
Pd/CNF-pure 0.9698 4.39 0.9998 50.05 0.089 21.38
Pd/KOH-CNF-800 0.9814 9.79 0.9609 260.85 0.065 15.54
Pd/PH-C14 0.9386 2.46 0.9870 16.72 0.328 78.44
Pd/KOH-PH-AC1 0.9060 1.19 0.9754 4.03 0.453 108.09
Pd/KOH-PH-AC8 0.9562 2.90 0.9701 20.72 0.226 53.93

introduction of higher total surface area for catalyst prepara-
tion (with the lower internal surface area) allowed the p-tol
yield to be getting worse off by a factor of 1.42 regardless of
reaction time. The selectivity and maximum yields of 4-HMBA
for all catalysts were compared with each other and presented
in Table 7. However, in all cases, addition of Pd leads to an
increment in the p-toluic acid and 4-HMBA production,
resulting in p-tol yields between of 31e70% and 4-HMBA yields
between 3 and 53%.
Investigation of chemical or diffusional regime by
WeiszePrater number and calculation for the prediction of
k1/k2 ratio over Pd catalysts
WeiszePrater number (NWeP) enables to identify the nature of
operating regime, i.e., chemical or diffusional regime and is
calculated by Refs. [30,31]:
r00  R2p  t
Nwp ¼
Vpore  D12  CS
where r00 is the initial specific rate (mol s1 g1), R2p is the radius
catalyst particle and is equal to 0.004 cm, Cs is the substrate
concentration on catalyst surface (mol cm3) that for
hydrogen calculated based on solubility of hydrogen in water
in a broad temperature [32] and for 4-CBA Calculated based on
the data of mH2 O and m4-CBA in Table 4. t is the tortuosity factor
of the pores and D12 is molecular diffusivity (cm2 s1) and is
defined by Ref. [33]:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4  M2
D12 ¼ 1:1728  1016  T  (7)
h2  V10:6
where T is the reaction temperature (T); 4 is the solvent as-
sociation parameter; M2 is the molecular weight of solvent
(g mol1); h2 is the viscosity of the solvent under the reaction
temperature and pressure (Pa s); and V1 is the molar volume of
the substrate solute at normal boiling point (m3 kmol1). WeP
criterion asserted that if NWeP < 0.3, rates for all reactions with
an order of 2 or less should have negligible mass transfer
(6) limitations, while a value for NWeP > 6 indicates the definite
diffusion control [34]. Table 8 shows that the calculated
WeiszePrater numbers (NWeP) of H2 all indicate insignificant
internal diffusion limitations for all catalysts whereas the
performance of all catalysts is clearly affected by diffusion
limitation for 4-CBA. This is in line with the differences in the
initial internal surface area, discussed above. Because, the
conversion of 4-CBA to 4-HMBA, p-tol and BA explained by the
first e order serieseparallel reaction (Denbigh reaction) ac-
cording to the following sequence-parallel of three irreversible
reactions:

k1 k2 9
4  CBA!4  HMBA!p  tol >
>
>
>
>
k3 =
4  CBA!BA k0 ¼ k1 þ k3 (8)
r4CBA ¼ k0 C4CBA >
>
>
>
r4HMBA ¼ k1 C4CBA  k2 C4HMBA >
;
rBA ¼ k3 C4CBA
Fig. 9 e Reaction constant for 4-CBA hydrogenation to 4-
HMBA over Pd/CNF-pure, Pd/KOH-CNF-800, Pd/PH-C14, Pd/ For convenience, by assuming that all reactions are first
KOH-PHAC1 and Pd/KOH-PH-AC8 catalysts. order and, additionally, unaffected by a change of volume,
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2981

Fig. 10 e Fitting curve of 4-CBA conversion vs. 4-HMBA yield over a) Pd/CNF-pure, b) Pd/KOH-CNF-800, c) Pd/PH-C14, d) Pd/
KOH-PHAC1 and e) Pd/KOH-PH-AC8 catalysts with Eqs. (10) and (11).

with k1, k2 and k3 being the intrinsic rate constants. The The elicited parameter (k3/k0 ratio) for all catalysts Pd/CNF-
apparent reaction rate constant (k0 ) was calculated from the pure, Pd/KOH-CNF-800, Pd/KOH-PH-AC1, Pd/KOH-PH-AC8 and
first-order plot of 4-CBA concentration with reaction time [35] Pd/PH-C14 was obtained 0.171, 0.212, 0.138, 0.172 and 0.157
and fitted in Fig. 9. Fig. 9 shows that Pd/KOH-PH-AC1 respectively.
possessed the highest reaction rate constant (k0 ) (approxi- Two models with any transport limitation Eq. (10) and
mately 0.046 min1) whereas the Pd/KOH-CNF-800 had the significant diffusion limitation Eq. (11) have been obtained
lowest observed activity (0.024 min1). Also, the apparent re- similar in the literature [36] for 4-HMBA yield over different
action rate constant (k3) was calculated from the plot of BA catalysts and operational conditions.
yield with reaction time according to Eq. (9) [35].    
S1
Y4HMBA ¼ ð1  XÞ ð1  XÞðS2 1Þ  1 (10)
k3 1  S2
YBA ¼ expð1  k0 tÞ (9)
k0
2982 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3

 a   
Y4HMBA ¼ ð1  XÞ ð1  XÞðb1Þ  1
b þ 1
   pffiffiffiffiffi
S1 S1
a¼  S2
1  S2 1  S2
pffiffiffiffiffi (11)
b ¼ S2
S1 ¼ k1 =k0
S2 ¼ k2 =k0

Eq. (10) assumes that the bulk concentration of 4-CBA and

4-HMBA at a point in the reactor are the same as those at the
catalyst surface. With a micro-porous catalyst, Eq. (11), how-
ever, the concentrations of 4-CBA and 4-HMBA within the
catalyst are in general different from the bulk concentrations
in the flowing gas. The reactant 4CBA will diffuse into the pore
structure where it will be reacted to the unstable intermediate
4-HMBA. An unstable molecule 4-HMBA formed within a pore
must make many random collisions with the pore walls
before it escapes out of a pore mouth into the gas phase. Thus
while trying to “find its way out” of the pore structure, the
intermediate 4-HMBA has a high probability of degenerating
into the product p-tol [36].
In Eqs. (10) and (11), X and Y4-HMBA are the instant con-
version of 4-CBA and the instant yield of 4-HMBA in the re-
action, k1, k2 and k0 are the intrinsic rate constants of the
reaction of 4-CBA to 4-HMBA, the reaction of 4-HMBA to p-tol
and the reaction of 4-CBA to BA and 4-HMBA, respectively. In
this work, it has been investigated for developing an intrinsic
rate constant for 4-HMBA consumption on the surface catalyst
from proposed models. The genetic algorithm was used for
the models and then was compared with the experimental
data. If the curve predicted by the model agrees with the
experimental one, the model may reasonably describe what is
occurring physically in the real system. In Eqs. (10) and (11) the
ratio of k1 and k2 again denotes the intrinsic selectivity factor.
Eq. (10) directly gives the point selectivity observed in absence
of transport limitations and Eq. (11) directly gives the point
selectivity observed in presence of transport limitations. The
relationship between instant yield of 4-HMBA and intrinsic
rate constant k1, k2 and instant conversion of 4-CBA (X),
assuming first order consecutive reaction, by Eqs. (10) and (11)
have been presented in Fig. 10. The elicited parameter (k1/k2
ratio) and R square from two suggested models have been
summarized in Table 8. The proposed model (Eq. (11)) corre-
sponds better l with the experimental data for all catalysts
except KOH-CNF-800 and indicated that significant diffusion
limitation almost occurs for the reaction of 4-HMBA to p-tol
(i.e. Thiele modus of 4-HMBA 4HMBA > 3). Table 8 shows the
results of calculation of the intrinsic selectivity factor (k1/k2)
for all catalysts using Eqs. (10) and (11). Rather similar values
around 0.9 are obtained for R square, confirming mass transfer
limitation has a deep influence especially for micro-porous
catalysts. Application of Eq. (11) on the data of Pd/KOH-PH-
AC1 results in a much lower value. The fact that the selec-
tivity factor (k1/k2) estimated using Eq. (11) results in about
4.033, significantly lower to the intrinsic value observed for
the other catalysts, confirms the deep effect of mass transfer
limitation on the selectivity of Pd/KOH-PH-AC1. The role of
diffusion resistance is very important through the selectivity
Fig. 11 e The relationship between p-tol and 4-HMBA
selectivity and micro surface area (%) for carbon based
catalysts.
effect described above. Jie Zhu et al. [12] investigated the in-
fluence of internal diffusion on selective hydrogenation of 4-
carboxybenzaldehyde over palladium catalysts supported on
carbon nanofiber coated monolith and Pd/AC catalysts. They
presented that the Pd/AC catalysts revealed a lower activity
toward to 4-HMBA, due to internal mass transfer limitation.
Fig. 11 presents the relationship between p-tol and 4-HMBA
selectivity and micro surface area (%) for carbon based cata-
lysts. When micro surface area was less than 10%, the 4-HMBA
selectivity was dominant factor. With increasing micro sur-
face area more than 65% the catalytic activity desired to p-
toluic production.
Conclusion
The results obtained in the present work reflect the efficacy of
different carbon structures (macro/micro) over selectivity of
Pd-catalyst in hydrogenation of 4-CBA. The present work
compares the performance and characterization of Pd/KOH-
modified carbon nanofiber as macro-pore structure and Pd/
KOH-activated carbon as micro-pore structure for selectivity
of catalysts. The majority of this paper has been focused on
the KOH-treatment of carbon nanofiber optimization, factors
affecting activation carbon production and performance of
catalyst. The physicochemical characteristics of these cata-
lysts were depended on the internal surface area content. The
desired selectivity to para-toluic acid was found to be deep
dependent on the micro-porous structure. Pd/KOH-PH-AC1
catalyst with more micro surface area (85.26%) achieved a
maximum yield to the intermediate product of 4-HMBA only
30% whereas the Pd/KOH-CNF-800 achieved typically 66%. The
calculated WeiszePrater numbers (NWeP) of H2 all indicate
insignificant internal diffusion limitations for all catalysts
whereas the performance of all catalysts is clearly affected by
diffusion limitation for 4-CBA and 4-HMBA. For all catalysts,
when micro surface area was less than 20%, the 4-HMBA
selectivity was dominant factor. With increasing micro sur-
face area more than 65% the catalytic activity desired to p-
toluic production.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 2 ( 2 0 1 7 ) 2 9 7 0 e2 9 8 3 2983

references highly stable bimetallic Ni-Cu catalyst for simultaneous

production of hydrogen and fish-bone carbon nanofibers:
optimization, effect of catalyst preparation methods and
deactivation. Int J Hydrogen Energy ( 2014;39:7765e79.
[1] Chen Y, Fulton JL, Partenheimer W. The structure of the
[18] Yoon Seong-Ho, Lim Seongyop, Song Yan, Ota Yasunori,
homogeneous oxidation catalyst, Mn(II)(Br(-1))x, in
Qiao Wenming, Tanaka Atsushi, et al. KOH activation of
supercritical water: an X-ray absorption fine-structure study.
carbon nanofiber. Carbon 2004;42:1723e9.
J Am Chem Soc 2005;127:14085e93.
[19] Romanos J, Beckner M, Rash T, Firlej L, Kuchta B, Yu P, et al.
[2] Wiley Series in Polymer, Science. In: Scheirs J, Long TE,
Nano space engineering of KOH activated carbon.
editors. Modern polyesters: chemistry and technology of
Nanotechnology 2011;23.
polyesters and copolyesters. England: John Wiley and Sons
[20] Otowa T, Nojima Y, Miyazaki T. Development of KOH
Ltd; 2003.
activated high surface area carbon and its application to
[3] Shahid Tondgooyan Petrochemical Company; Mahshahr
drinking water purification. Carbon 1997;35:1315e9.
Bandar, Iran.
[21] Ludlow Douglas K. Activated carbon adsorption. J Am Chem
[4] Jhung SH, Romanenko AV, Lee KH, Park YS, Moroz EM,
Soc 2006;128. 10630e10630.
Likholobov VA. Carbon-supported palladium-ruthenium
[22] Gun’ko VM, Turov VV, Kozynchenko OP, Nikolaev VG,
catalyst for hydropurification of terephthalic acid. Appl Catal
Tennison SR, Meikle ST, et al. Activation and structural and
A General 2002;225:131e9.
adsorption features of activated carbons with highly
[5] Li KT, Hsu MH, Wang I. Palladium coreeporous silica shell-
developed micromeso- and macroporosity. Adsorption
nanoparticles for catalyzing the hydrogenation of 4-
2011;17:453e60.
carboxybenzaldehyde. Catal Commun 2008;9:2257e60.
[23] Matranga KR, Myers AL, Glandt ED. Storage of natural gas by
[6] Chou TL. Kinetics of salt formation using terphethalic acid
adsorption on activated carbon. Chem Eng Sci
and N-methyl-2-Pyrolidinone. M.Sc. Thesis. Bozeman, MT:
1992;47:1569e79.
University of Montana State; 2004.
[24] Lillo-Rodenas MA, Juan-Juan J, Carzola-Amoros D, Linares-
[7] Jhung S, Park Y. Hydropurification of crude terephthalic acid
Solano A. About reactions occurring during chemical
over PdRu/Carbon composite catalyst. J Korean Chem Soc
activation with hydroxides. Carbon 2004;42:1371e5.
2002;46:57e63.
[25] Raymundo-Pinero E, Azaı̈s P, Cacciaguerra T, Carzola-
[8] Schroeder, H.; James, D. E. Purification of crude terephthalic guin F. KOH and NaOH
Amoros D, Linares-Solano A, Be
acid. US Patent 4,892,972 A, 1990.
activation mechanisms of multiwalled carbon nanotubes with
[9] Serp P, Corrias M, Kalck P. Carbon nanotubes and nanofibers
different structural organization. Carbon 2005;43:786e95.
in catalysis. Appl Catal A 2003;253:337e58.
[26] Romanenko; Anatoly Vladimirovich (Novosibirsk, RU),
[10] Menegazzo F, Fantinel T, Signoretto M, Pinna F. Metal
Likholobov; Vladimir Alexandrovich (Novosibirsk, RU),
dispersion and distribution in Pd- based PTA catalysts. Catal
Timofeeva; Maria Nikolaevna (Novosibirsk, RU), Hva; Jhung
Commun 2007;8:876e9.
Sung (Taejeon-Si, KR), Seok; Park Jun (Taejeon-Si, KR)
[11] Romanenko AV, Tyschishin EA, Moroz EM, Likholobov VA,
“Catalytic composition, method for manufacturing thereof
Zaikovskii VI, Jhung SH, et al. Influence of ruthenium
and method for the purification of terephthalic acid” United
addition on sintering of carbon-supported palladium. Appl
States Patent 6,753,290, (2004).
Catal A General 2002;227:117e23.
[27] Marsh H, Yan DS, O Gray TM, Wennerberg A. Formation of
[12] Zhu Jie, Wu Fan, Li Mingshi, Zhu Jianjun, van Ommen Jan G,
active carbons from cokes using potassium hydroxide.
Lefferts Leon. Influence of internal diffusion on selective
Carbon 1984;22:603e11.
hydrogenation of 4-carboxybenzaldehyde over palladium
[28] Ahmadpour A, Do DD. The preparation of active carbons
catalysts supported on carbon nanofiber coated monolith.
from coal by chemical and physical activation. Carbon
Appl Catal A General 2015;498:222e9.
1996;34:471e9.
[13] Fakeeha Anis Hamza, Khan Wasim Ullah, Al-Fatesh Ahmed
[29]  denas MA, Cazorla-Amoro
Lillo-Ro  s D, Linares-Solano A.
Sadeq, Abasaeed Ahmed Elhag, Naeem Muhammad Awais.
Understanding chemical reactions between carbons and
Production of hydrogen and carbon nanofibers from
NaOH and KOH: an insight into the chemical activation
methane over NieCoeAl catalysts. Int J Hydrogen Energy
mechanism. Carbon 2003;41:267e75.
2015;40:1774e81.
[30] Weisz PB, Prater CD. Interpretation of measurements in
[14] Bayat Nima, Rezaei Mehran, Meshkani Fereshteh. Hydrogen
experimental catalysis. Adv Catal 1954;6:143e96.
and carbon nanofibers synthesis by methane decomposition
[31] Albert Vannice M. Kinetics of catalytic reactions. US:
over NiePd/Al2O3 catalyst. Int J Hydrogen Energy
Springer; 2005.
2016;41:5494e503.
[32] Pray HA, Schweickert CE, Minnich BH. Solubility of hydrogen,
[15] Bayat Nima, Rezaei Mehran, Meshkani Fereshteh. Methane
oxygen, nitrogen, and helium in water at elevated
decomposition over Niee/Al2O3 catalysts for production of
temperatures. Ind Eng Chem 1952;44:1146e51.
COx-free hydrogen and carbon nanofiber. Int J Hydrogen
[33] Satterfield CN. Mass transfer in heterogeneous catalysis.
Energy 2016;41:1574e84.
Cambridge Mass: MIT Press; 1970.
[16] Maneerung Thawatchai, Hidajat Kus, Kawi Sibudjing. Co-
[34] Poling BE, Prausnitz JM, O'Connell JP. The properties of gases
production of hydrogen and carbon nanofibers from catalytic
and liquids. 5th ed. Beijing: Chem, Ind, Press; 2002.
decomposition of methane over LaNi(1x)MxO3a
[35] levenspiel Octave. Chemical reaction engineering. 1999.
perovskite (where M ¼ Co, Fe and X ¼ 0, 0.2, 0.5, 0.8, 1). Int J
[36] Frankenburg WG, Komarewsky VI. Advances in catalysis and
Hydrogen Energy 2015;40:13399e411.
related subjects, vol. 2; 1951.
[17] Izadi Nosrat, Rashidi Ali Morad, Zeraatkar Ahmad,
Varmazyar Heshmatollah, Rashtchi Maryam. Preparation of