Journal of Energy Chemistry 23(2014)645656

Catalytic performance of cement clinker supported nickel

catalyst in glycerol dry reforming
Hua Chyn Leea , Kah Weng Siewa , Maksudur R. Khana , Sim Yee China,b ,
Jolius Gimbuna,b, Chin Kui Chenga,b
a. Faculty of Chemical & Natural Resources Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang Kuantan, Pahang Malaysia;
b. Centre of Excellence for Advanced Research in Fluid Flow, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang Kuantan, Pahang Malaysia
[ Manuscript received February 17, 2014; revised April 29, 2014 ]

Abstract
The paper reports the development of cement clinker-supported nickel (with metal loadings of 5 wt%, 10 wt%, 15 wt% and 20 wt%) catalysts
for glycerol dry (CO2 ) reforming reaction. XRF results showed that CaO constituted 62.0% of cement clinker. The physicochemical characterization of the catalysts revealed 32-folds increment of BET surface area (SBET ) with the addition of nickel metal into the cement clinker, which
was also corroborated by FESEM images. Significantly, XRD results suggested different types of Ni oxides formation with Ni loading, whilst
Ca3 SiO5 and Ca2 Al0.67 Mn0.33 FeO5 were the main crystallite species for pure cement clinker. Temperature-programmed reduction analysis
yielded three domains of H2 reduction peaks, viz. centered at approximately 750 K referred to as type-I peaks, another peaks at 820 K denoted
as type-II peaks and the highest reduction peaks, type-III recorded at above 1000 K. 20 wt% Ni was found to be the best loading with the highest
XG and H2 yield, whilst the lowest methanation activity. Syngas with lower H2 /CO ratios (0.6 to 1.5) were readily produced from glycerol dry
reforming at CO2 -to-Glycerol feed ratio (CGR) of unity. Nonetheless, carbon deposit comprised of whisker type (Cv ) and graphitic-like type
(Cc ) species were found to be in majority on 20 wt%Ni/CC catalysts.
Key words
cement clinker; dry reforming; glycerol; nickel catalyst; syngas

1. Introduction

The global energy consumption is expected to attain 820

quadrillion British thermal units (Btu) by the year 2040 from
the current 524 quadrillion Btu [1]. If the same consumption pattern continues, 80% of the world energy usage will
still depend on the fossil fuel. Unfortunately, fossil fuel combustion for transportation and electricity purposes is a major source of the anthropogenic carbon dioxide (CO2 ) emission. CO2 concentration in atmosphere has once surpassed
400 parts per million (ppm), a level deemed detrimental to climate change [2]. International agreement on limiting warming
to 2 C will only be achieved if two-third of the fossil fuel reserve remains underground by the year 2050 as reported in the
2012 IEA World Energy Outlook report [3]. Significantly, a
combination of factors such as fossil fuel depletion and growing environmental concern have largely driven a search for
more sustainable and renewable resources. In particular, for
biomass-rich nation like Malaysia, the answer seems to be ob

vious. The Malaysian government has planned to fully implement B5 mandate (5% methyl ester in diesel) nationwide
in July 2014. There were 60 biodiesel manufacturing licenses
being approved by the government anticipating total annual
biodiesel production of 6.5 MT. It is widely known that during the transesterification, 10 wt% glycerol bio-waste is obtained [47]. Therefore, an annual production of 0.6 MT of
crude glycerol is anticipated. High impurity contents in the
crude glycerol necessitate an expensive purification process
[810]. This has confined the applications of purified glycerol to pharmaceutical, cosmetic and food industries. Hence
finding alternative use of crude glycerol is significant. Conversion of glycerol (bio-waste) into biofuel or chemical products, i.e. syngas (mixture of H2 and CO), acetone, acrolein,
ethers and methanol [1114], has been touted as a promising
route to provide a total solution to the biodiesel industries, as
well as energy sustainability.
Syngas is primarily being generated from natural gas, in
particular methane (CH4 ) reforming. Nonetheless, it is more
enticing to derive syngas from glycerol biowaste instead as

Corresponding author. Tel: +60-9-5492896; Fax: +60-9-5492889; E-mail: chinkui@ump.edu.my

This work was supported by Ministry of Education, Malaysia through MTUN (No. RDU121216).

Copyright2014, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. All rights reserved.
doi: 10.1016/S2095-4956(14)60196-0

646

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

glycerol (C3 H8 O3 ) possesses more H-elements than CH4 .

Glycerol steam reforming (Equation 1) has also been extensively studied over the past decade [1517]. However, it
produces high H2 /CO ratios (typically more than 10) under
excess steam conditions, as CO has undergone water-gasshift (WGS) reaction (Equation 2) which is unfavourable for
Fischer-Tropsch (FT) synthesis [18,19]. Another drawback of
glycerol steam reforming is significant production of CO2 .
C3 H8 + 6H2 O 3CO2 + 10H2

(1)

CO + H2O CO2 + H2

(2)

To mitigate CO2 emission, dry (CO2 ) reforming has

gained popularity as an alternative to steam reforming. Dry
reforming is highly selective to CO as compared with steam
reforming, hence lower H2 /CO ratio is more suitable for FT
synthesis and methanol synthesis without syngas conditioning. Dry reforming has been extensively reported for methane,
propane, ethanol and toluene [2025], yet little works have
been dedicated to glycerol dry reforming, which are limited
to the thermodynamically study by Wang and his co-workers
[26]. The overall glycerol dry reforming concept is expressed
as in Equation (3).
C3 H8 O3 + CO2 4CO + 3H2 + H2 O

(3)

Nickel-based catalysts have widely replaced the noble

catalysts such as Rh, Pt and Pd [2729] in various reforming technologies due to their availability and lower costs. Nibased catalysts possess high activity in hydrogenation and
dehydrogenation, which means the bond cleavages between
OH, CH2 , CC and CH3 are easily accomplished [30].
Nevertheless, it is well-known that Ni-based catalysts suffered
from the deactivation attributed to coke deposition. Coke
deposition is influenced by the types of support employed.
Modification of catalyst support by addition of strong Lewis
base oxide such as CaO and MgO has improved carbon deactivation resistance during reforming [3133]. Previous studies
[34,35] reported that cement clinker (CC) contains 63.28%
and 63.17% CaO respectively, hence exhibiting properties
suitable as catalyst synthesis. CC is an intermediate product of the cement industries. Halmann and Steinfeld [36]
have performed the thermo-chemical calculations for the coproduction of limestone (CaO) and syngas via CH4 dry reforming, using CO2 released from lime-kiln. Gimbun et al.
[37] have activated CC for biodiesel production and glycerol
was the by-product. Some of the researchers also reported that
carbon nanotubes deposited on Ni catalyst could mechanicalstrengthened the cementitious material. Therefore, it is feasible to conduct reforming of glycerol (from the biodiesel industry) over the CC utilizing the emitted CO2 from the cement kiln. This will provide an utmost solution to environment, cement and biodiesel industry. The current work serves
to characterize the catalysts using spectroscopic analysis and
also to investigate the activity of glycerol dry reforming over
CC-supported Ni catalysts with the aims of producing syngas
and understanding the carbon laydown behaviour.

2. Experimental
2.1. Materials and catalyst preparation
Analytical grade glycerol was supplied by Sigma Aldrich
and the CC was obtained from the Pahang Cement Sdn. Bhd.
CC was ground with mortar and sieved for the particle range
of 100200 m before wet-impregnated with 5 wt%, 10 wt%,
15 wt% and 20 wt% nickel (Ni)-metal, respectively, using
Ni(NO3 )2 6H2 O as precursor in 50 mL ultrapure water. Subsequently, the resulting slurry was stirred for 3 h at room temperature before oven-dried for 24 h at 403 K. The catalysts
were then air-calcined at 1073 K for 6 h at 5 Kmin1. Postcalcination, the catalysts were ground and sieved again to the
size of 90200 m for reaction studies.
2.2. Catalyst characterization
N2 physisorption isotherms of the catalysts were performed at 77 K using thermo scientific surface gas adsorption porosimeter. Before physisorption, the catalysts were degassed overnight at 573 K and 1104 Torr. The cumulative
pore size was determined using Barett-Joyner-Halenda (BJH)
equation. The oxide composition of the catalysts was determined by wavelength X-ray fluorescence (WD-XRF) spectrometry (Bruker, S8 Tiger Model, Germany). The crystalline behaviour of the catalysts was measured by powder
XRD measurement (Rigaku Miniflex II) with Cu K radia at 30 kV and 15 mA, from 2 of 10o to 80o
tion, = 1.5418 A
with a step size of 0.02o and a step time of 1 s. The oxide
crystalline size of the catalysts was determined using Scherrer
equation, d = 0.94/(d cos ), where d is the crystallite size,
is the wavelength of the radiation, d is the full-width at half
maximum (FWHM) of the diffraction peak and is the half
of the diffraction angle.
The surface structure of the catalysts was captured
by field emission electron microscopy (FESEM) model
JOEL/JSM-7800F, equipped with energy microscopy dispersive spectroscopy (EDS). Qualitatively, Fourier transform infrared (FTIR) spectra of the catalysts were recorded from 600
to 4000 cm1 with a Perkin Elmer (Model Spectrum 100)
FTIR spectrophotometer using KBr disks technique. Thermogravimetric analysis (TGA) was performed to observe the
gas-solid catalyst interaction in 60 mLmin1 of air with
40 mLmin1 of N2 carrier up to 1173 K employing ramping
rate of 10 Kmin1.
Temperature-programmed reduction (TPR) was performed using Thermo Finnigan TPD/R/O 1100 series unit
equipped with TCD detector to gauge the reducibility of the
catalysts. An approximately 0.1 g catalyst (with different Ni
loadings) was reduced with 20 mLmin1 5% H2 in N2 carrier
gas at ramping 20 Kmin1 up to 1173 K with holding period
of 30 min prior to cooling to room temperature.
After the reaction, temperature-programmed oxidation
(TPO) and FESEM were also performed on the used catalysts to examine the carbon deposition phenomenon and types

647

Journal of Energy Chemistry Vol. 23 No. 5 2014

of solid carbon deposited onto the catalysts. TPO analysis

was accomplished via using 60 mLmin1 oxygen (O2 ) with
40 mLmin1 N2 carrier.


ri

mol
gcat .s


=

yi Fi
, where i = H2 , CO and CH4
W

2.3. Catalyst activity

Figure 1 shows the experimental rig for the catalysts testing which comprised of stainless steel 316 fixed-bed reactor
(ID: 9.42 mm; OD: 12.7 mm; length: 400 mm). The catalytic
evaluation was carried out over 0.20 g catalyst sandwiched
between two layers of quartz wool. The catalyst was activated and reduced by 50% H2 /N2 gas (50 mLmin1 STP)
for 2 h and held at 1073 K. Glycerol was pumped into the
reactor by high performance liquid chromatography (HPLC)
pump (LabAlliance Series 1, Max Pressure: 3000 psi and
flowrate from 0.01 to 10.0 mLmin1) while the flowrates of
reactant and carrier gas (as diluent) were controlled by Alicat Scientific electronic mass flow controller (Model: MC500SCCM-D). The performance of each catalysts was evaluated at 1023 K with carbon dioxide-to-glycerol feed ratio (CGR) of unity and constant gas-hourly space velocity
1 at STP. Reactor outlet gases
(GHSV) of 3.6104 mLg1
cat h
were passed through a cold trap for liquid product capture
and then over a drierite (CaSO4 & CoCl2 ) bed (8 mesh).
The gaseous product was collected into 1-L Tedlar gas sampling bag and the flowrate of the exit gas was measured using
bubble meter. The composition of produced syngas was determined using Agilent gas chromatography (GC) with TCD
capillary columns, HP-MOLSIV (Model No. Agilent 19095P,
30.0 m530 m50.0 m) and HP-Plot/Q column (Model
No. Agilent 19095-Q04, 30.0 m530 m40.0 m). The total amount of gas entering the sample loop was 25 L. The
glycerol conversion to H2 , CO and CH4 gaseous products was
calculated based on H-atom balance as in Equation (4). The
yields of hydrogen (YH2 ) and carbon containing species (Yi )
were calculated according to Equations (5) and (6), respectively, whilst the rate of gaseous product formation was calculated based on Equation (7).
XG =

2 FH2 + 4 FCH4
100%
8 FC3H8 O3

(4)

2 FH2
100%
8 FC3H8 O3

(5)

YH2 =
Yi =

Fi
100%, where i = CO and CH4
3 FC3 H8 O3 ,in

(6)

Figure 1. Experimental rig for glycerol dry reforming studies

3. Results and discussion

3.1. Catalyst characterization
3.1.1. N 2 physisorption, FESEM and XRF analyses
Table 1 shows that the BET specific surface area (SBET)
and the cumulative pore volume of the catalysts increased with
the increasing Ni metal loading. SBET has increased 32-fold
from 0.55 gcm3 to 17.83 gcm3 with Ni loading of 20 wt%.
The low surface area of CC was attributed to the sintering
effect from calcination at 1723 K in the cement kiln [38]. The
density of the catalysts also increased with Ni loading, viz.
from 2.99 to 3.22 gcm3, which was consistent with the result of Taylor [39].
Figure 2 depicts the N2 physisorption isotherms of the
pure cement clinker (CC) and 20 wt% Ni/CC catalysts (as
a representative). The pore size of CC fell in the range of
macroporous (Figure 2a) and shifted towards mesoporous region (Figure 2b) upon Ni-loadings, indicating the formation of
new compound attributed to the alteration of the surface structure of the catalysts. This explained the increment of SBET of
the catalyst with Ni loading. As CC was a macroporous material, Ni dopant was unlikely to cover the pores of the support.

Table 1. Properties of fresh solid catalysts

Catalysts
Pure CC
5 wt% Ni/CC
10 wt% Ni/CC
15 wt% Ni/CC
20 wt% Ni/CC
a

SBET
(m2 g1 )
0.55
7.73
15.81
17.30
17.83

Density
(gcm3 )
2.99
3.03
3.12
3.12
3.22

Pore volume
(cm3 g1 )
0
0.0022
0.0053
0.0054
0.0057

The data presented was in wt% except this label in ppm

(7)

CaO
61.98
53.05
57.02
44.11
38.66

Oxide composition from XRF analysis (wt%)

NiO
SiO2
Al2 O3
Fe2 O3
62a
17.21
3.90
3.53
18.67
11.49
2.66
2.64
14.14
8.44
3.07
3.16
27.63
9.28
1.75
2.38
34.83
6.77
1.36
2.07

MgO
0.55
0.41
0.50
0.32
0.25

648

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

Hence, more Ni loading has resulted in more NiO formation

and therefore larger SBET .
Figure 3 shows the FESEM images of the catalysts. The
surface of pure CC (Figure 3a) was very smooth and lessporous, which was in agreement with SBET . The morphology
of the surface became rugged and bulkier with more Ni doping. Table 2 shows the element composition of the catalysts,

which exhibited similar trend to XRF results summarized in

Table 1.
XRF analysis (Table 1) shows that the locally-sourced CC
was a mixture of oxide metals with CaO as the major ingredient of 61.98%, corresponding to the earlier studies [34,35].
Besides CaO, silicates (SiO2 ), aluminate (Al2 O3 ) and ferrite
(Fe2 O3 ) were the major compounds in CC. The percentage of

Figure 2. N2 physisorption isotherms of pure cement clinker (a) and 20 wt% Ni/CC catalysts (b) as a representative
Table 2. Element composition estimated from EDS analysis
Catalysts
Pure CC
5 wt% Ni/CC
10 wt% Ni/CC
15 wt% Ni/CC
20 wt% Ni/CC

C
3.35
5.48
5.91
6.88
7.76

Al
0.54
2.23
0.44
0.18
0.62

Si
10.65
3.37
5.83
2.50
3.20

Elements composition (wt%)

S
K
Ca

45.69
0.54
0.81
36.97

0.45
35.52

16.55
0.55
0.67
8.95

Fe

2.81

0.45

Ni

8.52
11.65
36.10
37.93

O
39.77
39.27
40.21
37.78
39.87

Figure 3. FESEM images of different catalysts. (a) Fresh cement clinker support, (b) 5 wt% Ni/CC catalysts, (c) 10 wt% Ni/CC catalysts, (d) 15 wt% Ni/CC
catalysts, (e) 20 wt% Ni/CC catalysts

Journal of Energy Chemistry Vol. 23 No. 5 2014

CaO and SiO2 in catalysts decreased from 61.98% to 38.66%

and 17.21% to 6.77%, respectively, as Ni loadings were increased. Indeed, the doping of Ni has increased the amount
of NiO species (originally also present in CC support), contributed to a hike from 62 ppm to 34.83%. This result was
also qualitatively captured by the subsequent XRD diffraction
patterns.
3.1.2. X-ray dif fraction analysis
Figure 4 shows the XRD patterns of both undoped and Nidoped CC catalysts. The main mineralogical phase presenting
in pure CC was alite (Ca3 SiO5 ), judging from the peaks shown
in Table 3. The average crystalline size of alite was 33.65 nm.

Figure 4. XRD patterns of different catalysts. 1Bunsenite, NiO;

2Calcium nickel catena silicate, CaNiSi2 O6 ; 3Calcium nickel
silicate,
CaNiSi4 O10 ;
4Nickel magnesium manganese oxide,
(Ni0.9 MgO0.1 )6 MnO8 ; 5Calcium aluminum iron manganese oxide,
Ca2 Al0.67 Mn0.33 FeO5 ; 6Alite, Ca3 SiO5

649

The alite species in CC had monoclinic crystals system (as

opposed to pure alite that possessed triclinic structure) due
to the presence of some impurities [40]. Besides alite, another major phase presenting in CC was calcium aluminum
iron manganese oxide (Ca2 Al0.67 Mn0.33 FeO5 ) in CC, possesing orthorhombic system with crystalline size of 25.65 nm.
Upon Ni doping onto CC, some peaks of alite, i.e. 38.79o;
45.79o and 50o disappeared, whilst some additional peaks at
approximately 2 = 37.5o, 43.3o and 62.9o were found related
to Ni oxides formations. Table 3 shows that different percentage of Ni doping resulted in the formation of different Ni oxides species. For 5 wt% Ni loading, Ni was bonded with free
Mg and MnO to form nickel magnesium manganese oxide
(Ni0.9 MgO0.1)6 MnO8 , with crystalline size of 12.84 nm. Calcium nickel catena-disilicate (Nickeldiopside, CaNi(Si2 O6 ))
and calcium nickel silicate (CaNiSi4 O10 ), were found in
monoclinic and tetragonal systems, respectively, for 10 wt%
Ni/CC catalyst. The average crystalline size for CaNi(Si2 O6 )
and CaNiSi4 O10 were 29.08 nm and 27.45 nm, respectively.
The presence of both CaNi(Si2 O6 ) and CaNiSi4 O10 were attributed to the reactions in Equations (8) and (9) during the
calcination stage [41].
Ca3 SiO5 + 5SiO2 + 3NiO 3CaNiSi2 O6

(8)

Ca3 SiO5 + 11SiO2 + 3NiO 3CaNiSi4 O10

(9)

A pure NiO was only found for the catalysts with Ni

doping more than 10 wt%. By comparing both 15 wt% and
20 wt% Ni doping, XRD peaks shifted to higher angle of
diffraction with higher Ni doping, indicating the substitution
of Ni2+ ions into the support lattice [42]. The pure NiO formation was confirmed by peaks at 37.4o, 43.3o and 62.9o for both
15 wt% and 20 wt% Ni/CC catalysts with plane reflection of
(111), (200) and (220). The mean crystalline size of NiO
in 15 wt% and 20 wt% Ni/CC catalysts were 16.67 nm and
16.53 nm, respectively, hence no significant difference.

Table 3. Average crystallite size, dcrystallite of Ni oxides calculated from XRD patterns
Oxides/Ni oxides presents
2
Average dcrystallite (nm)
Ca3 SiO5
29.45o , 30.17o , 32.21o , 32.54o , 33.92o , 34.41o , 38.79o , 41.31o , 41.63o , 45.79o ,
33.65
46.84o , 50.00o , 51.90o , 56.37o , 59.96o , 62.28o , 63.52o
Ca2 Al0.67 Mn0.33 FeO5
29.45o , 32.21o , 33.92o , 34.41o , 41.63o , 47.38o , 59.96o , 63.52o
25.65
5 wt% Ni/CC
(Ni0.9 MgO0.1 )6 MnO8
18.2o , 37.51o , 43.43o , 46.91o , 62.35o
12.84
10 wt% Ni/CC
CaNi(Si2 O6 )
29.50o , 34.09o , 34.51o , 37.54o , 41.33o , 43.48o , 47.18o , 51.97o , 54.03o , 60.01o , 62.86o
29.08
CaNiSi4 O10
29.50o , 34.51o , 41.33o , 43.48o , 47.18o , 51.97o , 54.03o , 62.86o
27.54
15 wt% Ni/CC
NiO
37.38o , 43.29o , 62.90o
16.67
20 wt% Ni/CC
NiO
37.41o , 43.36o , 62.94 wto , 75.37o
16.53
Catalysts
Pure CC

3.1.3. Fourier transform infrared spectroscopy (FTIR)

FTIR spectra of calcined samples as depicted in Figure 5 shows razor-sharp peaks around 3642 cm1 which represents the free OH stretch for each catalyst that ascended
with Ni loading. This result mirrored the previous works

[43,44], which mentioned that OH stretching at frequency of

3645 cm1 was due to calcium hydroxide (Ca(OH)2 ). Broad
peaks at 3436 to 3451 cm1 represented the hydrogen-bonded
OH band (symmetric and asymmetric stretching (v1 and v3 )
vibration of water) [45] that appeared widened and peak intensity increased with Ni loading; its increases can be attributed

650

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

to the hydrophilic-nature of Ni. Besides, the bands centered at

1634 cm1 referred to the vibrations of H2 O (v2 ) molecules
adsorbed on the surface of CC [43,46,47]. Moreover, the
bands from 1410 to 1425 cm1 can be assigned to the vibration of carbonate ion on the catalysts (v3 CO2
3 ) [48,49] or
the presence of CO2 [50] attributed to atmospheric CO2 adsorption, forming CaCO3 compound. In addition, the band
at frequencies ranging from 990 to 910 cm1 indicated the
presence of SiO4
4 in C3 SiO5 after the hydration of SiO (v3 )
functional group in CC, consistent with the previous findings
[43,51,52]. Ylmen et al. [47] described that the dissolution
of C3 S phase in clinker could also occur at the spectra range
of 800 to 970 cm1 . The band in the region for the frequency
less than 1000 cm1 can be referred to NiO arising from interatomic vibrations [53,54].

linked to the formation of Ni(NO3 )2 2H2 O after water elimination and also Ni2 O3 , respectively. Previous studies [58,60]
mentioned that anhydrous Ni(NO3 )2 cannot be thermallyobtained and remained in basic nickel nitrate intermediate,
[Ni(NO3 )2 Ni(OH)2 ]. The release of NO2 , NO, O2 and H2 O
from 2[Ni(NO3 )2 ; Ni(OH)2 ] contributed to the formation of
Ni2 O3 (Equation 10). The sharp peak in the third segment
(570620 K) corresponded to further oxide decomposition of
Ni2 O3 2NiO+1/2O2 . Beyond 620 K, no formation of physically meaningful peaks was recorded.
2[Ni(NO3 )2 Ni(OH)2 ]
2Ni2 O3 + 2NO2 + 2NO + O2 + 2H2O

(10)

Figure 6. Derivative weight profile of Nickel (II) nitrate hexahydrate at

10 Kmin1 ramping rate under air blanket

Figure 5. FTIR spectra of pure CC and Ni/CC catalysts with different Ni

loadings

3.1.4. Temperature-programmed calcination prof iles

The synthesized catalysts (dried only, non-calcined), the
CC support, and the metal precursor itself viz. nickel
(II) nitrate hexahydrate (Ni(NO3 )2 6H2 O) were subjected to
temperature-programmed calcination studies to obtain the individual calcination characteristic. Figure 6 shows the thermal profile of Ni(NO3 )2 6H2 O in air blanket. Significantly,
the weight reduction profile shows numerous inflection points
symptomatic of metal-salt decomposition/oxidation, hence
solid structural phases evolution. The decomposition of pure
nickel nitrate salt under the air blanket consistently depicted
three major peak segments (Figure 6), viz. below 420 K,
from 420 to 520 K and a sharp peak centered at 582 K.
The peaks below 420 K can be assigned to the hydration
water elimination, from Ni(NO3 )2 6H2 O to Ni(NO3 )2 4H2 O
species at 334 K and 366 K [5559]. The two peaks centered at about 465 and 504 K in the second segment may be

Significantly, Figure 7 shows that CC support had greater

thermal stability with only two decomposition peaks at 720 K
and 917 K. The peak at 720 K can be attributed to the decomposition of calcium hydroxide (portlandite, Ca(OH)2 ) as water was readily adsorbed on the surface of CC support, which
comprised mainly of CaO (Equation 11). Another possibility causing the formation of portlandite is alite (Ca3 SiO5 ) in
the support which has been hydrated (Equation 12). This
finding was validated with the FTIR results which recorded
OH stretching at 3645 cm1 (Figure 5).
CaO + H2 O Ca(OH)2

(11)

2Ca3 SiO5 + 7H2 O 3CaO 2(SiO2) 4(H2 O) + 3Ca(OH)2

(12)
The peak at 917 K of CC support in Figure 7 could be attributed to the loss of CO2 from CaCO3 contained in the support as CaO presents a promising CO2 capture agent (Equation 13) at room temperature [61]. Once CaCO3 was formed
from CaO, it required high temperature to release the CO2 .
This finding was substantiated by the FTIR results at band
ranging from 1410 to 1425 cm1 .
CaO + CO2 CaCO3

(13)

651

Journal of Energy Chemistry Vol. 23 No. 5 2014

complex hydrated calcium silicate compounds. In addition,

the small peak around 925 K can be attributed to the release
of CO2 from CaCO3 formed in the support matrix. The presence of carbonate species was consistent with the FTIR results
in Figure 5.
3.1.5. Temperature-programmed reduction (TPR) prof iles
The reducibility of Ni/CC catalyst was determined by
temperature-programmed reduction (TPR) profiling. The
TPR profiles of the fresh catalysts (with 5 wt% to 20 wt%
Ni loadings) are depicted in Figure 8 while the summaries of
H2 uptake at each peak are presented in Table 4. The TPR
curves of the metal-impregnated catalysts showed multiple
major reduction peaks indicative of different Ni-support interaction degree [63].
Figure 7. Derivative weight profiles at 10 Kmin1 ramping rate under air
blanket

The doping of Ni (5 wt% to 20 wt%) onto CC support

has resulted in a near-similar decomposition trend for calcination under air blanket (Figure 7). The peaks observed at
below 500 K was most likely attributed to the elimination of
water. Indeed, only one peak was noticeable, centered at approximately 570 K which can be linked to NiO formation at
582 K due to the total metal precursor (nickel (II) nitrate hexahydrate) decomposition (Figure 6). The calcination profiles
corroborated the XRD patterns which showed the presence
of NiO. The formation of shoulder peak at 673 K indicated
a continuous thermal decomposition of complex mixture of
hydrated silicate and aluminate-type compounds [62]. The
peaks formed at 740 K could be due to the release of hydroxide group from the hydrated calcium silicate compounds in
the catalyst. Moreover, the right-shifting of the decomposition peaks as depicted in Figure 7 may be linked to the incorporation of Ni metal into the matrix of CC, consequently endowing it with higher capacity to capture water and forming

Figure 8. Temperature-programmed reduction (TPR) profiles of freshlycalcined catalysts

Table 4. Reduction peak temperatures and H2 uptake estimated from H2-TPR profiles
Ni loading
on CC
(wt%)
0 (pure CC)
5
10
15
20

Peak I
(K)

752
745
768
749, 777

Peak reduction temperature (K) and H2 uptake (molg1

cat )
H2 uptake
Peak II
H2 uptake
Peak III
(K)
(molg1
(K)
(molg1
cat )
cat )

982; 1036
519.11

1006
630.63
804
619.89
995
648.20
839
624.47
1012
596.03
844
708.22
1031

The profiles, in general, illustrated three domains of H2

reduction peaks, viz. centered at approximately 750 K referred to as type-I peaks, another peaks at 820 K denoted
as type-II peaks and the highest reduction peaks, type-III
recorded at above 1000 K. The low temperature type-I peaks
were generally due to H2 reduction of metal particles which
indicated the exposed fraction of Ni atoms/free Ni compound
as well as reduction of Ni-metal weakly interacted with the

H2 uptake
(molg1
cat )
205.89
615.12
720.02
957.31
912.34

Total
metal surface
(m2 )
0
6.96
8.90
14.10
13.27

support (primarily comprised of CaO) [25,64,65]. This was

also echoed by Richardson et al. [66]. Through their investigation, they have found that the reduction peaks occurred
between 635 K to 746 K with CaO as the additive to NiO catalyst. In addition, Choudhary et al. [67] also found via their
TPR studies that peak maximum temperatures for NiO-CaO
catalysts were 763 K and 1073 K (small hump). Overall, it
seemed that the H2 reduction peaks shifted to higher reduc-

652

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

tion temperature with the increasing Ni loading. This was

probably due to the stronger interaction between nickel and
matrices of CC support.
The reduction peaks from 995 to 1036 K (type-III) could
be due to the very strong interaction between NiO and the
compounds such as Al2 O3 and MgO present in CC [6870]. It
can therefore be surmised that different elements on CC have
interacted with NiO species, contributing to different degree
reduction peaks on TPR profiles. Likewise, TPR profile of
pure CC also yielded two reduction peaks (Figure 8) at 982
and 1036 K symptomatic of H2 desorption from subsurface
layers and/or spillover H2 . As shown in Table 4, 15 wt%
Ni/CC catalyst had higher metal surface than 20 wt% Ni/CC
catalyst. This may be due to the lower dispersion or agglomerations of Ni on the surface of 20 wt% Ni catalyst. As NiO
reduction still occurred at 1031 K, therefore 1073 K was selected as the in-situ reduction temperature for the catalysts
prior to reaction studies.

flowrate was obtained over 20 wt% Ni/CC catalyst, with an

average of 190 mLmin1 . At the 4th hour, the gaseous output
flowrate of 5 wt%, 10 wt% and 15 wt% Ni/CC catalysts were
almost similar to each other, approximately, 160 mLmin1.
The gaseous product comprised primarily of H2 , CO, CO2 and
CH4 . Figure 10 represents the gaseous product output (vol%)
at the 4th hour when steady state condition was attained. Significantly, this indicates that H2 and CO were the products
from the primary reaction (Equation 3) whilst CH4 which occupied the least percentage of no more than 5 wt% and decreased when Ni loading was increased, was a by-product that
most likely originated from the competing methanation process. The outlet composition of H2 gaseous increased with Ni
loading, whilst CO decreased. Besides, less output of gaseous
CO2 over higher Ni doped catalysts was observed, indicating
most of CO2 were being consumed for glycerol dry reforming.

3.2. Glycerol dry reforming studies

The catalysts with different Ni doping levels were
screened over continuous 4 h reaction at 1023 K (the highest reforming temperature employed in the current work) and
also CO2 -to-C3 H8 O3 -ratio (CGR) of unity with weight hourly
1
space velocity (WHSV) of 3.6104 mLg1
cat h . Blank tests
using the same feed with either an empty reactor or calcined
CC bed yielded negligible glycerol conversion. This suggests
that neither homogeneous gas phase glycerol dry reforming
nor reaction over sites on the support occurred at sufficient
rates.
Figure 9 shows the highest gaseous product output

Figure 9. Output flowrate of the gaseous product for Ni/CC catalysts with
different Ni loadings

Figure 10. H2 , CO, CH4 and CO2 product output volume (%), glycerol conversion (XG ) at CGR of unity, T = 1023 K at 4th-h reaction. The WHSV was
1
set at of 3.6104 mLg1
cat h

Table 5 summarizes the transient yield of gaseous product (%), glycerol conversion (XG ) and the H2 /CO ratio obtained for the 1st and 4th hour of screening tests. The catalytic performance at the initial stage of reaction (1st hour)
was ranked in the order of 10 wt%>20 wt%>15 wt%>5 wt%
in terms of glycerol conversion (XG ). After 4 h reaction, XG
attained range from 64% to 76% for both 15 wt% Ni and
20 wt% Ni catalysts. XG over 10 wt% catalysts decreased
to the lowest one at the 4th hour, which was 46.14% whilst
5 wt% catalyst showed an increase of XG with margin of
29%. 5 wt% Ni/CC possessed higher stability over 4 h reaction than 10 wt% Ni/CC catalysts, which might be attributed
to the smaller Ni oxide crystallize size (Table 3). 15 wt% and
20 wt% Ni/CC catalysts showed higher conversion; 5 wt% and
10 wt% Ni/CC catalysts showed improved catalytic stability

Table 5. Glycerol conversion and yield of gases products

Catalysts
5 wt% Ni/CC
10 wt% Ni/CC
15 wt% Ni/CC
20 wt% Ni/CC

Glycerol conversion, XG (%)

1h
4h
24.7
53.4
77.0
46.1
65.3
63.8
76.7
75.6

Yield of H2 , YH2 (%)

1h
4h
13.6
28.6
53.7
24.2
56.4
50.4
69.3
66.1

Yield of CO, YCO (%)

1h
4h
32.7
69.4
89.9
59.6
51.3
47.7
55.0
59.3

Yield of CH4 , YCH4 (%)

1h
4h
7.4
16.6
15.5
14.6
5.9
8.9
4.9
6.3

653

Journal of Energy Chemistry Vol. 23 No. 5 2014

which was due to higher number of Ni active sites (Table 4)

that involving scission of CH and CC bonds of glycerol.
In terms of YH2 (Table 5), 5 wt% and 10 wt% Ni/CC
catalysts showed the similar yield at the 4th hour reaction,
in the range of 24% to 28% while achieving steady state.
On the other hand, both 15 wt% and 20 wt% Ni/CC catalyst showed an insignificant decrease of YH2 over 4 h reaction. As depicted in Figure 11, the rate of H2 production (rH2 )
over 4 h reaction for 15 wt% and 20 wt% Ni/CC catalysts
1 and 1.14104 molg1 s1 , rewere 1.5104 molg1
cat s
cat
spectively, double from 5 wt% and 10 wt% Ni/CC catalysts.
This can be attributed to the higher SBET of 15 wt% and
20 wt% Ni catalyst, which means more reactants was in contact with the active metals, resulting in higher H2 production
rate. Moreover, stronger Ni-support interaction of 15 wt%
and 20 wt% Ni/CC observed from TPR analysis has contributed to the insignificant deterioration in catalytic activity,
hence better catalyst performance was obtained. The rates of
CO and CH4 production formation (rCO and rCH4 ) in general reduced with higher Ni dopant levels. Figure 12 illustrates that H2 /CO ratios obtained at CGR of unity over the
catalysts were in the range of 0.5 to 1.50, which were substantially lower than those obtained from glycerol steam reforming (ratios averaging 10.0). H2 /CO and CO/CH4 product ratios generally increased with Ni dopant amounts. Overall, 15 wt% and 20 wt% Ni/CC catalysts yielded H2 /CO ratios between 1.5 to 2.0 which is desirable for Fischer-Tropsch
(FT) synthesis. Among all the evaluated catalysts, 20 wt%
Ni/CC catalyst exhibited the best performance and stability
with the highest XG and YH2 , which could be attributed to
the stronger support-metal interaction and total metal surfaces
from TPR analysis. Besides, 20% Ni/CC catalysts yielded
the lowest methanation activity, satisfying H2 /CO product
ratio 1.5.

Figure 12. Product ratio of different catalysts (data were taken at 4 h on stream)

3.3. Used catalyst characterization

3.3.1. Temperature-programmed oxidation (TPO)
Dry reforming process normally suffered from catalyst
deactivation via coke deposition [71]. The used catalyst was
subjected to temperature-programmed oxidation (TPO) analysis to study the chemical nature of coke deposit. The use
of pure O2 during the oxidation ensured complete conversion
of coke into CO2 gas. Table 6 and Figure 13 list the total
amount of carbon deposits and its gasifying temperature by
oxidation. Coke deposition showed an increasing trend with
more Ni loading. The amount of coke deposit on 15 wt%
and 20 wt% Ni/CC catalysts showed no significant difference
(difference of 0.01 gcoke g1
cat ), as was also between 5 wt% and
10 wt% Ni/CC catalyst (difference of 0.03 gcoke g1
cat ). It can
therefore inferred that higher Ni loading produced higher activity in terms of catalytic cracking of CC bond, leading to
higher coke deposition. There were three gasifying peaks temperature regions, viz. Peak I, 884 to 900 K; Peak II, 907 to
941 K and Peak III, 960 to 1000 K, which can be categorized
into different types of carbon formed. The presence of those
peaks indicates the laydown of carbon deposits with dissimilar reactivity towards O2 species. Fundamentally, this posits
to entirely different chemical species of carbons, viz. Peak
I temperature region assigned to the fibers or whiskers type
coke (Cv ) can be easily removed at lower gasification, whilst
Peaks II and III temperature regions falls to the graphitic-alike
(Cc ) carbon species [72].
Table 6. Total amount of coke deposited on catalysts
after 4 h on stream, at 1 atm, CGR of unity and 1023 K
Catalysts

Figure 11. Rate of product formation at 4 h on stream

5 wt% Ni/CC
10 wt% Ni/CC
15 wt% Ni/CC
20 wt% Ni/CC

Total carbon
(gcoke g1
cat )
0.46
0.49
1.78
1.79

Carbon gasification temperature (K)

Peak I
Peak II
Peak III
890
941

885, 890
946
1000
884
907
960
900
921
960

654

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

Figure 13. TPO profiles of various catalysts with different Ni loadings

3.3.2. FESEM images

As aforementioned, 20 wt% Ni/CC catalyst suffered from
the highest coke deposition, therefore it was further characterized with FESEM. By comparing Figure 14(a) with 14(b),
it apparently showed some spherical deposits, indicating carbon clusters encapsulation occurred on the surface of the used
cement clinker support. This corresponded to the previous
work [73] onto the gasoline fueled exhaust catalyst which

underwent morphology transformation that looks similar to

Figure 14(a). There were several short hollows tube was observed on the encapsulated Ni surface, it was believed to be the
starting of carbon whiskers formation. Obviously, there were
different types of carbon, which appeared to whiskers type of
carbon in the case of used 20 wt% Ni/CC catalyst (Figure 14b)
as opposed to the fresh catalyst (Figure 3e). This finding
was consistent with Papadopoulou et al. [74] whereby they
described that in most Ni based catalysts, carbon whiskers

Figure 14. FESEM images of used cement clinker support (a) and used 20 wt% Ni/CC catalysts (b)

Journal of Energy Chemistry Vol. 23 No. 5 2014

predominate. Zooming into the circles zone of Figure 14(b),

the hollow formations at the tip of carbon filaments (nanotubes) were clearly seen. Ni particles on the surface of the
support were pushed away by the growing carbon whiskers
and Ni metals were eventually dropped onto the open dissemination over the support surface, leaving the hollows.
4. Conclusions
Cement clinker catalysts with Ni dopant were synthesized
for glycerol dry reforming reaction. The glycerol conversion obtained over the screening of the catalysts was in the
range of 46% to 75% over 4 h on stream test. Therefore, cement clinker could be the potential support of glycerol dry reforming. Physicochemical characterization has revealed that
cement clinker was a complex mixture of oxide compounds
with CaO and SiO accounted for more than half (62% and
17%, respectively). Although cement clinker was macroporous, the addition of Ni has significantly improved the BET
specific surface area (from 0.55 to 17.83 m3 g1 ). XRD examination of the calcined catalysts showed complex Ni oxide phase formation at different Ni loading. The screening
of the catalysts has determined that 20 wt% Ni is the best
doping percentage, as 20 wt% Ni/CC catalysts has the highest activity and stability in terms of glycerol conversion regardless the effect of coke deposition. Glycerol dry reforming activity over 20 wt% Ni/CC catalysts also yielded H2 /CO
1.5 at 1023 K, which is desirable for Fischer-Tropsch synthesis compared with glycerol steam reforming. In post dry
reforming, used catalysts characterization revealed the phenomenon of carbon deposition which was confirmed by the
whisker-type carbon laydown captured by FESEM technique.
Significantly, temperature-programmed oxidation studies indicated that the carbon deposit can be completely gasified
with O2 and deduced the possibility of catalyst recyclability.
Acknowledgements
Authors would like to acknowledge the financial support from
Ministry of Education, Malaysia through MTUN (No. RDU121216).
Both Hua Chyn Lee and Kah Weng Siew appreciate the Universiti
Malaysia Pahang for a provision of GRS120377 scholarship.

References
[1] EIA. EIA projects world energy consumption will increase 56%
by 2040. http://www.eia.gov/todayinenergy/detail.cfm?id =
12251.U.S.EnergyInformationAdministration, 2013
[2] Kunzig R. Climate Milestone: Earths CO2 Level Passes
400 ppm. News. http://news.nationalgeographic.com/news/energy
/2013/05/130510-earth-co2-milestone-400-ppm/.National
Geographic, 2013
[3] IEA. CO2 emissions from fuel combustion-highlights.
https://www. iea. org/co2highlights/co2highlights. pdf. International Energy Agency (IEA), Paris, 2013
[4] Ma F R, Hanna M A. Bioresour Technol, 1999, 70(1): 1
[5] Fukuda H, Kondo A, Noda H. J Biosci Bioeng, 2001, 92(5): 405
[6] Herseczki Z, Marton G, Varga T. Hung J Ind Chem, 2011, 39:

655

183
[7] Long Y D, Fang Z, Su T C, Yang Q. Appl Energy, 2014, 113:
1819
[8] Singhabhandhu A, Tezuka T. Energy, 2010, 35(6): 2493
[9] Abad S, Turon X. Biotechnol Adv, 2012, 30(3): 733
[10] Hunsom M, Saila P, Chaiyakam P, Kositnan W. Int J Renew Energy Res, 2013, 3(2): 364
[11] Adhikari S, Fernando S D, Haryanto A. Renew Energy, 2008,
33(5): 1097
[12] Lauriol-Garbay P, Millet J M M, Loridant S, Belli`ere-Baca V,
Rey P. J Catal, 2011, 280(1): 68
[13] van Bennekom J G, Kirillov V A, Amosov Y I, Krieger T,
Venderbosch R H, Assink D, Lemmens K P J, Heeres H J. J
Supercrit Fluids, 2012, 70: 171
[14] Babajide O. J Energy, 2013: 1
[15] Czernik S, French R, Feik C, Chornet E. Ind Eng Chem Res,
2002, 41(17): 4209
[16] Adhikari S, Fernando S, Gwaltney S, Filip To S D, Mark Bricka
R, Steele P H, Haryanto A. Int J Hydrog Energy, 2007, 32(14):
2875
[17] Buffoni I N, Pompeo F, Santori G F, Nichio N N. Catal Commun, 2009, 10(13): 1656
[18] Cheng C K, Foo S Y, Adesina A A. Catal Commun, 2010, 12(4):
292
[19] Montini T, Singh R, Das P, Lorenzut B, Bertero N, Riello P,
Benedetti A, Giambastiani G, Bianchini C, Zinoviev S, Miertus
S, Fornasiero P. ChemSusChem, 2010, 3(5): 619
[20] Djinovic P, Batista J, Pintar A. Int J Hydrog Energy, 2012, 37(3):
2699
[21] Albarazi A, Beaunier P, Da Costa P. Int J Hydrog Energy, 2013,
38(1): 127
[22] Raberg L B, Jensen M B, Olsbye U, Daniel C, Haag S,
Mirodatos C, Sjastad A O. J Catal, 2007, 249(2): 250
[23] Blanchard J, Oudghiri-Hassani H, Abatzoglou N, Jankhah S,
Gitzhofer F. Chem Eng J, 2008, 143(1-3): 186
[24] Hu X, Lu G X. Catal Commun, 2009, 10(13): 1633
[25] Chen T H, Liu H B, Shi P C, Chen D, Song L, He H P, Frost R
L. Fuel, 2013, 107: 699
[26] Wang X D, Li M S, Wang M H, Wang H, Li S R, Wang S P, Ma
X B. Fuel, 2009, 88(11): 2148
[27] Dantas S C, Escritori J C, Soares R R, Hori C E. Chem Eng J,
2010, 156(2): 380
[28] Bermudez J M, Fidalgo B, Arenillas A, Menendez J A. Fuel,
2012, 94(1): 197
[29] Zhang Y F, Zhang G J, Zhang B M, Guo F B, Sun Y L. Chem
Eng J, 2011, 173(2): 592
[30] Calles J A, Carrero A, Vizcano A J, Garca-Moreno L. Catal
Today, 2014, 227: 198
[31] Bellido J D A, De Souza J E, MPeko J C, Assaf E M. Appl
Catal A, 2009, 358(2): 215
[32] Al-Fatish A S A, Ibrahim A A, Fakeeha A H, Soliman M A,
Siddiqui M R H, Abasaeed A E. Appl Catal A, 2009, 364(1-2):
150
[33] Ruckenstein E, Hu Y H. Appl Catal A, 1995, 133(1): 149
[34] Kurdowski W. Cem Concr Res, 2002, 32(3): 401
[35] Tsakiridis P E, Agatzini-Leonardou S, Oustadakis P, Katsioti M,
Mauridou E. Cem Concr Res, 2005, 35(11): 2066
[36] Halmann M, Steinfeld A. Energy, 2006, 31(10-11): 1533
[37] Gimbun J, Ali S, Kanwal C C S C, Shah L A, Ghazali N H M,
Cheng C K, Nurdin S. Procedia Eng, 2013, 53: 13
[38] Li H X, Agrawal D K, Cheng J P, Silsbee M R. Cem Concr Res,
1999, 29(10): 1611

656

Hua Chyn Lee et al./ Journal of Energy Chemistry Vol. 23 No. 5 2014

[39] Taylor H. Cement Chemistry. 2nd ed. London: Academic Press,

1990
[40] Gross S. Israel Geol Sur Israel Bull, 1977: 1
[41] Masse S, Boch P, Vaissi`ere N. J Eur Ceram Soc, 1999, 19(1):
93
[42] Ahmed F, Arshi N, Anwar M S, Lee S H, Byon E S, Lyu N J,
Koo B H. Curr Appl Phys, 2012, 12: S174
[43] Trezza M. Mater Res, 2007, 10(4): 331
[44] Ylmaz B, Olgun A. Cem Concr Compos, 2008, 30(3): 194
[45] Pavia D L, Lampman G M, Kriz G S. Introduction to Spectroscopy. 4th ed. Belmont, CA: Brooks Cole, 2009
[46] Trezza M A, Lavat A E. Cem Concr Res, 2001, 31(6): 869
[47] Ylmen R, Jaglid U, Steenari B M, Panas I. Cem Concr Res,
2009, 39(5): 433
[48] Zaki M I, Knozinger H, Tesche B, Mekhemer G A H. J Colloid
Interface Sci, 2006, 303(1): 9
[49] Estokova A, Palasc a kova L, Singovszka E, Holub M. Procedia
Eng, 2012, 42: 123
[50] Singh N K, Mishra P C, Singh V K, Narang K K. Cem Concr
Res, 2003, 33(9): 1319
[51] Mollah M Y A, Kesmez M, Cocke D L. Sci Total Environ, 2004,
325(1-3): 255
[52] Puertas F, Garca-Daz I, Palacios M, Gazulla M F, Gomez M P,
Orduna M. Cem Concr Compos, 2010, 32(3): 175
[53] Qiao H H, Wei Z Q, Yang H, Zhu L, Yan X Y. J Nanomater,
2009, 795928
[54] Rahemi N, Haghighi M, Babaluo A A, Jafari M F, Khorram S.
Int J Hydrog Energy, 2013, 38(36): 16048
[55] Estelle J, Salagre P, Cesteros Y, Serra M, Medina F, Sueiras J E.
Solid State Ionics, 2003, 156(1-2): 233
[56] Mile B, Stirling D, Zammitt M A, Lovell A, Webb M. J Catal,
1988, 114(2): 217

[57] Deng X Y, Zhang Z K. J Qingdao Univ Sci Technol (Qingdao

Keji Daxue Xuebao), 2006, 27(1): 24
[58] Brockner W, Ehrhardt C, Gjikaj M. Thermochim Acta, 2007,
456(1): 64
[59] Loaiza-Gil A, Villarroel M, Balbuena J F, Lacruz M A,
Gonzalez-Cortes S. J Mol Catal A-Chem, 2008, 281(1-2): 207
[60] Paulik F, Paulik J, Arnold M. Thermochim Acta, 1987, 121: 137
[61] Mohamed M, Yusup S, Maitra S. J Eng Sci Technol, 2012, 7(1):
1
[62] Gabrovsek R, Vuk T, Kaucic V. Acta Chim Slov, 2006, 53(2):
159
[63] Lemonidou A A, Vasalos I A. Appl Catal A, 2002, 228(1-2):
227
[64] Joo O S, Jung K D. Bull Korean Chem Soc, 2002, 23(8): 1149
[65] Kumar P, Sun Y P, Idem R O. Energy Fuels, 2008, 22(6): 3575
[66] Richardson J T, Turk B, Twigg M V. Appl Catal A, 1996, 148(1):
97
[67] Choudhary V R, Rajput A M, Mamman A S. J Catal, 1998,
178(2): 576

[68] Avila-Neto
C N, Dantas S C, Silva F A, Franco T V, Romanielo
L L, Hori C E, Assis A J. J Nat Gas Sci Eng, 2009, 1(6): 205
[69] Al-Dalama K, Stanislaus A. Thermochim Acta, 2011, 520(1-2):
67
[70] Dong W S, Roh H S, Liu Z W, Jun K W, Park S E. Bull Korean
Chem Soc, 2001, 22(12): 1323
[71] Kale G R, Kulkarni B D. Fuel Process Technol, 2010, 91(5):
520
[72] Bartholomew C H. Appl Catal A, 2001, 212(1-2): 17
[73] Jia L W, Zhang J, Shen M Q, Wang J, Lin M Q. Catal Commun,
2005, 6(12): 757
[74] Papadopoulou C, Matralis H, Verykios X. In: Guczi L,
Erdohelyi A eds. Catalysis for Alternative Energy Generation.
New York: Springer, 2012. 57