Influence of niobium on carbon nanofibres based Cu/ZrO2catalysts for liquid phase hydrogenation of CO2to methanol

Article abstract of DOI:10.1016/j.cattod.2015.06.019

A series of carbon nanofibres supported Cu/ZrO₂/Nb₂O₅catalyst synthesized by deposition precipitation method were extensively investigated in relation to their performance in hydrogenation of CO₂to methanol. In order to study the promotion effect of niobium, catalysts were loaded with 0.4, 0.8 and 1.2 wt.% of Nb₂O₅. Incorporation of Nb₂O₅facilitated copper reduction by exhibiting a shift of highly reduced copper peak to lower temperature. Likewise, surface enrichment of copper was also enhanced by introduction of Nb₂O₅to the catalysts. Activity studies of Nb₂O₅doped Cu/ZrO₂catalysts were evaluated in a slurry reactor with a CO₂/H₂gas mixture of 1:3 volume ratio, 180^oC temperature and 3.0 MPa total pressure. The highest activity was achieved with the incorporation of 0.8 wt.% of Nb₂O₅, illustrating the high degree of CuO crystallization on the surface of catalyst are beneficial for the generation of copper catalyst with enhanced activities.

Full text of DOI:10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Influence of niobium on carbon nanofibres based Cu/ZrO catalysts for
2
liquid phase hydrogenation of CO to methanol
2
a
a,∗
b
c
d
Israf Ud Din , Maizatul S. Shaharun , Duvvuri Subbarao , A. Naeem , F. Hussain
a
Department of Fundamental and Applied Sciences, Universiti Teknologi PETRONAS, Malaysia
Department of Chemical Engineering, Universiti Teknologi PETRONAS, Malaysia
National Centre of Excellence in Physical Chemistry, University of Peshawar, Pakistan
Wuhan University of Technology, China
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
2 2 5
A series of carbon nanofibres supported Cu/ZrO /Nb O catalyst synthesized by deposition precipita-
Received 31 December 2014
Received in revised form 9 June 2015
Accepted 29 June 2015
Available online xxx
tion method were extensively investigated in relation to their performance in hydrogenation of CO2 to
methanol. In order to study the promotion effect of niobium, catalysts were loaded with 0.4, 0.8 and
1
.2 wt.% of Nb2O5. Incorporation of Nb2O5 facilitated copper reduction by exhibiting a shift of highly
reduced copper peak to lower temperature. Likewise, surface enrichment of copper was also enhanced
by introduction of Nb2O5 to the catalysts. Activity studies of Nb2O5 doped Cu/ZrO2 catalysts were eval-
Keywords:
Methanol synthesis
Slurry reactor
Carbon nanofibres
Copper based
o
uated in a slurry reactor with a CO2/H2 gas mixture of 1:3 volume ratio, 180 C temperature and 3.0 MPa
total pressure. The highest activity was achieved with the incorporation of 0.8 wt.% of Nb2O5, illustrating
the high degree of CuO crystallization on the surface of catalyst are beneficial for the generation of copper
catalyst with enhanced activities.
Carbon dioxide conversion
Niobium promoter
© 2015 Elsevier B.V. All rights reserved.
1
. Introduction
Incorporation of promoters like ZnO, Cr O3 and Nb O5 into the
2
2
Cu/ZrO catalysts have resulted in higher activity and selectivity to
2
Chemical transformation of carbon dioxide into petrochemi-
methanol in the CO2 hydrogenation reaction [2,7]. The improved
performance of the niobia-supported catalysts was attributed to
the formation of new active sites due to the greater interaction of
support with the metal atoms [8]. In addition, niobia-supported
catalysts were reported to have low acidity property and therefore
exhibited better activity and selectivity in hydrogenation reaction
of CO and CO2 [9–11].
cal products has been extensively investigated over the last few
decades [1]. At present, catalytic reduction of CO2 to methanol is
considered as a promising economical route which could contribute
to CO2 mitigation. In fact, current industrial scale methanol syn-
thesis is carried out over Cu–ZnO/Al O catalysts with a mixture of
2
3
◦
syngas (CO/H ) and CO2 at high operating temperature 300 C [2].
2
Furthermore, alumina based catalysts displayed poor activity for
In the above context, the specific objectives of this work are
CO hydrogenation to methanol due to its high water-affinity [3,4].
to investigate the effect of Nb O5 on the physical, structural and
2
2
These research outcomes have stimulated the development of new
catalytic system for hydrogenation of pure CO2 to methanol. Nev-
ertheless, the lack of probative evidence on the mechanism of CO2
hydrogenation is the major obstacle for the development of new
catalyst formulation [5].
activity of CNFs based Cu/ZrO2 catalysts in the CO2 hydrogenation
to methanol. The CNFs based Cu/ZrO catalyst was used as reference
2
in the catalytic tests.
Carbon nanofibres (CNFs) with a high surface area and hydro-
phobic nature could be a good alternative as a catalyst support
for hydrogenation of pure CO2 to methanol. ZrO2 doped Cu cata-
lysts have shown promising results for CO2 hydrogenation [4,6].
2. Experimental
2.1. Functionalization of carbon nanofibres (CNFs)
Surface of CNFs was modified by treating with 35 vol. %
nitric acid solution. The refluxing was continued for 16 h at ele-
◦
∗ Corresponding author.
E-mail address: maizats@petronas.com.my (M.S. Shaharun).
vated temperature of 90 C. After refluxing, CNFs were cooled to
room temperature and filtered by vacuum filtration. After washing
http://dx.doi.org/10.1016/j.cattod.2015.06.019
0
920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
2
several times with distilled water, oxidized CNFs (CNFs-O) were
dried overnight in oven at 100 C.
photoelectron spectroscope (XPS, Thermo-Fisher K-Alpha)
equipped with monochromitized AlK source having ultimate
energy resolution of ≤0.5 eV was used in XPS studies. Avantage
software was used for peak fitting and chemical state identification.
Surface basicity was examined by using CO2 temperature
programmed desorption (CO2-TPD). Prior to TPD analysis, sam-
◦
2
.2. Synthesis of Nb O5 doped Cu/ZrO /CNF (CZC-Nb) catalysts
2 2
Deposition precipitation method was utilized for the synthesis
of Nb O5 promoted Cu/ZrO /CNF catalysts [12,13]. A series of
◦
ples were subjected to temperature of 500 C for 60 min under
2
2
catalyst containing a constant loading of 15 wt.% Cu and 15 wt.%
inert atmosphere to desorb the surface moisture and other
adsorbed molecules. Pre-reduced catalysts were cooled to room
temperature and were saturated with pure CO2. The adsorp-
ZrO2 with varying Nb O5 content ranged from 0.4, 0.8 and 1.2 wt.%
2
were synthesized. A known quantity of zirconyl nitrate hydrate
◦
(
SIGMA-ALDRICH, USA) was added gradually to the solution of
tion of gases was continued for 1 h at 90 C and physiosorbed
Cu (NO ) ·3H O (R&M Chemicals, UK). When both nitrate salts
molecules were desorbed with He flow. The adsorbed gas was
desorbed in temperature range of 40–800 C. Desorption of CO2
at relative degree of temperature were quantified by calibrated
TCD.
3
2
2
◦
were completely dissolved, required quantity of CNFs-O was
added to the solution. The suspension was vigorously stirred and
◦
heated to 90 C, then the slurry solution was precipitated with
◦
0
.1 g/mL of urea solution. The precipitates were aged at 90 C for
2
0 h, cooled and filtered by vacuum filtration. The precipitates
◦
2.4. Catalytic Tests
were dried in oven at 110 C for overnight. The catalysts were
calcined in N2 flow at 450 C for 3 h and labelled as CZC-Nb0.4,
◦
Activity of catalysts in CO2 hydrogenation to methanol was
evaluated in autoclave slurry reactor slurry reactor (Parr 4593).
Prior to the activity studies, the catalysts were reduced for 6 h
in H2 with flow rate of 2000 cm /h at 380 C. A 0.5 grams of
reduced sample was suspended in 25 ml of ethanol placed in reac-
tion vessel. The reactor was purged at room temperature and
CZC-Nb0.8, and CZC-Nb1.2 catalysts. A reference CNF based
Cu/ZrO2 catalyst with a nominal composition of Cu:Zr = 50:50
(
wt.%) was prepared using similar technique described
3
◦
above.
2.3. Characterization
then pressurized with mixture of H /CO2 gases with 3:1 molar
2
ratio to the desirable pressure of 3.0 MPa. The reaction stud-
ies were performed at 180 C. Reaction mixture was agitated by
PANalytical model Empyrean X-ray diffractometer was
◦
employed for phase studies of catalyst components. PANalyt-
ical High Score Plus software was used for phase identification.
stirrer and a speed of 1300 rpm was selected to avoid mass diffu-
sion constrains. Analysis of reactants and products were carried
out on Agilent GC-6890 system chromatograph equipped with
a flame ionization and thermal conductivity detectors. Exper-
iments were repeated three time to check for reproducibility.
Measurements are in general reproducible within a maximum of
◦
The XRD data were measured at room temperature from 20 to
0 2ꢀ Bragg angle.
◦
8
Nitrogen adsorption–desorption isotherms technique was car-
ried out for investigations of catalysts surface area and pore volume
using Micrometrics ASAP 2020 [3].
1
0%. Turnover frequency of methanol was calculated by following
Copper metallic surface area (SCu), dispersion of Cu (DCu),
formula [22,23]:
average particle size (dCu) and distribution of Cu content (RCu
)
were determined by N O chemisorption technique [14–17]. Cat-
2
◦
A × Na
alysts were first reduced with H at 500 C. Reduced samples
TOFMeOH (s−1) =
2
◦
3600 · S · na
were cooled in He flow to 60 C and purged for 30 min. Then
Cu
N O gas was introduced for 1 h. Residual N O was flushed
2
2
where A represents methanol activity in mol/g h, Na is Avogadro’s
out by He flow for 1 h. Finally, the samples were reduced for
23
◦
number (6.023 × 10 ), SCu denotes metallic copper surface area
the second time at 500 C. Surface area and dispersion of Cu
2
were measured by assuming 1.46 × 10 Cuat/m2 surface atomic
19
in m /g and na designates number of Cu atoms in a monolayer
2
(
na = 1.469 × 1019 atoms/m )
density and Cu:N O = 2 stoichiometry, respectively. Average par-
2
ticle size (dCu) was obtained by a relationship displayed as
follows [3,18,19].
1
04
dCu(nm) =
DCu(%)
The distribution of Cu content was estimated by the following equa-
tion [20].
0
Cu surface area
Cu content × BET surface area
Transmission Electron Microscopy (TEM) was used to study
RCu
=
morphology and particle size measurement of the catalysts. Zeiss
LIBRA 200TEM with accelerating voltage of 200 kV was utilized for
this purpose [21].
Temperature Programmed Reduction (TPR) technique was used
to study the reduction behaviour of catalyst and metal support
interactions. TPDRO1100 MS equipped with thermal conductivity
◦
detector (TCD) was used in temperature range of 30–800 C with
◦
−1
heating rate of 10 C min . The analyses were performed in 5 vol.%
3
1
H /N flow with a flow rate of 20 cm min .
2
2
X-ray photoelectron spectroscopy was utilized to inves-
Fig. 1. XRD profile of (a) CNFs-O, (b) CZC, (c) CZC-Nb0.4, (d) CZC-Nb0.8 and (e)
tigate chemical nature and surface composition of Cu. X-ray
CZC-Nb1.2 catalysts.
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
3
Fig. 2. TEM images of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-Nb1.2 catalysts.
Table 1
Fig. 1. For comparison an XRD spectrum of CNFs-O and unloaded
CZC catalyst were also included. Hexagonal graphitic planes of CNFs
Average particle size of catalyst components.
◦
Catalysts
Copper average
particle size (nm)
Zirconia average
particle size (nm)
were identified by two prominent peaks at 2ꢀ values of 26 and
◦
4
4 (JCPDS No. 41-1487). Similarly, diffraction pattern with peaks
◦
◦
◦
◦
◦
◦
◦
◦
◦
◦
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
4
5
7
6
8
9
at 32.6 , 35.5 , 38.7 , 48.8 , 53.6 , 58.3 , 61.67 , 66.4 , 68.1 , 72.3
◦
and 75.1 on 2ꢀ scale was found which is indexed as monoclinic
phased tenorite CuO with JCPDS card files No. 48-1548 (a = 4.62 A,
b = 3.43 A˚ , and c = 5.06 A˚ ). A small diffraction peak around 31 was
˚
◦
found in XRD profile, indicating zirconia component of the catalyst.
However, it was disappeared with the incorporation of Nb con-
tent. This indicates transformation of zirconia to the amorphous
phase with the promotion of niobium loading. Highly dispersed
or very fine crystalline ZrO2 (crystallite size < 2 nm) could also
be possible reasons. However, peak corresponding to Nb2O5 or
any Cu-Nb composite was not observed indicating that Nb2O5
existed in amorphous or microcrystalline state, which could not
3
3
3
. Results and discussion
.1. Catalyst characterization
.1.1. Phase determination studies
Phase analysis of catalyst components were investigated by XRD
technique. XRD profile of Nb promoted CZC catalysts is shown in
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
4
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
Fig. 3. N2 adsorption desorption isotherms of (a) CZC, (b) CZC-Nb0.4, (c) CZC-Nb0.8 and (d) CZC-Nb1.2 catalysts.
Table 2
be detected by XRD technique due to low degree of crystalliza-
tion. An interesting observation was found in the XRD studies
whereby the intensification of CuO peaks is dependent on addi-
tion of Nb O5 up to 0.8 wt.% Nb O5 concentration. The increase
Textural properties of CZC and Nb promoted CZC catalysts.
Catalyst
Tot.
ads. gas
SBET
(m /g)
Pore
diameter
(nm)
Pore
2
volume
2
2
3
3
(
cm /g)
(cm /g)
in CuO peak intensity with increasing Nb O5 loading reflects
2
the growth in crystal size. However, the intensity of CuO peak
CZC
290
220
207
213
155
108
104
105
10.1
11.3
11.3
11.2
0.39
0.30
0.29
0.30
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
decreased with further rise in concentration of 1.2 wt.% Nb O5.
2
The decreasing CuO peak intensity with further rise in Nb O5
2
content indicates the transformation of crystalline CuO to the amor-
phous form. Such a trend of decreasing CuO peak intensity with
increasing copper concentrations was also reported by Rajaram
et al. [24].
3.1.3. BET surface area and pore size distribution
Nitrogen adsorption-desorption isotherms of CZC and Nb pro-
moted CZC catalysts are shown in Fig. 3. Each catalyst exhibited
a typical type-IV isotherm with H4 type hysteresis loops hav-
3.1.2. Morphology investigations
o
TEM images of CZC and Nb promoted CZC catalysts are presented
ing sharp inflection between p/p ranges of 0.90–0.94, indicating
in Fig. 2. Dark black spherical shaped particles were identified as
copper particles whereas tetragonal shaped light coloured particles
were recognized as zirconia particles. Particles of copper and zirco-
mesoporus nature of the synthesized catalysts. The textural prop-
erties of CZC catalysts are provided in Table 2. BET surface area
decreased remarkably from 155 to 108 m /g with the incorpora-
2
nia are clearly seen by TEM images. However, Nb O5 particles could
tion of 0.4 wt.% of Nb O5 to the CZC catalyst. Although, the surface
2
2
not be identified because of their low concentrations. The results
demonstrated a homogenous deposition of catalysts particles with
area reduced with further increase of Nb O5 but the change was
2
not very prominent. The decline of surface area with Nb O5 intro-
2
low concentrations of Nb O5. Nevertheless, agglomerations of
Cu particles were observed when Nb content exceeded 0.8 wt.%.
Average particle size of both metal oxides increased with increas-
duction is supported by a drastic decrease in magnitude of total
adsorbed gas from 290 to 220 (cm /g) for CZC and CZC-Nb0.8 cat-
2
3
alysts, respectively. Similarly, pore volume decreased from 0.39 to
3
ing Nb O5 content (Table 1) due to the agglomeration of metal
0.29 cm /g with the introduction of 0.8 wt.% of Nb O to CZC cata-
2
2
5
oxides.
lysts. The decline in BET surface area and pore volume could be due
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
5
Fig. 4. TPR profile of (a) CZC, (b) CZC-Nb0.4, (c) CZC-Nb0.8 and (d) CZC-Nb1.2 cata-
lysts.
Fig. 5. XPS of Cu 2p of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-Nb1.2 catalysts.
to pore-filling by the introduction of Nb O5. In contrast, the pore
2
diameter of the catalyst increased with addition of 0.4 wt.% Nb O5
to the CZC catalyst. However, it remained invariant with further
2
Table 4
XPS data of CZC catalysts with different Nb content.
addition of Nb O5.
2
Sample
Binding energy (eV)
FWHM (eV)
Atomic Cu/Zr
ratio
3
.1.4. Reducibility studies
The reduction behavior of the catalysts was studied by TPR tech-
Cu 2p3/2
Zr 3d5/2
Cu 2p3/2
Zr 3d5/2
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
934.10
934.24
934.28
182.16
182.22
182.18
3.92
4.11
4.08
4.57
4.47
4.45
1.22
1.31
1.21
nique. TPR profile of CZC and Nb-loaded CZC catalysts is presented
in Fig. 4. The quantitative evaluation of TPR analysis is shown in
Table 3. Two distinct TPR peaks were observed in unprompted CZC
◦
catalyst with reduction maxima at 260 and 340 C and denoted
as peak ␣ and peak ␤, respectively. TPR peak ␣ is recognized as
reduction peak due to highly dispersed copper while TPR peak ␤ is
ascribed to bulk–like CuO [25]. Similarly, according to López-Suárez
observed in TPR peak ␤. Shift of peak ␣ to lower temperature with
Nb2O5 loading suggests that incorporation of Nb2O5 facilitated the
reduction of Cu2+, in agreement with previous works [33–35]. On
the other hand, peak ␤ experienced a higher temperature shift from
2+
et al., peak ␣ is assigned to easily reduced form of Cu and peak
correspond to less-reducible CuO [26]. However in concomitant
to these two traditional peak, a new broad reduction peak (peak
◦
␤
340 to 420 C when CZC catalysts were doped with Nb2O5. The shif-
ting of reduction peak is due to the formation of new phase due
to strong interaction between Cu and the dopant Nb2O5 [36–38].
Guarido et al. reported similar TPR results involving Cu/Nb2O5 cata-
lyst used in ethanol reforming and partial oxidation reactions [39].
The results proved that incorporation of Nb2O5 in CZC catalysts
facilitates the reduction of dispersed Cu at one end but depresses
the reducibility of bulk Cu on the other end.
◦ ◦
) starting at 350 C with a tail at 490 C was observed in Nb pro-
␥
moted catalysts. Roma et al. investigated TPR profile of Cu/Nb O5
2
catalysts and observed a broad peak for CuO reduction around
◦
4
50 C [27]. Based on this observation, peak ␥ in this work could
be due to higher interaction of Cu or Cu-Nb O5 composites. Like-
2
wise the occurrence of more than two reduction peaks has also been
reported in the literature [28]. Nevertheless, an additional reduc-
◦
tion peak was observed at 375 C for CZC catalysts with maximum
3.1.5. Metal–metal interaction and surface analysis
Nb O5 concentration. TPR peak at the same magnitude of tempera-
ture was also recorded by Courtois et al. and attributed to reduction
Fig. 5 shows the Cu 2p XPS spectra for Nb promoted CZC cat-
alysts. Each catalyst demonstrated Cu 2p3/2 core electrons peak
at 934.4 eV associated with a broad satellite peak around 943 eV.
Similarly, core electrons peak for Cu 2p1/2 was observed at 954 eV
accompanied with a shakeup peak around 962 eV. The occurrence
of satellite peaks in concomitant to the parental peaks confirms
Cu as the predominant oxidation state of Cu in all studied cata-
lysts. However, this predominant oxidation state was only observed
in the calcined catalysts and not under reaction conditions.
The binding energies of Cu 2p3/2 and Zr 3d5/2 core electrons and
their full width at half maximum (FWHM) values along with Cu/Zr
2
of bulk CuO [29]. Reduction of Nb O5 is generally reported in the
2
◦
temperature range of 800–900 C in the literature [30–32]. There-
fore in the current study, all the reduction peaks attribute to the
CuO reduction. This can be further justified by the fact that mag-
2+
nitude of H /Cu < 1 was recorded for each niobium promoted CZC
2
catalysts as evident in Table 3.
Incorporation of Nb O5 provides interesting changes in TPR
2
profile of CZC catalysts. Position of TPR peak ␣ was shifted from
◦ ◦
60 to 230 C while a lower temperature shift of almost 10 C was
2
Table 3
TPR data of CZC calcined at different temperature.
◦
Sample
H2 cons. (␮mol/g)
H2/Cu
Red. temp. ( C) peak
H2 con. (␮mol/g) peak
␣
␤
␥
␣
␤
␥
CZC
2172
2210
2239
2268
0.92
0.94
0.95
0.96
266
226
235
235
340
330
342
355
–
1542
288
291
301
690
521
401
517
–
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
437
444
446
1401
1547
1450
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
6
Fig. 6. XPS Zr 3d spectra of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-Nb1.2 catalysts.
atomic ratio are listed in Table 4. As evident from the tabulated
data, binding energy of Cu 2p3/2 has shifted slightly from 934.10
to 934.24 eV by increasing Nb O5 concentration from 0.4 to 0.8%,
2
respectively. However, substantial variation was not observed with
further increase in Nb O5 content. Similarly, position of Zr 3d5/2
2
on binding energy scale remained almost invariant for different
Nb O5 concentrations. Nevertheless, magnitude of Cu 2p3/2 FWHM
2
increased slightly from 3.92 to 4.11 eV by increasing Nb O5 con-
2
centration from 0.4 to 0.8%. This indicates the formation of two
2
+
2−
2
2−
additional weak Cu –Cu bonding with neighbouring O ions,
+
leading to distortion of Cu ion coordination symmetry towards
a highly distorted octahedral symmetry [40]. On the other hand,
FWHM value of Zr 3d5/2 were not affected by Nb O5 concentra-
2
4+
tion. Therefore information on bonding and nature of Zr ions
could not be obtained. Furthermore, relative atomic ratio of Cu/Zr
increased with by increasing Nb O5 content from 0.4 to 0.8%, sug-
Fig. 7. XPS Cu 2p peak fitting curves of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-
Nb1.2 catalysts.
2
gesting enrichment of surface copper. However, further addition
of Nb to the parent catalyst declined the Cu/Zr atomic ratio. This
demonstrates depletion of surface Cu and subsequent enrichment
of surface Zr content. Agglomeration of Cu could be one of the
reasons for surface Cu depletion.
XPS profile of Nb 3d core electrons of Nb promoted CZC cata-
lysts is displayed in Fig. 8. Two major peaks were observed for Nb
3d region, Nb 3d5/2 peak centred at 207 eV and Nb 3d3/2 peak origi-
nated at 210 eV. In order to identify the different forms of niobium
oxide, Nb 3d region was deconvoluted into six different peaks.
Generally metallic Nb is characterized by its XPS peak at 202 eV
which was not found in the current case, indicating the absence of
Nb metal. However, XPS analysis revealed the existence of Nb in
different oxygenated forms. The two main XPS peaks at 207.1 and
Likewise, XPS profile of Zr 3d of CZC catalysts with different
Nb O5 content is depicted in Fig. 6. Zirconium ions were identified
2
by two XPS peaks as Zr 3d5/2 and Zr 3d3/2 with binding energies of
1
82.2 and 184.6 eV, respectively. The presence of two different XPS
peaks with 2.4 eV energy gap indicates the existence of two differ-
ent types of zirconium ions. Lower binding energy peak at 182.2 eV
4
+
ratifies the existence of Zr as ZrO whereas higher binding energy
2
peak at 184.6 eV represents Zr2+ species [41].
209.7 eV were recognized as Nb (2) peaks, indicating the presence
5+
Cu 2p3/2 peak of each catalyst was resolved into different peaks
to identify the different Cu species and is displayed in Fig. 7. Cupric
ion can easily be recognized due to its characteristic coupling phe-
nomenon between unpaired electrons. However, magnitudes of
binding energies of Cu and Cu+ are so closed that the two peaks
overlap each other, hence these two Cu species could not be differ-
entiated. In the current study, higher energy Cu 2p3/2 peak observed
at 933.7 eV was attributed to Cu2+ ion while a low energy peak cen-
of Nb O5 [42,43]. Another major contribution of niobium oxide
2
5+
form was recognized by Nb (1) peak with binding energies of
206.5, 209.16 and 210.09 eV indicating NbO6 octahedra. Besides,
the existence of these two oxides, XPS peak for NbO4 was mani-
fested by a small shoulder at 208.1 eV. [44]. The contribution of each
0
oxide to the total Nb O5 was calculated and listed in Table 5. Incor-
2
poration of Nb O5 content affected the contribution of each oxide.
2
Relative percentage of Nb O5 increased whereas contribution of
2
+
tred at 932.7 eV was assigned to Cu ion. The appearance of two
NbO6 declined with increasing Nb content from 0.4 to 0.8 wt.%. In
contrast, the opposite trend was observed when the Nb content
was further increased from 0.8 to 1.2 wt.%. However, contribution
different CuO species by XPS analysis strongly support the H -TPR
2
findings where step-wise reduction of CuO was detected.
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
7
Fig. 9. CO2-TPD profile of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-Nb1.2 catalysts.
Table 7
N2O chemisorption data of un-promoted and Nb oxide promoted catalysts.
2
Sample
SCu (m /g)
DCu (%)
dCu (nm)
RCu
CZC
8.0
14.4
15.9
15.4
22.0
15.8
17.2
16.5
4.7
6.6
6.0
6.3
0.34
0.88
1.00
0.97
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
◦
were classified as medium basic sites (below 450 C) and strong
◦
basic sites (above 450 C). Irrespective of the Nb concentrations,
basic sites were predominantly found in the higher temperature
region. This implies that strength of basic sites is not affected by
variation in Nb content.
Total number of basic sites, their densities and relative distri-
bution is provided in Table 6. Total number of basic sites remained
almost constant throughout the range of Nb O5 concentration. Sim-
2
ilarly the densities of total basic sites were also found invariant
Fig. 8. XPS Nb 3d peak fitting curves of (a) CZC-Nb0.4, (b) CZC-Nb0.8 and (c) CZC-
Nb1.2 catalysts.
with the variation of Nb O5 concentrations. Magnitude of medium
2
basic sites increased moderately when the concentration of Nb O5
2
Table 5
increased from 0.4 to 0.8 wt.%. However the magnitude of medium
Contribution of each oxide to total niobium oxide.
basic sites declined with further addition of Nb O5 content beyond
2
0
.8 wt.%.
Catalyst
iA (%)
Nb2O5
NbO6
NbO4
3
.2. Chemisorption studies
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
68
74
71
28
23
26
4
3
3
The surface area of metallic copper (SCu), Cu dispersion (DCu),
average particle size (dCu) and relative distribution of metallic cop-
per (RCu) measured by N O chemisorption are presented in Table 7.
iA (%) represents the ratio of Ai/ꢁAi (Ai is the area of each peak).
2
As evident from the tabulated data, increase in Nb O5 content has
2
of NbO4 remained almost invariant throughout the range of Nb
loadings.
a marked influence on the surface properties of Cu. The surface
area of metallic copper increased progressively with incorpora-
tion of Nb O5 content in the parent catalyst and a maximum S
2
2
Cu
3.1.6. Basicity studies
of 15.9 m /g was obtained for CZC-Nb0.8 catalyst. However, the
Fig. 9 shows CO TPD profile of Nb promoted CZC catalysts. Basic
sites were found in temperature range of 250–650 C. Basic sites
SCu decreased with further rise in concentration of Nb O5 content
beyond 0.8 wt.%. Unlike SCu, dispersion of Cu (DCu) was adversely
2
2
◦
Table 6
CO2 TPD data of Nb promoted CZC catalysts.
Catalyst
Number of total basic sites (␮mol/g cat)
Density of total basic sites (␮mol/m2)
Medium basic sites (␮mol/g cat)
Strong basic sites (␮mol/g cat)
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
333
331
334
1.51
1.60
1.57
33
39
34
300
292
300
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
8
Table 8
Activity data of CZC and Nb promoted CZC catalysts.
TOFMeOH (s 1
−
)
Catalyst
Methanol activity (g/kg cat h)
CO2 conversion (%)
Methanol selectivity (%)
CZC
32
37
40
38
14
10
9
75
89
87
88
1.42
0.92
0.89
0.88
CZC-Nb0.4
CZC-Nb0.8
CZC-Nb1.2
8
Table 9
Comparative study of the current catalyst for CO2 hydrogenation to methanol with the literature data.
◦
Catalyst
T ( C)
P (bar)
Methanol activity (g/kg h)
CO2 conversion (%)
Reference
CuO/ZnO/Al2O3
Pd/Zn/CNTs
Pd/ZnO/Al2O3
PdZnO/AC
170
250
250
250
230
250
250
220
220
220
200
200
180
30
30
30
30
80
20
20
80
80
80
9
20
37
24
28
27
57
62
40
19
15
23
48
40
5.8
6.3
4.4
4.9
–
16
13.7
2
2
2
1.7
5.8
9
[7]
[42]
[42]
[42]
[36]
[41]
[41]
[40]
[40]
[40]
[47]
[47]
Ag/ZrO2
Cu/B2O3/ZrO2
Cu/Ga2O3/ZrO2
Cu/ZnO/ZrO2
Au/ZnO/ZrO2
Ag/ZnO/ZrO2
Cu/ZrO2
Cu/ZrO2/ZnO
CZC-Nb0.8
9
30
Present work
Fig. 10. Correlation of SCu and methanol TOF.
Fig. 11. Correlation of dCu and methanol TOF.
affected by incorporation of Nb O5 to the parent CZC catalyst. The
methanol, ethane, hexane and carbon monoxide were also found
as by-products in the GC chromatogram of the reaction mixture.
In terms of methanol synthesis, a comparative study of the activ-
ity data of this novel catalyst with the reported literature revealed
2
DCu increased from 15.8% to 17.2% with the addition of Nb O5 of
2
0.4 to 0.8 wt.%. However, it remained almost invariant with further
addition of Nb O5. On the other hand, the copper particle size was
2
less affected by variation of Nb O5 content. Distribution of surface
that the current Nb O5 promoted CZC catalysts showed higher
2
2
copper (RCu) increased to a maximum with the incorporation of
activity for methanol yield and CO conversion as compared to that
2
0
.8 wt.% Nb O5.
recorded by Shaharun et al. over Cu/ZnO/Al O and Sloczynski et al.
2
2 3
over Ag/ZnO/ZrO2 and Au/ZnO/ZrO2 catalysts [7,45]. Similarly, the
results obtained in this study were very much comparable in terms
of methanol yield and CO2 conversion to the work of Liu et al.,
for CO2 hydrogenation over Cu/Ga O /ZrO catalysts [46]. Fur-
3
.3. Catalytic tests
Table 8 displays the variation of catalysts performance as a func-
2
3
2
tion of Nb O5 loading. Methanol synthesis rate increased from 32
thermore, a comparable activity data for methanol synthesis were
reported by Liang et al. for carbon nanotube-supported Pd/ZnO cat-
alysts [47]. A detailed comparative study of the current catalyst for
CO2 hydrogenation to methanol with the literature is displayed in
Table 9.
2
to 37 (g/kg cat h) with the incorporation of 0.4 wt.% Nb O5 to the
2
CZC parent catalyst. Further addition of Nb O5 to 0.8 wt.% improved
2
the catalysts activity to 40 (g/kg cat h). Nevertheless, methanol syn-
thesis rate decreased with further incorporation of Nb O5 content
2
(
1.2 wt.%). The increment in methanol synthesis rate by Nb O5 pro-
Incorporation of 0.4 wt.% of Nb O5 led to decline of CO2 conver-
2
2
motion could be due to the ease of Cu reduction as indicated by TPR
studies. Likewise the increase in methanol synthesis rate could be
justified by the improvement of methanol selectivity as a function
of Nb O5 introduction to the parent catalyst. On the other hand, CO
sion from 14 to 10%. However, CO2 conversion remained almost
constant by further increase in Nb O5 content. From the activity
2
pattern, it could be inferred that Nb O5 promotion has a suppor-
2
ting effect on the methanol selectivity. Similar observations were
2
2
conversion was adversely affected by Nb O5 addition. Apart from
also reported by Passos et al. in the n-heptane conversion to olefins
2
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019

G Model
CATTOD-9669; No. of Pages9
ARTICLE IN PRESS
I.U. Din et al. / Catalysis Today xxx (2015) xxx–xxx
9
using Nb O5 promoted platinum catalyst [48]. Similarly turnover
[5] F. Arena, G. Mezzatesta, G. Zafarana, G. Trunfio, F. Frusteri, L. Spadaro, Catal.
Today 210 (2013) 39–46.
2
frequency (TOF) of methanol synthesis reduced significantly from
[
6] P. Gao, F. Li, F. Xiao, N. Zhao, W. Wei, L. Zhong, Y. Sun, Catal. Today 194 (2012)
9–15.
[7] S. Shaharun, M.S. Shaharun, D. Mohamad, M.F. Taha, AIP Conf. Proc. 1621
2014) 3–9.
−1
1
.42 to 0.92 s by incorporation of 0.4 wt.% Nb O5.
2
In order to investigate the role of SCu in CO2 hydrogenation to
(
methanol, a graph was plotted between SCu and TOF of methanol
formation. As evident from Fig. 10, TOF decreased with increasing
copper surface area. Generally activity of the copper based cata-
lysts increases linearly with increasing SCu. However conflicting
results have also been reported in the literature [49,50]. The results
obtained in the current study reveal that SCu is not the only factor
to regulate the catalysts activity. Apart from SCu different factors
affecting the catalytic activity have been proposed by different
[
[
8] Z. Hu, K. Kunimori, T. Uchijima, Appl. Catal. 69 (1991) 253–268.
9] T. Iizuka, Y. Tanaka, K. Tanabe, J. Mol. Catal. A: Chem. 17 (1982) 381–389.
[10] A. Frydman, D.G. Castner, C.T. Campbell, M. Schmal, J. Catal. 188 (1999) 1–13.
[11] M. Schmal, A. Frydman, J. Catal. 188 (1999) 1–13.
12] M.L. Toebes, M.K. van der Lee, L.M. Tang, M.H. Huis in‘t Veld, J.H. Bitter, A.J.
van Dillen, K.P. de Jong, J. Phys. Chem. B 108 (2004) 11611–11619.
[13] M.K. van der Lee, J. van Dillen, J.H. Bitter, K.P. de Jong, JACS 127 (2005)
13573–13582.
[
[
14] X.-L. Du, Q.-Y. Bi, Y.-M. Liu, Y. Cao, H.-Y. He, K.-N. Fan, Green Chem. 14 (2012)
35–939.
9
researchers. The higher selectivity of the Nb O5 promoted cata-
2
[15] G. Sengupta, D.K. Gupta, M.L. Kundu, S.P. Sen, J. Catal. 67 (1981) 223–225.
[16] T.J. Osinga, B.G. Linsen, W.P. van Beek, J. Catal. 7 (1967) 277–279.
17] B. Dvo rˇ ák, J. Pa sˇ ek, J. Catal. 18 (1970) 108–114.
18] J.W. Evans, M.S. Wainwright, A.J. Bridgewater, D.J. Young, Appl. Catal. 7 (1983)
75–83.
lyst could be due to the role of Nb O5 to hold more metallic Cu in
2
[
[
promoted CZC catalyst. This can also be justified by increasing SCu
with the introduction of Nb O5 to the parent CZC catalyst. Similar
2
observations were reported by Yoshihara and Cambell for methanol
synthesis over Cu/ZnO catalysts [51].
[19] G. Bonura, M. Cordaro, C. Cannilla, F. Arena, F. Frusteri, Appl. Catal. B 152–153
2014) 152–161.
(
[
20] J. Agrell, M. Boutonnet, I. Melián-Cabrera, J.L.G. Fierro, Appl. Catal. A 253
(2003) 201–211.
To study structure sensitivity behavior of the CO2 hydrogena-
tion, a graph was plotted between copper particles size (dCu) and
TOF of methanol formation. As indicated by Fig. 11, a sharp decline
was observed in magnitude of TOF with increasing dCu from 5
to 6 nm. Nevertheless, it remained almost constant with further
growth of dCu. This is because of the small variation of dCu as a
[21] C. Sievers, O. Jiménez, R. Knapp, X. Lin, T.E. Müller, A. Türler, B. Wierczinski,
J.A. Lercher, J. Mol Catal. A: Chem. 279 (2008) 187–199.
[
22] H. Ahouari, A. Soualah, A. Le Valant, L. Pinard, P. Magnoux, Y. Pouilloux, React.
Kinet. Catal. Lett. 110 (2013) 131–145.
[23] Q. Sun, Y.-L. Zhang, H.-Y. Chen, J.-F. Deng, D. Wu, S.-Y. Chen, J. Catal. 167
(1997) 92–105.
[
24] K. Rajaram, E. Savarimuthu, S. Arumugam, Chem. Sci. Rev. Lett. 2 (2014)
26–629.
25] W.P. Dow, T.J. Huang, J. Catal. 147 (1994) 322–332.
consequent of Nb O5 addition.
2
6
[
4
. Conclusion
CZC catalysts with different concentrations of Nb O5 were syn-
[26] F.E. López-Suárez, A. Bueno-López, M.J. Illán-Gómez, Appl. Catal. B 84 (2008)
51–658.
6
[
27] M.N.S.C. Roma, D.S. Cunha, G.M. Cruz, A.J.G. Cobo, Braz. J. Chem. Eng. 17
(2000) 937–946.
2
thesized to study the promoting effect of Nb. The addition of Nb O5
significantly modified the XRD spectra of the CZC catalyst. Increas-
[28] X.-m.Z. Gong-Xin Qi, Jin-hua Fei, Zhao-yin Hou, Indian J. Chem. 40A (2001)
88–593.
2
5
[
29] V.P.X. Courtois, M. Primet, G. Bergeret, Stud. Surf. Sci. Catal. 130 (2000)
1031–1036.
ing the Nb O5 content, increased the intensity of the CuO peak
2
which reflect the growth of the Cu crystal size. When niobium oxide
was incorporated into the CZC catalyst, the small diffraction peak
around 31 suggested that the zirconia crystal was transformed to
[30] F.d.O. Cantão, W.d.C. Melo, L.C.A. Oliveira, A.R. Passos, A.C.d. Silva, Quim. Nova
3 (2010) 528–531.
3
[
[
31] M. Ziolek, Catal. Today 78 (2003) 47–64.
◦
32] C.G. Alonso, A.C. Furtado, M.P. Cantao, O.A.A. Dos Santos, N.R.C.
Fernandes-Machado, Int. J. Hydrogen Energy 34 (2009) 3333–3341.
33] S. Esposito, M. Turco, G. Bagnasco, C. Cammarano, P. Pernice, A. Aronne, Appl.
Catal. A 372 (2010) 48–57.
the amorphous phase at 0.4 to 1.2 wt.% Nb O5 loading. In addition,
2
[
the shift of TPR peak ␣ to lower temperature with Nb O5 load-
2
ing suggests that incorporation of Nb O5 facilitated the reduction
2
[34] P.H. Matter, D.J. Braden, U.S. Ozkan, J. Catal. 223 (2004) 340–351.
of dispersed Cu. Physicochemical and activity data revealed CZC
[35] L.C. Wang, Q. Liu, M. Chen, Y.M. Liu, Y. Cao, H.Y. He, K.N. Fan, J. Phys. Chem. C
111 (2007) 16549–16557.
catalysts with 0.8 wt.% Nb O5 content is the most efficient cata-
2
[
36] C.G. Alonso, A.C. Furtado, M.P. Cantão, O.A. Andreo dos Santos, N.R. Camargo
Fernandes-Machado, Int. J. Hydrogen Energy 34 (2009) 3333–3341.
lyst in the hydrogenation of CO2 to methanol in a slurry reactor.
High methanol activity up to 40 g/kg.cat.h can be achieved at low
temperature and total pressure of 453 K and 30 bar, respectively.
[37] F.B. Noronha, M. Schmal, M. Primet, R. Frety, Appl. Catal. 78 (1991) 125–139.
[38] M.P. Maia, M.A. Rodrigues, F.B. Passos, Catal. Today 123 (2007) 171–176.
[
39] C.E.M. Guarido, D.V. Cesar, M.M.V.M. Souza, M. Schmal, Catal. Today 142
2009) 252–257.
[40] Y.-W. Suh, S.-H. Moon, H.-K. Rhee, Catal. Today 63 (2000) 447–452.
(
Acknowledgements
´
[
[
[
[
[
[
41] R. Grabowski, J. Słoczyn s´ ki, M. Sliwa, D. Mucha, R.P. Socha, M. Lachowska, J.
This work was supported by Ministry of Science, Technology
and Innovation (MOSTI) and Government of Malaysia under the
Research Project No. 03-02-02-SF0084, Ministry of Higher Educa-
tion Malaysia, FRGS No: FRGS/1/2011/SG/UTP/02/13 and Universiti
Teknologi PETRONAS, Malaysia.
Skrzypek, ACS Catal. 1 (2011) 266–278.
42] N. Özer, T. Barreto, T. Büyüklimanli, C.M. Lampert, Sol. Energy Mater. Sol. Cells
36 (1995) 433–443.
43] C. Zhou, Y. Zhao, L. Shang, Y. Cao, L.-Z. Wu, C.-H. Tung, T. Zhang, Chem.
Commun. 50 (2014) 9554–9556.
44] G. Zhang, X. Zou, J. Gong, F. He, H. Zhang, S. Ouyang, H. Liu, Q. Zhang, Y. Liu, X.
Yang, B. Hu, J. Mol. Catal. A: Chem. 255 (2006) 109–116.
45] J. Słoczy n´ ski, R. Grabowski, A. Kozłowska, P. Olszewski, J. Stoch, J. Skrzypek,
M. Lachowska, Appl. Catal. A 278 (2004) 11–23.
46] X.-M. Liu, G.Q. Lu, Z.-F. Yan, Appl. Catal. A 279 (2005) 241–245.
References
[
[
[
[
1] W. Cai, P.R. de la Piscina, J. Toyir, N. Homs, Catal. Today 242 (Part A) (2015)
93–199.
2] G. Bonura, F. Arena, G. Mezzatesta, C. Cannilla, L. Spadaro, F. Frusteri, Catal.
Today 171 (2011) 251–256.
[47] X.-L. Liang, X. Dong, G.-D. Lin, H.-B. Zhang, Appl. Catal. B: Environ. 88 (2009)
315–322.
[48] F.B. Passos, D.A.G. Aranda, R.R. Soares, M. Schmal, Catal. Today 43 (1998)
3–9.
[49] H. Ahouari, A. Soualah, A. Le Valant, L. Pinard, P. Magnoux, Y. Pouilloux, React.
Kinet. Mech. Catal. 110 (2013) 131–145.
[50] H. Berndt, V. Briehn, S. Evert, Appl. Catal. A 86 (1992) 65–69.
[51] J. Yoshihara, C.T. Campbell, J. Catal. 161 (1996) 776–782.
1
3] F. Arena, K. Barbera, G. Italiano, G. Bonura, L. Spadaro, F. Frusteri, J. Catal. 249
(
2007) 185–194.
4] C. Yang, Z. Ma, N. Zhao, W. Wei, T. Hu, Y. Sun, Catal. Today 115 (2006)
22–227.
2
Please cite this article in press as: I.U. Din, et al., Influence of niobium on carbon nanofibres based Cu/ZrO2 catalysts for liquid phase
hydrogenation of CO2 to methanol, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.06.019