A study on methanol steam reforming to CO2 and H2 over the La2CuO4 nanofiber catalyst

Article abstract of DOI:10.1016/j.jssc.2007.10.029

The La₂CuO₄ crystal nanofibers were prepared by using single-walled carbon nanotubes as templates under mild hydrothermal conditions. The steam reforming of methanol (SRM) to CO₂ and H₂ over such nanofiber catalysts was studied. At the low temperature of 150 °C and steam/methanol=1.3, methanol was completely (100%, 13.8 g/h g catalyst) converted to hydrogen and CO₂ without the generation of CO. Within the 60 h catalyst lifespan test, methanol conversion was maintained at 98.6% (13.6 g/h g catalyst) and with 100% CO₂ selectivity. In the meantime, for distinguishing the advantage of nanoscale catalyst, the La₂CuO₄ bulk powder was prepared and tested for the SRM reaction for comparison. Compared with the La₂CuO₄ nanofiber, the bulk powder La₂CuO₄ showed worse catalytic activity for the SRM reaction. The 100% conversion of methanol was achieved at the temperature of 400 °C, with the products being H₂ and CO₂ together with CO. The catalytic activity in terms of methanol conversion dropped to 88.7% (12.2 g/h g catalyst) in 60 h. The reduction temperature for nanofiber La₂CuO₄ was much lower than that for the La₂CuO₄ bulk powder. The nanofibers were of higher specific surface area (105.0 m²/g), metal copper area and copper dispersion. The in situ FTIR and EPR experiments were employed to study the catalysts and catalytic process. In the nanofiber catalyst, there were oxygen vacancies. H₂-reduction resulted in the generation of trapped electrons [e] on the vacancy sites. Over the nanofiber catalyst, the intermediate H₂CO/HCO was stable and was reformed to CO₂ and H₂ by steam rather than being decomposed directly to CO and H₂. Over the bulk counterpart, apart from the direct decomposition of H₂CO/HCO to CO and H₂, the intermediate H₂COO might go through two decomposition ways: H₂COO=CO+H₂O and H₂COO=CO₂+H₂.

Full text of DOI:10.1016/j.jssc.2007.10.029

ARTICLE IN PRESS
Journal of Solid State Chemistry 181 (2008) 7–13
www.elsevier.com/locate/jssc
A study on methanol steam reforming to CO and H over
2
2
the La CuO nanofiber catalyst
2
4
a,
Ã
b
Lizhen Gao , Gebiao Sun , Sibudjing Kawi
b
a
School of Mechanical Engineering M050, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
b
Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, Singapore 117576, Republic of Singapore
Received 6 July 2007; received in revised form 15 October 2007; accepted 30 October 2007
Available online 5 November 2007
Abstract
2 4
The La CuO crystal nanofibers were prepared by using single-walled carbon nanotubes as templates under mild hydrothermal
conditions. The steam reforming of methanol (SRM) to CO and H over such nanofiber catalysts was studied. At the low temperature of
2
2
1
50 1C and steam/methanol ¼ 1.3, methanol was completely (100%, 13.8 g/h g catalyst) converted to hydrogen and CO
generation of CO. Within the 60 h catalyst lifespan test, methanol conversion was maintained at 98.6% (13.6 g/h g catalyst) and with
00% CO selectivity. In the meantime, for distinguishing the advantage of nanoscale catalyst, the La CuO bulk powder was prepared
and tested for the SRM reaction for comparison. Compared with the La CuO nanofiber, the bulk powder La CuO showed worse
2
without the
1
2
2
4
2
4
2
4
catalytic activity for the SRM reaction. The 100% conversion of methanol was achieved at the temperature of 400 1C, with the products
being H and CO together with CO. The catalytic activity in terms of methanol conversion dropped to 88.7% (12.2 g/h g catalyst) in
0 h. The reduction temperature for nanofiber La CuO was much lower than that for the La CuO bulk powder. The nanofibers were
2
2
6
2
4
2
4
2
of higher specific surface area (105.0 m /g), metal copper area and copper dispersion. The in situ FTIR and EPR experiments
were employed to study the catalysts and catalytic process. In the nanofiber catalyst, there were oxygen vacancies. H -reduction resulted
in the generation of trapped electrons [e] on the vacancy sites. Over the nanofiber catalyst, the intermediate H CO/HCO was stable and
2
2
was reformed to CO
direct decomposition of H
H COO ¼ CO+H O and H COO ¼ CO +H .
2
and H
2
by steam rather than being decomposed directly to CO and H
2
. Over the bulk counterpart, apart from the
2
CO/HCO to CO and H , the intermediate H COO might go through two decomposition ways:
2
2
2
2
2
2
2
r 2007 Elsevier Inc. All rights reserved.
2 4 2
Keywords: La CuO nanofiber; Methanol steam reforming; CO selectivity; In situ FTIR
1
. Introduction
Hydrogen has been envisaged to be the ideal clean
(ii) partial oxidation of methanol (CH OH+1/2O =CO
3 2 2
(or CO)+2H ) and (iii) steam reforming of methanol
(SRM) (CH OH+H O=CO +3H ) [7]. Amongst the
2
3
2
2
2
energy in the future [1,2]. The difficulty in wide usage of
hydrogen is its storage and distribution [3,4]. Before a
highly effective hydrogen storage material (more than
three ways, (i) and (ii) will produce a considerable amount
of CO as a product or by-product. In the hydrogen fuel
cell, even a trace of CO can deteriorate the Pt electrode
[5,6,8]. The SRM process produces H and CO with the
6
.5 wt% hydrogen can be stored and released under a mild
2
2
condition) is discovered, the on-board generation of
hydrogen from liquid fuel such as methanol with high
energy density will be an option [5,6]. Through three
ways methanol can be transformed to hydrogen, i.e. (i) the
direct decomposition of methanol (CH OH=CO+2H );
molar ratio of 3/1 (the highest amongst the three ways) as
well as a little CO as a by-product [7], so it has been
regarded as the most suitable on-board process to generate
hydrogen. So far the most extensively studied catalysts for
SRM are Cu/ZnO/Al O [9,10], Cu/ZrO [11], Al–Cu–Fe
3
2
2
3
2
[
12], etc., which are also the methanol synthesis catalysts
Ã
from syngas. The problems we have to overcome are (i) the
high reaction temperature to obtain the high conversion of
Corresponding author. Fax: +61 8 64881024.
E-mail address: lizhen@mech.uwa.edu.au (L. Gao).
0022-4596/$ - see front matter r 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2007.10.029

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
8
methanol. Even over the most effective catalyst, 100%
methanol conversion can only be achieved at temperatures
above 300 1C [7,13]; (ii) the formation of CO at the high
conversion of methanol at high temperatures. Apart from
the Cu–Zn–Al series catalysts, the perovskite-structured
complex cuprites (such as YBa Cu O ) have been demon-
the sample was heated (heating rate: 51/min) to 700 1C and
kept at 700 1C for 2 h. The effluent was monitored by GC.
The CNT oxidation caused CO2 formation. When we
could not observe CO formation, we believed that the
2
CNTs have been completely oxidized.
The SRM reaction was carried out on a fixed bed flow
reactor at atmospheric pressure by using 100 mg of
catalyst. The catalyst was reduced at 500 1C for 1 h in the
flow of H /N (5/95, v/v; flow rate ¼ 35 ml/min) prior to
2
3
7
strated to be good catalysts for methanol synthesis from
syngas [14]. Recently we have successfully synthesized
perovskite-like La CuO4 single-crystal nanofibers (ca.
2
2
2
30 nm in diameter and 3 mm in length) by using single-
walled carbon nanotubes (SWNTs; ca. 2 nm in inner
SRM reaction. The feed gas composition was MeOH/
H O ¼ 1/1.3 (molar ratio). Flow rate of reagent was
2
diameter; made via CH4 cracking over the catalyst of
Mg0.8Mo0.05Ni0.10Co0.05O at 800 1C) as templates under
0.04 ml/min (liquid). The products were analyzed online
by GC.
x
mild hydrothermal conditions and a temperature around
6
The contents of copper in different oxidation states were
estimated by means of iodometry according to the
procedures adopted by Harris and Hewston [16]. The
oxygen non-stoichiometry values were calculated from
0 1C [15]. Here we report our results that such La CuO4
2
nanofibers showed an excellent catalytic performance for
SRM reaction. The 100% methanol conversion could be
reached at the low temperature of 150 1C. There was no
significant drop in activity within 60 h reaction test on
stream. No CO was created below the temperature of
2
+
+
3+
the amount of Cu , Cu or Cu present, assuming that
3
+
the La was in its stable oxidation state [17]. The copper
surface area and dispersion were measured using a nitrous
oxide titration [18]. Prior to N O surface titration, samples
3
of the catalyst, the La CuO bulk powder counterpart was
00 1C. To distinguish the advantage of the nanofiber state
2
were reduced in a 5% H /He stream at 500 1C, followed by
2
4
2
prepared and tested for comparison as well. Techniques
such as TGA, in situ FTIR (diffuse reflectance infrared
Fourier transform spectroscopy—DRIFT) and EPR were
employed to study the catalysts and the SRM mechanism.
cooling to 150 1C under a flow of helium. A known volume
of N O was then injected in pulses by using a 6-port valve,
and the N and N O in the effluent were analyzed by a
2
2
2
GC–MS system. The copper metallic surface area and
1
9
dispersion were calculated by assuming 1.46 ꢁ 10 copper
2
2
. Experimental section
atoms/m and a molar stoichiometry N O/Cu ¼ 0.5,
2
s
where Cu implies the copper atom on the surface.
s
Detailed descriptions about the synthesis of SWNTs and
La CuO nanofibers by using SWNTs as templates were
The morphologies of La CuO nanofibers and bulk
2 4
powder were observed under a transmission electron
microscope (JEOL, JEM 2010) and a field emission
scanning electron microscope (FESEM) (JEOL, JSM
7600F). The thermogravimetric (TG) curves in hydrogen
atmosphere were obtained on a Shimadzu DTG-60 thermal
analysis instrument.
2
4
described in our previous publication [15]. Briefly, the
SWNTs were home-made by cracking of CH (CH /H /
4
4
2
He ¼ 1/1/8) at 800 1C over
a mixed-oxide catalyst
Mg Mo Ni0.10Co0.05O . The SWNT sample was pur-
0
.8
0.05
x
ified by nitric-acid washing repeatedly in an ultrasonic
bath. The carbon nanotubes (CNTs) were single walled
with a 2 nm in-average inner diameter. For hydrothermal
synthesis of La CuO single-crystal nanofibers by using
EPR spectra were recorded at ꢂ196 1C with a JEOL
spectrometer operating in the x-band and calibrated with a
DPPH standard (g ¼ 2.004). About 0.2 g of catalyst was
placed in a self-made quartz cell in which the sample could
be treated under different atmospheres at various tempera-
tures. Before performing the EPR studies over the samples,
the sample was He-purged (flow rate, 20 ml/min) at
25 1C for 1 h; then H (H /He ¼ 5/95, total flow rate ¼
2
4
SWNTs as templates, the mixed solution of the surfactant
poly(ethylene glycol)-block-poly(propylene glycol)-block-
poly(ethylene glycol), La(NO ) ꢀ 6H O and Cu(NO ) ꢀ
3
3
2
3 2
6
H O (according to the stoichiometric composition of
2
La CuO ), SWNTs and H O was dispersed ultrasonically
2
4
2
2
2
2
and was put into an autoclave for hydrothermal synthesis
at 60 1C for 20 h. The precipitation obtained from
hydrothermal synthesis was filtered and washed with
distilled water repeatedly and then was heated at 110 1C
for 1 h. Thus the La CuO nanofibers were synthesized.
30 ml/min) was introduced into the quartz cell at 500 1C for
1 h, followed by He-purging at the same temperature and
quenching (ꢂ196 1C) for EPR analysis.
In situ FTIR spectra (DRIFT) were collected on a
Nicolet series II magna–IR 550 spectrometer with a
SPECTRA TECH in situ cell. The catalyst powder
weighing approximately 50 mg was contained in a low-
dead volume infrared cell. The cell was heated by means of
an electrical resistance heater. Before FTIR spectrum
collection, the cell was pumped for 10 min to remove
2
4
The La CuO bulk powder catalyst was prepared by
4
2
heating the solution of La(NO ) ꢀ 6H O, Cu(NO ) ꢀ 6H O
3
3
2
3 2
2
(
molar ratio of La/Cu ¼ 2/1) and citric acid at 50 1C until
a syrup was formed, followed by heating in air at 700 1C
for 6 h.
To remove the CNTs in La CuO , the sample was put
gaseous CO2 and H . In situ absorbance spectra were
2
2
4
ꢂ1
into a quartz tube positioned in a tubular furnace. The air
flow (flow rate: 30 ml/min) was conducted into the tube and
obtained at 4 cm resolution. Methanol was injected into
the cell at the temperature of 350 1C. The spectra were then

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
9
referenced to a spectrum of the catalyst collected at the
same temperature under a H flow (H /N ¼ 5/95, v/v,
could be completely oxidized to CO2 at temperatures
around 650 1C [19,20].
2
2
2
2
0 ml/min.).
Pulse reaction was carried out on a pulse microreactor
The FESEM image of the 700 1C-treated nanofibers is
shown in Fig. 1C. The diameter of La CuO nanofibers is
around 30 nm. We deduce that the shrinkage of La CuO
2
4
system. Each pulse volume was 67.5 ml. During the reaction,
He was used as the carrier gas (flow rate, 20 ml/min). The
products were checked by a GC–MS system.
The BET specific surface area of the catalysts was
determined by nitrogen adsorption data at ꢂ196 1C on a
Nova 1200 system.
2
4
nanofibers is due to the combustion of SWNTs inside it.
Table 1 shows the composition, BET specific surface
area, Cu metal area and dispersion of the La CuO
nanofiber and bulk powder. In the La CuO bulk powder,
2
4
2
4
3
+
2+
there are Cu and Cu while in the nanofibers there are
2
+
+
Cu and Cu . In the bulk powder sample, the oxygen is
over-stoichiometric, while in the nanofiber counterpart,
there are the oxygen vacancies. The BET specific surface
areas of La CuO nanofibers (treated at 700 1C in air) and
3
. Results and discussion
The FESEM images of as-synthesized La CuO nanofi-
2
4
2
4
bers are shown in Fig. 1A in which we observe groups of
aligned microfibers. The length of the fiber is close to 3 mm.
Fig. 1B is the enlarged FESEM image for the individual
fibers. The average diameter of the fibers is around 60 nm.
XRD result has confirmed that these single-crystal fibers
are of La CuO crystal structure [15]. We are not sure
Table 1
Composition, BET specific surface area, Cu-metal area, Cu dispersion of
La CuO nanofiber and bulk powder
2
4
2
4
Catalyst
Composition BET surface Cu-metal
2
are (m /g)
Cu dispersion
2
area (m /g) (%)
whether or not there are SWNTs inside the La CuO
2
4
4
nanofibers. So we have treated the as-synthesized La CuO
2
2+
Cu0.88
Nanofibers
La
2
105.0
2.7
10.8
1.5
40.21
nanofibers in air at 700 1C for 2 h for further purification
before catalytic activity measurement. At 700 1C in air, the
CNTs would be burnt out completely as we cannot observe
+
Cu0.12O
3.94
2
+
Bulk powder La
2
Cu0.92
3+
5.78
Cu0.08O
4.04
any more CO formation. We have found that the CNTs
2
2 4
Fig. 1. FESEM images of La CuO single-crystal nanofibers (hydrothermal synthesis for 20 h). (A) Groups of highly oriented and uniform-diameter
fibers; (B) an enlarged image of some individual fibers; (C) after being treated in air at 700 1C for 1 h; (D) after being tested for SRM reaction lifespan
for 60 h.

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
10
Table 2
Activity and selectivity for steam reforming of methanol over nanofiber La
2
CuO
4
and bulk powder La
2
CuO
4
Catalyst
Reaction
temperature (1C)
Conv. of
MeOH (%)
Conv. of
MeOH
Selectivity (%)
Rate of H
2
production
(g/h g catalyst)
(
g/h g catalyst)
CO
2
CO
La
2
CuO
4
nanofibers
100
52.4
100.0
100.0
100.0
100.0
100.0
100.0
7.23
100.00
100.00
100.00
100.00
100.00
99.98
0
1.11
2.11
2.11
2.11
2.11
2.09
2.05
1
2
2
3
3
4
50
00
50
00
50
00
13.8
13.8
13.8
13.8
13.8
13.8
0
0
0
0
0.02
0.05
99.95
Bulk powder La
2
CuO
4
200
6.7
11.2
0.92
98.64
98.02
97.45
97.00
96.52
1.36
1.98
2.55
3.00
3.45
0.13
0.23
0.34
1.06
2.03
2
3
3
4
50
00
50
00
1.55
2.26
7.13
16.4
51.7
100.0
13.8
2
bulk powder are 105.0 and 2.7 m /g, respectively. So far the
highest specific area reported for La CuO catalyst is
1
102
2
4
100
2
3.0 m /g prepared by sol–gel method [21]. The copper
metal area and dispersion over the nanofibers are both
much higher than those over the bulk powder. The higher
dispersion of copper metal is beneficial for the methanol
steam reforming reaction.
98
9
9
6
4
Nanofiber
bulk powder
The catalytic activities of the La CuO nanofiber and
2
4
92
bulk powder catalysts for the SRM reaction are summar-
ized in Table 2. Over the nanofiber catalyst, at 100 1C, the
methanol conversion is 52.4% (7.23 g/h g catalyst). When
the temperature is elevated to 150 1C, 100% methanol is
converted (13.8 g/h g catalyst). At temperatures below
9
0
88
0
20
40
Reaction time (hour)
60
80
3
In SRM reaction, CO is generated by the reverse water–
00 1C, the products are only H and CO without CO.
2
2
Fig. 2. Methanol conversions as a function of time-on-stream over (E)
nanofiber La CuO catalyst at 250 1C and (m) bulk powder La CuO
catalyst at 400 1C.
2
4
2
4
gas-shift reaction: CO +H ¼ CO+H O. This means that
2
2
2
the nanofiber catalyst can suppress this reaction. When the
reaction temperatures are at 350 and 400 1C, only 0.02%
CO (CO /CO ¼ 99.98/0.02) and 0.05% CO (CO /CO ¼
finally drops down to 98.6% (13.6 g/h g catalyst) in the 60th
hour without CO formation. The CO selectivity is still
2
2
9
9.95/0.05) are detected, respectively. Over the bulk
2
powder La CuO catalyst, below 200 1C, little methanol
100%. Over the powder catalyst, in the beginning 17 h,
there is no activity drop, while in the following 43 h, the
activity declines to 88.7% (12.2 g/h g catalyst) of methanol
conversion with the CO selectivity of 4.2%. So in terms of
the lifespan, the nanofiber catalyst is also superior to the
bulk powder. With respect to the fact that the lifetime test
for the nanofibers is conducted at lower (250 1C) tempera-
ture than that (400 1C) for the bulk powder, we have
also examined the methanol conversion of La CuO
2
4
has been reformed to hydrogen and either CO or CO . At
2
temperatures of 200, 250, 300 and 350 1C, the methanol
conversion is 6.7% (0.92 g/h g catalyst), 11.2% (1.55 g/h g
catalyst), 16.4% (2.26 g/h g catalyst) and 51.7% (7.13 g/h g
catalyst), respectively. At the temperature of 400 1C, 100%
methanol (13.8 g/h g catalyst) can be reformed. In all the
tested temperatures, the products are H , CO and CO.
2
2
The CO selectivity is positively dependent on the tempera-
tures (200 1C: 1.36%; 400 1C: 3.45%). Obviously the
catalytic activity over nanofiber La CuO catalyst is much
2
4
nanofibers at 400 1C for 60 h and found that even at such
conditions, the methanol conversion was maintained at
98.0% (13.5 g/h g catalyst). Even using the Pd-based
methanol-reforming catalysts and microstructured reac-
tors, 85% methanol conversion was achieved at 310 1C
(steam/ethanol ¼ 1.9); CO concentration was 0.2–0.5%
[23]. Over the advanced plate-fin reformer, the device
incorporates methanol steam reforming (Cu/Zn/Al O )
2
4
better than that over its powder counterpart. CO concen-
tration increased with the rise of temperature [22].
The lifetimes of both the nanofiber (at 250 1C) and the
bulk powder (at 400 1C) catalysts are tested within 60 h and
the results are shown in Fig. 2. Over the nanofiber catalyst,
in the beginning 23 h, no activity drop can be observed and
afterwards the methanol conversion slightly decreases and
2
3
with catalytic combustion (Pt/Al O ); 100% methanol
2
3

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
11
conversion was obtained over 100 h at temperatures of
10–270 1C (steam/methanol ¼ 1.2–1.6). But the CO con-
centration steadily increased over this time from 0.4% to
about 100 La CuO molecule units. Such 1D nanofibers
2 4
2
are easier to be reduced than the bulk oxides. It is
reasonable as in the 1D La CuO sample there are more
2
4
+
1
.2% [24].
Fig. 1D is the FESEM image of a La CuO nanofiber
copper ions (may be Cu ) with lower valence (lower
oxygen coordinated) than those in the bulk powder
La CuO . The composition listed in Table 1 demonstrates
2
4
after being tested for catalytic activity lifespan for 60 h. No
significant sintering is observed from Fig. 1D, indicating
that the nanofiber catalyst is still quite fresh. The BET
2
4
2
+
+
that in the nanofibers there are Cu and Cu , while in the
+
bulk powder there are Cu
3
2+
and Cu . An agreement
specific surface area of the lifespan-tested nanofiber is
2
7.3 m /g, a little decrease in comparison with the original
has already been established for the methanol synthesis
from syngas and its reverse reaction—SRM over cupric
9
value of 105 m /g, while the specific surface area of the
2
+
o
o
oxides catalysts: (i) Cu /Cu (most likely Cu ) are the key
2
lifespan-tested La CuO bulk powder is only 0.9 m /g. The
+
o
catalytic active sites and (ii) the highly dispersed Cu /Cu
is beneficial to the catalytic reaction. There is an optimum
2
4
biggest problem with the Cu-based catalyst is the tendency
for copper crystallites to readily sinter at the temperatures
o
+
balance between metallic Cu and oxidized Cu for
maximum activity/selectivity and this is a function of not
only the catalyst preparation and composition but also the
feed and reaction conditions [23]. The La CuO nanofiber
4
300 1C [25]. On the other hand, the coke generated by the
side reactions of SRM may deactivate the powder catalyst.
The energy-dispersive X-ray analysis reveals that there is a
carbon element on the surface of the used powder catalyst.
Both the methanol synthesis reaction and the SRM
reaction over cupric catalysts follow a redox mechanism
2
4
fulfills these two requirements. It is easier to be reduced or,
2
in other words, the reduction of Cu /Cu in La CuO
+
+
2
4
nanofibers can happen at a lower temperature than that of
3
+
2+
[
La CuO nanofiber and powder. The TG curves of
14,26]. Therefore we studied the reducibility of the
Cu /Cu in La CuO bulk powder. After reduction, the
2 4
copper ions and metal copper can be highly dispersed on
the large nanofiber surface. The N O titration results listed
2
4
La CuO nanofiber and bulk powder in hydrogen atmo-
2
4
2
sphere are shown in Fig. 3A and B. Both the nanofiber and
powder La CuO catalysts show one weight-loss step. The
weight loss is due to the reduction of La CuO , i.e.
in Table 1 indicate that the metal copper area and
dispersion in the H -reduced La CuO nanofiber are both
higher than those in the H -reduced La CuO bulk powder.
2 2 4
2
4
2
2
4
2
4
The EPR spectra of the H -reduced bulk powder and
2
Cu þðCu3þÞ þ e ! Cu ðor Cu Þ.
2
þ
o
nanofiber La CuO samples as depicted in Fig. 4 exhibit a
2
4
strong anisotropic signal with extremes at g ¼ 2.232, 2.130
The nanofibers start to be reduced at the temperature of ca.
30 1C, whereas the bulk powder begins reduction at ca.
00 1C. For the nanofiber catalyst, about 2.2 wt% weight
loss is observed, while for the bulk one, the weight loss is
about 0.9%. These results indicate that the one-dimen-
sional nanofibers of La CuO are much easier to be
reduced than the bulk powder counterpart. The diameter
of fibers in Fig. 1C is around 30 nm. The La CuO lattice
2+
and 2.053; these are typical features of Cu species; the line
2
5
broadening may be from life time (spin–orbital interactions)
and spin–spin dipole interactions. These results indicate that
2+
after reduction, there is still certain amount of Cu ions in
both the bulk powder and nanofiber La CuO samples. The
2
4
2
4
2+
intensity of Cu signal in the reduced nanofiber La CuO is
2
4
weaker than that in the bulk powder La CuO sample,
2
4
2
4
indicating that nanofibers have been deeper reduced than the
bulk powder. A signal at g ¼ 2.007 appears in the reduced
sample of nanofiber La CuO ; this signal can be attributed to
parameters are a ¼ b ¼ 0.53 nm, c ¼ 1.30 nm [27]. So we
imagine that in the cross-section of the cubic fiber there are
2
4
e trapped in the oxygen vacancies [28]. The mechanism for
the generation of trapped electrons could be
H2 ! 2H;
H þ O ꢂ ! HO þ ½e ꢃ,
2
ꢂ
ꢂ
Fig. 3. Thermogravimetric (TG) curves of (A) nanofiber La
bulk powder La CuO in a H flow.
2
CuO
4
and (B)
Fig. 4. EPR spectra of H
powder.
2 2 4
-reduced La CuO (a) nanofibers and (b) bulk
2
4
2

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
12
ꢂ
are detected. CO2 was generated in both the La CuO
2
4
nanofibers sample and the La CuO bulk powder sample.
2
4
The CO species are only detected over the bulk powder
La CuO sample. The HCO/H CO is only detected in
the nanofiber sample. Without the H O steam, the decom-
2
4
2
2
position of HCOO/H COO could produce CO and H .
2
2
2
HCO/H CO may react with the lattice oxygen or other
2
surface oxygen species such as hydroxyl in the catalyst to
produce CO . As it is difficult to carry out the in situ FTIR
2
experiment of water steam adsorption, we perform the
methanol/H O pulse reaction over the nanofiber and bulk
2
powder samples at 3001C, respectively. After adsorption of
methanol, water steam is pulsed. Over the nanofiber catalyst,
only CO is generated. Over the bulk powder sample, CO
2
and CO2 are co-produced. We deduce that over the
nanofibers
H O þ H CO ! CO þ 2H ,
2
2
2
2
Fig. 5. In situ FTIR (DRIFT) spectra of methanol adsorption over the
-reduced La CuO (a) nanofibers and (b) bulk powder.
H
2
2
4
H2O þ HCO ! CO2 þ 3=2H2.
Normally the HCO or H CO is rapidly decarbonylated to
2
CO and H , and then partially transformed to CO and H
2
Table 3
Assignments of FTIR absorbance bands
2
2
through the secondary WGS reaction [36,37]. Over nanofi-
ber, we cannot observe the adsorbed CO vibration band, but
we find HCO/H CO. The oxygen vacancy with the trapped
Samples
Wavenumber
ꢂ1
Assignment
2
(cm
)
electron in the nanofiber catalyst may stabilize the HCO and
H CO and prevent them from decomposing to CO and H
Nanofibers
2950
1771, 1730
CH
CH
3
O
(a) (methoxide) [29,30]
2
2
2
O(a) (formaldehyde), HCO(a)
directly. In the spectrum of methanol adsorption over the
La CuO bulk powder, we cannot detect the H CO/HCO
species, but we find the CH OO and adsorbed CO species.
(
CO
formyl) [30]
ꢂ
2
4
2
1695
2
(antisymmetric stretch) [31]
2
Bulk powder
2956
CH (a) (methoxide) [29,30]
3
O
We assume that two possibilities may occur: firstly, the
H CO/HCO is extremely unstable and has already decom-
2870
b-CH OO(a) (bidentate
2
2
methylenebisoxy) [32]
CO(a)–Cu (linear adsorbed CO on
posed to CO and H ; secondly, the CH OO species may go
2
2085
2
o
Cu ) [33]
through the two decomposition ways:
ꢂ
1
695
600
CO
2
(antisymmetric stretch) [31]
1
H
2
O
(a), b-HCO3 (a) (bidentate
H COO ! CO þ H O;
2
2
bicarbonate) [34]
m-CO3(a) (monodentate carbonate)
1372
H2COO ! CO2 þ H2.
[35]
In the FTIR spectrum, the adsorbed H O is detected.
2
ꢂ
wherein [e ] represents a trapped electron located at an
4. Conclusion
2+
oxygen vacancy. Cu
+
can be reduced to Cu or Cu by
o
picking up a trapped electron. This small signal may also be
caused by the carbon impurities in the nanofibers. On the
La CuO bulk powder sample, no such trapped electrons
(i) The La CuO nanofiber is prepared by using CNTs as
2
4
templates. Compared with the La CuO bulk powder
2
4
sample, the nanofiber is of higher specific surface area.
3
2
4
+
The La CuO bulk powder is composed of Cu , Cu
2 4
2+
have been generated. Fig. 5 shows the in situ FTIR (DRIFT)
spectra of methanol adsorption over the bulk powder and
nanofiber La CuO catalysts. Table 3 lists the in situ FTIR
and extra-oxygen species while the La CuO nanofiber
2
4
2
+
+
is composed of Cu , Cu and oxygen vacancies. The
nanofiber can be reduced at lower temperature and to a
farther extent than the bulk powder counterpart. After
2
4
(DRIFT) spectra adsorption bands and their assignments
[29–35]. Over the nanofiber La CuO , the bands attributable
to CH O (2950 cm ), HCO/H CO (1771, 1730 cm ) and
2
4
ꢂ1
ꢂ1
o
being reduced, the Cu-metal area and Cu dispersion in
the nanofibers are higher than those in the bulk
powder.
3
2
ꢂ
ꢂ1
CO (1695 cm ) are observed. Over the bulk powder
2
La CuO sample, the bands attributable to CH O
2
4
ꢂ1
ꢂ1
3
ꢂ1
ꢂ1
ꢂ
(
(
2956 cm ), CH OO (2870 cm ), CO (2085 cm ), CO
2
(ii) For the reaction of SRM to produce CO2 and
hydrogen, the La CuO nanofiber shows a much better
2
ꢂ1
ꢂ1
1695 cm ), H O/HCO (1600cm ) and CO (1372 cm
3
)
2
3
2
4

ARTICLE IN PRESS
L. Gao et al. / Journal of Solid State Chemistry 181 (2008) 7–13
13
catalytic performance in terms of methanol conversion,
CO2 selectivity and catalyst lifespan. When the
temperature is 150 1C, 100% methanol is converted
[13] C.Z. Yao, L.C. Wang, Y.M. Liu, G.S. Wu, Y. Cao, W.L. Dai,
H.Y. He, K.N. Fan, Appl. Catal. A 297 (2006) 151.
[
[
14] L.Z. Gao, C.T. Au, J. Catal. 189 (2000) 1.
15] L.Z. Gao, X.L. Wang, H.T. Chua, S. Kawi, J. Solid State Chem. 179
(
products are only H and CO without CO. Over the
13.8 g/h g catalyst). At temperatures below 300 1C, the
(
2006) 2044.
2
2
[16] D.C. Harris, T.A. Hewston, J. Solid State Chem. 69 (1987) 182.
[17] L.Z. Gao, C.T. Au, J. Mol. Catal. A 168 (2001) 173.
nanofiber catalyst, the oxygen vacancy together with
the trapped electron may stabilize the intermediates of
HCO and H CO and prevent them from decomposing
[18] J.T. Li, W.D. Zhang, L.Z. Gao, P.Y. Gu, K.Q. Sha, H.L. Wan, Appl.
Catal. A 165 (1997) 411.
2
[19] B.C. Liu, L.Z. Gao, Q. Liang, S.H. Tang, M.Z. Qu, Z.L. Yu, Catal.
Lett. 71 (2001) 225.
to CO and H . Over the bulk catalyst, the intermediate
2
H COO may decompose to CO and H O.
2 2
[20] H. Li, Q. Li, L.Z. Gao, C.T. Au, Z.L. Yu, Catal. Lett. 74 (2001) 185.
[
[
21] N. Guilhaume, S.D. Peter, M. Primet, Appl. Catal. B 10 (1996) 325.
22] D.R. Palo, J.D. Holladay, R.A. Dagle, Y.-H. Chin, ACS Symp. Ser.
References
(
23] D.R. Palo, R.A. Dagle, J.D. Holladay, Chem. Rev. 107 (2007) 3992.
2005) 209.
[
[
[
1] Editorial, Nature 427(6976) (2004) 661.
2] M.Z. Jacobson, W.G. Colella, D. Golden, Science 308 (24) (2005)
901.
[24] L.W. Pan, S.D. Wang, Int. J. Hydrogen Energy 30 (2005) 973.
[25] M. Twigg, M. Spencer, Top. Catal. 22 (2003) 192.
[26] R.D. Cortright, R.R. Davda, J.A. Dumesic, Nature 418 (2002) 964.
[27] M.J. Akhtar, C.R.A. Catlow, S.M. Clark, W.M. Temmerman,
J. Phys. C 21 (1988) L917.
1
[
[
[
[
[
3] W. Grochala, P.P. Edwards, Chem. Rev. 104 (2004) 1283.
4] P. Chen, Z. Xiong, J. Luo, J. Lin, K.L. Tan, Nature 420 (2002) 302.
5] B.C.H. Steele, A. Heinzel, Nature 414 (2001) 345.
6] M. Schrope, Nature 414 (2001) 682.
[28] L.Z. Gao, C.T. Au, Appl. Catal. B 30 (2001) 35.
[29] G.J. Millar, C.H. Rochester, K.C. Waugh, J. Chem. Soc. Faraday
Trans. 88 (1992) 1033.
7] T. Shishido, Y. Yamamoto, H. Morioka, K. Takaki, K. Takehira,
Appl. Catal. A 263 (2004) 249.
[30] I.A. Fisher, A.T. Bell, J. Catal. 172 (1997) 222.
[31] G.J. Millar, C.H. Rochester, Top. Catal. 3 (1996) 103.
[32] E.M. Stuve, R.J. Madix, B.A. Sexton, Surf. Sci. 119 (1982) 279.
[33] D.B. Clarke, D.K. Lee, M.J. Sandoval, A.T. Bell, J. Catal. 142 (1993)
27.
[
[
8] B.A. Peppley, J.C. Amphett, L.N. Kearns, R.F. Mam, Appl. Catal.
A 179 (1999) 31.
9] J. Bravo, A. Karim, T. Conant, G.P. Lopez, A. Datye, Chem. Eng.
J. 101 (2004) 113.
[
[
[
10] S. Fukahori, T. Kitaoka, A. Tomoda, R. Suzuki, H. Wariishi, Appl.
Catal. A 300 (2006) 155.
[34] W. Hertl, Langmuir 5 (1989) 96.
[35] J.F. Edwards, J.L. Schrader, J. Phys. Chem. 88 (1984) 5620.
[36] N. Takezawa, N. Iwasa, Catal. Today 36 (1997) 45.
[37] N. Iwasa, T. Mayanagi, N. Wataru, M. Arai, T. Takewasa, Appl.
Catal. A 248 (2003) 153.
11] I. Ritzkopf, S. Vukojevic
F. Schuth, Appl. Catal. A 302 (2006) 215.
12] T. Tanabe, S. Kameoka, A.P. Tsai, Catal. Today 111 (2006) 153.
´
, C. Weidenthaler, J.D. Grunwaldt,
¨