Photoreduction of carbon dioxide with hydrogen over ZrO2
Yoshiumi Kohno,* Tsunehiro Tanaka, Takuzo Funabiki and Satohiro Yoshida
Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Sakyo-ku, Kyoto 606-01, Japan
Among many transition-metal oxides, zirconium oxide is
found to be active for photocatalytic reduction of carbon
dioxide to carbon monoxide with hydrogen in the gas
phase.
of CO was detected after 6 h photoirradiation, and no other
reduced product was detected by an FID gas chromatograph
(which can detect gaseous hydrocarbons and oxygenates) and a
mass spectrometer, as well as the on-line TCD gas chromato-
graph. ZrO
difference was found in Zr LIII edge XANES spectra of ZrO
before and after photoreaction. When the reaction was carried
out in the dark, no CO was detected, although CO is known to
be adsorbed on the surface of ZrO . This indicates that
2
is stable under photoirradiation and no appreciable
From the viewpoint of environmental problems, catalytic
reduction of carbon dioxide to more valuable compounds has
2
1–3
been investigated extensively. However, this reaction often
requires severe conditions of high pressure and/or high
temperature. In contrast, by photoirradiation, photocatalytic
reduction of carbon dioxide with hydrogen may occur even
under mild conditions.
2
2
photoirradiation is essential for this reaction. When the short
wavelength region of the irradiated light was cut off by a series
of filters which permits light with wavelength l > 370, 330,
310 and 290 nm, the amount of evolved CO after 6 h
photoirradiation changed from 1.0 mmol to 0, 0, 0.1 and
0.4 mmol, respectively. From these results, it can be concluded
that the light with wavelength < 300 nm was required to enable
this reaction to proceed.
As a photocatalyst for such types of reactions, noble metal/
titanium oxide systems are the most well known and have been
extensively investigated so far.4 Zirconium oxide has rarely
been reported to be active; however, Sayama et al. recently
reported that zirconium oxide showed excellent photocatalytic
activity for the decomposition of liquid water and photoreduc-
tion of aqueous carbonate.7 They also reported that metal
loading inactivated zirconium oxide, in contrast to the case of
titanium oxide. In the present study, we report that zirconium
oxide without metal loading is active for photocatalytic
reduction of gaseous carbon dioxide with hydrogen under mild
conditions of low pressure and room temperature. The wide
–6
2
Without introduction of CO no trace of CO was detected
,8
13
12
after UV irradiation. When CO
only 13CO was detected. These results show that CO was
produced from the CO admitted to the system.
Fig. 1 shows the time dependence of the amount of CO
formed and H consumed. The rate of CO formation mildly
decreased from 8 to ca. 40 h irradiation. On the other hand, the
consumption rate of H was very fast at the beginning, and then
slowed. The amount of consumed H was larger than the
amount of evolved CO at all times. This suggests that H may
2 2
was used instead of CO ,
2
2
band-gap of ZrO
2
(5.0 eV) permits the photoreduction of carbon
2
dioxide with hydrogen.
Zirconium oxide used in this study was prepared from an
aqueous solution of zirconium oxychloride by precipitation
2
2
9
be adsorbed onto the catalyst surface in a dissociated form or
incorporated into other compounds containing carbon atom(s).
We then tried to decompose any adsorbed species by heating the
catalyst sample after photoirradiation. More CO evolution was
with 25 mass% NH
with distilled water followed by filtration until the filtrate was
negative towards AgNO . The sample was dried at 373 K
3
(aq). The precipitate obtained was washed
3
overnight, grounded to powder, and calcined at 773 K for 5 h in
a dry air stream. The XRD pattern indicated that the resulting
powder was zirconium oxide in a mixture of monoclinic and
tetragonal phases. Other metal oxides were commercially
available ones and they were calcined at 773 K for 5 h prior to
use. Before reactions, the oxides were heated at 673 K for 30
min in the air and evacuated for 30 min at the same temperature,
2
observed, while the amount of desorbed H was small. After 6
h of photoirradiation, 1.0 mmol of CO was detected (as
mentioned above), and when the catalyst sample was heated at
673 K for 20 min after 2 min evacuation, 3.3 mmol of CO was
evolved. When the photoirradiation time was extended to
40.5 h, 3.2 mmol of CO was formed, and by heating the catalyst,
5.8 mmol of CO was evolved. This observation shows that
2
followed by treatment with 8 kPa O for 75 min and evacuation
for 30 min at 673 K.
3
2
1
0
CO formation
consumption
30
The photocatalytic reaction was carried out in a closed static
system connected to a vacuum line. A 500 W ultrahigh-pressure
mercury lamp (Ushio Denki USH-500) was used as the light
source. A metal oxide sample (0.3 g) was spread on the flat
bottom of a quartz reactor and subjected to illumination from
H
2
20
the bottom. A mixture of CO
2
(150 mmol) and H (50 mmol) was
2
admitted to the reactor, and the total pressure in the reactor was
ca. 25 kPa. After the irradiation, the gases were analysed by an
on-line TCD gas chromatograph (Shimadzu GC-8A) using
argon as a carrier gas, which can detect H
2
, CH
4
and CO.
10
0
Temperature programmed desorption (TPD) experiments were
carried out for zirconium oxide, and the desorbed gases were
detected by a quadrupole-type mass spectrometer (ULVAC
Massmate-100). To avoid the interference of nitrogen at the
signal m/z = 28, 13C labelled CO or CO
Activity for photoreduction of CO
TiO , ZrO , V , Nb , Ta , WO
only ZrO
showed no activity for this reaction at all. With ZrO
was used.
was investigated using
3
and ZnO. Among these,
2
0
10
20
t / h
30
40
2
2
2
2
O
5
2
O
5
2
O
5
Fig. 1 Time dependence of CO formation and H
consumption over ZrO
2
2
2
was found to be active, and other metal oxides
, 1.0 mmol
under irradiation; initial amounts of CO2 and H2 were 150 mmol and
50 mmol, respectively
2
Chem. Commun., 1997
841
Kohno, Yoshiumi
