Visible-light photocatalytic conversion of carbon monoxide to methane by nickel(II) oxide

Article abstract of DOI:10.1002/anie.201307851

Solar Fuels: Different n- and p-type semiconductors have been investigated for sustainable solar fuel production. p-Type semiconductors, such as NiO, Fe₃O₄, Co₃O₄, and CuO, are able to reduce carbon monoxide by water or hydrogen to methane (see picture). The highest CH₄ yield achieved was 17.26 mmol of CH₄ per gram of catalyst using NiO in an excess of H₂. Copyright

Full text of DOI:10.1002/anie.201307851

Angewandte
Chemie
DOI: 10.1002/anie.201307851
Photocatalysis
Hot Paper
Visible-Light Photocatalytic Conversion of Carbon Monoxide to
Methane by Nickel(II) Oxide**
Francesc Sastre, Avelino Corma,* and Hermenegildo Garcꢀa*
Research in the area of “solar fuels” obtained photocatalyti-
cally by sunlight is actual and active because these fuels are
sustainable and CO2-neutral. Although H2 generation from
H2O is the most widely studied photocatalytic process for
solar fuel production, the technical problems associated with
H2 storage and fuel-cell efficiency make also attractive other
chemicals obtained from CO2 such as CH3OH, CH4, and
CO.[1–7]
ethane. Formation of CO2 can be understood as arising from
photooxidation of CO by positive holes generated in the
charge separation on the semiconductor. The conversion of
CO using TiO2 promoted by solar light was very small (see
Table 1) in agreement with the well-documented low photo-
catalytic activity of TiO2 at solar light irradiation.[15]
Then we tested other first-row transition-metal oxides
that can absorb visible light, though they are generally
The production of solar fuels requires efficient photo-
catalysts with high visible-light photoresponse. While TiO2 is
the most widely studied photocatalyst,[8–10] this semiconductor
shows activity only upon UV-light exposure and a consider-
able effort aims at developing TiO2 photocatalysts with visible
response.[11–15] In contrast, the use of p-type semiconductors
absorbing visible light such as NiO, CuOx, FeOx, and CoOx
has been less explored in photocatalysis, because of their
presumably lower stability compared to TiO2.[16] In particular,
NiO has been frequently employed as co-catalyst in combi-
nation with titanates or other photocatalysts for visible-light
splitting of H2O rather than as active photocatalyst. In the
present report, we describe that p-type semiconducting metal
oxides, specifically NiO and FeOx, show high visible and
sunlight photocatalytic activity for the reduction of CO by
H2O vapor or H2 leading to the formation of substantial
amounts of CH4, with almost complete selectivity at full CO
conversion. NiO can be reused without change in the XRD
pattern, supporting its photostability. The novelty of our study
derives from the consideration of two points: 1) the unpre-
cedented room-temperature photocatalytic CO reduction by
H2O or H2 to CH4, and 2) the possibility of using sunlight and
visible light to activate the process.
Table 1: Photocatalytic activity for CO conversion, carbon content, and
formation of H2, CO2 and CH4 in the presence of metal-oxide semi-
conductors.[a]
Photocatalyst[b] CO Conversion
[%]
C [%][c] Production [mmolgꢁ1 cat.]
[d]
H2
CO2
CH4
[e]
TiO2
3
95
0.1
3.3
260 650
820 12200 1700
(10.2)
4 (0.5)
NiO[e]
CuO[e]
Co3O4
Co2O3
14
91
93
59
68
96
0.5
1.2
3.1
0.1
2.5
2.2
20
2900 270 (7.5)
[e]
[e]
1000 12300 600 (4.3)
990 12500 200 (1.3)
3620 8200 900 (9.8)
[e]
Fe3O4
Fe2O3
NiO[f]
[e]
3760 10400 20 (0.2)
[g]
–
11240 8387
(39.1)
[f]
[g]
[g]
[gj
[g]
[g]
[g]
Fe3O4
49
99.5
28
99.4
99.0
16
3.6
1.8
0.5
1.9
1.8
1.0
–
–
–
–
–
–
6590 733 (7.1)
NiO[h]
711
943
450
87
1386 1379
(38.3)
17266 (87)
580 (28.2)
20799 (92)
16030 (97)
[h]
Fe3O4
NiO[i]
NiO[j]
NiO[k]
[a] Blank controls: 1) illumination in the absence of any solid, 2) contact
to the solid and gas phase in the dark or 3) irradiation of the
photocatalyst in N2 did not lead to any product, C mass balances were
higher than 95%. [b] See the Supporting Information for the origin and
characterization data of these commercially available photocatalysts.
[c] Weight percentage of carbon on the solid after the photocatalytic
experiment. [d] The number in brackets corresponds to the selectivity of
the process to CH4 defined as number of moles formed of CH4 divided by
moles of converted CO in percentage. [e] Irradiation conditions: Mixture
of N2 (19.6 mmoles), H2O (2.5 mmol), and CO (20%, 4.9 mmol),
irradiation time 22 h, photocatalyst mass 250 mg, irradiation source
solar simulator (1000 Wmꢁ2). [f] Reaction conditions: Mixture of CO
(20%, 5 mmol), H2 (5 mmol), Ar (1.1 mmol), and N2 (13.9 or
0.125 mmol); photocatalyst mass 250 mg, reaction time 3 h, light source
solar simulator. [g] Reagent. [h] Reaction conditions: Mixture of CO
(20%, 5 mmol), H2 (18.75 mmol), Ar (1.1 mmol), and N2 (13.9 or
0.125 mmol); photocatalyst mass 250 mg, reaction time 3 h, light source
solar simulator. [i] Second test under conditions indicated in [h] and
activation of NiO at 4508C with air for 5 h. [j] Third test under conditions
indicated in [h] and activation of NiO at 4508C with air for 5 h.
[k] Irradiation with the solar simulator equipped with a cut-off filter
(l>400) under conditions indicated in [h].
Preliminary experiments using TiO2 as photocatalyst
showed that this n-type semiconductor leads mainly to
oxidation of CO by H2O vapor, forming quasi-stoichiometric
amounts of CO2 and H2 [Eq. (1)],
hn
ð1Þ
CO þ H O
CO þ H ðDGꢀ ¼ þ31:36 kJ molꢁ1
Þ
ƒƒƒƒƒƒ!
2
2
2
photocatalyst
accompanied by detectable, amounts of CH4, and traces of
[*] Dr. F. Sastre, Prof. Dr. A. Corma, Prof. Dr. H. Garcꢀa
Instituto Universitario de Tecnologꢀa Quꢀmica CSIC-UPV
Universidad Politꢁcnica de Valencia
Avenida De los Naranjos s/n, 46022 Valencia (Spain)
E-mail: acorma@itq.upv.es
hgarcia@qim.upv.es
[**] Financial support by the Spanish Ministry of Economy and
Competitiveness (MINECO) through several actions (Severo Ochoa
and grant number CTQ 2012-32315) is gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201307851.
Angew. Chem. Int. Ed. 2013, 52, 12983 –12987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12983

.
Angewandte
Communications
considered much less efficient photocatalysts than TiO2. The
results obtained for the p-type metal-oxide semiconductors
are summarized in Table 1, whereas Figure S1 (see the
Supporting Information) shows the temporal evolution of
CH4 upon irradiation in the presence of the p-type semi-
conductors using simulated solar light.
In contrast to experiments performed in the dark, where
no products are formed (see footnote [a] in Table 1), or in
contrast to the generation of CO2 using TiO2, p-type semi-
conducting metal oxides of Ni, Cu, Fe, and Co afford
significant amounts of CH4 in the gas phase upon sunlight
irradiation. Equation (2) provides the presumed stoichiom-
gradually. CH4 exhibits an induction period at initial times
and is formed as a secondary product. These temporal
profiles, particularly the fact that the rate of H2 decrease is
about one half that of the CH4 growth, provide kinetic
support to a CH4 formation mechanism involving prior
generation of H2 from H2O vapor and subsequent reaction
of H2 to form CH4.
With regard to Table 1, it is worth commenting on the high
percentage of carbon present on the solid photocatalyst,
particularly in the cases of NiO, Co2O3, and Fe2O3. This
carbonaceous residue adsorbed on the photocatalyst was
characterized by Raman and X-ray photoelectron spectros-
copy (XPS; see Figures S2 and S3), both spectra being
compatible with the assignment of this residue to graphitic
carbon generated in the photocatalytic process. The fresh
photocatalysts do not show any carbon-related Raman peak.
However, after irradiation, the characteristic 2D, G, and D
bands of carbon appearing between 2500–2900 and at 1600
and 1350 cmꢁ1 were observed.
XPS of the photocatalyst after being used shows the
presence of a peak at 285.12 eV that can be deconvoluted in
three major components corresponding to sp2 graphene
C atoms (67%), sp2 C atoms bonded to O (25%), and sp3
C atoms (8%), respectively. The percentage of carbonaceous
residue deposited on each material before and after irradi-
ation was determined by combustion elemental analysis of the
solid. The carbon content of the fresh photocatalysts was
negligible, whereas it was necessary to consider the amount of
carbon deposited on the photocatalyst after the reaction,
together with the CH4 and CO2 formed in the gas phase to
calculate satisfactorily the C mass balance (> 95%) in all
cases. This indicates that CO is converted not only into gas-
phase products, but it also forms carbon deposits on the solid.
In agreement with the presence of carbon on the used
photocatalyst, the colored metal oxides after the reaction
become in some cases visually black (see the inset in Figure S2
for NiO).
CO þ 2 H2O ! CH4 þ 1:5 O2 ðDG0 ¼ 541 kJ molꢁ1
Þ
ð2Þ
etry of the CH4 formation for the photocatalytic CO
reduction by water using simulated solar light. While TiO2 is
a white powder, all the rest of the transition metals in Table 1
are colored, the visible absorption bands being responsible for
the notable sunlight photoactivity (up to 33 times that of
TiO2). Particularly notable is the performance of NiO and
Fe3O4 (in its most abundant a-polymorph)[17] where the
production of CH4 is the highest (10.2% in the gas-phase
mixture in the case of NiO). These differences in photo-
catalytic activity in the series of p-type metal-oxide semi-
conductors are unlikely to arise from differences in their
surface area, since they show similar values (see Table S1). It
is more reasonable to attribute the results to the intrinsic
photocatalytic properties (redox potentials of electrons and
holes, efficiency of charge separation and lifetime of charge
separation among other parameters to be considered) of the
semiconductor.
Figure 1 shows the temporal evolution of the CO con-
version and the formation of CO2, CH4, and H2 using NiO as
photocatalyst and H2O as reducing agent under simulated
sunlight irradiation. As can be seen in Figure 1, H2 is formed
as primary, but unstable product and its proportion grows in
the first hour of irradiation, but subsequently decreases
The behavior observed for p-type semiconductors (for-
mation of large CH4 percentages in the gas phase and carbon
residues on the solid) sharply contrasts with the photocata-
lytic activity of TiO2 commented earlier [Eq. (1)]. Additional
tests using CeO2 as photocatalyst, another n-type semicon-
ductor, also show the preferential formation of CO2. Sunlight
conversion of CO using TiO2 and CeO2 is about 30 times
lower than the conversion achieved by p-type metal oxides.
This low activity of TiO2 and CeO2 at solar light irradiation is
well-known and derives from their limited visible-light
photoresponse of these wide-bandgap semiconductors.
Besides the advantageous photoaction spectra of n-type
semiconductors arising from the visible-light absorption of
these materials, the different photocatalytic behavior of p-
and n-type semiconductors could derive from the higher
conduction-band potential in p-type semiconductors (+ 0.8 V
versus the normal hydrogen electrode, NHE, for Ni)[18]
compared to TiO2 (ꢁ0.1 V versus NHE at pH 0)[19] and
CeO2 (ꢁ0.1 V).[12] A high conduction band potential favors
the reduction of CO to C and CH4 with respect to the
oxidation of CO to CO2 which is almost the sole process
observed for TiO2 and CeO2.
&
*
Figure 1. Photocatalytic CO ( ) conversion and formation of CO2 ( ),
~
!
CH4 ( ) and H2 ( ). Reaction conditions: gas-phase irradiation of
a moisture-saturated mixture of N2 (80%) and CO (20%, 4.9 mmol),
NiO 250 mg, irradiation source solar simulator (1000 Wꢂmꢁ2).
12984 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12983 –12987

Angewandte
Chemie
To establish if the photocatalytic reduction of CO to CH4
using p-type photocatalyst can be further favored using H2
instead of H2O as reducing agent present in the system,
additional experiments with simulated sunlight where NiO or
Fe3O4 were illuminated in contact with an atmosphere that
contains CO and H2 were carried out. H2 could be obtained
independently from steam reforming and the water-gas shift
reaction or generated photocatalytically from water. The
results obtained using two different CO/H2 molar ratios are
summarized in Table 1 (footnotes [f] and [h]), whereas the
corresponding time conversion plots are shown in Figure 2.
The photocatalytic CH4 formation increases considerably
using H2 as reducing agent. The high CO conversions
when the reaction time is prolonged (Figure S4). When the
irradiation is carried out in H2 excess, the CH4 production is
higher, the consumption of CO is complete, and the percent-
age of CO2 formed is considerably lower.
Comparison of the photocatalytic activity of NiO and
Fe3O4 indicates that the former is more efficient for the CH4
formation (Table 1). The highest CH4 yield achieved herein
was 17.26 mmol per gram of catalyst, corresponding to
a reaction carried out using NiO as photocatalyst with a 3.5-
fold excess of H2. Under these optimal conditions in excess of
H2 using simulated sunlight with a nominal power of
1000 Wꢀmꢁ2, the yield of CO2 is 24 times lower than the
global amount of alkanes. This efficiency using simulated
solar light is remarkable and without precedent.
To assess the visible-light photoactivity of NiO, the
photoaction spectrum for the CH4 generation was determined
(see the inset in Figure 2). The overlay of the apparent
quantum yield for the CH4 generation based on the incident
photon conversion (maximum Fapp 7.5% at 450 nm) with the
NiO absorption band shows a notable coincidence, although
the efficiency of the photocatalytic CH4 generation at long
wavelengths is much lower compared to the NiO absorption
spectrum. This could be related to the higher efficiency of the
ligand-to-metal electronic transition appearing around
410 nm with respect to the less efficient Ni2+-centered d–d
absorption characteristic of transition-metal ions with incom-
plete d shell appearing at longer wavelengths in the visible
region.
In cases in which H2 is present together with CH4,
significant amounts of ethane and propane are also formed.
Formation of these light alkanes decreases somewhat the CH4
selectivity. This behavior is reminiscent of the Fischer–
Tropsch process, where a mixture of CO and H2 gives rise to
the formation of linear hydrocarbon molecules at elevated
temperatures using Ni or Fe catalysts. It is, however, pertinent
to remind that the current process is performed at temper-
atures below 368C.
~
&
*
Figure 2. CO ( ) conversion and formation of CO2 ( ) and CH4 ( ),
!
O2 ("), ethane (3) and propane ( ). Reaction conditions: Mixture of
CO (20%, 5 mmol), H2 (18.75 mmol), Ar (1.1 mmol), and N2 (13.9 or
0.125 mmol); photocatalyst mass 250 mg, reaction time 3 h, light
source solar simulator, illumination source solar simulator
(1000 Wꢂmꢁ2). The inset shows the apparent quantum yields (F) for
the CH4 formation (a) and the UV/Vis absorption spectrum of NiO (b;
A=absorbance).
Formation of elemental carbon on the photocatalyst
corresponds to CO reduction and this elemental carbon
could act as an intermediate in the formation of CH4. In order
to verify this possibility, a test was carried out in which CO
diluted in N2 was submitted to irradiation until the NiO
photocatalyst became visually black. At this time, the
irradiation was stopped and the photoreactor outgassed to
remove CO. Sampling of the photocatalyst revealed 1.7 wt%
carbon content. Then, H2 was admitted into the photoreactor
and the irradiation continued. CH4 evolution was observed in
the gas phase in the absence of CO. This experiment can be
easily interpreted considering that upon irradiation in the
absence of H2 carbon deposits are formed on the NiO surface.
Subsequently, this carbon residue is able to react photo-
catalytically with H2 to form CH4, regardless of the absence of
CO. A control experiment following an identical protocol in
the dark shows that no CH4 is formed.
achieved using H2 allow the quantification of the O2
evolution, which shows a trend similar to the temporal profile
of the CH4 evolution (Figure 2) according Equation (3).
CO þ 2 H2 ! CH4 þ 0:5 O2ðDG0 ¼ þ35:6 kJ molꢁ1
Þ
ð3Þ
Both in the absence and presence of H2, formation of CO2 is
observed although with considerably lower yields when H2 is
present (see Table 1 and Figure 2). This higher activity using
H2 as reducing agent can be easily understood considering the
high endoergonicity of the reaction using H2O as reducing
agent [Eq. (2)] and the more complex reaction mechanism
using H2O that requires prior H2 generation. The percentage
of elemental carbon deposited on NiO is reduced by one half
when the irradiation is carried out using H2 as reducing agent
compared to H2O vapor.
A mechanistic proposal for the photoreduction of CO by
H2 using a p-type photocatalyst is shown in Scheme 1.
According to this proposal, CO can act as a hole-trapping
agent, leading to the formation of CO2. If another electron
donor agent, such as H2O or H2, is also present, then these
For the formation of CH4 two equivalents of H2 should be
consumed. Thus, for those irradiation reactions where H2 and
CO are present in a 1:1 molar ratio, the generation of CH4
reaches asymptotically a value that does not grow further
Angew. Chem. Int. Ed. 2013, 52, 12983 –12987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12985

.
Angewandte
Communications
CH4 selectivity (from 87 to 97%) were measured after the
three tests. Moreover, XRD of the NiO sample before and
after the three tests was also coincident, ruling out changes in
the crystal structure of NiO. XPS, which is a specific surface
analysis technique, did not show differences in the binding
energy related to the Ni 2p peaks (Figure S5). This informa-
tion indicates the stability of NiO as photocatalyst under the
present conditions. A similar stability was observed by XPS
for Fe3O4 and Co3O4 (Figures S6–S8)
The previous photocatalytic experiments were carried out
using simulated sunlight containing about 5% UV radiation.
The high photocatalytic activity in terms of CO conversion
using NiO as catalyst compared to TiO2 or CeO2 has been
attributed to the visible-light photoresponse of NiO in
accordance to the photoaction spectrum. Additional experi-
ments on the photocatalytic reduction of CO to CH4 by H2
using NiO were also carried out under visible-light irradiation
using a filter that cuts off radiation wavelengths shorter than
400 nm (see the spectrum of the cut-off filter in Figure S9).
The results of the visible-light photocatalytic activity of NiO
(see Table 1) show that although a large percentage of the
catalytic activity of NiO is related to the UV region, NiO is
still active under visible-light illumination.
Scheme 1. Contrasting photocatalytic behavior of n- and p-type semi-
conductors based on the product distribution observed.
compounds will compete with CO for oxidation. CO can
alternatively undergo reduction by the semiconductor con-
duction-band electrons, forming elemental C and O2. This CO
reduction process will be more favorable as the reduction
potential of the conduction-band edge of the photocatalyst
increases (case of p-type semiconductors), as has been
observed in related precedents.[20] Thus, one of the main
factors influencing the outcome of the photocatalytic CO
reduction should be the position of the semiconductor
conduction-band edge.
In agreement with this rationalization, a stoichiometric
growth in the concentration of molecular O2 was detected
when H2 is the reducing agent. When H2O is used as electron
donor, the lower formation of CH4 makes the quantification
of O2 uncertain because of the air contamination. Elemental
carbon is deposited on the surface of the photocatalyst as
evidenced visually (inset in Figure S2) and determined by
combustion chemical analysis (Table 1). Elemental carbon
undergoes efficient photocatalytic H2 reduction, giving rise to
CH4.
To provide some support to the mechanistic proposal,
additional experiments were carried out. A mixture of CO
and H2 was irradiated in the presence of methanol as
sacrificial electron donor agent, whereby total inhibition of
the formation of CO2 was observed, while CH4 was still
formed. This is in accordance with the proposed origin of CO2
arising from the photooxidation of CO on the surface of the
semiconductor as indicated in Scheme 1.
The photocatalytic irradiation of CO and H2 using
simulated sunlight was also performed in the presence of
CH4 and NiO as photocatalyst, and under these conditions
ethane as primary irradiation product was observed, similar to
the Fischer–Tropsch mechanisms and the reported photo-
catalytic CH4 conversion.[21]
One of the main reasons frequently invoked to disregard
p-type semiconducting metal oxides as photocatalysts is their
presumed lack of stability and the occurrence of corrosion.
Photocorrosion tests have been carried out mainly in liquid
water. Since these conditions are different from those
employed in the present study in the gas phase, we have
verified the stability of the photocatalytic activity of NiO by
performing three consecutive tests of the same sample. After
each test, the NiO sample was submitted to calcination at
4508C in air to effect the combustion of possible carbona-
ceous residue. As can be seen in Table 1, a minor decay in the
CO conversion (from 99.5 to 99%) and even an increasing
In conclusion, here we have shown that p-type transition-
metal oxide semiconductors are efficient photocatalysts for
the visible-light or solar-light, room-temperature reduction of
CO to CH4, provided that H2O vapor or H2 are present as
reducing agents. This photochemical behavior contrasts with
that of TiO2 and CeO2 (n-type semiconductors) that promote
under the same conditions highly selectively the oxidation of
CO to CO2. Formation of elemental carbon deposited on the
photocatalyst surface is also observed. NiO was the most
efficient photocatalyst among the series of first-raw transi-
tion-metal oxide semiconductors. Under optimal conditions
using simulated sunlight and molecular H2 as reducing agent,
a selectivity towards the formation of CH4 of 97% at
complete CO conversion was obtained. Our results open
new opportunities for the photocatalytic production of solar
fuels at high conversion and selectivity.
Received: September 6, 2013
Published online: October 24, 2013
Keywords: methane formation · metal oxides · photocatalysis ·
.
semiconductors · solar fuels
[1] G. R. Dey, J. Nat. Gas Chem. 2007, 16, 217.
[2] S. Navalꢁn, A. Dhakshinamoorthy, M. ꢂlvaro, H. Garcia,
ChemSusChem 2013, 6, 562.
[3] J.-S. Hwanga, J.-S. Changa, S.-E. Parkb, K. Ikeuec, M. Anpo,
Top. Catal. 2005, 35, 311.
[4] T. Inoue, A. Fujishima, S. Konishi, K. Honda, Nature 1979, 277,
637.
´
[5] K. Koci, L. Obalovꢃ, L. Matejovꢃ, D. Plachꢃ, Z. Lacny, J.
´
Jirkovsky, O. Solcovꢃ, Appl. Catal. B 2009, 89, 494.
[6] W. Leitner, Angew. Chem. 1995, 107, 2391; Angew. Chem. Int.
Ed. Engl. 1995, 34, 2207.
[7] K. Mori, H. Yamashita, M. Anpo, RSC Adv. 2012, DOI: 10.1039/
c2ra01332k.
[8] M. A. Fox, M. T. Dulay, Chem. Rev. 1993, 93, 341.
12986 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 12983 –12987

Angewandte
Chemie
[9] M. R. Hoffmann, S. T. Martin, W. Y. Choi, D. W. Bahnemann,
Chem. Rev. 1995, 95, 69.
[15] C. G. Silva, R. Juarez, T. Marino, R. Molinari, H. Garcia, J. Am.
Chem. Soc. 2011, 133, 595.
[16] J. M. Herrmann, ACS Symp. Ser. 1986, 298, 200.
[17] G. Carraro, C. Maccato, A. Gasparotto, T. Montini, S. Turner,
O. I. Lebedev, V. Gombac, G. Adami, G. Van Tendeloo, D.
Barreca, P. Fornasiero, Adv. Funct. Mater. 2013, n/a.
[18] M. P. Dare-Edwards, J. B. Goodenough, A. Hamnett, N. D.
Nicholson, J. Chem. Soc. Faraday Trans. 2 1981, 77, 643.
[19] M. Grꢄtzel, Nature 2001, 414, 338.
[20] Y. P. Xie, G. Liu, L. Yin, H.-M. Cheng, J. Mater. Chem. 2012, 22,
6746.
[21] L. Yuliati, H. Yoshida, Chem. Soc. Rev. 2008, 37, 1592.
[10] D. W. Bahnemann, Res. Chem. Intermed. 2000, 26, 207.
[11] A. Primo, A. Corma, H. Garcia, Phys. Chem. Chem. Phys. 2011,
13, 886.
[12] A. Primo, T. Marino, A. Corma, R. Molinari, H. Garcia, J. Am.
Chem. Soc. 2011, 133, 6930.
[13] S. Sakthivel, M. Janczarek, H. Kisch, J. Phys. Chem. B 2004, 108,
19384.
[14] S. Sakthivel, M. V. Shankar, M. Palanichamy, B. Arabindoo,
D. W. Bahnemann, V. Murugesan, Water Res. 2004, 38, 3001.
Angew. Chem. Int. Ed. 2013, 52, 12983 –12987
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12987