Photoelectrochemical Reduction of CO2 Coupled to Water Oxidation Using a Photocathode with a Ru(II)-Re(I) Complex Photocatalyst and a CoOx/TaON Photoanode

Article abstract of DOI:10.1021/jacs.6b09212

Photoelectrochemical CO₂ reduction activity of a hybrid photocathode, based on a Ru(II)-Re(I) supramolecular metal complex photocatalyst immobilized on a NiO electrode (NiO-RuRe), was confirmed in an aqueous electrolyte solution. Under half-reaction conditions, the NiO-RuRe photocathode generated CO with high selectivity, and its turnover number for CO formation reached 32 based on the amount of immobilized RuRe. A photoelectrochemical cell comprising a NiO-RuRe photocathode and a CoO_x/TaON photoanode showed activity for visible-light-driven CO₂ reduction using water as a reductant to generate CO and O₂, with the assistance of an external electrical (0.3 V) and chemical (0.10 V) bias produced by a pH difference. This is the first example of a molecular and semiconductor photocatalyst hybrid-constructed photoelectrochemical cell for visible-light-driven CO₂ reduction using water as a reductant.

Full text of DOI:10.1021/jacs.6b09212

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
2
Photoelectrochemical Reduction of CO Coupled to
Water Oxidation Using a Photocathode With a Ru(II)–
x
Re(I) Complex Photocatalyst and a CoO/TaON Photoanode
Go Sahara, Hiromu Kumagai, Kazuhiko Maeda, Nicolas Kaeffer,
Vincent Artero, Masanobu Higashi, Ryu Abe, and Osamu Ishitani
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b09212 • Publication Date (Web): 02 Oct 2016
Downloaded from http://pubs.acs.org on October 6, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Photoelectrochemical Reduction of CO2 Coupled to Water Oxidation
Using a Photocathode With a Ru(II)–Re(I) Complex Photocatalyst
and a CoOx/TaON Photoanode
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Go Sahara,a Hiromu Kumagai,a Kazuhiko Maeda,a Nicolas Kaeffer,b Vincent Artero,b Masanobu
Higashi,c Ryu Abe,*,c and Osamu Ishitani*,a
aDepartment of Chemistry, School of Science, Tokyo Institute of Technology, O-okayama 2-12-1-NE-1, Meguro-ku,
Tokyo 152-8550, Japan.
bLaboratoire de Chimie et Biologie des Métaux, Université Grenoble Alpes, CNRS UMR 5249, CEA, F-38000, Gre-
noble, France
cDepartment of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Katsura,
Nishikyo-ku, kyoto 615-8510, Japan
ABSTRACT: Photoelectrochemical CO2 reduction activity of a hybrid photocathode, based on a Ru(II)–Re(I) supramolec-
ular metal complex photocatalyst immobilized on a NiO electrode (NiO–RuRe) was confirmed in an aqueous electrolyte
solution. Under half-reaction conditions, the NiO–RuRe photocathode generated CO with high selectivity, and its turno-
ver number for CO formation reached 32 based on the amount of immobilized RuRe. A photoelectrochemical cell com-
prising a NiO–RuRe photocathode and a CoOx/TaON photoanode showed activity for visible-light-driven CO2 reduction
using water as a reductant to generate CO and O2, with the assistance of an external electrical (0.3 V) and chemical bias
(0.10 V) produced by a pH difference. This is the first example of a molecular and semiconductor photocatalyst hybrid-
constructed photoelectrochemical cell for visible-light-driven CO2 reduction using water as a reductant.
INTRODUCTION
fully constructed, and some have achieved CO2 reduction
with conversion of photon energy to chemical energy
using methanol as an electron donor; however, the utili-
zation of water as a reductant has not been reported.15,19
And there are two other reports of hybrid systems com-
prising two or more semiconductor materials as light
absorbers and metal complexes that exhibit relatively
high activities to form formic acid and O2 from CO2 and
water.20–22 One was constructed with two semiconductor
photoelectrodes and a cathodic side hybridized with a
Ru(II) catalyst for CO2 reduction. The other system used
the combination of an amorphous silicon–germanium
solar cell with the same Ru catalyst. The former system
required the use of TiO2 or SrTiO3 as photoanodes neces-
sitating UV light for excitation. The latter system used a
triple-junction solar cell for producing both high-energy
electrons and holes. In both cases, the use of extremely
expensive materials, i.e., single-crystalline InP or the tan-
dem solar cell, was required. In addition, the metal com-
plexes were only used as catalysts for CO2 reduction and
not for photocatalysts in these hybrid systems. Thus the
hybrid photocatalytic system composed of both molecular
and semiconductor photocatalysts for visible-light-driven
CO2 reduction using water as a reductant has not yet been
realized.
Photocatalytic CO2 reduction to produce high-energy
compounds using sunlight as an energy source, so-called
artificial photosynthesis, is a promising means for con-
structing a sustainable and carbon-neutral society. Metal
complex photocatalysts have been intensively investigat-
ed as key players in photochemical CO2 reduction sys-
tems.1–3 In particular, supramolecular metal complexes
comprising both photosensitizer and catalyst units con-
nected with bridging ligands have demonstrated high
selectivity, durability, and quantum efficiency for CO2
reduction under visible light irradiation.4–14 These metal
complex photocatalysts required a sacrificial reductant for
driving the CO2 reduction due to the low oxidizing ability
of the photosensitizer unit, which is insufficient strength
to utilize water as a reductant.
Aiming to achieve visible-light-driven photocatalytic
CO2 reduction using water as a reductant, hybrid photo-
catalytic systems combined two kinds of photocatalysts,
i.e., metal complexes and semiconductor materials, have
been proposed.15–18 In these systems, two photons are
required for a Z-scheme-type stepwise excitation to pro-
duce both an electron with strong reducing power for CO2
reduction and a hole with strong oxidizing power for non-
sacrificial electron donor such as methanol or water. Re-
ported hybrid photocatalytic systems have been success-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 9
Recently, we reported a novel hybrid photoelectrode reaction chamber was conducted using a gas chromato-
1
2
3
4
5
6
7
8
consisting of a Ru(II)–Re(I) supramolecular metal com-
plex photocatalyst (RuRe) and a NiO electrode for pho-
toelectrochemical CO2 reduction in a non-aqueous solu-
tion.23 Visible light irradiation of the immobilized RuRe
was shown to drive highly selective CO2 reduction using
electrons supplied from an external electric circuit
through the p-type NiO semiconductor electrode, of
which excitation was not necessary. If a cell, which used a
counter photoanode composed of visible-light-driven
semiconductor photocatalyst for water oxidation, e.g.
TaON,24,25 was constructed with the above mentioned
photocathode, CO2 reduction could proceed under visible
light irradiation using the electrons generated through
water oxidation. Moreover, in such a tandem cell design,
the photocatalyst for CO2 reduction can be separated
from the photoanode so as to suppress the back oxidation
of products. Finally, the photocatalytic materials for CO2
reduction and water oxidation can be individually de-
signed and optimized in such a total tandem system.
Therefore, developing these type of hybrid photoelectro-
chemical cells could be a first step in the realization of
CO2 reduction with simultaneous water oxidation using
visible light as the energy source.
graph equipped with a mass spectrometer (Shimadzu,
QP-2010-Ultra, Molsieve5A capillary column (30 m)). X-
ray photoelectron spectroscopy (XPS) measurement was
conducted using ESCA-3400 (Shimazu). The binding
energy of the impurity carbon (1s) peak was adjusted to
284.6 eV to correct the chemical shifts of each element.
Attenuated total reflection infrared spectroscopy (ATR-
IR) was measured using FT/IR-4600 (JASCO).
9
Preparation of the NiO-metal Complex Hybrid
Electrodes. NiO-metal complex hybrid electrode was
prepared as reported before.23 Details of structural charac-
terization of the electrode are shown in our previous
paper.23 Briefly, precursor solution consisting of
Ni(NO3)2·6H2O (1.0 g, 99.95% KANTO chemicals) and
Pluronic F-88 (0.5 g) dissolved in water/ethanol (4.5 g,
1:2, w/w) was deposited on a cleaned FTO glass (AGC
fabritech, 15ꢀ50 mm, 12 Ω/sq) by the doctor blade meth-
od with a glass rod using Scotch mending tape as a spac-
er. The obtained sample was calcined at 773 K for 30 min
in air. This deposition-calcination cycle was repeated four
times. The electrode was cut in half before use and im-
mersed in a solution of acetonitrile with metal complex (4
mL, 10 μM) overnight. The electrode was washed with
acetonitrile after hybridization. The adsorption amount of
the metal complex was estimated by the difference in the
absorbance of the metal to ligand charge transfer (MLCT)
absorption band of the solution (at 461 nm) between
before and after hybridization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Herein, we report a novel hybrid photoelectrochemical
cell for CO2 reduction that uses water as the electron
donor, composed of a NiO–RuRe hybrid photocathode
and a CoOx/TaON photoanode. First, we clearly demon-
strate that the NiO–RuRe photocathode can drive the
photocatalytic reduction of CO2 in an aqueous electrolyte
solution. Next, the photocathode was successfully com-
bined with the CoOx/TaON photoanode to drive water
oxidation under visible light irradiation in a functional
photoelectrochemical cell.25
Preparation of the CoOx/TaON Electrode.
CoOx/TaON electrode was prepared as reported proce-
dure.25 Details of structural characterization of the elec-
trode are shown in our previous paper.25 TaON powder
was prepared by heating Ta2O5 powder under NH3 flow
(20 mL min–1) at 1123 K for 15 h. CoOx nanoparticles
(5 wt% as metal) were loaded on TaON particles by im-
pregnation from an aqueous Co(NO3)2 solution, followed
by heating at 673 K for 30 min in air. As-prepared CoOx-
loaded TaON particles were deposited on Ti substrate by
electrophoretic deposition. The electrodes were treated
with 50 μL of TaCl5 methanol solution (10 mM) and then
dried in air at room temperature. After this process was
performed five times, the electrode was heated in NH3
flow (10 mL min-1) at 723 K for 30 min.
EXPERIMENAL SECTION
Materials. Acetonitrile was distilled three times over
P2O5 and once over CaH2 immediately before use. Pluron-
ic® F-88 was kindly supplied by BASF. 13CO2 (13C = 99%)
was purchased from CIL. NaH13CO3 (13C = 98%) was pur-
chased from Aldrich. H2 O (18O = 98%) was purchased
18
from CMR and diluted with non-labeled water to 30% as
H218O. Other reagents and solvents were commercial
grade quality and used without further purification.
Estimation of Electrochemically Active RuRe Spe-
cies on NiO-RuRe. The estimations of electrochemically
active RuRe species on the photoelectrode were conduct-
ed by the cyclic voltammetry using a three-electrode set-
up. An Ar-purged CH3CN containing 0.1 M Et4NBF4 was
used as electrolyte. A Pt wire and a Ag wire in 0.01 M
AgNO3 acetonitrile solution were employed as the coun-
ter and reference electrodes, respectively. Scan rate was
set to 50 mV s-1 and the amount of electrochemically ac-
tive Ru unit was calculated using Eq.1 from 2nd cycles of
each sample.
General Procedure. (Photo)electrochemical meas-
urements were conducted using a potentiostat (ALS/CHI
620 or ALS/CHI 760e). A 300 W Xe lamp (Asahi Spec-
trum, MAX-302) with an IR-cut mirror module and a
band pass filter (Asahi Spectrum, MX0460 for irradiation
of
λ
= 460 nm) or cut-off filters (HOYA, L42 and Y48 for
> 400 nm and > 460 nm, respectively)
irradiation of
λ
λ
was utilized for light irradiation. Product analysis of CO
and H2 in the gas phase was performed by a means of gas
chromatography (Inficon, MGC3000A). HCOOH in the
liquid phase was checked using a capillary electrophoresis
system (Otsuka Electronics, Capi-3300I). Evolved oxygen
in liquid phase was measured by Clark-type oxygen sensor
(Unisense OX-N). GC–MS analysis of the gas phase of the
n= S / F v (Eq.1)
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
where n is amount of electrochemically active Ru unit
(mol), S: area of oxidation peak of Ru(II/III) (A V), F:
Faraday constant (96485 C mol-1), v: sweep rate (V s-1).
for the anode chamber. After bubbling with He instead of
18
36Ar which has the same molecular weight as O2 into the
1
2
3
4
5
6
7
8
anode chamber and with CO2 into the cathode chamber
for 30 min, visible light was irradiated for 60 min with the
bias of –0.3 V by using the same system as described
above. The products in the gas phase of the anode cham-
ber were analyzed by GC–MS by sampling using a syringe
through a septum after the irradiation.
Photoelectrochemical CO2 Reduction using NiO-
RuRe. Photoelectrochemical CO2 reduction was conduct-
ed using an O-ring-sealed Pyrex cell and a stirring tip.
The total volume of the cell without the stirring tip and
electrodes was ca. 130 mL and 84 mL of a CO2-saturated
50 mM NaHCO3 (KANTO chemical) aqueous solution
(pH 6.6) was used as the electrolyte; therefore, the esti-
mated gas phase was ca. 46 mL. The reaction was con-
ducted after purging with CO2 for more than 40 min. A
three-electrode setup was used through the photoelectro-
chemical measurement and CO2 reduction reaction. A Pt
wire and Ag/AgCl in saturated KCl aqueous solution were
employed as the counter and reference electrodes, respec-
tively. The counter electrode was separated from the
reaction solution by Vycor® glass to prevent the influence
of oxidization reaction taking place on the counter elec-
trode, of which oxidation product(s) did not be measured.
The geometric area of working electrode was ca. 2.5 cm2.
9
RESULTS AND DISCUSSION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the hybrid photocathode, NiO–RuRe, RuRe with
two methylphosphonic acid groups as anchors (Chart 1)
was immobilized on the surface of a p-type semiconduc-
tor NiO electrode consisting of fluorine doped tin oxide
(FTO) glass coated with a NiO film (active area: ca. 2.5
cm2).23 Photoelectrochemical measurements using the
NiO–RuRe photocathode were conducted to examine its
activity for reducing CO2 in an aqueous solution. Figure 1
shows the current–potential curve of the NiO–RuRe
photocathode under irradiation at λ = 460 nm, at which
the photoexcitation of NiO cannot be occurred, in a CO2-
purged aqueous solution containing 50 mM NaHCO3 (pH
= 6.6); a photocathodic response was clearly observed
starting at approximately −0.1 V vs. Ag/AgCl. This behav-
ior was similar to that observed in an organic solution,
i.e., a 5:1 v/v N,N’-dimethylformamide and triethanola-
mine mixture containing Et4NBF4 (0.1 M) as an electro-
lyte,23 and the photoresponse can be attributed to elec-
tron transfer from the valence band of the NiO electrode
to the excited Ru photosensitizer unit in RuRe. These
results strongly suggest that photoelectrochemical CO2
reduction proceeded using the NiO–RuRe photocathode
in the aqueous solution.
Photoelectrochemical Water Oxidation using
CoOx/TaON. Photoelectrochemical water oxidation was
conducted using O-ring-sealed Pyrex cell with the total
volume of ca. 16 mL. 10 mL of an Ar-purged 50 mM Na-
HCO3 aqueous solution (pH 8.3) was used as the electro-
lyte. The reaction was conducted after purging with Ar for
more than 30 min. A three-electrode setup was used
through the photoelectrochemical measurement and
water oxidation reaction. A Pt wire and Ag/AgCl in satu-
rated KCl aqueous solution were employed as the counter
and reference electrodes, respectively. The counter elec-
trode was separated from the reaction solution by Nafion®
membrane (Aldrich, Nafion 117). The active geometric
area of working electrode was ca. 1.1 cm2.
2+
Br
Re
CO
N
H2O3P
H2O3P
N
N
N
N
N
N
CO
CO
Ru
Photoelectrochemical CO2 Reduction using Hybrid
Electrochemical Cell with Z-scheme Configuration.
The reaction was conducted using the electrochemical
cell with two chambers (ca.16 mL for each chamber) di-
vided by Nafion® membrane (Aldrich, Nafion 117) and the
photoelectrodes were installed into each chamber indi-
vidually. The active geometric area of each electrode was
ca. 1.1 cm2. 10 mL of a 50 mM NaHCO3 aqueous solution
was added as an electrolyte to both chambers and purged
with CO2 for the cathode chamber (pH = 6.6) and with Ar
for the anode chamber (pH = 8.3), respectively. Light
irradiation was conducted from the back side of the NiO–
RuRe photocathode and transmitted light through NiO–
RuRe and the Nafion membrane was irradiated to the
CoOx/TaON photoanode. (Figure 6) An electrical bias of -
0.3 V vs. CoOx/TaON was applied for NiO–RuRe using a
potentiostat to accelerate the reaction. We note that pre-
reaction using CoOx/TaON was conducted under 400 nm
light with an applied potential of +0.5 V vs. Ag/AgCl up to
10 mC as the electric charge passing the circuit in order to
remove organic substance on the electrode surface and
improve the Faraday efficiency for water oxidation. An
isotope tracer experiment was conducted using water
N
RuRe
2+
N
H2O3P
N
N
N
Ru
N
H2O3P
N
Ru
Re
Chart 1. Structures of metal complexes.
ꢀ5
ꢀ10
λ = 460 nm
ꢀ15
50 mV secꢀ1
ꢀ20
ꢀ0.8
ꢀ0.6
ꢀ0.4
ꢀ0.2
Potential / V vs.Ag/AgCl
Figure 1. Current–potential curve of NiO–RuRe with
intermittent irradiation (λ = 460 nm) in a CO2-purged 50-
18
containing 30% H2 O and NaHCO3 (50 mM) as a solvent
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 9
mM NaHCO3 aqueous solution (pH = 6.6). Dotted line
indicates I–E curve under dark conditions.
under 13CO2
in NaH13CO3
under CO2
in NaHCO3
1
2
3
4
5
6
7
8
9
m/z = 28
m/z = 29
m/z = 28
m/z = 29
As a typical run for the photoelectrochemical reactions,
the NiO–RuRe photocathode with an applied bias of
−0.7 V vs. Ag/AgCl in an aqueous solution was irradiated
at λ > 460 nm using a 300 W Xe lamp with a cutoff filter.
Figure 2 shows the time course of the generated products
that consist mainly of CO with a very small amount of H2
and half of the electrons which passed through the outer
circuit of photoelectrochemical cell; 361 nmol of CO was
produced with 12 h irradiation yielding a turnover number
(TONCO) of 32 based on the amount of RuRe adsorbed on
the NiO electrode. The selectivity of CO (SLCO) was very
high, especially in the first stage (SLCO > 95%). After an
induction period, the formation of H2 increased but total
SLCO was still 91% for 12 h of irradiation. Other gaseous
products were not detected. The electrons which could
not be assigned for the CO formation might be consumed
for reduction of the trivalent nickel ions (Ni3+) in NiO at
the initial stage of the reaction, which was observed in
similar photoelectrochemical reactions in both organic
and aqueous solutions,23,26 and another possibility is for-
mation of a small amount of formic acid via CO2 reduc-
tion which should be lower than the detection limit of
formate with the capillary electrophoresis analyzer (about
2 μmol in the solution). Formation of CO did not proceed
without irradiation or under an Ar only atmosphere (en-
try 7 in Table 1).
9.0 9.2 9.4 9.6 9.8 10.0
Retention time / min
9.0 9.2 9.4 9.6 9.8 10.0
Retention time / min
Figure 3. GC–MS chart of evolved CO in gas phase dur-
ing photoelectrochemical CO2 reduction using NiO–
RuRe in an aqueous solution containing 50 mM NaHCO3
under CO2 (left) or 50 mM NaH13CO3 under 13CO2 (right).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1 summarizes the results of the photoelectro-
chemical reactions conducted under different applied
biases; the corresponding I-t curves are shown in Figure
S1. The photocurrent and CO evolution are relatively
small at −0.3 V (entry 2) compared to those at −0.7 V
(entry 1), and become almost zero at 0 V (entry 3). Corre-
lation between photocurrent and CO evolution clearly
suggests a mechanism in which electrons are supplied to
the immobilized RuRe through the NiO electrode for the
reduction of CO2. Formation of CO did not proceed in the
cases using the model mononuclear complexes of either
the photosensitizer Ru(II) unit (Ru in Chart 1) or the
Re(I) catalyst unit (Re) instead of the RuRe (entries 4 and
5) complex. An induction period for the generation of H2
using NiO–RuRe was observed for the first 2 h (Figure
S2), and the H2 formation was also observed even using
NiO–Ru instead of NiO–RuRe (entry 4 in Table 1, TONH2
= 2). These results strongly suggest that the H2 was de-
rived from photocatalysis of the degraded product of the
Ru photosensitizer, such as [RuII(N^N)2(L)2]n+ (L = sol-
vent), during the irradiation. Such phenomenon has been
reported in homogeneous photocatalytic systems using a
Ru(II)–Re(I) binuclear complex in an aqueous solution.14,27
For clarifying the carbon source(s) of the produced CO,
a similar photoelectrochemical reaction was conducted
with NaH13CO3 as an electrolyte under a 13CO2 atmosphere
(C-13 labeled gas and electrolyte were used to prevent
equilibrium redistribution of the labeled isotope). GC–MS
analysis revealed that 13CO was produced and 12CO was
not clearly detected (Figure 3). Therefore, the carbon
source of CO was confirmed to be CO2 either as intro-
duced or formed from the NaHCO3 electrolyte.
Table 1. Photoelectrochemical CO2 reduction using var-
ious electrodesa.
eꢀ/2
600
En
Com-
plex/
nmol
CO/
nmol
(TONCO
Poten-
tial/ Vb
H2/
nmol
Fred/
%
try Sample
CO
H2
400
200
0
)
1
241 (22)
13
36c
59
64c
45
‒
NiO–RuRe
11.2
−0.7
361 (32)c
111 (11)
n.d.
2
NiO–RuRe
NiO–RuRe
NiO–Ru
10.3
9.6
11.5
15.9
10.2
9.4
−0.3
0
10
3
n.d.
21
0
4
8
12
Time / h
4
−0.7
−0.7
−0.7
−0.7
n.d.
21
‒
Figure 2. Time course of gaseous products during irra-
diation of NiO–RuRe photocathode at −0.7 V vs. Ag/AgCl
5
NiO–Re
n.d.
n.d.
n.d.
11
6d
7e
NiO–RuRe
NiO–RuRe
n.d.
‒
under a CO2 atmosphere (λ > 460 nm).
n.d.
56
a
Electrode was irradiated at λ > 460 nm for 5 h in 50-
mM NaHCO3 aq. under a CO2 atmosphere. vs. Ag/AgCl
(sat. KCl). 12 h irradiation. In the dark. Under an Ar
atmosphere.
b
c
d
e
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
The results described above demonstrated that the
acid anchor groups from the metal oxide surfaces has
been reported, and was shown to be accelerated by irradi-
ation.30,31 Hence, suppression of photodesorption of the
metal complex from the electrode should be a way to
improve both the activity and stability of the hybrid pho-
tocathode.
1
2
3
4
5
6
7
8
NiO–RuRe hybrid photocathode is capable of driving
photoelectrochemical CO2 reduction with high selectivity
under visible light in an aqueous solution; however, the
TONCO of the NiO–RuRe photocathode (32) was lower
than the reported value of 130 for the homogeneous sys-
tem that used a sacrificial electron donor in an aqueous
solution.14 In order to investigate the source of this differ-
ence, the electrochemically active amount of RuRe on the
NiO–RuRe electrode was estimated using cyclic voltam-
metry in an acetonitrile solution (Figure 4 left and entry 1
in Table 2). In the first scan, a reversible redox system and
an irreversible oxidation wave were observed, attributable
to the Ru(II/III) of the Ru unit and Re(I/II) of the Re unit,
respectively. The second scan using the same electrode
showed only a reversible wave, consistent with previous
reports that indicate oxidation of [ReI(diimine)(CO)3X]+-
type complex to induce decomposition of the complex.28
Therefore, only the Ru unit should be electrochemically
active during the second scan. Additional scans did not
affect the shape or area of the reversible redox peak. Ow-
ing to the simplicity of the CV wave, we chose the second
wave for estimating the amount of the electrochemically
active RuRe, which should be almost the same as the
remaining Ru unit in the second scan. The amount of
electrochemically active RuRe on the NiO electrode was
estimated to be 6.5 nmol, lower than the total adsorbed
amount of RuRe on the NiO electrode before the photo-
catalytic reaction (9.7 nmol), indicating only two-thirds of
RuRe in NiO–RuRe was present to accept electrons from
the NiO electrode; in other words, only two-thirds of the
starting complexes is active for the photocatalytic CO2
reduction. Cyclic voltammetry using electrode after the
photoelectrochemical reaction for 5 h was also conducted
using the same procedure (Figure 4 right and entry 2 in
Table 2).29 The amount of electrochemically active RuRe
was reduced to 2.9 nmol. These results suggest that one-
third of the total amount of RuRe immobilized on the
NiO electrode are natively inactive, even before irradia-
tion, likely due to low conductivity in some areas of the
NiO electrode, and half of the active RuRe is lost to pho-
tocatalysis during irradiation for 5 h. XPS measurements
were also conducted to reveal the electronic state and
abundance ratio of elements on the NiO–RuRe surface
before and after the photoelectrochemical reaction. (Fig-
ure S3). No obvious change of the electronic state was
observed. Unfortunately, the peak of the Ru 3d partially
overlapped with that of the carbon 1s (285 eV), making
assignment difficult, but the Re/Ni ratio decreased from
0.028 to 0.013 after the photoelectrochemical reaction.
The decrease of the Re complex on the electrode was
consistent with that of the Ru complex obtained by using
the cyclic voltammetry as discussed above. ATR-IR meas-
urement confirmed that the tricarbonyl structure of Re
unit on the electrode was maintained after the photoelec-
trochemical reaction; however, the strength of the band
became weaker (Figure S4). Therefore, desorption of
RuRe from the NiO surface and photochemical decom-
position of RuRe are likely sources for the relatively low
TONCO. Desorption of metal complexes with phosphonic
1st scan
2nd scan
1st scan
2nd scan
600
400
200
0
600
400
200
0
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ200
ꢀ400
ꢀ200
ꢀ400
0.4 0.6 0.8 1.0 1.2
Potential / V vs. Ag/AgNO3
0.4 0.6 0.8 1.0 1.2
Potential / V vs. Ag/AgNO3
Figure 4. Cyclic voltammograms of NiO–RuRe (left)
and after reaction (right) in an Ar-purged CH3CN contain-
ing 0.1 M Et4NBF4 as electrolyte. Scan rate was 50 mV s−1.
Table 2. Estimated amounts of RuRe on the NiO–
RuRe electrodes.
Estimated amount of RuRe/nmol
(CV)/
(Absorb-
ance)
Adsorbed
(from absorb-
ance)
Electrochemi-
cally active
(from CV)
Entry
1 (as pre-
pared)
9.7
9.2
6.5
0.67
0.32
2
(after
2.9
reaction)
In order to construct a full cell with the NiO–RuRe
photocathode, photoelectrochemical water oxidation
using a CoOx/TaON photoanode, which has been report-
ed as an efficient photoanode for the water splitting reac-
tion,25 was examined in an aqueous solution containing
sodium bicarbonate (50 mM, pH = 8.3). The CoOx/TaON
electrode was prepared as previously reported.25
Figure 5 shows the current–potential curve of the
CoOx/TaON photoanode in a NaHCO3 aqueous solution
under an Ar atmosphere. The anodic photocurrent, which
corresponds to water oxidation, was observed at a poten-
tial more positive than −0.7 V vs. Ag/AgCl, which is al-
most equivalent to the reported value measured in a
Na2SO4 aqueous solution (pH8).25 Under these reaction
conditions, formation of O2 was observed in the liquid
phase, which was detected using an O2 sensor, and its
total Faraday efficiency was 89% after 5 min of irradiation
(Figure S5). Of note, we limited the time of the photoelec-
trochemical reaction to 5 min to avoid contamination
from air. During this period, most of the evolved O2 re-
mained in the liquid phase. Based on these data, we con-
cluded that the CoOx/TaON photoanode is active for
water oxidation even in a NaHCO3 aqueous solution.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
77 nmol with 68% Faradaic efficiency after 60 min of
600
500
400
300
200
100
0
20 ꢁA cmꢀ2
1
2
3
4
5
6
7
8
10 mV secꢀ1
irradiation. Note that the O2 detected in the liquid phase
decreased after irradiation stopped, as shown in Figure 7.
This indicates that diffusion of the O2 from the liquid
phase to the gas phase proceeds but the rate was slow,
even after 60 min of irradiation (38 nmol h−1). The pro-
duced CO was simultaneously detected in the gas phase
in the NiO–RuRe photocathode chamber; 79 nmol of CO
was evolved for 60 min of irradiation, corresponding to a
TONCO = 17 based on the amount of immobilized RuRe
before irradiation. Total Faradaic efficiency of the cathod-
ic reaction, including H2 evolution (6 nmol), was 37%.
The difference of the Faradaic efficiencies between the
anodic and cathodic reactions could be mainly caused by
the reduction of trivalent nickel ions (Ni3+) that originally
existed in NiO on the photocathode, observed in similar
photoelectrochemical reactions in both organic and
aqueous solutions.23,26 The photoelectrochemical cell
without the RuRe photocatalyst did not produce any
reduction products of CO2 in the photocathode chamber
and increase of O2 in the photoanode chamber was very
small (<20 nmol) for 60 min irradiation under the same
reaction condition.
ꢀ0.8 ꢀ0.6 ꢀ0.4
λ > 400 nm
ꢀ100
ꢀ0.8
9
ꢀ0.4
0.0
0.4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Potential / V vs. Ag/AgCl
Figure 5. Current–potential curve of CoOx/TaON pho-
toanode with intermittent irradiation (λ > 400 nm) in an
Ar-purged 50 mM NaHCO3 aqueous solution (pH = 8.3).
Figure 6 shows a schematic representation of the hybrid
photoelectrochemical cell consisting of the NiO–RuRe
photocathode and the CoOx/TaON photoanode, with a Z-
scheme configuration for CO2 reduction using water as a
reductant. In the photoelectrochemical cell, the pho-
toanode was separated from the photocathode using a
Nafion® membrane. During the reaction, both CO and H2
in the gas phase of the photocathode chamber were de-
tected using micro gas chromatography. The evolved O2
in the liquid phase of the photoanode chamber was de-
tected using the O2 sensor; however, the exact O2 concen-
tration in the gas phase could not be determined because
of small amounts of contamination of air from outside the
chamber. Visible light irradiation (λ > 400 nm) was con-
ducted from the backside of the NiO–RuRe photocath-
ode; transmitted light penetrated through the NiO–RuRe
and the Nafion membrane to be irradiated to the
CoOx/TaON photoanode. A bias of −0.3 V vs. CoOx/TaON
was applied to NiO–RuRe using a potentiostat to acceler-
ate the reaction. We note that there was also a chemical
bias of 0.10 V from the pH difference between two elec-
trodes.
140
On
eꢀ/4
O2
Off
50
40
30
20
10
0
120
100
80
60
40
20
0
On
Off
Off On
Off
On
0
0
20 40 60 80
Time / min
20 40 60 80
Time / min
Figure 7. Time course of photocurrent of hybrid photo-
electrochemical cell comprising NiO–RuRe and
CoOx/TaON under visible light irradiation (λ > 400 nm)
(left) and amount of evolved oxygen in liquid phase
(right). A combined electrical (0.3 V) and chemical bias
(0.10 V) was applied.
An isotope tracer experiment using water containing
30% H218O and NaHCO3 (50 mM) as the solvent was con-
ducted. This solution was added to the anode chamber of
the cell, and after bubbling He into the anode chamber
and CO2 into the cathode chamber for 30 min, both elec-
trodes were irradiated with visible light for 60 min with a
bias of −0.3 V using the same system described above
(Figure 6). The products in the gas phase of the anode
chamber were analyzed by GC–MS. The results are sum-
marized in Figure 8. The ratios of 16O18O and 18O2 were
18
significantly larger when using 30% H2 O (16O18O: 1.9%,
18O2: 0.38%) than those for the reaction using non-labeled
water (16O18O: 0.44%, 18O2: <0.01%); however, they were
different than the theoretically expected ratios if all the
O2 was produced from water (16O18O: 42%, 18O2: 9%) with-
out any exchange of the oxygen atoms with other sub-
strates during the photoelectrochemical reaction. In-
volvement of O atoms from the CoOx oxygen evolution
Figure 6. Schematic image of hybrid photoelectro-
chemical cell with Z-scheme configuration.
Figure 7 shows the time course of the photocurrent and
observed amounts of O2 in the reaction solution in the
chamber of the CoOx/TaON photoanode with intermit-
tent irradiation. Evolution of O2 was observed only when
the photoelectrochemical cell was irradiated, reaching
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
photocathode catalytically generated CO with high selec-
catalyst is a possible explanation for such a result.32 Ex-
change of NaHC16O3 by 18O from H218O, should also pro-
ceed during the photoelectrochemical reaction, and con-
tamination from air when the syringe was used for sam-
pling the gas phase in the anode chamber through the
septum cannot be excluded. In photoelectrochemical
experiment (Figure 7), on the other hand, an oxygen sen-
sor in the liquid phase was used for determining the pro-
duced amount of O2 in solution to prevent overestimation
of O2. These results suggest that O2 was, at least partially,
generated by a water oxidation reaction driven by visible
light irradiation.
tivity in aqueous solution. The hybrid photoelectrochemi-
cal cell consisting of the NiO–RuRe hybrid photocathode
and the CoOx/TaON photoanode showed activity for
visible-light-driven catalytic reduction of CO2 and oxida-
tion of water to generate CO and O2. To the best of our
knowledge, this is the first example of a visible-light-
driven CO2 reduction system using water as the reduct-
ant, based on molecular photocatalysts hybridized with
semiconductor photocatalysts.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information.
in H2O
in 30 % H218O
Time course of photocurrent of NiO-RuRe at various applied
potential, time course of evolved H2 during photoelectro-
chemical CO2 reduction using NiO-RuRe, XPS spectra, ATR-
IR spectra, time courses of photocurrent and amount of
evolved oxygen using a CoOx/TaON, the spectrum of irradi-
ated light, and calculation method of light-energy conversion
efficiency for hybrid photoelectrochemical cell. This material
is available free of charge via the Internet at
http://pubs.acs.org.
16O2
m/z = 32
m/z = 32
16O18
O
m/z = 34
m/z = 34
m/z = 36
×100
×500
×100
×500
18O2
m/z = 36
*
4.2 4.4 4.6 4.8 5.0 5.2
Retention Time / min
4.2 4.4 4.6 4.8 5.0 5.2
Retention Time / min
AUTHOR INFORMATION
Corresponding Author
Figure 8. GC–MS chart of O2 in gas phase of anode
chamber during photoelectrochemical CO2 reduction
using Z-scheme configuration in H2O (left) or 30% H218O
(right). * indicates 36Ar derived from contamination of air
through sampling.
Ryu Abe: ryu-abe@scl.kyoto-u.ac.jp
Osamu Ishitani: ishitani@chem.titech.ac.jp
ACKNOWLEDGMENT
This work was supported by Strategic International Collabo-
rative Research Program (PhotoCAT project) from both
Japan Science and Technology Agency (JST) and the French
National Research Agency (ANR-14-JTIC-0004-01) and
CREST (Molecular Technology project, JST). This is also
partially supported by a Grant-in-Aid for Scientific Research
on Innovative Area “Artificial Photosynthesis (Project
24107005) from the Japan Society for the Promotion of Sci-
ence (JSPS) and the Photon and Quantum Basic Research
Coordinated Development Program (MEXT, Japan). N.K. and
V.A. acknowledge Labex program, ARCANE, ANR-11-LABX-
0003-01 (French National Research Agency). K.M. acknowl-
edges the Noguchi Institute, and the Murata Science Founda-
tion.
Therefore, this photoelectrochemical cell is confirmed
to drive CO2 reduction to obtain CO and O2 with visible
light irradiation (Equation 2).
CO2 → CO + 1/2O2
(Eq. 2)
Despite the use of external electrical and chemical bias-
es, this is the first instance of successful visible-light-
driven CO2 reduction using water as a reductant, accom-
plished by use of a molecular photocatalyst hybridized
with a semiconductor photocatalyst. The light energy
conversion efficiency of the reaction was calculated to be
1.6 × 10−5, considering the electrical bias of 0.3 V and
chemical bias of 0.10 V from the pH difference between
the two electrolytes. The details of the calculation are
described in the Supporting Information. The low per-
formance of the photoelectrochemical system mainly
arises from the photocathode. The NiO electrode sub-
strate itself is problematic because of its low conductivity,
and presence of electrochemically inactive surface areas.
In addition, desorption and photochemical degradation of
RuRe from the NiO surface is the main reason of limited
durability. Researches to address such problems are now
under way in our laboratory.
REFERENCES
(1) Morris, A. J.; Meyer, G. J.; Fujita, E. Acc. Chem. Res. 2009,
42 (12), 1983–1994.
(2) Sahara, G.; Ishitani, O. Inorg. Chem. 2015, 54 (11), 5096–
5104.
(3) Yamazaki, Y.; Takeda, H.; Ishitani, O. J. Photochem. Photo-
biol. C Photochem. Rev. 2015, 2–12.
(4) Gholamkhass, B.; Mametsuka, H.; Koike, K.; Tanabe, T.;
Furue, M.; Ishitani, O. Inorg. Chem. 2005, 44 (7), 2326–2336.
(5) Sato, S.; Koike, K.; Inoue, H.; Ishitani, O. Photochem. Pho-
tobiol. Sci. 2007, 6 (4), 454–461.
(6) Koike, K.; Naito, S.; Sato, S.; Tamaki, Y.; Ishitani, O. J. Pho-
tochem. Photobiol. A Chem. 2009, 207, 109–114.
(7) Tamaki, Y.; Morimoto, T.; Koike, K.; Ishitani, O. Proc. Natl.
Acad. Sci. 2012, 109 (39), 15673–15678.
(8) Tamaki, Y.; Watanabe, K.; Koike, K.; Inoue, H.; Morimoto,
T.; Ishitani, O. Faraday Discuss. 2012, 155, 115–127.
CONCLUSION
We successfully developed a NiO–RuRe hybrid photo-
cathode for photoelectrochemical CO2 reduction in an
aqueous solution and constructed a hybrid photoelectro-
chemical cell with a CoOx/TaON photoanode that can
drive water oxidation. Using visible light, the NiO–RuRe
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 9
(9) Tamaki, Y.; Koike, K.; Morimoto, T.; Ishitani, O. J. Catal.
2013, 304, 22–28.
(25) Higashi, M.; Domen, K.; Abe, R. J. Am. Chem. Soc. 2012,
134 (16), 6968–6971.
1
2
3
4
5
6
7
8
(10) Tamaki, Y.; Koike, K.; Morimoto, T.; Yamazaki, Y.;
Ishitani, O. Inorg. Chem. 2013, 52 (20), 11902–11909.
(11) Kato, E.; Takeda, H.; Koike, K.; Ohkubo, K.; Ishitani, O.
Chem. Sci. 2015, 6 (5), 3003–3012.
(12) Tamaki, Y.; Koike, K.; Ishitani, O. Chem. Sci. 2015, 6 (12),
7213–7221.
(13) Nakada, A.; Koike, K.; Nakashima, T.; Morimoto, T.;
Ishitani, O. Inorg. Chem. 2015, 54 (4), 1800–1807.
(14) Nakada, A.; Koike, K.; Maeda, K.; Ishitani, O. Green Chem.
2016, 18 (1), 139–143.
(15) Sekizawa, K.; Maeda, K.; Domen, K.; Koike, K.; Ishitani, O.
J. Am. Chem. Soc. 2013, 135 (12), 4596–4599.
(16) Yoshitomi, F.; Sekizawa, K.; Maeda, K.; Ishitani, O. ACS
Appl. Mater. Interfaces 2015, 7 (23), 13092–13097.
(17) Kuriki, R.; Matsunaga, H.; Nakashima, T.; Wada, K.;
Yamakata, A.; Ishitani, O.; Maeda, K. J. Am. Chem. Soc. 2016, 138
(15), 5159–5170.
(18) Muraoka, K.; Kumagai, H.; Eguchi, M.; Ishitani, O.;
Maeda, K. Chem. Commun. 2016, 52 (50), 7886–7889.
(19) Nakada, A.; Nakashima, T.; Sekizawa, K.; Maeda, K.;
Ishitani, O. Chem. Sci. 2016, 7 (7), 4364–4371.
(20) Sato, S.; Arai, T.; Morikawa, T.; Uemura, K.; Suzuki, T. M.;
Tanaka, H.; Kajino, T. J. Am. Chem. Soc. 2011, 133 (39), 15240–
15243.
(26) Kaeffer, N.; Massin, J.; Lebrun, C.; Renault, O.; Chavarot-
Kerlidou, M.; Artero, V. J. Am. Chem. Soc. 2016, 138 (38), 12308–
12311.
(27) Kuramochi, Y.; Ishitani, O. Inorg. Chem. 2016, 55 (11),
5702–5709.
(28) Bullock, J. P.; Carter, E.; Johnson, R.; Kennedy, A. T.; Key,
S. E.; Kraft, B. J.; Saxon, D.; Underwood, P. Inorg. Chem. 2008, 47
(17), 7880–7887.
(29) We note that there is no oxidation peak corresponding to
the Re unit for the NiO–RuRe after the photoelectrochemical
reaction, while that for as prepared unit was clearly observed in
Figure 4 (left). The results suggest that a large positive shift of
the oxidation potential of the Re unit was derived from ligand
substitution. It could be generally considered that the Re catalyst
(Re(N^N)(CO)3Br) forms coordinately unsaturated intermedi-
ates by dissociation of the Br ligand during the reaction and
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
could
terminally
change
to
solvent
coordinates
([Re(N^N)(CO)3L]n+).he positive shift in the oxidation potential
in Figure 4 (right) was possibly caused by formation of coordi-
nate compounds with acetonitrile ([Re(N^N)(CO)3MeCN]+),
which has lower energy for the t2g orbital compared to the origi-
nal complex, through cyclic voltammetry in an acetonitrile solu-
tion.
(30) Hanson, K.; Brennaman, M. K.; Luo, H.; Glasson, C. R. K.;
Concepcion, J. J.; Song, W.; Meyer, T. J. ACS Appl. Mater. Inter-
faces 2012, 4 (3), 1462–1469.
(21) Arai, T.; Sato, S.; Kajino, T.; Morikawa, T. Energy Environ.
Sci. 2013, 6 (4), 1274.
(22) Arai, T.; Sato, S.; Morikawa, T. Energy Environ. Sci. 2015, 8
(7), 1998–2002.
(23) Sahara, G.; Abe, R.; Higashi, M.; Morikawa, T.; Maeda, K.;
Ueda, Y.; Ishitani, O. Chem. Commun. 2015, 51 (53), 10722–10725.
(24) Abe, R.; Higashi, M.; Domen, K. J. Am. Chem. Soc. 2010,
132 (34), 11828–11829.
(31)
Hanson, K.; Brennaman, M. K.; Ito, A.; Luo, H.; Song,
W.; Parker, K. A.; Ghosh, R.; Norris, M. R.; Glasson, C. R. K.;
Concepcion, J. J.; Lopez, R.; Meyer, T. J. J. Phys. Chem. C 2012, 116
(28), 14837–14847.
(32) Koroidov, S.; Anderlund, M. F.; Styring, S.; Thapper, A.;
Messinger, J. Energy Environ. Sci. 2015, 8 (8), 2492–2503.
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC
9
ACS Paragon Plus Environment