Synergistic Cocatalytic Effect of Carbon Nanodots and Co3O4 Nanoclusters for the Photoelectrochemical Water Oxidation on Hematite

Article abstract of DOI:10.1002/anie.201600918

Cocatalysis plays an important role in enhancing the activity of semiconductor photocatalysts for solar water splitting. Compared to a single cocatalyst configuration, a cocatalytic system consisting of multiple components with different functions may realize outstanding enhancement through their interactions, yet limited research has been reported. Herein we describe the synergistic cocatalytic effect between carbon nanodots (CDots) and Co₃O₄, which promotes the photoelectrochemical water oxidation activity of the Fe₂O₃ photoanode with a 60 mV cathodically shifted onset potential. The C/Co₃O₄-Fe₂O₃ photoanode exhibits a photocurrent density of 1.48 mA cm^-2 at 1.23 V (vs. reversible hydrogen electrode), 78 % higher than that of the bare Fe₂O₃ photoanode. The slow reaction process on the single Co^III-OH site of the Co₃O₄ cocatalyst, oxidizing H₂O to H₂O₂ with two photogenerated holes, could be accelerated by the timely H₂O₂ oxidation to O₂ catalyzed on CDots.

Full text of DOI:10.1002/anie.201600918

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201600918
German Edition: DOI: 10.1002/ange.201600918
Water Splitting
Synergistic Cocatalytic Effect of Carbon Nanodots and Co3O4
Nanoclusters for the Photoelectrochemical Water Oxidation on
Hematite
Peng Zhang+, Tuo Wang+, Xiaoxia Chang, Lei Zhang, and Jinlong Gong*
Abstract: Cocatalysis plays an important role in enhancing the
activity of semiconductor photocatalysts for solar water
splitting. Compared to a single cocatalyst configuration,
a cocatalytic system consisting of multiple components with
different functions may realize outstanding enhancement
through their interactions, yet limited research has been
reported. Herein we describe the synergistic cocatalytic effect
between carbon nanodots (CDots) and Co3O4, which promotes
the photoelectrochemical water oxidation activity of the Fe2O3
photoanode with a 60 mV cathodically shifted onset potential.
The C/Co3O4-Fe2O3 photoanode exhibits a photocurrent den-
sity of 1.48 mAcmÀ2 at 1.23 V (vs. reversible hydrogen
electrode), 78% higher than that of the bare Fe2O3 photoanode.
The slow reaction process on the single CoIII-OH site of the
Co3O4 cocatalyst, oxidizing H2O to H2O2 with two photo-
generated holes, could be accelerated by the timely H2O2
oxidation to O2 catalyzed on CDots.
iridium and ruthenium oxides, cobalt-based materials are
earth-abundant and cost-effective.[7] Cobalt phosphate (Co-
Pi) showed a prominent promotive effect on the activity of
hematite (a-Fe2O3) for photoelectrochemical water oxidation
with a substantial cathodic onset potential shift and an
enhanced photocurrent density.[8] It was demonstrated that
the Co-Pi could increase the lifetime of photogenerated holes
and retard the recombination of charge carriers.[9] The
photocatalytic water-oxidation activity of Fe2O3 photoanode
could also be effectively improved by integration of an
inorganic Co3O4 cocatalyst.[10]
Metal-free materials are also investigated as cocatalysts
for solar water splitting. Graphitic carbon nanodots (CDots)
were demonstrated to enhance the photocatalytic water-
oxidation activates of TiO2 nanotube arrays,[11] ZnO nanowire
arrays,[12] and Ag3PW12O40 photocatalyst.[13] Most recently,
a CDots/C3N4 composite photocatalyst showed an outstand-
ing activity for solar water splitting with an overall solar-
energy conversion efficiency of 2.0%.[14] It was proposed that
the CDots performed as highly active catalysts for the
decomposition of H2O2, which was considered as the inter-
mediate spices from a two-step-two-electron water-splitting
process on C3N4. Comparing with the one-step-four-electron
water-splitting route, such a reaction pathway is kinetically
favored, which leads to the high activity of the CDots/C3N4
photocatalyst.
Cocatalysts ranging from metal oxides to metal-free
materials have been widely studied for improving the activity
of semiconductor photocatalysts. However, relevant research
mostly focuses on single-component cocatalyst, whose coca-
talytic effect is restricted to the physical, chemical and
electronic properties of the material. As complex processes
are involved in the water-oxidation reaction, single cocatalyst
can hardly realize the acceleration of all these processes
simultaneously. Thus, to achieve better a performance it is
desirable to fabricate a cocatalytic system with different
functions for synergistic enhancement. Fe2O3 is selected to
demonstrate the effectiveness of cocatalyst systems owing to
its notoriously slow surface reaction kinetics for water
oxidation.[15] By comparing the activity of Fe2O3 photoanodes
with different cocatalysts (i.e., CDots, Co3O4, and both of
them), we demonstrated that Co3O4 and CDot cocatalysts
showed a synergistic cocatalytic effect. Ex situ detection of
H2O2 from electrolyte was conducted to investigate the
mechanism of the effect.
H
ydrogen production via solar water splitting is one of the
most promising approaches to develop sustainable energy
resources.[1] Semiconductor photocatalysts with suitable
band-gap structures, good charge-carrier conductivities, and
fast surface reaction kinetics are desired to achieve high solar-
energy conversion efficiencies.[2] The surface reaction kinetics
of photocatalysts could be enhanced by integrating cocata-
lysts, which can provide surface reaction sites with lower
overpotentials.[3] Moreover, cocatalysts could act as selective
trapping sites for photogenerated electrons/holes and sup-
press the recombination of charge carriers.[4] Thus, strategies
of cocatalysts loading have been extensively developed to
enhance the activity of semiconductor photo-absorbers with
slow surface reaction kinetics. Much attention has been paid
to the water oxidation half reaction since it is regarded as the
rate determining and kinetically unfavored process (with four
electrons involved to produce one O2 molecule).[5]
Cobalt-based compounds have been developed as coca-
talysts for photocatalytic water oxidation.[6] Comparing with
[*] Dr. P. Zhang,[+] Prof. T. Wang,[+] X. Chang, Prof. L. Zhang, Prof. J. Gong
Key Laboratory for Green Chemical Technology of Ministry of
Education, School of Chemical Engineering and Technology, Tianjin
University, Collaborative Innovation Center of Chemical Science and
Engineering
Tianjin 300072 (China)
E-mail: jlgong@tju.edu.cn
[+] These authors contributed equally to this work.
The Fe2O3 photoanode was prepared by a previously
reported hydrothermal approach (see Supporting Informa-
tion for experimental details).[16] The Fe2O3 photoanode
obtained shows a one-dimensional wormlike structure (Fig-
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under http://dx.doi.org/10.
1002/anie.201600918.
Angew. Chem. Int. Ed. 2016, 55, 5851 –5855
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5851

Angewandte
Communications
Chemie
Figure 2. TEM images of C/Co3O4–Fe2O3 photoanodes at a) low and
b) high magnification. White circles in (a) indicate CDots.
(Figure S4). As indicated by X-ray photoelectron spectrosco-
py (XPS), the O1s peaks at 531.2 eV for C-Fe2O3 and C/
Figure 1. Scanning electron microscopy (SEM) images of a) Fe2O3,
b) C-Fe2O3, c) Co3O4–Fe2O3, and d) C/Co3O4–Fe2O3 photoanodes.
À
=
[18]
Co3O4–Fe2O3 photoanodes correspond to the -C OH, -C O,
= À
and O C O- groups on the surface of CDots (Figure S5a).
The XPS peaks at 796.4 eVand 780.4 eV for Co3O4–Fe2O3 and
C/Co3O4–Fe2O3 photoanodes corresponds to the Co2p1/2 and
Co2p3/2 spin-orbits in Co3O4, respectively (Figure S5c).[19] The
peaks at 724.1 eV and 710.6 eV represent the Fe2p1/2 and
Fe2p3/2 spin-orbits in Fe2O3,[20] which exhibit lower peak
intensities without any peak shift upon the integration of
cocatalysts (Figure S5b). This phenomenon indicates that the
cocatalysts partly cover the surface of Fe2O3 photoanodes.
The activities for photocatalytic water oxidation of differ-
ent photoanodes were examined by measuring the photo-
current density versus bias (J–V) curves in NaOH electrolyte
(1m, pH 13.6) under simulated air mass 1.5 global (AM 1.5G)
irradiation at 100 mWcmÀ2 (Figure 3a). The applied bias
photon-to-current conversion efficiency (ABPE) and incident
photon-to-current conversion efficiency (IPCE) of different
photoanodes were also measured (Figure 3b,c). The IPCE
spectra were collected with a bias of 1.23 V (vs. reversible
hydrogen electrode (RHE)) and integrated over the AM
1.5G spectrum to obtain photocurrent densities, which are
close to the values from J–V curves, confirming the accuracy
of the results (Table S1).[21]
The as-prepared Fe2O3 photoanode exhibits a photocur-
rent density of 0.83 mAcmÀ2 at 1.23 V (vs. RHE). The onset
potential (determined as the crossing point of a tangent to the
inflection part of the photocurrent density with the dark
current[22]) is about 0.85 V (vs. RHE), which is relatively low
as a result of the partial removal of surface states by the high-
temperature annealing.[23] The ABPE of Fe2O3 is 0.067% at
1.0 V (vs. RHE). The IPCE of Fe2O3 is about 20% in the 325–
375 nm range. CDot cocatalysts do not show an obvious
promotive effect, as the photocurrent density of C-Fe2O3
photoanode at 1.23 V (vs. RHE) does not change, and there
is little change in the ABPE and the IPCE. However, the
onset potential of the C-Fe2O3 photoanode shows a cathodic
shift of 40 mV, which may be caused by the high conductivity
of the CDots. The Co3O4 cocatalysts can enhance the activity
of the Fe2O3 photoanode, leading to a photocurrent density of
1.23 mAcmÀ2 at 1.23 V (vs. RHE) as well as a 20 mV cathodic
shift of the onset potential. The ABPE and the IPCE of
Co3O4–Fe2O3 photoanode are 0.14% and 25% (at 375–
ure 1a). The thickness of the Fe2O3 films is 224 Æ 22 nm
(Figure S1 in the Supporting Information). Upon the inte-
gration of CDots, the surface of Fe2O3 becomes slightly
rougher (C-Fe2O3; Figure 1b). However, CDot particles
could not be observed since their size is extremely small.
Particulate Co3O4 is observed on the surface of the Co3O4–
Fe2O3 photoanode (Figure 1c). When both CDots and Co3O4
were loaded onto the Fe2O3 photoanode its surface became
rough with particles clearly shown on the surface (Figure 1d),
which indicates the successful preparation of C/Co3O4–Fe2O3
photoanode.
Transmission electron microscopy (TEM) images were
obtained to better illustrate the configurations of the coca-
talysts on different photoanodes. CDot with a lattice spacing
of 0.202 nm (corresponding to the (101) lattice planes of
graphitic carbon) was tightly attached on the surface of C-
Fe2O3 photoanode (Figure S2a). The CDot exhibits a size of
about 5 nm. Fe2O3 is well crystalized showing a clear lattice
with a spacing of 0.262 nm (corresponding to the (104) lattice
planes of Fe2O3). The Co3O4 cocatalysts with a lattice spacing
of 0.293 nm (corresponding to the (220) lattice planes of
Co3O4) were loaded on Co3O4–Fe2O3 photoanode (Fig-
ure S2b). The Co3O4 particles have a size of about 8 nm. On
C/Co3O4–Fe2O3 photoanode, CDots with a lattice spacing of
0.202 nm and Co3O4 with a lattice spacing of 0.244 nm
(corresponding to the (311) lattice planes of Co3O4) are
distributed without aggregation on the surface (indicated with
white circles in Figure 2).[14] Moreover, the random loading
leads to the two types of cocatalysts being adjacent to each
other (Figure 2b), which accounts for their synergistic coca-
talytic effect.
Only the characteristic X-ray diffraction (XRD) peaks of
hematite at 36.38 and 64.68 could be observed for different
samples owing to the relatively small sizes and low loading
amounts of the cocatalysts (Figure S3).[17] However, the
ultraviolet/visible (UV/Vis) light absorption spectra of differ-
ent photoanodes reveal that the loading of CDots could
enhance the light absorption of Fe2O3 photoanodes, while
Co3O4 could only slightly improve the light-absorption ability
5852
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 5851 –5855

Angewandte
Communications
Chemie
by the lower charge-transfer resistance of Co3O4–Fe2O3
photoanode, which is in accordance with the J–V curves.
When both CDots and Co3O4 cocatalysts are loaded on to
Fe2O3, the value of Rct decreases further, demonstrating
their synergistic cocatalytic effect.
As the synergistic cocatalytic effect between CDots
and Co3O4 has been demonstrated through the activity
tests and the EIS characterization, it is worthwhile
understanding the mechanism of such a synergistic
effect. The intermediates of water oxidation on Co3O4
had been investigated through time-resolved Fourier-
transform infrared spectroscopy, suggesting that two kinds
of reaction sites exist on the surface.[26] The fast-reaction
site, containing two linked CoIII–OH groups, could
directly oxidize H2O to molecular O2 with four photo-
generated holes (Figure 4a). On the other hand, the single
=
CoIII–OH slow-reaction site is initially oxidized to CoIV
O
by one photogenerated hole. After the attack and
Figure 3. a) J–V curves, b) applied bias photon-to-current conversion effi-
ciency (ABPE), c) incident photon-to-current conversion efficiency (IPCE),
and d) electrochemical impedance spectroscopy (EIS) in Nyquist plots of
Fe2O3, C-Fe2O3, Co3O4–Fe2O3, and C/Co3O4–Fe2O3 photoanodes. The sym-
bols shows the measured data and the lines shows the fitting results in (d).
Inset of (d) shows the equivalent circuit of the photoanodes.
deprotonation of H2O molecule, the slow-reaction site
II
À
changes to Co OOH. Then another H2O molecule in the
II
electrolyte reacts with Co OOH to form H2O2,[27] when
À
II
À
Co OOH is oxidized by the second photogenerated hole
to restore the CoIII OH site. Thus, a complete cycle is
À
established to oxidize H2O to H2O2 at the slow-reaction
site of Co3O4. Recently, CDots have been shown to be an
325 nm), respectively. When CDots and Co3O4 are loaded
simultaneously, the activity is extensively enhanced. The
photocurrent density of the C/Co3O4–Fe2O3 photoanode is
1.48 mAcmÀ2 at 1.23 V (vs. RHE), which is 78% higher than
Table 1: EIS Fitting results of Rtrap and Rct for Fe2O3, C-Fe2O3, Co3O4–
Fe2O3, and C/Co3O4–Fe2O3 photoanodes.
Sample
Fe2O3
C-Fe2O3
Co3O4–Fe2O3
C/Co3O4–Fe2O3
that of the bare Fe2O3 photoanode. This enhancement is the
largest reported value at 1.23 V (vs. RHE) for the cocatalyst
strategy in non-sacrificial photoelectrochemical water oxida-
tion by Fe2O3 photoanodes. Moreover, the onset potential
shows a cathodic shift of 60 mV, indicating a lower over-
potential for water oxidation as a result of the synergistic
cocatalytic effect. The onset potential of the Fe2O3 photo-
anode would be further reduced by the optimization of the
loading ratios and the amounts of different cocatalysts as well
as the strengthening of the connection between the cocata-
lysts and the Fe2O3 photoanode. The ABPE of the C/Co3O4–
Fe2O3 photoanode is 0.20% at 1.0 V (vs. RHE), which is
2.46 times that of the bare Fe2O3 photoanode. Moreover, the
IPCE is about 28% in the 375–325 nm range.
Rtrap [W]
Rct [W]
670Æ10
592Æ139
690Æ147
600Æ10
417Æ12
678Æ43
230Æ37
1330Æ13
Electrochemical impedance spectroscopy (EIS) charac-
terization was conducted under AM 1.5G (Figure 3d). Two
clear semicircles are observed for the bare Fe2O3 photoanode,
indicating the existence of two capacitances.[24] Thus, a typical
equivalent circuit model was used to fit the EIS data (inset of
Figure 3d), which involved a series resistors (Rs), a surface
states charge trapping resistance (Rtrap), a charge transfer
resistance (Rct), a depletion layer capacitance (CPEbulk), and
a surface states trapped charge capacitance (CPEss). Constant
phase elements were used owning to the non-ideal feature of
the capacitances caused by the nanostructure of the electro-
des.[25] The fitting results of Rtrap for different photoanodes are
similar (Table 1), because the Fe2O3 substrates of different
samples are alike. However, the values of Rct change
dramatically when Fe2O3 is loaded with cocatalysts. Co3O4
shows a better cocatalytic effect than the CDots, as reflected
Figure 4. a) Schematic illustration of the fast/slow reaction processes
on Co3O4 cocatalyst and the two-step-two-electron reaction pathway for
photocatalytic water oxidation on the C/Co3O4-Fe2O3 photoanode.
b) Photoluminescence (PL) spectra of the scopoletin assay of the
electrolytes after solar water-splitting reactions with different photo-
anodes.
Angew. Chem. Int. Ed. 2016, 55, 5851 –5855
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5853

Angewandte
Communications
Chemie
excellent catalyst for H2O2 decomposition.[14,28] Therefore, the
Acknowledgements
H2O2 formed at the slow-reaction site on Co3O4 is further
oxidized to molecular O2 on CDots, which results in the two-
step-two-electron reaction pathway for water oxidation (Fig-
ure 4a). The prompt consumption of H2O2 in the two-step-
two-electron reaction pathway causes the acceleration of the
slow-reaction process on Co3O4. Comparing with the four-
electron water-oxidation reaction, the two-step-two-electron
reaction pathway is kinetically favored, requiring a lower
concentration of photogenerated holes on the surface.[29] This
two-step-two-electron reaction pathway explains the syner-
gistic cocatalytic effect between CDots and Co3O4. To confirm
this hypothesis, ex situ detection of the H2O2 in the reaction
electrolyte was conducted.
We thank the National Natural Science Foundation of China
(21525626, U1463205, 21222604, and 51302185), Specialized
Research Fund for the Doctoral Program of Higher Educa-
tion (20120032110024 and 20130032120018), the Scientific
Research Foundation for the Returned Overseas Chinese
Scholars (MoE), the Program of Introducing Talents of
Discipline to Universities (B06006), and the Natural Science
Foundation of Tianjin City (13JCYBJC37000) for financial
support.
Keywords: carbon nanodots · Co3O4 · cocatalysts · iron ·
water splitting
As the amount of H2O2 in the electrolyte is low, a sensitive
chemiluminescence detection method by scopoletin assay was
performed.[30] The fluorescent scopoletin could be oxidized by
H2O2 at the presence of horseradish peroxidase, leading to the
loss of its fluorescence.[31] Thus, the higher intensity of the
emission peak for scopoletin at 430 nm in the photolumines-
cence (PL) spectra indicates a lower concentration of H2O2 in
the reaction electrolyte, and vice versa (Figure 4b). Compar-
ing the results between the reference (experiment without
electrodes, i.e., only electrolyte was irradiated by light) and
the Fe2O3 photoanode, it can be inferred that a small amount
of H2O2 is produced during the photoelectrochemical water
oxidation on Fe2O3. When only CDot cocatalyst is loaded on
Fe2O3, the increased intensity of the emission peak indicates
that the H2O2 produced could be decomposed in the presence
of CDots, confirming their catalytic effect for H2O2 decom-
position. However, in the case of the Co3O4–Fe2O3 photo-
anode, much more H2O2 is detected to quench the fluores-
cence, confirming the existence of the slow-reaction site on
Co3O4. When both CDots and Co3O4 are loaded on the Fe2O3
photoanode, the intensity of the emission peak rises back to
a similar level as for bare Fe2O3 photoanode, which shows that
the H2O2 produced on the Co3O4 cocatalyst is further oxidized
on the CDot cocatalysts. Thus, the two-step-two-electron
reaction path way and the synergistic cocatalytic effect of
CDots and Co3O4 are confirmed by this chemiluminescence
detection of H2O2 produced during photoelectrochemical
water oxidation.
How to cite: Angew. Chem. Int. Ed. 2016, 55, 5851–5855
Angew. Chem. 2016, 128, 5945–5949
[1] a) M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q.
Mi, E. A. Santori, N. S. Lewis, Chem. Rev. 2010, 110, 6446 – 6473;
b) X. Chen, S. Shen, L. Guo, S. S. Mao, Chem. Rev. 2010, 110,
6503 – 6570; c) T. Wang, J. Gong, Angew. Chem. Int. Ed. 2015, 54,
10718 – 10732; Angew. Chem. 2015, 127, 10866 – 10881.
[2] a) Y. Ma, X. Wang, Y. Jia, X. Chen, H. Han, C. Li, Chem. Rev.
2014, 114, 9987 – 10043; b) P. Zhang, J. Zhang, J. Gong, Chem.
Soc. Rev. 2014, 43, 4395 – 4422.
[3] J. Ran, J. Zhang, J. Yu, M. Jaroniec, S. Z. Qiao, Chem. Soc. Rev.
2014, 43, 7787 – 7812.
[4] J. Yang, D. Wang, H. Han, C. Li, Acc. Chem. Res. 2013, 46, 1900 –
1909.
[5] J. D. Blakemore, R. H. Crabtree, G. W. Brudvig, Chem. Rev.
2015, 115, 12974 – 13005.
[6] X. Chang, T. Wang, P. Zhang, J. Zhang, A. Li, J. Gong, J. Am.
Chem. Soc. 2015, 137, 8356 – 8359.
[7] J. Wang, W. Cui, Q. Liu, Z. Xing, A. M. Asiri, X. Sun, Adv. Mater.
2015, 28, 215 – 230.
[8] a) D. K. Zhong, J. Sun, H. Inumaru, D. R. Gamelin, J. Am.
Chem. Soc. 2009, 131, 6086 – 6087; b) K. J. McDonald, K.-S.
Choi, Chem. Mater. 2011, 23, 1686 – 1693.
[9] B. Klahr, S. Gimenez, F. Fabregat-Santiago, J. Bisquert, T. W.
Hamann, J. Am. Chem. Soc. 2012, 134, 16693 – 16700.
[10] L. Xi, P. D. Tran, S. Y. Chiam, P. S. Bassi, W. F. Mak, H. K.
Mulmudi, S. K. Batabyal, J. Barber, J. S. C. Loo, L. H. Wong, J.
Phys. Chem. C 2012, 116, 13884 – 13889.
In summary, we have demonstrated the synergistic
cocatalytic effect between CDots and Co3O4 to enhance the
activity of the Fe2O3 anode for photoelectrochemical water
oxidation. When the two cocatalysts are loaded onto the
anode, a kinetically favored two-step-two-electron reaction
partway is established. Initially, the water molecule is
oxidized to H2O2 at the slow-reaction site on Co3O4. The
H2O2 produced is then further oxidized to O2 on the CDot
cocatalysts. Thus, the slow reaction process on Co3O4 is
accelerated. The kinetically favored two-step-two-electron
reaction partway and the enhanced surface reaction rate
result in the improved solar water-oxidation activity of the
Fe2O3 photoanodes. The discovery of this synergistic effect
between different cocatalysts can help design effective
cocatalytic systems with multiple functional components, an
effect which is also applicable to other energy conversion
applications.
[11] X. Zhang, F. Wang, H. Huang, H. Li, X. Han, Y. Liu, Z. Kang,
Nanoscale 2013, 5, 2274 – 2278.
[12] C. X. Guo, Y. Dong, H. B. Yang, C. M. Li, Adv. Energy Mater.
2013, 3, 997 – 1003.
[13] J. Liu, H. Zhang, D. Tang, X. Zhang, L. Yan, Y. Han, H. Huang,
Y. Liu, Z. Kang, ChemCatChem 2014, 6, 2634 – 2641.
[14] J. Liu, Y. Liu, N. Liu, Y. Han, X. Zhang, H. Huang, Y. Lifshitz, S.-
T. Lee, J. Zhong, Z. Kang, Science 2015, 347, 970 – 974.
[15] a) K. Sivula, F. Le Formal, M. Grätzel, ChemSusChem 2011, 4,
432 – 449; b) I. Cesar, A. Kay, J. A. Gonzalez Martinez, M.
Grätzel, J. Am. Chem. Soc. 2006, 128, 4582 – 4583; c) Y. Lin, G.
Yuan, S. Sheehan, S. Zhou, D. Wang, Energy Environ. Sci. 2011,
4, 4862.
[16] L. Vayssieres, N. Beermann, S.-E. Lindquist, A. Hagfeldt, Chem.
Mater. 2001, 13, 233 – 235.
[17] B. M. Klahr, A. B. F. Martinson, T. W. Hamann, Langmuir 2011,
27, 461 – 468.
[18] H. Sun, A. Zhao, N. Gao, K. Li, J. Ren, X. Qu, Angew. Chem. Int.
Ed. 2015, 54, 7176 – 7180; Angew. Chem. 2015, 127, 7282 – 7286.
5854
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 5851 –5855

Angewandte
Communications
Chemie
[19] B. Varghese, C. H. Teo, Y. Zhu, M. V. Reddy, B. V. R. Chowdari,
[26] M. Zhang, M. de Respinis, H. Frei, Nat. Chem. 2014, 6, 362 – 367.
A. T. S. Wee, V. B. C. Tan, C. T. Lim, C. H. Sow, Adv. Funct.
Mater. 2007, 17, 1932 – 1939.
[27] G. Duca, Homogeneous Catalysis with Metal Complexes,
Springer, Berlin, 2012.
[20] R. Morrish, M. Rahman, J. M. D. MacElroy, C. A. Wolden,
ChemSusChem 2011, 4, 474 – 479.
[28] H. Sun, N. Gao, K. Dong, J. Ren, X. Qu, ACS Nano 2014, 8,
6202 – 6210.
[21] Y. Li, L. Zhang, A. Torres-Pardo, J. M. Gonzµlez-Calbet, Y. Ma,
P. Oleynikov, O. Terasaki, S. Asahina, M. Shima, D. Cha, L.
Zhao, K. Takanabe, J. Kubota, K. Domen, Nat. Commun. 2013,
4, 2556.
[22] F. Le Formal, N. TØtreault, M. Cornuz, T. Moehl, M. Grätzel, K.
Sivula, Chem. Sci. 2011, 2, 737 – 743.
[29] F. Le Formal, E. Pastor, S. D. Tilley, C. A. Mesa, S. R. Pendle-
bury, M. Grätzel, J. R. Durrant, J. Am. Chem. Soc. 2015, 137,
6629 – 6637.
[30] H. Perschke, E. Broda, Nature 1961, 190, 257 – 258.
[31] J. T. Corbett, J. Biochem. Biophys. Methods 1989, 18, 297 – 307.
[23] O. Zandi, T. W. Hamann, J. Phys. Chem. Lett. 2014, 5, 1522 –
1526.
[24] B. Klahr, T. Hamann, J. Phys. Chem. C 2014, 118, 10393 – 10399.
[25] T. Lopes, L. Andrade, F. Le Formal, M. Grätzel, K. Sivula, A.
Mendes, Phys. Chem. Chem. Phys. 2014, 16, 16515.
Received: January 27, 2016
Published online: March 24, 2016
Angew. Chem. Int. Ed. 2016, 55, 5851 –5855
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5855