Effect of Redox Cocatalysts Location on Photocatalytic Overall Water Splitting over Cubic NaTaO3 Semiconductor Crystals Exposed with Equivalent Facets

Article abstract of DOI:10.1021/acscatal.5b02503

In the semiconductor photocatalyst system for overall water splitting, cocatalysts play crucial roles because they provide not only redox active sites but also charge separation function for photogenerated electrons and holes. In this work, we have investigated the cubic structured NaTaO₃ with six equivalent {001} facets to address the following two important questions: Can charge separation occur among the equivalent facets? How can photogenerated charges be separated on the equivalent surface for photocatalytic reactions? Charge location probe experiments by photodepsotion of noble metals and metal oxides show that no spatial charge separation occurs among the six equivalent facets of NaTaO₃. However, observation of efficient overall water-splitting reaction upon loading of well-known cocatalyst NiO on the NaTaO₃ clearly demonstrates that photogenerated electrons and holes could still be well-separated. In-situ formation of Ni and NiO cocatalysts during the water-splitting process was revealed by X-ray photoelectron spectroscopy and synchrotron X-ray absorption spectroscopy, confirming the role of dual cocatalysts Ni/NiO, where nickel serves as an electron trap (catalytic sites for proton reduction) and NiO serves as a hole trap (catalytic sites for water oxidation). Such vicinal charge separation by dual cocatalysts leads to efficient overall water splitting. (Chemical Equation Presented).

Full text of DOI:10.1021/acscatal.5b02503

Subscriber access provided by Chicago State University
Article
The Effect of Redox Cocatalysts Location on
Photocatalytic Overall Water Splitting over Cubic NaTaO3
Semiconductor Crystals Exposed with Equivalent Facets
Qiao Zhang, Zheng Li, Shengyang Wang, Rengui Li, Xianwen
Zhang, Zhen-xing Liang, Hongxian Han, Shijun Liao, and Can Li
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b02503 • Publication Date (Web): 03 Feb 2016
Downloaded from http://pubs.acs.org on February 4, 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.
ACS Catalysis 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 32
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
The Effect of Redox Cocatalysts Location on
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
2
3
4
Photocatalytic Overall Water Splitting over Cubic
NaTaO Semiconductor Crystals Exposed with
3
Equivalent Facets
1,2
2
2
2
2
1
5
6
Qiao Zhang , Zheng Li ,Shengyang Wang , Rengui Li , Xianwen Zhang , Zhenxing Liang ,
2
,3*
1*
2,3*
Hongxian Han , Shijun Liao , Can Li
7
8
9
1. Key Laboratory on Fuel Cell Technology of Guangdong Province, School of Chemistry and
Chemical Engineering, South China University of Technology, Guangzhou 510640, China
1
0
1
2. State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of
Sciences; Division of Solar Energy, Dalian National Laboratory for Clean Energy, Dalian
116023, China
1
12
1
3
3. The Collaborative Innovation Center of Chemistry for Energy Materials (iChEM).
1
1
1
4
5
6
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
1
2
7
8
9
0
ABSTRACT
In the semiconductor photocatalyst system for overall water splitting, cocatalysts play crucial
roles because they provide not only redox active sites but also charge separation function for
photogenerated electrons and holes. In this work, we have investigated the cubic structured
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
NaTaO with six equivalent {001} facet exposure to address the following two important
3
2
2
3
questions: if charge separation can occur among the equivalent facets, and how photogenerated
charges can be separated on the equivalent surface for photocatalytic reactions. Photodeposition
of noble metals and metal oxides show no preferential deposition, indicating that no spatial
2
24
2
5
6
charge separation occurs among the six equivalent facets of NaTaO . However, observation of
3
2
efficient overall water splitting reaction upon loading of wellꢀknown cocatalyst NiO on the
27
NaTaO clearly demonstrates that photogenerated electrons and holes could still be wellꢀ
3
28
29
30
31
separated. Inꢀsitu formation of Ni and NiO cocatalysts during the water splitting process was
revealed by XPS and XAFS analysis, confirming the role of dual cocatalysts Ni/NiO, where
nickel serves as an electron trap (catalytic sites for water reduction) and NiO serves as a hole trap
(catalytic sites for water oxidation). Such kind of vicinal charge separation by dual cocatalysts
can also lead to efficient overall water splitting.
32
3
3
4
3
KEYWORDS: Charge separation, Equivalent facets, Dual Cocatalysts, Overall water splitting,
Photocatalysis, Hydrogen production
35
36
3
7
38
ACS Paragon Plus Environment
2

Page 3 of 32
ACS Catalysis
1
2
3
3
9
INTRODUCTION
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
0
1
Photocatalytic splitting of water has been extensively studied in recent years since it is a viable
1
4
means to harvest and store abundant solar energy in the clean chemical energy form of H . In
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
42
order to achieve efficient overall water splitting with sunlight, a photocatalyst or photocatalyst
system must simultaneously meet the requirements of efficient light absorption, charge
separation and catalytic reactions, which are the three key factors of photocatalytic water
4
4
4
4
3
4
5
6
2
splitting reactions. The crucial steps of charge separation and surface reaction can be principally
tuned by semiconductor’s interface and surface engineering. For example, fabrication of
3
4
5
47
heterojunction, pꢀn junction, and phase junction have been well adopted for achieving efficient
4
8
9
charge separation. Loading cocatalysts on the surface of a semiconductor is also a generally
1
a, 2b
4
applied strategy for efficient charge separation and creating surface redox reaction sites.
50
Recently, we reported that photogenerated electrons and holes can be spatially separated on the
6
5
1
2
different facets of BiVO semiconductor crystals, which opens up a new avenue for the
4
5
assembly of crystalꢀbased photocatalytic system by selectively loading redox cocatalysts on the
7
53
different facets of a semiconductor crystal. Due to synergetic effect of facet charge separation
54
55
56
57
and provision of redox active sites by cocatalysts spatially on the different facets, dramatic
7
improvement of photocatalytic activity can be achieved. Though facet charge separation are
8
6ꢀ7,
now well accepted and studied on the different semiconductor crystals, such as TiO , BiVO ,
2
4
9
10
11
12
13
14
7
15
Cu WS , Cu O, BiOCl, Sr Nb O , BaTi O and BaLa Ti O
15
etc, the mechanism of
2
4
2
2
2
7
4
4
4
58
facet charge separation and its relation with the surface reaction sites as well as cocatalysts are
not well understood.
5
9
0
6
NaTaO is a typical nꢀtype semiconductor for overall water splitting. It was reported that
3
61
NiO/NaTaO :La photocatalyst could achieve 56% apparent quantum efficiency (AQE) for
3
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
16
6
6
6
6
2
3
4
5
overall water splitting at 270 nm UV light irradiation. The high photocatalytic activity was
ascribed to the separation of H and O evolution reaction sites by the unique nanostep structure
2
2
of NiO ultrafine nanoparticles. Essentially, cocatalyst introduces two positive effects on the
photocatalysis: (i) promotion of charge separation and (ii) construction of active sites for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
b
66
reduction and oxidation reactions. However, the actual functional species of nickel oxide
cocatalsyts is still under debate. It has been proposed that nickel oxide is a proton reduction
cocatalyst for various nꢀtype semiconductor photocatalysts, whereas water oxidation occurs at nꢀ
6
7
8
6
1
7
69
type semiconductor surface. And recently, this mechanism was revised to suggest that nickel
oxide should be reformulated as Ni and NiO component, and the metallic Ni actually functions
7
0
1
18
7
as proton reduction site and NiO functions as water oxidation site. These controversial
mechanisms are rather bewildering and needs to be further addressed. It would be rather
interesting to investigate further on how photogenerated electrons and holes are separated on the
semiconductor crystals with equivalent facets, which is also a case of many semiconductors.
72
73
74
75
76
In this work, NaTaO with six equivalent (100) crystal facets was synthesized and used as a
3
model semiconductor photocatalyst to investigate the charge separation effect on equivalent
facets and the methodologies for achieving efficient charge separation on semiconductors with
equivalent facets. Photoꢀreductive deposition of noble metals and photoꢀoxidative deposition of
metal oxides were used respectively for the probe of the preferential locations of photogenerated
77
7
8
9
7
80
electrons (reduction sites) and holes (oxidation sites) on the surface of NaTaO . In the meantime,
3
81
82
83
84
impregnation method for randomly distribution of noble metals and metal oxides on the surface
were also applied for comparison as control experiments. Following these probe experiments for
photogenerated electrons and holes, more detailed photocatalytic water splitting reactions were
carried out by loading NiO or Ni/NiO cocatalysts. The existence of Ni and NiO dual cocatalysts
ACS Paragon Plus Environment
4

Page 5 of 32
ACS Catalysis
1
2
3
8
8
8
5
6
7
was confirmed by combination of XPS and XAFS analyses, and their important functional roles
in achieving overall water splitting reactions on semiconductors with equivalent facets were also
investigated and discussed.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
88
89
EXPERIMENTAL SECTION
Preparation of NaTaO . Cubic NaTaO was synthesized by a microwaveꢀassisted hydrothermal
3
3
1
9
9
0
1
method as reported in the previous literature. Ta O (purchased from Amresco) and NaOH
2 5
9
(AR) were employed as the starting materials. Firstly, 0.54 g (1.2 mmol) of Ta O was added to
2
5
92
75 mL of 1.75 M NaOH solution to form a mixed suspension under magnetic stirring at room
temperature for 30 min. The resultant reaction mixture was transferred to a 100 ml Teflonꢀlined
vessel (model XPꢀ1500, CEM Corp.), which was then placed in a MARSꢀ5 microwave digestion
93
94
95
96
97
98
99
o
system operated at 2.45GHz (EMCorp.). The reaction mixture was heated to 220 C in 60 min
o
and maintained at 220 C for 3 hours. Then the products were simply collected by centrifugal
separation and washed with deionized water for 4ꢀ5 times. Finally, the resultant samples was
o
dried in an oven at 60 C for overnight.
Deposition of metal and metal oxide nanoparticles. Photoꢀreductive deposition of noble
100
metals (Pt and Ag) was carried out as follows. To a suspension of NaTaO (0.3 g) aqueous
3
1
01
02
solution containing 20 vol% methanol solution, a calculated amount (5 wt%) of Pt precursor
H PtCl or Ag precursor AgNO were added. The resulting suspension was stirred and irradiated
1
2
6
3
103
under a 450 W Hg lamp for 1 hour. For photoꢀoxidative deposition of MnO or PbO , a
x x
1
1
1
1
04
05
06
07
calculated amount (5 wt%) of the Mn precursor Mn(NO ) or the Pb precursor Pb(NO ) were
3 2 3 2
added into the suspension of NaTaO in the 500 mL NaIO solutions (0.01M).The resulting
3
3
suspension was stirred and irradiated under a 450 W Hg lamp for 1 hour. For metal and metalꢀ
oxide simultaneous photodeposition, H PtCl (AgNO ) and Mn(NO ) (Pb(NO ) ) were chosen
2
6
3
3 2
3 2
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
1
1
08
09
10
11
as precursors and reactions were carried out without any sacrificial reagent. After
photodeposition, the final reaction mixture was well filtered, washed with deionized water for
o
more than 3 times, and finally dried at 60 C for overnight.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Deposition of NiO cocatalyst. NiO cocatalyst was loaded on the surface of cubic NaTaO by
3
112
both photodeposition and impregnation methods.
1
13
14
In the photodeposition method, NaTaO powder (0.3 g) suspension in 500 mL 20%
3
1
methanol/water solution containing an appropriate amount of Ni(NO ) .6H O was prepared in an
3
2
2
115
inner irradiation cell made of quartz. The mixture was stirred while irradiation with a light source
of a 450 W highꢀpressure mercury lamp (UshioꢀUM452). After 3 hours reaction time, the
suspension was filtered and washed with deionized water for more than 5 times. The sample was
1
16
17
1
o
118
dried at 60 C for overnight and further calcined at 543K for 2 hours. Thus prepared sample is
119
120
121
122
denoted as NaTaO ꢀNiO (hv).
3
For comparison, NiO was also deposited on the surface of NaTaO by impregnation method.
3
Typically, a suspension of NaTaO powder in an aqueous solution (2~3 mL) containing an
3
appropriate amount of Ni(NO ) .6H O was prepared on a porcelain crucible. The suspension was
3
2
2
o
123
then stirred at 100 C using a glass rod during water evaporation. The dried powder was calcined
1
24
25
at 543K for 2 hours. Such prepared sample is denoted as NaTaO ꢀNiO (imp).
3
1
Photocatalytic Reactions. The photocatalytic activities of the prepared samples were carried out
in a closed gas circulation and evacuation system under irradiation with a 450W highꢀpressure
Hg lamp (UshioꢀUM452). Typically, 0.3 g of sample (either NaTaO ꢀNiO (imp) or NaTaO ꢀ
126
127
128
129
3
3
NiO(hv)) was suspended in 500 mL deionized water. Before irradiation, the reaction system was
thoroughly degassed by evacuation in order to drive off the air inside. The amounts of evolved
ACS Paragon Plus Environment
6

Page 7 of 32
ACS Catalysis
1
2
3
1
1
1
1
30
31
32
33
H and O gases were determined by an onꢀline gas chromatograph (Agilent, GCꢀ7890, TCD, Ar
2
2
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
carrier).
Characterization. All samples were characterized by powder Xꢀray diffraction (XRD, Rigaku
D/maxꢀ2500PC) equipped with a graphiteꢀfiltered Cu/Kα (λ=0.154 nm) radiation source, and the
applied current and voltage were 100 mA and 40 kV, respectively. The morphology of the
samples were examined using a scanning electron microscope (SEM, FEI Quanta200F) operated
at 20 kV, a transmission electron microscope (TEM, FEI Tecnai Spirit) and a high resolution
transmission electron microscope (HRTEM, FEI TecnaiF30) operated at 120 kV and 300 kV,
respectively. The Xꢀray photoelectron spectroscopy (XPS) measurements were carried out on a
VGESCALAB MK2 spectrometer with monochromatized Al/Kα excitation, and the C1s peak
(284.6 eV) was used as the reference. ICPꢀAES analyses were carried out on an inductively
coupled plasma atomic spectrometry (ICPSꢀ8100 Shimadzu). The XAFS spectra of Ni Kꢀedge
were measured at beamline 14W station of Shanghai Synchrotron Radiation Facility (SSRF)
equipped with 150 MeV linear accelerator, 3.5 GeV booster and 3.5 GeV storage ring. The
output energy was tuned for Ni measurement using a Si(111) crystal monochromator. The
samples were sealed with Kapton films in glove box and stored in argon atmosphere before
measurements. The data were collected in fluorescence mode at room temperature using 32ꢀ
element Canberra/XIA Ge detector system. Ni foil was also measured with the samples
simultaneously using the third ionization chamber so that the energy calibration could be
performed by scan. Athena software package was employed to process and analysis of the XAFS
data.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
134
1
35
36
1
137
1
38
39
1
140
1
1
1
1
41
42
43
44
145
1
46
47
1
148
1
49
50
1
1
51
RESULTS AND DISCUSSION
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
1
1
52
53
54
55
Crystal structure of NaTaO . As shown in SEM images of Fig. 1a, the synthesized NaTaO3
3
samples are cubic morphology crystals. The particle size ranges from 130ꢀ430 nm with the mean
value of ca. 270 nm. The normally seen crystalline structures of NaTaO are: (a) the cubic phase;
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(b) the monoclinic phase; (c) the orthorhombic phase. The 2θ positions of the principal
diffraction peaks for these three phases are nearly identical, which leads to difficulty in
identifying the different phase structures of the synthesized NaTaO To identify phase structure
156
157
158
159
160
3
.
of the synthesized NaTaO , we have conducted the XRD measurement at slow scanning rate of
3
o
0.02 /min and the spectra are shown in Fig 1b. Some principal peaks, such as those designated as
A, B and C, were identified to be comprised of several subꢀpeaks. By comparing with the
161
standard JCPDS, we concluded that the NaTaO materials prepared in this work are the
3
2
0
1
62
63
orthorhombic phase (Pcmn a=5.5319 Å, b=7.8054 Å and c = 5.4943 Å). The full TEM images
1
(Figure 1c) in 100 nm resolution shows that the synthesized NaTaO are cubic morphology
3
164
crystals, which is consistent with the result of the SEM images. The SAED patterns (Figure 1d)
of the area shown with a circle in Figure 1c further indicates that the synthesized NaTaO3
belongs to an orthorhombic phase with space group Pcmn. The calculated lattice parameters of a0,
b and c are 5.5319 Å, 7.8054 Å and 5.4943 Å, respectively. However, the main exposed
1
1
65
66
167
0
0
168
169
170
171
crystallographic planes are not the typical orthorhombic phase of (100) and (001), because it
shows six crystallographic planes with four equivalent crystallographic planes of (101), (10ꢀ1), (ꢀ
101), and (ꢀ10ꢀ1) indexed by {101} set together and two equivalent crystallographic planes of
(010) and (0ꢀ10) indexed by {010} set together. It should be noted that the six crystallographic
172
planes are actually corresponding to the cubic {100} facet of NaTaO . Furthermore, HRTEM
3
1
73
74
image (Figure 1e) shows that the crystal features with lattice distances of 3.92 Å and 7.84Å
calculated from FFT analysis, which are corresponding to the crystal planes of (101) and (010),
1
ACS Paragon Plus Environment
8

Page 9 of 32
ACS Catalysis
1
2
3
1
1
1
75
76
77
respectively. Therefore, it is clear that the synthesized NaTaO is a standard cubic crystal with
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
six equivalent {100} facets exposed on the surface.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
178
179
Figure 1 (a) SEM image, (b) XRD patterns, (c) TEM image, (d) SAED pattern, and (e) HRTEM of
1
1
80
81
synthesized cubic crystals of NaTaO3.
1
82
Probing electron and hole sites on the facets. Excitation of a semiconductor with an
appropriate light source can generate photoexcited electrons and holes, which may migrate to
some particular sites of the semiconductor surface to participate in the surface redox reactions.
Thus, photoꢀreductive deposition of metals or photoꢀoxidative deposition of metal oxides are
convenient methods to probe the surface electron sites or hole sites, respectively.
183
184
185
186
187
The six facets of the cubic NaTaO are smooth as revealed by the SEM image shown in Figure
3
2
+
ꢀ
188
2a. Photoꢀoxidation deposition of Mn ions on NaTaO was performed with IO as the electron
3 3
1
89
90
acceptor. It can be seen from the SEM image of the Figure 2b that the MnO nanoparticles are
x
1
deposited randomly on the six facets of NaTaO without any facet selectivity. This demonstrates
3
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
1
1
91
92
93
94
that there is no preferential facet to capture the photogenerated holes upon photoꢀirradiation. The
6
photodeposition reactions can be described as follows:
2
+
+
+
Mn + xH O + (2xꢀ2) h
MnO + 2x H
(1)
2
x
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
ꢀ
ꢀ
ꢀ
(2xꢀ2)H + (xꢀ1)/3IO +(2xꢀ2) e
(xꢀ1)/3 I + (xꢀ1)H O (2)
3
2
195
Photodeposition of Pt was carried out using H PtCl as Pt precursor in the presence of methanol
2 6
1
96
97
as hole sacrifice agent. It can be seen from the SEM image of Figure 2c that Pt particles are also
1
randomly deposited on the six faces of NaTaO without any facet selectivity. This demonstrates
3
198
that there is no preferential facet to host the photogenerated electrons upon photoꢀirradiation. The
1b
1
99
00
photoꢀreduction of metal ions into metal nanoparticles can be described as follows:
2
ꢀ
ꢀ
ꢀ
2
PtCl6 + 4e
Pt + 6Cl
(1)
(2)
+
+
201
2/3CH OH + 2/3H O+ 4h
2/3 CO + 4H
2
3
2
202
203
204
205
Simultaneous photodeposition of Pt and MnO was also carried out without using sacrificial
x
agents. The SEM image (Fig. 2d) shows that Pt and MnO are also randomly distributed on each
x
facet of NaTaO . To demonstrate the universality of the observed nonꢀfacet selective deposition
3
phenomena, photodepositions of other metals (e.g., Ag) and metal oxides (e.g., PbO ) were also
2
206
tested. And again the SEM images (Fig. S1) revealed that these metals and metal oxides were
2
07
08
also randomly deposited on the surface of NaTaO , suggesting nonꢀfacet selectivity is not due to
3
2
some particular properties of deposited metals or metal oxides but is due to the intrinsic nature of
209
the NaTaO crystal with six equivalent facets.
3
ACS Paragon Plus Environment
10

Page 11 of 32
ACS Catalysis
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
210
211
Figure 2 SEM images of (a) NaTaO ; (b) MnO ꢀNaTaO ; (c) PtꢀNaTaO ; (d) PtꢀMnO ꢀNaTaO .
3
x
3
3
x
3
2
12
13
2
Overall, the probing of electron and hole sites on NaTaO revealed that the photogenerated
3
214
electrons and holes cannot be spatially separated on the equivalent {100} facets of NaTaO3,
which is quite different compared with the results of spatial separation of photogenerated
2
15
16
2
electrons and holes on the nonequivalent {001} and {101} facets of BiVO . The present work
4
217
adequately demonstrates that the prerequisite of the facet charge separation in semiconductor
crystals is the existence of nonequivalent facets.
218
219
220
221
Overall water splitting by loading NiO cocatalyst. Since the key process of charge separation
cannot be achieved by facet charge separation on NaTaO with six equivalent facets, question
3
then remains on how to achieve efficient charge separation on crystals with equivalent facets for
222
overall water splitting. Since NiO is a known cocatalyst matching with NaTaO in energy level
3
1
6
2
23
24
for water splitting, it was used as a cocatalyst for the investigation of charge separation on
2
NaTaO with six equivalent facets during photocatalytic overall water splitting.
3
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
2
2
25
26
27
28
For comparison, NiO was loaded by both photodeposition (Fig. 3a) and impregnation (Fig. 3b)
methods. Firstly, the photocatalytic water splitting activities of NaTaO ꢀNiO (hv) and NaTaO ꢀ
3
3
NiO (imp) samples with highest amount of NiO (2 wt%) were subject for the investigation.
Under the same experimental conditions, these two photocatalysts showed similar photocatalytic
activity for overall water splitting as shown in the photocatalytic reaction time courses in Fig. 3c
and Fig. 3d. The fact of the similar photocatalytic activities is a strong indication that the NiO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
229
2
30
31
2
cocatalyst might be randomly deposited on each facets of NaTaO in a similar manner in spite of
3
232
the different cocatalyst loading methods. This is not surprising since the synthesized NaTaO3
with six equivalent facets has no facet charge separation effect as confirmed in the previous
sections. Nevertheless, the successful splitting of water demonstrates that the photogenerated
electrons and holes indeed could be efficiently separated once NiO were loaded on the surface of
2
33
34
2
235
236
237
238
239
NaTaO either by photodeposition or impregnation method, implying that the cocatalyst plays an
3
important role in achieving efficient charge separation for overall water splitting.
To know the nature of the NiO species in these two samples, ICPꢀAES, XPS, TEM and XRD
patterns analyses were performed. Per ICPꢀAES, the amount of Ni in NaTaO ꢀNiO (hv) and
3
240
NaTaO ꢀNiO (imp) is 1.05 wt % and 1.12 wt %, respectively, which are nearly the same within
3
2
41
42
the experimental error. XPS spectra (Figure S2) show two main peaks at 855.4 and 872.7 eV
and shoulder peaks at 861.2 and 879.4 eV, which correspond to Ni 2p3/2 and Ni 2p1/2,
respectively. All peaks are the distinctive feature of oxidized nickel, i.e. NiO. However, a small
2
243
244
245
246
247
shoulder peak at ca. 853.0 eV in NaTaO ꢀNiO (hv) is relevant to the metallic Ni species. Despite
3
this, observation of almost identical binding energies of Ni 2p3/2 and 2p1/2 for both of the samples
demonstrates that the similar NiO species were formed on the surface of NaTaO though
3
21
different deposition methods were applied. The morphology and the particle size of NiO
ACS Paragon Plus Environment
12

Page 13 of 32
ACS Catalysis
1
2
3
2
2
2
48
49
50
particles deposited by different methods are also quite similar, as visualized by the almost
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
complete coverage of NiO on the surface of NaTaO in TEM images (Fig. 4). The XRD patterns
3
of NaTaO with and without doped NiO are shown in Fig. S3. We found that the 2θ values of the
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
251
252
253
most intense peak at 32.5° did not change before and after NiO loading, which means that
loading of NiO coꢀcatalysts on the surface of NaTaO did not obviously change the crystal
3
structure of NaTaO .These results demonstrate that there is no significant difference of the NiO
3
254
species in terms of loading amount, valence state and morphology for the NaTaO samples with
3
2
55
56
2 wt% NiO loading either by photodeposition or impregnation methods. It can be tentatively
concluded that achievement of overall water splitting is not due to charge separation in a way of
spatial facet charge separation in this system. Efficient charge separation for overall water
splitting might be related to the presence of the NiO cocatalyst species.
2
257
2
58
59
2
260
261
262
263
Figure 3 The scheme of NaTaO with NiO cocatalyst deposited by (a) photodeposition and (b) impregnation
3
methods, and (c, d) the time courses of the photocatalytic water splitting activity over NaTaO ꢀ2% NiO(hv)
3
and NaTaO ꢀ2% NiO(imp) , which are corresponding to (a) and (b), respectively.
3
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
264
265
Figure 4 The TEM images of (a,b) NaTaO ꢀNiO(hv) and (c,d) NaTaO ꢀNiO(imp) with 2 wt% NiO loading.
3
3
2
66
67
2
It should be noted that the H and O rates in Fig. 3c and Fig. 3d are not in stoichiometric ratios.
2
2
268
Such kind of deviation may be related to many factors, including the reactions with sacrificial
reagents, adsorption of the evolved gases on the catalyst, consumption of the evolved gases by
impurities, etc. But in most cases, it reflects that the photogenerated electrons and holes are not
equally consumed by water splitting reaction, which may be related to the nature of the surface
redox active sites for H and O evolution. To better understand the nonꢀstoichiometric ratio of
2
69
70
2
271
2
72
73
2
2
2
H and O evolution, the effect of the NiO cocatalyst was further examined by varying its
2
2
274
loading amount from 0 wt% to 2 wt%. The reaction rates of evolved H and O are shown in Fig
2 2
275
276
277
278
5. It can be seen that the photocatalytic water splitting activities are quite different for NaTaO3ꢀ
NiO(imp) and NaTaO ꢀNiO(hv). The NaTaO ꢀNiO (imp) samples show a kind of volcanic trend.
3
3
At high NiO loading, the H /O ratio is much higher than 2. And the highest activity with H /O
2
2
2
2
ratio approximately to 2:1 can be reached when NiO loading was 0.2 wt%. Further decrease of
the NiO loading decreases the O evolution rate, and no O evolution was observed without NiO
279
2
2
ACS Paragon Plus Environment
14

Page 15 of 32
ACS Catalysis
1
2
3
2
2
2
2
80
81
82
83
loading. On the other hand, the NaTaO ꢀNiO (hv) also showed that the highest photocatalytic
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
water splitting activity with H /O ratio more than 4 was reached when the NiO loading was ca.
2
2
1.6 wt%. Further decrease of the NiO loading decrease the O evolution rate and no oxygen
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
evolution was observed. It can be seen that the NiO loading amount largely affects the
284
photocatalytic activities in both cases. Especially the O evolution rate was largely inhibited
2
2
85
86
when low amount of NiO was loaded by photodeposition method. This implies that though the
2
same amounts of NiO cocatalysts were loaded on the surface of NaTaO , the photodeposition
3
287
method may give much less O evolution sites to consume the photogenerated holes, especially
2
2
88
when NiO was in low amount loading.
289
290
291
Figure 5 Photocatalytic water splitting into H and O over (a) NaTaO ꢀNiO (imp); (b) NaTaO ꢀNiO(hv) with
2
2
3
3
different amounts of NiO loading.
2
92
293
To gain a deep insight into the effect of NiO loading, the typical samples of NaTaO ꢀNiO (imp)
3
2
94
95
(which shows both H and O evolution activity) and NaTaO ꢀNiO (hv) (which shows only H
2
2
2
3
2
evolution activity) with the same amount of NiO loading (0.2 wt%) were also subject to ICPꢀ
AES, XPS, HRSEM investigations. The ICPꢀAES analyses show that the actual amounts of Ni in
NaTaO ꢀNiO (imp) and NaTaO ꢀNiO (hv) are similar, which are 0.15 wt% and 0.13 wt%,
296
97
2
3
3
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
3
3
98
99
00
01
respectively. HRSEM and SEM images (Fig. 6) show that the NiO particles with size ranging
from 7 to 9 nm are all randomly distributed on the surface of each sample. XPS (Fig. S3) spectra
show almost identical binding energies of Ni 2p1/2 and Ni 2p3/2 for both samples. The shoulder
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
peaks at ca. 852.5 eV is due to the metallic Ni, which is more intense in NaTaO ꢀNiO (hv) than
3
302
that in NaTaO ꢀNiO (imp). Ni and NiO are the main component of the Ni species in both
3
3
03
04
samples though the relative ratio of Ni and NiO may be different. Further, to address the
3
questions whether the cocatalysts change the facial orientation of NaTaO , we have carefully
3
305
analyzed the TEM and HRTEM images of NaTaO ꢀNiO (imp) and NaTaO ꢀNiO (hv) with 0.2 wt%
3 3
3
06
07
NiO loading. It can be seen from Figure 7 that the lattice fringe distance of NiO (111) plane was
kept same, no matter what kinds of the loading methods were employed. This provides sound
base for the comparison of NiO effects on different loading methods. Furthermore, HRTEM
images of NaTaO also clearly show that the synthesized NaTaO crystals always keep the lattice
3
308
309
310
311
312
3
3
distances of crystal planes of (101) to be 3.92 Å in the presence and absence of NiO. From these
observation, we concluded that the presence of NiO cocatalyst does not change the facial
orientation of NaTaO . All of these characterization results indicate that there is no great
3
313
difference of the originally loaded NiO species in NaTaO ꢀNiO (imp) and NaTaO ꢀNiO (hv).
3
3
3
14
15
This is a little hint that certain amount of H or O evolution sites may not present at the initial
2 2
3
stage of the photocatalytic reaction but be derived inꢀsitu during the photocatalytic reaction.
ACS Paragon Plus Environment
16

Page 17 of 32
ACS Catalysis
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
3
3
3
16
17
18
Figure 6: The SEM and the expanded SEM images of (a,b) NaTaO ꢀNiO(hv) and (c,d) NaTaO ꢀNiO(imp)
3
3
with 0.2 wt% NiO loading
3
19
3
3
20
21
Figure 7: The TEM and HRTEM images analysis of (a , b) NaTaO ꢀNiO (imp); (c, d) NaTaO ꢀNiO (hv). Both
3 3
samples were loaded with 0.2 wt% loading.
3
22
23
3
The evolution of Ni/NiO dual cocatalysts. In order to know if new types of active Ni species
324
were generated during the photocatalytic reactions, we further examined the Ni valance state
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
3
3
25
26
27
28
before and after photocatalytic water splitting reaction by XPS (Fig.8), EXAFS (Fig. 9) diffuse
reflectance spectra (Fig. S5).
We have measured the XPS of NiO only (Figure 8a), and compared it with XPS data of NaTaO3ꢀ
NiO(imp) with 0.2wt% loading before (Figure 8b) and after (Figure 8c) water splitting reactions.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
329
The XPS spectrum of NiO obtained by calcination of Ni(NO ) at 270 C for 4 h (Figure 8a)
3
3
3
30
31
shows similar XPS spectrum as the commercial NiO (Fig. S6). First, the peaks centered at 853.7
eV and 855.5 eV as well as those at 872.9 eV and 870.9 eV can be ascribed to two multipletꢀsplit
peaks of Ni2p3/2, and the broad peaks centered at 860.9 eV and at 879.1 eV can be ascribed to
the other two satellite peaks of NiO. However, it can be seen from Figure 8b that metallic Ni is
formed after photocatalytic water splitting reactions, as evidenced by the clear observation of Ni
2p1/2 and Ni 2p3/2 at 855.5 eV and 872.9 eV in addition to the shoulder peaks at 861.0 eV and
879.1 eV, respectively. From Figure 8b,it can be seen that NiO is the main component for
3
332
3
33
34
3
335
336
337
338
339
NaTaO ꢀNiO(imp) with 0.2wt% loading before water splitting reaction, as evidenced by the clear
3
observation of Ni 2p1/2 and Ni 2p3/2 at 855.5 eV and 872.9 eV with the shoulder peaks at 861.0
eV and 879.1 eV, respectively. However, after photocatalytic water splitting, the peak at 855.5
eV was dramatically broadened, and appearance of additional peak at low binding energy of
852.4 eV is a strong indication that reduced Ni species, mostly probably metallic Ni species,
were formed (Fig. 8c). So it is reasonable to deduce that the actual cocatalyst during
photocatalytic reaction is Ni and NiO dual cocatalysts, where Ni was inꢀsitu formed by
photoreduction of NiO during photocatalytic reactions.
340
3
41
42
3
343
44
3
ACS Paragon Plus Environment
18

Page 19 of 32
ACS Catalysis
1
2
3
4
5
6
7
8
9
Ni 2p
a)
after
(
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
(
b)
before
(c)
NiO
850
855
860
865
870
875
880
885
Binding Energy (eV)
3
3
3
3
45
46
47
48
Figure 8: The Ni 2p XPS spectra of NaTaO ꢀNiO(imp) with 0.2 wt% NiO loading (a) before, (b) after water
3
splitting reaction and (c) NiO .
349
In order to further elucidate the oxidation state changes of the Ni species, EXAFS spectra were
3
50
51
recorded for Ni foil, NiO, and the NaTaO ꢀNiO(imp) before and after photocatalytic water
3
3
splitting reactions (Figure 9a). In the XANES spectra, the position of the Kꢀedge corresponds to
the oxidation state of the measured elements, the higher energy position it is, the higher
oxidation state of the elements, and vise versa. Though the high energy shift of the Kꢀedge of the
NiO species compared to the metallic Ni can be directly visualized, the Kꢀedge spectra of the
352
3
3
53
54
355
NaTO ꢀNiO(imp) before and after photocatalytic water splitting reaction look rather similar,
3
3
3
3
3
56
57
58
59
indicating that NiO on NaTaO were well reserved after photocatalytic water splitting reaction.
3
The Kꢀedge positions of these samples were further analyzed using the derivative spectra as
shown in the inset of Figure 9b, since the Kꢀedge positions can be more clearly judged by the
peak positions in the derivative spectra. Compared to the pure NiO, the Ni Kꢀedge position of
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
3
3
60
61
62
63
NaTaO ꢀNiO(imp) before photocatalytic water splitting reaction was slightly shifted toward low
3
energy compared to pure NiO, indicating that some reduced Ni species might be present. After
photocatalytic water splitting reaction, the Ni Kꢀedge was further shifted toward the low energy
position, indicating that more reduced Ni species might be formed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
364
The Ni Kꢀedge K weighted EXAFS spectra as well as the Fourierꢀtransformed relative radial
3
65
66
distance spectra are shown in Fig 9c and Fig. 9d, respectively. It can be seen that the Ni species
3
in NaTaO ꢀNiO (imp) before and after photocatalytic water splitting reaction were mainly in the
3
367
form of NiO states, because all of the peak positions are all well correlated with those of pure
NiO in both spectra. However, careful inspection on the Fourierꢀtransformed relative radial
distance spectra shows some minor but clearly detectable changes of the Ni species after
photocatalytic water splitting reaction. Firstly, the second shell of NiꢀNi was slightly shifted (ca.
0.02 Å) toward the first shell of the metallic Ni, while the peaks of the third and the fourth shells
were also simultaneously shifted (0.06Å and 0.13Å, respectively) toward the peak positions of
the third and forth shells of the metallic Ni. This is a strong indication that there must be some
3
68
69
3
370
371
372
373
374
16
reduced Ni species formed. Combining with the XPS results, we can infer that the inꢀsitu
375
formed metallic Ni by photoreduction of NiO is not stable, which can be readily oxidized back to
NiO. The reduced Ni species deduced from the EXAFS analysis might be some stable metallic
Ni species randomly distributed on NiO or surface of NaTaO3.
3
76
77
3
ACS Paragon Plus Environment
20

Page 21 of 32
ACS Catalysis
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
378
3
3
79
80
Figure 9: (a) Ni Kꢀedge spectra of Ni foil, NiO, NaTaO ꢀNiO(imp) before photocatalytic water splitting
3
reaction; NaTaO ꢀNiO(imp) after photocatalytic water splitting reaction. (b) corresponding derivative spectra
3
381
382
383
for the clarify of the Ni Kꢀedge shift. (c) Ni Kꢀedge K3 weighted EXAFS spectra and (d) Fourier Transformed
spectra of Ni foil, NiO, NaTaO ꢀNiO(imp) before photocatalytic water splitting reaction; NaTaO ꢀNiO(imp)
3
3
after photocatalytic water splitting reaction.
3
84
85
3
This infers that the actual cocatalyst during photocatalytic reaction is dual cocatalyst, Ni and
NiO. At the initial stages of the photocatalytic water splitting reactions, the photogenerated
electrons might reduce NiO inꢀsitu to Ni at reduction sites instead of the reduction of proton to
H . Before reaching the balance, nonꢀstoichiometric amounts of H and O was generated, as
386
3
87
88
3
2
2
2
389
observed in Fig. S8 that the ratio of H and O was ca. 3.2 at a reaction time of 15 min. With the
2 2
3
90
91
reaction proceeds, more NiO was reduced to metallic Ni, and finally reach the reaction balance
to give H and O evolution ratio in 2:1. Diffuse reflectance spectrum of the NaTaO ꢀNiO (imp)
3
2
2
3
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
3
3
3
92
93
94
95
with 0.2wt% NiO loading (Fig. S5) was similar to that of the naked NaTaO before
3
photocatalytic water splitting reaction because of the small amount of NiO loaded. After water
splitting, the spectrum of NaTaO ꢀNiO (imp) became similar to that of the NaTaO ꢀNi
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
photocatalyst with 5wt% Ni loading by photoreduction. The additional absorption at ca. 400~800
nm was observed after reaction which is consistent with the color change from white to grey,
396
3
3
97
98
indicating that the metallic nickel fine particles were formed in situ when NaTaO ꢀNiO (imp)
3
photocatalyst was photoirradiated. These inꢀsitu photoreduced Ni species may play important
role in photocatalytic water splitting reaction.
399
4
00
01
4
The Illustration of dual cocatalysts in overall water splitting reaction on the Ni/NiO
402
403
404
NaTaO photocatalyst.
3
Based on the above experimental results, Scheme 1 describes the promotion of the photocatalytic
water splitting reaction by the formation of Ni and NiO dual cocatalysts on the cubic NaTaO3
with equivalent facets.
405
4
06
07
In the assembly of the semiconductor photocatalyst system for overall water splitting, the energy
level matching between light absorbing semiconductor photocatalyst and the cocatalysts is
vitally important. As shown in Scheme 1a, the energy diagram shows that photogenerated
4
408
4
09
10
electrons on NaTaO could reduce NiO to metallic Ni, since the conduction band level of the
3
4
NaTaO photocatalyst is roughly estimated to be ꢀ1.06 eV, which is more negative than that of
3
2
2
411
NiO (ꢀ0.96 eV). This is quite different compared with the behavior of the NiO on the surface of
4
4
4
4
12
13
14
15
SrTiO , which showed no reduction of NiO to metallic Ni, because the lower conduction band
3
17d, 18b
level of SrTiO is more positive than that of the NiO.
In case of NaTaO , inꢀsitu
3
3
photoreduced metallic Ni species may also involve in the photocatalytic water splitting reaction,
and the balance between Ni and NiO as well as their location sites are crucial factors influencing
ACS Paragon Plus Environment
22

Page 23 of 32
ACS Catalysis
1
2
3
4
4
4
4
16
17
18
19
the overall water splitting activity (Scheme 1b). Scheme 1c and 1d give the illustration of two
possible processes for the evolvement of the Ni and NiO dual cocatalysts in the water splitting
reaction and the overall water splitting on the dual cocatalysts. The early stage of the progresses
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
in the Scheme1c and 1d are the same, photogenerated electrons on NaTaO could reduce NiO to
3
420
metallic Ni. The difference is that situation between NiO and inꢀsitu formation of Ni dual
cocatalyst in the following stage for overall water splitting, and Scheme 1c and 1d give two
possible illustrations.
4
21
22
4
423
In the synthesis of NaTaO ꢀNiO (hv) with low Ni content (e.g., 0.2 wt% NiO), the Ni precursor
3
4
24
25
Ni(NO ) was photoreduced to metallic Ni by photogenerated electrons, and then oxidized to
3
3
o
4
NiO at 270 C. In such processes, the NiO should mainly reside at the reduction sites of NaTaO .
3
426
Because it is at the reduction sites that such deposited NiO could be reduced to metallic Ni
(highly efficient proton reduction cocatalyst) again by photoirradiation during the photocatalytic
water splitting reactions, hence greatly enhance the hydrogen evolution activity to give nonꢀ
stoichiometric ratio of H and O evolution during the photocatalytic water splitting reaction. As
427
428
429
430
2
2
for the NaTaO ꢀNiO (imp) with same low Ni content (e.g., 0.2 wt% NiO), though the synthesis
3
431
procedure was same as that of the NaTaO ꢀNiO(hv), but NiO cocatalyst may be located
3
4
32
33
randomly on both reduction and oxidation sites of NaTaO . The deposition of NiO on NaTaO
3
3
4
by impregnation method is simply by physical adsorption thus no redox site selectively. Thus,
434
photoirradiation of NaTaO ꢀNiO (imp) during photocatalytic water splitting reaction may also
3
435
436
437
438
reduce partial NiO on the reduction sites to metallic Ni while still maintaining large amount of
NiO on the oxidation sites. Metallic Ni species produced at the reduction sites serves as efficient
proton reduction cocatalyst and the NiO reserved at the oxidation sites serves as water oxidation
cocatalyst. This may be the reason for that the NaTaO ꢀNiO (imp) with 0.2 wt% NiO loading can
3
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
4
4
4
39
40
41
42
evolve stoichiometric amounts of H and O . The similar photocatalytic water splitting activities
2
2
of NaTaO ꢀNiO (hv) and NaTaO ꢀNiO (imp) with high Ni loading (2 wt% NiO) may also be
3
3
easily understood. Because of high NiO loading, prolonged time is needed for the reduction of
NiO to metallic Ni till reaching the balance between Ni and NiO. However, due to relatively high
NiO loading in both cases, well balance between Ni and NiO could be reached which may ensure
both H and O evolution effectively, leading to stoichiometric generation of H and O .
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
443
4
44
45
2
2
2
2
4
Meanwhile, we have performed additional O and H evolution experiments on bare NaTaO ,
2
2
3
446
NiO/NaTaO , and NiO only in the presence of sacrificial agents, and the results are summarized
3
4
47
48
in Table S1. It can be seen that NiO itself showed negligible O or H evolution activities, but
2
2
4
both O and H evolution reactions were promoted over NaTaO ꢀ0.2 wt% NiO (imp) compared
2
2
3
449
to those of the bare NaTaO . It should be noted that ratio of the separate H and O evolution
3 2 2
450
451
452
453
rates in the presence of sacrificial agents were increased approximately 4.7 and 2.7 times,
respectively. Since both H and O evolution reactions can be promoted by NiO loading, and
2
2
considering the synergetic effects of redox reactions (promotion of oxidation reaction by
promoting the corresponding reduction reaction, vise versa), it is difficult to determine whether
NiO is a proton reduction cocatalyst or water oxidation cocatalyst directly by separate oxygen
evolution experiments. However, because water oxidation is the rateꢀdetermining step in overall
water splitting reaction, and the enhancement of water oxidation activity by NiO cocatalyst
loading was indeed observed by the separated water oxidation reaction experiment, we
454
4
55
56
4
457
458
459
460
461
tentatively prefer that NiO on NaTaO more likely serves as a water oxidation cocatalyst in
3
overall water splitting in our work. Thus, it is believed that they must be at the vicinity of the
NiO on the surface of NaTaO since major part of the metallic Ni are originated from the
3
reduction of NiO inꢀsitu by photoirradiation.
ACS Paragon Plus Environment
24

Page 25 of 32
ACS Catalysis
1
2
3
4
62
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4
4
4
4
4
4
4
63
64
65
66
67
68
69
70
Scheme 1: Schematic description of overall water splitting over Ni/NiO dual cocatalyst loaded cubic NaTaO3
with equivalent facets. (a) Energy diagram showing inꢀsitu formation of Ni/NiO dual cocatalysts at the initial
stage of the photocatalytic water splitting reaction. (b) Energy diagram showing the functional role of Ni and
NiO in photocatalytic water splitting reactions. Metallic Ni serves as proton reduction site and NiO serves as
the water oxidation site. (c, d) Illustration of two possible processes for the evolvement of the Ni and NiO dual
cocatalysts in the water splitting reaction and the overall water splitting on the dual cocatalysts.
4
71
72
Dual cocatalysts play important roles in the achievement of overall photocatalytic water splitting
1
a, 2b, 18b, 23
4
activity. The concept of the dual cocatalysts is now well accepted and applied.
But the
473
optimal distance of spatial separation of reduction and oxidation cocatalysts is still not clear. For
example, Domen’s group designed a core/shell structure photocatalysts with spatially separated
reduction and oxidation sites to block the reverse reaction. The high photocatalytic water
splitting activity was attributed to the core/shell structure and spatially separated reduction and
oxidation cocatalysts assisting separation and collection of photogenerated electrons and holes at
4
4
4
4
74
75
76
77
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
23a
4
4
4
4
78
79
80
81
the respective cocatalysts. The dual cocatalysts on the different facets of BiVO were also
4
rationally designed and prepared. The separation of reduction and oxidation sites and the
synergetic effect of dual cocatalysts lead to significantly enhanced photocatalytic performances
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6ꢀ7
in water oxidation. However, recent experimental results also show that the closely intimate
482
location of the reduction and the oxidation cocatalysts may also lead to efficient charge
separation for promoting photocatalytic reactions. For example, the PdS/CdS photocatalyst was
found to be more active than Pt/CdS for H production in Na S–Na SO aqueous solution. The
4
83
84
4
2
2
2
3
485
TEM study clearly showed that Pd reduction cocatalyst and PdS oxidation cocatalyst were
24
4
86
87
connected to each other at the atomic level. In PdꢀIrO /TiO photocatalytic system, it was also
2
2
4
found that the intimate contact between the Pd reduction cocatalyst and the IrO oxidation
2
2
3b
488
cocatalyst can lead to the enhancement of photocatalytic activity of H production. Present
2
489
490
491
492
work further demonstrated that efficient charge separation can also occur even on the same facets
of the cubic NaTaO semiconductor photocatalyst with the equivalent facets instead of spatial
3
separation of reduction and oxidation cocatalysts on the different facets. Presence of the dual
cocatalysts might be a prerequisite for promoting the redox reactions on the same surface.
4
93
CONCLUSIONS
4
94
95
Using cubic NaTaO with six equivalent {001} facets as a model semiconductor photocatalyst,
3
4
no preferential deposition of noble metals (Pt and Ag) and metal oxides (MnO and PbO ) on a
2
2
496
particular facet confirms that it could not be the facet charge separation that lead to
photocatalytic overall water splitting. However, observation of efficient overall water splitting
upon loading of wellꢀknown cocatalyst NiO clearly demonstrates that photogenerated electrons
and holes could still be wellꢀseparated. Impregnation and photoreduction methods were used
respectively for random (reduction and oxidation sites) and selective (reduction sites) deposition
497
498
499
500
ACS Paragon Plus Environment
26

Page 27 of 32
ACS Catalysis
1
2
3
5
5
5
5
01
02
03
04
of various amounts of NiO cocatalyst on the surface of NaTaO , which suggest that the location
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of cocatalyst affects the charge separation. XPS and XAFS analyses of the samples before and
after photocatalytic water splitting reactions revealed that reduced Ni species (metallic Ni) were
formed inꢀsitu by photoreduction of NiO during the photocatalytic water splitting reactions,
demonstrating that the actual efficient water splitting cocatalysts are metallic Ni and NiO located
on the same facets. The enhancement of the overall photocatalytic water splitting activity upon
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
505
5
06
07
5
NiO loading on the NaTaO with equivalent facets is due to the synergetic effect between NiO
3
508
and metallic Ni, which are derived during the photocatalytic reaction. The metallic nickel serves
as an electron trap for proton reduction and NiO serves as a hole trap for water oxidation.
Presence of the dual cocatalyst Ni/NiO is a prerequisite for the promotion of the photocatalytic
5
09
10
5
511
overall water splitting, specially for the semiconductors, like NaTaO crystal with equivalent
3
5
5
5
12
13
14
facets exposed.
ASSOCIATED CONTENT
5
15
The Supporting Information is available free of charge on the Internet at http://pubs.acs.org.
516
517
AUTHOR INFORMATION
Corresponding Author
5
18
*(Can Li). Email: canli@dicp.ac.cn, Tel: 86ꢀ411ꢀ84379070; Fax: 86ꢀ411ꢀ84694447;
Homepage: http:// www.canli.dicp.ac.cn
519
5
20
*(Hongxian Han). Email: hxhan@dicp.ac.cn, Tel: 86ꢀ411ꢀ84379760
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 32
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
21
*(Shijun Liao). Email: chsjliao@scut.edu.cn.
522
ACKNOWLEDGMENT
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
23
24
We thank Prof. Minrun Li from Dalian Institute of Chemical Physics for useful discussions and
5
HRTEM characterization of the structure of NaTaO .This work was financially supported by the
3
525
National Natural Science Foundation of China (No. 21473189), the 973 National Basic Research
Program of China (No. 2014CB239401).
5
5
26
27
REFERENCES
528
(1)
(a) Hisatomi, T.; Kubota, J.; Domen, K. Chem. Soc. Rev. 2014, 43, 7520ꢀ7535; (b) Chen,
X.; Shen, S.; Guo, L.; Mao, S. S. Chem. Rev. 2010, 110, 6503ꢀ6570; (c) Kudo, A.; Miseki, Y.
Chem. Soc. Rev. 2009, 38, 253ꢀ278.
5
29
30
5
5
31
32
(2)
(3)
(4)
(5)
(a) Ma, Y.; Wang, X.; Jia, Y.; Chen, X.; Han, H.; Li, C. Chem. Rev. 2014, 114, 9987ꢀ
10043; (b) Yang, J.; Wang, D.; Han, H.; Li, C. Acc. Chem. Res. 2013, 46, 1900ꢀ1909.
5
533
(a) Hou, Y.; Zuo, F.; Dagg, A.; Feng, P. Nano Lett. 2012, 12, 6464ꢀ6473; (b) Su, J.; Guo,
L.; Bao, N.; Grimes, C. A. Nano Lett. 2011, 11, 1928ꢀ1933.
5
34
35
5
Böer, K. W., Survey of Semiconductor Physics: Volume II Barriers, Junctions, Surfaces,
and Devices; Springer Science & Business Media, 2012, 2 , 739ꢀ764.
536
537
538
539
(a) Zhang, J.; Xu, Q.; Feng, Z.; Li, M.; Li, C. Angew. Chem., Int. Ed. 2008, 47, 1766ꢀ
1769; (b) Wang, X.; Xu, Q.; Li, M.; Shen, S.; Wang, X.; Wang, Y.; Feng, Z.; Shi, J.; Han,
H.; Li, C. Angew. Chem., Int. Ed. 2012, 51, 13089ꢀ13092.
ACS Paragon Plus Environment
28

Page 29 of 32
ACS Catalysis
1
2
3
5
40
41
(6)
Li, R.; Zhang, F.; Wang, D.; Yang, J.; Li, M.; Zhu, J.; Zhou, X.; Han, H.; Li, C. Nat.
commun. 2013, 4, 1432.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
5
42
43
(7)
(8)
Li, R.; Han, H.; Zhang, F.; Wang, D.; Li, C. Energy Environ. Sci. 2014, 7, 1369ꢀ1376.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
(a) Zheng, Z.; Huang, B.; Lu, J.; Qin, X.; Zhang, X.; Dai, Y. Chem. Eur. J. 2011, 17,
15032ꢀ15038; (b) Pan, J.; Liu, G.; Lu, G. Q. M.; Cheng, H. M. Angew. Chem., Int. Ed. 2011,
50, 2133ꢀ2137; (c) Ohno, T.; Sarukawa, K.; Matsumura, M. New J. Chem. 2002, 26, 1167ꢀ
1170.
544
5
5
45
46
547
(9)
Zhu, J.; Fan, F.; Chen, R.; An, H.; Feng, Z.; Li, C. Angew. Chem., Int. Ed. 2015, 127,
5
48
49
9239ꢀ9242.
5
(10) Li, N.; Liu, M.; Zhou, Z.; Zhou, J.; Sun, Y.; Guo, L. Nanoscale. 2014, 6, 9695ꢀ9702.
5
50
51
(11) (a) Zhang, L.; Shi, J.; Liu, M.; Jing, D.; Guo, L. Chem. Commun. 2014, 50, 192ꢀ194; (b)
Li, R.; Tao, X.; Chen, R.; Fan, F.; Li, C. Chem. Eur. J. 2015, 21,1ꢀ6.
5
552
(12) Jiang, J.; Zhao, K.; Xiao, X.; Zhang, L. J. Am. Chem. Soc. 2012, 134, 4473ꢀ4476.
(13) Kudo, A.; Kato, H.; Nakagawa, S. J. Phys. Chem. B. 2000, 104, 571ꢀ575.
(14) Kohno, M.; Ogura, S.; Sato, K.; Inoue, Y. Chem. Phys. Lett. 1997, 267, 72ꢀ76.
5
53
54
5
5
5
55
56
(15) Iizuka, K.; Wato, T.; Miseki, Y.; Saito, K.; Kudo, A. J. Am. Chem. Soc. 2011, 133,
20863ꢀ20868.
557
(16) Kato, H.; Asakura, K.; Kudo, A. J. Am. Chem. Soc. 2003, 125, 3082ꢀ3089.
ACS Paragon Plus Environment
29