Development of Plasmonic Photocatalyst by Site-selective Loading of Bimetallic Nanoparticles of Au and Ag on Titanium(IV) Oxide

Article abstract of DOI:10.1002/cctc.202000366

Morphology-controlled rutile titanium(IV) oxide (TiO₂) and anatase TiO₂ were prepared by a hydrothermal method and their surfaces were selectively loaded with Au, Ag and AuAg bimetallic nanoparticles (NPs) by photo-deposition in order to obtain visible-light responsive photocatalysts. With the help of local surface plasmon resonance (LSPR) of noble metal NPs, the photocatalytic activity of noble metal-loaded TiO₂ under visible-light irradiation was improved compared to that of bare TiO₂. The enhancement of LSPR on the photocatalytic activity of rutile TiO₂ was larger than that of anatase TiO₂ with optimum amount of Au or AuAg. In addition, the reusability (stability) of AuAg-loaded TiO₂ was much better than that of Ag-loaded TiO₂. Double-beam photoacoustic results confirmed different trap energy level distribution lead to different electron transfer ways, resulting in different oxidizing capabilities and they gave one of important strategies for designing visible-light responsive TiO₂.

Full text of DOI:10.1002/cctc.202000366

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Development of Plasmonic Photocatalyst by Site-selective
Loading of Bimetallic Nanoparticles of Au and Ag on Titanium(IV)
Oxide
Authors: Zhi Zheng, Naoya Murakamia, Jingjing Liu, Zhenyuan Teng,
Qitao Zhang, Yu Cao, Honghui Cheng, and Teruhisa Ohno
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202000366
Link to VoR: https://doi.org/10.1002/cctc.202000366
0
1/2020

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
Development of Plasmonic Photocatalyst by Site-selective
Loading of Bimetallic Nanoparticles of Au and Ag on Titanium(IV)
Oxide
[
a,b]
[a,c]
[b]
[a]
[d]
[a]
Zhi Zheng , Naoya Murakami , Jingjing Liu* , Zhenyuan Teng , Qitao Zhang , Yu Cao ,
[
b]
[a]
Honghui Cheng , Teruhisa Ohno*
[
a]
Zheng, Dr. Murakami, Teng, Cao and Dr. Ohno
Department of Applied Chemistry, Faculty of Engineering
Kyushu Institute of Technology
1
-1 Sensuicho, Tobata, Kitakyushu 804-8550, Japan.
E-mail: zheng.zhi308@mail.kyutech.jp
Dr. Liu and Dr. Cheng
[
[
[
b]
c]
d]
School of Mechanical Engineering
Yangzhou University
Yangzhou, 225127, PR China.
Dr. Murakami
Graduate School of Life Science and Systems Engineering
Kyushu Institute of Technology
2
-4 Hibikino, Wakamatsu-ku, Kitakyushu 808-0196, Japan.
Dr. Zhang
International Collaborative Laboratory of 2D Materials for Optoelectronics Science and Technology of Ministry of Education, Institute of Microscale
Optoelectronics
Shenzhen University
Shenzhen, 518060, China
Supporting information for this article is given via a link at the end of the document.
Abstract: Morphology-controlled rutile titanium(IV) oxide (TiO
2
) and
energy, which results in low efficiency of sunlight utilization.[1-6]
The development of TiO that is responsive to visible light with
anatase TiO were prepared by a hydrothermal method and their
2
2
surfaces were selectively loaded with Au, Ag and AuAg bimetallic
nanoparticles (NPs) by photo-deposition in order to obtain visible-
light responsive photocatalysts. With the help of local surface
plasmon resonance (LSPR) of noble metal NPs, the photocatalytic
high efficiency and enables stable degradation of organic
pollutants is therefore needed.
Due to extensive studies and the development of visible-light
2
responsive photocatalysts, the light absorption of TiO can be
activity of noble metal-loaded TiO
improved compared to that of bare TiO
on the photocatalytic activity of rutile TiO
anatase TiO with optimum amount of Au or AuAg. In addition, the
reusability (stability) of AuAg-loaded TiO was much better than that
of Ag-loaded TiO . Double-beam photoacoustic results confirmed
2
under visible-light irradiation was
. The enhancement of LSPR
was larger than that of
extended from UV to visible light by various methods including
chemical doping, photosensitization and formation of oxygen
2
2
2
vacancies.[7] Chemical doping can narrow the band gap of TiO ,
2
which change in the energy demand of photoexcitation. But it
generates new recombination centers that reduce the efficiency
2
2
of
photosensitization proceeds by electron injection from
photoactive molecules absorbing visible light into TiO . However,
a photocatalytic reaction. Photocatalytic reaction using
different trap energy level distribution lead to different electron
transfer ways, resulting in different oxidizing capabilities and they
gave one of important strategies for designing visible-light
2
this kind of photocatalysts is characterized by poor
photocatalytic reaction due to low stability and oxidation ability of
the photoactive molecules. In the case of oxygen vacancy-type
2
responsive TiO .
2
TiO , its oxidization ability is not suitable for decomposition of
organics due to the fact that the energy level of oxygen
vacancies usually lies at 0.75~1.18 eV below the bottom of
conduction band.[8-18] Local surface plasmon resonance
Introduction
2
Titanium(IV) oxide (TiO ) has a wide range of applications
(
LSPR) by noble metal nanoparticles (NPs) provides an efficient
way to develop visible-light responsive TiO because LSPR
induced charge separation at the interface between TiO and
including degradation of environmental pollutants, carbon
dioxide (CO reduction and water splitting; however,
photocatalytic reaction on TiO requires ultraviolet (UV) light
which accounts for only a small portion of sunlight) because the
large band gap of TiO demand photoexcitation having high
2
2
)
a
2
2
noble metal NPs under visible light irradiation without losing
activity under UV irradiation.[19]
(
2
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
Among noble metal possessing LSPR in the visible-light region,
Au provides a stable LSPR and Ag supplies a strong LSPR
which results in highly efficient photocatalysis, but not vice versa.
A
[
20-26] Bimetallic nanoparticles of Au and Ag (AuAg) have been
further developed for enhancing the efficiency and stability of
plasmonic photocatalysts. The photocatalytic activity for CO
2
reduction to fuels was significantly enhanced due to the
synergistic effect in alloy NPs of Au and Ag.[27] Furthermore,
the AuAg-TiO
2
composites have better performance as
plasmonic photocatalysis which is used for degrading a standard
molecule salicylic acid.[28] Therefore, it is of great importance to
explore synergistic effects of AuAg-loaded TiO
applications.[29, 30]
2
for visible light
In this study, morphology-controlled rutile TiO
TiO were loaded with Au, Ag and AuAg for visible-light
response. Owing to separation of redox sites on the
morphology-controlled TiO [31], the noble metal was site-
2
and anatase
2
B
2
selectively loaded by a photo-deposition method. This site-
selective loading is expected to also provide separation of redox
site under plasmonic photocatalytic reaction, as shown in the
previous study [32]. Photocatalytic activity test revealed that
2
AuAg-loaded TiO shows superior photocatalytic activity under
visible-light irradiation with high stability. When the
photocatalytic activities of all the samples were compared, it is
found that the activity of Au- and AuAg-loaded rutile TiO
much higher than that of Au- and AuAg-loaded anatase TiO
This is an opposite tendency that anatase TiO shows higher
activity in many reactions under UV irradiation than rutile
TiO .[33] By double-beam photoacoustic spectroscopy (DB-
PAS), we analyzed the electron transfer processes of Au- and
AuAg-loaded rutile TiO and anatase TiO under visible light
irradiation and confirmed the reason why Au- and AuAg-loaded
rutile TiO exhibited higher oxidation ability.
2
was
2
.
2
2
2
2
Figure 1. (A) XRD patterns of bare and AuAg-Rutile. (B) XRD patterns of bare
and AuAg-Anatase.
2
Scanning transmission electron microscope (STEM)
observations and energy-dispersive X-ray spectroscopy
Results and Discussion
(
EDS) mapping
Fig. 2 shows scanning transmission electron microscope
STEM) images for 1 wt% AuAg-Rutile and STEM, scanning
X-ray diffraction (XRD) patterns
(
Fig. 1 shows XRD patterns of bare and AuAg-loaded TiO
2
transmission electron microscope equipped with a high-angle
annular dark-field (STEM-HAADF) images and energy-
dispersive X-ray spectroscopy (EDS) mapping for AuAg on the
sample. It can be clearly seen from Fig. 2A the rod-like structure
samples, revealing the crystalline structures and phases
composition of the samples. All of the diffraction peaks for bare
rutile sample in Fig. 1A is assigned to rutile form (JCPDS:21-
1
276) without any other impurity peaks, the strongest diffraction
with
a length of approximately 220 nm and width of
peak attributed to Au and/or Ag (Au JCPDS:65-2870, Ag
o
approximately 40 nm has a flank and a triangular end. In
addition, these results agree well with Fig. S2. It has been
reported that the flank and triangular end of single-crystalline
JCPDS:04-0873) was observed at around 38.1 for 2 wt% AuAg-
loaded TiO
are loaded on the surface of rutile TiO
Fig. 1B for bare anatase sample is assigned to anatase TiO
JCPDS:21-1272) without other impurity peaks. The strongest
2
.[34, 35] These results suggest that Au and/or Ag
2
. The diffraction peaks in
rutile TiO
corresponding to reduction and oxidation sites, respectively.[36,
7] It can also be seen in Fig. 2A that there are some shiny
2
are assigned to {110} and {111} crystal faces
2
(
3
diffraction peak attributed to Au and/or Ag was not observed,
possibly due to overlapping of the strongest peak and anatase
black dots on the {110} crystal face, indicating AuAg was
photocatalytically deposited on the {110} crystal face. This is
reasonable results because {110} crystal face is reduction sites
(
200) peak. Fig. S1 shows the XRD patterns of Au- and Ag-
loaded TiO The diffraction peaks assigned to Au were
observed in XRD patterns for rutile TiO sample loaded with
more than 2 wt% Au, and any other peaks was not observed
apart from rutile TiO and anatase TiO . The possible reason
2
.
2
of rutile TiO . To investigate the structure of AuAg, STEM,
2
STEM-HAADF and EDS mapping were performed. Fig. 2B
presents a spherical particle with its spacing lattices exposed
which is a shiny black dot in Fig. 2A and the spherical particle
connected a flank with its spacing lattices exposed. The
spherical particle exhibiting two spacing lattices and the flank
surface exhibiting one spacing lattice were measured. The two
spacing lattices of the spherical particle is determined to be
2
2
why the peaks assigned to Ag was not observed for Ag-Rutile is
possibly due to low stability and small amount of Ag.
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
0
.235 nm (Au {111}) and 0.238 nm (Ag {111}) and the spacing
The EDS mapping showed that the Au and Ag elements were
located in the same particle. These results proved that AuAg
lattice of the flank is determined to be 0.351 nm (rutile TiO
2
{
110}), confirming the spherical particle is AuAg bimetal NP and
were formed on the reduction site of anatase TiO
photo-deposition.
2
surface by
2
that the flank is the reduction site of rutile TiO . Heavier Au
atoms (atomic number: Z = 79) and Ag atoms (atomic number: Z
47) usually provide a brighter image than do lighter Ti atoms
Z = 22) since the brightness is proportional to the square of the
A
B
C
=
0.238 nm
(
atomic number (atomic resolution Z-contrast imaging) in a
STEM-HAADF image. Fig. 2C therefore reveals that AuAg are
0.235 nm
0
.352 nm
(101)
anatase
supported on the surface of rutile TiO
2
. Fig. 2D~F show EDS
mapping for 1 wt% AuAg loaded on rutile TiO
2
, indicating that Au
1
10 0 nn mm
2nm
2nm
and Ag elements are located on the same particle. These results
also confirm that AuAg were formed on the reduction site of
D
E
F
2
rutile TiO by the photo-deposition method.
aA
bB
C
0.235 nm
2nm
2nm
2nm
0.238 nm
0.351 nm
(
110)
Figure 3. (A) STEM images for 0.5 wt% AuAg-Anatase. (B) STEM and (C)
STEM-HAADF images for AuAg on the sample. EDS mapping images of AuAg
on the sample for (D) Ti+Ag+Au, (E) Ag and (F) Au.
rutile
50nm
5nm
5nm
D
E
F
Optical characterization
The optical properties of bare and AuAg-loaded TiO were
2
revealed by ultraviolet–visible (UV-Vis) diffuse reflectance
spectra, and it is shown in Fig. 4. For bare rutile and bare
anatase, the thresholds of the absorption wavelength were at
approximately 415 nm and 387 nm, respectively. These results
also indicate that it is necessary to control wavelength of
photoirradiation for photocatalytic reaction in order to discuss
photocatalytic reaction induced by LSPR excitation because
5
nm
5nm
5nm
Figure 2. (A) STEM images for 1 wt% AuAg-Rutile. (B) STEM and (C) STEM-
HAADF images for AuAg on the sample. EDS mapping images of AuAg on the
sample for (D) Ti+Ag+Au, (E) Ag and (F) Au.
rutile TiO
00~415 nm of visible light irradiation.
Visible-light absorption for rutile TiO
2
itself is likely to show photocatalytic reaction under
4
2
and anatase TiO
2
In analogy with AuAg rutile TiO
2
, AuAg loaded on the anatase
sample increased as an increase with loading amount, and
peaks were changed depending on loading amount. For AuAg-
Rutile (Fig. 4A) absorption peak was located at 540~560 nm
while AuAg-Anatase showed absorption peaks at 480~510nm
TiO was studied by using STEM, STEM-HAADF and EDS
2
mapping. Fig. 3A shows a polyhedral projection structure with an
average diameter of 13 nm which corresponds to Fig. S3. On
the basis of previous studies [38], the polyhedral projection is
(
Fig. 4B). Fig. S4 shows UV-Vis diffuse reflectance spectra of
Au- and Ag-loaded TiO . Au-loaded anatase TiO and rutile TiO
showed similar absorption peak positions at 540~570 nm. In the
case of Ag-loaded TiO , absorption peak positions at 510~540
nm for Ag-Rutile while Ag-Anatase TiO shows 450~550 nm of
attributed to decahedral anatase TiO
2
with most stable two
2
2
2
exposed crystal faces {101} and {001} corresponding to
reduction and oxidation sites. There are also some small
2
spherical black NPs on the {101} crystal faces of anatase TiO
indicating that AuAg were site-selectively loaded on reduction
sites of anatase TiO .[37, 38] For further study on AuAg on the
anatase TiO , STEM and STEM-HAADF observation and EDS
2
,
2
absorption peak. It is well known that wavelength of absorption
peaks by LSPR depends on kind, shape and dispersity of novel
metal, and the absorption peaks of Au, Ag and AuAg-loaded
2
2
mapping were performed. Fig. 3B shows a spherical black NP
with two kinds of spacing lattices (0.235 nm and 0.238 nm)
loaded on a flank with a spacing lattice in 0.352 nm. The result
reveals that AuAg loaded by photo-deposition are connected
TiO
1] These results suggest that peak wavelength of Ag-loaded
samples was shorter than of Au-loaded samples. Moreover, it is
notable that AuAg-loaded TiO have the ability for absorption
light in a wide range of 425~650 nm, possibly due to a
synergistic effect of LSPR of Au and Ag for TiO , resulting in a
great improvement of the light absorption of noble metal on TiO
2
in the present study agree with that in previous studies.[39-
4
2
well with the reductions sites of anatase TiO
and the flank can be attributed to AuAg and {101} crystal face of
anatase TiO , respectively. Fig. 3C and 3D~F shows a STEM-
HAADF image and EDS mapping of AuAg loaded on the sample.
2
since the black NP
2
2
2
.
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
because noble metal NPs covered TiO
2
active sites or
A
absorption sites for reaction. In Fig. 6, it is notable that the
magnitudes of enhancement by the LSPR effect of Ag on the
photocatalytic activities of rutile TiO
2 2
and anatase TiO are
almost at the same level but that the enhancement by the LSPR
effect of Au and AuAg on the photocatalytic activity of rutile TiO
2
is greater than that on the photocatalytic activity of anatase TiO
2
.
A
B
B
Figure 4. UV-Vis diffuse reflectance spectra of (A) bare and AuAg-Rutile and
(
B) bare and AuAg-Anatase.
Photocatalytic activity tests
2
The photocatalytic activity of TiO samples was evaluated by
oxidation of 2-propanol in gas phase.[42, 43] All samples show
acetone generation as oxidation product of 2-propanol during
early period of photoirradiation, and acetone decrease for some
Au-Rutile was observed with longer irradiation due to oxidation
Figure 5. (A) Time courses of acetone evolution from bare and AuAg-Rutile
and (B) bare and AuAg-Anatase.
from acetone to CO
the surface of TiO
accuracy of the amount of CO
2
. Because organic substance remaining on
during photo-deposition process affects the
production, photocatalytic activity
2
2
Several studies indicate that photocatalytic reaction using
was estimated from acetone generation during the period. Fig. 5
and Fig. S5 show time courses of acetone evolution from bare
LSPR excitation over Au-TiO
from Au into TiO and then reduction proceeds on TiO
oxidation proceeds on electron-deficient Au.[20,45,46,47]
Therefore, numbers of injected electrons in rutile TiO is
expected to be larger than that in anatase TiO , judging only
from more positive potential of conduction band of rutile TiO
However, this mechanism is not sufficient to explain much
higher activity of rutile TiO as observed in Fig. 6 because
difference of conduction band potential between rutile TiO and
anatase TiO is not so large (only 0.2 eV). Moreover, it is shown
2
is induced by electron injection
2
2
while
2
and Au-, Ag-, AuAg-loaded TiO samples under visible light
irradiation. It can be clearly seen that bare rutile and bare
anatase exhibited a visible-light response. The weak visible-light
response is thought to result from surface defects in as-prepared
2
2
2
.
2 2
rutile TiO and anatase TiO .[44] It is notable that the acetone
evolution for all samples increased almost linearly for 6 h of
photoirradiation. By linear fitting, the reaction rate of acetone
evolution was calculated and is shown in Fig. 6. The
photocatalytic reaction rate was improved by Au, Ag and AuAg
2
2
2
that water oxidation, which requires relatively high oxidation
ability, is performed using LSPR excitation [48,49] despite that
oxidation potential of electron-deficient Au is estimated to be not
so high. Therefore, in order to discuss difference of activity, the
loading on rutile TiO
The results suggested that the LSPR effect enhanced the
photocatalytic activity under visible-light irradiation.
2 2
and anatase TiO for almost all samples.
Photocatalytic activity of noble metal loaded sample depended
on the loading amount, and excess amount of the loading
caused a decrease in photocatalytic activity. This reason is
2
behavior of electrons in TiO was observed by double-beam
photoacoustic spectroscopy (DB-PAS) measurement[50,51],
and the detailed mechanism is proposed in the next section.
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
+
0.817 V vs. NHE, pH 7) induced by AuAg (or Au) as co-catalyst.
The reason why anatase TiO
2
does not show this kind of
rutile
anatase
reaction can be explained by difference of energy level
distribution of electron trapping state.[53-56] The mechanism is
summarized in Scheme 1. No large difference between Au and
AuAg was observed in result of DB-PAS measurements, which
is attributable to optical response, as suggested in Fig. 4 and S4.
Figure 6. Photocatalytic reaction rates of acetone evolution for 6
h of
photoirradiation.
Detection of electron behavior under photoirradiation by
DB-PAS analysis
Fig. 7A shows time-course curves of photoacoustic (PA)
intensity for Au-loaded, AuAg-loaded, and bare anatase TiO
under visible-light irradiation. In the case of Au- and AuAg-
loaded anatase TiO , PA intensity increased under visible-light
2
2
irradiation whereas PA intensity for bare anatase showed no
changes. This increase is attributed to trapped electrons such as
3+
trivalent titanium (Ti ) species, which is produced by injected
electrons from Au and AuAg. Thus, an increase in PA intensity is
attributed to the amount of injected electrons. After stopping
visible-light irradiation, PA intensity decreased possibly due to
+
electron consumption by reduction of H .
Fig. 7B shows time course curves of PA intensity for Au-
loaded, AuAg-loaded, and bare rutile TiO
irradiation. Surprisingly, PA intensity decreased under visible-
light irradiation in the case of Au- and AuAg-loaded rutile TiO
while that of bare rutile TiO did not change. This result is
contrary to that of anatase TiO , and this decrease in PA
intensity is presumably caused by transfers of trapped electron
in TiO into electron-deficient Au and AuAg. In detail, this
2
under visible-light
Figure 7. Time course curves of PA signals of (A) anatase TiO
TiO with and without Au- and AuAg loading. Double arrows denote duration of
visible-light irradiation in the range of 470~750 nm.
2
and (B) rutile
2
2
2
2
2
mechanism can be explained by three processes as follows; (1)
excited electrons are generated by LSPR absorption of Au and
AuAg, and they are consumed by reduction. (2) electrons in the
trap energy level transfer to the electron-deficient Au and AuAg,
and Au and AuAg return to the original state. (3) holes are
2
generated in the valence band of TiO by electron excitation
from valence band to unoccupied state of energy level of
trapped electron, and the holes oxidize organic substrates.
Actually, Ti3+ species were confirmed in X-ray photoelectron
spectrometry (XPS) spectrum for AuAg-Rutile (Fig. S6).
2
Because oxidation proceeds on TiO by valence-band holes in
this case, excellent oxidation ability is available. Similar reaction
mechanism using two-step excitation has been reported.[50-52]
Scheme 1. Proposal mechanism on photocatalytic reaction over Au- and
AuAg-loaded anatase TiO and rutile TiO through LSPR excitation.
2
2
According to conduction band potential of rutile TiO
2
(−0.413 V
vs. NHE, pH 7), a counterpart reaction of the oxidation is
−
possibly one-electron reduction of oxygen (O
−(aq), −0.284 V vs. NHE; O
O + H + e = HO
2 2 2
2
) (e.g., O
(aq), −0.459 V vs.
NHE, pH 7) if the reaction proceeds at pH 7. The other
2
+ e =
Stability of noble metal-loaded TiO
2
+
−
The durability of a photocatalyst is an important factor for
evaluating whether it can be used for practical applications. Fig.
S7 shows photocatalytic cyclic abilities of Au- and Ag- loaded
possibility is multi-electron reduction (e.g., O
(aq), +0.269 V vs. NHE, pH 7; O
+ 4H+ + 4e− = 2H
2
+ 2H+ + 2e−
=
H
2
O
2
2
2
O,
TiO
2
. Au-loaded TiO
2
had good photocatalytic cycle ability, and
and anatase
photocatalytic efficiency of Au-loaded rutile TiO
2
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
TiO
respectively (Fig. S7a and b). By contrast, Ag-loaded TiO
poor photocatalytic cycle ability, and photocatalytic efficiency of
Ag-loaded TiO after 3 cycles decreased to 37.5 % and 21.4 %,
respectively (Fig. S7c and d). Therefore, neither of them can be
used for practical applications because of low photocatalytic
2
after 3 cycles were maintained at 85.1 % and 95.1 %,
consists of two components which attributed to metallic Ag0
(368.6 eV and 374.2 eV) and Ag (367.9 eV and 373.9 eV). For
+
2
had
the sample after the photocatalytic activity test, the ratio of peak
intensity for Ag to that for Ag0 become slightly stronger,
compared to that before the photocatalytic reaction. This
suggests that Ag element included in AuAg was partially
oxidized, but Ag0 is maintained to some extent even after
photocatalytic activity test.[37,57] Fig. S8 shows Ag 3d XPS
spectra of 0.5 wt% AuAg-Anatase. Although the noisy was
relatively high, a similar tendency could still be observed. Fig. 9B
shows Ag 3d XPS spectra of 0.5 wt% Ag-Rutile. Before
+
2
reaction rate for Au-loaded TiO
2
and poor durability for Ag-
2
having superior photocatalytic
loaded TiO . AuAg-loaded TiO
2
activity under visible light were chosen for assessment of cyclic
performance in order to investigate the durability. Fig. 8 shows
the disabilities of 1 wt% AuAg-Rutile and 0.5 wt% AuAg-Anatase
under visible-light irradiation. As can be seen in Fig. 8A, that the
+
photocatalytic reaction, strong intensity of Ag peaks reveal that
acetone evolution for AuAg-Rutile showed
a
decrease in
2
Ag NPs on the surface of TiO surface were partially oxidized.
+
photocatalytic efficiency after 3 cycles, and the efficiency was
well maintained at 85.1 %. AuAg-Anatase also showed a good
cyclic ability, and the efficiency was maintained at 85.3 % (Fig.
After photocatalytic reaction, the intensity of Ag becomes lower
than before indicating severe oxidation during photocatalysis.
These results suggest that the stability of AuAg was still much
better than Ag, and Au in AuAg is beneficial for the stability of
8
B) after 3 cycles. On the basis of these results, it can be
0
concluded that the combination of Au and Ag provides good
stability and a high level of ability for oxidizing organics under
visible-light irradiation.[27]
Ag state. These results agree with the cyclic ability of
photocatalytic test (Fig. 8 and S7).
A
A
B
B
Figure 9. Ag 3d XPS spectra of (A) 1 wt% AuAg-Rutile and (B) 0.5 wt% Ag-
Rutile before and after the photocatalytic activity test.
Figure 8. The photocatalytic cyclic abilities of (A) 1 wt% AuAg-Rutile and (B)
0
.5 wt% AuAg-Anatase photocatalysts.
Action spectrum experiments
In order to determine whether the photocatalytic reaction
occurs via photo absorption of AuAg or by a thermocatalytic
process, an action spectrum was measured. To obtain an action
spectrum, the amount of acetone produced by oxidizing 2-
propanol over 1 wt% AuAg-Rutile was measured under light
irradiation with different wavelengths. Fig. 10 shows action
spectrum of apparent quantum efficiency (AQE) for acetone
evolution from 2-propanol decomposition over 1 wt% AuAg-
Rutile. The action spectrum corresponds well to the absorption
spectrum calculated by Kubelka-Munk function, revealing that
X-ray photoelectron spectrometry (XPS) measurements
As is well known, Ag NPs are very sensitive to oxidation and
are easily oxidized. In order to confirm the oxidation states of
Ag element included in AuAg, XPS measurements were
performed on AuAg-loaded samples before and after
photocatalytic tests.
Fig. 9A shows Ag 3d XPS spectra of 1 wt% AuAg-loaded rutile
2
TiO . Both spectra of the samples before and after photocatalytic
activity test exhibit two peaks at 368.6 eV and 374.2 eV of
binding energies, which are attributed to typical values of Ag
3d5/2 and Ag 3d3/2. These peaks indicated that each peak
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
-1
the behavior of the photocatalytic reaction is induced by photo
absorption of LSPR of AuAg.
the supernatant was lower than 10 µS cm . The residue was dried under
reduced pressure at 60 oC. Finally, obtained powders was ground and
irradiated by UV light in order to remove organic compounds on its
surface.
Synthesis of anatase TiO2
A solution containing 10 mL of titanium(IV) ethoxide, 50 mL of ethanol
and 3 mL of deionized water was stirred for 30 min. The suspension was
centrifuged, and the precipitate was dried in a vacuum oven at 60 oC.
Powders obtained by grinding with agate mortar were dispersed in 150
mL of a solution containing deionized water and 50 mL of hydrogen
peroxide, and a yellow-colored solution was obtained by stirring the
solution. The pH of the solution was adjusted to 7 by addition of ammonia
solution, and poly(vinyl alcohol) was added to the solution as an anatase
formation inducer. After stirring for 24h, the solution was put in a Teflon
cup sealed with Teflon-lined autoclave reactor and heated at 200 oC for
4
8 h in an oven.[38] The precipitate was centrifuged and washed several
times and then dried in a vacuum oven at 60 oC. Finally, the organic
compounds that remained on the surfaces of particles were removed by
UV light irradiation.
Figure 10. Action spectrum of AQE for acetone evolution from 2-propanol
decomposition over 1wt% AuAg-Rutile.
Preparation of Au-, Ag- and AuAg-loaded TiO2
Conclusion
Au, Ag and AuAg NPs were loaded on as-prepared TiO2 by photo-
deposition. An aqueous suspension containing 60 mL of water and 200
By photo-deposition on morphology-controlled TiO
and AuAg NPs were site-selectively loaded on reduction sites of
TiO . The AuAg-loading on TiO provides not only good activity
2
, Au, Ag
mg of as-prepared TiO powder was sonicated. Then, 20 mL of methanol
2
was added to the suspension and it was bubbled with nitrogen for 30 min.
Before photoirradiation with a mercury lamp, a specific amount of 0.029
mol/L HAuCl4 or 0.05 mol/L AgNO3 was injected into the suspension and
it was stirred for 30 min to achieve a homogeneous mixture. The
suspension was irradiated with a light intensity of 220 mW cm-2 for 1.5 h,
and Au-loaded TiO2 or Ag-loaded TiO2 powder was obtained after
filtration and drying under reduced pressure.
2
2
performance for oxidizing 2-propanol but also stability under
visible-light. DB-PAS analysis revealed that Au- and AuAg-
loaded rutile TiO
photocatalytic reaction under visible light compared to Au- and
AuAg-loaded anatase TiO . Due to the different distribution of
electron trap energy levels of rutile TiO and anatase TiO , the
electron transfer processes and where the oxidation reaction
occurs of Au- and AuAg- loaded rutile TiO and anatase TiO are
different. The Au- and AuAg- loaded rutile TiO oxidation site
has more positive valence band position compared to the Au-
and AuAg- loaded anatase TiO oxidation site, which means that
the Au- and AuAg- loaded rutile TiO has a stronger oxidation
ability than the Au- and AuAg- loaded anatase TiO . The results
confirmed the outstanding performance of 1 wt% AuAg-loaded
rutile TiO and a possible way to develop highly efficient
2
exhibited different electrons transfer ways for
2
For preparation of AuAg-loaded TiO2, a suspension containing water
and as-prepared TiO2 was sonicated, and it was bubbled with nitrogen
after methanol addition. A specific amount of AgNO3 was injected into
suspension and it was irradiated for 45 min. And then, HAuCl4 was
injected into the solution and the suspension was irradiated for another
2
2
2
2
2
4
5 min. Finally, AuAg-loaded TiO2 powder was obtained after filtration
and drying. By quantitative analysis of noble metals on the TiO2 surface,
it can be confirmed that all noble metals are supported on the TiO2
surface. In this article, Au, Ag, AuAg loaded rutile TiO2 and anatase TiO2
2
2
2
are
named
as
Au-Rutile/Anatase,
Ag-Rutile/Anatase,
AuAg-
Rutile/Anatase. The amounts of Au, Ag and AuAg loaded on the surface
of TiO2 were estimated by analysis of filtrate solution after
photodeposition by inductively coupled plasma atomic emission
spectroscopy (ICP-AES). The ICP-AES results showed that 99.8~ 99.9%
of Au and Ag were loaded on the TiO2, and the ratio of Au and Ag in
AuAg bimetal modified TiO2 was almost 1:1 in a weight ratio.
2
photocatalysts based on LSPR.
Experimental Section
Characterization
Chemicals
The crystalline phases of prepared samples were identified by using a
powder X-ray diffraction (XRD) instrument (MiniFlex II, Rigaku Co.). UV-
vis diffuse reflectance spectra were acquired by a spectrometer equipped
with an integral sphere (UV-2600, Shimadzu Co.), and relative
reflectance was converted to absorbance using Kubelka−Munk function.
The morphology of samples were characterized by using a field emission
high-resolution scanning transmission electron microscope (STEM;
FEI Titan Cubed Themis G2 300) equipped with a high-angle annular
dark-field (HAADF) detector and energy-dispersive X-ray spectroscopy
Chemical reagents including titanium(IV) chloride, titanium(IV) ethoxide,
ethanol, methanol, ammonia solution, hydrogen peroxide solution and
poly(vinyl alcohol) (partially hydrolyzed, average degree of
polymerization: 3,500) and acetone were purchased at a reagent level
from Wako Chemicals Co., and they were used in experiments without
further purification.
Synthesis of rutile TiO2
(
EDS) for elemental mapping. The specific surface areas of prepared
A
50-mL solution containing 15 mL of titanium(IV) chloride and
samples were measured with a surface area analyzer (Nova 4200e,
Quantachrome Co.) by the Brunauer-Emmett-Teller (BET) equation. The
specific surface areas of the prepared samples are summarized in Table.
S1. X-ray photoelectron spectrometry (XPS) spectra were measured in
order to confirm oxidation state of AuAg loaded on TiO2 by using an
deionized water was stirred for 30 min and then put into a Teflon cup. It
was sealed with a stainless jacket and heated at 200 oC for 24 h in an
oven. After the hydrothermal treatment, the precipitate in the Teflon cup
was washed and centrifuged several times until the ionic conductivity of
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
electron spectrometer (ESCALAB 250X, Thermo Co.). The Al kα
excitation line was used, and the shift of binding energy was calibrated
by using the C 1s level at 284.8 eV as an internal standard.
[14] T. Ohno, T. Tsubota, K. Nishijima, Z. Miyamoto, Chemistry Letters 2004,
33, 750-751.
[15] T. Ohno, Water Science and Technology 2004, 49, 159-163.
[
16] R. Sadowski, A. Wach, M. Buchalska, P. Kuśtrowski, W. Macyk, Applied
Surface Science 2019, 475, 710-719.
Photocatalytic decomposition of 2-propanol
[
[
17] A.S. Hoffman, Macromol. Symp. 1996, 101, 443-454.
18] D. Sung, S. Park, S. Jon, Langmuir 2012, 28, 4507-4514.
The photocatalytic activity of TiO2 was evaluated by determining the
changes in the concentrations of evolved acetone and CO2. A powder
sample (200 mg) was dispersed in a glass dish (4 cm ) and the dish was
put into a Tedlar bag with a volume of 125 ml mixed air (79 % N2, 21 %
O2, <0.1 ppm of CO2, 500 ppm of 2-propanol). A 500-W xenon lamp
[19] N. Zhang, C. Han, X. Fu, Y. Xu, Chem 2008, 4, 1832-1861.
[20] Y. Tian, T. Tatsuma, J. Am. Chem. Soc. 2005, 127, 7632-7637.
[21] W. Hou, W.H. hung, P. Pavascar, A. Goeppert, M. Aykol, S.B. Cronin, J.
ACS Catalysis. 2011, 1, 929-936.
2
(
Ushio, SX-UI501XQ) was used as
a light source, and the light
[
22] A. Tanaka, Y. Nishino, S. Sakaguchi, T. Yoshikawa, K. Imamura, K.
Hashimoto, H. Kominami, Chem. Commun. 2013, 49, 2551-2553.
23] Y. Sugano, Y. Shiraishi, D. Tsukamoto, S. Ichikawa, S. Tanaka, T. Hirai,
Angew.Chem. Int. Ed. 2013, 53, 5295-5299.
wavelength was controlled by a UV-47 filter and a heat-absorbing filter
-2
and light intensity were adjusted to 110 mW cm . Light irradiation was
started at room temperature after the sample reached an absorption
equilibrium. After starting the irradiation, the concentration of evolved
acetone and CO2 were measured by gas chromatography (3000 Micro
GC, Agilent/Inficon) with a PLOTU column and an OV-1 column.
[
[24] K. Fuku, R. Hayashi, S. Takakura, T. Kamegawa, K. Mori, H. Yamashita,
Angew. Chem. Int. Ed. 2013, 52, 7446-7450.
For measurement of action spectrum, apparent quantum efficiency
[
[
[
25] P. Dong, F. Yang, X. Cheng, Z. Huang, X. Nie, Y. Xiao, X. Zhang,
Materials Science and Engineering: C 2019, 96, 197-204.
(
AQE) was calculated by acetone generation and the amount of incident
photons using a xenon lamp quipped with a band-pass filter (Asahi
Spectra Co.) centered at 425 nm, 450 nm, 500 nm, 550 nm, 600 nm and
26] X. Cheng, P. Dong, Z. Huang, Y. Zhang, Y. Chen, X. Nie, X. Zhang,
Journal of CO
27] M. Tahir, B. Tahir, N.A.S. Amin, Appl. Catal. B: Environ. 2017, 204, 548-
60.
2
Utilization 2017, 20, 200-207.
6
50 nm.
5
Double-beam photoacoustic spectroscopic measurement
[28] S. Kamimura, T. Miyazaki, M. Zhang, Y. Li, T. Tsubota, T. Ohno, Appl.
Catal. B: Environ. 2016, 180, 255-262.
A gas-exchangeable photoacoustic (PA) cell equipped with two valves for
gas flow was used, and a sample was placed in the cell. The atmosphere
was controlled by a flow of nitrogen-containing ethanol vapor, and the
measurements were conducted after shutting off the valves, i.e., in a
closed system at room temperature. A light-emitting diode (LED) emitting
light at ca. 940 nm (SFH 4751, OSRAM) was used as a probe light for
detection of electron accumulation, and the output intensity was
modulated at 40 Hz. Visible light in the range of 470~750 nm was
extracted using a white LED (CXA1512-0000-000F0HM240F, CREE)
and a sharp cut filter (SCF-50S-48Y, SIGMA KOKI), it was also used as
simultaneous continuous irradiation for photoexcitation of LSPR in
addition to the modulated light. The PA signal acquired by a MEMS
microphone (INMP401, TDK InvenSense) buried in the cell was amplified
and monitored by a digital lock-in amplifier (LI5640, NF). Detailed setups
of DB-PAS measurements have been reported.[58, 59]
[29] N. Zhou, L. Polavarapu, N. Gao, Y. Pan, P. Yuan, Q. Wang, Q.H. Xu,
Nanoscale 2013, 5, 4236-4241.
[30] Y. Qu, R. Cheng, Q. Su, X. Duan, J. Am. Chem. Soc. 2011, 133, 16730-
16733.
[31] F. Dufour, S. Pigeot-Remy, O. Durupthy, S. Cassaignon, V. Ruaux, S.
Torelli, L. Mariey, F. Mauge, C. Chaneac, Appl. Catal. B: Environ. 2015,
174-175, 350-360.
[32] M. Qi, Y. Li, F. Zhang, Z. Tang, Y. Xiong, Y Xu, ACS Catalysis 2020, 10,
3194-3202.
[33] O.O.P. Mahaney, N. Murakami, R. Abe, B. Ohtani, Chem. Lett. 2009, 38,
238-239.
[34] C.B. Gao, Y.X. Hu, M.S. Wang, M.F. Chi, Y.D. Yin, J. Am. Chem. Soc.
2014, 136, 7474-7479.
[35] D.Q. Zhang, M.C. Wen, S.S. Zhang, P.J. Liu, W. Zhu, Appl. Catal. B:
Environ. 2014, 147, 610-616.
[
[
36] E. Bae, T. Ohno, Applied Catalysis B: Environmental 2009, 91, 634–639.
37] T. Ohno, K. Sarukawa, M. Matsumura, New J. Chem. 2002, 26, 1167-
1
170.
[
38] N. Murakami, Y. Kurihara, T. Tsubota, T. Ohno, J. Phys. Chem. C 2009,
113, 3062-3069.
2
Keywords: • electron transfer • morphology-controlled TiO •
surface plasmon resonance • visible-light responsive
[39]
Y. Ma, K. Kobayashi, Y. Cao, T.Ohno, Appl. Catal. B: Environ. 2019,
45, 681-690.
40] E. Kowalska, O.O.P. Mahaney, R. Abe, B. Ohtani, Phys. Chem. Chem.
2
[
[
[
[
1]
2]
A. Fujishima, K. Honda, Nature 1972, 238, 37-38.
M.R. Hoffmann, S.T. Martin, W. Choi and D.W. Bahnemann, Chem. Rev.
995, 95, 69-96.
Phys. 2010, 12, 2344-2355.
41]
K. Awazu, M. Fujimaki, C. Rockstuhl, J. Tominaga, H. Murakami, Y.
Ohki, N. Yoshida, T. Watanabe, J. Am. Chem. Soc. 2008, 130, 1676-
1
[
[
3]
4]
D.F. Ollis, Comptes Rendus de l'Académie des Sciences - Series IIC -
Chemistry 2000, 3, 405-411.
1
680.
42] Y. Ohko, K. Hashimoto, A. Fujishima, J. Phys. Chem. A 1997, 101,
057-8062.
[
[
[
[
[
[
[
[
T. Ohno, T. Higo, N. Murakami, H. Saito, Q.T. Zhang, Y. Yang, T.
Tsubota, Applied Catalysis B: Environmental 2014,152, 309-316.
T.E. Mallouk, J. Phys. Chem. Lett. 2010, 1, 2738-2739.
8
43]
H. Yamashita, M. Honda, M. Harada, Y. Ichihashi, M. Anpo, J. Phys.
Chem. B 1998, 102. 10707-10711.
[
[
5]
6]
A.C. Benniston, A. Harriman, Artificial photosynthesis, Mater. Today.
44] F. Zuo, L. Wang, T. Wu, Z. Zhang, D. Borchardt, P. Feng, J. Am. Chem.
Soc. 2010, 132, 11856−11857.
2
008, 11, 26-34.
D. Zhang, X. Ma, H. Zhang, Y. Liao, Q. Xiang, Materials Today Energy
018, 10, 132-140.
N. Sharotri, D. Sharma, D. Sud, J MATER RES TECHNOL 2019, 8,
995-4009.
[
[
[
[
7]
8]
9]
45] A. Furube, L. Du, K. Hara, R. Katoh, M. Tachiya, J. Am. Chem. Soc.
2
2
007, 129, 14852−14853.
46] A. Tanaka, S. Sakaguchi, K. Hashimoto, H. Kominami, ACS Catal. 2013,
, 79-85.
47] K. Kimura, S. Naya, Y. Jin-nouchi, H. Tada, J. Phys. Chem. C 2012, 116,
111-7117.
3
3
I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihara, K. Takeuchi,
J Mol Catal A Chem 2000, 161, 205-212.
7
10] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 2001, 293,
69-271.
48]
S. Wang, Y. Gao, S. Miao, T. Liu, L. Mu, R. Li, F. Fan, C. Li, J. Am.
Chem. Soc. 2017, 139, 11771-11778.
2
[
[
11] N. Sharotri, D. Sud, Desalin Water Treat 2016, 57, 8776-8788.
12] T. Ohno, M. Akiyoshi, T. Umebayashi, K. Asai, T. Mitsui, M. Matsumura,
Applied Catalysis A: General 2004, 265, 115-121.
49] A. Tanaka, K. Teramura, S. Hosokawa, H. Kominami, T. Tanaka, Chem.
Sci. 2017, 8, 2574-2580.
[
13] T. Ohno, T. Mitsui, M. Matsumura, Chemistry letters 2003, 32, 364-365.
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
[
[
[
[
50] A. Tanaka, K. Hashimoto, H. Kominami, J. Am. Chem. Soc. 2014, 136,
86-589
5
51] K.K. Paul, P.K. Giria, H. Sugimoto, M. Fujii, B. Choudhury, Sol. Energy
Mater. Sol. Cells 2019, 201, 110053-110064.
52] T. Kashyap, S. Biswasi, A.R. Pal, B. Choudhury, ACS Sustainable Chem.
Eng. 2019, 7,19295−19302
53] S. Ikeda, N. Sugiyama, S. Murakami, H. Kominami, Y. Kera, H. Noguchi,
K. Uosaki, T. Torimoto, B. Ohtani, Phys. Chem. Chem. Phys. 2003, 5,
7
78-783.
54] A. Yamakata, J.J.M. Vequizo, H. Matsunaga, J. Phys. Chem. C 2015,
19, 24538-24545
[
[
1
55] A. Nitta, M. Takase, M. Takashima, N. Murakami, B. Ohtani, Chem.
Comm. 2016, 52, 12096-12099
[
[
56] T. Shinoda, N. Murakami, J. Phys. Chem. C 2019, 123, 12169-12175
57]
S. Kamimura, S. Yamashita , S. Abe , T. Tsubota, T. Ohno, Applied
Catalysis B: Environmental 2017, 211, 11-17.
[
[
58] N. Murakami, O.O.P. Mahaney, R. Abe, T. Torimoto, B. Ohtani, J. Phys.
Chem. C 2007, 111, 11927-11935
59]
N. Murakami, T. Chiyoya, T. Tsubota, T. Ohno, Appl. Catal. A: Gen.
008, 348, 148-152.
2
9
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.202000366
FULL PAPER
Entry for the Table of Contents
Different electron transfer ways of Au- or AuAg- modified rutile TiO
2
and anatase TiO
2
for photocatalytic reaction under visible light
were revealed. The distribution of trapped energy level determines the transfer and consumption ways of photo-generated electrons
and determines the oxidization for photocatalytic reaction.
1
0
This article is protected by copyright. All rights reserved.