Control of Surface Plasmon Resonance of Au/SnO2 by Modification with Ag and Cu for Photoinduced Reactions under Visible-Light Irradiation over a Wide Range

Article abstract of DOI:10.1002/chem.201504606

Gold particles supported on tin(IV) oxide (0.2 wt % Au/SnO₂) were modified with copper and silver by the multistep photodeposition method. Absorption around λ=550 nm, attributed to surface plasmon resonance (SPR) of Au, gradually shifted to longer wavelengths on modification with Cu and finally reached λ=620 nm at 0.8 wt % Cu. On the other hand, the absorption shifted to shorter wavelength with increasing amount of Ag and reached λ=450 nm at 0.8 wt % Ag. These Cu- and Ag-modified 0.2 wt % Au/SnO₂ materials (Cu-Au/SnO₂ and Ag-Au/SnO₂) and 1.0 wt % Au/SnO₂ were used for mineralization of formic acid to carbon dioxide in aqueous suspension under irradiation with visible light from a xenon lamp and three kinds of light-emitting diodes with different wavelengths. The reaction rates for the mineralization of formic acid over these materials depend on the wavelength of light. Apparent quantum efficiencies of Cu-Au/SnO₂, Au/SnO₂, and Ag-Au/SnO₂ reached 5.5 % at 625 nm, 5.8 % at 525 nm, and 5.1 % at 450 nm, respectively. These photocatalysts can also be used for selective oxidation of alcohols to corresponding carbonyl compounds in aqueous solution under visible-light irradiation. Broad responses to visible light in formic acid mineralization and selective alcohol oxidation were achieved when the three materials were used simultaneously.

Full text of DOI:10.1002/chem.201504606

DOI: 10.1002/chem.201504606
Full Paper
&
Photocatalysts
Control of Surface Plasmon Resonance of Au/SnO by
2
Modification with Ag and Cu for Photoinduced Reactions under
Visible-Light Irradiation over a Wide Range
[a, b]
[a]
[a]
Atsuhiro Tanaka,
Keiji Hashimoto, and Hiroshi Kominami*
Abstract: Gold particles supported on tin(IV) oxide (0.2 wt%
with visible light from a xenon lamp and three kinds of
light-emitting diodes with different wavelengths. The reac-
tion rates for the mineralization of formic acid over these
materials depend on the wavelength of light. Apparent
quantum efficiencies of Cu-Au/SnO , Au/SnO , and Ag-Au/
Au/SnO ) were modified with copper and silver by the multi-
2
step photodeposition method. Absorption around l=
5
50 nm, attributed to surface plasmon resonance (SPR) of
Au, gradually shifted to longer wavelengths on modification
with Cu and finally reached l=620 nm at 0.8 wt% Cu. On
the other hand, the absorption shifted to shorter wave-
length with increasing amount of Ag and reached l=
2
2
SnO reached 5.5% at 625 nm, 5.8% at 525 nm, and 5.1% at
2
450 nm, respectively. These photocatalysts can also be used
for selective oxidation of alcohols to corresponding carbonyl
compounds in aqueous solution under visible-light irradia-
tion. Broad responses to visible light in formic acid minerali-
zation and selective alcohol oxidation were achieved when
the three materials were used simultaneously.
4
50 nm at 0.8 wt% Ag. These Cu- and Ag-modified 0.2 wt%
Au/SnO2 materials (Cu-Au/SnO2 and Ag-Au/SnO2) and
.0 wt% Au/SnO were used for mineralization of formic acid
1
2
to carbon dioxide in aqueous suspension under irradiation
Introduction
ties, which are associated with their strong photoabsorption in
the visible light region, due to surface plasmon resonance
[
3]
Titanium(IV) oxide (TiO ) is a favorable photocatalyst because it
(SPR). The photoabsorption peaks due to SPR of Au, Ag, and
Cu nanoparticles are generally observed around 550, 450, and
600 nm, respectively. Recently, Au nanoparticles supported on
TiO and cerium(IV) oxide (CeO ) have been applied as a new
2
consists of abundant elements (Ti and O) and is a stable and
safe compound. However, a weakness TiO as a photocatalyst
2
is that it requires irradiation with UV light to induce photocata-
lytic reactions due to its relatively large bandgap (ca. 3.2 eV).
Since UV light accounts for only about 5% of the total solar
energy, sufficient utilization of solar energy cannot be achieved
2
2
[
4]
type of photocatalysts responding to visible light. Previously,
we found that action spectra in photocatalytic reactions over
Au/TiO and Au/CeO were in good agreement with their ab-
2
2
by a TiO photocatalyst. Therefore, development of photocata-
sorption spectra, which suggested that photocatalytic reac-
tions were induced by photoabsorption due to SPR of the sup-
ported Au nanoparticles. In contrast to reports on supported
Au nanoparticles, there have been few reports on chemical re-
actions induced by SPR of Ag and Cu nanoparticles, probably
due to their instability under working conditions. To utilize the
unique SPR property of Cu, we examined modification of
2
lysts responding to visible light is an important topic from
a practical point of view, because visible light accounts for
about 50% of the total solar energy. Most of the visible-light-
[
1]
responsive photocatalysts, such as doped (or modified) TiO
2
[
2]
and narrow-bandgap photocatalysts with a cocatalyst, work
under irradiation of light with wavelengths shorter than l=
4
80 nm. The development of photocatalysts that operate
CeO -supported Au nanoparticles with Cu and found that pho-
2
[
5]
under irradiation of light with wavelengths longer than l=
toabsorption due to SPR was shifted to longer wavelength.
4
80 nm is important in order to utilize solar energy efficiently.
Nanoparticles of metals such as gold, silver, and copper
Over thus-prepared Cu-Au/CeO2 materials, organic acids in
aqueous solutions were mineralized to CO even under irradia-
2
[
5]
have been studied extensively because of their unique proper-
tion with visible light of l=700 nm. These results suggest
that photoabsorption due to SPR of supported Au nanoparti-
cles can be controlled by modification with Ag and Cu and
that the modified metal nanoparticles show a wide response
to visible light, which results in efficient utilization of solar
energy.
[
a] Dr. A. Tanaka, Dr. K. Hashimoto, Dr. H. Kominami
Department of Applied Chemistry, Faculty of Science and Engineering
Kindai University, Kowakae, Higashiosaka, Osaka 577-8502 (Japan)
E-mail: hiro@apch.kindai.jp
[
b] Dr. A. Tanaka
In this study, we examined control of absorption due to SPR
Department of Applied Chemistry, College of Life Sciences
Ritsumeikan University, 1-1-1 Noji-higashi, Kusatsu, Shiga 525-8577 (Japan)
by modification of Au nanoparticles supported on SnO with
2
Ag and Cu. SnO attracted our interest for use as a support for
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201504606.
2
metal nanoparticles having photoabsorption due to SPR since
Chem. Eur. J. 2016, 22, 4592 – 4599
4592
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
it is a wide-bandgap (3.2 eV) photocatalyst that cannot absorb
visible light, the contribution of which is completely eliminated
from the photoabsorption of modified samples in the range of
visible light. The position of the conduction band edge of
and the average diameter of the Au particles was determined
to be 8.2 nm. This indicates that MSPD is applicable for prepa-
ration of Au/SnO with larger Au particles as well as Au/CeO .
2
2
Since MSPD has also been shown to be effective for introduc-
tion of a second metal (Cu) onto Au particles loaded on
[
6,7]
SnO , about 0.5 eV lower than those of TiO and ZnO,
indi-
2
2
[
5]
cates that one-electron reduction of O , that is, formation of
CeO2, it was used for modification of Au/SnO with second
2
2
À
À
À
2
O C radical (O +e !O C ), does not occur. Two-electron re-
metals (Ag and Cu) in this study. As shown in Figure 1c and d,
larger particles were also observed in the TEM images of
Ag(0.8)-Au(0.2)/SnO and Cu(0.8)-Au(0.2)/SnO , and the average
2
2
À
+
duction of O (O +2e +2H !H O ) and four-electron reduc-
2
2
2
2
À
+
tion of O (O +4e +4H !2H O) may occur. Since H O is
2
2
2
2
2
2
2
probably decomposed to water and O over Au nanoparticles,
diameters of the particles were determined to be 8.6 and
8.5 nm, respectively (Supporting Information, Figure S1). In-
creases in particle size after introduction of Ag and Cu clearly
indicate that the second metals were deposited on the Au par-
ticles, and that MSPD is useful for preparation of supported
Ag–Au and Cu–Au bimetallic materials.
2
active oxygen species would be removed from the reaction
system, which also eliminates the contribution of active
oxygen species to photoinduced reactions. Herein, we report
1
) mineralization of formic acid and selective oxidation of alco-
hols in aqueous suspensions of Au/SnO , Ag-Au/SnO , and Cu-
2
2
Au/SnO under irradiation with visible light and 2) simultane-
2
ous use of all three materials to achieve a wide response to
visible light.
Photoabsorption properties
Figure 2 shows absorption spectra of Au(0.2)/SnO , Ag(0.8)-
2
Au(0.2)/SnO , Au(1.0)/SnO , and Cu(0.8)-Au(0.2)/SnO . In the
2
2
2
spectra of Au(0.2)/SnO2 and Au(1.0)/SnO , photoabsorption
2
Results and Discussion
was observed around l=550 nm (Figure 2d,b), which was at-
[
4,5,8,9]
TEM observations
tributed to SPR of the supported Au particles,
and more
intense photoabsorption was achieved by increasing the Au
Figure 1 shows TEM images of unmodified and modified Au/
SnO samples. Figure 1a shows a TEM image of Au(0.2)/SnO .
[
8]
contents of Au/SnO and Au/CeO . Both a shift of photoab-
2
2
2
2
sorption and an increase in intensity were achieved by intro-
Au particles were observed in the TEM image and the average
diameter was determined to be 3.8 nm, that is, Au nanoparti-
cles were successfully deposited on the surface of SnO2 by
using the photodeposition (PD) method. Previously, we found
duction of Ag and Cu into Au(0.2)/SnO by MSPD (Figure 2a,c).
2
The photoabsorption at 550 nm of Ag(0.8)-Au(0.2)/SnO2 and
that of Cu(0.8)-Au(0.2)/SnO at 550 nm were less intense than
2
that of the Au(0.2)/SnO mother material before modification.
2
that the size of Au nanoparticles loaded on CeO gradually in-
2
These results suggest that the SPR properties of Ag(0.8)-
Au(0.2)/SnO and Cu(0.8)-Au(0.2)/SnO are not inherited from
[8]
creased with repeating Au photodeposition. In the multistep
photodeposition (MSPD) method, the metal source is reduced
and deposited as the metal on previously deposited metal par-
ticles acting as reduction sites, which results in the growth of
2
2
Au(0.2)/SnO and originate from the properties of Ag and Cu
2
themselves. In other experiments, however, we did not suc-
ceed in the preparation of Au-free Cu/SnO exhibiting SPR by
2
metal particles on CeO under light irradiation.
2
PD and MSPD. In addition, it is known that Cu nanoparticles
In the present study, we used MSPD for preparation of Au/
SnO2 with larger Au particles. As expected, larger particles
were observed in the TEM image of Au(1.0)/SnO (Figure 1b),
2
Figure 1. TEM images of a) Au(0.2)/SnO
SnO , and d) Cu(0.8)-Au(0.2)/SnO
2
, b) Au(1.0)/SnO
2
, c) Ag(0.8)-Au0.2)/
Figure 2. Absorption spectra of a) Ag(0.8)-Au(0.2)/SnO
c) Cu(0.8)-Au(0.2)/SnO , and d) Au(0.2)/SnO
2 2
, b) Au(1.0)/SnO ,
2
2
.
2
2
.
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4593
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
are easily oxidized and lose their SPR gradually under ambient
[
10]
conditions. Therefore, Au particles are clearly indispensable
for preparation of stable Cu-based particles supported on SnO2
exhibiting intense photoabsorption at 630 nm due to SPR. The
Au particles probably function as a kind of template and stabi-
lizer for Cu and Ag metals. However, we did not observe
a core–shell structure in TEM images of Ag(0.8)-Au(0.2)/SnO2
and Cu(0.8)-Au(0.2)/SnO2.
Figure 3a,b shows the influence of Ag and Cu contents (X
and Y wt%, respectively) in the Ag(X)-Au(0.2)/SnO and Cu(Y)-
2
Au(0.2)/SnO samples on the top wavelength due to SPR ltop
2
and photoabsorption (one-reflection) at ltop. The photoabsorp-
tion at ltop increased with increasing Ag and Cu contents. The
maximum peaks gradually shifted to shorter and longer wave-
lengths with increasing X and Y and reached l=450 nm at X=
Figure 4. Time courses of evolution of CO
pensions of Au(1.0)/SnO
2
from formic acid in aqueous sus-
(circles), Ag(0.8)-Au(0.2)/SnO (squares), and
2
2
Cu(0.8)-Au(0.2)/SnO (diamonds) under irradiation with visible light from
2
À2
green, blue, and red LEDs (1.7 mWcm ).
0
.8 wt% and l=630 nm at Y=0.8 wt%. Thus, the peak posi-
tion of photoabsorption due to SPR can be controlled by
modification of Au(0.2)/SnO with Ag and Cu by MSPD.
2
Figure 3. Influence of a) Ag loading X and b) Cu loading Y on Au(0.2)/SnO
on the top of the wavelength due to SPR ltop and one-reflection photoab-
sorption at ltop
2
.
Figure 5. Spectra of visible light from blue, green, and red LEDs
Photocatalytic mineralization of formic acid in aqueous sus-
pensions of Au(1.0)/SnO , Ag(0.8)-Au(0.2)/SnO , and Cu(0.8)-
À2
(1.7 mWcm ) used for photoinduced reactions.
2
2
Au(0.2)/SnO2
Figure 4 shows time courses of evolution of CO from formic
with green, blue, and red lights were determined to be 0.86,
2
À1
acid in aqueous suspensions of Au(1.0)/SnO , Ag(0.8)-Au(0.2)/
0.93, and 2.1 mmolh , respectively. Owing to the strong pho-
2
SnO , and Cu(0.8)-Au(0.2)/SnO under irradiation from green,
toabsorption of Cu(0.8)-Au(0.2)/SnO at 630 nm (Figure 2c), it
2
2
2
blue, and red light-emitting diodes (LEDs) with maximum
wavelengths of 475, 530, and 640 nm, respectively (Figure 5).
In the presence of Au(1.0)/SnO (circles in Figure 4), CO was
exhibited the highest rate under red-light irradiation. In the
case of Ag(0.8)-Au(0.2)/SnO (squares in Figure 4), the CO for-
2
2
mation rates under irradiation with green, blue, and red lights
2
2
À1
evolved after irradiation with green, blue, and red lights, and
were determined to be 0.62, 2.4, and 0.11 mmolh , respective-
CO formation continued linearly with irradiation time, indicat-
ly. The highest rate for Ag(0.8)-Au(0.2)/SnO was obtained with
2
2
ing zero-order kinetics over Au(1.0)/SnO2.
blue-light irradiation, as expected from the photoabsorption
properties. These results indicate that the materials exhibit the
best performance under irradiation with visible light overlap-
ping with their SPR.
On the other hand, the absence of gas evolution in the dark
indicated that no thermocatalytic mineralization of formic acid
occurred under the present conditions. From the slopes of
time courses of CO evolution, rates over Au(1.0)/SnO under ir-
The stability of Au(1.0)/SnO2, Ag(0.8)-Au(0.2)/SnO2, and
2
2
radiation with green, blue, and red lights were determined to
Cu(0.8)-Au(0.2)/SnO in the mineralization of formic acid was
2
À1
be 2.9, 0.55, and 0.29 mmolh , respectively. As shown in Fig-
evaluated by prolonged use under irradiation with light from
green, blue, and red LEDs, respectively, and all three materials
continuously decomposed formic acid without losing their ac-
ures 2b and 5, photoabsorption due to SPR of Au(1.0)/SnO2
overlapped well with light from the green LED, and the greater
utilization of light (photons) explains the highest rate under
green-light irradiation. Similarly, rates of CO2 formation over
tivity (Figure 6). After irradiation for about 50 h, CO evolution
2
was almost saturated at about 100 mmol, which corresponds to
the initial amount of formic acid (100 mmol) and indicates that
Cu(0.8)-Au(0.2)/SnO (diamonds in Figure 4) under irradiation
2
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4594
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 7. Absorption spectra measured with barium sulfate as a reference
Figure 6. Time courses of evolution of CO
2
from formic acid (100 mmol) in
(squares),
(left axis) and action spectra (circles) in formic acid mineralization (right
axis). a) Ag(0.8)-Au(0.2)/SnO , b) Au(1.0)/SnO , and c) Cu(0.8)-Au(0.2)/SnO .
2 2 2
aqueous suspensions of Au(1.0)/SnO (circles), Ag(0.8)-Au(0.2)/SnO
2
2
and Cu(0.8)-Au(0.2)/SnO
green, blue and red LEDs (1.7 mWcm ), respectively.
2
(diamonds) under irradiation with visible light from
À2
Selective oxidation of various substrates in aqueous suspen-
sions of Ag(0.8)-Au(0.2)/SnO , Au(1.0)/SnO , and Cu(0.8)-
Au(0.2)/SnO
formic acid was almost completely decomposed to CO2
2
2
(
HCOOH+0.5O !CO +H O) over Au(1.0)/SnO2, Au(0.2)-
2
2
2
2
Ag(0.8)/SnO , and Au(0.2)-Cu(0.8)/SnO under irradiation with
2
2
green, blue, and red light, respectively. The amounts of CO2
evolved were about fifteen times larger than the total amounts
Previously, we found almost quantitative oxidation of benzyl
alcohols to benzaldehydes in an aqueous suspension of Au/
[
8]
(
2.5, 4.2, and 6.8 mmol, respectively) of Au, Ag-Au, and Cu-Au
CeO under irradiation with green light from an LED. Herein,
2
loaded on SnO , that is, mineralization of formic acid was not
Ag(0.8)-Au(0.2)/SnO , Au(1.0)/SnO , and Cu(0.8)-Au(0.2)/SnO
2
2
2
2
a quantitative reaction but a photocatalytic reaction.
were used for oxidation of benzyl alcohol under irradiation
with light from three LEDs to evaluate their performance in
photocatalytic conversions other than mineralization. No oxida-
tion of benzyl alcohol occurred over metal-free SnO ; thus, visi-
2
Action spectra
ble light from the LEDs did not cause bandgap excitation of
SnO . On the other hand, Ag(0.8)-Au(0.2)/SnO , Au(1.0)/SnO ,
The action spectrum is a powerful tool for determining wheth-
er an observed reaction occurs by a photoinduced or a thermo-
catalytic process. To obtain action spectra in this reaction
system, mineralization of formic acid in aqueous suspensions
of Au(1.0)/SnO , Ag(0.8)-Au(0.2)/SnO , and Cu(0.8)-Au(0.2)/SnO
2
2
2
and Cu(0.8)-Au(0.2)/SnO were active in the oxidation of benzyl
2
alcohol and yielded benzaldehyde with quite high selectivity
(>99%) at greater than 99% conversion of benzyl alcohol
after 20 h when blue, green, and red LEDs were used, respec-
tively (Supporting Information, Table S1). We confirmed that
2
2
2
was carried out at 298 K under irradiation with monochromat-
ed visible light from an Xe lamp with a light width of Æ5 nm.
The apparent quantum efficiency (AQE) at each centered wave-
length of light was calculated from the ratio of the amount of
a photocatalyst, irradiation with visible light, and O were indis-
2
pensable for oxidation of benzyl alcohol. Since the amount of
benzaldehyde increased linearly with photoirradiation time of
the three materials, the formation rates were determined from
slopes of the time courses of benzaldehyde formation
CO and the amount of incident photons (see Supporting In-
2
formation) by using Equation (1).
(
Figure 8). The highest reaction rates were obtained when the
amount of CO2
amount of incident photons
irradiating light overlapped well with photoabsorption due to
SPR, as in the case of formic acid mineralization.
AQE ¼
 100
ð1Þ
The use of Ag-Au/SnO , Au/SnO and Cu-Au/SnO for selec-
2
2
2
The results are shown in Figure 7. Wavelength dependences
of AQEs over Au(1.0)/SnO , Ag(0.8)-Au(0.2)/SnO , and Cu(0.8)-
tive oxidation under visible-light irradiation was extended to
various water-soluble alcohols as substrates. Ag(0.8)-Au(0.2)/
SnO , Au(1.0)/SnO and Cu(0.8)-Au(0.2)/SnO were also effective
2
2
Au(0.2)/SnO were similar to those of their photoabsorption
2
2
2
2
(
one-reflection), and this indicates that formation of CO from
in oxidation of a secondary aliphatic alcohol (2-pentanol), an
alicyclic alcohol (cyclohexanol), a secondary aromatic alcohol
(1-phenylethanol), and a primary aromatic alcohol (2-phenyl-
ethanol) to the corresponding aldehyde or ketone with selec-
tivities higher than 99%, which indicates that these photocata-
2
formic acid in aqueous suspensions of the three materials was
induced by photoabsorption due to their SPR. AQEs exceeding
5
% were obtained for all three materials when they were irra-
diated with light matching their SPR.
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4595
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 9. Time course of evolution of CO
2
from formic acid in an aqueous
solution in the simultaneous presence of Ag(0.8)-Au(0.2)/SnO
and Cu(0.8)-Au(0.2)/SnO under irradiation with visible light from green,
blue, and red LEDs (1.7 mWcm ).
2 2
, Au(1.0)/SnO ,
2
À2
Figure 8. Rates of formation of benzaldehyde from benzyl alcohol in aque-
ous suspensions of SnO , Ag(0.8)-Au(0.2)/SnO , Au(1.0)/SnO , and Cu(0.8)-
2
2
2
Linear evolution of CO2 with photoirradiation was ob-
served, and the rates under irradiation with green, blue,
Au(0.2)/SnO
2
under irradiation with visible light from green, blue, and red
À2
LEDs (1.7 mWcm ).
and red lights were determined to be 1.9, 1.3, and
À1
1
.2 mmolh , respectively. Figure 9 also shows that no gas
Table 1. Oxidation of various alcohols in aqueous suspensions of Ag(0.8)-
evolved in the dark, that is, no thermocatalytic CO forma-
2
Au(0.2)/SnO
2 2 2
(Ag-Au), Au(1.0)/SnO (Au), and Cu(0.8)-Au(0.2)/SnO (Cu-Au)
tion occurred under the present conditions. Figure 10
shows an action spectrum for mineralization of formic acid
in an aqueous solution containing Au(0.2)-Ag(0.8)/SnO2,
Au(1.0)/SnO , and Au(0.2)-Cu(0.8)/SnO . A wide action spec-
under irradiation with visible light from blue, green, and red LEDs in the pres-
[a]
2
ence of O .
[
b]
[b]
Photocatalyst Substrate
Product
LED t [h] Conv. [%] Sel. [%]
2
2
Ag-Au
Au
Cu-Au
Ag-Au
Au
Cu-Au
Ag-Au
Au
blue 32 32
green 32 38
>99
>99
>99
>99
>99
>99
>99
>99
>99
trum was obtained in the presence of all three materials,
and the AQE reached 3.3% at 450 nm, 3.8% at 550 nm, and
red
32 42
3
.2% at 650 nm; thus, a broad response to visible light was
blue 30 42
green 30 51
achieved by simultaneous use of the three materials.
red
30 38
blue 48 32
green 48 34
Cu-Au
red
48 29
Ag-Au
Au
Cu-Au
blue 68 19
green 68 25
>99
>99
>99
Photocatalytic selective oxidation of various alcohols in
aqueous solution in the simultaneous presence of Au-Ag/
SnO , Au/SnO , and Au-Cu/SnO
red
68 18
2
2
2
3
[
a] Photocatalyst: 50 mg, substrate: 50 mmol, water: 5 cm , O
2
: 1 atm, light in-
À2
Samples of all three materials (20 mg each) were simultane-
ously used for selective oxidation of benzyl alcohol to ben-
zaldehyde in an aqueous solution under irradiation with
visible light from an Xe lamp equipped with L-42, Y-50, and
R-62 cut filters (for the resulting spectra, see Figure 11 a).
The rates of benzaldehyde formation under irradiation with
visible light passing through L-42, Y-50, and R-62 cut filters
tensity: 1.7 mWcm . [b] Determined by GC with an internal standard. Conv.:
conversion, Sel.: selectivity.
lysts can be used for selective oxidation of various alcohols
under visible-light irradiation (See Table 1).
À1
were 3.6, 2.6, and 1.2 mmolh , respectively (Figure 11 b). The
order L-42>Y-50>R-62 corresponds to the amount of pho-
tons provided to the reaction system, and this indicates that si-
multaneous use of three materials is effective for efficient uti-
lization of solar light in selective oxidation of benzyl alcohol to
benzaldehyde.
Photocatalytic mineralization of formic acid in aqueous solu-
tions in the simultaneous presence of Au-Ag/SnO , Au/SnO ,
and Au-Cu/SnO2
2
2
Since Au-Ag/SnO , Au/SnO , and Au-Cu/SnO worked most ef-
2
2
2
fectively under irradiation with blue, green, and red light, re-
spectively, we had the idea of simultaneously using these ma-
terials to effectively utilize solar light. Figure 9 shows time
Applicability of the simultaneous use of three materials for
photocatalytic oxidation of alcohols to the corresponding car-
bonyl compounds was investigated under irradiation with visi-
ble light passing through L-42, Y-50, and R-62 cut filters
(Table 2). Carbonyl compounds were obtained with high selec-
tivity (>99%) for all alcohols. A similar effect of cut filters on
yields was observed in selective oxidation of these alcohols.
courses of the evolution of CO from formic acid in aqueous
2
solutions containing Au(0.2)-Ag(0.8)/SnO , Au(1.0)/SnO , and
2
2
Au(0.2)-Cu(0.8)/SnO (20 mg each) under irradiation with light
2
from green, blue, and red LEDs.
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4596
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Oxidation of various alcohols in aqueous solutions in the simul-
taneous presence of Ag(0.8)-Au(0.2)/SnO
Au(0.2)/SnO under irradiation with visible lights from an Xe lamp with L-
2, Y-50, and R-62 cut filters in the presence of O
2 2
, Au(1.0)/SnO , and Cu(0.8)-
2
[a]
4
2
.
[b]
[b]
Substrate
Product
Filter
t [h]
Conv. [%]
Sel. [%]
L-42
Y-50
R-62
L-42
Y-50
R-62
L-42
Y-50
R-62
65
41
25
76
37
21
55
28
12
>99
>99
>99
>99
>99
>99
>99
>99
>99
5
5
5
5
L-42
Y-50
R-62
48
23
9
>99
>99
>99
Figure 10. Absorption spectra of Ag(0.8)-Au(0.2)/SnO
2 2
, Au(1.0)/SnO , and
3
[
a] Photocatalyst: 50 mg, substrate: 50 mmol, water: 5 cm , O
2
: 1 atm, irra-
Cu(0.8)-Au(0.2)/SnO measured with barium sulfate as a reference (left axis)
2
diation time: 5 h. [b] Determined by GC with an internal standard. Conv.:
conversion, Sel.: selectivity.
and action spectrum in formic acid mineralization in an aqueous solution in
the simultaneous presence of all three materials (right axis).
Figure 12. Proposed reaction mechanism for oxidation reactions over plas-
monic metal/SnO materials under irradiation with visible light.
2
the semiconductor film takes place under visible-light irradia-
[
11]
tion. The following reaction mechanism is proposed: 1) inci-
dent photons are absorbed by metal nanoparticles through
SPR excitation, 2) electrons are injected from metal nanoparti-
cles into the conduction band of SnO , 3) the resultant elec-
2
tron-deficient metal nanoparticles could oxidize formic acid or
alcohols to be returned to the original metallic state, and
4
) electrons in the conduction band are consumed for reduc-
tion of oxygen. Electrons in the conduction band of SnO are
2
trapped by O , and H O (two-electron reduction step) and/or
2
2
2
H O (four-electron reduction step) is formed. Very high selectiv-
2
ity for carbonyl compounds in the present system suggests
that active oxygen species were removed from the reaction
system.
Figure 11. a) Spectra of visible light from an Xe lamp with L-42, Y-50, and R-
6
2 cut filters. b) Rates of formation of benzaldehyde from benzyl alcohol in
an aqueous solution in the simultaneous presence of Ag(0.8)-Au(0.2)/SnO
Ag-Au), Au(1.0)/SnO , and Cu(0.8)-Au(0.2)/SnO under irradiation with visible
2
(
2
2
light from a Xe lamp with L-42, Y-50, and R-62 cut filters.
Conclusions
SnO -supported Au particles exhibiting strong photoabsorp-
2
Proposed working mechanism
tion due to SPR of Au were prepared by MSPD. Modification of
The proposed working mechanism for oxidation over plasmon-
ic photocatalysts under irradiation with visible light is shown
in Figure 12. Rapid electron transfer from SPR nanoparticles to
0.2 wt%Au/SnO with Cu and Ag by MSPD resulted in forma-
2
tion of Cu-Au/SnO and Ag-Au/SnO . The absorption of the
2
2
Au/SnO around l=550 nm gradually shifted to longer and
2
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4597
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
shorter wavelengths with increasing contents of Cu and Ag, re-
spectively. Cu-Au/SnO and Ag-Au/SnO containing 0.8 wt% Cu
tance measurement unit (ISR-2000, Shimadzu). The morphology of
photocatalysts was observed under a JEOL JEM-3010 transmission
electron microscope operated at 300 kV in the Joint Research
Center of Kindai University. LEDs (HDMS8R, Hayashi Watch Works,
Tokyo) and a 500 W Xe lamp (Ushio, Tokyo) were used as light
sources. Spectra and intensities of light from blue, green, and red
LEDs and monochromated light from the Xe lamp were deter-
mined with a USR-45D spectroradiometer (Ushio, Tokyo).
2
2
and Ag showed strong absorption at around l=450 and
20 nm, respectively. These samples were active for mineraliza-
6
tion of formic acid and selective oxidation of alcohols to car-
bonyl compounds under visible-light irradiation and exhibited
the best performance when irradiated by light overlapping
with their SPR. Wide responses to visible light in mineralization
and selective oxidation were achieved by simultaneous use of
Cu-Au/SnO , Au/SnO , and Ag-Au/SnO .
Mineralization of formic acid in aqueous suspensions of Au/
SnO , Ag-Au/SnO , and Cu-Au/SnO under irradiation with
2
2
2
2
2
2
visible light
Mineralization of formic acid to CO over a photocatalyst under ir-
Experimental Section
2
radiation with visible light was chosen as the test reaction
[4c]
Preparation of 0.2 wt% Au/SnO2
[Eq. (2)].
Commercial SnO (Nanotech, Kanto Chemical) powder was calcined
at 1273 K for 1 h in a box furnace. A TEM image of the calcined
2
HCOOH þ 0:5 O ! CO þ H O
ð2Þ
2
2
2
SnO powder is shown in Figure S1 of the Supporting Information.
2
The dried Au/SnO , Ag-Au/SnO , or Cu-Au/SnO powder (50 mg)
2
2
2
Gold was loaded onto SnO by PD. The SnO powder (198 mg) was
2
2
3
3
was suspended in distilled water (5 cm ), and the suspension bub-
bled with O2 and sealed with a rubber septum. Formic acid
(100 mmol) was injected into the suspension, which then irradiated
with visible light from a blue, green, or red LED (HDMS8B,
HDMS8G, and HDMS8R, respectively, Hayashi Watch Works, Tokyo)
with magnetic stirring in a water bath continuously kept at 298 K.
The spectra of the visible light irradiating the reaction system is
shown in Figure 5, and the maximum wavelength of the light was
determined to be 475, 530, and 640 nm, respectively. The amount
of CO2 in the gas phase of the reaction mixture was measured
with a gas chromatograph (GC-8A, Shimadzu) equipped with Pora-
pak QS columns. To obtain an action spectrum, the full arc from
a Xe lamp was monochromated with an SM-100 monochromator
suspended in 10 cm of an aqueous solution of methanol
50 vol%) in a test tube (height: 180 mm, inner diameter: 18 mm,
(
3
total volume: 35 cm ), and the test tube was sealed with a rubber
septum under argon. An aqueous solution of tetrachloroauric acid
À3
(
corresponding to 0.2 wt% Au as metal, 4.0 gdm ) was injected
into the sealed test tube and then photoirradiated at l>300 nm
by a 400 W high-pressure mercury arc lamp (Eiko-sha, Osaka) with
magnetic stirring in a water bath continuously kept at 298 K. The
Au source was reduced by photogenerated electrons, and Au
metal was deposited on the SnO2 particles to form 0.2 wt%Au/
[8,9]
SnO . The 1.0 wt% Au/SnO was prepared by MSPD,
whereby
2
2
PD of 0.2 wt% Au was repeated five times (1.0 wt%=0.2 wt%”5).
These samples were designated Au(0.2)/SnO and Au(1.0)/SnO , re-
2
2
(
Bunkoukeiki, Tokyo).
spectively. Analysis of the liquid phase after PD revealed that the
Au source had been almost completely (>99.9%) deposited as Au
metal on the SnO particles. The resultant powder was washed re-
peatedly with distilled water and then dried in air at 310 K over-
night.
Oxidation of alcohols in aqueous suspensions of Au/SnO2,
Ag-Au/SnO , and Cu-Au/SnO under irradiation with visible
light
2
2
2
Oxidation of alcohols to carbonyl compounds (aldehydes or ke-
tones) over photocatalysts under irradiation with visible light was
Preparation of Ag-Au/SnO and Cu-Au/SnO2
2
[8]
chosen as test reaction. The dried Au/SnO , Ag-Au/SnO , or Cu-
MSPD was also used for preparation of Ag-Au/SnO2 and Cu-Au/
SnO by modification of Au(0.2)/SnO with Ag and Cu. An aqueous
2
2
3
Au/SnO powder (50 mg) was suspended in distilled water (5 cm ),
2
2
2
À3
and the suspension bubbled with O2 and sealed with a rubber
septum. Alcohols were injected into the suspensions, which were
then irradiated with visible light from blue, green, and red LEDs.
The amounts of alcohols and carbonyl compounds in the liquid
phase were determined with a Shimadzu GC-14B gas chromato-
graph equipped with a DB-1 capillary column (30 m, 0.25 mm). The
solution of silver sulfate (Ag: 4.0 gdm ) or copper sulfate (Cu:
.0 gdm ) was injected into an aqueous methanolic suspension of
À3
4
Au(0.2)/SnO and the mixture was photoirradiated with a mercury-
2
arc lamp under the same conditions as for the preparation of
Au(0.2)/SnO . The amount of Ag and Cu loaded per PD was fixed
2
at 0.2 wt%, and PD of Ag and Cu was repeated for additional Ag
and Cu loadings of Ag-Au(0.2)/SnO and Cu-Au(0.2)/SnO . For ex-
3
reaction solution (1 cm ) was added to diethyl ether/water (2/1 v/v,
2
2
3
3
1
cm ). Toluene was used as internal standard (injection amount:
ample, Ag PD was repeated four times for the preparation of
.8 wt%Ag-Au(0.2)/SnO . Hereafter, this material is designated
00 mL, toluene/2-propanol, 50 mL/5 mL). After the mixture had
0
2
been stirred for 10 min, alcohols and carbonyl compounds in the
ethereal phase were analyzed. The amounts of alcohols and car-
bonyl compounds were determined from the ratios of the peak
areas of alcohol and carbonyl compound to the peak area of tolu-
ene (see Supporting Information for details).
Ag(0.8)-Au(0.2)/SnO , and that modified with 0.8 wt%Cu as Cu(0.8)-
2
Au(0.2)/SnO . Analysis of the liquid phase after PD revealed that
2
the Ag and Cu sources had been almost completely (>99.9%) de-
posited as Ag and Cu metals on Au(0.2)/SnO . The resultant
2
powder was washed repeatedly with distilled water and then dried
in air at 310 K overnight.
Acknowledgements
Characterization
Diffuse-reflectance spectra were obtained with a UV/Vvis spectrom-
eter (UV-2400, Shimadzu, Kyoto) equipped with a diffuse-reflec-
This work was partly supported by JSPS KAKENHI Grant Num-
bers 26289307, 26630415, and 15K18269. This work was also
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4598
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
supported by MEXT-Supported Program for the Strategic Re-
search Foundation at Private Universities 2014–2018, subsidy
from MEXT and Kindai University. H.K. is grateful for financial
support from the Faculty of Science and Engineering, Kindai
University.
2010, 12, 2344–2355; c) H. Kominami, A. Tanaka, K. Hashimoto, Chem.
Commun. 2010, 46, 1287–1289; d) X. Zhou, G. Liu, J. Yu, W. Fan, J.
Mater. Chem. 2012, 22, 21337–21354; e) S. Linic, P. Christopher, D. B.
Ingram, Nat. Mater. 2011, 10, 911 –921; f) C. Wang, D. Astruc, Chem. Soc.
Rev. 2014, 43, 7188–7216; g) Q. Xiao, E. Jaatinen, H. Zhu, Chem. Asian J.
2014, 9, 3046–3064.
[
5] A. Tanaka, K. Hashimoto, H. Kominami, ChemCatChem 2011, 3, 1619–
1
623.
[6] Z. Sun, M. Liang, J. Chen, Acc. Chem. Res. 2015, 48, 1541–1550.
Keywords: gold · nanoparticles · photochemistry · supported
catalysts · surface plasmon resonance
[
[
[
7] P. Docampo, P. Tiwana, N. Sakai, H. Miura, L. Herz, T. Murakami, H. J.
Snaith, J. Phys. Chem. C 2012, 116, 22840–22846.
8] A. Tanaka, K. Hashimoto, H. Kominami, J. Am. Chem. Soc. 2012, 134,
[
[
1] a) R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 2001, 293,
69–271; b) H. Kisch, L. Zang, C. Lange, W. F. Maier, C. Antonius, D.
Meissner, Angew. Chem. Int. Ed. 1998, 37, 3034–3036; Angew. Chem.
998, 110, 3201–3203; c) T. Ohno, M. Akiyoshi, T. Umebayashi, K. Asai, T.
Mitsui, M. Matsumura, Appl. Catal. A 2004, 265, 115–121; d) S. Kitano, K.
Hashimoto, H. Kominami, Appl. Catal. B 2011, 101, 206–211.
14526–14533.
2
9] a) A. Tanaka, S. Sakaguchi, K. Hashimoto, H. Kominami, Catal. Sci. Tech-
nol. 2012, 2, 907–909; b) A. Tanaka, S. Sakaguchi, K. Hashimoto, H. Ko-
minami, Catal. Sci. Technol. 2014, 4, 1931–1938.
1
[
[
10] a) I. Pastoriza-Santos, A. Sµnchez-Iglesias, B. Rodríguez-Gonzµlez, Liz-
L. M. Marzµn, Small 2009, 5, 440–443; b) P. Kanninen, C. Johans, J.
Merta, K. Kontturi, J. Colloid Interface Sci. 2008, 318, 88–95.
2] a) R. Abe, H. Takami, N. Murakami, B. Ohtani, J. Am. Chem. Soc. 2008,
1
30, 7780–7781; b) T. Arai, M. Horiguchi, M. Yanagida, T. Gunji, H. Sugi-
11] A. Furube, L. Du, K. Hara, R. Katoh, M. Tachiya, J. Am. Chem. Soc. 2007,
hara, K. Sayama, Chem. Commun. 2008, 5565–5567; c) A. Tanaka, K. Ha-
shimoto, H. Kominami, Chem. Lett. 2011, 40, 354–356; d) Y. Sasaki, A.
Tanaka, K. Hashimoto, H. Kominami, Chem. Lett. 2013, 42, 419–421.
3] M. A. Garcia, J. Phys. D 2011, 44, 283001–283020.
4] a) E. Kowalska, R. Abe, B. Ohtani, Chem. Commun. 2009, 241–243; b) E.
Kowalska, O. O. P. Mahaney, R. Abe, B. Ohtani, Phys. Chem. Chem. Phys.
1
29, 14852–14853.
[
[
Received: November 16, 2016
Published online on February 16, 2016
Chem. Eur. J. 2016, 22, 4592 – 4599
www.chemeurj.org
4599
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim