Selective reductions using visible light photocatalysts of supported gold nanoparticles

Article abstract of DOI:10.1039/c2gc36542a

Photocatalytic synthesis using visible light is a desirable chemical process because of its potential to utilize sunlight. Supported gold nanoparticles (Au-NPs) were found to be efficient photocatalysts and the effects of the supports were identified including CeO₂, TiO₂, ZrO₂, Al₂O₃, and zeolite Y. In particular Au/CeO₂ exhibited the high catalytic activity to reduce nitro-aromatics to azo compounds, hydrogenate azobenzene to hydroazobenzene, reduce ketones to alcohols, and deoxygenate epoxides to alkenes at ambient temperatures, under irradiation of visible light (or simulated sunlight). The reactive efficiency depends on two primary factors: one is the light adsorption of catalysts and the other is the driving ability of catalysts corresponding to the reactants. The light absorption by Au-NPs is due to the surface plasmon resonance effect or interband electron transition; this is related to the reduction ability of the photocatalysts. Irradiation with shorter wavelengths can excite the conduction electrons in Au-NPs to higher energy levels and, as a result, induce reduction with more negative reduction potentials. It is known that when irradiated with light the Au-NPs can abstract hydrogen from isopropanol forming Au-H species on the Au-NP surface. Hence, we proposed that the active Au-H species will react with the NO, NN, CO double bonds or epoxide bonds, which are weakened by the interaction with the excited electrons in the Au-NPs, and yield the final reductive products. The reacting power of the Au-H species depends on the energy of the excited electrons in Au-NPs: the higher the electronic energy, the stronger the reduction ability of the Au-H species. This finding demonstrates that we can tune the reduction ability of the photocatalysts by manipulating the irradiation wavelength.

Full text of DOI:10.1039/c2gc36542a

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Selective reductions using visible light photocatalysts
of supported gold nanoparticles†
Cite this: Green Chem., 2013, 15, 236
Xuebin Ke,* Xingguang Zhang, Jian Zhao, Sarina Sarina, John Barry and
Huaiyong Zhu*
Photocatalytic synthesis using visible light is a desirable chemical process because of its potential to
utilize sunlight. Supported gold nanoparticles (Au-NPs) were found to be efficient photocatalysts and
2 2 2 2 3
the effects of the supports were identified including CeO , TiO , ZrO , Al O , and zeolite Y. In particular
Au/CeO exhibited the high catalytic activity to reduce nitro-aromatics to azo compounds, hydrogenate
2
azobenzene to hydroazobenzene, reduce ketones to alcohols, and deoxygenate epoxides to alkenes at
ambient temperatures, under irradiation of visible light (or simulated sunlight). The reactive efficiency
depends on two primary factors: one is the light adsorption of catalysts and the other is the driving
ability of catalysts corresponding to the reactants. The light absorption by Au-NPs is due to the surface
plasmon resonance effect or interband electron transition; this is related to the reduction ability of the
photocatalysts. Irradiation with shorter wavelengths can excite the conduction electrons in Au-NPs to
higher energy levels and, as a result, induce reduction with more negative reduction potentials. It is
known that when irradiated with light the Au-NPs can abstract hydrogen from isopropanol forming Au–
H species on the Au-NP surface. Hence, we proposed that the active Au–H species will react with the
NvO, NvN, CvO double bonds or epoxide bonds, which are weakened by the interaction with the
excited electrons in the Au-NPs, and yield the final reductive products. The reacting power of the Au–H
species depends on the energy of the excited electrons in Au-NPs: the higher the electronic energy, the
stronger the reduction ability of the Au–H species. This finding demonstrates that we can tune the
reduction ability of the photocatalysts by manipulating the irradiation wavelength.
Received 30th September 2012,
Accepted 13th November 2012
DOI: 10.1039/c2gc36542a
www.rsc.org/greenchem
2
chemicals only achieved limited progress. Titania (TiO ) and
2
Introduction
silica-supported titania are active for photocatalytic oxidation
The photocatalytic process has been extensively applied to of light alkanes and alkenes under ultraviolet (UV) irradiation
2
degrade environmental pollutants, develop new solar cells, at ambient temperatures. UV light accounts for only about 4%
1
and conduct homogeneous reactions of photochemistry. The of the solar energy received by the earth while visible light
2
d
process has salient advantages: it is generally conducted at (wavelength 400–700 nm) accounts for 43%. The efficiency of
ambient temperatures with the potential to be driven by sun- the photocatalytic synthesis becomes a primary concern.
light—an inexhaustible and clean energy source. If the photo- Various methods have been developed to produce visible light
catalytic process is applied to synthesize fine organic photocatalysts. For example, some research groups reported
chemicals, it can bring about significant benefits such as the synthesis of a series of aldehydes from the corresponding
achieving better selectivity to the desired product via a green alcohols by photocatalytic oxidation with dye-sensitized TiO
2
3
approach. Moreover, the substances that are unstable at high or molecular oxygen on TiO under visible light. The dye-sen-
2
temperatures may be synthesized through this unique light- sitized photocatalysts can absorb visible light and this process
excited progress. However, the application of the sunlight- achieved high conversion and product selectivity. Nonetheless,
driven photocatalytic process to synthesize fine organic this approach is not applicable to catalyze reduction reactions.
Given that reduction is a major process for the synthesis of
organic fine chemicals, it will be of great interest if photocata-
School of Chemistry, Physics and Mechanical Engineering, Faculty of Science and
Engineering, Queensland University of Technology, Brisbane QLD4001, Australia.
E-mail: x.ke@qut.edu.au, hy.zhu@qut.edu.au; Fax: +61 7 31381804;
Tel: +61 7 31389197, +61 7 31381581
lytic reduction processes can be applied to synthesize fine
chemicals under visible light. However, the synthesis of
organic compounds via visible light photocatalytic reduction is
even more challenging than that via visible-light photocatalytic
oxidation. Currently there are a few reports on this topic,
†
Electronic supplementary information (ESI)
0.1039/c2gc36542a
available. See DOI:
1
2
36 | Green Chem., 2013, 15, 236–244
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
which focus on the reduction of nitro-aromatics, such as the as dyes in the polymer, textile and food industries as well as in
photoreduction of nitrobenzene, using photocatalysts of the production of medicines to treat inflammation and a type
1
0
2
cadmium sulphide or modified TiO , and the main products of arthritis. The supported Au-NPs catalysts exhibited excel-
4
are aniline or imines. The core issue regarding photocatalytic lent activity for these photocatalytic syntheses. Moreover, the
reduction is to devise efficient photocatalysts that can catalyze reduction potentials of the four reductions vary from −0.8 V to
1
,11
the reduction reactions with superior product selectivity under −2.4 V.
Hence, it is possible to examine the impact of the
visible light irradiation. If we can find photocatalysts that are irradiation wavelength on the ability of the supported Au-NPs
able to absorb visible light and activate chemical reactions, catalysts to drive the reductions with different reduction poten-
photocatalytic processes driven by sunlight can be developed.
tial. Indeed, we found that the irradiation of shorter wave-
Since the 1980s Au-NPs have been recognized as a fascinat- length can drive the reduction of more negative reduction
ing material that shows great potential in sustainable chemical potential. Since the irradiation wavelength is related to the
5
synthesis. In the photocatalytic process on gold science, we energy of the electrons excited by the light, this finding means
found that the supported Au-NPs exhibited superior perform- that the higher the energy of the excited electrons in the
ance in the photoredox catalysis of fine organic compounds photocatalysts, the stronger the reduction ability of the cata-
6
under both visible and UV irradiation. It highlights that sus- lyst. Thus a tentative mechanism of the photochemical
tainable sunlight resources can be used to drive some chemi- reduction can be proposed.
cal process with high efficiency. Au-NPs absorb visible light
7
mainly due to the surface plasmon resonance (SPR) effect.
The SPR effect is the collective oscillation of conduction elec- Results and discussion
trons in the NPs, which resonate with the electromagnetic
1
. Effects of Au-NPs on different supports
field of the incident light, and this resonance gives rise to sig-
nificant field enhancement of the local electromagnetic fields
near surfaces of the Au-NPs. The SPR absorption for typical
spherical Au-NPs is generally observed between 520 and
The size and dispersion of the Au-NPs should influence the
photocatalytic performance, which can be readily tuned by
adjusting the ratios of sodium borohydride to HAuCl and
HAuCl to supports. Fig. 1 shows the typical transmission
electron microscopy (TEM) images of the Au-NPs on the
various supports including CeO , TiO , ZrO , Al O , and
2 2 2 2 3
zeolite Y (namely Au/CeO , Au/TiO , Au/ZrO , Au/Al O , and
Au/Y). The Au-NPs are well dispersed and have narrow size dis-
tributions. The sizes of the Au-NPs on CeO , TiO , and ZrO
2
are estimated to be in the range of 5 nm to 8 nm while those
on Al O are smaller (about 3 nm) and those on zeolite Y are
bigger (about 10 nm). The X-ray photoelectron spectroscopy
(XPS) spectra of the photocatalysts are shown in Fig. 2, Au 4f
spectra of all Au-NPs photocatalysts exhibit similar patterns.
The binding energies of Au 4f7/2 and Au 4f5/2 electrons are 84.0
and 87.7 eV, respectively. It is consistent with the reports on
gold metal, which indicates that Au-NPs on the supports exist
4
1
2
6
,7
4
5
50 nm, and the conducting electrons at the NP surfaces
gain energy from the absorption. The energetic electrons may
strongly interact with the molecules on the Au-NP surface and
2
2
2
2 3
6
,7b
even migrate to the molecules absorbed on the support
and
will relax back to their thermal equilibrium states and release
heat to the Au-NP lattice, resulting in rapid heating of the Au-
2
2
7
c
NPs. Together, the two effects can induce chemical reactions
of molecules on the Au-NPs. The mechanism of the photocata-
lysis with Au-NPs is distinctly different from that of semicon-
2
3
2
,6
ductor photocatalysts.
This poses an opportunity and
question that it may be possible to use the supported Au-NPs
to catalyze reduction reactions under visible and UV light. If
successful, we can drive important reactions using sunlight,
gain insight into the interaction between light, Au-NPs and
reactant molecules and understand the mechanism of photo-
catalytic reductions, which are valuable for designing new
photocatalytic processes.
1
2
in the metallic state.
In the present study, gold nanoparticles (Au-NPs) supported
on CeO , TiO , ZrO , Al O , and zeolite Y were prepared and
2 2 2 2 3
the effects of the supports were identified. In particular we
investigated four typical reduction processes on supported Au-
NP photocatalysts under visible light (or simulated sunlight):
reduction of nitro-aromatics to azo compounds, hydrogenation
of azobenzene to hydroazobenzene, reduction of ketones to
alcohols, and deoxygenation of epoxides to alkenes. These
reduction reactions are important in organic synthesis and
8
–10
biological chemistry.
For example, in the vitamin K cycle
the recycling of the oxidized vitamin K back to the reduced
8
form is a key step currently relying on enzyme catalysis, which
involves deoxygenation of epoxides. The reduction of ketones
to alcohols is a significant transformation in organic synthesis
and the chemical industry. Azo compounds are widely used bars in the images is 10 nm.
Fig. 1 TEM images of Au-NPs on different supports. The length of the scale
9
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 236–244 | 237

View Article Online
Paper
Green Chemistry
1
3
UV/Vis spectra of Au-NPs on different supports as well as Au/TiO
supports themselves were collected and shown in Fig. 3. An though Au-NPs on different supports have similar size (Fig. 1),
absorption band at about 520 nm is observed in the UV/Vis the strong light absorption of Au/CeO may be attributed to
spectra of the photocatalysts, which is the characteristic SPR the strong interface action between Au-NPs and CeO The
2
, indicating a stronger SPR effect of Au/CeO
2
.
Even
2
1
3
2
.
absorption by Au-NPs. The light absorption of the Au-NPs in light absorption is important to the photocatalytic perform-
the photocatalysts varies with the change of the supports, such ance, but the catalysis activity is also dependent on other
as CeO
exhibit little absorption of light with wavelengths longer than and supports as the photocatalytic reactions take place on the
00 nm (Fig. 3). From the UV/Vis spectra, the TiO support can Au-NP surface with the aid of supports. Some researchers
2 2 2 2 3
, TiO , ZrO , Al O , and zeolite Y, while the supports factors, particularly the interface action between the Au-NPs
4
2
absorb most light with wavelengths less than 400 nm, and propose electron transitions between Au-NPs and the support
CeO can absorb light with wavelengths less than 500 nm. On (TiO ), which is not perfect to clarify the functions of supports,
2
2
the other hand, the Au/CeO
2
2
catalyst shows a little shift to the particularly in the photocatalytic process. Au/CeO has a stron-
long-wavelength direction, and has higher intensity relative to ger interaction between Au-NPs and CeO
2
and has a stable
structure with Au-NPs incorporated into CeO , which has been
2
1
4
identified previously. Thus in a reactive system the aggrega-
tion of Au-NPs can be avoided and intensive light absorption
is maintained. Au-NPs were also loaded on TiO and ZrO and
2
2
we found that for reduction of styrene oxide the selectivity to
the product styrene is moderately lower than that of the Au/
CeO (Table 1). Nevertheless, all the supports themselves show
2
trivial activities in these photocatalytic reductions. It appears
that the support materials have a limited impact on the con-
version of photocatalytic reductions. In view of the strong
linkage between Au-NPs and CeO , the Au-NPs on the CeO
2 2
support were selected to be an efficient and stable
photocatalyst.
Fig. 2 X-ray photoelectron spectra (XPS) of Au-NPs on different supports.
2. Effects of gold loads for the reduction of styrene oxide
The photocatalytic efficiency of supported Au-NPs depends on
two primary aspects: the light absorption of catalysts to the
emission light and the driving capability of catalysts with
regard to the reactants. In a certain reactive system, for
instance, the reduction of styrene oxide, light absorption of
catalysts plays a significant role in the enhancement of reactive
activity due to the contribution of light absorption of Au-NPs.
Fig. 4A exhibits the UV/Vis diffuse reflectance spectra of Au/
2
CeO with various amounts of gold loads. It appears that the
light absorption of Au/CeO at about 520 nm increases with
2
the increasing loadings of Au-NPs. The results of the photoca-
talytic reduction of styrene oxide under visible light irradiation
are shown in Fig. 4B. High conversions are achieved for this
reaction and the photocatalytic efficiency is significantly
enhanced with the increase of gold loading. Meanwhile, the
Fig. 3 UV-Visible (UV/Vis) diffuse reflectance spectra of Au-NPs on different corresponding conversions in the dark are found to maintain
supports: (a) Au/CeO ; (b) Au/TiO ; (c) Au/ZrO ; (d) Au/Al ; (e) Au/Y.
2
2
2
2
O
3
a minimal increase. For instance, when the gold loading
Table 1 Influences of supports on reduction of styrene oxide with supported Au-NPs
Visible light
No light
Catalysts
Reaction time (h)
Conversion (%)
Selectivity (%)
Conversion (%)
Selectivity (%)
Au/CeO
Au/ZrO
Au/TiO
2
16
16
16
16
16
19.7
17.4
22.5
23.9
4.5
88
72
73
60
62
4.4
7.8
7.8
5.8
1.6
72
75
75
51
55
2
2
2 3
Au/Al O
Au/Y
2
38 | Green Chem., 2013, 15, 236–244
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
increased from 1.5% to 3%, the conversion under visible light reactants were observed. This indicates that the reduction reac-
increased from 12% to 20% while the conversion in the dark tions can be driven by visible light irradiation and may follow
had negligible improvement. This indicates that the light different mechanisms from those of thermal processes and
absorption of catalysts is a crucial factor to enhance the photo- photocatalytic reductions using other catalysts under UV
4
,8–10
catalytic efficiency. To minimize the influence of the reduction irradiation.
reaction in the dark, 3% Au/CeO was selected to be an exem- example (entry 1), the SPR effect of Au-NPs plays a key role in
plar photocatalyst for the study of four types of reductions. assisting activation of the NvO bonds, which is distinct from
Take the reduction of nitrobenzene as an
2
6
a
the case under UV light irradiation. Thus, the product is
mainly azoxybenzene rather than aniline and the mechanism
of this process will be discussed later.
3
.
2
Performance of Au/CeO for the four photocatalytic
reductions
The results of the four photocatalytic processes on the photo-
4.
Effects of light wavelength
catalyst of Au/CeO under visible light irradiation are summar-
2
The wavelength of the irradiation has influence on the photo-
ized in Table 2. High conversion rates were achieved for these
reactions. Typically, 43.5% of nitrobenzene was reduced to azo-
benzene in 2 h with a selectivity of 96%. 40% of azobenzene
was reduced to hydroazobenzene in 6 h with a selectivity of
1
5
catalytic process. In photocatalytic experiments, glass filters
were employed to block the irradiation below a certain wave-
length (e.g. <550 nm), this wavelength is called the cut-off
wavelength. With the higher cut-off wavelength, i.e. the light
source focuses on longer wavelength, all the reactions declined
gradually even under other identical experimental conditions.
Therefore, it can be deduced that the SPR effect has a primary
influence on these photocatalytic reductions due to the strong
absorption around 520 nm (Fig. 3). When the light with wave-
length longer than 550 nm was used—by filtering out wave-
length shorter than 550 nm—the reduction of nitrobenzene
still proceeded even though the conversion (35%) decreased
7
8%. 31% of acetophenone transformed into benzyl ethanol
with a selectivity of 99%. 20% of styrene oxide was converted
to styrene in 16 h with a high selectivity of about 88%. These
reactions were also conducted in the dark with other exper-
imental conditions unchanged—very low conversions of the
(
Fig. 5a). This was similarly observed in the reduction of azo-
benzene of which the conversion decreased to 22% as well as
the hydrogenation of acetophenone of 9% (Fig. 5b and 5c).
However, the reduction of styrene oxide could not be activated
under such irradiation at the cut-off wavelength of 550 nm,
shown in Fig. 5d. Furthermore, when the cut-off wavelength of
the light source was around 600 nm, the reduction of azoben-
zene still proceeded (20% of conversion), but acetophenone
and styrene oxide could not be reduced.
Fig. 4 A. UV/Vis diffuse reflectance spectra of Au/CeO
sions of Au/CeO for the reduction of styrene oxide with various amounts of
gold loads: (a) 0 (CeO ); (b) 0.5%; (c) 1.5%; (d) 3%; (e) 5%.
2
. B. Catalytic conver-
We recognized a relationship between the cut-off wave-
length of the irradiation and the reduction potentials of the
2
2
2
Table 2 Performance of 3 wt% Au/CeO for the four photocatalytic reductions
Visible light
Conv. (%)
43.5
No light
Conv. (%)
8.1
Entry
1
Reagents
Products
Time (h)
Temp. (°C)
30
Sel. (%)
96
Sel. (%)
>99
2
2
6
30
40
78
10
79
3
4
24
16
30
25
31
20
>99
88
1
>99
72
4.4
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 236–244 | 239

View Article Online
Paper
Green Chemistry
Fig. 5 Influences of cut-off threshold wavelength on these demonstrated reac-
Fig. 6 Influences of light intensity on these demonstrated reactions: (a)
reduction of nitrobenzene; (b) hydrogenation of azobenzene; (c) reduction of
acetophenone; and (d) deoxygenation of styrene oxide.
tions: (a) reduction of nitrobenzene at 40 °C; (b) hydrogenation of azobenzene;
(c) reduction of acetophenone; and (d) deoxygenation of styrene oxide.
view of thermodynamics that reductions with more negative
reductive potentials could not proceed because the reactant
molecules could not gain the photo-excited electrons at lower
energy levels. This may imply that the reactant molecules in
the reduction with more negative reduction potentials can
only gain the excited electrons of Au-NPs at the higher energy
levels (Scheme 1), which are excited by light of shorter wave-
length. The wavelength of the light determines the reduction
power of the photocatalysts. This finding highlights important
opportunities that the conversion and product selectivity of
the reduction reactions can be optimized and the by-products
may be avoided by manipulating the irradiation wavelength. As
the SPR absorption of the Au-NPs is very weak in the range
longer than 650 nm (Fig. 3), the conversion of all the four reac-
tants is negligible. This confirms again that the SPR absorp-
tion of Au-NPs is crucial for the catalytic activity.
Scheme 1 Schematic band structure of supported Au-NPs with different
irradiation wavelengths and reduction potential of typical reactants: (a) nitro-
benzene; (b) azobenzene; (c) acetophenone; (d) styrene oxide.
5. Effects of light intensity
In an attempt to clarify the influence of light irradiation inten-
sity, the irradiation intensity was reduced from 0.40 to 0.30
and 0.15 W cm− , while other experimental conditions were
unchanged. We found that the conversion of nitrobenzene
decreased from 48% to 40% and then to 28%, correspond-
ingly. Similar changes were observed for the reduction of the
styrene oxide, azoaromatics and the ketones (Fig. 6). Such a
tendency reveals that these photocatalytic reductions are
dependent on the light intensity, as a higher intensity can
excite more energetic electrons in Au-NPs and enhance the
interaction with reactants absorbed on the Au-NPs under
visible light irradiation. These results further confirm that the
reduction reactions were driven by visible light.
reactants. The reduction potentials of the four reactants are
2
−
0.8 eV (nitrobenzene), −1.1 eV (azobenzene), −1.9 eV (aceto-
1
,11
phenone), and −2.4 eV (styrene oxide).
Correlating these
reduction potentials with the conversions under the
irradiation of different cut-off wavelengths, it can be seen that
a shorter cut-off wavelength is required to reduce a compound
with a more negative reduction potential. Conduction elec-
trons in the Au-NPs gain energy from the light absorption and
their distribution over the energy levels shifts to higher levels
6
,7
in the 6sp band. As shown in Scheme 1, the cut-off wave-
length determines the energy of the excited electrons as well as
6
their distribution over the high energy levels. The shorter the
wavelength is, the higher the energy of the excited electrons is
6.
Apparent activation energy of photocatalysis and thermal
reactions
benzene (reduction potential −0.8 eV) and azobenzene The contribution of visible light irradiation to reducing the
reduction potential −1.1 eV), but cannot induce the reduction apparent activation energies of the reductions is also investi-
of more negative reduction potentials. It is rational from the gated. These reduction reactions were conducted both under
(Scheme 1). The electrons excited by the light with wavelength
longer than 600 nm are able to induce the reduction of nitro-
(
2
40 | Green Chem., 2013, 15, 236–244
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
irradiation (photocatalytic process) and in the dark (thermal for the photocatalytic reduction and 30.7 kJ mol− for the
process) at various temperatures to examine the reduction kin- thermal reduction. The difference between the two activation
etics. According to the first-order reaction rate equation, the energies (12.9 kJ mol− ) reveals the contribution of light
values of the rate constants were calculated. Furthermore, the irradiation (Fig. 7b). It indicates that the light irradiation plays
Arrhenius equation was applied to derive the apparent acti- a major role in activating the reduction reaction.
vation energies of the reductions via photocatalytic and
1
1
6
d
thermal processes. For one reduction reaction, the difference
between the activation energies of the two processes indicates
the contribution of light irradiation to reducing the apparent 7. Photocatalytic reductions with simulated sunlight
activation energy. The results of four reactions are shown in
Photocatalysis with simulated sunlight was conducted under
the same experimental conditions but using a different light
Fig. 7. For azobenzene, the activation energy is 17.8 kJ mol−
1
source (model: xenon lamp, 150 W, Farrand Optical Co.). The
results with simulated sunlight are roughly the same corre-
sponding to those with the visible light, which further confirm
that light absorption and the reaction system are two primary
aspects of the photocatalytic efficiency. As light intensity and
the catalytic system are unchanged, the minimal deviations in
conversions may be attributed to the differences of wavelength
ranges in the light sources. These reduction reactions exhib-
ited the great potential to excite the photocatalytic reactions
utilizing sunlight (Table 3).
Table 3 Photocatalytic reduction reactions conducted with simulated light
Time
(h)
Temperature
(°C)
Conversion
(%)
Selectivity
(%)
Reactions
Reduction of
nitrobenzene
Hydrogenation of
azobenzene
Reduction of
acetophenone
3
6
40
30
40
30
66
33
31
18
95
79
94
89
24
Fig. 7 Contributions to activation energy of light irradiation on these demon-
strated reactions: (a) reduction of nitrobenzene; (b) hydrogenation of azoben-
zene; (c) reduction of acetophenone; and (d) deoxygenation of styrene oxide.
Deoxygenation of 20
styrene oxide
Scheme 2 Schematic mechanisms of the reaction pathways in the photocatalytic reductions with supported Au-NPs.
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 236–244 | 241

View Article Online
Paper
Mechanistic discussions
Green Chemistry
8
.
these production processes can utilize sunlight for chemical
synthesis.
The fact that light irradiation substantially reduces the appar-
ent activation energy of the reduction process also suggests
that the photocatalytic process follows a different mechanism
from that of the thermal process (Scheme 2). Generally, isopro- Experimental
panol is a hydrogen donor in the studied reduction reactions.
Raw materials
Under visible light irradiation, the Au-NPs are able to abstract
4
Chemicals of chloroauric acid (99.999%, HAuCl ), L-lysine
hydrogen from isopropanol forming the transient Au–H spe-
6
,13b
(98.0%), sodium borohydride (99.99%, NaBH ), hydrochloric
cies.
The active Au–H species attack the double bonds
4
acid (37%, HCl), nitrobenzene (99.0%), potassium hydroxide
(NvO, CvO, NvN) or the epoxide bond leading to the hydro-
(
99.0%, KOH), isopropanol (99.5%), acetophenone (99.0%),
genation or deoxygenation, in which the hydrogen of Au–H
species is consumed and the final reductive products form
without releasing hydrogen molecules. In a control exper-
iment, an efficient hydrogen-abstracting reagent, 2,2′,6,6′-tetra-
methylpiperidine N-oxyl (TEMPO)—which can abstract
hydrogen from the surface of gold and other metals to form
hydroxylamine—was added into the reaction system while the
azobenzene (98.0%) and styrene oxide (97.0%) were purchased
from Sigma-Aldrich. The inert gas of argon was from the BOC
(
9
99.99%). The supports of cerium(IV) oxide (nanopowder,
9.95%, CeO ), zirconium(IV) oxide (nanopowder, <100 nm of
2
particle size, ZrO ), titanium(IV) oxide (nanopowder, 99.5%),
2
2 3
and aluminum oxide (Al O ) were also purchased from Sigma-
1
6
Aldrich and used without further purification. The zeolite Y
was synthesized by the hydrothermal method reported in the
other experimental conditions were unchanged. No final
reductive products were detected. This confirms the important
role of the surface Au–H species in the reductions. The SPR
effect also enhances the local electromagnetic fields near
1
7
literature.
rough surfaces of the Au-NPs and this could assist activation Catalyst preparation
of the double bonds or the epoxide bond when the reactant
In a glass beaker with a magnetic stirring bar, 2.5 g of support
powders, such as CeO , were dispersed into the 100 ml of 3.8 ×
M HAuCl aqueous solution. Subsequently, 20 ml of 0.53
M lysine was added dropwise to the mixture under stirring,
and after that the stirring was prolonged for 30 min. To this
molecules are adsorbed on the surface of Au-NPs. The ability
of the Au–H species to react with the double bonds (NvO,
CvO, NvN) or the epoxide bond should depend also on the
energy of the excited electrons in Au-NPs, which is higher
when the irradiation wavelength is shorter as discussed pre-
viously. The Au–H species can release hydrogen to reactant
molecules. Moreover, higher conversions were generally
achieved for reactants with an aromatic ring than those
without the ring. It is believed that the aromatic reactants have
stronger affinity to the Au-NP surface.
2
−3
10
4
suspension, 10 ml of 0.35 M NaBH solution was added drop-
4
wise, followed by adding 10 ml of 0.3 M hydrochloric acid (to
achieve a pH = 9.5). The mixture was stirred for 1 h and aged
for 24 h. Then the solid was separated, washed firstly with
600 ml deionized water three times in a 250 ml centrifuge
bottle to remove the chlorine ions, and the residual solution
was tested by 0.1 M AgNO solution and no AgCl precipitate
3
was observed. The solids were collected, dried at 60 °C for
Conclusions
16 h, and used directly as the photocatalyst.
In summary, this study demonstrated that the four reductions,
reduction of nitro-aromatics to azo compounds, hydrogenation
Photocatalytic reduction test
of azobenzene to hydroazobenzene, reduction of ketones to The reaction was conducted in a 50 ml round-bottomed Pyrex
alcohols, and deoxygenation of epoxides to alkenes, can be glass flask with a sealed spigot and a magnetic stirrer. The
efficiently catalyzed by the Au/CeO photocatalyst under visible reaction temperature was controlled by an air-conditioner
2
light irradiation or simulated sunlight. The reduction power of within a sealed wooden box. A 500 W halogen lamp was used
the photocatalysts can be optimized by manipulating the as the incandescent light source. The light intensity in the
irradiation wavelength. Irradiation with shorter wavelength reaction position was set at 0.4 W cm− and could be adjusted
can induce the more difficult reduction with more negative by changing the distance between the reactor and the light
reduction potentials. The efficiency of the photocatalytic source. The wavelength range was tuned by using various glass
reductions increased with the increasing irradiation intensity filters to cut off the irradiation below a certain value of
under ambient conditions. The kinetic study showed that the wavelength.
2
apparent activation energy can be significantly reduced under
visible light irradiation compared with that in the thermal- conducted under the argon atmosphere. Typically, 3 mmol
driven processes. The photocatalytic reductions on Au/CeO acetophenone, 30 ml isopropanol as the solvent, and 3 ml of
Catalytic reduction of nitrobenzene to azoxybenzene was
2
followed a mechanism distinctly different from that on the 0.1 M KOH solution in isopropanol were mixed in the reactor,
semiconductor photocatalysts and gave an insight into the for- followed by adding 100 mg of the catalyst and purging with
mation of Au–H species and interaction with the reactants. argon gas, and then stirred during the reaction and illumi-
Such photocatalytic syntheses are sustainable and clean as nated with the incandescent light.
2
42 | Green Chem., 2013, 15, 236–244
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
To hydrogenate azobenzene, generally 2 mmol azobenzene
was dissolved in 20 ml isopropanol in the reactor. 100 mg of
the photocatalysts and a 2 ml solution of 0.1 M KOH in isopro-
panol were then added. After that, air in the flask was elimi-
nated with argon. The reaction was conducted under the
illumination of an incandescent light and the reaction mixture
was stirred by a magnetic stirrer.
Z. F. Zheng, S. E. Bottle and J. Chang, et al., Chem. Sci.,
2012, 3, 2138; (c) S. Malato, P. Fernández-Ibáňez,
M. I. Maldonado, J. Blanco and W. Gernjak, Catal. Today,
2009, 147, 1; (d) M. Fagnoni, D. Dondi, D. Ravelli and
A. Albini, Chem. Rev., 2007, 107, 2725; (e) M. Montalti,
A. Credi, L. Prodi and M. T. Gandolfi, Handbook of Photo-
chemistry, CRC Press, Boca Raton, 3rd edn, 2006.
Catalytic reduction of ketones to alcohols was conducted
under the argon atmosphere. Typically, 3 mmol acetophenone
was dissolved in 30 ml isopropanol, and a 3 ml solution of 0.1
M KOH in isopropanol and 100 mg of the catalyst were added
into the mixture that was then stirred during the reaction and
illuminated with the incandescent light.
2 (a) X. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011,
331, 746; (b) M. Bowker, Green Chem., 2011, 13, 2335;
(c) G. Palmisano, E. García-López, G. Marcí, V. Loddo,
S. Yurdakal, V. Augugliaro and L. Palmisano, Chem.
Commun., 2010, 46, 7074; (d) Z. G. Zou, J. H. Ye, K. Sayam
and H. Arakawa, Nature, 2001, 414, 625.
Catalytic deoxygenation of epoxides to alkenes was con-
ducted under the argon atmosphere. Typically, 3 mmol styrene
oxide was dissolved in 30 ml isopropanol, and 3 ml of a 0.1 M
KOH solution in isopropanol and 100 mg of the catalyst were
added into the mixture. The mixture was stirred during the
reaction and illuminated with the incandescent light.
During these reactions, 0.5 ml aliquots were collected at
given irradiation time intervals and filtered through a Milli-
pore filter (pore size 0.45 μm) to remove the catalyst particu-
lates. Then the flask was purged with argon again for more
than 3 min to remove air and then sealed. The filtrates were
analyzed by an Agilent 6890 gas chromatograph with the HP-5
column. Quantities of the products and reactants were calcu-
lated from the peak areas of the compounds using calibration
curves. An Agilent HP5973 mass spectrometer was used to
determine and analyze the product compositions.
3 (a) A. Tanaka, K. Hashimoto and H. Kominami, Chem.
Commun., 2011, 47, 10446; (b) H. Kominami, A. Tanaka
and K. Hashimoto, Chem. Commun., 2010, 46, 1287;
(c) S. Higashimoto, N. Kitao, N. Yoshida, T. Sakura,
M. Azuma, H. Ohue and Y. Sakata, J. Catal., 2009, 266, 279;
(d) M. Zhang, C. C. Chen, W. H. Ma and J. C. Zhao, Angew.
Chem., Int. Ed., 2008, 47, 9730.
4 (a) B. Pal, T. Torimoto, K. Okazaki and B. Ohtani, Chem.
Commun., 2007, 483; (b) A. Maldotti, L. Andreotti,
A. Molinari, S. Tollari, A. Penoni and S. Cenini, J. Photo-
chem. Photobiol., A, 2000, 133, 129; (c) S. O. Flores, O. Rios-
Bernij, M. A. Valenzuela, I. Córdova, R. Gómez and
R. Gutiėrrez, Top. Catal., 2007, 44, 507; (d) H. Tada,
T. Ishida, A. Takao and S. Ito, Langmuir, 2004, 20, 7898;
(e) F. Mahdavi, T. C. Bruton and Y. Li, J. Org. Chem., 1993,
58, 744; (f) J. L. Ferry and W. H. Glaze, Langmuir, 1998, 14,
3551; (g) O. Rios-Bernÿ, S. O. Flores, I. Córdova and
Characterization
M. A. Valenzuela, Tetrahedron Lett., 2010, 51, 2730.
5
(a) T. Uchiyama, H. Yoshida, Y. Kuwauchi, S. Ichikawa,
S. Shimada, M. Haruta and S. Takeda, Angew. Chem., Int.
Ed., 2011, 50, 10157; (b) H. Zhang, T. Watanabe,
M. Okumura, M. Haruta and N. Toshima, Nat. Mater.,
TEM images were taken with a JEOL JEM-2100 transmission
electron microscope employing an accelerating voltage of 200
kV. The specimens were sample powders deposited onto a
copper microgrid coated with a holey carbon film. The
element composition of some samples was determined by an
energy-dispersive X-ray spectroscope attached on an FEI
Quanta 200 scanning electron microscope. The diffuse reflec-
tance UV-Visible spectra of the samples were recorded on a
Cary 5000 spectrometer. The XPS data were recorded on an
ESCALAB 250 spectrometer and Al Kα radiation was used as
the X-ray source. The C1s peak at 284.5 eV was used as a refer-
ence for the calibration of the binding energy scale.
2012, 11, 49; (c) Y. Zhang, X. Cui, F. Shi and Y. Deng, Chem.
Rev., 2012, 112, 2467.
6
(a) H. Y. Zhu, X. B. Ke, X. Z. Yang, S. Sarina and H. W. Liu,
Angew. Chem., Int. Ed., 2010, 49, 9657; (b) H. Y. Zhu,
X. Chen, Z. F. Zheng, X. B. Ke, E. Jaatinen, J. C. Zhao,
C. Guo, T. F. Xie and D. J. Wang, Chem. Commun., 2009,
7
524; (c) X. B. Ke, S. Sarina, J. Zhao, X. G. Zhang, J. Chang
and H. Y. Zhu, Chem. Commun., 2012, 48, 3509;
d) X. G. Zhang, X. B. Ke and H. Y. Zhu, Chem.–Eur. J., 2012,
8, 8048.
(a) S. Linic, P. Christopher and D. B. Ingram, Nat. Mater.,
011, 10, 911; (b) P. Mulvaney, Langmuir, 1996, 12, 788;
(
1
7
8
Acknowledgements
2
X. Ke is indebted to the Queensland Government for a Smart
Futures Fellowship. The authors also gratefully acknowledge
financial support from the Australia Research Council.
(c) D. K. Roper, W. Ahn and M. Hoepfner, J. Phys. Chem. C,
2007, 111, 3636; (d) M. Jakob, H. Levanon and P. V. Kamat,
Nano Lett., 2003, 3, 353.
(a) T. Mitsudome, A. Noujima, Y. Mikami, T. Mizugaki,
K. Jitsukawa and K. Kaneda, Angew. Chem., Int. Ed., 2010,
4
9, 5545; (b) F. Zaccheria, F. Santoro, R. Psaro and
Notes and references
N. Ravasio, Green Chem., 2011, 13, 545; (c) J. Ni, L. He,
Y. M. Liu, Y. Cao, H. Y. He and K. N. Fan, Chem. Commun.,
2011, 47, 812.
1
(a) Y. Lu, H. Yu, S. Chen, X. Quan and H. Zhao, Environ.
Sci. Technol., 2012, 46, 1724; (b) S. Sarina, H. Y. Zhu,
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 236–244 | 243

View Article Online
Paper
Green Chemistry
9
(a) F. Z. Su, L. He, J. Ni, Y. Cao, H. Y. He and K. N. Fan, 13 (a) A. Primo, T. Marina, A. Corma, R. Molinari and
Chem. Commun., 2008, 3531; (b) Y. Zhu, H. Qian,
B. A. Drake and R. Jin, Angew. Chem., Int. Ed., 2010, 49,
H. García, J. Am. Chem. Soc., 2011, 133, 6930;
(b) A. Maldotti, A. Molinari, R. Juárez and H. García, Chem.
Sci., 2011, 2, 1831; (c) H. Kominami, A. Tanaka and
K. Hashimoto, Appl. Catal., A, 2011, 397, 121.
1
295.
1
1
1
0 (a) E. Merino, Chem. Soc. Rev., 2011, 40, 3835;
(
2
b) H. M. Nanjundaswamy and M. A. Pasha, J. Chem. Res., 14 (a) A. Furube, L. Du, K. Hara, R. Katoh and M. Tachiya,
005, 12, 772; (c) A. Grirrane, A. Corma and H. Carcía,
J. Am. Chem. Soc., 2007, 129, 14852; (b) M. Murdoch,
G. I. N. Waterhouse, M. A. Nadeem, J. B. Metson and
M. A. Keane, et al., Nat. Chem., 2011, 3, 489.
Science, 2008, 322, 1661.
1 (a) S. Kohtani, E. Yoshioka, K. Saito, A. Kudo and
H. Miyabe, Catal. Commun., 2010, 11, 1049; (b) K. Boujlel, 15 C. G. Silva, R. Juárez, T. Marino, R. Molinari and H. García,
P. Martigny and J. Simonet, J. Electroanal. Chem., 1983, 144,
37.
J. Am. Chem. Soc., 2011, 133, 595.
4
16 (a) J. P. Roth, J. C. Yoder, T. J. Won and J. M. Mayer,
Science, 2001, 294, 2524; (b) L. Tebben and A. Studer,
Angew. Chem., Int. Ed., 2011, 50, 5034.
2 (a) A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int.
Ed., 2006, 45, 7896; (b) M. D. Hughes, Y. J. Xu, P. Jenkins,
P. McMorn, P. Landon and D. I. Enache, Nature, 2005, 437, 17 B. A. Holmberg, H. T. Wang, J. M. Norbeck and Y. S. Yan,
132; (c) M. Haruta, Catal. Today, 1997, 36, 153–166. Microporous Mesoporous Mater., 2003, 59, 13.
1
2
44 | Green Chem., 2013, 15, 236–244
This journal is © The Royal Society of Chemistry 2013