Efficient photocatalytic Suzuki cross-coupling reactions on Au-Pd alloy nanoparticles under visible light irradiation

Article abstract of DOI:10.1039/c4gc00588k

We report herein highly efficient photocatalysts comprising supported nanoparticles (NPs) of gold (Au) and palladium (Pd) alloys, which utilize visible light to catalyse the Suzuki cross-coupling reactions at ambient temperature. The alloy NPs strongly absorb visible light, energizing the conduction electrons of NPs which produce highly energetic electrons at the surface sites. The surface of the energized NPs activates the substrates and these particles exhibit good activity on a range of typical Suzuki reaction combinations. The photocatalytic efficiencies strongly depend on the Au:Pd ratio of the alloy NPs, irradiation light intensity and wavelength. The results show that the alloy nanoparticles efficiently couple thermal and photonic energy sources to drive Suzuki reactions. Results of the density functional theory (DFT) calculations indicate that transfer of the light-excited electrons from the nanoparticle surface to the reactant molecules adsorbed on the nanoparticle surface activates the reactants. The knowledge acquired in this study may inspire further studies of new efficient photocatalysts and a wide range of organic syntheses driven by sunlight. the Partner Organisations 2014.

Full text of DOI:10.1039/c4gc00588k

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Efficient photocatalytic Suzuki cross-coupling
reactions on Au–Pd alloy nanoparticles under
visible light irradiation†
Cite this: Green Chem., 2014, 16,
4272
Qi Xiao,a Sarina Sarina,a Esa Jaatinen,a Jianfeng Jia,b Dennis P. Arnold,a
Hongwei Liuc and Huaiyong Zhu*a
We report herein highly efficient photocatalysts comprising supported nanoparticles (NPs) of gold (Au)
and palladium (Pd) alloys, which utilize visible light to catalyse the Suzuki cross-coupling reactions at
ambient temperature. The alloy NPs strongly absorb visible light, energizing the conduction electrons of
NPs which produce highly energetic electrons at the surface sites. The surface of the energized NPs acti-
vates the substrates and these particles exhibit good activity on a range of typical Suzuki reaction combi-
nations. The photocatalytic efficiencies strongly depend on the Au : Pd ratio of the alloy NPs, irradiation light
intensity and wavelength. The results show that the alloy nanoparticles efficiently couple thermal and photonic
energy sources to drive Suzuki reactions. Results of the density functional theory (DFT) calculations indicate
that transfer of the light-excited electrons from the nanoparticle surface to the reactant molecules adsorbed
on the nanoparticle surface activates the reactants. The knowledge acquired in this study may inspire further
studies of new efficient photocatalysts and a wide range of organic syntheses driven by sunlight.
Received 3rd April 2014,
Accepted 24th June 2014
DOI: 10.1039/c4gc00588k
www.rsc.org/greenchem
key issues for sustainable development of the chemical indus-
try. Hence huge efforts have been made to develop hetero-
Introduction
Cross-coupling reactions are powerful and versatile tools in geneous Pd catalysts for Suzuki reactions,8 including the
modern organic synthesis for the formation of carbon–carbon immobilization or stabilization of Pd NPs on different sup-
bonds.1 Among various cross-coupling reactions, the Suzuki ports, such as high surface-area silica,9 carbon nanotubes,10
reaction, discovered by the 2010 Nobel laureate Akira Suzuki,2 polymers,11 metal oxides,12 double hydroxides,13 dendrimers14
is of great significance in synthetic chemistry for biaryl com- and magnetic nanomaterials.15 For instance, gold–palladium
pounds, which have many applications as intermediates in the bimetallic NPs were reported as highly active catalysts for
preparation of materials,3 natural products,4 and bioactive Suzuki reactions.16 Nonetheless, many of these reported pro-
compounds.5 The use of well-tolerated functionalized aryl cesses with heterogeneous catalysts require relatively elevated
halides to react with innocuous boronic acids, which are gen- reaction temperature and prolonged reaction time. In some
erally non-toxic and thermally, air- and moisture-stable, is the cases, the active sites on the porous solid supports are not
practical advantage of the Suzuki reaction, relative to many readily accessible for the reactants.8 Therefore, highly active,
other cross-coupling processes.6 In the past, the Suzuki reac- easily separable and reusable catalyst systems, combined with
tions were typically performed under homogeneous conditions mild, green chemistry techniques, are still considered as an
using phosphine ligand/palladium as the catalytic system, important objective for Suzuki reactions.
which has shown relatively high activity and selectivity.7
Photocatalysis is particularly intriguing in the realm of
However, the separation and recovery of the catalyst represent green chemical science, because it combines the efficiency of
catalysis with the potential use of sunlight.17 Photocatalysis
driven by visible light is an ideal process, because of the abun-
dance of visible light, its benign environmental impact, and
aSchool of Chemistry, Physics and Mechanical Engineering, Science and Engineering
Faculty, Queensland University of Technology, Brisbane, QLD 4001, Australia.
sustainability. Recently, visible light photocatalytic reactions
have shown great potential in organic synthesis.18 We have
been particularly interested in using photocatalysis as a new
means to drive chemical reactions for organic synthetic trans-
formations with high activity and selectivity, such as visible
light induced oxidation and reduction of aryl aromatics using
E-mail: hy.zhu@qut.edu.au; Fax: +61 7 3138 1804; Tel: +61 7 3138 1581
bSchool of Chemistry and Material Science, Shanxi Normal University,
Linfen 041004, China
cSydney Microscopy & Microanalysis, Australian Centre for Microscopy &
Microanalysis, The University of Sydney, Sydney, NSW 2006, Australia
†Electronic supplementary information (ESI) available: Fig. S1–S5, DFT calcu-
lations. See DOI: 10.1039/c4gc00588k
4272 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
gold NP photocatalysts.19 Very recently, we have incorporated
Pd into Au NPs to form Au–Pd alloy NPs and utilized the
ensemble properties of the alloy NPs to drive several reactions,
including Suzuki reactions, with visible light at ambient temp-
erature.20 It was found that the intrinsic catalytic activity of Pd
can be significantly enhanced under light irradiation of alloy
NPs due to the electronic heterogeneity at the alloy NP surface
and the energy absorbed from the incident light. Two other
research groups also found, almost at the same time, that it is
an effective approach to use irradiation to accelerate Suzuki
cross-coupling reactions using Au–Pd bimetallic nano-
structures.21 In our previous study, we believed that the con-
duction electrons of the alloy NPs gain the energy of the
incident light, generating electrons at high energy levels
(“energetic electrons”). These energetic electrons are available
at the surface Pd sites of the alloy NPs. The surface sites have
good affinity for the reactant molecules, and the energetic elec-
trons at these sites enhance the intrinsic ability of the metal
sites to activate the reactant molecules. The charge hetero-
geneity of the alloy NP surface, due to the different electrone-
gativities of gold and palladium, also plays a key role in the
catalytic reactions. This tentatively proposed mechanism of
light-excited electrons on alloy NPs and the interaction of the
reactant molecules adsorbed on the NPs with the excited elec-
trons and the NP surface are still by no means well under-
stood. The delineation of detailed steps involved in the
reaction is quite a profound challenge.
In this study, we focused on a systematic investigation of
the Au–Pd alloy NP catalysed Suzuki reaction under visible
light irradiation. We found that the performance of the alloy
catalyst depends on the alloy composition, light intensity and
wavelength. Theoretical calculations show that the alloying of
gold and palladium enhances the interaction between the reac-
tant molecules and the alloy NPs. The strong interaction facili-
tates the transfer of light-excited electrons on the alloy NPs to
the reactant molecules adsorbed on the NPs, and such electron
transfer weakens the C–I bond of the reactant molecules and
facilitates the reactions. Understanding this mechanism is
useful for developing photocatalytic versions of other cross-
coupling reactions.
Table 1 Metal content, the Au : Pd ratio and specific surface area of the
photocatalysts
Au : Pd
[molar ratio]
Surface areaa
Catalyst
Au [wt%]
Pd [wt%]
[m2 g−1
]
Alloy-1
Alloy-2
Alloy-3
Alloy-4
Au-ZrO2
Pd-ZrO2
1.63
1.11
5.72
5.03
3.20
0
1.43
3.07
0.99
2.66
0
1 : 1.62
1 : 5.14
1 : 0.33
1 : 0.99
1 : 0
10.54
10.35
10.59
10.55
13.38
9.99
3.12
0 : 1
a The specific surface area of the pure ZrO2 support was measured to
be 11.17 m2 g−1
.
the samples were derived from the metal content and are also
listed in Table 1. The specific surface areas of the photocata-
lysts were estimated from N2 physical sorption data using the
Brauner–Emmet–Teller (BET) method.
The BET specific surface areas of the samples are similar to
that of the pure ZrO2 support (Table 1). The ZrO2 support has
a moderate specific surface area, and loading with the metal
NPs does not cause significant change in the overall specific
surface area of samples, and the overall specific surface area
seems not to influence substantially the photocatalytic activity
of the samples in the present study.
Transmission electron microscopy (TEM) analysis of the
NPs (Fig. 1A) shows Au–Pd alloy NPs uniformly dispersed on
the surfaces of the ZrO2 crystals. The mean diameters of the
pure Au, pure Pd and the Au–Pd alloy NPs are less than 7 nm.
High resolution TEM (HR-TEM) in Fig. 1B and 1C also con-
firmed the formation of the Au–Pd alloy NP on the ZrO2 crystal
surface. Line profile analysis of the energy dispersive X-ray
(EDX) spectrum for a typical Au–Pd alloy NP shows that the NP
consists of both Au and Pd distributed spherically around a
Results and discussion
Catalyst preparation and characterization
Photocatalysts made of Au and Pd alloy NPs with various
Au : Pd ratios on the ZrO2 support were prepared via an
impregnation–reduction procedure as described in our pre-
vious paper.20 The different Au : Pd ratios were achieved by
using different quantities of HAuCl4 or PdCl2 aqueous solu-
tions under otherwise unchanged experimental conditions.
The Au and Pd contents in the alloy NPs are given in Table 1.
The metal element contents of the catalysts were determined
by energy-dispersive X-ray spectroscopy (EDS). For comparison,
catalysts of the pure Au NP or the pure Pd NP on ZrO2 were
Fig. 1 Catalyst characterization. (A) TEM image of the Au–Pd alloy NPs.
(B) HR-TEM image of the typical Au–Pd alloy NPs. (C) HR-TEM image of
an alloy particle indicated in Fig. 1B (green square). (D) Line profile ana-
lysis of the EDX spectrum of a typical Au-Pd NP indicated in Fig. 1B (blue
dotted line) providing the information on the elemental composition
also prepared by a similar method. The Au–Pd molar ratios of and Au/Pd distribution of the NP.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4273

View Article Online
Paper
Green Chemistry
common centre, which means that the two metals exist as the metal NPs. The absorption peak at 520 nm in the spectrum
binary alloy NPs in this sample (Fig. 1D). of the pure Au NP sample is due to the characteristic localized
Fig. 2 shows the X-ray diffraction (XRD) patterns of the cata- surface plasmon resonance (LSPR) absorption of Au NPs.22
lysts with different Au : Pd ratios on ZrO2. All diffraction peaks The presence of the support and its interaction with the Au
can be indexed to a monoclinic structure of the ZrO2 crystal NPs can shift and broaden this peak. The LSPR absorption
(JCPDS, no. 65-2357), no reflection peaks of Au and Pd were band of Pd NPs is deep within the UV wavelength range, so its
observed by the XRD patterns, because the metal content is light absorption at solar wavelengths occurs through both
low and the metal diffraction peaks may interfere with the LSPR and interband electron transition contributions.23 The
diffraction peaks of ZrO2; this result suggests that the detec- spectrum of the Au–Pd alloy NP sample is clearly different
tion of metal NP signals in XRD patterns is also closely related from the spectra of the pure metal NPs. The dielectric constant
to the supporting materials.
of the NP changes and the plasmon resonance are determined
The formation of Au–Pd alloy NPs is also supported by the by the dielectric constant of the alloy NPs. In the spectra of the
light absorption properties of the samples, as shown in Fig. 3. alloy NP samples, the characteristic Au LSPR absorption peak at
ZrO2 has a band-gap of about 5 eV,19a and exhibits weak 520 nm is much weaker compared with the spectrum of the
absorption of visible light of wavelengths above 400 nm and pure Au sample. For example in Alloy-1 the molar ratio of
therefore, the ZrO2 support itself does not contribute to the Au : Pd is 1 : 1.62. The number of Pd atoms in this alloy is much
photocatalytic activity. In contrast, all the NP photocatalysts larger than that of Au atoms. Thus, the absorption due to the
display strong absorption in the UV and visible ranges of the LSPR effect of gold nanostructure has limited contribution to
spectrum, indicating that solar energy is strongly coupled to the overall light absorption. Nonetheless, the absorption of the
alloy NPs in the visible range is more intense than that observed
for the pure Pd sample. This means that the alloy NPs have
better ability to gain light energy, which is important to their
catalytic performance under visible light irradiation.
Catalyst screening in Suzuki cross-coupling reactions
The photocatalytic activity of Alloy-1 for the Suzuki cross-coup-
ling reaction between 3-iodotoluene and phenylboronic acid
was tested at 30 °C and the results are summarized in Table 2
(entry 1). An excellent yield (96%) of the desired cross-coupling
product 3-methylbiphenyl was achieved under visible light
Table 2 Catalyst screening in Suzuki cross-coupling reactionsa
Fig. 2 XRD patterns of catalysts with different Au : Pd ratios on ZrO2.
Vertical bars represent the standard diffraction data for the monoclinic
ZrO2 (JCPDS, no. 65-2357).
Incident
light
Conv.b
[%]
TOFc
[h−1
Q.Y.d
[%]
Entry
Catalyst
Alloy-1
Alloy-2
Alloy-3
Alloy-4
Au-ZrO2
Pd-ZrO2
TONc
]
1
2
3
4
5
6
Visible
Dark
Visible
Dark
Visible
Dark
Visible
Dark
Visible
Dark
Visible
Dark
96
37
40
10
28
6
55
17
2
87
34
34
8
33
6
56
17
3
14.5
5.7
5.7
1.3
5.5
1.0
9.3
2.8
0.5
0
4.4
1.8
1.3
2.5
0.1
1.2
0
26
11
0
18
8
3.0
1.3
a Reaction conditions: 3-iodotoluene (1 mmol), phenylboronic acid
(1.5 mmol), photocatalysts (50 mg, containing 3% of metals), base
K2CO3 (3 mmol), 20 mL of the solvent N,N-dimethylformamide (DMF)–
H2O (V : V = 3 : 1), 30 °C, 6 h under an argon atmosphere, light
intensity 0.5 W cm−2 b GC conversion: (aryl iodide converted)/(initial
.
amount of aryl iodide) × 100. c TON and TOF were calculated based on
the total amount of metal(s). d Q.Y.: quantum yield, the calculation
method of Q.Y. is given in ESI.
Fig. 3 UV-vis spectra of Au–Pd NPs with different Au : Pd ratios on
ZrO2.
4274 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
irradiation within 6 h; only a trace of the homo-coupling sion. Thus the Au–Pd alloy NP photocatalyst is much greener
product of phenylboronic acid was detected. Control experi- and efficient.
ments showed that the yield decreased to 37% when the reac-
tion was conducted in the dark while other experimental
conditions remained identical. This confirms that visible light
The influence of reaction conditions
The influence of several critical reaction conditions in the
cross-coupling reaction of 3-iodotoluene with phenylboronic
acid, such as solvents, bases, and reaction atmosphere, has
been investigated using the Alloy-1 catalyst under visible light
irradiation. First, the reaction was carried out under argon,
oxygen and air atmospheres, respectively, and we observed
that inert gas atmosphere promoted the reaction markedly
(Table 4, entries 1–3). Under the oxygen atmosphere, more
homocoupling and oxidation products of phenylboronic acid
were detected by GC-MS. Several commonly used solvents were
used for the reaction while other reaction conditions were
maintained unchanged. The polar aprotic solvents, such as N,
N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO) and
dioxane afforded low yields of the coupling product (even trace
yield in dioxane), the nonpolar solvent toluene also gave a
trace yield (Table 4, entries 12, 13, 16 and 17). In contrast,
much higher yield was observed in the polar protic solvents
such as ethanol (Table 4, entry 14). Interestingly, we found
that the reactions in both pure DMF and H2O were sluggish
(Table 4, entries 12, 15); however, a mixed solvent of DMF and
H2O led to a profound increase in activity. Excellent yield and
irradiation significantly facilitates the Suzuki cross-coupling
reactions using the alloy photocatalyst.
To investigate the effect of alloy NP photocatalyst compo-
sition, the coupling of 3-iodotoluene with phenylboronic acid
was used as a model reaction to screen the effect of catalysts
with different compositions. Photocatalysts with varying Au–
Pd compositions show differences in the reaction conversion
rates (Table 2, entries 1–6). The highest yield of the target
product was found when using Alloy-1 (Au : Pd molar ratio =
1 : 1.62) as the catalyst, showing relatively high turnover fre-
quency (TOF) and photo quantum yield (Q.Y.). The pure Pd
catalyst showed poor performance (26%) and the pure Au cata-
lyst exhibited little activity (with a conversion of 2%). There-
fore, palladium is an indispensable active component of the
catalyst for Suzuki cross-coupling reactions. The photocatalytic
properties of Au–Pd alloy NPs are much better than those of
the pure metal NPs of either component metal or that of a
mechanical mixture of the pure metal NPs of the two elements
in the same ratio and quantity. The underlying cause of this
apparent improvement in reactivity and its dependence on
alloy composition is of great interest, and is believed to be
related to the electron redistribution between the two metals,
which will be further discussed below.
Table 4 Influence of the reaction conditions for Suzuki cross-coupling
reactionsa
Compared with some literature reported reaction con-
ditions of heterogeneous Pd catalysts for Suzuki cross-coup-
lings (Table 3), it can be seen that Pd NPs supported on
various supports can drive the Suzuki coupling under elevated
temperatures, while Au–Pd alloy NPs can drive the reaction at
only 30 °C with visible light irradiation achieving high conver-
Conv.b
[%]
Sel.c
[%]
Entry
Base
Solvent
Atmosphere
Table 3 Comparison of the reaction conditions and achieved conver-
sion of heterogeneous Pd based catalysts reported in the literature for
Suzuki cross-coupling reactions using iodobenzene and phenylboronic
acid as substrates
1
2
3
4
5
6
7
8
K2CO3
K2CO3
K2CO3
Na2CO3
Cs2CO3
NaOAc
KI
Na3PO4
NaF
NaOH
KOH
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF–H2O (3 : 1)
DMF
DMSO
EtOH
H2O
Toluene
Dioxane
EtOH–H2O (1 : 1)
DMSO–H2O (3 : 1)
Ar
O2
air
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
Ar
96
25
45
79
86
13
8
>99
78
91
90
90
70
0
>99
>99
92
90
69
96
45
>99
—
Conv.
[%]
Entry
Catalysts
Conditionsa
89
4
0.1 mol% Pd/MCM-419c
LDH-DS-Pd13b
XL-Pd11a
78 °C, 5 h
80 °C, 5 h
115 °C
93%
9
1
2
3
4
5
6
7
8
10
11
12
13
14
15
16
17
18
19
75
86
10
34
73
5
Trace
Trace
60
96
93%
86%
71%
98%
47%
100%
97%
99%
95%
100%
95%
89%
100%
98%
24
G3-OH(Pd)10
Reflux, 24 h
78 °C, 18 h
Reflux, 24 h
60 °C, 24 h
95 °C, 4 h
100 °C, 2 h
Reflux, 2 h
80 °C, 5 h
G4-OH Pd NPs25
Pd-G-314a
0.01 mol% Pd-CD26
Pd-Ln27
Silica-APTS-Pd28
3% Pd/MWCNT29
Polymer-anchored Pd30
Pd-NP-131
—
>99
95
9
10
11
12
13
14
15
Reflux, 4 h
80 °C, 12 h
20 °C, 1 d
a Reaction conditions: 3-iodotoluene (1 mmol), phenylboronic acid
(1.5 mmol), Alloy-1 photocatalysts (50 mg, containing 3% of metals),
base (3 mmol), 20 mL of solvent, environment temperature 30 °C,
Pd-S-GaAs(001)32
PdTNs33
Au–Pd alloy NPs in
the present study
30 °C, 6 h visible light
reaction time 6 h, light intensity 0.5 W cm−2 b GC conversion: (aryl
.
iodide converted)/(initial amount of aryl iodide) × 100. c Selectivity:
(target product formed)/(total amount of product formed) × 100.
a Reaction conditions: reaction temperature, reaction time.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4275

View Article Online
Paper
Green Chemistry
selectivity were achieved when the volume ratio of DMF and 1). Na3PO4 also furnished the coupling products in good activity
H2O was 3 : 1 (Table 4, entry 1). Other mixed solvents such as (Table 4, entry 8). In contrast, Na2CO3, Cs2CO3 and hydroxide
EtOH–H2O and DMSO–H2O were also tested, and exhibited bases led to relatively lower yields and selectivities (Table 4,
decent activity (Table 4, entries 18, 19). The superior function entries 4, 5, 10 and 11) and other weak bases such as NaOAc, KI
of the mixed solvent may be due to the increased solubility of and NaF gave very poor activity (Table 4, entries 6, 7 and 9).
the reactants and bases, as the ability to dissolve bases in
The scope of Suzuki cross-coupling reactions with different
substituents
water for activating arylboronic acid can enhance the rate of
the reaction in the aqueous medium. The influence of the
mixed solvent on the Suzuki cross-coupling reactions has been
reported in the literature.34
For the various bases added into the reaction system, the
best results were obtained with K2CO3 as the base in the DMF–
H2O (V : V = 3 : 1) solvent at a given temperature (Table 4, entry
With the optimized conditions in hand, the photocatalytic per-
formance of Alloy-1 for various substrates was investigated. All
the reactions were carried out at 30 °C in argon using K2CO3
as the base and DMF aqueous solution (DMF–H2O = 3 : 1) as
the reaction medium. The results are summarized in Table 5.
Table 5 The scope of Suzuki cross-coupling reactions with different substituents
Entry
1
Reactant 1
Reactant 2
Product
Conv.a [%] (dark)
98b (45)
Sel.a [%] (dark)
>99 (99)
2
3
85b (30)
98 (99)
96b (37)
>99 (90)
4
93b (48)
95c (65)
100c (100)
90b (50)
23b (6)
98 (86)
>99 (99)
98 (99)
98 (99)
>99 (99)
>99
5
6
7
8
9
44b (trace)
98c (58)
97d (65)
80e (55)
10
11
12
>99 (99)
>99 (96)
68 (60)
13
70 f (55)
99 (96)
a GC conversion: (aryl iodide converted)/(initial amount of aryl iodide) × 100; selectivity: (target product formed)/(total amount of product
formed) × 100. The values in the parentheses are the data for dark reactions. Reaction conditions: aryl iodide (1 mmol), aryl boronic acid
(1.5 mmol), Alloy-1 photocatalysts (50 mg, containing 3% of metals), base K2CO3 (3 mmol), 20 mL of the solvent N,N-dimethylformamide (DMF)–
H2O (V : V = 3 : 1), environment temperature 30 °C, argon atmosphere, light intensity 0.5 W cm−2 b Reaction time 6 h. c Reaction time 2 h.
.
d Reaction time 4 h. e Reaction time 22 h. f 1,4-Diiodobenzene (1 mmol), phenylboronic acid (2 mmol), reaction time 6 h, the other reaction
conditions kept identical.
4276 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
The cross-coupling reactions of aryl iodides with electron dence. We calculated the contributions of the light irradiation
withdrawing and electron donating substituents and phenyl- to the conversion efficiency by subtracting the conversion of
boronic acid afforded the desired biaryls in good yields in the reaction in the dark from the overall conversion observed
various reaction periods. The catalytic system is also effective when the system was irradiated, with both reactions occurring
for electronically diverse arylboronic acids (Table 5, entries at identical reaction temperature. Here the conversion of the
10–12). Most of the reactions can be accomplished within 6 h, reaction in the dark is regarded as the contribution of thermal
but the coupling reaction of 4-iodoaniline (Table 5, entry 8) effect. The relative contributions of light and thermal pro-
and 4-iodophenol (Table 5, entry 9) with phenylboronic acid cesses to the conversion efficiencies are shown in Fig. 4. We
led only to modest conversions. In most cases, the formation can see that the higher the light intensity, the greater the con-
of homo-coupling products was not observed, and the extent tribution of irradiation to the overall conversion rate. When
of this side reaction did not exceed 2% except for the case of the light intensity is 0.1 W cm−2, the light contribution was
4-(N,N-dimethylamino)phenylboronic acid with iodobenzene only 27%, and when the light intensity increased to 0.5 W
(Table 5, entry 12).
cm−2, 67% of the conversion is due to irradiation. A stronger
We also tried to use aryl bromides as substrates to achieve light intensity will induce a larger population of electrons in
the reaction under visible light irradiation; however, the usual high energy levels and create a stronger electromagnetic field
reaction conditions above achieved poor conversion (<5%). around the NPs (field enhancement effect), as reported for Au
Thus we tried to improve the conditions and found that using NPs.35 The field enhancement effect also contributes to a
a stronger base NaOH combined with cetyltrimethylammo- stronger interaction between the NPs and reactant molecules,
nium bromide (CTAB) and H2O as the solvent can drive the and thus enhanced catalytic activity of the coupling
reaction effectively. CTAB helps in bringing bromobenzene reactions.20
into the aqueous reaction solution.21a As can be seen from
In addition, the photocatalytic reactions were conducted
Table 6, we can efficiently enhance the catalytic activity of the under high intensity white LED irradiation (the LED light
reactions using aryl bromide substrates with alloy NPs under intensity is 5.0 W cm−2, and the wavelength range is between
light irradiation.
400 nm and 700 nm as shown in Fig. S2, ESI†). There is no
apparent difference in conversions between the two kinds of
light sources when using Alloy-1, Alloy-3 and Alloy-4 catalysts
(Fig. 5). In contrast, when using Alloy-2 (much higher Pd
content, Au : Pd molar ratio = 1 : 5.14) for the photocatalytic
procedure, the reaction under high intensity LED (67%)
showed much better conversion than the normal halogen
lamp (40%). We investigated the difference of wavelength
range between the LED and halogen lamps and it is possible
that the absence of the wavelength range near 500 nm for the
The influence of light intensity and wavelength
We investigated the dependence of catalytic activity on light
intensity and the results are depicted in Fig. 4. When the
irradiation intensity was increased from 0.1 to 0.2, 0.3, 0.4 and
0.5 W cm−2 with other reaction conditions unchanged, the
conversion of the reactions on the Au–Pd alloy NPs increased.
There is a positive relationship between the intensity and reac-
tion rate. The results clearly show an almost linear depen-
Table 6 Photocatalytic Suzuki cross-coupling reactions using aryl bromides as substrates
Entry
1
Reactant 1
Reactant 2
Product
Conv.a [%] (dark)
80 (25)
Sel.a [%] (dark)
>99 (99)
2
3
4
25 (0)
87 (20)
28 (0)
98 (−)
>99 (95)
>99 (−)
5
78 (50)
>99 (99)
a GC conversion: (aryl bromide converted)/(initial amount of aryl bromide) × 100; selectivity: (target product formed)/(total amount of product
formed) × 100. The values in the parentheses are the data for dark reactions. Reaction conditions: aryl bromide (1 mmol), aryl boronic acid
(1 mmol), Alloy-1 photocatalysts (50 mg, containing 3% of metals), NaOH (3 mmol), cetyltrimethylammonium bromide (CTAB) (1 mmol), 10 mL
H2O, environment temperature 30 °C, reaction time: 3 h, light intensity 0.5 W cm−2
.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4277

View Article Online
Paper
Green Chemistry
5 nm, 470 5 nm, 530 5 nm, 590 5 nm and 620 5 nm,
respectively, were used and the rates converted to the apparent
quantum efficiencies (AQE). The AQE was calculated as:
AQE (%) = [(Ylight − Ydark)/(the number of incident photons)] ×
100, where Ylight and Ydark are the amounts of products formed
under light irradiation and dark conditions, respectively. The
plot of the AQE versus the respective wavelengths is the action
spectrum of the reaction. As shown in Fig. 6, the action spec-
trum of the Suzuki reaction matches well with the absorption
spectrum of Au–Pd alloy NPs, and higher activity is observed at
the wavelengths where the Au–Pd alloy NPs strongly absorb the
light. When this reaction was conducted in the dark both
Pd NPs and Au NPs exhibited much lower catalytic activities
than the alloy NPs (Table 2). This can be attributed to the
higher surface charge heterogeneity of the alloy NPs, which
contributes positively to the reaction.20 When irradiated with
light, the reaction yield increases significantly. The most sig-
nificant enhancement is observed from the alloy NPs, which
tracks the light absorption by the alloy NPs. Furthermore, the
shorter wavelengths with higher energy photons are able to
excite electrons to higher energy levels that can then be trans-
ferred to the adsorbed reactant molecules more readily as pro-
posed in the Mechanism section. Therefore when the light
intensity is constant, the shorter the wavelength, the higher
AQE. Generally the light absorption of short wavelengths
mainly excites inter-band electron transition,23b which
enhances the yield of the reaction significantly, and the overall
enhancement tracks the light absorption by the alloy NPs.
The dependence of photocatalytic activity on light intensity
and wavelength indicates that energetic electrons excited by
light absorption are responsible for the observed photo-
catalytic activity.37 Since the reaction rate is expected to
depend on the population of electrons with sufficient energy
to initiate the reaction of the reactant molecules, one can
increase the number of energetic electrons by applying higher
light intensity. Tuning the irradiation wavelength can also
Fig. 4 The dependence of the catalytic activity of Au–Pd alloy NPs for
Suzuki reaction on the intensity of the light irradiation. The numbers
with percentages show the contribution of the light irradiation effect.
Reaction conditions: 3-iodotoluene (1 mmol), phenylboronic acid
(1.5 mmol), Alloy-1 photocatalysts (50 mg, containing 3% of metals),
K2CO3 (3 mmol), 20 mL of the solvent DMF–H2O (V : V = 3 : 1), 30 °C, 4 h
under an argon atmosphere. In the reactions to determine the light-
intensity dependence, a photometer was used to measure the light
intensity; the other experimental conditions were kept identical.
white LED light may strongly reduce the LSPR effect of Au NPs
in the alloy NPs, thus Alloy-1, Alloy-3 and Alloy-4 show sluggish
differences when using high intensity LED irradiation. The
underlying cause of the apparent improvement in reactivity of
Alloy-2 is believed to be a synergistic effect of light induced
inter-band electron transitions of Pd NPs and the charge re-
distribution between gold and palladium. These results imply
that energetic electrons derived from Pd sites on the alloy
surface may play a significant role in the photocatalytic
process.
A useful tool for determining whether an observed reaction
occurs via a photo-induced process or a thermo-catalytic
process is the action spectrum, which should show one-to-one
mapping between the wavelength-dependent photocatalytic
rate and the light extinction spectrum.36 In the present study,
the reaction rates of the photocatalyic Suzuki reaction using
Alloy-1 at 30 °C under irradiation with different wavelengths
were determined. Five LED lamps with wavelengths 400
Fig. 6 Action spectrum for Suzuki cross-coupling reactions (red
points). The light absorption spectrum (left axis) is the diffuse reflec-
tance UV-visible (DR-UV-vis) spectrum of the Au–Pd alloy NPs (blue
curve). 3-Iodotoluene was used as the aryl halide substrate to react with
phenylboronic acid under visible light irradiation.
Fig. 5 The reaction conversion comparison of varying catalysts using
different light sources.
4278 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
increase the number of energetic electrons and may also assist increase catalytic activity.40 Furthermore, the Fermi level in
us to understand the mechanism of the reactions.
alloy NP (Φalloy) is higher than that in pure Pd NP (ΦPd), so
that the transfer of electrons at the Fermi level of the alloy NPs
to reactant molecules adsorbed on the NPs is easier compared
with that from the Fermi level of pure Pd NPs to the adsorbed
molecule. The light absorption of gold results in energetic con-
The influence of the Au : Pd composition ratio
The results in Table 2 reveal the strong dependence of photo-
catalytic performance of Au–Pd alloy catalysts on the Au : Pd
molar ratio for the Suzuki reaction. Of the four alloys studied,
it was found that the highest yield of target products was
achieved when the alloy NPs have the Au : Pd molar ratio of
1 : 1.62. At this molar ratio, the ratio of the electrons in Au to
those in Pd is 1 : 0.89 (near 1 : 1). Alloy NPs with other Au : Pd
ratios exhibited much lower activities. A possible explanation
for the superior catalytic activities of Au–Pd alloy NPs to NPs
consisting of either pure component is that Au can isolate
active Pd sites within bimetallic systems.38 However, this does
not explain why optimal catalytic activity was observed when
the Au : Pd electron ratio is near 1 : 1. It is possible that a
charge redistribution on the alloy particle surface can contrib-
ute to the increase of the photocatalytic activity.20 Such a
charge redistribution stems from the electronic structure of
the metals. As illustrated in Scheme 1, pure palladium has a
slightly larger work function (ΦPd ∼ 5.6 eV) than pure gold
(ΦAu ∼ 5.3 eV),39 so once the two metals are in contact, conduc-
tion electrons will flow between gold and palladium until equi-
librium is reached (Φ*) with the electron chemical potential
equal everywhere in the alloy NP. A consequence of such a
charge-redistribution is that the palladium atoms at the NP’s
surface will be electron rich sites and the gold atoms at the
surface will be slightly positively charged. The heterogeneity in
charge distribution may enhance interaction between reactant
molecules and the alloy NPs when the reactant molecules are
electrophilic or nucleophilic.20 The enhanced interaction is
able to lower the activation energy of the reaction and thus
duction electrons, which are in even higher energy level Φalloy
*
and have a higher driving force to migrate to the Pd sites on
the surface. This further increases the possibility of electron
transfer from the alloy NPs to the reactant molecules. The
detailed transfer of the light-induced electrons to the adsorbed
reactant molecule is further discussed below.
The underlying cause of this apparent improvement in reac-
tivity and its dependence on alloy composition is of great inter-
est, and is believed to be related to the electron redistribution
between the two metals. The electron redistribution can be
estimated by using a free electron-gas model.41 The analysis
reveals that the number of electrons transferred between the
two metals (ΔN) is maximum when the ratio of the electrons
of the two metals in the alloy NPs is approximately equal
(details are provided in ESI text S1†). As shown in Fig. 7 the
electron transfer (ΔN) predicted by the model is a function of
the gold electron concentration (%) in the Au–Pd alloy NPs,
and/or the Au : Pd molar ratio. More importantly, there is a
strong correlation between the number of electrons transferred
ΔN and the conversion efficiency of reactants over the alloy NP
photocatalysts: the photocatalytic conversions (refer to the axis
on the right hand side) of the photocatalysts for the Suzuki
reaction are given (the red symbol) in Fig. 7. As discussed for
the NPs investigated here, alloy NPs with a Au : Pd electron
ratio near 1 : 1 (Au : Pd molar ratio: 1 : 1.62) possess the largest
electron transfer number ΔN and display significant Au–Pd
ionic bond character because of Pd rich surface of the NP.
Scheme 1 The conduction electrons of Au NPs exist at the NP’s surface and a small fraction of the conduction electrons distribute in the energy
levels above the Fermi level at ambient temperature. The surface electronic properties of the alloy NPs are different from those of pure gold NPs as
there are Pd islets on the alloy NP’s surface. The Pd sites are electron-rich because Pd (ΦPd ∼ 5.6 eV) has a slightly larger work function than gold
(ΦAu ∼ 5.3 eV) and electrons will flow from gold to palladium until equilibrium is reached (the chemical potentials of the electrons are equal in the
two metals, being Φalloy). The Fermi level Φalloy in the alloy NPs is higher than that in pure palladium. Au NPs strongly absorb the visible light mainly
due to the LSPR effect and inter-band electron transitions, by which the conduction electrons gain the energy of light irradiation (more conduction
electrons distribute to high energy levels). The energetic conduction electrons in gold can migrate to the Pd sites on the surface. It follows that the
light flux to the NPs predominantly results in a surface (indicated by red colour) with high energy electrons. The surface Pd sites with energetic elec-
trons could exhibit significantly enhanced catalytic activity.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4279

View Article Online
Paper
Green Chemistry
irradiation. The apparent activation energy is ∼49.2 kJ mol−1
for the cross-coupling reaction in the dark, while it is ∼33.7 kJ
mol−1 for the photocatalytic reaction under visible light. The
activation energy of the photocatalytic cross-coupling reaction
is 15.5 kJ mol−1 lower as illustrated in Fig. 8. This further con-
firms that the enhanced charge heterogeneity is able to lower
the activation energy of the reaction and thus increase the
catalytic activity.
Leaching tests
Although various metal NPs have been applied in Suzuki reac-
tion catalysis, the nature of the true metal species in catalytic
cycles remains to be elucidated.8,43 It is debated if the catalysis
stems from the NPs themselves (in other words, heterogeneous
catalysis process), or from trace Pd leached into solution
(homogeneously catalysis process).16a It is a highly challenging
task to discriminate the actions of homogeneous Pd-complex
catalysts with single Pd atoms from soluble NP catalysts in the
present study of NP catalysis. This can be ascribed to the poss-
ible dynamic exchanges linking different types of metal
species in solution and the NPs acting as reservoirs.16a,44
There is support for both of these pathways because they may
be intertwined, since leaching may be assisted by one or
several steps of the catalytic cycle.44 Current studies employ
mainly the “filtrate transfer” test,45 elemental analysis of the
reaction solution, or both, to examine catalyst leaching.
Although there is no strict rule for the metal NP catalyzed reac-
tions, the absence of evidence against catalyst leaching may
lead to questioning of reaction heterogeneity as well as the
identity of actual catalytic species.46
Herein, a sequential reactivity study (“filtrate transfer” test)
was developed to pinpoint the main origin of the efficient cata-
lysis. Two identical catalytic reactions were initiated using the
normal procedure: the mixture for the reaction conducted
under visible light was labelled as A, and that for the dark reac-
tion was labelled as B. After 0.5 h, the reactions were inter-
rupted and the catalysts were separated from the solutions by
centrifugation. The clarified supernatant solution from the
photocatalytic reaction was divided into two equal parts to give
Fig. 7 Electron transfer from gold to palladium in the alloy NPs,
expressed as ΔN, varies with the composition of the alloy NPs (the
curve). ΔN reaches a maximum when the Au : Pd molar ratio is 1 : 1.62.
The Au : Pd molar ratio in the alloy NPs (horizontal axis) and photo-
catalytic conversions (red symbols) of the Suzuki cross-coupling reac-
tions in the present study (vertical axis on the right) are given
respectively, the reaction conversions were based on the average values
of 3 runs for each experiment.
These electronic properties lead to the strongest interaction
between the reactant and the metal NPs facilitating the
photocatalytic reaction. This could explain why the activity of
Alloy-1 is much higher than that of the catalysts with other
Au : Pd ratios.
We also carried out simulations using the density func-
tional theory (DFT) for electronic states with and without light
irradiation; the irradiation wavelength range between 532 and
535 nm was chosen, which is around the LSPR absorption of
Au. Calculation capacity limitations of our DFT simulation
necessitated the examination of a Pd32, Au32, and Au12Pd20
cluster. The Au : Pd ratio of the Au12Pd20 cluster is 1.67, close
to the ratio of 1 : 1.62 for the optimal Alloy-1 photocatalyst.
The detailed calculation method and the calculated Mulliken
charge distributions are given in ESI.† The DFT simulation
results confirm that charge heterogeneity exists even in the
monometallic Pd clusters and monometallic Au clusters, and
the alloy structure of Au and Pd increases the charge hetero-
geneity of the NP surface. This result is consistent with that of
the free electron-gas model analysis and previous reports.42
Furthermore, light irradiation can strongly promote the charge
heterogeneity in Au–Pd alloy NPs compared with pure Au NPs
(Fig. S4, ESI†).
From point of view of reaction kinetics, the visible light
absorption by the photocatalyst can contribute to reducing the
activation energy. Hence, the photocatalytic process has a
lower activation energy compared to the corresponding reac-
tion under heating (thermal process). In the present study, the
cross-coupling of 3-iodotoluene with phenylboronic acid was
used as the model reaction to investigate the kinetics at
different temperatures: 20, 30, 40 and 50 °C, and the Arrhenius
equation was applied to derive the apparent activation energies
of the reaction under light irradiation and thermal reaction
in the dark. The difference between the activation energy
Fig. 8 Apparent activation energy reduction of the Suzuki reaction
of the two processes indicates the contribution of the light caused by the light irradiation on the Au–Pd alloy NP photocatalyst.
4280 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 7 The study of the role of homogeneous Pd2+ ions in the photo-
catalysis process
solutions A1 and A2, and similarly the supernatant solution
from the dark reaction was divided into solutions B1 and B2.
Then A1 and B1 were re-irradiated for 2.5 h, A2 and B2 were
left in the dark to continue reaction for 2.5 h.
Reaction conditionsa
Conversiond (%)
PdCl2 dark reactionb
81.5
81.1
81.4
82.5
As shown in Fig. 9, the photocatalytic reaction achieved
96% conversion in our typical reaction, which is much higher
than those reactions interrupted and proceeding without cata-
lysts. This suggests that the heterogeneous Au–Pd alloy NPs
are essential in the photocatalytic reaction to achieve high
yield. On the other hand, we found that for both A and B
series solutions, the reaction still proceeded to some extent
after the solid catalyst was removed from the reaction mixture
regardless of whether the reaction was initially conducted in
the dark or under irradiation. The conversion achieved after
the removal of the solid catalyst, indicated by the increases in
conversion, is attributed to the trace component peeled off
from the catalyst during the first 0.5 h reaction. The greatest
conversion rate was achieved in A1. Nonetheless, the conver-
sion rates achieved by the peeled trace component in the
supernatant are much less than that of the typical reaction
under light irradiation. This indicates that the reaction is pre-
dominately catalysed by the supported alloy NPs.
In supernatant left in the dark, the reaction is catalysed by
the leached component or alloy NPs peeled off from the cata-
lyst (A2 and B2). The increases in the conversion rate of the
two supernatants are similar, being 17% (= 45%–28%) and
15% (= 25%–10%), respectively. Interestingly, the conversion
rates of the irradiated supernatants are higher than those of
the corresponding supernatants in the dark (comparisons of
B1 with B2 and A1 with A2). This fact suggests that light
irradiation can enhance the catalytic performance of the
leached component or peeled alloy NPs, as well.
PdCl2 photoreactionb
PdCl2 + Au-ZrO2 dark reactionc
PdCl2 + Au-ZrO2 photoreactionc
a General
reaction
conditions:
3-iodotoluene
(0.5
mmol),
phenylboronic acid (0.75 mmol), catalysts, K2CO3 (1.5 mmol), 10 mL
of the solvent DMF–H2O (V : V = 3 : 1), 30 °C, reaction time: 20 min.
b Catalyst: 0.5 mL PdCl2 (0.01 M) solution. c Catalysts: 0.5 mL PdCl2
(0.01 M) solution + 25 mg Au-ZrO2 (1.5% wt). d GC conversion: (aryl
iodide converted)/(initial amount of aryl iodide) × 100.
reactions were conducted both under light irradiation and in
the dark, and the conversions were nearly the same. Au-ZrO2
was also added to the PdCl2 catalysed reaction system, but it
did not affect the catalytic performance (Table 7). Hence, the
differences between the conversions of B1 and B2 (20%) and
between those of A1 and A2 (5%) are not contributed by Pd2+
ions, but by the peeled Au–Pd NPs in the supernatant. We
repeated the experiments but removed the solid catalyst after
reaction proceeded for one hour, and compared the results
with those shown above (see ESI Fig. S5†). The extension of
the initial reaction stage from 0.5 h to 1.0 h gave similar
results.
Reusability tests
The reusability of the Au–Pd alloy NP catalysts was examined
using bromobenzene as substrate under visible light
irradiation. The results with Alloy-1 are illustrated in Fig. 10.
The catalyst was used for five runs with each run condition
kept identical. After each run of the experiment, the catalyst
was separated by centrifugation, exhaustively washed with
water and ethanol twice, and then dried at 60 °C for reuse. The
results show that the catalyst can be reused without losing
activity significantly and the product selectivity can be main-
tained >99%. The Au–Pd alloy NPs are reusable photocatalysts
for Suzuki reactions. These results also suggest that the leach-
ing of the alloy NPs from the photocatalyst during the photo-
catalytic reactions should be negligible.
The trace component peeled off from the solid catalyst
could be fine alloy NPs or Pd2+ ions, but the centrifugation
could not separate the species from liquids. We investigated
the reaction using Pd2+ ions (Table 7). The PdCl2 catalysed
Mechanism
It is well known that light-excited electrons of plasmonic metal
NPs can populate unoccupied orbitals of the molecules
adsorbed on the NPs yielding transient anionic species.47 We
propose tentative reaction pathways for the photocatalytic
Suzuki reaction using the Au–Pd alloy NP photocatalyst. When
the alloy NPs are irradiated with light, the conduction elec-
trons are elevated into excited states (hot electrons) through
absorption of light energy. This increases the NPs’ ability to
induce reactions involving the adsorbed reactant molecules,
due to the enhanced surface charge heterogeneity. Several pre-
vious studies suggested that the heterogeneous metal NP’s
surface was the true active catalyst, and that activation of aryl
halides was possible due to synergistic anchimeric (adsorption
Fig. 9 The conversion rates of the supernatants without the solid cata-
lyst. The reaction conversions were based on the average values of 2
runs for each experiment.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4281

View Article Online
Paper
Green Chemistry
remaining steps may proceed following the well-accepted
mechanism.50 Reaction with the base gives the intermediate,
which via transmetalation with the boronate complex forms
diaryl organopalladium species. Finally, reductive elimination
of the desired product restores the initial alloy surface com-
pleting the photocatalysis cycle (Scheme 2B).
Conclusions
In summary, it is found that visible light can efficiently
enhance the performance of Au–Pd alloy NPs supported by
ZrO2 for Suzuki cross-coupling reactions at low temperatures.
The combination of light absorption of alloy NPs, the
enhanced interaction between the reactants and the NPs as
well as the intrinsic catalytic activity of the transition metal
leads to a unique structure where the absorption of visible
light can yield energetic electrons available at catalytically
active transition metal sites on the NP surface promoting the
reactions of the molecules adsorbed on the NPs. The photo-
catalytic Suzuki reactions achieved superior activity when the
ratio of the number of electrons from Au and Pd in alloy NPs
is nearly equal, and showed good feasibility on a range of sub-
strates. The dependence of photocatalytic activity on light
intensity and wavelength indicates that energetic electrons
excited by light absorption are responsible for the observed
Fig. 10 The reusability of the Au–Pd alloy NPs for the Suzuki reaction
using bromobenzene as the substrate under visible light irradiation.
Reaction conditions: bromobenzene (1 mmol), phenylboronic acid
(1 mmol), Alloy-1 photocatalysts (50 mg, containing 3% of metals),
NaOH (3 mmol), cetyltrimethylammonium bromide (CTAB) (1 mmol),
10 mL H2O, environment temperature 30 °C, reaction time 3 h, light
intensity 0.5 W cm−2
.
of the aryl moiety to the NP surface can serve as an anchor,
enhancing the chemical interaction of the carbon–halogen
bond with a separate but nearby active site) and electronic (the
adsorption of the aryl moiety can influence the electron
density of the catalytic surface, which can facilitate the
carbon–halogen bond activation) effects occurring in the pres-
ence of adsorbed species on the catalyst surface.44,48 The oxi-
dative addition is known as the rate-determining step in
Suzuki reactions, and it is said that this step in the heterocata-
lysis involves the electron transfer from the Pd atoms to the
halide atoms.21a Furthermore, the redox properties of tran-
sition metals are known to be influenced by irradiation with
light.49 Palladium(0) d10 complexes are relatively well studied
and are known to possess long-lived (triplet) excited states
both in solution and in the solid state. Due to their excited
state lifetimes in the microsecond range, these complexes can
undergo bimolecular photochemical reactions, including elec-
tron transfer and atom transfer reactions with halocarbons.49
Thus, they can readily react with halogenated hydrocarbons
and aromatic compounds under irradiation with visible light.
We therefore performed DFT calculations on the transfer of
light-excited electrons from the NP’s surface to the reactant
iodobenzene molecule (the detailed calculation methods and
results are given in ESI†). The simulation suggests that when
one electron enters an unoccupied orbital, the C–I bond will
elongate to 0.300 nm from 0.214 nm (Scheme 2A), so the clea-
vage of the C–I bond will be much easier. For the reactions in
which the rate-determining step is the C–I bond activation,
irradiation will facilitate electron transfer from the NP to the
adsorbed aryl iodide molecule, yielding a transient radical
anion. Loss of the iodide ion will afford either an adsorbed
phenyl radical or a true organometallic aryl-palladium iodide
complex on the surface. Once the reaction is initiated, all the
Scheme 2 (A) Light irradiation excites electrons of an alloy NP to high
energy levels, and the transfer of the excited electrons with sufficient
energy from the alloy NP to the LUMO of molecules adsorbed on the
NP can take place, activating the reaction. The electrons finally return to
the metal NP at a lower energy.47a DFT calculations show that the C–I
bond is elongated due to the electron transfer. (B) Proposed catalytic
cycle for Suzuki reactions using the Au–Pd alloy NPs under light
irradiation.
4282 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
photocatalytic activity. The results of both the free electron-gas and 40 mA) were used. DR-UV-vis spectra of the sample
model analysis and DFT simulation indicate that the Au–Pd powders were obtained using a Varian Cary 5000 spectrometer.
alloy nanostructure increases the charge heterogeneity of the
General procedure for photocatalytic reactions
NP surface, which enhances interaction between the alloy NPs
and the reactant molecules adsorbed on the NPs. A mechan-
ism is proposed based on the results of the DFT calculations
and experimental observation: light absorption of the alloy
NPs generates energetic electrons and the excited electrons
with sufficient energy are able to transfer to the reactant mole-
cules adsorbed on the NPs, weakening the chemical bonds of
the molecules and facilitating the reactions. Using NP photo-
catalysts and visible light to drive the synthesis of biaryl com-
A 25 mL Pyrex round bottom flask was used as the reaction
container, after adding reactants and the catalyst, the flask was
sealed with a rubber septum cap. The flask was then trans-
ferred into a reactor chamber and irradiated with magnetic
stirring using a halogen lamp (Nelson, wavelength in the range
400–750 nm) as the visible light source. The light intensity was
measured to be 0.5 W cm−2, light illuminance 94 000 lux, the
light intensity was kept constant in all photocatalytic processes
except for the experiments investigating the impact of inten-
sity. The light intensity can be adjusted by manipulating the
distance between the lamp and the reaction flask, a photo-
meter (TES 1332A) was used to measure the light intensity.
The temperature of the reaction system was carefully con-
trolled with an air conditioner attached to the reaction
chamber. The environmental temperature in the reactor
chamber was maintained at 30 °C. The solution temperature
pounds represents
a
new controlled, simplified, and
sustainable process in the realm of green chemistry. The
knowledge acquired in this study may inspire further studies
in new efficient photocatalysts of gold and other transition
metals for a wide range of organic syntheses driven by
sunlight.
Experimental section
Catalysts preparation
in the flask was measured to be 31
1.5 °C. At given
irradiation time intervals, 2 mL aliquots were collected, centri-
fuged, and then filtered through a Millipore filter (pore size
0.45 μm) to remove the catalyst particulates. The filtrates were
analyzed by an Agilent 6890 gas chromatograph with a HP-5
column. An Agilent HP5973 mass spectrometer was used to
determine and analyze the product compositions. For the reac-
tions using H2O as the solvent, the product was extracted with
dichloromethane (CH2Cl2) before GC analysis. All the dark
reactions were conducted using a water bath placed above a
magnetic stirrer to control the reaction temperature, identical
to the photocatalytic reactions; the reaction flask was wrapped
with aluminium-foil to prevent ingress of light.
Catalysts with 3 wt% of pure gold nanoparticles on ZrO2, 3 wt
% of pure palladium nanoparticles on ZrO2 and Au–Pd alloy
photocatalysts with different Au : Pd ratios on ZrO2 were pre-
pared by the impregnation–reduction method. For example,
Alloy-1 (1.5 wt% Au–1.5 wt% Pd supported on ZrO2) was pre-
pared by the following procedure: ZrO2 powder (2.0 g) was dis-
persed into HAuCl4 aqueous solution (0.01 M, 15.2 mL) and
PdCl2 aqueous solution (0.01 M, 28.3 mL) was added while
stirring magnetically. Aqueous lysine (0.53 M, 20 mL) was then
added with vigorous stirring, which was continued for 30 min.
To this suspension, aqueous NaBH4 solution (0.35 M, 10 mL)
was added dropwise over 20 min, followed by an addition of
hydrochloric acid (0.3 M, 10 mL). During the reduction
process, the white ZrO2 powder became black, then dark grey.
The mixture was allowed to stand for 24 h and then the solid
was separated by centrifugation, washed with water and
ethanol, and dried at 60 °C. The dried solid was used directly
as the catalyst. Catalysts with other Au : Pd ratios were prepared
by a similar method but using different quantities of the
HAuCl4 aqueous solution or the PdCl2 aqueous solution.
Acknowledgements
The authors gratefully acknowledge financial support from the
Australian Research Council (ARC DP110104990).
Notes and references
1 (a) A. de Meijere and F. Diederich, Metal-Catalyzed Cross-
Coupling Reactions, Wiley-VCH, Weinheim, 2004, vol. 1–2;
(b) G. C. Fu and A. F. Littke, Angew. Chem., Int. Ed., 2002,
41, 4176–4211.
2 (a) N. Miyaura, K. Yamada and A. Suzuki, Tetrahedron Lett.,
1979, 20, 3437–3440; (b) For the 2010 Nobel Prize,
Catalyst characterization
TEM study and line profile analysis by energy dispersion X-ray
spectrum technique of the photocatalysts were carried out on
a Philips CM200 TEM with an accelerating voltage of 200 kV.
The Au and Pd contents of the prepared catalysts were deter-
mined by EDS technology using the attachment to a FEI
Quanta 200 Environmental SEM. The element line scanning
see:
http://www.nobelprize.org/nobel_prizes/chemistry/
laureates/2010/.
was conducted on
a
Bruker EDX scanner attached to
3 S. Kotha and K. Lahiri, Eur. J. Org. Chem., 2007, 8,
1221–1236.
4 A. Fihri, P. Meunier and J. C. Hierso, Coord. Chem. Rev.,
2007, 251, 2017–2055.
JEOL-2200FS TEM with a scanning beam diameter of 1.0 nm.
X-ray diffraction (XRD) patterns of the sample powders were
collected using a Philips PANalytical X’pert Pro diffractometer.
CuKα radiation (λ = 1.5418 Å) and a fixed power source (40 kV
5 C. J. Li, Chem. Rev., 2005, 105, 3095–3166.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4283

View Article Online
Paper
Green Chemistry
6 A. F. Littke, C. Dai and G. C. Fu, J. Am. Chem. Soc., 2000,
122, 4020–4028.
7 (a) S. L. Buchwald, Acc. Chem. Res., 2008, 41, 1439;
(b) L. Yin and J. Liebscher, Chem. Rev., 2007, 107, 133–173;
(c) N. Miyaura and A. Suzuki, Chem. Rev., 1995, 95, 2457–
2483.
2011, 47, 8679–8681; (f) P. V. Pham, D. A. Nagib and
D. W. C. MacMillan, Angew. Chem., Int. Ed., 2011, 50, 6119–
6122; (g) C. Dai, J. M. R. Narayanam and
C. R. J. Stephenson, Nat. Chem., 2011, 3, 140–145;
(h) R. S. Andrews, J. J. Becker and M. R. Gagné, Org. Lett.,
2011, 13, 2406–2409.
8 A. Fihri, M. Bouhrara, B. Nekoueishahraki, J. M. Basset and 19 (a) X. Chen, H. Y. Zhu, J. C. Zhao, Z. F. Zheng and
V. Polshettiwar, Chem. Soc. Rev., 2011, 40, 5181–5203.
9 (a) P. Han, X. Wang, X. Qiu, X. Ji and L. Gao, J. Mol. Catal.
A: Chem., 2007, 272, 136–141; (b) Z. Zheng, H. Li, T. Liu
and R. Cao, J. Catal., 2010, 270, 268–274; (c) D. D. Das and
A. Sayari, J. Catal., 2007, 246, 60–65.
X. P. Gao, Angew. Chem., Int. Ed., 2008, 47, 5353–5356;
(b) H. Y. Zhu, X. B. Ke, X. Z. Yang, S. Sarina and H. W. Liu,
Angew. Chem., Int. Ed., 2010, 49, 9657–9661; (c) S. Sarina,
H. Y. Zhu, Z. F. Zheng, S. Bottle, J. Chang, X. B. Ke,
J. C. Zhao, Y. N. Huang, A. Sutrisno, M. Willans and
G. R. Li, Chem. Sci., 2012, 3, 2138–2146.
10 V. I. Sokolov, E. G. Rakov, N. A. Bumagin and
M. G. Vinogradov, Fullerenes, Nanotubes, Carbon Nanos- 20 S. Sarina, H. Zhu, E. Jaatinen, Q. Xiao, H. Liu, J. Jia,
truct., 2010, 18, 558–563.
C. Chen and J. Zhao, J. Am. Chem. Soc., 2013, 135, 5793–
11 (a) R. Najman, J. K. Cho, A. F. Coffey, J. W. Davies and
5801.
M. Bradley, Chem. Commun., 2007, 5031–5033; 21 (a) F. Wang, C. Li, H. Chen, R. Jiang, L. D. Sun, Q. Li,
(b) I. P. Beletskaya, A. N. Kashin, I. A. Khotina and
A. R. Khokhlov, Synlett, 2008, 1547–1552; (c) G. Wei,
W. Zhang, F. Wen, Y. Wang and M. Zhang, J. Phys. Chem. C,
2008, 112, 10827–10832.
J. Wang, J. C. Yu and C. H. Yan, J. Am. Chem. Soc., 2013,
135, 5588–5601; (b) X. Huang, Y. Li, Y. Chen, H. Zhou,
X. Duan and Y. Huang, Angew. Chem., Int. Ed., 2013, 52,
6063–6067.
12 (a) M. L. Kantam, S. Roy, M. Roy, B. Sreedhar and 22 (a) P. Mulvaney, Langmuir, 1996, 12, 788–800; (b) S. Eustis
B. M. Choudary, Adv. Synth. Catal., 2005, 347, 2002–2008;
(b) A. Gniewek, J. J. Ziolkowski, A. M. Trzeciak,
and M. A. El-Sayed, Chem. Soc. Rev., 2006, 35, 209–217;
(c) P. V. Kamat, J. Phys. Chem. B, 2002, 106, 7729–7744.
M. Zawadzki, H. Grabowska and J. Wrzyszcz, J. Catal., 2008, 23 (a) K. Patel, S. Kapoor, D. P. Dave and T. Mukherjee, Res.
254, 121–130.
Chem. Intermed., 2006, 32, 103–113; (b) S. Sarina, H. Y. Zhu,
Q. Xiao, E. Jaatinen, J. Jia, Y. Huang, Z. Zheng and H. Wu,
Angew. Chem., Int. Ed., 2014, 53, 2935–2940.
13 (a) M. L. Kantam, S. Roy, M. Roy, B. Sreedhar,
B. M. Choudary and R. L. De, J. Mol. Catal. A: Chem., 2007,
273, 26–31; (b) S. Y. Liu, Q. Z. Zhou, Z. N. Jin, H. J. Jiang 24 Y. Li and M. A. El-Sayed, J. Phys. Chem. B, 2001, 105, 8938–
and X. Z. Jiang, Chin. J. Catal., 2010, 31, 557–561. 8943.
14 (a) K. R. Gopidas, J. K. Whitesell and M. A. Fox, Nano Lett., 25 M. Pittelkow, K. Moth-Poulsen, U. Boas and
2003, 3, 1757–1760; (b) L. Wu, B. L. Li, Y. Y. Huang, J. B. Christensen, Langmuir, 2003, 19, 7682–7684.
H. F. Zhou, Y. M. He and Q. H. Fan, Org. Lett., 2006, 8, 26 J. D. Senra, L. F. B. Malta, M. E. H. M. da Costa,
3605–3608.
R. C. Michel, L. C. S. Aguiar, A. B. C. Simas and
15 (a) S. E. Garcia-Garrido, J. Francos, V. Cadierno,
O. A. C. Antunes, Adv. Synth. Catal., 2009, 351, 2411–2422.
J. M. Basset and V. Polshettiwar, ChemSusChem, 2011, 4, 27 S. L. Huang, A. Q. Jia and G. X. Jin, Chem. Commun., 2013,
104–111; (b) V. Polshettiwar and R. S. Varma, Tetrahedron,
2010, 66, 1091–1097.
16 (a) P. P. Fang, A. Jutand, Z. Q. Tian and C. Amatore, Angew.
49, 2403–2405.
28 L. Zhang, L. Wang, H. Li and P. Li, Synth. Commun., 2008,
38, 1498–1511.
Chem.,
(b) T. S. A. Heugebaert, S. De Corte, T. Sabbe, T. Hennebel,
W. Verstraete, N. Boon and C. V. Stevens, Tetrahedron Lett., 30 S. M. Islam, P. Mondal, K. Tuhina, A. S. Roy, S. Mondal
Int.
Ed.,
2011,
50,
12184–12188; 29 H. B. Pan, C. H. Yen, B. Yoon, M. Sato and C. M. Wai,
Synth. Commun., 2006, 36, 3473–3478.
2012, 53, 1410–1412.
17 (a) G. Palmisano, V. Augugliaro, M. Pagliaro and
and D. Hossain, J. Inorg. Organomet. Polym., 2010, 20,
264–277.
L. Palmisano, Chem. Commun., 2007, 3425–3437; 31 V. Chandrasekhar, R. S. Narayanan and P. Thilagar, Organo-
(b) M. Fagnoni, D. Dondi, D. Ravelli and A. Albini, Chem. metallics, 2009, 28, 5883–5888.
Rev., 2007, 107, 2725–2756; (c) S. Navalon, R. Martin, 32 N. Hoshiya, N. Isomura, M. Shimoda, H. Yoshikawa,
M. Alvaro and H. Garcia, ChemSusChem, 2011, 4, 650–657.
18 (a) D. A. Nicewicz and D. W. C. MacMillan, Science, 2008,
Y. Yamashita, K. Iizuka, S. Tsukamoto, S. Shuto and
M. Arisawa, ChemCatChem, 2009, 1, 279–285.
322, 77–80; (b) D. A. Nagib, M. E. Scott and 33 F. Lu, J. Ruiz and D. Astruc, Tetrahedron Lett., 2004, 45,
D. W. C. MacMillan, J. Am. Chem. Soc., 2009, 131, 10875– 9443–9445.
10877; (c) H. W. Shih, M. N. Vander Wal, R. L. Grange and 34 (a) L. F. Liu, Y. H. Zhang and B. W. Xin, J. Org. Chem., 2006,
D. W. C. MacMillan, J. Am. Chem. Soc., 2010, 132, 13600–
13603; (d) M. Rueping, C. Vila, R. M. Koenigs, K. Poscharny
71, 3994–3997; (b) Y. Li, X. M. Hong, D. M. Collard and
M. A. El-Sayed, Org. Lett., 2000, 2, 2385–2388.
and D. C. Fabry, Chem. Commun., 2011, 47, 2360–2362; 35 S. Sarina, E. R. Waclawik and H. Zhu, Green Chem., 2013,
(e) M. Rueping, S. Zhu and R. M. Koenigs, Chem. Commun.,
15, 1814–1833.
4284 | Green Chem., 2014, 16, 4272–4285
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
36 (a) A. Tanaka, S. Sakaguchi, K. Hashimoto and
H. Kominami, ACS Catal., 2013, 3, 79–85; (b) E. Kowalska,
R. Abea and B. Ohtania, Chem. Commun., 2009, 241–243.
37 P. Christopher, H. L. Xin and S. Linic, Nat. Chem., 2011, 3,
467–472.
catalyst (usually by filtration or centrifugation); (3) treat
the filtrate under normal reaction conditions without any
addition of heterogeneous catalyst; and (4) if the reactant
conversion in the filtrate remains unchanged over the
usual reaction completion time, then the active catalyst
is considered to be heterogeneous with minimal
leaching.
38 M. S. Chen and D. W. Goodman, Science, 2004, 306,
252–255.
39 Handbook of Chemistry and Physics, 91st edn, 2010–2011, 46 H. Cong and J. A. Porco Jr., ACS Catal., 2012, 2, 65–70.
section 12–114.
47 (a) L. Brus, Acc. Chem. Res., 2008, 41, 1742; (b) M. Turner,
V. B. Golovko, O. P. Vaughan, P. Abdulkin, A. Berenguer-
Murcia, M. S. Tikhov, B. F. Johnson and R. M. Lambert,
Nature, 2008, 454, 981; (c) P. Christopher, H. L. Xin,
A. Marimuthu and S. Linic, Nat. Mater., 2012, 11, 1044.
40 C. S. Cho, W. X. Ren and S. C. Shim, Bull. Korean Chem.
Soc., 2005, 26, 1611–1613.
41 K. Yamada, K. Miyajima and F. Mafun, J. Phys. Chem. C,
2007, 111, 11246–11251.
42 (a) A. Kotsifa, T. I. Halkides, D. I. Kondarides and 48 C. R. LeBlond, A. T. Andrews, Y. K. Sun and J. R. Sowa, Org.
X. E. Verykios, Catal. Lett., 2002, 79, 113; (b) P. P. Fang, Lett., 2001, 3, 1555–1557.
J. F. Li, X. D. Lin, J. R. Anema, D. Y. Wu, B. Ren and 49 D. M. Roundhill, Photochemistry and photophysics of metal
Z. Q. Tian, J. Electroanal. Chem., 2012, 665, 70. complexes, Plenum Press, New York, 1994.
43 J. A. Widegren and R. G. Finke, J. Mol. Catal. A: Chem., 50 (a) M. B. Thathagar, J. E. ten Elshof and G. Rothenberg,
2003, 198, 317–341.
Angew. Chem., Int. Ed., 2006, 45, 2886–2890;
(b) A. V. Gaikwad, A. Holyuigre, M. B. Thathagar, J. E. ten
Elshof and G. Rothenberg, Chem. – Eur. J., 2007, 13, 6908–
6913; (c) S. MacQuarrie, J. H. Horton, J. Barnes,
K. McEleney, H. P. Loock and C. M. Crudden, Angew.
Chem., Int. Ed., 2008, 47, 3279–3282; (d) Z. Niu, Q. Peng,
Z. Zhuang, W. He and Y. Li, Chem. – Eur. J., 2012, 18, 9813–
9817.
44 (a) R. G. Finke and S. Ozkar, Coord. Chem. Rev., 2004, 248,
135–146; (b) N. T. S. Phan, M. Van Der Sluys and
C. W. Jones, Adv. Synth. Catal., 2006, 348, 609–679.
45 The “filtrate transfer” test refers to the following experi-
mental conditions: (1) initiate a nanoparticle-catalyzed
reaction under normal conditions; (2) interrupt the reac-
tion before completion by removing the heterogeneous
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 4272–4285 | 4285