Palladium nanoparticles bonded to two-dimensional iron oxide graphene nanosheets: A synergistic and highly reusable catalyst for the Tsuji-Trost reaction in water and air

Article abstract of DOI:10.1002/chem.201402545

Low cost, high activity and selectivity, convenient separation, and increased reusability are the main requirements for noble-metal-nanocatalyst- catalyzed reactions. Despite tremendous efforts, developing noble-metal nanocatalysts to meet the above requirements remains a significant challenge. Here we present a general strategy for the preparation of strongly coupled Fe₃O₄ and palladium nanoparticles (PdNPs) to graphene sheets by employing polyethyleneimine as the coupling linker. Transmission electron microscopic images show that Pd and Fe₃O₄ nanoparticles are highly dispersed on the graphene surface, and the mean particle size of Pd is around 3 nm. This nanocatalyst exhibits synergistic catalysis by Pd nanoparticles supported on reduced graphene oxide (rGO) and a tertiary amine of polyethyleneimine (Pd/Fe₃O₄/PEI/rGO) for the Tsuji-Trost reaction in water and air. For example, the reaction of ethyl acetoacetate with allyl ethyl carbonate afforded the allylated product in more than 99% isolated yield, and the turnover frequency reached 2200 h^-1. The yield of allylated products was 66% for Pd/rGO without polyethyleneimine. The catalyst could be readily recycled by a magnet and reused more than 30 times without appreciable loss of activity. In addition, only about 7.5% of Pd species leached off after 20 cycles, thus rendering this catalyst safer for the environment. Picking up speed: A multifunctional Pd nanocatalyst was synthesized by in situ growth of palladium nanoparticles (PdNPs) and the assembly of Fe₃O₄ NPs on reduced graphene oxide (rGO) by employing polyethyleneimine as the coupling linker. This nanocatalyst exhibits synergistic catalysis by the amine on the same rGO surface as the PdNPs for acceleration of the Tsuji-Trost reaction in water and air (see figure).

Full text of DOI:10.1002/chem.201402545

DOI: 10.1002/chem.201402545
Full Paper
&
Synergistic Catalysis
Palladium Nanoparticles Bonded to Two-Dimensional Iron Oxide
Graphene Nanosheets: A Synergistic and Highly Reusable Catalyst
for the Tsuji–Trost Reaction in Water and Air
Jian Liu, Xing Huo, Tianrong Li, Zhengyin Yang, Pinxian Xi, Zhiyi Wang, and Baodui Wang*[a]
Abstract: Low cost, high activity and selectivity, convenient
separation, and increased reusability are the main require-
ments for noble-metal-nanocatalyst-catalyzed reactions. De-
spite tremendous efforts, developing noble-metal nanocata-
lysts to meet the above requirements remains a significant
challenge. Here we present a general strategy for the prepa-
ration of strongly coupled Fe3O4 and palladium nanoparti-
cles (PdNPs) to graphene sheets by employing polyethylene-
imine as the coupling linker. Transmission electron micro-
scopic images show that Pd and Fe3O4 nanoparticles are
highly dispersed on the graphene surface, and the mean
particle size of Pd is around 3 nm. This nanocatalyst exhibits
synergistic catalysis by Pd nanoparticles supported on re-
duced graphene oxide (rGO) and a tertiary amine of polyeth-
yleneimine (Pd/Fe3O4/PEI/rGO) for the Tsuji–Trost reaction in
water and air. For example, the reaction of ethyl acetoace-
tate with allyl ethyl carbonate afforded the allylated product
in more than 99% isolated yield, and the turnover frequency
reached 2200 hꢀ1. The yield of allylated products was 66%
for Pd/rGO without polyethyleneimine. The catalyst could be
readily recycled by a magnet and reused more than 30 times
without appreciable loss of activity. In addition, only about
7.5% of Pd species leached off after 20 cycles, thus render-
ing this catalyst safer for the environment.
Introduction
protocol to minimize aggregation and maximize the chemically
accessible area. Despite numerous efforts to use magnetic NPs
and graphene as supports for NMNPs for catalytic applications,
an easily controlled architecture of magnetic NPs/graphene
hybrid materials as a magnetically recoverable catalyst support
on which NMNPs are covalently bonded by using a stabilizer
and that exhibit high dispersion has not been synthesized and
studied. Such ternary conjugates should serve as new multi-
functional heterogeneous catalysts to enhance activity and sta-
bility, and accordingly, reduce their cost and minimize environ-
mental contamination.
Currently the design of multifunctional heterogeneous metal
catalysts that preserve high metal dispersion and stability
during the catalytic reactions and easily separate from the re-
action mixture are of great importance. Noble-metal nanoparti-
cles (NMNPs) with ultrafine size have attracted more and more
attention owing to their superior catalytic activity in carbon–
carbon cross-coupling, water splitting, methanol oxidation, and
so forth.[1] However, NMNPs with small particle size in the reac-
tion conditions tend to aggregate and lose their catalytic activ-
ities.[2] Moreover, the high cost and limited resources also
hinder their practical application. To solve these problems, the
particles are often immobilized on various solid supports, such
as carbon materials, metal oxides, polymers, and so on.[3]
Among all solid supports studied thus far, magnetic nanoparti-
cles[4] and graphene[5] have attracted particular interest be-
cause magnetic NPs have efficient magnetic separation capaci-
ty for easy separation of the catalyst from the reaction solution
and graphene has a high surface area, which offers an effective
Given the rise of green chemistry, organic reactions conduct-
ed in water have received considerable attention since water is
nontoxic, nonflammable, and the cheapest solvent in the
world.[6] Palladium allylic substitution, also called the Tsuji–
Trost reaction, is a reliable and widely used method for carry-
ing out CꢀC and CꢀO bond formation in organic syntheses.[7]
Usually, these reactions are carried out in organic and often
toxic solvents. However, only a limited variety of supported
palladium catalysts have been reported up to now for aqueous
heterogeneous Tsuji–Trost reactions.[8] The catalysts reported
so far are recovered by tedious filtration or centrifugation, and
the recycling of the catalysts has not been a consideration.
Therefore, more efficient systems are still in demand.
[a] J. Liu, Prof. X. Huo, Dr. T. R. Li, Prof. Z. Y. Yang, Prof. P. X. Xi, Z. Y. Wang,
Prof. B. D. Wang
Key Laboratory of Nonferrous Metal Chemistry and
Resources Utilization of Gansu Province and
State Key Laboratory of Applied Organic Chemistry
Lanzhou University Gansu, Lanzhou, 730000 (P.R. China)
Fax: (+86)931-8912582
In this study, we demonstrate an efficient and universal strat-
egy for the in situ growth of palladium nanoparticles (PdNPs)
and the assembly of Fe3O4 NPs on reduced graphene oxide
(rGO) by employing polyetheylenimine (PEI) as the coupling
linker among the three components. In these nanocomposites,
GO could provide a large surface area on which the active
E-mail: wangbd@lzu.edu.cn
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402545.
Chem. Eur. J. 2014, 20, 11549 – 11555
11549
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
phase is finely dispersed and covalently anchored. These nano-
composites have a well-defined 2D morphology, confined
PdNPs and Fe3O4 NPs in the PEI, controllable particle size, and
high specific surface areas. Inspired by the monodisperse, ul-
trafine, and pristine properties of PdNPs on the same rGO sur-
face as the tertiary amine of polyethyleneimine (Pd/Fe3O4/PEI/
rGO), we sought to determine whether Pd/Fe3O4/PEI/rGO ex-
hibits synergistic catalysis for the
PEI/rGO nanocomposites, the PdꢀN absorption bands ap-
peared at 624 cmꢀ1, and FeꢀO absorption bands appeared at
594 and 445 cmꢀ1, respectively.[11b] The above characteristics
were also found for Pd/DIB/PEI/rGO and Fe3O4/DIB/PEI/rGO, re-
spectively. These indicate that PdNPs and Fe3O4 NPs are conju-
gated by means of the DIB/PEI/rGO. Figure 1b–e shows the
TEM images of the PEI/rGO, Pd/PEI/rGO, and Pd/Fe3O4/PEI/rGO
acceleration of the Tsuji–Trost re-
action in water and air. The Pd/
Fe3O4/PEI/rGO catalyst could be
easily separated from the reac-
tant with the simple application
of an external magnetic field,
and its catalytic efficiency was
still high even after recycling
30 times in water.
Results and Discussion
Particle preparation and char-
acterization
The overall synthetic strategy of
Pd/Fe3O4/PEI/rGO nanocompo-
sites is illustrated in Figure 1a.
First, PEI was chemically modi-
fied on the GO surface by amide
bonds, which led to the forma-
tion of PEI-grafted rGO (PEI/rGO).
Second, the resulting PEI/rGO
nanosheets were treated with
3,4-dihydroxybenzaldehyde (DIB)
to introduce DIB onto rGO for
the coordination assembly of
Fe3O4 NPs.[9] The resulting prod-
uct was stirred with PdCl2 in a so-
lution of ethanol for two hours
to ensure that the metal ion
thoroughly bonded to the N
atoms of the PEI. NaBH4 was
then added and stirred for one
hour to generate PdNPs in
situ.[10] Finally, the obtained Pd/
PEI/rGO composites were react-
ed with the as-synthesized Fe3O4
NPs to form Pd/Fe3O4/PEI/rGO.
Surface catechol groups and
amino groups therefore en-
hanced the stability and dispersi-
ty of Fe3O4 NPs and PdNPs on
rGO, respectively.
The composition of Pd/Fe3O4/
PEI/rGO nanocomposites was
confirmed by FTIR. (Figure S2 in
Figure 1. a) Synthetic route of Pd/Fe3O4/PEI/rGO. TEM images of b) PEI/rGO, c) Pd/PEI/rGO, and d) Pd/Fe3O4/PEI/
rGO. The inset shows the nanoparticle size distribution of Pd at rGO. e) HRTEM image and the SAED pattern of
the Supporting Information). In
the FTIR spectrum of Pd/Fe3O4/ Pd/Fe3O4/PEI/rGO nanocomposites.
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org 11550
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
nanohybrids. The monodisperse Fe3O4 NPs and the relatively
uniform PdNPs were well dispersed on the surface of the rGO
without clear aggregation. The morphology of Fe3O4 NPs was
also unchanged after chemical modification onto rGO. The
sizes of Pd particles are uniform, and the mean diameters are
(4.31ꢁ0.40) nm (inset of Figure 1d). The high-resolution TEM
(HRTEM) image (Figure 1e) reveals that the Pd nanocrystals
and Fe3O4 NPs are spherical with resolved lattice fringes.[2,11]
The selected-area electron diffraction (SAED) pattern (inset of
Figure 1e) also confirms the presence of PdNPs and Fe3O4 NPs.
The typical energy-dispersive X-ray (EDX) pattern (Figure S3 in
the Supporting Information) indicates that the Pd/Fe3O4/PEI/
rGO composites contain Pd, Fe, O, N, and C, whereas all of the
elemental Cu and some of the C come from the carbon-coated
copper grids. Inductively coupled plasma (ICP) results indicate
that the loading capacities of the Pd on the magnetic support
are 1.93 wt%, and the ratio of amine groups to Pd is 66:1.
Scanning TEM and an elemental mapping investigation of
Pd/Fe3O4/PEI/rGO were performed to illustrate the distribution
of carbon, nitrogen, palladium, iron, and oxygen components
in the hybrids (Figure S4c–g in the Supporting Information).
The C distribution appears to be uniform across the 2D nano-
sheets. The N and O distributions are similar to the C distribu-
tion but with a weaker energy-dispersive X-ray spectroscopy
(EDS) signal. In contrast, the palladium and iron species are
only identified within the nanometer-sized domains, which are
homogeneously dispersed on the 2D nanosheets. These results
imply that PdNPs and Fe3O4 NPs are formed in the polymer
and not on the GO during the in situ growth and assembly,[12]
respectively, and the four components (PEI, Pd, Fe3O4, and gra-
phene) are strongly coupled together in the nanocomposites.
The Pd/Fe3O4/PEI/rGO nanocomposites were also character-
ized by X-ray photoelectron spectroscopy (XPS) and powder X-
ray diffraction (XRD) analysis. The sharp N and O peaks in Fig-
ure 2a and c demonstrate the presence of PEI, Fe3O4, and DIB
in the Pd/Fe3O4/PEI/rGO nanocomposites. The XPS spectrum of
Pd 3d can be fitted into a main doublet peak, as shown in Fig-
ure 2b. The binding energy of the doublet peaks at 335.4 eV
(assigned to Pd0 3d5/2) and 340.7 eV (assigned to Pd0 3d3/2) can
be attributed to the Pd0 state.[2,13] Similarly, the XPS spectrum
of the Fe 2p region (Figure 2c) clearly indicates that Fe3O4 NPs
also exist on the PEI/rGO nanocomposites.[11] Figure 2d shows
the XRD pattern of the as-synthesized Pd/Fe3O4/PEI/rGO nano-
composites. The peaks located at 29.91, 35.41, 43.11, 57.21,
and 62.818 (2q) indicate the presence of Fe3O4.[11] The weak
peak appears around 2q=248, which is the characteristic peak
of residual unoxidized graphite.[13] Moreover, new peaks that
correspond to the (111) reflection of Pd crystal are observed
for Pd/Fe3O4/PEI/rGO, thus reconfirming the successful growth
of metal particles on the surface of magnetic hybrids again.[2,3g]
The magnetic properties of the obtained samples were in-
vestigated by using a vibrating sample magnetometer (VSM)
system at room temperature. The magnetization curves in Fig-
ure 2e show that Pd/Fe3O4/PEI/rGO nanocomposites are super-
paramagnetic at room temperature. The saturation magnetiza-
tion values for Fe3O4 NPs and Pd/Fe3O4/PEI/rGO are 48.2 and
18.4 emugꢀ1, respectively. The decrease in saturation magneti-
Figure 2. a) XPS spectra of Pd/Fe3O4/PEI/rGO. b) Spectrum in the Pd 3d
region. c) Spectrum in the Fe 2p region. d) XRD pattern of Pd/Fe3O4/PEI/rGO.
e) Magnetic behavior of Fe3O4 NPs and Pd/Fe3O4/PEI/rGO at 300 K. The pho-
tographs demonstrate that Pd/Fe3O4/PEI/rGO in aqueous solution can be at-
tracted and arranged vertically by a magnetic bar.
zation is caused by the increasing amount of noble metal, PEI,
and rGO in the hybrids. The photographs in Figure 2e demon-
strate that the Pd/Fe3O4/PEI/rGO composites in aqueous solu-
tion can be harvested and separated by a magnetic field,
which is important for the magnetic separation and reusability
of the Pd/Fe3O4/PEI/rGO composites from the reaction mixture.
The Tsuji–Trost allylation of 1,3-dicarbonyl compounds with
allyl ethyl carbonate
Recent societal demands to develop sustainable chemistry
have increasingly led to chemists considering water as
a benign solvent. Given that the Pd/Fe3O4/PEI/rGO catalyst is
highly dispersed in water, we performed the Tsuji–Trost reac-
tion in water and air. In a preliminary catalysis study, ethyl ace-
toacetate 1 as a model substrate with the allyl ethyl carbonate
2 as the nucleophile was chosen to optimize the reaction con-
ditions (Table S1 in the Supporting Information). Herein, Pd/
Fe3O4/PEI/rGO catalyst containing 1.8 mmol Pd was used in all
the reactions if not stated otherwise. We screened various
amounts of PPh3 and found that 7.2 mmol PPh3 (four times the
amount of Pd) gave the corresponding diallylated products 4
in more than 99% yield in one hour (Table S1 of the Support-
ing Information, entry 3). A decrease in the temperature from
100 to 308C resulted in a lower yield (<10%) (Table S1 of the
Supporting Information, entry 6). When PPh3 was omitted, the
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org
11551
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
allylic alkylation yield was less than 41% (Table S1 of the Sup-
porting Information, entry 1). When K2CO3 was used in this
system in the absence of PPh3, 73% isolated yield was ob-
tained (Table S1 of the Supporting Information, entry 8). How-
ever, in the presence of both K2CO3 and PPh3, the allylated
product yield was 79% (Table S1 of the Supporting Informa-
tion, entry 9). The low yield might be due to hydrolysis of ethyl
acetoacetate 1 in basic conditions. Thus, Tsuji–Trost reactions
were performed in water and air in the presence Pd/Fe3O4/PEI/
rGO catalyst that contained 1.8 mmol Pd and 7.2 mmol PPh3.
Synergistic catalysis of a Pd species and base on a nonstruc-
tured solid surface has been recently reported in the Tsuji–
Trost reaction.[14] To test the potency of Pd/Fe3O4/PEI/rGO as
a synergistic catalyst, we examined the Tsuji–Trost reaction be-
tween ethyl acetoacetate 1 and allyl ethyl carbonate 2 in
water under optimized conditions. As shown in
Table 1. Tsuji–Trost reaction using various catalysts in H2O and air.[a]
[b]
Entry Catalyst
t [h] Yield [%] TOF [hꢀ1
]
3/4 [%][c]
1
2
3
4
Pd/Fe3O4/PEI/rGO
0.5 >99
>2200
733
–
40:60
30:70
–
Pd/rGO
1.0
1.0
66
–
Fe3O4/PEI/rGO[d]
Fe3O4/PEI/rGO[d] +Pd/rGO 1.0
73
811
15:85
[a] Reaction conditions: Pd catalyst (Pd: 1.8 mmol), PPh3 (7.2 mmol),
n
1 (2.0 mmol), 2 (5.0 mmol), H2O (3.0 mL), 1008C. [b] TOF [hꢀ1]=n
.
ðproductÞ
ðPdÞ ꢂt ½hꢃ
[c] Yields of isolated products determined by 1H and 13C NMR spectrosco-
py. Tetramethylsilane was used as internal standard. [d] PEI/Fe3O4/rGO
(Fe: 0.045 mmol) was used.
Table 1, the use of Pd/Fe3O4/PEI/rGO resulted in more
Table 2. The Tsuji–Trost reaction of various 1,3-dicarbonyl compounds with allyl ethyl
than 99% yield of the total allylated products in half
an hour (Table 1, entry 1). The yield of allylated prod-
ucts was 66% for Pd/rGO (the size of Pd is about
4 nm, which is close to the size of Pd on the Pd/
Fe3O4/PEI/rGO; Figure S5b in the Supporting Informa-
tion) without PEI (Table 1, entry 2). The Tsuji–Trost re-
action that used Fe3O4/PEI/rGO (Figure S5c in the
Supporting Information) did not proceed (Table 1,
entry 3). These results indicate that the tertiary amine
on Pd/Fe3O4/PEI/rGO accelerates the Pd-catalyzed
Tsuji–Trost reaction. To further prove this conclusion,
the reaction with PEI/rGO was conducted by using
Pd/rGO (Table 1, entry 4). The product yield was
much lower than that for the reaction that used Pd/
Fe3O4/PEI/rGO (Table 1, entry 1).
carbonate in H2O and air.[a]
Entry
1
1,3-Dicarbonyl
Catalyst
t [h]
Mono/Di
35:65
Yield [%][b]
Pd/Fe3O4/PEI/rGO
0.5
>99
2
3
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.0
1.0
0:100
13:87
>99
68
4
Pd/Fe3O4/PEI/rGO
0.5
40:60
>99
5
6
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.0
1.0
0:100
30:70
>99
66
7
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.0
1.0
1.0
1.0
0:100
2:98
>99
59
8
9
Pd/Fe3O4/PEI/rGO
Pd/rGO
0:100
13:87
>99
26
10
11
Pd/Fe3O4/PEI/rGO
1.0
30:70
>99
To further explore the scope of the synthetic ap-
plicability of the Pd/Fe3O4/PEI/rGO catalyst, other al-
lylic substitution reactions were also examined in
water and air under the optimized conditions. As
shown in Table 2, nearly all of the allylic substitution
reactions finished within three hours. In the reaction
of 1,3-dicarbonyls, the results for Pd/rGO without PEI
are also shown (Table 2). The Pd/Fe3O4/PEI/rGO cata-
lyst showed good activity for the reaction of 1,3-die-
ster (Table 2, entries 19 and 21) and 1,3-diketone
(Table 2, entries 23 and 25). Bulky alkyl esters, such as
isopropyl acetoacetate, tert-butyl acetoacetate, and
isobutyl acetoacetate, were also suitable substrates
for this allylic alkylation process; they afforded the
corresponding products in greater than 99% yields
(Table 2, entries 7, 9, and 11). The high product yields
indicate that the Pd/Fe3O4/PEI/rGO catalyst exhibits
high catalytic activity for the bulky ester. For the
phenyl keto ester, 81% yield of products was ob-
tained (Table 2, entry 17). Interestingly, the ethyl-2-
methyl-3-oxobutanoate with steric bulk also gave
more than 99% product (Table 2, entry 15). For all
1,3-dicarbonyls used for comparison, Pd/Fe3O4/PEI/
rGO showed higher product yields than Pd/rGO as
well as other reported catalytic systems.[14b]
12
13
14
Pd/rGO
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.0
1.0
1.0
13:87
42:58
2:98
53
94
26
15
16
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.5
1.5
–
–
>99
31
17
Pd/Fe3O4/PEI/rGO
4.0
0:100
81
18
19
20
21
22
Pd/rGO
Pd/Fe3O4/PEI/rGO
Pd/rGO
4.0
3.0
3.0
3.0
3.0
0:100
42:58
2:98
43
91
28
88
26
Pd/Fe3O4/PEI/rGO
Pd/rGO
42:58
2:98
23
24
25
26
Pd/Fe3O4/PEI/rGO
Pd/rGO
1.0
1.0
1.5
1.5
0:100
0:100
–
>99
43
Pd/Fe3O4/PEI/rGO
Pd/rGO
83
–
27
[a] Reaction conditions: Pd catalyst (Pd: 1.8 mmol), PPh3 (7.2 mmol), 1 (2.0 mmol), 2
(5.0 mmol), H2O (3.0 mL), 1008C. [b] Yields of isolated products determined by 1H
and13C NMR spectroscopy. Tetramethylsilane was used as internal standard.
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org
11552
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Mechanistic studies
Generally, the mechanism of homogeneous Pd-mediated allylic
alkylation is an allylic leaving group forming the h3-allylpalladi-
um species, which reacts with various nucleophiles to furnish
allylic compounds.[7b,8b,c] To evaluate if the Pd/Fe3O4/PEI/rGO-
mediated allylic alkylation involved the same mechanism,
1H NMR spectroscopic experiments were conducted. Pd/Fe3O4/
PEI/rGO exhibited a signal at d=3.60 ppm, which was from PEI
of Pd/Fe3O4/PEI/rGO. After treatment of Pd/Fe3O4/PEI/rGO with
allyl ethyl carbonate 2, two signals appeared at d=4.1 ppm
(a-H) and d=3.45 ppm (b-H) that were enhanced relative to
the signal at d=3.6 ppm in the 1H NMR spectrum (Fig-
ure S6b,c in the Supporting Information), thus indicating for-
mation of the h3-allylpalladium species.[15] On the basis of
these observations and the different yields of allylated prod-
ucts using both Pd/rGO and Pd/Fe3O4/PEI/rGO, we conclude
that Pd/Fe3O4/PEI/rGO-mediated allylic substitution in water
proceeds through a mechanism close to that of homogeneous
Pd catalysis. That is to say, in the Tsuji–Trost reaction, the
PdNPs activate the allyl methyl carbonate to form a propylene/
Pd nanocomposite while the tertiary amines simultaneously ac-
Figure 3. a) Reusability of Pd/Fe3O4/PEI/rGO for ethyl acetoacetate with allyl
ethyl carbonate in water and air. b) TEM image of Pd/Fe3O4/PEI/rGO nano-
composites after 20 cycles. c) Recycling abilities of different nanocomposites
in water and air (30 min).
tivate 1,3-dicarbonyls by abstraction of an
(Scheme 1).
a
proton
for achieving high activity and
catalyst reusability. To address
this possibility, the evolution of
Pd leaching was measured by
means of inductively coupled
plasma atomic emission spec-
troscopy
(ICP-AES).
After
20 cycles of reaction and mag-
netic separation of the catalyst,
a total of 7.6% Pd remained in
the liquid phase of water. Addi-
tionally, the morphologies of
PdNPs of the recovered Pd/
Scheme 1. Possible mechanism of Pd/Fe3O4/PEI/rGO in the Tsuji–Trost reaction.
Fe3O4/PEI/rGO were examined by
TEM (Figure 3c). The images indi-
Recycling of the catalyst
cate that the morphology and
Isolation, convenient recycling, and reuse of a catalyst are
known to be crucial requirements to develop an efficient
“green” system. For the recycling study, the Tsuji–Trost reaction
was performed with ethyl acetoacetate 1 and allyl ethyl car-
bonate 2 and by maintaining the same reaction conditions as
described above except that the recovered catalyst was used.
In our systems, 30 reaction cycles were tested for the catalyst.
After each run, the catalysts could be easily separated from
the reaction mixture with the help of a magnet (Figure 2e,
inset). Upon washing with ethyl acetate, the catalysts were di-
rectly used for the next round of reaction. As shown in Fig-
ure 3a, the catalyst can be reused 30 times and the yield still
remains above 92%. The high recycle counts and yields have
been discovered in our system for the first time and are very
important for applications in industry. However, catalytic activi-
ties of other catalysts decrease very quickly after six cycles (Fig-
ure 3b). Similarly, the stability of the catalyst is highly desirable
size of PdNPs are highly dispersed on graphene surfaces after
the twentieth reuse, which indicates that PEI can provide well-
defined anchor sites for PdNPs and Fe3O4 NPs on the graphene
surface and protect Pd nanoparticles from agglomeration. As
a result, Pd/Fe3O4/PEI/rGO retains excellent catalytic activity
during the recycling process. That could be the reason why
the prepared catalyst has higher catalytic activity and stability
than that reported previously.[16]
Conclusion
We have developed a universal strategy for the preparation of
strongly coupled Fe3O4 NPs and a noble-metal nanocatalyst on
graphene sheets by employing PEI as the coupling linker.
These ternary nanocomposites have a well-defined 2D mor-
phology, controllable size, high dispersion of PdNPs and Fe3O4
NPs, and high specific surface area. We have demonstrated for
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org
11553
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
above dispersion. After stirring for another 12 h at room tempera-
ture, the solvent was evaporated by half under reduced pressure.
The product was separated by centrifugation and washed three
times with ethanol and diethyl ether (1:4 v/v) and then dispersed
into ethanol (15 mL) as a stock solution.
the first time that the present catalyst exhibits synergistic cat-
alysis for the acceleration of the Tsuji–Trost reaction in water
and air. More importantly, the noble-metal nanocatalyst on the
magnetic support has a very low leaching loss and allows one
to reach at least 92% yield in Tsuji–Trost catalysis over 30 cy-
cling tests. All these characteristics are desirable in terms of
cost and environmental protection. Since PEI can chelate vari-
ous transition-metal cations, our work provides a general
methodology for the preparation of other noble-metal/Fe3O4/
PEI/rGO nanocomposites with high stability, high catalytic ac-
tivity, and good magnetic recycling properties for industrial ap-
plications.
Synthesis of Pd/PEI/rGO
Ethanol (40 mL) was added to the stock solution of DIB/PEI/rGO
(3 mL, 2 mg GO) and sonicated for 10 min to obtain a suspension.
The solution of PdCl2 (4.0 mg, 0.023 mmol) in ethanol (5 mL) was
added dropwise into the suspension and stirred for 2 h at room
temperature to ensure the metal ions thoroughly bound to the N
atoms of PEI. Then freshly dissolved NaBH4 (5.2 mg, 0.14 mmol) in
ethanol (2 mL) was added dropwise under an N2 atmosphere into
the above solution over 10 min and stirred for another 2 h. The
raw product was centrifuged after adding hexane (40 mL) and
washed three times with excess amounts of ethanol and hexane.
The resulting solid was then redispersed into ethanol and chloro-
form (40 mL, 2:1 v/v).
Experimental Section
Chemicals
Graphite flake, branched PEI, PdCl2, DIB, [Fe(acac)3] (acac=acetyla-
cetonate; 99.9%), benzyl ether, oleylamine, and triphenylphos-
phine (PPh3) were purchased from Sigma–Aldrich. Potassium per-
manganate (KMnO4), sulfuric acid (H2SO4, 98%), hydrogen chloride
(HCl, 37%), and hydrogen peroxide (H2O2, 30%) were obtained
from Tianjin Med. All chemicals were used without further purifica-
tion. The H2O used throughout the entire experimental process
was deionized H2O. Graphene oxide (GO) was prepared from
Synthesis of Pd/Fe3O4/PEI/rGO
Fe3O4 NPs (4 mg) in a fresh mixture of ethanol and chloroform
(4 mL, 2:1 v/v) was added into as-prepared Pd/PEI/rGO stock solu-
tion and stirred overnight at room temperature. The product was
precipitated by adding hexane and collected by means of an exog-
enous magnet. The residual solid was washed with hexane three
times. Finally, the solid product was obtained by vacuum drying
under an N2 atmosphere.
graphite flake powder according to
a modified Hummers
method,[17] and Fe3O4 nanoparticles were prepared according to
a previous work.[18]
Pd/Fe3O4/PEI/rGO catalyst for the Tsuji–Trost reaction
Instrumentation
For the Tsuji–Trost reaction, 1,3-dicarbonyl (2.0 mmol), allyl ethyl
carbonate (5.0 mmol), PPh3 (7.2 mmol), and H2O (3 mL) were used.
The amount of catalyst used in each reaction was 10 mg (Pd:
1.8 mmol), and the reaction mixture was heated under reflux condi-
tions at 1008C under air. The reaction process was monitored by
thin layer chromatography (TLC). After completion of the reaction,
the mixture was cooled to room temperature and separated by
magnetic separation. The residual mixture was washed with ethyl
acetate (3ꢂ1 mL). The product was separated by column chroma-
tography and determined by 1H and 13C NMR spectroscopy and
MS. After the first cycle of the reaction, another 3 mL of H2O,
7.2 mmol of PPh3, and the corresponding substrates were added
for the next cycle.
TEM measurements were carried out with a JEM-2100 (200 kV) in-
strument under ambient conditions through the deposition of
hexane or ethanol dispersions of the nanomaterials on amorphous
carbon-coated copper grids. X-ray powder diffraction patterns of
the particles were recorded using a Bruker AXS D8 Advance diffrac-
tometer with CuKa radiation (l=1.5418 ꢁ). X-ray photoelectron
spectroscopy (XPS) measurements were performed using a PHI-
5702 multifunctional spectrometer with AlKa radiation. Raman spec-
tra were collected using a confocal microprobe Raman system (Re-
nishaw, RM2000). Magnetic properties were studied using a Lake-
shore 7404 high-sensitivity vibrating sample magnetometer (VSM)
with fields up to 1.5 tesla at room temperature. IR spectra were re-
corded using a Nicolet FT-170SX spectrometer. TGA was carried
out using a TA 2901 instrument under a nitrogen atmosphere at
a heating rate of 108Cminꢀ1 from 30 to 6008C to determine the
1
content of the PEI on the GO. H and 13C NMR spectra were gath-
Acknowledgements
ered using a JEOL ESC 400M instrument. Mass spectrometry was
performed using a TPRACE DSO instrument.
The work was supported by the National Natural Science Foun-
dation of China (21271093), the NCET (13-0262), and the Fun-
damental Research Funds for the Central Universities (lzujbky-
2013-56 and lzujbky-2014-k06).
Synthesis of DIB/PEI/rGO
PEI/rGO was prepared according to the literature with a little modi-
fication:[19] GO (10 mg) was dispersed into deionized water
(100 mL) to form a light brown dispersion by sonication for about
30 min, then a solution that contained PEI (1 g, Mr: 25000) in dis-
tilled water (100 mL) was added. The reaction was maintained at
608C for 12 h. After that, the solvent was removed, and the prod-
uct was separated by centrifugation and washed several times
with ethanol and diethyl ether (1:4 v/v). The product was then re-
dispersed into ethanol (100 mL). To prepare DIB/PEI/rGO, DIB
(55 mg, 0.4 mmol) in ethanol (20 mL) was added dropwise into the
Keywords: graphene · palladium · nanoparticles · synergistic
catalysis · Tsuji–Trost reaction
[1] a) M. Lamblin, L. Nassar-Hardy, J. C. Hierso, E. Fouquet, F. X. Felpin, Adv.
Synth. Catal. 2010, 352, 33–79; b) S. U. Son, Y. J. Jang, J. Park, H. B. Na,
H. M. Park, H. J. Yun, J. Lee, T. Hyeon, J. Am. Chem. Soc. 2004, 126,
5026–5027; c) S. W. Kim, M. Kim, W. Y. Lee, T. Hyeon, J. Am. Chem. Soc.
2002, 124, 7642–7643; d) H. M. Chen, C. K. Chen, R. S. Liu, L. Zhang, J. J.
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org
11554
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Zhang, D. P. Wilkinson, Chem. Soc. Rev. 2012, 41, 5654–5671; e) Z. J.
Han, F. Qiu, R. Eisenberg, P. L. Holland, T. D. Krauss, Science 2012, 338,
1321–1324; f) E. Formo, Z. M. Peng, E. Lee, X. M. Lu, H. Yang, Y. N. Xia, J.
Phys. Chem. C. 2008, 112, 9970–9975; g) J. H. He, I. Ichinose, T. Kunitake,
A. Nakao, Y. Shiraishi, N. Toshima, J. Am. Chem. Soc. 2003, 125, 11034–
11040; h) S. H. Joo, S. J. Choi, I. Oh, J. Kwak, Z. Liu, O. Terasaki, R. Ryoo,
Nature 2001, 412, 169–172.
[8] a) Y. Uozumi, K. Shibatomi, J. Am. Chem. Soc. 2001, 123, 2919–2920;
b) Y. Uozumi, Pure Appl. Chem. 2007, 79, 1481–1489; c) F. X. Felpin, Y.
Landais, J. Org. Chem. 2005, 70, 6441–6446.
[9] a) C. J. Xu, K. M. Xu, H. W. Gu, R. K. Zheng, H. Liu, X. X. Zhang, Z. H. Guo,
B. Xu, J. Am. Chem. Soc. 2004, 126, 9938–9939; b) J. Xie, C. J. Xu, N.
Kohler, Y. L. Hou, S. H. Sun, Adv. Mater. 2007, 19, 3163–3166.
[10] A. J. Amali, R. K. Rana, Green Chem. 2009, 11, 1781–1786.
[11] a) T. Yang, C. Shen, Z. Li, H. Zhang, C. Xiao, S. Chen, Z. Xu, D. Shi, J. Li, H.
Gao, J. Phys. Chem. B 2005, 109, 23233–23236; b) J. Xie, K. Chen, H. Lee,
C. Xu, A. R. Hsu, S. Peng, X. Chen, S. Sun, J. Am. Chem. Soc. 2008, 130,
7542–7543.
[2] Z. Jin, D. Nackashi, W. Lu, C. Kittrell, J. M. Tour, Chem. Mater. 2010, 22,
5695–5699.
[3] a) K. Kçhler, R. G. Heidenreich, J. G. E. Krauter, J. Piestsch, Chem. Eur. J.
2002, 8, 622–631; b) G. Z. Hu, F. Nitze, T. Sharifi, H. R. Barzegar, T. Wꢃg-
berg, J. Mater. Chem. 2012, 22, 8541–8548; c) Y. R. Choi, M. S. Gu, J. N.
Park, H. K. Song, B. S. Kim, Adv. Energy Mater. 2012, 2, 1510–1518;
d) P. H. Li, L. Wang, L. Zhang, G. W. Wang, Adv. Synth. Catal. 2012, 354,
1307–1318; e) B. H. San, S. Kim, S. H. Moh, H. Lee, D. Y. Jung, K. K. Kim,
Angew. Chem. 2011, 123, 12130–12135; Angew. Chem. Int. Ed. 2011, 50,
11924–11929; f) R. M. Crooks, M. Q. Zhao, L. Sun, V. Chechik, L. K.
Yeung, Acc. Chem. Res. 2001, 34, 181–190; g) S. Mandal, D. Roy, R. V.
Chaudhari, M. Sastry, Chem. Mater. 2004, 16, 3714–3724; h) A. J. Amali,
R. K. Rana, Chem. Commun. 2008, 35, 4165–4167.
[4] a) S. Shylesh, V. Schꢄnemann, W. R. Thiel, Angew. Chem. 2010, 122,
3504–3537; Angew. Chem. Int. Ed. 2010, 49, 3428–3459; b) J. Liu, S. Z.
Qiao, Q. H. Hu, G. Q. Lu, Small 2011, 7, 425–443.
[5] a) X. Huo, J. Liu, B. D. Wang, H. L. Zhang, Z. Y. Yang, X. G. She, P. X. Xi, J.
Mater. Chem. A 2013, 1, 651–656; b) C. V. Rao, C. R. Cabrera, Y. Ishikawa,
J. Phys. Chem. C 2011, 115, 21963–21970.
[12] S. Li, D. Q. Wu, C. Cheng, J. Z. Wang, F. Zhang, Y. Z. Su, X. L. Feng,
Angew. Chem. 2013, 125, 12327–12331.
[13] M. Y. Zhu, G. W. Diao, J. Phys. Chem. C 2011, 115, 24743–24749.
[14] a) H. Noda, K. Motokura, A. Miyaji, T. Baba, Angew. Chem. 2012, 124,
8141–8144; Angew. Chem. Int. Ed. 2012, 51, 8017–8020; b) H. Noda, K.
Motokura, A. Miyaji, T. Baba, Adv. Synth. Catal. 2013, 355, 973–980.
[15] H. Kinoshita, H. Shinokubo, K. Oshima, Org. Lett. 2004, 6, 4085–4088.
[16] T. Mitsudome, K. Nose, K. Mori, T. Mizugaki, K. Ebitani, K. Jitsukawa, K.
Kaneda, Angew. Chem. 2007, 46, 3288–3290.
[17] W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 80, 1339.
[18] Z. C. Xu, C. M. Shen, Y. L. Hou, H. J. Gao, S. H. Sun, Chem. Mater. 2009,
21, 1778–1780.
[19] a) D. S. Yu, L. M. Dai, J. Phys. Chem. Lett. 2010, 1, 467–470; b) X. H.
Zhou, Z. X. Chen, D. H. Yan, H. B. Lu, J. Mater. Chem. 2012, 22, 13506–
13516.
[6] R. N. Butler, A. G. Coyne, Chem. Rev. 2010, 110, 6302–6337.
[7] a) J. Tsuji, Acc. Chem. Res. 1969, 2, 144–152; b) B. M. Trost, M. L. Crawley,
Chem. Rev. 2003, 103, 2921–2943.
Received: March 11, 2014
Published online on July 22, 2014
Chem. Eur. J. 2014, 20, 11549 – 11555
www.chemeurj.org
11555
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim