The palladium-catalyzed Suzuki coupling reaction of aryl halides with arylboronic
acids represents one of the most important methods for the formation of a carbon–carbon
bond.[1 ] Significant progress has been achieved in this area with a variety of homogeneous
palladium catalysts.[2 ] Although homogeneous catalysts have attractive properties such as high reactivity
and selectivity, difficulties are encountered in the separation and recovery of these
expensive catalysts from reaction mixtures for reuse.[3 ] Moreover, homogeneous catalysts tend to lose their activity owing to palladium metal
agglomeration and precipitation.[4 ] Such disadvantages lead to serious economic and environmental concerns in large-scale
synthesis. Therefore, the successful development of homogeneous catalysts has been
often followed by attempts to immobilize the catalysts on insoluble supports.[5 ] A desirable method of supporting homogeneous metal complexes involves the formation
of a chemical bond between the surface of a solid support and a ligand group in the
metal complex. A variety of highly active, heterogeneous palladium catalysts for Suzuki
couplings have been developed.[6 ]
It should be noted that the support may play an important role, affecting the interactions
between pendant catalytic sites and substrates, together with the affinity between
the support and substrates. The most typical supports include porous inorganic solids
such as silica, zeolite, mesoporous materials, and others.[7 ] However, these materials have high sensitivity towards acid and base conditions.[8 ] Carbon materials as supports have attracted a fast-growing interest in the recent
past due to better corrosion resistance in both alkaline and acidic environments.[9 ] This is important from a durability point of view, since support corrosion is one
of the important factors for the aggregation of catalytic metal particles, which causes
considerable loss of catalytic activity. Carbon nanotube supported palladium nanoparticles
have been prepared for various catalytic applications.[10 ]
Graphene oxide (GO) nanosheets, as a form of carbon with a high specific surface area
as well as chemical and thermal stability, arise as an excellent candidate as a catalyst
support.[11 ] Homogeneous metal complexes can be easily immobilized through the formation of covalent
bonds between the surface of GO nanosheets and silylated ligands, as the nanosheets
contain a range of reactive oxygen functional groups on their surface.[12 ] Another important advantage of GO nanosheet support is concerned with the horizontal
open structure. From the point of view of structure openness, GO nanosheets seem to
be a unique catalyst support which can lead to enhanced catalytic activity. Although
GO nanosheet supported palladium nanoparticles have been developed in recent years,[13 ] so far there are few reports of the immobilization of homogeneous catalysts on a
GO nanosheet support. Wu and co-workers have reported an efficient GO-supported palladium
catalyst (aryldiimine as ligand) for Suzuki coupling reactions.[6a ]
It has long been known that N-heterocyclic carbene–palladium (NHC-Pd) complexes act
as very effective catalysts in homogeneous couplings.[14 ] The advantages of adopting NHCs as ligands in Pd-catalyzed couplings are as follows:
(1) the strong σ-donating ability of NHCs into the Pd center facilitates the oxidative
addition process; (2) the steric bulkiness of NHC substituents renders reductive elimination;
and (3) the strong NHC–Pd bond makes it more attractive. Our interest in this area
led us to explore NHC-Pd complexes immobilized on commercial GO nanosheets which may
have advantages over the homogeneous counterparts. Such heterogeneous catalysts with
excellent catalytic performance under mild conditions and long-term durability would
be highly desired. Herein, we present a triazine-tethered NHC-Pd heterogeneous catalyst
as a useful catalyst for the room temperature Suzuki coupling of aryl halides with
arylboronic acids.
Scheme 1 Synthesis of complex 4 . Reagents and conditions : a) APTES, toluene, reflux, 10 h; b) cyanuric chloride, THF, r.t.; c) 1-methylimidazole,
75 °C, 2 h; d) Pd(OAc)2 , THF, 75 °C, 2 h.
As illustrated in Scheme [1 ], GO nanosheet supported NHC-Pd complex 4 was easily prepared in four steps. Aminopropyl-functionalized GO 1 was simply obtained by refluxing a mixture of GO nanosheets and (3-aminopropyl)triethoxysilane
(APTES) in anhydrous toluene. Reaction of 1 with cyanuric chloride in THF gave cyanuric chloride functionalized GO 2 . Subsequently, treatment of 2 with 1-methylimidazole led to the formation of imidazolium-functionalized GO 3 . Complex 4 was obtained by treatment of 3 with Pd(OAc)2 in THF at 75 °C. Loading amounts of Pd on the surface can be controlled in this step.
Complex 4 with a Pd content of 0.19 mmol/g was prepared for this study; the value was confirmed
by inductively coupled plasma atomic optical emission spectrometry (ICP-OES).
To investigate the elemental composition and chemical states of the compounds shown
in Scheme [1 ], X-ray photoelectron spectroscopy (XPS) measurements were performed. As expected,
there are only C 1s and O 1s peaks observed in the XPS survey spectrum of GO nanosheets
(Figures [1a ] and S1, see Supporting Information). The N 1s, Si 2p, and Si 2s peaks found in the
XPS spectrum of 1 demonstrate the effective silylation of GO nanosheets. Meanwhile, Pd 3d peaks are
present at the surface of complex 4 , revealing the successful incorporation of Pd on the GO nanosheets. To further investigate
the chemical composition of the surface of the compounds, further study was carried
out on the N 1s and Pd 3d XPS spectra.
In the high-resolution N 1s XPS spectrum, 1 displays one peak at 399.8 eV corresponding to primary amine (RNH2 ) (Figure [1b ]).[15 ] In the expanded XPS spectrum of 2 , the N 1s peaks at 399.46 and 401.1 eV can be assigned to nitrogen in the form of
triazine (C=N) and secondary amine (R2 NH), respectively (Figure [1c ]).[16 ] In Figure [1d ], the N1s XPS spectrum of 3 consists of three peaks at 399.46, 401.1, and 398.9 eV, corresponding to the binding
energies of triazine (C=N), R2 NH and N+ (imidazole), and R3 N (imidazole), respectively.[17 ]
Figure 1 N 1s XPS spectra of (a) GO, (b) 1 , (c) 2 , (d) 3 , and (e) catalyst 4 ; (f) Pd 3d XPS spectrum of complex 4
After the addition of Pd moieties, the triazine (C=N) and imidazole (R3 N) bands shifted to higher binding energies of 399.66 and 399.09 eV, respectively
(Figure [1e ]). This phenomenon could be explained by donor–acceptor interactions between the
nitrogen in ligands and Pd moieties.[18 ] Moreover, the intensity of the peak at 401.1 eV decreased and the peak at 399.09
eV increased owing to the conversion of N+ into R3 N-type nitrogen after anchoring by Pd moieties. Finally, we attempted to determine
the oxidation state of Pd in complex 4 . The Pd 3d core level XPS spectrum of complex 4 exhibits main peaks at 337.95 and 343.25 eV, which are assigned to Pd(II) (Figure
[1f ]).[18 ]
Elemental maps in energy-dispersive X-ray spectroscopy (EDX) of complex 4 show that N and Pd atoms were homogeneously dispersed on the surface of GO nanosheets.
The data also clarify that the Pd moieties are anchored on the triazine-tethered NHC
ligand (Figure [2 ]).
Figure 2 (a) TEM image; EDX mapping images of (b) carbon, (c) nitrogen, and (d) palladium
for fresh complex 4
In addition, structural elucidations of complex 4 and reused catalyst were performed using transmission electron microscopy (TEM) analysis
(Figure [3 ]). For complex 4 , it is difficult to find agglomerated Pd nanoparticles, suggesting that most of the
Pd moieties are homogeneously dispersed on the surface of the catalyst (Figure [3a,b ]). TEM images obtained from recovered catalyst 4 show that some Pd nanoparticles are formed at the surface of the catalyst (Figure
[3c,d ]).
Figure 3 (a) Low-resolution and (b) high-resolution TEM images of fresh complex 4 ; (c) low-resolution and (d) high-resolution TEM images of catalyst 4 after 12 runs
The results for N2 adsorption–desorption containing the Brunauer–Emmett–Teller (BET) surface area (SBET ) and the total pore volumes (Vtotal ) of GO nanosheets and complex 4 are summarized in Table S1 (see Supporting Information). The isotherms exhibited
type IV hysteresis for GO nanosheets and complex 4 (Figure S2, see Supporting Information). The values for SBET and Vtotal of complex 4 were slightly decreased compared to those of the starting material GO nanosheets.
With complex 4 as the heterogeneous catalyst in hand, we carried out the Suzuki coupling of 4-bromoanisole
with phenylboronic acid as a model reaction (see Table S2, Supporting Information).
The reaction was initially performed in the presence of 0.2 mol% 4 and 2.0 equivalents of K2 CO3 in EtOH/H2 O, and an excellent yield was obtained within 1 hour at room temperature (entry 1).
High catalytic activity was observed with different bases in the presence of 0.1 mol%
4 . K2 CO3 proved to be the most effective in terms of yield (entries 1–4), whereas other inorganic
bases gave moderate yields (entries 5–7) and organic bases were less favorable (entries
8–10). Next, we investigated solvents in the presence of 0.1 mol% 4 . A brief survey revealed that EtOH/H2 O was the best solvent among those evaluated (entries 11–17).
To further expand the scope of our catalytic system, we next investigated the coupling
of various aryl halides with arylboronic acids in the presence of 0.1 mol% 4 at room temperature (Table [1 ]). Bromobenzene and activated aryl bromides such as 1-bromo-2-nitrobenzene and 4-bromobenzonitrile
were very well coupled in excellent yields within 0.6 hour (entries 1–3). Deactivated
aryl bromides including 2-bromotoluene, 4-bromotoluene, 2-bromoanisole, and 4-bromoanisole
were found to furnish the corresponding biaryl products in excellent yields (entries
4–7). Moreover, 4 was efficient for coupling 1-bromonaphthalene and 9-bromoanthracene (entries 8, 9).
Importantly, it was also found to be active for unprotected 3-bromoaniline (entry
10). Electron-poor aryl chlorides are far more difficult to activate than aryl bromides.[19 ] Nevertheless, 4 showed satisfactory efficiency toward 4-chlorobenzonitrile and 1-chloro-2-nitrobenzene
at 50 °C for 6 hours (entries 11, 12). Complex 4 was also efficient for the coupling of deactivated aryl halides with electron-rich
arylboronic acids at room temperature (entries 13, 14).
Table 1 Suzuki Coupling of Aryl and Heteroaryl Halides with Arylboronic Acidsa
Entry
(Hetero)aryl halide
Product
Time (h)
Yield (%)b
1
0.5
99
2
0.6
99
3
0.5
99
4
0.8
99
5
1
99
6
1
99
7
1
99
8
1
97
9
1
95
10
2
93
11
24c
51c
12
24c
65c
13
2
91
14
1
96
15
4c
83c
16
4c
87c
17
4c
91c
18
4c
89c
19
4c
87c
20
2c
87c
a Reaction conditions: aryl halide (1.0 mmol), arylboronic acid (1.5 mmol), 4 (0.1 mol%), K2 CO3 (2.0 mmol), TBAB (0.5 mmol), H2 O (2 mL), EtOH (2 mL), r.t.
b GC yield, determined using n -dodecane as an internal standard.
c The reaction was carried out at 50 °C.
The catalytic system was further extended to the coupling of heteroaryl bromides with
phenylboronic acid. The Suzuki coupling of heteroaryl halides is of particular interest
to the pharmaceutical industry because many biologically active compounds are accessed
through this methodology.[20 ] Despite their importance, the coupling of heteroaryl bromides remains a challenge,
especially at low temperature. It is noteworthy that 3-bromopyridine, 2-bromopyridine,
sterically hindered 2-bromo-3-methylpyridine, 3-bromothiophene, 2-bromothiophene,
and 5-bromo-2-thiophenecarboxaldehyde could be coupled in good yield at room temperature,
although longer time was required (Table [1 ], entries 15–20). In addition, the heteroaryl halides underwent the coupling at 50 °C
to afford the corresponding desired products in high yields in 1 hour. The excellent
catalytic activity of 4 could be explained by the fact that the triazine-tethered NHC could increase the
electron density of the center Pd, consequently to greatly enhance the performance
of 4 .[14a ]
To demonstrate the support effect on catalytic activity, we compared Pd supported
on the most common support SiO2 (Pd-NHC@SiO2 ) with 4 (Figure S3, see Supporting Information). Pd-NHC@SiO2 was prepared using the same method as for catalyst 4 , except for the support. Firstly, we coupled activated 4-bromobenzonitrile and phenylboronic
acid. After 30 minutes, the reaction showed complete conversion with 4 , in comparison to 95% with Pd-NHC@SiO2 . Secondly, we selected deactivated 4-bromoanisole as aryl halide. Within 50 minutes,
4 and Pd-NHC@SiO2 yielded the coupling product in 99% and 84% yield, respectively. In the case of 4-bromoanisole,
the difference in activity between the catalysts was much more conspicuous. Finally,
we selected 4-chlorobenzonitrile as aryl halide and the reaction was carried out at
50 °C. After 6 hours, the yields were 87% and 58% for 4 and Pd-NHC@SiO2 , respectively. On the basis of these findings, we could carefully assert that the
structural openness of GO facilitates the catalysis. It is also noteworthy that homogeneous
triazine-based NHC-Pd catalysts require somewhat higher temperatures and longer reaction
times to achieve effective couplings[21 ] and would suffer from catalyst reuse and recycling problems.
Catalyst reuse is one of the most attractive advantages of the heterogeneous reaction.
The reusability of 4 was investigated in the reaction of deactivated 4-bromoanisole with phenylboronic
acid. As shown in Figure [4 ], after 10 times reuse, the excellent catalytic efficiency was still retained without
loss of catalytic activity. The catalytic activity gradually decreased from the 11th run. TEM studies have confirmed that Pd nanoparticles are formed at the surface of
the GO nanosheets after 12 runs (Figure [3c,d ]).
Figure 4 Reuse of catalyst 4 in the Suzuki coupling of 4-bromoanisole with phenylboronic acid
A leaching test was carried out in the coupling of 3-bromoaniline with phenylboronic
acid. No further reaction took place after catalyst 4 was removed by filtration (Figure S4, see Supporting Information). Also, ICP-OES
analysis of the solution indicated that little Pd metal (0.11 ppm) leached into the
solution during the reaction. These results demonstrated that the high sustainability
of the catalyst could be explained by a strong binding of the triazine-tethered NHC
to Pd.
We have compared the results obtained in this work with various reported heterogeneous
catalysts (see Table S3, Supporting Information), which verified that 4 is highly efficient in room-temperature Suzuki coupling reactions. Along with the
inherent nature of the free Pd complex, the excellent reusability and reactivity of
4 could be explained by the strong binding of triazine-tethered electron-rich NHCs
to the metal center and the structural openness of GO nanosheets.
In summary, we have developed a highly active, recyclable, GO nanosheet supported,
triazine-bridged NHC catalyst for Suzuki coupling reactions. A wide range of aryl
and heteroaryl bromides were very well coupled at room temperature. The catalytic
system is also effective for less reactive aryl chlorides. In particular, the catalytic
activity remained high after several reuses. Along with the inherent nature of the
free Pd complex, the high efficiency appears to be due to the good stability of the
catalytically active Pd species on GO nanosheets. The simple preparation of 4 , in combination with its high catalytic activity and recyclability, may hopefully
stimulate future research in the field of heterogeneous catalysis. Further studies
of other coupling reactions are currently in progress.
