K.-i. Shimizu et al. / Journal of Catalysis 227 (2004) 202–209
203
catalysts in air were not well studied, and a large amount
of the catalysts were generally used for these heterogeneous
catalytic systems whose turnover numbers were several or-
ders of magnitude lower than those of the homogeneous
catalysts. These aspects limit the industrial application of
heterogeneous catalysts for Suzuki reaction.
and dolomite) without a decomposition of sepiolite itself.
[Pd(NH3)4]2 -exchanged sepiolite clay (Pd–sepiolite) was
prepared by exchanging the sepiolite with an aqueous so-
lution of [Pd(NH3)4]Cl2 at 298 K for 48 h, followed by
centrifuging and washing with deionized water, and subse-
quently drying in vacuo at 298 K. A similar procedure was
+
2
+
The stability of the catalytic Pd species is also important
for large-scale industrial applications of the coupling reac-
tions. It is generally assumed that the deactivation of the Pd
catalyst occurs via clustering of palladium intermediates in
the catalytic cycle. It is well known that phosphine ligands
stabilize Pd(0) intermediates and prevent the aggregation of
them to inactive clusters [5–8,16]. The cation-exchangeable
porous inorganic supports may be a candidate for the sta-
bilizer of Pd(II) precursors and Pd(II) intermediates formed
used to prepare [Pd(NH3)4] -exchanged NaY zeolite (Pd–
NaY) and [Pd(NH3)4]2 -exchanged mica (Pd–mica) using
NaY (JRC-Z-Y 5.6, a reference catalyst of the Catalysis
Society of Japan, SiO2/Al2O3 = 5.6) and Na–mica (Na–
fluorotetrasilicic mica, Somasif ME-100, with ideal formula
of NaMg2.5Si4O10F2, COOP Chemicals Co. Ltd.), respec-
tively. [Pd(NH3)4]Cl2 dispersed on the amorphous silica
(Pd–SiO2) was prepared by an impregnation method, fol-
lowed by drying in vacuo at 298 K.
+
during the reaction. Indeed, [Pd(NH3)4]2 -exchanged NaY
zeolite has been shown to be effective for Heck [22,23] and
Suzuki [15,17] reactions. For the Heck reaction, it was spec-
+
XRD patterns were taken by MX Labo (MAC Science)
with CuKα radiation (40 kV, 25 mA). Diffuse reflectance
spectra were obtained with a UV-Vis spectrometer (Jasco;
V-550) and were converted from reflection to absorbance
by the Kubelka-Munk method. Transmission electron mi-
croscopy (TEM) was carried out on a JEOL JEM-2010 op-
erating at 200 kV. Pd K-edge XAFS spectra were taken at
BL-10B of the Photon Factory in High Energy Accelerator
Research Organization in Tsukuba (Japan), with a ring en-
ergy of 2.5 GeV and stored current of 250–350 mA. The
spectra were recorded in a transmission mode at room tem-
perature with a Si(311) channel-cut monochromator. The
ionization chambers for I0 (17 cm) and for I (31 cm) were
filled with Ar. Pd LIII K-edge XANES (X-ray absorption
near-edge structures) spectra were obtained at the BL-9A
station of the Photon Factory with a ring energy of 2.5 GeV
and a stored current of 250–350 mA. The spectra were
recorded in a transmission mode at room temperature with a
Si(111) double-crystal monochromator. High-energy X-rays
from high-order reflections were removed by a pair of flat
quartz mirrors coated with Rh/Ni that were aligned in paral-
lel. The ionization chambers filled with N2(30%)–He(70%)
for I0 (17 cm) and N2(100%) for I (31 cm) were used. For
Pd K and Pd LIII XAFS, the energy was defined by as-
signing the first inflection point of the Cu foil spectrum to
8980.3 eV.
0
ulated that dissolved active Pd species catalyze the reaction
in the channels or cages by a homogeneous mechanism [22].
However, very little attention has been paid to the structural
analysis of the active Pd species before and after the reac-
tion [23], and hence the structure of the active Pd species
during the reaction and the role of inorganic supports on the
stabilization of Pd species are still speculatively discussed.
We chose a sepiolite as a support to stabilize the active Pd
species. Sepiolite is an inexpensive natural clay mineral with
ideal formula of Mg8Si12O30(OH)4 ·4H2O·nH2O. Its struc-
ture is formed of talc-like ribbons arranged in such a way
that the tetrahedral sheet is continuous but inverts in api-
cal directions in adjacent ribbons, generating uniform size
parallel-piped intracrystalline tunnels (10.8 × 4.0 Å) along
the fiber [24]. The magnesium ion on the tunnel wall can
be substituted with various cations [25–27], and highly dis-
persed metal cations act as the catalytic active site [28–30].
Previously, we have reported preliminary results that the
Pd–sepiolite catalyst is highly active and stable for Suzuki
cross-coupling reactions of aryl bromides using DMF as
a solvent at relatively high temperature under an inert at-
mosphere [20].
In this paper, we report that the Pd–sepiolite shows high
TON for Suzuki-type coupling reactions in water. The cata-
lyst is air stable and can be reused without noticeable loss of
activity. The structure of the catalyst before and after the re-
action is well characterized, and the results will be discussed
to reveal the nature of the Pd species required for the stable
catalysis in the Suzuki reaction in water.
2.2. Catalysts tested
The reagents were obtained from commercial sources and
were used without further purification. A typical experi-
mental procedure for Suzuki reactions is as follows. A 20-
mL round-bottom flask was charged with 4-bromophenol
(
1 mmol) and phenylboronic acid (1.2 mmol) or sodium tet-
2
. Experimental
raphenylborate (0.27 mmol), Na2CO3 (3.0 mmol for phenyl-
boronic acid or 1.75 mmol for sodium tetraphenylborate),
catalyst (0.5 µmol, 0.05 mol% Pd), and 10 mL of water.
The mixture was stirred at room temperature in air or in N2.
After the reaction, the basic solution was neutralized by di-
lute HCl solution, and products were extracted with ethyl
acetate (3 mL) three times. After adding n-nonane as an
2
.1. Catalyst preparation and characterization
According to our previous report [31], natural sepio-
lite (Konan, China) was treated with dilute HCl aqueous
solution (0.59 mol dm ) to eliminate impurities (calcite
−3