Pd nanoparticles embedded in sponge-like porous silica as a Suzuki-Miyaura catalyst: Similarities and differences with homogeneous catalysts

Article abstract of DOI:10.1016/j.jcat.2007.07.027

Pd nanoparticles of homogeneous size were prepared by PdCl₂ reduction with hydrazine in an inverse micelle microemulsion. These Pd nanoparticles were stabilized with dodecanethiol and 3-mercaptopropyltrimethoxysilane. The latter ligands were co-condensed with tetraethylorthosilicate, yielding an amorphous, sponge-like silica framework embedding the Pd nanoparticles. The resulting npPd@SiO₂ exhibits significant catalytic activity for the demanding Suzuki coupling of electron-rich 4-bromoanisole and phenylboronic acid. The solid can be recovered and reused with only a gradual decrease in catalytic activity. Using the cross-coupling of 4-bromoacetophenone and phenylboronic acid, a total turnover number of 63,000 was obtained. npPd@SiO₂ can considered one of the most active phosphine-free Pd solid catalysts reported to date.

Full text of DOI:10.1016/j.jcat.2007.07.027

Journal of Catalysis 251 (2007) 345–353
www.elsevier.com/locate/jcat
Pd nanoparticles embedded in sponge-like porous silica as a Suzuki–
Miyaura catalyst: Similarities and differences with homogeneous catalysts
∗
Gerolamo Budroni, Avelino Corma , Hermenegildo García, Ana Primo
Instituto de Tecnología Química, UPV-CSIC, Universidad Politécnica de Valencia, Av. de los Naranjos s/n, 46022 Valencia, Spain
Received 23 May 2007; revised 17 July 2007; accepted 30 July 2007
Available online 14 September 2007
Abstract
Pd nanoparticles of homogeneous size were prepared by PdCl reduction with hydrazine in an inverse micelle microemulsion. These Pd
2
nanoparticles were stabilized with dodecanethiol and 3-mercaptopropyltrimethoxysilane. The latter ligands were co-condensed with tetraethyl-
orthosilicate, yielding an amorphous, sponge-like silica framework embedding the Pd nanoparticles. The resulting npPd@SiO exhibits significant
2
catalytic activity for the demanding Suzuki coupling of electron-rich 4-bromoanisole and phenylboronic acid. The solid can be recovered and
reused with only a gradual decrease in catalytic activity. Using the cross-coupling of 4-bromoacetophenone and phenylboronic acid, a total
turnover number of 63,000 was obtained. npPd@SiO can considered one of the most active phosphine-free Pd solid catalysts reported to date.
2
©
2007 Elsevier Inc. All rights reserved.
Keywords: Heterogeneous catalysis; Supported palladium catalyst; Suzuki coupling; Cross-coupling; Metal–organic–inorganic; Nanostructured materials;
Pd nanoparticles
1
. Introduction
hand, gold complexes are more stable under reaction conditions
but generally give lower reactivity then Pd for more demanding
carbon–carbon forming reactions [16–18].
The heterogeneous Pd catalysts could be a solution to avoid
metal agglomeration. However, it has been shown that the cat-
alytic activity of supported palladium catalysts is significantly
lower than that achieved with palladium phosphine ligands
Catalysis by palladium and gold has become an important
topic in chemistry [1,2]. Although the development of ex-
tremely active soluble palladium complexes over the last few
years has been impressive [3], further work on developing spe-
cific ligands and increasing the stability of the palladium com-
plexes is needed. In almost all cases reported to date, the Pd
complexes end up forming palladium particles during the reac-
tion, and the activity of the initial catalyst decreases as the parti-
cle size grows beyond a few tens of nanometers. Thus, whereas
some phosphine ligands, such as derivatives of biphenyl [4–7]
or xanthyl [8–10], form palladium complexes and exhibit very
high catalytic activity for cross-coupling carbon–carbon bond-
forming reactions, they present intrinsic problems derived from
the availability of these ligands and the tendency of the com-
plexes to decompose under reaction conditions, leading eventu-
ally to deactivation of the catalytic system [11–15]. On the other
[
19,20]. For instance, palladium-exchanged zeolites and Pd/Cu
hydrotalcites are active catalysts for the Suzuki, Heck, Sono-
gashira, and cycloisomerization reactions [21–25].
Attempts to combine the advantages of palladium complexes
with those derived from using insoluble supports have been
made by synthesizing periodic mesoporous organosilicas con-
taining carbapalladacycle subunits in their structure. This hy-
brid organic–inorganic solid is active for Suzuki coupling but
rapidly deactivates on reuse [26,27]. It appears, then, that in all
reported cases there is much room for improvement, and a rele-
vant task could be to develop palladium-containing solids with
catalytic activity for the most challenging substrates, at least
comparable with those of the current state-of-the-art homoge-
neous phosphine palladium complexes.
*
Corresponding author. Fax: +34 963877809.
E-mail addresses: Gerbudro@itq.upv.es (G. Budroni), Acorma@itq.upv.es
With this goal in mind, here we report on the synthesis and
the catalytic activity of palladium nanoparticles embedded in an
(A. Corma), Hgarcia@qim.upv.es (H. García), Aprimo@itq.upv.es (A. Primo).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.07.027

346
G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
amorphous silica matrix (npPd@SiO2). The catalysts present a
homogeneous distribution of palladium nanoparticles (average
particle size 4 or 7 nm) included in an “open”, highly porous,
amorphous silica matrix surrounding the palladium nanoparti-
cles and thwarting their agglomeration, stabilizing their size,
and allowing easy mass transfer through the pores. The cat-
alytic activity of the material derives from the small size of the
palladium nanoparticles and its stability within the embedding
silica matrix. In this regard, it has been reported that small pal-
ladium nanoparticles can exhibit catalytic activity comparable
to that exhibited by some palladium complexes [28–30]. It also
has been proposed that these palladium nanoparticles can act as
reservoirs for small atomic palladium clusters that go into the
solution, promoting carbon–carbon coupling reactions in the
homogeneous phase [19,29,31–37]. Then, even if the activity
of Pd-supported solid catalyst were due mainly to the leached
Pd, the results could be of general interest provided that the to-
tal TON was as high as it was with the homogeneous catalysts
but without requiring the sometimes costly and time-consuming
synthesis of ligands.
acterized by transmission electron microscopy (Philips CM300
FEG system with an operating voltage of 100 kV), quanti-
tative atomic absorption spectroscopy (Varian-Spectra A-10
Plus), N2 absorption (Micrometrics ASAP 2000), and thermo-
gravimetric analysis (Mettler TGA/SDTA 851e).
2.2. General procedure for the Suzuki coupling of
4-bromoanisole and phenylboronic acid [4]
4-Bromoanisole (3.6 mmol), phenylboronic acid (5.4 mmol)
and K2CO3 (10.8 mmol) in 4 mL of DMF were added at the
open air to a 10 mL round bottom flask equipped with a stir-
◦
ring bar. Then, the mixture was heated in an oil bath to 135 C.
When this temperature was reached, the catalyst was added. Af-
ter 24 h of reaction under vigorous stirring, the mixture was
filtered under vacuum and the catalyst washed with CH2Cl2.
The combined extracts were analyzed by CG using nitroben-
zene as the external standard.
2.3. General procedure for the Suzuki coupling of
4
-bromoacetophenone and phenylboronic acid
2
. Experimental
A 25-mL round-bottomed flask was charged with 4-bromo-
2
.1. Catalyst preparation
acetophenone (3.5 mmol), phenylboronic acid (5.25 mmol), and
K3PO4 (7 mmol) using toluene (10.7 mL) as a solvent. Then
the mixture was heated in an oil bath at 100 C under vigorous
◦
Palladium nanoparticles (npPd) were prepared in a water-in-
oil microemulsion [38]. The three-phase system comprised wa-
ter, isooctane, and AOT [sodium bis(2-ethylhexyl)sulfosuccin-
ate], with a molar ratio of 4.5:60:1 [39]. Two milliliters of 0.1 M
PdCl2 in 0.2 M HCl was added to the previously formed AOT-
isooctane emulsion stirred for 20 min. After another 15 min of
stirring in N2 atmosphere, the ionic Pd was reduced with an ex-
cess of hydrazine (100 µL), forming metallic Pd nanoparticles
in the aqueous phase.
stirring, and the catalyst, npPd@SiO2 (0.001 mol% Pd), was
added. After 19 h of reaction, the mixture was filtered under
vacuum, and the catalyst was washed with CH2Cl2. The com-
bined organic extracts were analyzed by CG using nitrobenzene
as an external standard.
2.4. Three-phase test
The npPd was stabilized with a mixture of thiols (60 µL)
composed of 1-dodecanethiol (DT) and a small amount of
A sample of 4-bromobenzamide covalently bound to sil-
ica was obtained by reacting in toluene at reflux temperature
a suspension of commercial 3-aminopropyl-functionalized sil-
ica (Aldrich) and 4-bromobenzoyl chloride in the presence of
a few drops of pyridine. The functionalized silica solid was
found to be stable under the Suzuki reaction conditions in the
absence of palladium catalyst. Blank controls established that
4-bromobenzoic acid can be recovered from the solid in 60%
efficiency by refluxing a suspension of the 4-bromobenzamide
containing silica in 0.5 M aqueous solution of NaOH for 3 days.
The corresponding 4-bromobenzamide or its derivative could
be isolated after blank controls. The resulting silica-bound
4-bromobenzamide was added to a mixture of phenylboronic
acid and 4-bromoanisole in DMF under analogous conditions
to those specified previously. At the end of the reaction, the
solids (catalyst, inorganic salts, and functionalized silica) were
collected, washed with CH2Cl2, and submitted to digestion at
the reflux temperature with an aqueous 0.5 M solution of NaOH
for 3 days. After this period, the solution was acidified and the
products were extracted with CH2Cl2. The combined liquid so-
lutions were concentrated under reduced pressure and analyzed
by GC. The formation of benzoic acid and 4-carboxybiphenyl
was determined by GC–MS.
3
-mercaptopropyltrimethoxysilane (MPMS) (molar ratio 18:1).
npPd values of 2.9 and 1.5 nm were obtained by adding the thiol
mix 5 min after or 2 min before the hydrazine, respectively. To
destabilize the microemulsion and promote precipitation of the
Pd particles, 40 mL of water and about 0.8 L of ethanol were
added to the microemulsion. Under these conditions, a single
liquid phase was observed and a suspension was formed. Af-
ter 12 h, the solid was filtered using a 0.45-µm Whatman nylon
membrane and washed with H2O and ethanol. The precipitate
was finally redispersed in 12 mL of ethanol.
The supported Pd catalysts were prepared following a sol–
gel process [40] in the presence of thiol-capped npPd. Then
5
.2 mL of tetraethyl orthosilicate (TEOS) and 0.1 mL of H2O
were added to 12 mL of the ethanol solution containing the
capped npPd. The hydrolysis was catalyzed by 1.8 mL of
a 0.0434 M aqueous solution of NH4F. After 40 min under
continuous stirring, the sol became a gel that slowly con-
verted to a light-brown powder. The final npPd@SiO2 catalyst
◦
was prepared treating the dried powder at 450 C for 4 h in
−1
◦
1
1
00 mL min of air and subsequently at 450 C for 3 h in
00 mL min of 20% H2 80% N2. The materials were char-
−1

G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
347
Fig. 1. Schematic representation of npPd capped with DT and MPMS and the npPd@SiO structure. The MPMS (thick lines) provides the organic links between
2
the metal particles and the inorganic sponge-like silica.
3
. Results and discussion
cles and consequently favoring accessibility of the reactants to
the catalytic sites (Fig. 1).
3
.1. Catalyst preparation
The final npPd@SiO catalysts were prepared from this
2
metal–organic–inorganic material, eliminating the organic ma-
◦
The npPd@SiO2 catalyst was prepared from a metal–
terial by calcination at 450 C and then reducing the metal
◦
organic–inorganic material [41] formed by palladium nanopar-
ticles (npPd) linked to the inorganic silica through organic
bridges of 3-mercaptopropylsilyl. The synthesis involved a
sol–gel process performed in presence of npPd coordinated
to MPMS as ligand and using TEOS as an additional source
of silica. The npPd were prepared in an inverse microemul-
sion system, reducing PdCl2 with hydrazine and stabilizing the
metal particles with a convenient mixture of two thiols, DT and
MPMS, at a ratio of 18:1. Thus, these particles will be covered
with an excess of DT and a smaller amount of MPMS (Fig. 1).
This method, based on the use of microdroplets in a con-
fined reactor to form metal nanoparticles, was recently used
to prepare Au-SiO2 catalyst [41]. It allows us to obtain small
nanoparticles with a diameter of few nm and a narrow particle
size distribution. In this work, we prepared npPd with average
diameters of 2.9 and 1.5 nm (Figs. 2a and 2c) by adding the
stabilizing agents after and before the hydrazine, respectively.
In the former case, the npPd were first formed in the water
droplets and then stabilized to prevent agglomeration. In the
latter case, smaller particles were obtained because the thiols
interacted with the Pd before the reduction, providing a driving
force against particle growth. It has been reported that, differ-
ent from Au, in Pd, the addition of thiols before the reducing
agent may promote the formation of polymeric complexes with
a [Pd(SR)2]n structure [42] in which the Pd is in the reduced
form but not aggregated in particles. In our study, nanoparticles
were observed in the TEM images of the unsupported Pd sam-
ples (Fig. 2c). Some formation of [Pd(SR)2]n complex cannot
be completely ruled out. However, we note that the final ther-
mal treatment would eliminate the organic component, leading
in all cases to organic-free Pd metal particles.
at 450 C. In particular, two catalysts were prepared starting
from the 2.9- and 1.5-nm npPd. The TEM images of these
supported catalysts show that after thermal treatment, Pd sin-
tering occurred, and the average particle size grew to 6.8 nm
(Fig. 2e) and 3.9 nm (Fig. 2g). These two catalysts are desig-
nated npPd@SiO (7 nm) and npPd@SiO (4 nm). The thermal
2
2
decomposition of the organic DT and MPMS can be observed in
◦
the thermogravimetric profile between 250 and 350 C (Fig. 3).
The npPd@SiO catalysts showed BET surface areas of
2
2
−1
≈320 m g , with the shape of the N isotherms correspond-
2
ing to a no-ordered mesoporous material with pore size dis-
tribution centered at 7 to 8 nm (Fig. 4). The metal loading,
as measured by atomic absorption spectroscopy, was 1.4% for
npPd@SiO and 0.7% for npPd@SiO . These different load-
2
2
ings reflect the effect of the preparation procedure on the for-
mation and stabilization of the palladium nanoparticles.
3.2. Catalytic tests
Suzuki coupling between 4-bromoanisole with phenyl-
boronic acid was chosen to test the catalytic activity of
npPd@SiO2. This reaction has been proposed as a benchmark
for comparing the efficiency of various palladium catalysts
[43,44]. The presence of a methoxy substituent in the 4-position
makes the aryl bromide more reluctant to undergo carbon–
carbon coupling, thus providing an opportunity to rank the
palladium catalysts according to activity. The reaction was car-
ried out in DMF using K2CO3 as a base and a substrate/Pd
molar ratio of 0.4% at two different temperatures. The results
are indicated in Scheme 1.
As can be seen, working at moderate temperatures, the cross-
coupling product was formed with a high selectivity and a
turnover frequency (TOF), calculated from the time conversion
The two molecules, DT and MPMS, selected as ligands to
stabilize the nanoparticles, play specific roles in the hydrolysis
of TEOS and formation of the supported catalyst. The siloxane
groups of MPMS form a link between npPd and silica, whereas
DT acts as a spacer between npPd and silica, preventing com-
plete occlusion of the particle. The final material can be viewed
as an “open” silica that will limit the sintering of the particles at
high temperature while preventing total occlusion of the parti-
−
1
−1
plot at low conversions, of 1521 mol4-bromobiphenyl mol
h .
Pd
To determine the maximum TOF, the reaction was performed
under the conditions listed in Table 1, but using a substrate/Pd
molar ratio of 0.2%.
It is noteworthy that high temperatures led to considerably
lower selectivity, due mainly to formation of the debromination

348
G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
Fig. 2. (a and c) TEM micrographs of the stabilized npPd of 2.9 nm and 1.5 nm average size, respectively. (e and g) TEM micrographs of the catalysts npPd@SiO
2
(
7 nm) and npPd@SiO (4 nm). Bars = 50 nm. Plots (b), (d), (f) and (h) show the corresponding particle size distribution calculated over 200–400 particles. The
2
average Pd particles size (PS) for each sample is also indicated.
product. A likely explanation for this negative influence of tem-
perature on selectivity is related to the occurrence of leaching
from the solid catalyst to the solution, because higher tempera-
tures favor the presence in the liquid phase of palladium species
that can effect debromination without promoting C–C cross-
coupling. Indeed, the literature includes reports of studies using
the three-phase test that have established that leached palla-
dium species can produce dehalogenation but are inefficient in

G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
349
a
Temperature
Conversion
(%)
Selectivity (%)
◦
(
C)
1
2
3
1
65
61
62
51
>95
43
–
6
–
135
a
Reaction conditions: 4-bromoanisole (682 mg), phenylboronic acid
Fig. 3. Thermogravimetric profile of the metal–organic–inorganic precursor of
npPd@SiO (4 nm).
(
1.5 eq.), K CO (3 eq.), DMF (4 mL). Catalytic data measured at 24 h.
2 3
2
Scheme 1.
Table 1
Results of the Suzuki cross-coupling between 4-bromoanisole and phenyl-
boronic acid catalyzed by npPd@SiO . Reaction conditions: 4-bromoanisole
2
(682 mg using fresh catalyst and same substrate to Pd ratio for consecutive
reuses), phenylboronic acid (1.5 eq.), DMF (4 mL using fresh catalyst, same
◦
substrate to solvent ratio for consecutive reuses), reaction temperature 135 C,
reaction time 24 h
Run Catalyst
Catalyst treatment
Yield
%)
(
1
npPd@SiO (7 nm) Fresh catalyst activated under H at
62
2
2
◦
450 C
2
3
npPd@SiO (7 nm) Catalyst from run 1 without reactivation 22
2
(a)
npPd@SiO (7 nm) Catalyst from run 2 reactivated with
2
42
76
water (pH 5) followed by calcination
◦
under H at 450 C
2
4
npPd@SiO (7 nm) Catalyst from run 3 reactivated with
2
water washings (1 L at pH 1), followed
◦
by calcination under O at 200 C and
2
◦
subsequently under H at 200 C
2
5
6
7
8
9
npPd@SiO (7 nm) Catalyst from run 4 using the same
2
33
96
70
52
63
reactivation procedure as run 4
npPd@SiO (4 nm) Fresh catalyst activated under H at
2
2
◦
a
2
00 C
npPd@SiO (4 nm) Catalyst from run 6 using the same
2
reactivation procedure as run 4
npPd@SiO (4 nm) Catalyst from run 7 using the same
2
(
b)
reactivation procedure as run 4
Fig. 4. Absorption/desorption isotherm cycle and pore size distribution of
npPd@SiO (4 nm) Catalyst from run 6 using analogous
2
npPd@SiO (4 nm).
2
reactivation as run 4 but water washings
(
1 L) at pH 5
promoting the Suzuki reaction [25]. We suggest here that the
size of the palladium species migrating to the solution increases
with increasing temperature, whereas the catalytic activity for
the coupling reaction decreases with increasing particle size.
We will discuss this later on the basis of further experimental
results.
Most previous reports on heterogeneous solid palladium cat-
alysts have noted lower activity and TOF [45–55] than the val-
ues we report here. In other cases in which the solid catalyst
exhibits high activity, the solid contains an atomic palladium
10
11
npPd@SiO (4 nm) Catalyst from run 9 using the same
2
50
60
reactivation procedure as run 9
npPd@SiO2 (4 nm) Catalyst from run 10 using analogous
reactivation as run 9 but water washings
(300 mL) at pH 11
12
npPd@SiO (4 nm) Not calcined catalyst
2
51
31
13
npPd@SiO (4 nm) Catalyst from run 11 reactivated with
2
washings with distilled water (250 mL)
a
◦
The reduction temperature was reduced to 200 C because no differences in
activity were observed activating at higher temperature.

350
G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
complex [56–58] rather than palladium nanoparticles as in the
present case. Considering that the stability of palladium com-
plexes is generally lower than desired and that they are gener-
ally considered to act as precatalysts [59], the use of supported
palladium nanoparticles has the conceptual advantage of avoid-
ing complex and ligands. As mentioned previously, ligands are
sometimes not easily available and their preparation requires
dedicated, time-consuming synthesis, and, moreover, their use
in the preparation of pharmaceuticals and food additives may
introduce additional problems.
calcination under oxygen to burn off any organic material be-
fore the final reduction of palladium. This regeneration protocol
led to optimal recovery of the fresh catalyst activity (Table 1,
runs 4, 7, and 8). We note, however, that extensive acid wash-
ings can lead to palladium depletion; for this reason, the pH
of the washing water was maintained at 5 (Table 1, runs 9
and 10).
3.4. Influence of the size of Pd nanoparticles
It has been reported that palladium nanoparticles sup-
ported on clays exhibited extremely high activity for the cross-
coupling of 4-bromoanisole (TON = 56,000) combined with
some reusability (two uses reported), but, disappointingly,
also exhibited an unexpected “reverse” catalytic behavior and
lower activity than expected for supposedly more reactive aryl
halides, including iodobenzene [60]. In our case, we observed
normal catalytic behavior for npPd@SiO2, with reactivity fol-
lowing the expected chemical behavior of the reactants.
In any case, the results of Table 1, particularly the TOF
values obtained using the minimum palladium amount, rank
npPd@SiO2 (7 nm) at the top of the list of solid catalysts ex-
hibiting high activity for C–C cross-coupling reactions.
Le Bars et al. [63] reported that Heck cross-coupling
catalyzed by PVP-stabilized Pd nanoparticles is structure-
sensitive. In particular, higher reaction rates were observed with
smaller nanoparticles. The authors related the higher activity to
the low-coordination surface Pd. To determine whether there
is any benefit to decreasing the particle size in Suzuki cross-
coupling, we prepared the npPd@SiO2 (4 nm) catalyst. TEM
analysis of npPd@SiO2 (4 nm) after calcination demonstrated
a particle size of around 4 nm, significantly smaller than that of
npPd@SiO2 (7 nm).
We evaluated the catalytic activity of this new npPd@SiO2
(4 nm) sample for the cross-coupling of 4-bromoanisole and
phenylboronic acid under the same conditions used in the pre-
vious study. The results, also given in Table 1 (runs 6–11), show
that npPd@SiO2 (4 nm) has considerably higher catalytic ac-
tivity compared with npPd@SiO2 (7 nm). The solid could be
reused in the previously described reactivation procedure, with
a gradual decrease in catalytic activity on reuse. This decreased
catalytic activity of npPd@SiO2 (4 nm) is due mainly to the de-
pletion of palladium during water washings. To determine the
TON, a specific run was performed with a substrate-to-Pd mo-
lar ratio of 7000. Working under these conditions, the TON was
1632, which falls in the range of the most active homogeneous
Pd catalyst [5]. This TON value was calculated considering that
all Pd atoms act as active catalytic sites. If only the fraction of
Pd located at the surface of the 4-nm particles are considered
(∼25% of the total Pd atoms), then the TON for the accessible
Pd atoms should be 6448. We note that homogeneous cata-
3
.3. Catalyst recovery and regeneration
As noted in Section 1, an advantage of using solid catalysts is
the possibility of their recovery and reuse. In particular, it must
be demonstrated that the final productivity (i.e., kg of prod-
uct per g of palladium atom of the solid catalyst) on reuse is
higher than the productivity of homogeneous catalysts. In other
words, the mere fact that the solid catalyst can be recovered
does not represent any advantage versus an extremely active
homogeneous catalyst that can act at a much higher substrate-
to-palladium ratio, making its recovery for the reaction unnec-
essary. We note, however, that the substrate-to-palladium molar
ratio used in this work is generally 0.4%, a value not uncommon
for many homogeneous catalysts [5,61,62].
◦
To evaluate the reusability of npPd@SiO2 (7 nm), we per-
formed a series of consecutive runs after reaction by simple
filtration of the catalyst. Working under these conditions (run
lysts are normally tested at temperatures below 100 C, because
high temperatures are avoided to prevent rapid decomposition.
However, TONs can be fairly compared also when obtained at
different temperatures.
2
in Table 1), the solid catalyst became considerably deacti-
vated. However, we believe that this deactivation is due mainly
to the presence on the solid catalyst of inorganic and organic
reagents and products acting as poisons. To test this possibil-
ity, we carried out a reactivation procedure involving water
washings to remove inorganic poisons (excess of K2CO3, as
well as KBr and K3BO3 formed as reaction byproducts) fol-
lowed by calcinations to degrade/oxidize the adsorbed organic
material and a thermal reduction to reform palladium metal.
In addition to 4-bromoanisole, we also performed the
Suzuki coupling using 4-bromoacetophenone and phenyl-
boronic acid as reagents. Taking into account that a consid-
erable amount of information is available about the reactiv-
ity of 4-bromoacetophenone, the use of npPd@SiO2 (4 nm)
as a catalyst for the coupling of 4-bromoacetophenone pro-
vides another opportunity to compare the catalytic activity of
npPd@SiO2 with that of other systems reported in the litera-
ture [4]. In particular, performing the reaction under very low
substrate-to-catalyst molar ratios and using biphenyl phosphane
ligands, Buchwald et al. reported a TON as high as 100,000 for
this reaction (Scheme 2) [4]. Working under the same experi-
mental conditions as reported for homogeneous catalysis with
biphenylphosphine but using the npPd@SiO2 (4 nm) catalyst,
we obtained a TON of 37,249 at 19 h in a single run. After one
According to this regeneration scheme, water washings at pH
◦
5
and calcination under hydrogen at 450 C were performed.
As shown in Table 1 (run 3), this reactivation conditions were
unsatisfactory, leading to only partial recovery of the activity.
To increase the activity of used npPd@SiO2 to the level of
that of fresh catalyst, the pH of the washing water was de-
creased to ensure the removal of any base, followed by the

G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
351
reuse, the total TON was 62,980. This TON was calculated on
3.5. Pd leaching tests
the basis of all Pd atoms present in the material. However, if,
as mentioned before, this number is corrected to account exclu-
sively for the activity of the external Pd atoms of the particles,
then the resulting value (TON 251,920) compares very favor-
ably even to those reported for the most active homogeneous
phosphine Pd complex, but without the need for any phosphane
ligand.
Although, as described previously, it was possible to recover
the npPd@SiO2 catalyst and reuse it in consecutive runs with
a significant residual activity, it is very likely that during the
course of the reaction, some Pd species migrate from the solid
to the solution, and catalysis also may occur in the liquid phase.
In fact, most of the heterogeneous Pd catalysts are suspected
to act as reservoirs of leachable Pd species that would migrate
to the solution. Several tests are available to specifically address
the occurrence and extent of leaching in the npPd@SiO2 (7 nm)
solids; one of the most commonly used involves filtration of the
solid catalyst while the reaction mixture is hot and using the
liquid filtrate for further reaction in the absence of solid. We
performed this “filtration while hot” test and observed that the
conversion of 4-bromoanisole stopped completely after filtra-
tion of the solid catalyst. But Davies et al. [64] have challenged
the validity of these types of leaching test, arguing that the un-
avoidable temperature decrease during filtration of the solid is
sufficient to effect a complete redeposition of Pd species on
the surface of the solid, thus disguising the leaching test. These
authors proposed a more elaborate alternative test to safely de-
termine the occurrence of leaching when using Pd-supported
catalysts. In their three-phase test, a compound analogous to
Conditions: 4-bromoacetophenone (1 eq.), phenylboric acid (1.5 eq.), K PO
3
4
◦
(2 eq.), 0.001% molar of Pd, toluene (3 mL/mmol
), 100 C, 19 h of
)
Pd
halide
reaction
a
Catalyst
TON (mol
/at g
product
100,000
npPd@SiO (4 nm)
2
First use
Second use
37,249
62,980
a
Data of catalytic activity for the biphenylphosphine ligand taken from
Ref. [4].
Scheme 2.
Scheme 3. Preparation of 4-bromobenzoyl-bonded silica.
Scheme 4. Three-phase test.

352
G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
one of the substrates is covalently anchored to an insoluble sup-
port and, after reaction under the usual conditions, the solid is
analyzed to determine whether or not the anchored compound
has reacted. For truly heterogeneous catalysis, no reaction of
the covalently anchored compound should be observed.
highly demanding reactions. Moreover, we note that in some
cases, preparation of the ligands for the most active catalysts re-
quires at least three synthesis steps, and the presence of oxygen
and moisture should be carefully excluded during synthesis and
handling of the phosphanes to avoid oxidation to phosphine ox-
ide. These requirements do not apply to our case, because when
using the silica-embedded Pd catalyst, the reactions can be car-
ried out under ambient atmosphere.
To perform the three-phase test, we proceeded to form a
peptidic bond between aminopropyl-functionalized silica and
4
-bromobenzoic acid (Scheme 3) to obtain BrPhCO@SiO2.
A preliminary control was performed, optimizing the basic con-
ditions that allow hydrolyzation of the peptidic bond and recov-
ery of the largest possible part (60% recovery) of the anchored
benzoic acid.
Overall, our findings demonstrate that materials containing
stabilized nanoparticles of sufficiently small size can be highly
active for palladium-catalyzed reactions and can provide an
complementary alternative to the use of specially designed lig-
ands under homogeneous conditions.
When performing the reaction of 4-bromoanisole and phen-
ylboronic acid using npPd@SiO2 (7 nm) as a catalyst, we added
BrPhCO@SiO2 solid as a third phase. At the end of the reac-
tion, we collected the solids, washed them with ethanol, and
submitted them to hydrolysis with NaOH. This treatment al-
lowed us to detect 4-carboxybiphenyl and benzoic acid in the
organic extracts, thereby demonstrating the presence of leached
Pd in the solution during Suzuki coupling (Scheme 4) and/or the
fact that some coupling occurred under the basic aqueous con-
ditions of the dissolution. In any case, we note that leached Pd
species was more selective for dehalogenation than for cross-
coupling, as demonstrated by the greater amount of benzoic
acid formation compared with 4-carboxybiphenyl.
Acknowledgments
Financial support was provided by CICYT (MAT2006-
14274-C02-01 and CTQ 2006-05768). A.P. thanks the Spanish
Ministry of Education for a postgraduate scholarship.
References
[
[
[
[
1] P. Van de Weghe, Lett. Org. Chem. 2 (2005) 113.
2] A. Corma, P. Serna, Science 313 (2006) 332.
3] F.E. Hahn, M.C. Jahnke, T. Pape, Organometallics 26 (2007) 150.
4] J.P. Wolfe, R.A. Singer, B.H. Yang, S.L. Buchwald, J. Am. Chem.
Soc. 121 (1999) 9550.
Recently, Vries [65,66] proposed ligand-free Pd catalysts,
such as simple Pd(AcO)2, as another useful and economi-
cal alternative to the homogeneous complexes. Vries showed
that high activity can be obtained as long as a precise reac-
tant/catalyst ratio is used and proposed that the ligand-free
[5] K.W. Anderson, S.L. Buchwald, Angew. Chem., Int. Ed. 44 (2005) 6173.
[
[
[
6] J. Yin, M.P. Rainka, X.X. Zhang, S.L. Buchwald, J. Am. Chem. Soc. 124
(2002) 1162.
7] T.E. Barder, S.D. Walker, J.R. Martinelli, S.L. Buchwald, J. Am. Chem.
Soc. 127 (2005) 4685.
8] R.J. van Haaren, K. Goubitz, J. Fraanje, G.P.F. van Strijdonck, H. Oev-
ering, B. Coussens, J.N.H. Reek, P.C.J. Kamer, P.W.N.M. van Leeuwen,
Inorg. Chem. 40 (2001) 3363.
0
Pd species are the real active catalysts. Pd salt catalysts also
have been reported for use in Suzuki cross-coupling; in partic-
ular, PdCl2 in a ionic liquid [67] and Pd(OAc)2 stabilized by
DABCO [68] have been tested for phenyl-bromide substrate.
The latter work found remarkable performance similar to that
obtained in the present work; however, it must be noted that no
stabilizing agent is required when using supported Pd catalyst.
Thus, the results for npPd@SiO2 confirms that ligand-free Pd
can provide a convenient alternative to the elaborate homoge-
neous complexes and that the active species can be efficiently
generated from a solid precatalyst as well.
[
9] P. Wehman, H.M.A. van Donge, A. Hagos, P.C.J. Kamer, P.W.N.M. van
Leeuwen, J. Organomet. Chem. 535 (1997) 183.
[
[
10] J. Yin, S.L. Buchwald, J. Am. Chem. Soc. 124 (2002) 6043.
11] K.Q. Yu, W. Sommer, J.M. Richardson, M. Weck, C.W. Jones, Adv. Synth.
Catal. 347 (2005) 161.
[
[
[
12] W.J. Sommer, K.Q. Yu, J.S. Sears, Y. Ji, X. Zheng, R.J. Davis, C.D. Sher-
rill, C.W. Jones, M. Weck, Organometallics 24 (18) (2005) 4351.
13] C.J. Mathews, P.J. Smith, T. Welton, J. Mol. Catal. A Chem. 206 (2003)
77.
14] J. Vicente, J.A. Abad, J. Lopez-Serrano, P.G. Jones, C. Najera, L. Botella-
Segura, Organometallics 24 (2005) 5044.
[
[
15] T.E. Pickett, F.X. Roca, C.J. Richards, J. Org. Chem. 68 (2003) 2592.
16] C. González-Arellano, A. Corma, M. Iglesias, F. Sanchez, J. Catal. 238
4
. Conclusion
(
2006) 497.
In the present work, we have shown that Pd nanoparticles
[
17] C. González Arellano, A. Abad, A. Corma, H. Garcia, M. Iglesia, F. San-
embedded in a porous sponge-like silica is a suitable catalyst for
the Suzuki–Miyaura coupling of electron-rich aryl bromides.
The TON measured after reuse compares well with the val-
ues reported in the literature for the most active phosphine Pd
complexes. The three-phase test experiments have revealed that
the catalytic activity is due, at least in part, to the leaching of
Pd species from the solid to the solution. Taking into account
what is claimed when using homogeneous catalysts based on
noble metals (i.e., the decisive factor is not catalyst recovery,
but rather catalyst TON), the TON obtained with the catalyst
presented here (considering the Pd remaining) is higher than
that of any of the homogeneous catalysts reported to date for
chez, Angew. Chem. Int. Ed. 46 (2007) 1536.
[18] A.S.K. Hashmi, G.J. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896.
19] N.T.S. Phan, M. Van Der Sluys, C.W. Jones, Adv. Synth. Catal. 348 (6)
[
(2006) 609.
[
[
[
20] Y. Ji, S. Jain, R.J. Davis, J. Phys. Chem. B 109 (2005) 17232.
21] L. Djakovitch, K. Koehler, J. Am. Chem. Soc. 123 (2001) 5990.
22] A. Corma, H. Garcia, A. Leyva, Appl. Catal. A 236 (2002) 179.
[23] M. Dams, L. Drijkoningen, D. De Vos, P. Jacobs, Chem. Commun. 10
2002) 1062.
(
[
24] M. Dams, L. Drijkoningen, B. Pauwels, G. Van Tendeloo, D.E. De Vos,
P.A. Jacobs, J. Catal. 209 (1) (2002) 225.
[
[
25] A. Corma, H. Garcia, A. Primo, J. Catal. 241 (2006) 123.
26] C. Baleizao, A. Corma, H. Garcia, A. Leyva, Chem. Commun. 5 (2003)
606.

G. Budroni et al. / Journal of Catalysis 251 (2007) 345–353
353
[
27] J.M. Notestein, A. Katz, Chem. Eur. J. 12 (2006) 3954.
[51] A. Papp, D. Toth, A. Molnar, React. Kinet. Catal. Lett. 87 (2006) 335.
[52] Y. Wang, D.R. Sauer, Org. Lett. 6 (2004) 2793.
[53] G.W. Kabalka, R.M. Pagni, L. Wang, V. Namboodiri, C.M. Hair, Green
Chem. 2 (2000) 120.
[54] D.A. Conlon, B. Pipik, S. Ferdinand, C.R. LeBlond, J.R. Sowa, B. Izzo,
P. Collins, G.J. Ho, J.M. Williams, Y.J. Shi, Y.K. Sun, Adv. Synth.
Catal. 345 (2003) 931.
[28] S. Jansat, M. Gomez, K. Philippot, G. Muller, E. Guiu, C. Claver, S. Cas-
tillon, B. Chaudret, J. Am. Chem. Soc. 126 (2004) 1592.
[
[
29] M.T. Reetz, E. Westermann, Angew. Chem. Int. Ed. 39 (2000) 165.
30] I. Yuranov, P. Moeckli, E. Suvorova, P. Buffat, L. Kiwi-Minsker, A. Ren-
ken, J. Mol. Catal. A Chem. 192 (2003) 239.
[31] J.G. de Vries, Dalton Trans. (2006) 421.
[32] F. Zhaoa, K. Murakamib, M. Shiraia, M. Arai, J. Catal. 194 (2) (2000)
[
[
[
55] B.M. Choudary, S. Madhi, N.S. Chowdari, M.L. Kantam, B. Sreedhar,
J. Am. Chem. Soc. 124 (2002) 14127.
56] J.J.E. Hardy, S. Hubert, D.J. Macquarrie, A.J. Wilson, Green Chem. 6
4
79.
33] K. Kohler, R.G. Heidenreich, J.G.E. Krauter, M. Pietsch, Chem. Eur. J. 8
2002) 622.
[
(
(
2004) 53.
57] C. Baleizao, A. Corma, H. Garcia, A. Leyva, J. Org. Chem. 69 (2004)
39.
[
[
34] A. Biffis, M. Zecca, M. Basato, J. Mol. Catal. A Chem. 173 (2001) 249.
35] M.B. Thathagar, J.E. ten Elshof, G. Rothenberg, Angew. Chem. Int. Ed. 45
4
(2006) 2886.
[58] A. Corma, H. Garcia, A. Leyva, Tetrahedron 60 (2004) 8553.
[36] M.B. Thathagar, P.J. Kooyman, R. Boerleider, E. Jansen, C.J. Elsevier,
G. Rothenberg, Adv. Synth. Catal. 347 (2005) 1965.
[59] J. Spencer, D.P. Sharratt, J. Dupont, A.L. Monteiro, V.I. Reis, M.P.
Stracke, F. Rominger, I.M. McDonald, Organometallics 24 (2005)
[
[
37] K. Kohler, S.S. Prockl, W. Kleist, Curr. Org. Chem. 10 (2006) 1585.
38] S. Eriksson, U. Nylen, S. Rojas, M. Boutonnet, Appl. Catal. A 265 (2004)
5665.
[
[
[
60] K. Shimizu, T. Kan-no, T. Kodama, H. Hagiwara, Y. Kitayama, Tetrahe-
dron Lett. 43 (2002) 5653.
61] R.B. De Vasher, L.R. Moore, K.H. Shaughnessy, J. Org. Chem. 69 (2004)
207.
[
[
39] D.H. Chen, C.J. Chen, J. Mater. Chem. 12 (2002) 1557.
40] E. Reale, A. Leyva, A. Corma, C. Martinez, H. Garcia, F. Rey, J. Mater.
Chem. 15 (2005) 1742.
7919.
62] T. Brenstrum, D.A. Gerristma, G.M. Adjabeng, C.S. Frampton, J. Brit-
ten, A.J. Robertson, J. McNulty, A. Capretta, J. Org. Chem. 69 (2004)
[
[
41] G. Budroni, A. Corma, Angew. Chem. Int. Ed. 45 (2006) 3328.
42] M. Chen, Y.G. Feng, L.Y. Wang, L. Zhang, J.Y. Zhang, Colloids Surf.
A 281 (2006) 119.
7
635.
63] J. Le Bars, U. Specht, J.S. Bradley, D.G. Blackmond, Langmuir 15 (1999)
621.
64] I.W. Davies, L. Matty, D.L. Hughes, P.J. Reider, J. Am. Chem. Soc. 123
2001) 10139.
65] A.H.M. de Vries, F.J. Parlevliet, L. Schmieder-van de Vondervoort, J.H.M.
Mommers, H.J.W. Henderickx, A.M. Walet, J.G. de Vries, Adv. Synth.
Catal. 344 (2002) 996.
[
[
[
[43] N. Miyaura, A. Suzuki, Chem. Rev. 95 (1995) 2457.
[44] J. Dupont, C.S. Consorti, J. Spencer, Chem. Rev. 105 (2005) 2527.
[45] J.R. Ruiz, C. Jimenez-Sanchidrian, M. Mora, Tetrahedron 62 (2006) 2922.
[46] J.R. Ruiz, C. Jimenez-Sanchidrian, M. Mora, J. Fluorine Chem. 127
7
(
(2006) 443.
[47] E. Paetzold, G. Oehme, H. Fuhrmann, M. Richter, R. Eckelt, M.M. Pohl,
H. Kosslick, Microporous Mesoporous Mater. 44–45 (2001) 517.
[48] K. Shimizu, S. Koizumi, T. Hatamachi, H. Yoshida, S. Komai, T. Kodama,
Y. Kitayama, J. Catal. 228 (2004) 141.
[
[
[
66] A.H.M. de Vries, J.M.C.A. Mulders, J.H.M. Mommers, H.J.W. Hender-
ickx, J.G. de Vries, Org. Lett. 5 (2003) 3285.
67] G. Zou, Z.Y. Wang, J.R. Zhu, J. Tang, M.Y. He, J. Mol. Catal. A Chem. 206
[
[
49] R.S. Varma, K.P. Naicker, Tetrahedron Lett. 40 (1999) 439.
50] M. Mora, C. Jimenez-Sanchidrian, J.R. Ruiz, J. Colloid Interface Sci. 302
(2003) 193.
68] J.H. Li, W.J. Liu, Y.X. Xie, J. Org. Chem. 70 (2005) 5409.
(2006) 568.