Room-temperature Suzuki-Miyaura reaction catalyzed by Pd supported on rare earth oxides: Influence of the point of zero charge on the catalytic activity

Article abstract of DOI:10.1007/s10562-013-1008-4

Five Pd/REO (rare earth oxide) compounds (Pd/La₂O₃, Pd/CeO₂, Pd/Pr₆O₁₁, Pd/Sm₂O ₃, and Pd/Gd₂O₃) have been successfully used as precatalysts in the Suzuki-Miyaura C-C cross-coupling reaction of aryl bromides with arylboronic acids in ethanol/water at r.t. All Pd/REO showed high activity and selectivity. The effective catalyst arises from palladium leaching from the REO support. We have demonstrated that the activity of each Pd/REO (the decreasing order is Pd/La₂O₃ > Pd/Pr₆O ₁₁ > Pd/Gd₂O₃ > Pd/Sm₂O ₃ ? Pd/CeO₂) is strictly related to the corresponding PZC (point of zero charge) value. Accordingly, it can be reasonably argued that the metal is released in solution in the form of Pd²⁺, and higher is the amount of positive charge on the surface in catalytic conditions, higher is the concentration of Pd²⁺ ions in solution and faster is the formation of the coupling product. Also the recycling of each Pd/REO is strictly connected with the PZC value of the REO support. As a matter of fact, Pd/CeO₂, which has the lowest PZC value (6.7), shows the best reusability, according to its lower tendency to release Pd²⁺ ions in solution.

Full text of DOI:10.1007/s10562-013-1008-4

Catal Lett (2013) 143:547–554
DOI 10.1007/s10562-013-1008-4
Room-Temperature Suzuki–Miyaura Reaction Catalyzed by Pd
Supported on Rare Earth Oxides: Influence of the Point of Zero
Charge on the Catalytic Activity
Francesco Amoroso
Alessandro Del Zotto
Alessandro Trovarelli
•
Sara Colussi
Jordi Llorca
•
•
•
Received: 22 January 2013 / Accepted: 13 April 2013 / Published online: 27 April 2013
Ó Springer Science+Business Media New York 2013
Abstract Five Pd/REO (rare earth oxide) compounds
Pd/La O , Pd/CeO , Pd/Pr O , Pd/Sm O , and Pd/
1 Introduction
(
2
3
2
6
11
2 3
Gd O ) have been successfully used as precatalysts in the
2
The metal catalyzed cross-coupling process has been pro-
ven to be amongst the most useful and widely applied
3
Suzuki–Miyaura C–C cross-coupling reaction of aryl bro-
mides with arylboronic acids in ethanol/water at r.t. All Pd/
REO showed high activity and selectivity. The effective
catalyst arises from palladium leaching from the REO
support. We have demonstrated that the activity of each Pd/
REO (the decreasing order is Pd/La O [ Pd/Pr O [ Pd/
2
2
2
methods for C(sp )–C(sp ) and C(sp )-heteroatom bonds
formation [1–4]. The exponential growth of studies in this
field is mainly due to the exceptional relevance of the Ar–
Ar moiety present in several fine chemicals, drugs and
natural products. Even more than others, the Suzuki–Mi-
yaura (SM) cross-coupling method has found increasing
application for the production of asymmetric biphenyls [3,
5–8]. The development of new palladium-containing het-
erogeneous catalysts is one of the main goals in this area in
view of the recovery and reuse goal [2, 9]. However, it is
now well recognized that Pd-containing heterogeneous
systems employed as catalysts in the cross-coupling reac-
tion generally act as precursors of active forms of palla-
dium through a leaching process [10–17], even if there are
a few studies suggesting that the catalytic process takes
place (or may also occur) on the surface of the solid [18–
23]. Emblematic is the case of Pd/C where conflicting
interpretations have been put forward concerning the role
of Pd leaching on the catalytic activity [13, 23]. Starting
from precatalysts based on palladium supported on metal
oxides, it is generally accepted that the released, catalyti-
cally active, metal particles originate from amorphous PdO
2
3
6 11
Gd O [ Pd/Sm O ꢀ Pd/CeO ) is strictly related to the
2
3
2
3
2
corresponding PZC (point of zero charge) value. Accord-
ingly, it can be reasonably argued that the metal is released
2
in solution in the form of Pd , and higher is the amount of
?
positive charge on the surface in catalytic conditions,
2
?
higher is the concentration of Pd ions in solution and
faster is the formation of the coupling product. Also the
recycling of each Pd/REO is strictly connected with the
PZC value of the REO support. As a matter of fact, Pd/
CeO , which has the lowest PZC value (6.7), shows the
2
best reusability, according to its lower tendency to release
2
Pd ions in solution.
?
Keywords Palladium ꢁ Suzuki–Miyaura reaction ꢁ Metal
leaching ꢁ C–C bond formation ꢁ Rare earth oxides ꢁ Point
of zero charge
[
24–27]. The same is true also starting from various forms
F. Amoroso ꢁ S. Colussi ꢁ A. Del Zotto (&) ꢁ A. Trovarelli
Dipartimento di Chimica, Fisica e Ambiente, Sezione di Scienze
e Tecnologie Chimiche, Universit a` di Udine, via Cotonificio
of (PdO contaminated) metallic palladium [28]. As a
matter of fact, in the case of Pd/Al O catalyst, a reductive
2
3
2
?
pretreatment using hydrogen that partially converts Pd
into Pd(0) afforded a much less catalytically active species
25]. In addition, a neat decrease of catalytic efficiency has
108, 33100 Udine, Italy
e-mail: alessandro.delzotto@uniud.it
[
J. Llorca
been also observed when a high crystalline form of PdO
was provided by prolonged thermal treatment in air [28].
Such a finding is supported by the fact that thermally
Institut de T e` cniques Energ e` tiques and Centre for Research
in NanoEngineering, Universitat Polit e` cnica de Catalunya,
Diagonal 647, 08028 Barcelona, Spain
123

5
48
F. Amoroso et al.
treated Pd/Al O3 precatalyst showed reduced catalytic
2
catalytic studies on the Pd/CeO system to other Pd/REOs
2
activity in the Suzuki coupling [25].
(rare earth oxide), namely, Pd/La O , Pd/Pr O , Pd/
3
2
6 11
Efficiency and reusability of different Pd/supported
catalyst have been correlated to the occurrence of a
leaching/re-deposition mechanism [25, 27]. Accordingly,
at the end of the catalytic organic transformation, Pd(0) is
supposed to deposit back on the surface of the support. In a
recent paper, we have described the SM reaction catalyzed
by Pd/CeO2 [14]. Clear evidences for re-adsorption of
palladium onto the surface at the end of the catalytic pro-
cess were not achieved. The HRTEM analysis of fresh and
used (four times) precatalyst as well as particle size dis-
tribution did not show any remarkable modification.
Sm O , and Pd/Gd O . On the bases of previous results
2 3 2 3
obtained using Pd/CeO [14], our first goal was to gain
2
2
?
insights about the possibility that Pd ions are leached
from Pd/REO and then reduced to Pd(0). According to such
a scenario, all features that favour a weaker palladium-
2
?
surface interaction with consequent dissolution of Pd
ions should result eventually in an increase of the reaction
rate. In particular, we focused our study on the degree of
surface charging as indirectly estimated by measuring the
point of zero charge (PZC) [30–32] of each REO, thus
2
?
correlating the tendency of Pd to enter into solution to
Despite this, Pd/CeO showed to be reusable at least ten
2
the degree of positively/negatively charged surface. Then,
2
?
times without evident loss of catalytic activity. In principle,
these results can be roughly interpreted in terms of heter-
ogeneous coupling, but experimental results provided clear
and unambiguous evidence for the occurrence of a homo-
geneous catalytic process [14]. Literature data strongly
support this conclusion [10, 24, 25].
Pd ions in solution should be easily and rapidly reduced
to Pd(0), the effective catalyst that undergoes oxidative
addition by the aryl halide substrate, as the first step of the
catalytic cycle. If this mechanism is true, then it is rea-
2
?
sonable that the higher the amount of leached Pd ions,
the higher is the rate of formation of the coupling product,
unless aggregation into catalytically inactive form causes a
decrease of the amount of active metal centres.
Another intriguing issue is the paramount importance of
the nature of the base, absolutely necessary for the reaction
to occur. For example, catalytic systems that showed high
efficiency in the presence of NaOH, gave poor results using
Na CO [28]. Analogously, K CO was a better choice
2
3
2
3
2
Experimental Section
t
with respect to KF or KO Bu [14]. Notably, not only the
nature of the anion is important for obtaining high yields of
the coupling product, but also that of the cation. In a pre-
liminary base screening of the present work, K CO
2.1 Materials
2
3
The salts La(NO ) ꢁ6H O, Ce(NO ) ꢁ6H O, Pr(NO ) ꢁ6H O,
3
3
2
3 3
2
3 3
2
showed to be much more helpful than Na CO . By con-
2
3
Sm(NO ) ꢁ6H O, and Ga(NO ) ꢁ6H O were purchased
3
3
2
3 3
2
trast, using palladium wire as the catalyst, the activity in
the presence of NaOH or KOH was about the same [28].
Finally, very recently, the dramatic effect of the (polar or
non polar) nature of the solvent on the formation of a well-
defined species catalytically active in the SM reaction has
been clearly evidenced [29].
from Aldrich. The Pd precursor was a Pd(NO ) solution
3
2
(10 wt% in 10 wt% nitric acid, Pd 99.999 %) purchased
from Aldrich. All solvents were reagent grade and used
without further purification.
In this contradictory context, the recent finding that pal-
ladium particles released by Au@Pd nanoparticles (NPs) are
more active for a second catalytic run than the residual
palladium present in the recovered NP is very interesting
2
.2 Synthesis of Pd/REO 1–5
The Pd/REO catalysts Pd/La O 1, Pd/CeO 2 Pd/Pr O 3,
2
3
2
6 11
Pd/Sm O 4, and Pd/Gd O 5 were synthesized by a two-
2
3
2 3
[
30]. The authors also found that there is not a close relation
steps procedure as follows.
between the amount of noble metal present in solution and
the catalytic activity. Thus, most probably, only a part of the
total palladium released by the NP is present in a catalyti-
cally active soluble form. But the most striking feature of the
study is the demonstration of the exclusive and cooperative
role played by the base and the arylboronic acid on the for-
mation of catalytically active soluble Pd(0). A previous
study based on Pd/C catalyst showed, on the contrary, that
both the aryl halide and the arylboronic acid, but not the base,
promote palladium leaching [13].
First step: the rare earth oxides were prepared starting
from the corresponding nitrate precursor. The lanthanide
nitrate (3 g) was dissolved in water (100 mL) and the
mixture was heated at 333 K under stirring to facilitate
dissolution. Addition of a concentrated solution of NH
3
(5 mL) to the cooled solution resulted in the precipita-
tion of the rare earth hydroxide. The stirring was stopped
and the precipitate was allowed to stand for 1 h, then the
suspension was filtered off and the solid was washed
Stimulated by the need for new investigations that could
give light into this scientific fix, we have extended the
with water to remove traces of NH
. The product was
3
dried at 393 K for 15 h and then calcined at 873 K for
1
23

Room-Temperature SM Reaction Catalyzed
549
4
h with a temperature ramp of 10 K/min to afford the
water, followed by extraction with dichloromethane
rare earth oxide.
(2 9 1 mL). The solution was dried over Na SO and
2
4
Second step: the Pd/REO catalysts were prepared
according to the incipient wetness technique. The Pd
solution was added drop by drop to the REO support to
obtain a Pd/REO catalyst with a nominal 2 wt%
palladium loading. The solid was dried at 393 K for
analysed by GC after purification on a microcolumn filled
with silica gel. Analogous procedure was adopted when the
Pd loading was 0.1 or 0.05 mol%.
2.5 Determination of the Palladium Amount
in Solution
1
h and impregnated again. Once all the precursor had
been absorbed on REO, the solid was dried at 393 K for
5 h and then calcined at 873 K for 4 h with a
1
Four independent experiments were run with both Pd/
temperature ramp of 10 K/min.
La O 1 and Pd/CeO 2.
2
3
2
In the case of Pd/La O , a sample with a nominal
3
(a) In a thermostated bath at 25 °C, a 10 mL Schlenk was
charged in air with a magnetic stir bar, Pd/La O 1
2
0
.5 wt% palladium has been prepared following the same
procedure.
A sample of Pd/La O 1 has been reduced by treatment
2
3
(1.5 mg, 0.05 mol%), 1-naphthylboronic acid
(0.6 mmol), K CO (0.6 mmol), ethanol (1.5 mL)
2
3
2
3
for 3 h at 573 K with H (4 % in Ar) in a Carbolite oven.
2
and H O (0.5 mL), and the reaction was started upon
2
Cooling of the sample was performed at 10 °C/min cooling
rate in He atmosphere.
addition of 1-bromo-4-nitrobenzene (0.5 mmol).
After 2 min, the suspension was poured into a
100 mL round bottom vessel containing ethanol
(
30 mL) and H O (10 mL) to quench the reaction
2
2
.3 PZC Determinations
and then filtered immediately to eliminate the solid
material. The palladium content of the solution was
analyzed by ICP. Pd amount was 13.2 lg.
The PZC value of each REO was measured according to a
published procedure [31]. Water (100 mL) was thermo-
stated under stirring at 298 K and a 0.12 M solution
(
(
b) The reaction was started as in point a) and was
allowed to reach completion (20 min). The solution
was then filtered in order to eliminate the solid and
the palladium content of the solution was analyzed by
ICP. Pd amount was found to be 9.5 lg.
(
0.8 mL) of NaNO was added. The pH of the solution was
3
measured and a solution of NaOH 1 M was added drop by
drop until the pH reached a basic value. At this point, a
crop of REO (about 15–20 mg) was added and the mixture
was left under stirring for 2 min. After this time, the pH of
suspension was measured and another crop of REO was
added. After 2 min of stirring, the pH was measured again.
This procedure was repeated until the pH of the solution
remained unchanged.
c) As in b). The organic product [1-(4-nitrophenyl)naph-
thalene] was extracted with dichloromethane
(
2 9 10 mL). After elimination of the solvent, the
palladium content of the recovered crude solid
117 mg, yield 94 %) was analyzed by ICP. Pd
(
amount was 8.7 lg.
d) As in b) but Pd/CeO 2 (1.4 mg, 0.05 mol%) was
The whole procedure was repeated starting from a dif-
ferent initial basic pH. Analogously, the method was
applied to two starting solutions whose pH was adjusted to
an acidic value by adding, drop by drop, a solution of 1 M
HCl. The PZC value was obtained graphically from the
four curves pH versus REO total mass.
(
2
used instead of Pd/La O 1. Pd amount detected by
2
3
ICP was 1.2 lg.
2.6 Catalyst Recycling
2
.4 Catalytic Runs
In a thermostatted bath at 25 °C, a 8 mL conical centrifuge
test tube was charged in air with a magnetic stir bar, Pd/
La O 1 (30.2 mg), 4-methylphenylboronic acid (0.6 mmol),
The following general procedure was adopted for all
2
3
n
reactions catalyzed by 1–5. In a thermostated bath at
2
diethylene glycol di butyl ether (GC internal standard,
0.5 mmol), K CO (0.6 mmol), ethanol (1.5 mL) and H O
98 K, a 10 mL Schlenk flask was charged in air with a
2
3
2
magnetic stir bar, catalyst (1 mol% Pd), arylboronic acid
n
0.6 mmol), diethylene glycol di butyl ether (GC internal
(0.5 mL), and the reaction was started upon addition of
1-bromo-4-nitrobenzene (0.5 mmol). After 1 h, the sus-
pension was centrifuged for 3 min at 3,500 rpm and the
supernatant was removed and analyzed by GC. The solid
was washed with water and ethanol, dried on air and reused
for the second test, which was performed upon addition in
sequence of solvents, 4-methylphenylboronic acid, GC
(
standard, 0.5 mmol), K CO (0.6 mmol), ethanol (1.5 mL)
2
3
and H O (0.5 mL). Then the reaction was started by the
2
addition of aryl bromide (0.5 mmol). For the GC analysis,
about 0.1 mL of mixture was extracted from the flask by
means of a syringe and to the sample was added 0.5 mL of
123

5
50
F. Amoroso et al.
internal standard, base and 1-bromo-4-nitrobenzene, as
indicated above. Again, the reaction was stopped after 1 h
by centrifugation and the successive workup was analogous
to that previously described. Iteration of this procedure was
continued for four consecutive reuses of the catalyst.
Table 1 Physical and PZC data for compounds 1–5
a
b
c
d
Compound
% Pd
SA
(
PV
(cm /g)
PZC
(pH)
2
m /g)
3
Pd/La
Pd/CeO
Pd/Pr
Pd/Sm
2
O
3
1
1.76
1.93
1.93
1.85
1.99
16.6
31.5
22.4
11.9
10.2
0.24
0.18
0.36
0.21
0.09
8.8
6.7
7.8
7.4
7.5
0
0
2
2
Yields of 4-methyl-4 -nitro-1,1 -biphenyl from the GC
measurements were the following: 99.9 (first use), 99.7,
6
O
11
3
2
O
3
4
5
8
4.5, 20.8, 5.4.
2 3
Pd/Gd O
The same procedure was adopted when Pd/CeO2 2
27.5 mg) was used as the catalyst. In this case the reaction
a
(
Determined by ICP
Surface area
0
b
c
d
time was prolonged to 6 h. Yields of 4-methyl-4 -nitro-
0
1
,1 -biphenyl from the GC measurements were the fol-
Pore volume
lowing: 99.8 (first use), 99.3, 99.1, 98.6, 97.3.
Point of zero charge measured by a procedure described in the
After the fourth reuse, the residual palladium loading of
Pd/La O 1 was 0.04 %, while that of Pd/CeO 2 was
literature [31]. The reported value is the mean of three independent
determinations
2
3
2
0
.58 % (measurements by ICP).
Table 2 XPS data for compounds 1 and 2
2
.7 Instruments
Compound Binding Energy (eV)
%
0
2?
The GC–MS analyses, run to control the identity of the
compounds obtained in the catalytic trials, were carried out
with a Fisons TRIO 2000 gaschromatograph-mass spec-
trometer working in the positive ion 70 eV electron impact
mode. Injector temperature was kept at 250 °C and the
3d5/2
3d5/2
2?
Pd
3d3/2
3d3/2
2?
Pd
Pd
Pd
0
0
Pd
Pd
a
b
a
b
1
1
2
2
a
335.5
335.3
335.7
335.3
336.9
337.0
336.6
336.8
340.7
340.3
341.1
340.4
342.5
342.8
342.6
342.9
36.2 63.8
74.6 25.4
21.8 78.2
61.4 38.6
Ò
column (Supelco SE-54, 30 m long, 0.25 mm i.d., coated
with a 0.5 lm phenyl methyl silicone film) temperature
was programmed from 60 to 280 °C with a gradient of
Fresh sample
b
Sample used 4 times
1
0 °C/min. The GC analyses were run on a Fisons GC
000 Series gaschromatograph equipped with a Supelco
Ò
8
PTA-5 column (30 m long, 0.53 mm i.d., coated with a
.0 lm poly(5 % diphenyl-95 % dimethylsiloxane) film).
The data are shown in Table 1. Furthermore, compounds
Pd/La 1 and Pd/CeO 2 were subjected to X-ray Pho-
toelectron Spectroscopy analysis. The XPS data have been
3
2
O
3
2
Injector and column temperatures were as indicated above.
The X-ray spectra were recorded on a Philips X’Pert dif-
fractometer equipped with Cu K-alpha radiation source.
Step size was 0.01° with a time-per-step of 80 s. SEM
0
2?
collected in Table 2. The co-presence of both Pd and Pd
was clearly evidenced in the two samples, with a neat
prevalence of the latter.
(
Scanning Electron Microscopy) measurements were run
on a ZEISS EVO scanning electron microscope. X-ray
Photoelectron Spectroscopy was carried out with a SPECS
system equipped with an Mg anode XR50 source operating
at 200 W and a Phoibos MCD-9 detector. XP spectra were
recorded with pass energy of 25 eV at 0.1 eV steps at a
3.2 Catalytic Investigation
Preliminary catalytic tests were carried out according to
Scheme 1. The coupling reaction was between 1-bromo-4-
nitrobenzene and 4-methylphenylboronic acid in the pres-
-
9
pressure below 10 mbar.
ence of K CO , in a 3:1 ethanol/H O mixture, at 298 K in
2 3 2
air, according to a well-established protocol [14, 15]. Ini-
tially, the reaction was performed using 1 mol% catalyst
(with respect to the aryl bromide), then the catalyst amount
was decreased to 0.1 mol%, and finally to 0.05 mol%. The
results of the last series of catalytic tests are collected in
Table 3. Among all precatalysts, Pd/La O 1 gave the best
3
Results and Discussion
3
.1 Synthesis and Characterization of 1–5
2
3
Pd/REO 1–5 used in this work were easily prepared by a
two-step synthesis, starting from nitrate salts of the rare
earths and Pd(NO ) . The detailed procedure is described
catalytic performances. A slightly slower reaction was
observed in the case of Pd/Pr
O
11 3, Pd/Sm
O
4, and Pd/
2 showed to be one order of
6
2
3
Gd
O
3
5. Noticeably, Pd/CeO
3
2
2
2
in Sect. 2.2. Characterization of 1–5 was done by means of
BET and XRD measurements and Pd elemental analysis.
magnitude less effective. Analogous trend was observed by
reacting other substrates.
1
23

Room-Temperature SM Reaction Catalyzed
551
Scheme 1 The Suzuki–
Miyaura model reaction
Pd/REO, K2CO3
O2N
Br + CH3
B(OH)2
O2N
CH3
C2H5OH/H2O, 298 K
Table 3 Comparison of the catalytic activity of 1–5
previous investigation based on the application of both Pd/
CeO [14] and PdO hydrate [15] to the SM cross-coupling,
2
a
b
c
d
Entry
Catalyst
0.05 mol%)
Time
(min)
Yield
(%)
TON
TOF
-
1
(h )
we have found that the catalytic activity is completely
suppressed in the presence of a diamine arm grafted on
silica (‘‘heterogeneous chelation test’’). It is likely that the–
NHCH CH NH functionality is able to capture by chela-
(
1
2
3
4
5
6
7
Pd/La
Pd/La
Pd/La
2
O
2
O
2
O
3
1
13
420
360
260
17
68
0
1980
1980
1980
1980
1980
1980
1980
9,130
280
e
3
2
2
2
f
3
2
?
0
330
tion the Pd ions present in solution. However, in prin-
ciple, the much less probable co-ordination of Pd(0) atoms
cannot be excluded [33, 34].
Pd/CeO
Pd/Pr
Pd/Sm
Pd/Gd
2
2
2
450
6
O
11
3
60
52
56
6,980
5,650
5,940
2
O
3
4
5
21
In literature it is reported that the properties of the support,
in particular the surface charging that depends on its PZC,
play a major role in the adhesion of metal nanoparticles onto
the surface [35, 36]. For this reason we decided to measure
the PZC of each REO under conditions close to that used in
the catalytic runs. A reliable determination of PZC,
according to the method of Kallay and coworkers [31, 32],
was done in pure water (an example of PZC determination is
shown in Fig. 1). The PZC values of 1–5 are collected in the
2
O
3
20
Reagents and conditions: 1-bromo-4-nitrobenzene (0.5 mmol),
-methylphenylboronic acid (0.6 mmol), K CO (0.6 mmol), cata-
lyst, 1.5 mL ethanol, 0.5 mL water, T = 298 K
4
2
3
a
0
0
Time necessary to achieve [99 % yield of 4-methyl-4 -nitro-1,1 -
biphenyl, determined by GC using diethylene glycol di butyl ether as
n
internal standard
b
Product yield after 2 min of reaction
c
Turnover number calculated for 99 % GC yield
last column of Table 1. The values obtained for CeO and
2
d
Turnover frequency (mol of product per mol of catalyst per hour)
La O , 6.7 and 8.8, respectively, are very close to those
2 3
e
0
analysis)
.025 mol% using nominal 0.5 wt% Pd (0.47 wt% Pd from ICP
previously reported [37], while no data have been found in
the literature for the other REOs here investigated. It should
be noted that the PZC value measured for a given REO is the
same obtained for the corresponding Pd/REO, that is, the
presence of ca. 2 wt% of palladium on the surface does not
affect appreciably PZC, as expected.
f
2
Catalyst 1 left for 3 h at 573 K under a stream of H (4 % in Ar)
The scope of the catalyzed reaction has been explored.
In particular, in the presence of Pd/La O 1, about twenty
2
3
different combinations of aryl bromides and arylboronic
acids always resulted in the quantitative formation of the
cross-coupling product (isolated yields were in the range
The effective relation between the catalytic activity
shown by 1–5 and their PZC values is clearly evidenced in
Fig. 2, that shows a surprisingly good linear correlation
between the yield of the coupling product in the early
stages of the reaction (after 2 min) and the proton molar
concentration corresponding to the related PZC
8
9–97 %). Selected demonstrative results are presented in
Table 4. The whole data will be reported in a forthcoming
paper. Poor yields of the coupling product were observed
starting from aryl chlorides. For example, the reaction
between 1-chloro-4-nitrobenzene and 4-methylphenylbo-
?
-PZC
([H O ] = 10
3
). Similar good linear correlation
0
?
between product yield and [H O ] was also obtained at
ronic acid gave only 38 % (GC yield) of 4-methyl-4 -nitro-
0
3
1
,1 -biphenyl at 80 °C (Table 4, entry 8). On the other
longer reaction times (5 and 10 min). The higher reactivity
of Pd/La O is therefore correlated to a marked degree of
0
0
hand, only 3-chloro-3 -methoxy-1,1 -biphenyl has been
detected in solution by GC–MS when 1-chloro-3-bromo-
benzene reacted with 3-methoxyphenylboronic acid at
2
3
2
Pd leaching from the surface of La O , favoured by its
?
2
3
less negatively charged surface. According to the obser-
vations of Feltes et al. [38] in fact, at basic pH such as that
of the catalytic reaction (~10.5), high PZC oxides are less
negatively charged than low PZC ones. Within the systems
under investigation, La O shows the higher PZC and thus
2
5 °C (Table 4, entry 4). This clearly indicates that, in the
given experimental conditions, the C–C coupling selec-
tively occurs at the C–Br bond.
2
3
3
.3 Insights into the Formation of the True Catalyst
the lower density of negatively charged sites onto the
surface. On the contrary, the lowest activity observed with
It is reasonable to expect that that the sequence of activity
Pd/CeO is correlated to the higher degree of negatively
2
2
?
observed over the REO support (Pd/La O 1 [ Pd/Pr O
charged surface which help keeping Pd ions anchored to
the support, as schematically represented in Fig. 3.
To confirm our hypothesis, two sets of measurements on
the model reaction were carried out using modified
2
3
6 11
3
[ Pd/Gd O 5 [ Pd/Sm O 4 ꢀ Pd/CeO 2) have to be
2
3
2
3
2
related to the tendency of each Pd-REO combination to
deliver different amounts of palladium in solution. In a
123

5
52
F. Amoroso et al.
Table 4 Scope of the catalyzed reaction
1
2
a
Entry
1
X
R
R
Product
Yield
Br
4-NO
2
4-CH
–
3
96
91
89
2
3
Br
Br
4-CHO
4-CN
3-COCH
3
4
5
Br
Br
3-Cl
3-OCH
3
91
93
4-CHO
3,5-CF
3
b
6
Br
4-NO
2
96
c
d
7
8
Br
Cl
–
97
e
38
4-NO
2
4-CH
3
Reagents and conditions: arylbromide (0.5 mmol), arylboronic acid (0.6 mmol), K
ethanol, 0.5 mL water, 25 °C (unless otherwise stated)
2
CO
3
(0.6 mmol), Pd/La
O
2 3
1 (26.6 mg, 1 mol%), 1.5 mL
a
Isolated yield
b
1
-Naphthylboronic acid
c
d
e
5
-Bromopyrimidine
Reaction run at 60 °C
GC yield
Pd/La O samples. The first one was prepared with a much
2
deactivation of this precatalyst following recycling. Thus,
only 5.4 % yield of the coupling product was obtained
upon the fourth reuse of 1. Not surprisingly, the recovered
catalyst showed only 0.04 % residual palladium, as
checked by ICP elemental analysis. Conversely, catalyst
3
lower metal loading (nominal 0.5 wt% Pd), the second one
was obtained by reducing catalyst 1 at 573 K for 3 h under a
stream of H . A very slow reaction rate was observed using
2
nominal 0.5 wt% Pd (0.025 mol% Pd with respect to 1-
bromo-4-nitrobenzene), see entry 2 in Table 3. In fact, the
reaction was completed (yield[99 %) only after 7 h. Thus,
the lower the noble metal loading on the REO surface, the
Pd/CeO 2 showed better reusability, as almost quantitative
2
formation of the coupling product (GC yield 97.3 %) was
detected upon the fourth reuse. In this case, the ICP anal-
ysis of the recovered catalyst showed that the Pd content
was still high (0.58 %).
2
?
lower the release of Pd ions in solution. Consistently, the
H -reduced Pd/La O sample also showed a very low cat-
2
2 3
alytic activity due to the prevailing presence of Pd(0) on the
surface of the support [25], see entry 3 in Table 3.
Suitably designed experiments have been carried out in
order to assess the Pd content in solution during catalysis,
at the end of the reaction, and in the isolated organic
product. In the coupling reaction between 1-bromo-4-
2
The higher tendency of Pd to enter in solution in the
?
case of Pd/La O3 1 is also confirmed by the rapid
2
1
23

Room-Temperature SM Reaction Catalyzed
553
10,5
Pd2+
10
9
8
7
6
5
,5
9
Pd2+
Pd2+
-
-
- + -
- + - - - + -
,5
8
CeO2
PZC = 6.7
,5
7
Pd2+
,5
6
Pd2+
Pd2+
+
- + - + - + -
+ - + - + -
,5
0
50
100
150
200
La O3
2
weight (mg)
PZC = 8.8
Fig. 1 PZC determination of La
method [26]. Experimental conditions: water (100 mL) was thermo-
stated at 298 K and a 0.12 M solution (0.8 mL) of NaNO was added
under stirring. The pH of the solution was adjusted to a desired
starting acidic or basic value by adding drop by drop 1 M solutions of
2 3
O by adopting the mass titration
2?
Fig. 3 Representation of adhesion and release of Pd from the REO
surface under catalytic conditions (pH ~ 10.5)
3
re-deposition onto the REO support involves less than 30 %
3
HNO or NaOH. Then, a crop of REO (about 15–20 mg) was added
and the mixture was left under stirring for 2 min. After this time, the
pH of the suspension was measured. This procedure was repeated
until the pH values reached a plateau
of dissolved palladium. Interestingly, when Pd/CeO 2 was
2
used as the precatalyst, the amount of Pd in solution at the
end of the reaction (ca. 4 h) was only 1.2 lg, in accordance
with its much lower catalytic activity due to a reduced
palladium leaching. In the case of 1, we have also measured
the amount of Pd on the isolated organic product. It was
found that 1-(4-nitrophenyl)naphthalene was contaminated
by 8.7 lg of Pd (of the residual 9.5 lg). Thus, more than
90 % of the noble metal detected in solution at the end of
the reaction is present in the isolated crude coupling
product.
The XPS analysis of samples of 1 and 2 recovered after
2
the fourth reuse showed a neat increase of the Pd(0)/Pd
?
ratio in the residual noble metal present on the surface of
the REO support (Table 2). This may be seen as a further
2
?
proof that the release of Pd ions from Pd/REO is the
crucial step for the formation of the actual catalyst. On the
other hand, it may be due also to partial re-deposition of
palladium on the surface of REO in the form of Pd(0).
Furthermore, from the XPS analysis of both fresh and used
precatalysts 1 and 2, Pd/La and Pd/Ce ratios have been
obtained. Very interestingly, while in the case of 2, the Pd/
Ce value remained unchanged (fresh: 0.066, used 4 times:
0.064), in the case of 1 the Pd/La ratio notably increased
(fresh: 0.050, used 4 times: 0.073). Thus, albeit the amount
of palladium on the La O surface strongly decreases
0
0
Fig. 2 Yield of 4-methyl-4 -nitro-1,1 -biphenyl (after 2 min of
?
-PZC
reaction) versus [H
Reagents and conditions: 1-bromo-4-nitrobenzene (0.5 mmol), 4-
methylphenylboronic acid (0.6 mmol), K CO (0.6 mmol), catalyst,
.5 mL ethanol, 0.5 mL water, T = 298 K
3
O ] concentration (10
) for Pd/REOs 1–5.
2
3
1
nitrobenzene and 1-napthylboronic acid catalyzed by Pd/
La O 1 (0.05 mol%, 26.4 lg Pd), the amount of Pd in
solution was 9.5 lg (36 %) at the end of the reaction
20 min) and 13.2 lg (50 %) after 2 min, when about 65 %
2
3
(
2
3
of the total coupling product was already formed. This
findings are in line with those previously reported by
K o¨ hler and coworkers [25] using Pd/Al O precatalyst at
during repeated uses of 1, the residual noble metal on the
surface shows higher dispersion. At low temperature this
may be ascribed mainly to a re-deposition process.
In a previous work from our group, it was demonstrated
2
3
6
0 °C, though they found a higher ratio between the two
values (ca. 5 vs. 1.5). Their conclusion was that at the end
of the reaction palladium was almost completely re-
deposited onto the support. Reasonably, if it is assumed that
that a sample of Pd/CeO calcinated at higher temperature
2
(1027 instead of 827 K) gave still good performances upon
the tenth reuse [14]. Thus, the higher the temperature of
calcination of Pd/REO, the higher the reusability, but
conversely the lower the reaction rate. Reasonably, a
13.2 lg may be about the maximum amount of leached
palladium in the first use of 1, then it can be concluded that
123

5
54
F. Amoroso et al.
thermal treatment of the precatalyst at 1,027 K results in an
increase of the grade of crystallinity of PdO, thus making
2. Yin L, Liebscher J (2007) Chem Rev 107:133
3
4
. Alonso F, Beletskaya IP, Yus M (2008) Tetrahedron 64:3047
. Slagt VF, de Vries AHM, de Vries JG, Kellogg RM (2010) Org
Proc Res & Develop 14:30
2
?
more difficult the displacement of Pd
surface of the support.
ions from the
5. Miyaura N, Suzuki A (1995) Chem Rev 95:2457
6. Suzuki A (2002) In: Negishi E (ed) Handbook of organopalla-
dium chemistry for organic synthesis, vol 1. Wiley, New York,
pp 249–262
Finally, a comparative SEM/EDXS analysis of freshly
prepared and used (four times) Pd/La O 1 did not show
2
3
appreciable differences in the morphology of the two
samples. However, it should be noted that this result may
be due to limitation of the resolution. On the contrary, the
elemental analysis confirmed the extremely low amount of
palladium on the exhaust catalyst and revealed the presence
of potassium (~ 10 %) on the La O surface. This finding
7
8. Bai L, Wang J-X (2005) Curr Org Chem 9:535
. Kotha S, Lahiri K, Kashinath D (2002) Tetrahedron 58:9633
9. Lamblin M, Nassar-Hardy L, Hierso J-C, Fouquet E, Felpin F-X
(
2010) Adv Synth Catal 352:33
0. Pr o¨ ckl SS, Kleist W, Gruber MA, K o¨ hler K (2004) Angew Chem
Int Ed 43:1881
1
2
3
11. Phan NTS, Van Der Sluys M, Jones CW (2006) Adv Synth Catal
348:609
2. Richardson JM, Jones CW (2007) J Catal 251:80
3. Chen J-S, Vasiliev AN, Panarello AP, Khinast JG (2007) Appl
Catal A: Gen 325:76
can be of importance regarding the potential role, if any,
played by the cation of the base (K CO ) on the abstraction
1
1
2
3
2
?
of Pd ions from the REO support. Studies are in due
course to explore such intriguing issue.
14. Amoroso F, Colussi S, Del Zotto A, Llorca J, Trovarelli A (2010)
J Mol Catal A: Chem 351:197
1
5. Amoroso F, Colussi S, Del Zotto A, Llorca J, Trovarelli A (2011)
Catal Commun 12:563
4
Conclusions
1
6. Conlon DA, Pipik B, Ferdinand S, Leblond CR, Sowa JR Jr, Izzo
B, Collin P, Ho G-J, Williams JM, Shi Y-J, Sun Y (2003) Adv
Synth Catal 345:931
7. MacQuarrie S, Horton JH, Barnes J, McEleney K, Loock H-P,
Crudden CM (2008) Angew Chem Int Ed 47:3279
8. Huang Y, Zheng Z, Liu T, Lu J, Lin Z, Li H, Cao R (2011) Catal
Commun 14:27
In summary, different Pd/REO compounds have been pre-
pared and tested for the first time as catalysts in the SM cross-
coupling reaction. All of them have shown to effectively
promote the coupling of aryl bromides with arylboronic acids
1
1
2?
at room temperature. Leaching of Pd ions from the surface
of Pd/REO seems to be the first step for achieving the true
soluble catalyst. We have shown that a close relation does
exist between the charge of the surface (measured trough the
PZC value) and the extent of palladium leaching at the pH of
catalysis. Thus, more positively charged is the surface (this is
the case of Pd/La O ), more consistent is the leaching process,
19. Ellis PJ, Fairlamb IJ, Hackett SFJ, Wilson K, Lee AF (2010)
Angew Chem Int Ed 49:1820
2
2
0. Kurokhtina AA, Schmidt AF (2009) ARKIVOC xi:185
1. Gniewek A, Zi o´ łkowski JJ, Trzeciak AM, Zawadzki M, Grab-
owska H, Wrzyszcz J (2008) J Catal 254:121
22. Webb JD, MacQuarrie S, McEleney K, Crudden CM (2007) J
Catal 252:97
2
3. Maegawa T, Kitamura Y, Sako S, Udzu T, Sakurai A, Tanaka A,
Kobayashi Y, Endo K, Bora U, Kurita T, Kozaki A, Monguchi Y,
Sajiki H (2007) Chem Eur J 13:5937
2
3
and therefore higher is the reaction rate. Consistently, a much
slower reaction is observed when Pd/CeO is used as the
2
24. K o¨ hler K, Heidenrich RG, Soomro SS, Pr o¨ ckl SS (2008) Adv
Synth Catal 350:2930
catalyst, which shows the lowest PZC value within the series
of Pd/REO 1–5. This is confirmed by the much higher amount
of palladium measured in solution for Pd/La O with respect
2
5. Soomro SS, Ansari FL, Chatziapostolou K, K o¨ hler K (2010) J
Catal 273:138
2
3
26. Joucla L, Cusati G, Pinel C, Djakovitch L (2010) Adv Synth
Catal 352:1993
to Pd/CeO . As a consequence, prolonged recycling is pos-
2
2
7. Tarabay J, Al-Maksoud W, Jaber F, Pinel C, Prakash S, Dja-
kovitch L (2010) Appl Catal A 388:124
sible inthe caseof Pd/CeO . Onthe contrary, after fourreuses,
2
almost all palladium is disappeared from the surface of more
active Pd/La O . Thus, the use of Pd/CeO as the precatalyst
2
8. Amoroso F, Cersosimo U, Del Zotto A (2011) Inorg Chim Acta
375:256
29. Proutiere F, Schoenebeck F (2011) Angew Chem Int Ed 50:8192
2
3
2
should be preferred for two reasons: (i) it can be recycled
several times without decrease of activity, and (ii) the organic
product is poorly contaminated by the noble metal. Such
considerations are important in view of potential utilization of
these systems in the production of drugs.
3
0. Fang P–P, Jutand A, Tian Z-Q, Amatore C (2011) Angew Chem
Int Ed 50:12184
1. Preo cˇ anin T, Kallay N (1998) Croatica Chem Acta 71:1117
3
32. Kallay N, Preo cˇ anin T, Iv sˇ i c´ T (2007) J Colloid Interface Sci
309:21
3
3. Trzeciak AM, Mieczy n´ ska E, Zi o´ łkowski JJ, Bukowski W,
Bukowska A, Nowor o´ l J, Okal J (2008) New J Chem 32:1124
4. Zhao SF, Zhou RX, Zheng XM (2004) J Mol Catal A: Chem
3
2
11:139
35. D’Souza L, Regalbuto RJ, Miller JT (2011) J Catal 254:157
6. Hao X, Barnes S, Regalbuto RJ (2011) J Catal 279:48
References
3
1
. Miyaura N (2004) In: de Meijere A, Diederich F (eds) Metal-
catalyzed cross-coupling reactions, vol 1. Wiley-WCH, Wein-
heim, pp 41–123
37. Kosmulski M (2011) J Colloid Interface Sci 353:1
38. Feltes TE, Espinosa-Alonso L, de Smit E, D’Souza L, Meyer RJ,
Weckhuysen BM, Regalbuto RJ (2010) J Catal 270:95
1
23