The effect of the metal precursor-reduction with hydrogen on a library of bimetallic Pd-Au and Pd-Pt catalysts for the direct synthesis of H2O2

Article abstract of DOI:10.1016/j.cattod.2014.02.021

Two sets of bimetallic Pd-Pt (Pd: 1.0; Pt: 0.25-1.0%, w/w) and Pd-Au (Pd: 1.0; Au: 0.25-1.0%, w/w) catalysts have been used, with no added promoter, in the catalytic direct synthesis (CDS) of hydrogen peroxide from its elements at 2 °C with a CO₂/O₂H₂ mixture (72/25.5/2.5%, respectively). The catalysts were supported on the commercial macroreticular ion-exchange resin Lewatit K2621 and were obtained from the reduction with H₂ of ion-exchanged cationic precursors at 5 bar and at 60 °C. The addition of Pt or Au to Pd produced an increase of the initial overall catalytic activity in comparison with monometallic Pd with both the second metals, but with Pt the increase was much higher than with Au. Moreover, the addition of 0.25% (w/w) Pt, or more, invariably made all the Pd-Pt catalysts less selective with respect to Pd alone. In the case of Au, by contrast, the addition of 0.25% w/w produced an increase, albeit small, of the selectivity. As the result, the most active and productive Pd-Pt catalyst was 1Pd025PtK2621 with 1891 mol_H2) mol(Pd+Pt)-1 h^-1 initially consumed, 1875 mol_H2O2) mol(Pd+Pt)-1 h^-1 initially produced, a 45% selectivity towards H₂O₂ at 50% conversion of H₂. In the case of the Pd-Au bimetallic catalysts, 1Pd025AuK2621 was the best one, with 1184 mol_H2) mol(Pd+Pt)-1 h^-1 initially consumed, 739 mol_H2O2) mol(Pd+Pt)-1 h^-1 initially produced, a 55% selectivity towards H₂O₂ at 50% conversion of H₂. Although the characterization of the Pd-Pt and Pd-Au catalysts with TEM showed that the morphology of the nanostructured metal phases in the Pd-Pt and Pd-Au catalysts was very different from each family to the other, no clear correlation between the size of the nanoparticles and their distribution and the catalytic performance was apparent. These catalysts were also generally different, especially the Pd-Au ones, from previously reported related materials obtained from the same support and the same precursor, but with a different reducing agent (formaldehyde).

Full text of DOI:10.1016/j.cattod.2014.02.021

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
The effect of the metal precursor-reduction with hydrogen on a library
of bimetallic Pd-Au and Pd-Pt catalysts for the direct synthesis of H O
2
2
a,b,∗
b,c
a
a
Stefano Sterchele
, Pierdomenico Biasi , Paolo Centomo , Sandro Campestrini ,
c
d
b,c
b
Andrey Shchukarev , Anne-Riikka Rautio , Jyri-Pekka Mikkola , Tapio Salmi ,
a,∗
Marco Zecca
a
Dipartimento di Scienze Chimiche, Università degli Studi di Padova via Marzolo 8, I35131 Padova, Italy
Department of Chemical Engineering, Laboratory of Industrial Chemistry and Reaction Engineering, Process Chemistry Centre (PCC), Åbo Akademi
b
University, Biskopsgatan 8, FI-20500 Åbo-Turku, Finland
c
Department of Chemistry, Technical Chemistry, Chemical-Biochemical Centre (KBC), Umeå University, UMEÅ, SE-90187, Sweden
d
Department of Electrical Engineering, Faculty of Technology, Microelectronics and Materials Physics Laboratories, EMPART Research Group of Infotech
Oulu, University of Oulu, FI-90014 Oulu, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
Two sets of bimetallic Pd-Pt (Pd: 1.0; Pt: 0.25-1.0%, w/w) and Pd-Au (Pd: 1.0; Au: 0.25-1.0%, w/w) catalysts
have been used, with no added promoter, in the catalytic direct synthesis (CDS) of hydrogen peroxide
from its elements at 2 C with a CO2/O2/H2 mixture (72/25.5/2.5%, respectively). The catalysts were sup-
Received 27 November 2013
Received in revised form 30 January 2014
Accepted 13 February 2014
Available online xxx
◦
ported on the commercial macroreticular ion-exchange resin Lewatit K2621 and were obtained from the
◦
reduction with H2 of ion-exchanged cationic precursors at 5 bar and at 60 C. The addition of Pt or Au to Pd
produced an increase of the initial overall catalytic activity in comparison with monometallic Pd with both
the second metals, but with Pt the increase was much higher than with Au. Moreover, the addition of 0.25%
Keywords:
Bimetallic Pd-Pt catalysts
Pd-Au catalysts
(w/w) Pt, or more, invariably made all the Pd-Pt catalysts less selective with respect to Pd alone. In the case
of Au, by contrast, the addition of 0.25% w/w produced an increase, albeit small, of the selectivity. As the
H2O2 Direct Synthesis
−
1
−1
result, the most active and productive Pd-Pt catalyst was 1Pd025PtK2621 with 1891 mol(H2) mol
h
(Pd+Pt)
−1
−1
initially consumed, 1875 mol(H2O2) mol
h
initially produced, a 45% selectivity towards H2O2 at
0% conversion of H2. In the case of the Pd-Au bimetallic catalysts, 1Pd025AuK2621 was the best one,
(Pd+Pt)
5
−
1
−1
−1
−1
with 1184 mol(H2) mol
h
initially consumed, 739 mol(H2O2) mol
h
initially produced, a 55%
(Pd+Pt)
(Pd+Pt)
selectivity towards H2O2 at 50% conversion of H2. Although the characterization of the Pd-Pt and Pd-Au
catalysts with TEM showed that the morphology of the nanostructured metal phases in the Pd-Pt and Pd-
Au catalysts was very different from each family to the other, no clear correlation between the size of the
nanoparticles and their distribution and the catalytic performance was apparent. These catalysts were
also generally different, especially the Pd-Au ones, from previously reported related materials obtained
from the same support and the same precursor, but with a different reducing agent (formaldehyde).
©
2014 Elsevier B.V. All rights reserved.
1
. Introduction
catalytic direct synthesis (CDS) from H /O2 with catalysts based
2
on Pd could potentially lower the cost to less than half its cur-
H O is a desirable clean oxidant for many chemical applica-
rent value and make the use of H O2 feasible for processes such as
2
2
2
tions, but its industrial use is limited due to its high cost. The
primary and waste water treatments, cleaning of semiconductors
and oxidations in industrial chemical synthesis [1]. The CDS is
only seemingly a simple reaction, because several side-reactions
(
Scheme 1) yield water, the most stable molecule in the reaction
∗ Corresponding author at: Department of Chemical Engineering, Laboratory of
network, and adversely affect the selectivity. The latter typically
decreases with H2 conversion, limiting the final concentration of
H O . Safety is also a critical issue in the CDS, because H /O mix-
tures are explosive in a wide composition range (4–96%, v/v). Due
to its higher cost, H2 is usually the limiting reagent and for safety
reasons it is generally very diluted in the reaction mixture [2]. This
Industrial Chemistry and Reaction Engineering, Process Chemistry Centre (PCC), Åbo
Akademi University, Biskopsgatan 8, FI-20500 Åbo-Turku, Finland; Dipartimento di
Scienze Chimiche, Università degli Studi di Padova via Marzolo 8, I35131 Padova,
Italy.
2
2
2
2
E-mail addresses: ssterche@abo.fi (S. Sterchele), marco.zecca@unipd.it
M. Zecca).
(
http://dx.doi.org/10.1016/j.cattod.2014.02.021
920-5861/© 2014 Elsevier B.V. All rights reserved.
0
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
2
capacity = 1.92 mmol/g) was used as a metal/catalyst support. First,
3
the support was washed with deionized water (300 cm for 10 g of
material) and rinsed with methanol (100 cm3 for 10 g material).
Unless otherwise stated, all the other reagents and mate-
rials were used as received from the supplier. [Au(en) ]Cl3
2
(
[
en = 1,2-diaminoethane) was synthesized according to ref. [20];
Pd(NH ) ]SO and [Pt(NH ) ](NO ) were purchased from Alfa
3
4
4
3
4
3 2
Aesar; sodium thiosulfate pentahydrate (99.5%), potassium iodide,
starch, concentrated sulphuric acid and methanol were purchased
from Sigma-Aldrich; HPLC grade methanol (99.99%) from J.T. Baker;
H2, O2 and CO2 (99.999% mol/mol purity) from AGA (Linde group).
Methanol for Karl Fischer titration, Hydranal composite 2 and
ammonium molybdate tetrahydrate were purchased from Fluka.
ICP-MS measurements were carried out with a PerkinElmer
Sciex, ICP Mass Spectrometer 6100 DRC Plus according to the quan-
titative standard mode. Samples were prepared upon diluting the
mother liquor from the metalation experiments (aqueous solutions
of the unreacted metal precursor, see section 2.2) up to a known
fixed volume.
Scheme 1. Reaction network of hydrogen peroxide direct synthesis (side reactions
shadowed).
can severely limit the final concentration of H O and calls for very
2
2
selective catalysts. As a matter of fact the design of the catalyst has
apparently been so far the main tool to solve the dilemma opposing
safety to performance, with many literature reports on this topic in
the last decade.
Particular attention has been paid to bimetallic palladium-
platinum and palladium-gold catalysts supported on inorganic
oxides, such as SiO2 [3], TiO2 [4], ZrO2 and CeO2 [5,6], carbon [7]
and ion-exchange resins (sulfonated polystyrene-divinylbenzene)
[
8,9].
For the bimetallic Pd-Pt catalysts, good performances have been
Karl Fisher titrations were carried out with a Titrino GP 736 from
Metrohm.
reported in both the patent [10,11] and the open literature [5,9,12].
For instance, Pd-Pt/SiO2 catalysts (Pd/Pt = 95/5, mol/mol) showed
excellent results, with a 2.5-fold increase in the production of
H O and only a slight decrease in selectivity in comparison with
monometallic Pd catalyst [12]. However, the addition of higher
amounts of platinum resulted in a significant decrease in the selec-
tivity. The addition of Au as promoter to Pd catalysts has also been
thoroughly investigated and reported in many papers to have ben-
eficial effects [7,9,13].
In this context we have already shown that reduced palladium
catalysts (1%, w/w) and bimetallic Pd-Pt and Pd-Au supported
Lewatit K2621 are interesting catalysts in CDS. In the absence
of promoters (acids or halides) they perform better than e.g. the
The nanocluster size and the distribution thereof, as well as the
structure of the support, were assessed by energy filtered transmis-
sion electron microscopy (EFTEM, LEO 912 OMEGA, LaB6 filament,
120 kV).Samples were prepared by suspending a few milligrams
of the powdered materials in high purity isopropyl alcohol. After
sonication (30 s), a small droplet (5 ␮l) of the suspension was trans-
ferred onto a holey-carbon film coated Cu grids and was eventually
introduced into the microscope.
2
2
All XPS spectra were recorded with Kratos Axis Ultra electron
spectrometer equipped with a delay line detector. A monochro-
mated Al K␣ source operated at 150 W, hybrid lens system with
2
magnetic lens, proving an analysis area of 0.3 × 0.7 mm and a
charge neutralizer were used for the measurements. The binding
energy (BE) scale was referenced to the C1s line of aliphatic carbon,
set at 285.0 eV. Processing of the spectra was accomplished with
the Kratos software.
Pd/SiO2 catalyst in terms of H O2 selectivity [1,9,14].
2
Lewatit K2621 is a the macroreticular sulfonated polystyrene-
divinylbenzene resin (S-PS/DVB), which has been already used as
the support for CDS [9,14,15]. S-PS/DVB resins show at least two
interesting features. The metal precursors can be easily introduced
into the support by simply exchanging the counter-ions of the sul-
2.2. Ion exchange of Pd and Pt
+
+
In a typical experiment, K2621 (1.0 g) was suspended in 10 cm3
of distilled water and let to stand for 2 h. Then aqueous solutions
of [Pd(NH3)4]SO4 and of [Pt(NH3)4](NO3)2 were added to the sus-
pension. The total amount of the [Pd(NH3)4]SO4 was always the
same (0.094 mmol) and corresponded to 1% (w/w) palladium in
the final catalysts. The amount of [Pt(NH3)4](NO3)2 changed from
one experiment to another in order to achieve 0.25, 0.5 and 1%
(w/w) nominal Pt loadings, respectively (Table 1). After adding the
metal precursors, the suspension was let to react overnight under
mechanical stirring (swirling plate). After filtration it was carefully
fonic groups (H or Na , for instance) with metal containing cations.
Moreover, the reduction of the metal precursors takes place inside
porous system of the polymer, thus controlling the size and the size
distribution of the metal nanoparticles [15–18].
In this paper we report on the preparation, characterization and
testing in CDS of two sets of bimetallic Pd-Pt and Pd-Au supported
on K2621, featured by a fixed palladium loading (1%, w/w) and dif-
ferent amounts of either platinum or gold (0, 0.25, 0.5, 1%, w/w,
respectively). A new protocol for the synthesis of catalytic mate-
rials has been developed allowing to obtain metal nanoparticles
upon the reduction of the metal precursors with hydrogen in an
3
washed on the filter with distilled water (3 × 10 cm ) and eventu-
◦
◦
autoclave (5 bar) at 60 C.
ally dried overnight at 110 C. The mother liquor (filtrate and water
2+
2+
The use of the cationic complexes [Pd(NH ) ] , [Pt(NH ) ]
from washing) was analyzed for the unreacted metal by ICP-MS,
which was always found to be less than 0.1% of the initial amount
for both metals. Accordingly, the experimental metal loading was
equal to the nominal value in all the catalysts.
3
4
3 4
3
+
and [Au(en)2] allows their straightforward immobilization in
K2621 by ion-exchange from aqueous solutions. In order to eval-
uate the intrinsic properties of the pristine catalysts, we decided
to prepare and study these new catalysts with neither halide ions
−
−
(
Br , Cl ) nor mineral acids added as selectivity enhancers [19] in
2.3. Ion exchange of Pd and Au
the reaction mixture.
Before introducing the metals, K2621 was transformed into its
neutral sodium form (K2621Na) by flowing a 0.1 M NaOH solution
(
2
. Experimental
ca. 200 cm3 in 4 h) though a glass column where 20 g of the resin
2
.1. Materials and apparatus
had been packed into. K2621Na was carefully washed with dis-
tilled water until neutral pH of the eluate and dried overnight in
◦
A
batch of Lewatit K2621 (sulfonated polystyrene-
oven at 110 C. In a typical experiment K2621Na (1.0 g) was sus-
divinylbenzene macroreticular ion-exchange resin; exchange
pended in 10 cm3 of distilled water and let to stand for 2 h. Then
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
3
Table 1
Selected analytical on the bimetallic Pd-Pt and Pd-Au catalysts supported by K2621 (experimental loading: Pd 1%, w/w; Pt or Au, 0.25, 0.5, 1.0%, w/w).
Pd loadinga
(%, w/w)
M loading
ꢀ
a,b
M /(Pd + M )
(mol/mol)
ꢀ
ꢀ b
Final catalyst
K2621 (g)
[Pd(NH3)4]SO4
g)
[Pt(NH3)4](NO3)2
or [Au(en)2]Cl3 (g)
(
(%, w/w)
1
1
1
1
PdK2621
1.0120
1.0243
1.0196
1.0173
0.0285
0.0262
0.0274
0.0283
–
1.1
1.0
1.1
1
–
0
Pd025PtK2621
Pd05PtK2621
Pd1PtK2621
0.0052
0.0101
0.0205
0.25
0.5
1
0.12
0.22
0.36
1
1
1
Pd025AuK2621
Pd05AuK2621
Pd1AuK2621
1.0420
1.0359
1.0359
0.0292
0.0285
0.0275
0.0062
0.0120
0.0312
1.1
1.1
1.0
0.27
0.5
1.4
0.12
0.21
0.42
ꢀ
a: Experimental amount; b: M = Pt or Au.
◦
aqueous solutions of [Pd(NH ) ]SO and of [Au(en) ]Cl3 (en = 1,2-
the pressure and the temperature (2 C), both the stirrer and the
3
4
4
2
diaminoethane) were added to the suspension. The total amount of
the [Pd(NH ) ]SO was always the same (0.094 mmol) and corre-
pump were switched off and H was fed up to the desired amount. It
2
was always the limiting reagent and its partial pressure was always
kept low enough to be below the lower explosion and flammability
limit [21,22] of the CO , O , H and MeOH mixture. The hydro-
3
4
4
sponded to 1% (w/w) palladium in the final catalysts. The amount
of [Au(en) ]Cl changed from one experiment to another in order
2
3
2
2
2
to achieve 0.25, 0.5 and 1% (w/w) nominal Pt loadings, respectively
Table 1). After adding the metal precursors, the suspension was
gen amount introduced into the reactor (nH O) was calculated
from the difference of hydrogen pressure in the pre-cylinder [23].
2
(
let to react overnight under mechanical stirring (swirling plate).
After filtration it was carefully washed on the filter with distilled
Thus, the composition of the gases was 2.5% H , 25.5% O2 and 72%
2
CO , respectively. These conditions have been previously shown to
2
3
◦
water (3 × 10 cm ) and eventually dried overnight at 110 C. The
mother liquor (filtrate and water from washing) was analyzed for
the unreacted metal by ICP-MS, which was always found to be less
than 0.1% of the initial amount for both metals. Accordingly, the
experimental metal loading was equal to the nominal value in all
the catalysts.
ensure operations under kinetic regime [23].
Samples for the analysis of the product were regularly taken and
the concentration of H O and H O were determined by iodometric
2
2
2
and Karl Fischer titrations, respectively. The initial concentration of
water, [H O] , was also estimated from the Karl Fischer titration of
2
0
the neat solvent and was used to calculate the concentration of
water actually produced by the reaction, [H O]:
2
2
.4. Reduction of bimetallic materials in autoclave with 5 bars of
ꢀ
[
H O] = [H O] − [H O]
2
2
2
0
◦
hydrogen at 60 C
ꢀ
where [H O] is the uncorrected value directly obtained from the
2
titration. These data were used to monitor the progress of the reac-
tion and the changes of the selectivity with time. The former was
represented as the cumulative yield:
The beige solid recovered from the ion-exchange reaction was
suspended with 50 cm3 of THF in a glass vessel inserted into the
autoclave. After flushing three times with H , the autoclave was
2
filled with hydrogen and the reduction was carried at 5 bar of
hydrogen and 60 C for 5 h. After cooling and venting the autoclave,
VL
◦
C(%)H2,t = 100 · {[H O ] + [H O] } ·
2
2
t
2
t
n
H2,0
the catalyst was recovered as a black material by vacuum filtration
3
and carefully washed on the filter with THF (3 × 10 cm ). The solid
If not otherwise stated, the selectivity towards hydrogen perox-
ide at time t (SH ,t) was calculated as:
◦
was dried in oven at 110 C overnight (Table 1) and then ground
2O
2
with pestle and mortar.
[
H O ]
2 2
t
S(%)H2O2,t = 100 ·
{
[H O ] + [H O] }
2
2
t
2
t
2.5. Catalytic tests
The initial productivities (rates of H O and H O production per
2
2
2
total mol of metal) were calculated in a similar way as in reference
9]. The points of the initial sections of the kinetic plots were inter-
polated with a linear fit and the slopes divided by the sum of Pd
and M (Pt or Au) moles. The initial cumulative rates were obtained
from the interpolation with the linear fit of the initial section of the
H2 and O2 mixtures are explosive in the 4–96% range [21]; in
[
the presence of methanol the mixtures can be flammable even
outside of the H /O explosion range [22]. For safe operations the
2
2
H /O /MeOH mixtures must be kept outside of the explosion and
2
2
flammability range by dilution with an inert gas (CO in this work).
2
hydrogen consumption plots. The initial selectivity towards H O2
Not doing so, deliberately or accidentally, increases the fire and
2
was calculated as the initial slope from the interpolation of the plot
explosion hazards, with possible severe consequences.
Catalytic tests were carried out in a 600 cm3 stainless steel,
of hydrogen peroxide concentration vs time divided by the slope of
the plot of hydrogen consumption vs time. According to the mass
balance of hydrogen, the latter curve was obtained simply as the
tailor made jacketed batch reactor with a maximum working pres-
sure of 200 bar, thermostated at the desired temperature by a
cryogenic unity and equipped with a Heidolph RZR 2021 rotor
plot of the sum of H O2 and H O concentrations as a function of
2
2
time. This method will be hereafter referred to as RISLI (ratio of
initial slopes from linear interpolation).
(
200–1000 rpm), an HP-pump (High Pressure pump) for the recir-
culation of the liquid phase and a K-type thermocouple, for continu-
ous temperature detection. Further details can be found in ref. [23].
In order to test each catalyst for CDS, the autoclave was charged
with the catalyst (0.15 g) and flushed four times with 3 bars of car-
3. Results and discussion
bon dioxide. Then CO2 (18.4 bar) and O (6 bar) were fed directly
from the cylinders at 25 C and 420 cm (VL) of methanol were fed
3.1. Synthesis and characterization
2
◦
3
with an HPLC pump, at room temperature. The temperature was
The successful preparation of a number of Pd [8,14] and bimetal-
lic Pd-Au and Pd-Pt catalysts [9] for the production of hydrogen
peroxide from the gaseous elements has already been achieved in
◦
allowed to decrease to 2 C, followed by stirring and recirculation
3
(
1000 rpm and 4 cm /min, respectively). After the stabilization of
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
4
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
Table 2
XPS data on the fresh catalysts.
Sample
Photoelectron line
BE (eV)
FWHM (eV)
1
PdK2621
Pd 3d5/2
Pd 3d5/2
Pt 4f7/2
336.4 Pd(0)
1.05
1.75
337.9 Pd(II)
1Pd1PtK2621
335.8 Pd(0)
1.3
1.45
337.4 Pd(II)
71.7 Pt(0)
1.85
1.75
73.0 Pt(II)
1Pd025PtK2621
Pd 3d5/2
Pt 4f7/2
336.0 Pd(0)
1.05
1.8
337.7 Pd(II)
72.7 Pt(0)
1.55
2.05
74.0 Pt(II)
1
1
Pd1AuK2621
Pd05AuK2621
Pd 3d5/2
335.2 Pd(0)
1.05
1.8
337.6 Pd(II)
Au 4f7/2
Pd 3d5/2
84.0 Au(0)
1.05
336.2 Pd(0)
2.35
2
338.0 Pd(II)
Au 4f7/2
Pd 3d5/2
Au 4f7/2
84.0 Au(0)
1.05
1.45
85.0 Au(I)
1Pd025AuK2621
335.4 Pd(0)
1.05
1.75
337.8 Pd(II)
84.0 Au(0)
1.0
by LogNormal fit is at 4.8 nm with a skewness coefficient (SC) of 0.91
24]. The metal nanoparticles of 1PdK2621 are smaller than those
[
of other 1% (w/w) Pd catalysts supported on K2621, which were
obtained upon reduction of either the same immobilized precur-
sor with formaldehyde [9] or immobilized palladium acetate with
a refluxing mixture of methanol and water [14].
In the Pd-Pt catalysts (Fig. 1E and F) the metal nanoparticle
size are significantly smaller than in 1PdK2621 (Fig. 2; distribution
analyzed on a set of more than 540 nanoparticles). Moreover the
increase of the Pt amount from 0.25 to 1% (w/w) made the nanopar-
ticles smaller and narrowed their size distribution, as already
reported in the literature [25,26].
Fig. 1. Low resolution TEM images of 1PdK2621 (A, 160 K), 1Pd025AuK2621 (B,
160 K), 1Pd05AuK2621 (C, 250 K), 1Pd1AuK2621 (D, 250 K) 1Pd025PtK2621 (E,
50 K), 1Pd1PtK2621 (F, 250 K) samples.
2
In the Pd-Au catalysts the metal nanoparticles (Fig. 1B-D) are
completely different in comparison with the Pd-Pt ones. Upon
increasing the amount of gold from 0.25 to 1 wt.% their size and
size distributions (estimated from the analysis of more than 250
nanoparticles) get larger and broader: whereas in 1Pd025AuK2621
the center of the size distribution (LogNormal fit) is at 8.2 nm
(SC = 1.24), it is at 11.2 nm in 1Pd1AuK2621 (SC = 2.05). Moreover
in 1Pd1AuK2621 very large aggregates of metal nanoparticles are
also present (Fig. 1D).
Table 2 shows the results of XPS analysis on fresh catalysts.
The oxidation states were assessed from peak fitting, but the metal
loading was too low for the reliable quantitative determination of
the metal surface concentration.
The deconvolution of the Pd 3d5/2 signals provided two com-
ponents. The first has a binding energy (BE) in the 335.2–336.4 eV
range and can be assigned to zerovalent Pd; the second has a BE
in the 337.4–338.0 eV range and can be attributed to Pd(II) species.
Whereas the Pd(II) component is relatively little sensitive to the
presence of the second metal, the BE observed for Pd(0) is gener-
ally lower in the bimetallic catalysts. This suggests that the second
metal acts as an electron donor towards palladium. Gold seems
more efficient than platinum in this respect. This could depend
on the fact that whereas platinum is never completely reduced,
gold is generally present only as Au(0), with the exception of
1Pd05AuK2621 in which the BE of Pd(0) is almost as high as in
1PdK2621. The Pt 4f7/2 signals of all the Pd-Pt catalysts could be
separated into two components. The first has a BE ranging from
our laboratory. These catalysts were obtained by supporting the
metal nanoparticles the commercial S-PS/DVB macroreticular ion-
exchange resin Lewatit K2621, which proved to be an efficient
support for CDS [9,14,15]. Its ability to exchange cations is a very
useful feature, which allows the straightforward incorporation of
metals. Under the conditions required to prepare 1% (w/w) metal
catalysts, with the metal precursors as the limiting reagents, the
2+
2+
di-cationic complexes [Pd(NH ) ] and [Pt(NH ) ] and the tri-
3
4
3 4
3
+
cationic [Au(en)2] complexes were quantitatively taken up by the
resin, as shown by the mass balance of the metals based on ICP
analysis. The counter ions of the cationic precursors (SO
Cl ), repelled by the fixed anion charges of the ion-exchanger,
were removed from the polymeric framework by careful washing,
together with any cation (H+ or Na ) released in the ion exchange
process. As the result the ion-exchanged stocks of K2621 contained
2−
−
, NO3 ,
4
−
+
1
% (w/w) of palladium and from 0.25 to 1% (w/w) of the second
metal (Table 1).
Figs. 1 and 2, respectively, show the Transmission Electron
Microscopy (TEM) images and the particle size distributions of the
Pd, Pd-Pt and Pd-Au catalysts after the reduction with H2 at 5 bar
of the precursors in the materials suspended in THF. The low res-
olution images of the monometallic sample 1PdK2621 (Fig. 1A)
illustrate that the nanoparticles are well separated and their diam-
eter ranges from 1.5 to 8.5 nm (Fig. 2, upper left panel). The
particle size distribution, estimated by sampling more than 1000
nanoparticles, was relatively symmetric and the center estimated
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 2. Particle size distributions for 1PdK2621, 1Pd025PtK2621, 1Pd1PtK2621 (left), 1Pd025AuK2621, 1Pd05AuK2621, 1Pd1AuK2621 (right) samples (average diameter
estimated from LogNormal fit of the size distribution; SC calculated from the LogNormal parameters of size distribution [24]).
7
1.7 to 72.7 eV, close to 71.1 of bulk platinum [27], and can be
to H O2 hydrogenation and dismutation). It ends in approximately
2
assigned to Pt(0). The second component has a BE of ca. 74.0 eV
and can be attributed to Pt(II) species (Table 2). In the case of gold
the signals from the 4f5/2 can overlap with those of the Pd 4s ones
and, accordingly, only the 4f7/2 peak was analyzed for Au. In this
case we found a single peak at BE equal to 84.0 eV, which can be
attributed to Au(0). Only in the case of 1Pd05AuK2621 a component
at 85.4 eV, which can be attributed to Au(I), was also observed.
The shift to lower values of the BE for Pd(0) 3d5/2 in the bimetal-
lic catalysts is suggestive of the increase of the electron density of
palladium due to donation from the second metal. This implies the
electronic communication between Pd and either Pt or Au and is
consistent with the hypothesis of nanoalloying between the metals
20–40 min, i.e. when the H2 conversion exceeds 50%. In the
second stage, hydrogen is almost completely consumed and the
main active reaction, if any, is expected to be the dismutation of
hydrogen peroxide (see Scheme 1). However, with these catalysts
the conversion of H2 is always slightly lower than 100%, so that in
the case of the Pd-Au catalysts some H O2 is likely still produced
2
and the decrease of its concentration, if any, is little in comparison
to the monometallic and the bimetallic Pd-Pt catalysts.
The initial activity of the catalysts was assessed as the cumu-
lative amount of H O2 and H O per mole of either palladium or
2
2
total metal (Pd + Pt) and (Pd + Au) and unit time (Table 3). This
corresponds to rates of H2 consumption, which were provided by
the ratios of the slopes obtained from the linear interpolation of
[
28]. A negative shift of BE of the Pd 3d signals has already been
reported for Pd-Au alloys with Pd:Au higher than 1:1 (ca. 0.5 by
weight), as it is in this work. In any case, the nanostructured metal
phase contained at least oxidized Pd. The data presently available
do not allow to suggest the structure of the nanostructured metal
phase(s), because we cannot even know the surface composition.
For this purpose other techniques are required to investigate the
composition of the metal surface, such as Low Energy Ion Scatter-
ing (LEIS) [29], or of the whole metal phase and could be the subject
of further investigation.
3.2. Catalytic tests
The behavior of the Pd-Pt and Pd-Au catalysts is illustrated
in Figs. 3–6, which represent the concentrations of the products
H O , H O; Figs. 3 and 5) and the conversion f H and the selectiv-
(
2
2
2
2
ity towards H O (Figs. 4 and 6) as functions of the reaction time
2
2
over the Pd-Pt (Figs. 3 and 4) and Pd-Au (Figs. 5 and 6) catalysts.
It is possible to separate all the catalytic runs into two stages,
as already reported previously [9]. The first one is observed when
the hydrogen concentration is still relatively high (2.5–1.25 mol%)
in the gas phase. This initial stage is featured by a relatively
high hydrogen consumption rate and in general a fairly linear
change of H2 conversion with time. It is also characterized by
relatively fast H O formation and slow H O2 consumption (due
Fig. 3. Concentration of water and hydrogen peroxide produced over the Pd-Pt
catalysts as a function of time.
2
2
2
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
6
Table 3
Results of the catalytic runs with the Pd-Pt and Pd-Au bimetallic catalysts supported on K2621.
Cat.
Employed
Initial cumul.
rateb
Initial cumul.
productivity
Initial H2O2
productivity
Initial selectivitye
Selectivity at t1/2g
metal amounta
c,f
d,f
Pd + M f
ꢀ
vs Pd
vs Pd + M
ꢀf
Pd
1
1
1
1
PdK2621
1.45
1.44
1.44
1.42
–
29.3
73.8
77.7
81.3
ndh
2151
2267
2406
–
ndh
875
783
694
ndh
46
44
50 (38)
45 (25)
45 (22)
45 (21)
Pd025PtK2621
Pd05PtK2621
Pd1PtK2621
1.64
1.84
2.2
1891
1774
1553
45
1
1
1
Pd025AuK2621
Pd05AuK2621
Pd1AuK2621
1.43
1.42
1.42
1.63
1.81
2.19
45.9
45.1
55.4
1349
1334
1637
1184
1047
1062
739
249
588
62
24
55
55 (28)
22 (31)
52 (25)
5
−3 −1
−1
−1
−1
−1
ꢀ
a: mol·10 ; b: mmol(H2O2 + H2O)·dm
h
; c: mol(H2
O
2+H2O) mol
ꢀ
h
; d: mol(H2
O
2) mol
ꢀ
h
; e: %; f: M = Pt or Au; g: selectivity at 50% of H2 conversion (t1/2 min,
(Pd+M )
(Pd+M )
in parentheses); h: not determined (see text).
Fig. 6. H2 conversion and H2O2 selectivity over the Pd-Au catalysts as a function of
time.
Fig. 4. H2 conversion and H2O2 selectivity over the Pd-Pt catalysts as a function of
time.
the initial sections of the cumulative yield plots. Their compari-
son allows to analyze the influence of platinum and gold addition
on the activity of the catalysts. In addition, the catalyst selectivity
was compared both at the beginning of the reaction and at 50% H2
conversion (Table 3, “Selectivity at t1/2” column).
consumption. Accordingly the initial cumulative productivity (per
unit amount of total metal) undergoes a slight decrease with the
increase of the Pt content of the catalyst, because the relative
change of the absolute initial cumulative rates is smaller than the
relativechangein the sum of Pd and Pt amounts. In comparisonwith
There is only a little difference in the H2 conversion curves
over the Pd-Pt catalysts hence in the absolute initial rate of H2
1
PdK2621 the Pd-Pt bimetallic catalysts are initially more active, as
shown in Fig. 3 (due to the presence of an induction time in the plot
of the H2 conversion as a function of time, the RISLI method was
not applied to calculate the initial cumulative productivity).
The changes of the size and size distribution of the metal nano-
clusters from 1Pd05PtK2621 (3.3 nm) to 1Pd1PtK2621 (3.0 nm and
narrower distribution) seem to have only a little effect and, if any,
not the expected one. The reduction of the size, for instance, is not
associated to an increase of the activity. However, the productiv-
ity of the Pd-Pt catalysts is generally appreciably higher than that
of the monometallic catalyst, as also indicated by the shorter t1/2
.
The change of the Pt amount affects more or less equally the H O
2
and H O2 productivity, so that the selectivity of the Pd-Pt catalysts
2
appears to be almost independent of the composition of the metal
phase (Fig. 4). This holds in the comparison of both the initial selec-
tivity calculated with the RISLI method (Table 3, “Initial selectivity”
column) and of the values at 50% of H2 conversion (Table 4, right-
most column). From the point of view of selectivity the addition of
platinum has an adverse effect in this case. The initial selectivity of
1
PdK2621 was not calculated with the RISLI method, because the
concentrations of H O and H O did not increase linearly with time
2
2
2
at the beginning of the reaction due to the presence of an induction
time for both. However, bimetallic catalysts were always initially
less selective than the monometallic one, as shown by Fig. 4 (upper
panel), and at 50% conversion of H2 as well (Table 3). Apparently the
Fig. 5. Concentration of water and hydrogen peroxide produced over the Pd-Au
catalysts as a function of time.
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
7
addition of 0.25% (w/w) platinum, or more, suppressed the induc-
tion times for the production of both H O and H O . As the former
was longer than the latter, this made the bimetallic Pd-Pt cata-
lysts less selective. The related bimetallic Pd-Pt catalysts obtained
from the reduction of the precursors with formaldehyde were ini-
tially somewhat more productive, but even less selective (<40%)
for the latter set of catalysts to compare with, it can be appreciated
that 1Pd05AuK2621 is featured by the presence of some Au(I) in the
nanostructured metal phase and this could be also connected to the
selectivity decrease. On the other hand 1Pd1AuK2621 is practically
as selective as 1Pd025AuK2621. The presence of larger nanoclusters
in 1Pd1AuK2621, which has larger size and broader size distri-
bution than 1Pd025AuK2621, could at least in part explain this
recovery of selectivity upon increasing the gold content from 0.5
to 1% (w/w) [9].
To summarize, our data show that the addition of platinum to
the polymer-supported Pd catalysts yielded more active, but less
selective, materials when the platinum amount is 0.25% (w/w)
or higher. It is not clear, however, why the changes in the metal
composition of Pd-Pt catalyst, in the size and size distribution of
the metal nanoparticles did not influence appreciably the selectiv-
ity. We also found some differences between the catalysts (both
monometallic Pd and bimetallic Pd-Pt) obtained upon reduction of
2
2
2
[
9] than their counterparts reduced with H2 described herein. Also
the monometallic catalyst reduced with formaldehyde was pretty
less selective than 1PdK2621. From the point of view of the selec-
tivity, H2 seems therefore to be better than formaldehyde as the
agent for the reduction of the precursors of Pd and Pt. Also in the
case of the formaldehyde reduction route the bimetallic catalysts in
general were never more selective than the related monometallic
one, with the only exception of the material with 0.1% Pt [9]. This
relatively low level of Pt was not used in the present work; never-
theless the selectivity data of the new catalysts obtained from the
H2 reduction route confirm that platinum is useful to improve the
catalytic performance only in very little amount, in agreement with
our own [9] and other previous literature findings [12].
In comparison with platinum, the addition of gold has a similar
effect on the activity, which was higher than in the monometal-
lic catalyst, as shown by the kinetic plots of Fig. 6 (lower panel).
However the productivities are smaller than those of the bimetal-
lic Pd-Pt. This is reflected also in the values of t1/2, which are
intermediate between those of the Pd-Pt catalysts and the one
of the monometallic catalyst. In comparison with 1PdK2621, only
the metal precursors with H , described in this work, and related
2
materials which were obtained upon formaldehyde reduction of
the same precursors [9]. Whereas the catalysts reduced with H2
are less active, but generally more selective, at the highest Pt
contents, the bimetallic catalysts from the formaldehyde reduc-
tion route are practically as selective as the related monometallic
material. Moreover, the differences in the catalytic performance of
the Pd-Pt catalysts from the formaldehyde reduction route were
comparatively larger and were tentatively explained by product
poisoning (in particular by H O ). In the case of the H -reduced
1
Pd025AuK2621 was somewhat more selective. In fact its selectiv-
2
2
2
ity at 50% H2 conversion was higher and inspection of Fig. 6 (upper
panel) shows that it was likely just a bit more selective also in the
first 10–15 min of the reaction. With higher Au amounts, the Pd-
Au bimetallic catalysts described herein are never more selective
than 1PdK2621, unless relatively long reaction time are considered.
This is in line with the observation that relatively little amounts
of gold are required to have an effective promotion of palladium,
although positive effects have been observed up to a molar Pd:Au
ratio of 2:1 (corresponding approximately to 1% (w/w) for both
in the catalyst) [4,5,7,16]. As pointed out above, the conversion of
H2 over these Pd-Au catalysts is never quantitative and some pro-
duction of H O in the final stage still balanced its dismutation,
catalysts differences are too small to provide either further sup-
port or disputation of this hypothesis. These circumstances are in
agreement with previous reports [6,12,30] that the overall catalytic
performance seems to be much more dependent by the preparative
method used for synthetize the metal nanoparticles compared to
their size and size-distribution. This could be related to changes
in the topology or other features of the surface, when different
protocols of reduction are employed, possibly arising from differ-
ent morphological properties of the nanostructured metal phase.
To unravel these properties an in-depth investigation, which was
beyond the scope of this work, is required and will be the subject
of future work.
2
2
hence no final decrease of its concentration was observed. As the
consequence, 1Pd025AuK2621 and 1Pd1AuK2621 were apprecia-
bly more selective than 1K2621 when t > 60 min. In this respect,
the plots in the upper panel of Fig. 6 show that the selectivity of
The bimetallic Pd-Au catalysts described herein are even more
different from those obtained with formaldehyde as the reducing
agent of the (same) metal precursors. In the latter case, the catalyst
with 0.25% (w/w) Au added was appreciably less active than
the related monometallic catalyst, but much more selective. By
contrast 1Pd025AuK2621 is more active than 1PdK2621, but only
slightly more selective. Moreover, in the formaldehyde reduced
catalysts the activity and the selectivity underwent a steady
increase and decrease, respectively, as the gold content was
increased. As shown above this was not the case with the catalysts
1
PdK2621 decreased faster than that of all the Pd-Au catalysts. The
result is that by the time that H O2 reached its respective top con-
2
centrations over 1PdK2621, 1Pd025AuK2621 and 1Pd1AuK2621,
the selectivity of the bimetallic catalysts had surpassed that of
the monometallic one. With the Pd-Au catalysts too there is not
a clear correlation between the size and size distribution of the
metal nanoparticles and the catalytic performance. Also in the case
of Pd-Au bimetallic catalysts obtained upon reduction of the metal
precursors with formaldehyde [9] we did not find compelling evi-
dence that the nanoparticle size was distinctly more important
than other factors in directing the catalysts’ performance, in spite
of the fact that they had sizes and size distributions pretty much
reduced with H described in this work. It is not clear at the moment
2
why different reducing agents give catalysts performing differ-
ently, starting from the same support and the same precursors.
TEM characterization did not provide useful information and again
the reasons could be connected to differences in the structures
of the metal nanoparticles, which could arise, for instance, from
different reduction rates of the precursors with different reducing
agents with different ability to diffuse inside the support [31].
different from those of the H -reduced catalysts investigated in the
2
present work. However, we argued there that the presence of large
aggregates could have beneficial effects on the selectivity [9].
This could help in explaining, at least in part, some differences
in the behavior of the Pd-Au catalysts obtained with H (this work)
4. Conclusion
2
and with formaldehyde [9] as the reducing agent. With the latter,
we observed a steady decrease of the initial selectivity with the
increase of the Au amount. In line with this, 1Pd05AuK2621 is also
less selective than 1Pd025AuK2621. The difference in selectivity
here is higher than in the case of the related catalysts from the
formaldehyde reduction route. Although we do not have XPS data
Two sets of bimetallic Pd-Pt (Pd: 1.0; Pt: 0.25–1.0%, w/w) and
Pd-Au (Pd: 1.0; Au: 0.25–1.0%, w/w) catalysts supported on the
macroreticular ion-exchange resin Lewatit K2621 were prepared
by means of a simple ion-exchange procedure in water from
suitable precursors and reduction under hydrogen pressure (5 bar)
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2

G Model
CATTOD-8912; No. of Pages8
ARTICLE IN PRESS
S. Sterchele et al. / Catalysis Today xxx (2014) xxx–xxx
8
◦
at 60 C. It was found that the addition of a small amount of gold
(
[2] J. Garcia-Serna, T. Moreno Rueda, P. Biasi, M.J. Cocero, J.-P. Mikkola, T. Salmi,
Green Chem. (2014) http://dx.doi.org/10.1039/C3GC41600C
0.25%) makes the catalyst more active and slightly more selective
than a 1%, w/w, palladium catalyst on K2621. 1Pd025AuK2621 was
[
3] Q. Liu, J.C. Bauer, R.E. Schaak, J.H. Lunsford, Angew. Chem., Int. Ed. 47 (2008)
221–6224.
6
−
1
−1
the best one, with 1184 mol(H2) mol
h
initially consumed,
initially produced, a 55% selectivity
towards H O at 50% conversion of H . This performance is
[4] J.K. Edwards, N.E. Ntainjua, A.F. Carley, A.A. Herzing, C.J. Kiely, G.J. Hutchings,
Angew. Chem., Int. Ed. 48 (2009) 8512–8515.
[5] G. Bernardotto, F. Menegazzo, F. Pinna, M. Signoretto, G. Cruciani, G. Strukul,
Appl. Catal. A: Gen. 358 (2009) 129–135.
(Pd+Pt)
−
1
−1
7
39 mol(
mol
h
H2O2
)
(Pd+Pt)
2
2
2
[
[
[
[
6] F. Menegazzo, P. Burti, M. Signoretto, M. Manzoli, S. Vankova, F. Boccuzzi, F.
Pinna, G. Strukul, J. Catal. 257 (2008) 369–381.
7] J.K. Edwards, B. Solsona, N.E. Ntainjua, A.F. Carley, A.A. Herzing, C.J. Kiely, G.J.
Hutchings, Science (Washington, DC, U.S.) 323 (2009) 1037–1041.
8] C. Burato, P. Centomo, M. Rizzoli, A. Biffis, S. Campestrini, B. Corain, Adv. Synth.
Catal. 348 (2006) 255–259.
lower than that obtained with a related catalyst obtained by
using formaldehyde as the reducing agent, which was at the
same time more active and selective. No obvious trend appeared
when the amount of Au was increased to 0.5 and 1.0% (w/w).
In fact 1Pd05AuK2621 (0.5% Au, w/w) was poorly selective, pos-
sibly because Au was not completely reduced. Accordingly, no
clear connection between the nanocluster size and size distri-
bution and the catalytic performance was found, although the
presence of relatively high proportions of large nanoclusters in
Pd1AuK2621 could help in explaining why it is almost as selective
as 1Pd025AuK2621. On the other hand 1Pd025PtK2621, the best
Pd-Pt catalyst, initially consumed 1891 mol(
produced 875 mol(H2O2) mol−1
9] S. Sterchele, P. Biasi, P. Centomo, P. Canton, S. Campestrini, T. Salmi, M. Zecca,
Appl. Catal. A: Gen. 468 (2013) 160.
[10] L.W. Gosser, US Patent 4,681,751 (1987).;
L.W. Gosser, J.T. Schwartz, US Patent 4,772,485 (1988).
[
[
11] G. Paparatto, R. D’Aloisio, G. De Alberti, R. Buzzoni, US Patent 6,630,118 (2003).
12] Q. Liu, J.C. Bauer, R.E. Schaak, J.H. Lunsford, Appl. Catal. A: Gen. 339 (2008)
130–136.
1
[
13] P. Biasi, F. Menegazzo, F. Pinna, K. Eranen, T.O. Salmi, P. Canu, Chem. Eng. J.
176–177 (2011) 172–177.
−
1
−1
mol
h
and
H2
)
(Pd+Pt)
[14] C. Burato, S. Campestrini, Y.-F. Han, P. Canton, P. Centomo, P. Canu, B. Corain,
Appl. Catal. A: Gen. 358 (2009) 224–231.
15] G. Blanco-Brieva, E. Cano-Serrano, J.M. Campos-Martin, J.L.G. Fierro, Chem.
Commun. (2004) 1184–1185.
−1
h
, but its selectivity towards
(Pd+Pt)
[
H O at 50% conversion of H2 was 45%. Further increase of the
2
2
platinum amount changed the initial rates values to more or less
the same extent, so that at 50% conversion of H2 all the Pd-Pt
catalysts turned out to be equally selective.
[16] P. Centomo, P. Canton, M. Ferroni, M. Zecca, New J. Chem. 34 (2010) 2956.
[
[
17] B. Corain, M. Zecca, P. Canton, P. Centomo, Phil. Trans. R. Soc. 368 (2010) 1495.
18] B. Corain, K. Je rˇ abek, P. Centomo, P. Canton, Angew. Chem., Int. Ed. 43 (2004)
959.
It can be finally recalled that reactions under batch-wise condi-
tions imply non-steady-state kinetics. Therefore semi-continuous
or continuous reactors could be best suited to better assess, under
steady-state conditions, the effect of Pt or Au addition to Pd.
[19] N.E. Ntainjua, M. Piccinini, J.C. Pritchard, J.K. Edwards, A.F. Carley, J.A. Moulijn,
G.J. Hutchings, Chem. Sus. Chem. 2 (2009) 575–580.
[
[
20] B.P. Block, J.C. Bailar, J. Am. Chem. Soc. 73 (1951) 4722–4725.
21] T. Inoue, M.A. Schmidt, K.F. Jensen, Ind. Eng. Chem. Res. 46 (2007) 1153–1160.
[22] C.M. Piqueras, J. Garcia-Serna, M.J. Cocero, J. Supercrit. Fluids 56 (2012) 33–40.
[
[
[
[
23] P. Biasi, N. Gemo, J.R. Hernandez Carucci, K. Eranen, P. Canu, T.O. Salmi, Ind.
Eng. Chem. Res. 51 (2012) 8903–8912.
24] J. Aitchison, J.A.C. Brown, The LogNormal Distribution, Cambridge University
Press, 1957.
Acknowledgments
25] S. Melada, F. Pinna, G. Strukul, S. Perathoner, G. Centi, J. Catal. 237 (2006)
Lanxess is kindly acknowledged for providing the K2621 resin.
S. Sterchele gratefully acknowledges the Johan Gadolin Scholarship
for financial support. P. Biasi gratefully acknowledges the Kempe
Foundations for financial support. In Sweden, also the Bio4Energy
program is acknowledged. This work is also a part of the activities of
the Process Chemistry Centre (PCC), a centre of excellence financed
213–219.
26] Y. Liu, M. Chi, V. Mazumder, K.L. More, S. Soled, J.D. Henao, S. Sun, Chem. Mater.
23 (2011) 4199–4203.
[27] J. Pollmann, R. Franke, J. Hormes, H. Bnnemann, W. Brijoux, A. Schulze Tilling,
J. Electron. Spectrosc. Relat. Phenom. 94 (1998) 219.
[
28] B. Pawelec, A.M. Venezia, V. La Parola, E. Cano-Serrano, J.M. Campos-Martin,
J.L.G. Fierro, Appl. Surf. Sci. 242 (2005) 380.
˚
[29] A.J. Renouprez, A. Malhomme, J. Massardier, M. Cattenot, G. Bergeret, Stud. Surf.
Sci. Catal. 130 (2000) 2579.
by Abo Akademi University.
[
30] Y. Ye, J. Chun, S. Park, T.-J. Kim, Y.-M. Chung, S.-H. Oh, I.-K. Song, J. Lee, Korean
J. Chem. Eng. 29 (2012) 1115–1118.
References
[
31] K. Je rˇ ábek, M. Zecca, P. Centomo, F. Marchionda, L. Peruzzo, P. Canton, E. Negro,
[1] J.M. Campos-Martin, G. Blanck-Brieva, J.L.G. Fierro, Angew. Chem., Int. Ed. 45
V. Di Noto, B. Corain, Chem. Eur. J. 19 (2013) 9381.
(
2006) 6962.
Please cite this article in press as: S. Sterchele, et al., The effect of the metal precursor-reduction with hydrogen on a library of bimetallic
Pd-Au and Pd-Pt catalysts for the direct synthesis of H O , Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.02.021
2
2