K. Pattamakomsan et al. / Catalysis Today 164 (2011) 28–33
29
◦
◦
in O2 at 350 C for 2 h, and reduced in H2 at 500 C for
According to our previous study [9], the MA alumina consisted
2
h.
of 80 wt% -phase and 20 wt% of ␣-phase alumina while AA was
1
00 wt% ␣-phase alumina. The XRD characteristic peaks corre-
sponding to Pd, Pd–Sn intermetallic phase, and SnO2 could not be
detected for all the catalyst samples due probably to low metal
loading and/or the very small particle size. Chemical analysis (ICP)
showed that Pd wt% of Pd/MA and Pd–Sn/MA were similar at
3
. Catalyst characterization
The XRD patterns were collected using an X-ray diffractometer,
◦
SIEMENS XRD D5000, with Cu K␣ radiation with a Ni filter in 10–80
0
.5 wt%. A slightly less Pd content was found on the catalysts sup-
2
ꢀ angular regions. The composition of catalysts was determined
using an inductively coupled plasma-optical emission spectroscopy
ICP-OES) Activa (Jobin-Yvon). The BET surface area of Al O3 sup-
ported on AA (∼0.3–0.4 wt%). However, Sn wt% for Pd–Sn catalysts
on both alumina supports were 0.2% yielding different atomic ratios
(
2
(Pd/Sn = 3 for Pd–Sn/MA and Pd/Sn = 2 for Pd–Sn/AA).
ports were determined by N physisorption using a Micrometritics
2
TEM micrographs illustrating the morphology of Pd and Pd–Sn
ASAP 2000 automated system. The particle size and particle size
distribution were determined by transmission electron microscopy
using a JEOL JEM 2010, operating at 200 kV, equipped with a LaB6
tip, a high resolution pole piece (0.196 nm point resolution), and a
Pentafet-LinK ISIS EDX spectrometer (Oxford Instruments). Local
chemical composition of the nanoparticles was determined by
energy-dispersive X-ray (EDX) spectroscopy.
XPS analysis was performed with a Kratos Axis Ultra DLD spec-
trometer, equipped with a hemispherical analyzer, and a state of
the art delay line detector. A monochromated Al–K␣ X-ray source
with charge neutralization was used. In order to reduce the sample
a reaction chamber coupled to the ultra-high vacuum XPS chamber
◦
supported on MA and AA alumina after reduction at 500 C for 2 h
are shown in Fig. 1. The particle size distributions were examined
over 250 particles. The nanoparticles supported on MA alumina
were for both samples (Pd/MA and Pd–Sn/MA) around 2 nm with all
particles fitting in a very narrow range (1–3 nm). The nanoparticles
supported on AA alumina are not so homogeneous in size. Indeed
we can observe that together with smaller particles (1–3 nm)
we also have a distribution of larger particles. These ranges are
between 4 and 6 nm for Pd/AA and 4 and 10 nm for Pd–Sn/AA.
EDX analysis was performed on individual particles in the case
of the larger ones and on small groups of (3–5) nanoparticles for the
smaller particles. These EDX analysis results are shown in Fig. 2 for
Pd–Sn/MA and Pd–Sn/AA. The results suggest that alloy nanopar-
◦
allowed to heat the samples in H2 flow at 500 C for 2 h and then
cool down to room temperature before analysis. The samples can
thus be transferred for XPS analysis without air contact. The Al 2p
line was taken as an internal standard at 73.4 and 73.6 eV for AA
and MA respectively. The error in BE measurements was ± 0.2 eV
for all catalysts.
Surface investigation was performed by in situ FT-IR spectro-
scopies of adsorbed CO. 40 mg of reduced catalysts were deposited
directly on the CaF plate. The samples were pretreated as follows:
ticles with Pd Sn stoichiometry were formed for the bimetallic
3
catalysts regardless of alumina support. However the chemical
composition of the bimetallic nanoparticles was much more homo-
geneous on MA alumina than on AA alumina in which the size
distribution was larger. For Pd–Sn/MA we could never detect Sn
that was not associated with Pd which strongly suggests that no
tin or tin oxide nanoparticles were formed independently of the
bimetallic nanoparticles. Conversely, we cannot rule out the possi-
bility of the presence of independent Sn species on the support for
Pd–Sn/AA.
2
−
4
(
i) outgassing at room temperature for 15 min to 1.33 × 10 Pa, (ii)
◦
◦
heating under vacuum from 20 to 500 C at heating rate 20 C/min,
◦
(
iii) reduction under H2 at 500 C for 15 min, (iv) outgassing for
The surface composition and electronic properties of Pd and Sn
on the various alumina supported Pd and Pd–Sn catalysts were
investigated by XPS. The XPS spectra for Pd 3d of all the catalysts
3
0 min, and then cool down to room temperature before introduc-
◦
tion of CO 667 Pa at 20 C. Carbon monoxide temperature program
desorption (CO-TPD) was carried out after evacuation at room tem-
perature for 1 h at heating rate 5 C/min and recorded every 3 min
◦
◦
after reduction at 500 C in hydrogen for 2 h are shown in Fig. 3(a).
The binding energy of Pd 3d5/2 is in the range of 334.9–335.3 eV,
indicating the presence of metallic palladium species [5]. For the
Pd monometallic supported on MA and AA, the binding energy
of Pd 3d5/2 were similar with a maximum at 334.9 eV, indicating
that the phase of alumina and Pd particle size had no influence
on the electronic property of palladium catalysts [9]. However,
the binding energy of Pd 3d5/2 for Pd–Sn bimetallic on both alu-
mina supports is slightly shifted towards higher binding energy
with a maximum at 335.3 eV suggesting that tin addition modi-
fies the electronic properties of Pd catalysts and that bimetallic
alloy particles are formed. The deconvoluted XPS spectra of Sn 3d5/2
for Pd–Sn/MA and Pd–Sn/AA are shown in Fig. 3(b). Both catalysts
show two similar components of tin species. The higher binding
energy (486.4–486.7 eV) could be related to 32–33 wt% of tin oxide
until CO were completely desorbed. IR spectra were recorded using
−
1
a Fourier-transform Nexus Thermo Nicolet at a resolution of 2 cm
.
CO gas phase absorbance and background were subtracted from all
spectra.
4
. Reaction study
The selective hydrogenation of 1,3-butadiene was carried out
in a fixed bed flow stainless steel reactor at atmosphere pressure.
Before reaction, 50 mg of catalyst was reduced in situ with hydro-
◦
gen by heating from room temperature to 500 C at a heating rate
◦
of 5 C/min. Then the reactor was cooled down in helium to room
temperature. The feed flow rate was adjusted to 100 ml/min with
automatic flow controller (Brooks Instruments) and composed of
remaining even if the catalysts were reduced in hydrogen flow at
2
% 1,3-butadiene, 2% H , and balance with helium. The reaction
◦
2
5
00 C for 2 h. This observation together with the fact that no Sn
◦
temperature was adjusted in the range 20–70 C. The feed and prod-
uct composition were analyzed by a gas chromatograph equipped
with an FID detector (Intersat IGC 120 FB, pack column 0.19% Picric
Acid on Graphpac(tm)-GC, 80/100, 7 ft × 1/8 in. SS Tubing).
was found without Pd by EDX suggests that the remaining unre-
duced tin is within the bimetallic nanoparticles. On the other hand,
6
7–68 wt% of tin component observed at the lower binding energy
in the range 484.8–484.9 eV could be assigned to the presence of tin
alloy [10].
5
. Results and discussion
The IR spectra of adsorbed CO 667 Pa at room temperature
(Fig. 4) and the IR spectra of CO desorbed at different temper-
5
.1. Catalyst characterization
atures in vacuum (not given here) were investigated to further
characterize the catalysts and determine the adsorption strength
and adsorbed species on Pd and Pd–Sn catalysts supported on MA
and AA.
The Pd and Pd–Sn bimetallic catalysts supported on MA and AA
are denoted herein as Pd/MA, Pd–Sn/MA, Pd/AA, and Pd–Sn/AA.