108
T. Buntara et al. / Catalysis Today 210 (2013) 106–116
preparation of other Rh-based bimetallic catalysts, except for
Rh–ReOx/C, where the calcination step was omitted.
dispersive microanalysis. The mean metal particle size (W) was
determined using Gatan Digital Micrograph software.
Rh–ReOx/SiO2 catalysts with different rhodium loadings and
The density of OH groups was determined by a method reported
by Blin and Carteret [44] using thermogravimetric analysis (TGA).
The TGA patterns were recorded using a Mettler-Toledo analyzer
(TGA/SDTA851e) using fused alumina crucibles and a flow of syn-
thetic dry air of 100 mL/min (STP). The temperature was increased
from 30 to 900 ◦C at 5 ◦C/min with a plateau of 1 h at 130 ◦C. Blank
curve subtraction using an empty crucible was employed.
Textural analysis of the fresh and spent catalysts was carried
out by N2 physisorption analyses at −196 ◦C on a Micromeritics
ASAP 2420 analyzer. Prior to the measurements all samples were
degassed under vacuum at 200 ◦C for 10 h for the fresh materials.
The surface area was calculated by the BET method [45] in the rel-
ative pressure range of 0.06–0.25 (SBET). The single point total pore
volume (VT) was estimated from the amount adsorbed at a relative
pressure of 0.98 in the desorption branch. The pore size distribu-
tion was derived from the BJH model [46]. Micropore volume (Vꢀ)
was derived from the t-plot method [47].
Temperature-programmed reduction (TPR) with H2 was carried
out in a Micromeritics Autochem II instrument, using a mixture of
5 vol% H2 in Ar to reduce the samples by mixing high purity H2 and
Ar. Silver oxide (99.999%) was used for calibration. The sample was
not dried before analysis as the gas at the reactor outlet was passed
through a cold trap consisting of acetone–dry ice; this removes both
the physisorbed water as well as the water produced during the
reduction before entering the thermal conductivity detector.
X-ray photoelectron spectroscopy (XPS) was used to study
the chemical composition and the oxidation state of the ele-
ments on the catalyst surface. The XPS instrument, a VG Escalab
200 R spectrometer with a MgK␣ X-ray source (hꢁ = 1253.6 eV;
1 eV = 1.602 × 10−19 J), was equipped with a pre-treatment cham-
ber with controlled atmosphere and temperature in which the
catalyst samples could be treated under various conditions.
Re/Rh molar ratio were also prepared.
A combination of
rhodium(III) chloride hydrate and ammonium perrhenate(VII) of
(a) 171 mg (0.8 mmol) and 29 mg (0.1 mmol), (b) 183 mg (0.9 mmol)
and 235 mg (0.9 mmol), (c) 41 mg (0.2 mmol) and 27 mg (0.1 mmol),
and (d) 302 mg (1.4 mmol) and 193 mg (0.7 mmol) was used to
prepare catalysts with (a) 4 wt% Rh and Re/Rh of 0.13, (b) 4 wt%
Rh and Re/Rh of 1, (c) 1 wt% Rh and Re/Rh of 0.5, and (d) 6.5 wt% Rh
and Re/Rh of 0.5.
Monometallic Rh/SiO2 (4 wt%) and Re/SiO2 (4 wt%) catalysts
were also produced using 169 mg (0.8 mmol) of rhodium(III)
chloride hydrate and 120 mg (0.4 mmol) of ammonium perrhen-
ate(VII), respectively. Additional types of precursors of Rh and Re
were also investigated. A combination of (a) 176 mg (0.8 mmol)
of rhodium(III) chloride hydrate and 163 mg perrhenic(VII) acid
solution (0.4 mmol perrhenic acid) and (b) 243 mg (0.8 mmol) of
rhodium(III) nitrate hydrate and 106 mg (0.4 mmol) of ammonium
perrhenate(VII) was used to produce Rh–ReOx/SiO2 catalysts with
4 wt% Rh and Re/Rh molar ratio of 0.5. For some Rh–ReOx/SiO2 cat-
alysts (as stated in the text), the silica was precalcined at 773 K for
3 h prior to the impregnation.
Catalysts with various metal combinations were prepared
as well. Rhodium(III) chloride hydrate was combined with: (a)
chromium(III) nitrate nonahydrate (164 mg, 0.4 mmol), (b) man-
ganese(II) nitrate tetrahydrate (103 mg, 0.4 mmol), (c) tin(II)
chloride dihydrate (94 mg, 0.4 mmol), (d) ammonium molybdate
tetrahydrate (131 mg, 0.1 mmol), and (e) ammonium tungsten
oxide (30 mg, 0.1 mmol) to produce the corresponding bimetallic
catalysts. Non-Rh bimetallic catalysts were prepared from (a) tetra-
ammine platinum(II) nitrate (169 mg, 0.4 mmol) and ammonium
perrhenate(VII) (58 mg, 0.2 mmol) and (b) tetra-ammine palla-
dium(II) nitrate solution (1.7 g) and ammonium perrhenate(VII)
(109 mg, 0.4 mmol) providing Pt–ReOx/SiO2 and Pd–ReOx/SiO2 cat-
alysts, respectively.
2.2.4. DFT calculations
Gradient-corrected density functional theory (DFT) calculations
were used to determine the gas phase carbenium ion reaction
energies as well as reaction energies and activation barriers over
Rh200Re1OH clusters which are comprised of 200 Rh atoms, 1 Re
atom and 1 OH group. The gas phase energies were carried out
using DFT calculations implemented in DMol3 [48,49]. Numerical
basis sets of double numerical quality (DNP) with d-type polar-
ization functions were used for wave functions with the cutoff of
2.2.2. General procedure for the reaction of 2–3
2 (100 mg, 0.76 mmol), the Rh–ReOx/SiO2 catalyst (25 mg),
water (2 mL) and a Teflon stirring bar were added to an 8 mL glass
vial capped with a septum; the septum was pierced with a small
needle. The vial was placed in a stainless-steel autoclave, the auto-
clave was closed and the stirring was started at 1000 rpm After
three times pressurizing with first nitrogen and then hydrogen, the
autoclave was pressurized with hydrogen to 10 bar and the tem-
perature was raised to 120 ◦C. After 1 h, the pressure was raised to
80 bar and the reaction was continued for 4 h. Then, the autoclave
was allowed to cool to ambient temperature and the pressure was
released. The reactor content was filtered over a PTFE membrane to
remove the catalyst and the filtrate was subjected to GC analysis.
Adjustments were made for some experiments (when stated in
the text): (1) 1-propanol as the solvent, (2) temperature of 80 and
180 ◦C, (3) reaction time of 20 h, (4) catalyst intake of 10 wt% and
(5) no pre reduction step for 1 h at 10 bar hydrogen pressure.
˚
3.5 A. The Perdew-Wang 91 [50] (PW91) form of generalized gra-
dient approximation (GGA) was used to model the correlation and
exchange energies and gradient corrections. The electronic density
was converged to within 1 × 10−5 au. The energy in each geometry
optimization cycle was converged to within 2 × 10−5 Hartree with
the maximum displacement of 4 × 10−3 A and the maximum force
˚
of 3 × 10−3 Hartree/Å.
The gas-phase reaction energies for formation of carbenium ion
for ring structures were calculated as:
ROR + H+ → RORH+
2.2.3. Characterization of the catalyst and supports
where ROR and RORH+ refer to the initial ring structure and the
ring opened carbenium ion.
The metal particle size was determined by TEM analysis. The
samples were ground and the resulting fine powder was mixed
with isopropanol. One drop was dispersed onto a carbon TEM-grid
and left to dry for 15 min. Micro structural characterization was
carried out using a JEOL 2010F Transmission Electron Microscope
(TEM). The TEM micrographs were recorded with a Gatan digital
camera with Digital Micrograph. The X-rays were recorded by a
Bruker 5000 series of liquid N2-free XFlash® Silicon Drift Detec-
tors (SDD) and ESPRIT software to provide quantitative energy
The calculations on the Rh200Re1OH clusters were carried out
using periodic plane wave DFT calculations implemented in the
Vienna Ab Initio Simulation Program (VASP) [51–53]. The PW91
form of the GGA was used to determine gradient corrections to
the exchange and correlation energies. Wave functions were con-
structed by expanding a series of plane waves within a cutoff energy
of 400 eV Vanderbilt ultrasoft pseudopotentials [54] were used to
model the interactions between the core and valence electrons.