P. Rouge, et al.
MolecularCatalysis471(2019)21–26
[23–27]. Nevertheless, the impregnation procedure does not lead to the
exclusive formation of bimetallic particles. The co-existence of isolated
tin and platinum monometallic phase may give low selectivity in de-
sired product in particular for the dehydrogenation of hexanes by iso-
merisation and aromatisation reactions [27,28]. Preparation techniques
play a significant role in controlling the type of materials obtained.
Surface organometallic chemistry on oxide (SOMC/Oxide) and metals
(SOMC/M) leads to well-defined bimetallic catalysts whose active sites
have high homogeneity in their nature [8]. On the other hand, im-
proving selectivity also requires inhibiting the acid-catalyzed reactions
on the support (cracking, isomerisation, aromatisation and poly-
merisation). A most common method is to introduce additives, such as
basic alkaline elements.
The reaction of a tin organometallic compound with metallic sur-
faces has been pioneered by Yermakov and Ryndin in the 1970s
[29,30]. This surface organometallic chemistry approach has been ex-
tensively studied in our laboratory for isobutane dehydrogenation. In
fact platinum and rhodium nanoparticles carried on silica have been
succesfully modified by tin or germanium [31–34]. The latter synthetic
approach affords selective localisation of the Sn or Ge atoms on the Pt
nanoparticles, as previously confirmed by Mössbauer and EXAFS
spectroscopies. The resulting catalysts, prepared by SOMC on metal,
display high selectivity and stability during isobutane dehydrogenation
under hydrogen due to well-defined bimetallic Pt-Sn phase. In the lit-
erature, there are only few examples reporting the selective dehy-
drogenation of 2,3-dimethylbutane [35–37]. However, no example re-
ports the conversion of 2,3-dimethylbutane into 2,3-dimethylbutenes
on γ-alumina supported bimetallic PtSn catalyst prepared by SOMC.
In this paper, we demonstrate a stable and selective catalyst for the
dehydrogenation of dimethylbutanes to 2,3-dimethylbutenes based on
Pt nano-particles supported on alumina. Furthermore, different pro-
moting agents have been employed through the SOMC methodology on
alumina and on the platinum nanoparticle in order to identify the most
performing catalyst.
CM120 instrument (localised at Centre Technologique des Micro-
structures, Université Claude Bernard, Lyon 1) with an acceleration
voltage up to 120 kV. The particle size distribution was obtained from
TEM pictures calculating the surface average particle diameter from dp
= Σnidi3/Σnidi2. Nitrogen adsorption/desorption isotherms of samples
were obtained using a volumetric adsorption analyzer (ASAP2020) at
−196 °C (77 K). Before the adsorption analysis, the samples were de-
gassed for 3 h at 200 °C. The O2/H2 titration has been performed ac-
cording to published procedure [39] using a Belcat B device from BEL
Japan. After reduction in H2 at 400 °C, the experiment was conducted at
80 °C. The dispersion in the monometallic catalyst was calculated from
the value of hydrogen uptake assuming that the surface stoichiometry
of the irreversibly chemisorbed hydrogen is unity. The average particle
size of platinum was calculated from the dispersion values, assuming
cubic shapes for the platinum particles.
2.2. Grafting of Pt(acac)2 on alumina and characterisation of Pt/Al2O3
A solution of 0.116 g de Pt(acac)2 (Aldrich) in 10 ml of toluene was
added to 2.9 g of γ-Al2O3-(500). After stirring the suspension overnight at
room temperature, the solvent was distillated and the material dried
under high vacuum (10−5 mbar). The solid was next calcinated at
400 °C under synthetic air and reduced under hydrogen at 550 °C for 4 h
(heating rate 1 °C/min). Elemental analysis: 2 wt%Pt.
2.3. Preparation and characterisation of Li-Al2O3
A solution of 1 mL de n-butyl lithium (Aldrich) (2 M in cyclohexane)
in 10 ml of cyclohexane was added to 3 g of Al2O3-500. After stirring the
suspension overnight at room temperature, the solvent was washed 3
times before distillation under high vacuum (10−5 mbar). Elemental
analysis: 2.8 wt%Li.
2.4. Grafting of Pt(acac)2 on Li-Al2O3) and characterisation of Pt/Li-Al2O3
2. Experimental
A solution of 0.115 g de Pt(C5H7O2)2 (Aldrich) in 10 ml of toluene
was added to 2.9 g of Li2.8%-Al2O3-(500). After stirring the suspension
overnight at room temperature, the solvent was distillated and the
material dried under high vacuum (10−5 mbar). The solid was next
calcinated at 400 °C under air and reduced under hydrogen at 550 °C for
4 h (heating rate 1 °C/min). Elemental analysis: 2 wt%Pt and 2.8 wt%Li
2.1. General procedures
All experiments were carried out by using standard Schlenk and
glove-box techniques. Solvents were purified and dried according to
standard procedures. Al2O3-(500) was prepared with Alumina from
Evonik (specific area of 100 m2 g−1), which was partly dehydroxylated
at 500 °C under high vacuum (10-5 Torr) for 15 h. Pt(acac)2, BuLi in
hexane, Sn(Bu)4 and Ge(Bu)4 were purchased from Aldrich. Gas-phase
analyses were performed on a Hewlett-Packard 5890 series II gas
chromatograph equipped with a flame ionization detector and an
Al2O3/KCl on fused silica column (50 m × 0.32 mm). Elemental ana-
lyses were performed by the Mikroanalytisches Labor Pascher,
Remagen, Germany. IR spectra were recorded on a Nicolet 6700 FT-IR
spectrometer by using a DRIFT cell equipped with CaF2 windows. The
samples were prepared under Ar within a glove-box. Typically, 64 scans
were accumulated for each spectrum (resolution 4 cm-1). Solid-state
NMR spectra were acquired on Bruker Avance 500 and Bruker Avance
III 800 spectrometers (1H: 800.13 MHz,). For 1H NMR experiments
(18.8 T), the spinning frequency was 20 kHz, the recycle delay was 10 s,
and 16 scans were collected using a 90° pulse excitation of 2.25 μs. The
13C CP-MAS NMR spectrum was obtained at 11.74 T using the CP-MAS
pulse sequence and a high power 1H decoupling at a RF field amplitude
of 70 kHz. 1H-1H Double Quantum Magic-Angle Spinning spectra were
2.5. Preparation and characterisation of PtSn/Li-Al2O3
In double schlenck, 1.2 g of Pt2%/Li2.8%-Al2O3 was suspended in
5 mL heptane. Then, a solution of 20 μL of Sn(C4H9)4 (Aldrich) dis-
solved in 15 mL heptane was added. After freezing of the liquid in liquid
nitrogen and evacuation of argon, 500 Torr of H2 was introduced. The
solution was stirred for 24 h at 80 °C. The solid was then washed 3 times
with heptane before distillation under high vacuum (10−5 mbar). The
material was finally reduced under hydrogen for 2 h at 550 °C (heating
rate 1 °C/min). Elemental analysis: 2 wt%Pt, 2.8 wt%Li, 1.32 wt%Sn.
2.6. Preparation of PtGe/Li-Al2O3
In double schlenck, 0.9 g of Pt/Li-Al2O3 was suspended in 5 mL
heptane. Then, a solution of 12 μL of HGe(C4H9)3 (Aldrich) dissolved in
15 mL heptane was added. After freezing of the liquid in liquid nitrogen
and evacuation, 500 Torr of H2 was introduced. The solution was stirred
for 24 h at 80 °C. Next the solid was washed 3 times with heptane before
distillation under high vacuum (10−5 mbar). The material was finally
reduced under hydrogen for 2 h at 550 °C (heating rate 1 °C/min).
Elemental analysis: 2 wt%Pt, 2.8 wt%Li, 0.8 wt%Ge.
5
acquired at 20 kHz spinning speed using the R122 symmetry-based
recoupling scheme [38] applied for 200 μs at an RF field strength of
40 kHz. The recycle delay was set to 10 s and 16 transients were added
for each of the 300 t1 increments. Chemical shifts are given in ppm with
respect to TMS as external reference for 1H and 13C. Transmission
electron microscopy (TEM) observations were carried out on a Philips
22