2
C.M.A. Parlett et al. / Catalysis Communications xxx (2013) xxx–xxx
transport. An alternative approach to stabilising highly dispersed
palladium is to select a support material with a greater affinity for
the element, such as alumina [25] or silica-alumina [26], wherein
Al-rich supports favour sub-2 nm (or even atomically-dispersed)
palladium.
2.3. Materials characterisation
Nitrogen porosimetry was performed on a Quantachrome Nova 1200
porosimeter using NovaWin 2 v2.2 analysis software. Samples were
degassed at 120 °C for 2 h prior to N2 physisorption. Powder X-ray dif-
fraction (XRD) patterns were recorded on a PANalytical X'Pert PRO dif-
fractometer fitted with an X'Celerator detector, using a Cu Kα (1.54 Å)
excitation source, with patterns calibrated against a Si standard. Trans-
mission electron microscopy (TEM) and high angle annular dark field
scanning transmission electron microscopy (HAADF STEM) images
were recorded on a FEI Tecnai F20 FEG TEM operated at 200 kV and
equipped with a Gatan Orius SC600A CCD camera. Samples were pre-
pared by dispersion in methanol and drop-casting onto a copper grid
coated with a holey carbon support film (Agar Scientific Ltd). Images
were analysed using ImageJ 1.41 software. Actual Pd loadings were de-
termined by MEDAC Analytical and Chemical Consultancy Service Ltd,
with samples digested in HF prior to ICP analysis. X-ray photoelectron
spectroscopy (XPS) was performed on a Kratos Axis HSi X-ray photo-
electron spectrometer fitted with a charge neutralizer and magnetic fo-
cusing lens, using monochromated Al Kα radiation (1486.6 eV). Spectral
fitting was performed with CasaXPS version 2.3.14. Binding energies
were referenced to the adventitious C 1 s peak at 284.6 eV. Pd 3d XP
spectra were fitted adopting a common asymmetric peak shape deter-
mined from that of a palladium oxide standard, and a spin-orbit doublet
separation of 5.3 eV in agreement with literature values from the NIST
surface database. Pd dispersion was measured via CO pulse chemisorp-
tion on a Quantachrome ChemBET 3000 system. Samples were
outgassed at 150 °C under flowing He (20 cm3 min−1) for 1 h, prior
to reduction at 100 °C under flowing hydrogen (20 cm3 min−1) for
1 h before analysis at room temperature.
In contrast to silicas and carbons, the synthesis of high surface area
mesostructured aluminas is relatively undeveloped, with early
surfactant-templating approaches employing carboxylic acids achiev-
ing high surface areas (760 m2 g−1), but small (2 nm diameter) ran-
domly ordered mesopores [27]. Large mesopore (10 nm diameter)
γ-alumina has been fabricated through the use of room temperature
ionic liquids as a co-solvent and template, however this material
comprises randomly debundled nanofibres embedded in a wormlike
porous network, and hence presents a disordered environment for
the preparation of uniformly dispersed catalytic centres and reac-
tant/product diffusion. Evaporation-induced self assembly (EISA)
utilising poly(alkylene oxide) block copolymers affords a simple
route to highly ordered 2D hexagonal (p6mm) mesoporous materials
[28], and has been successfully applied to synthesise ordered
mesoporous aluminas with high thermal stability [29,30], large
pores up to 7.5 nm [31], diverse transition metal dopants [32,33], or
hierarchical architectures containing complementary macropores
via colloidal crystal co-templating with polystyrene microspheres
[34].
Here we exploit the EISA route to produce a high surface area,
highly ordered mesoporous alumina in order to stabilise dispersed
palladium nanoparticles in an oxidised form, with the goal of
enhancing the selective oxidation of allylic alcohols. The resulting
Pd/meso-Al2O3 catalysts are active towards diverse allylic alcohols,
and exhibit significantly higher turnover frequencies (TOFs) than
their silica counterparts, with initial rates inversely proportional to
palladium loading, reflecting a higher density of surface PdO active
species and stronger Pd-alumina interaction.
2.4. Allylic alcohol selox
Catalyst screening was performed using a Radleys Starfish carousel
batch reactor on a 10 cm3 scale at 90 °C under a bubbled O2 flow
(3 cm3 min−1 at 1 bar). Catalyst (50 mg) was added to a reaction mix-
ture of allylic alcohol (8.4 mmol, Sigma-Aldrich purity of all N 95%),
mesitylene (0.1 cm3, Sigma-Aldrich 99%) as an internal standard, and tol-
uene (10 cm3, Fisher Scientific 99.8%). Control reactions in the absence of
any solid phase, or presence of bare alumina supports, were conducted in
parallel with tests on Pd/meso-Al2O3 and gave negligible conversion of
any alcohols. Reactions were periodically sampled, with aliquot
(0.25 cm3) withdrawn, filtered, and diluted with toluene (1.75 cm3, Fish-
er Scientific 99.8%) for triplicate analysed on a Varian 3900GC with
CP-8400 autosampler (CP-Sil5 CB column, 15 m x 0.25 mm x 0.25 μm).
Initial rates were calculated from the initial linear region of the alcohol
conversion profiles (typically 0.3–1 h reaction), with selectivity and over-
all mass balances calculated using calibrated response factors for reac-
tants and products. Conversion and selectivity values are reported
within 3 % error, with mass balances in all cases ≥95 % during the
first hour and ≥90 % over 24 h. Catalyst recyclability was assessed by
screening a spent quantity of catalysts from a scaled-up (by a factor of
2.5) to ensure significant catalyst recovery by hot filtration. Spent cata-
lysts were stirred in 50 ml toluene at 90 °C for 10 min (three times) be-
fore drying at 120 °C for 2 h and subsequent catalytic testing under
identical conditions to those stated above.
2. Experimental
2.1. Mesoporous alumina synthesis
Highly ordered mesoporous alumina (meso-Al2O3) was prepared
adopting the procedure of Yuan and co-workers [29]. Pluronic P123
surfactant (3 g, Sigma-Aldrich) was dissolved in anhydrous ethanol
(60 cm3, Sigma-Aldrich N 99.5%) under agitation, and nitric acid
(4.5 cm3, Fisher Scientific 65 wt%) and aluminium isopropoxide
(6.2 g, Sigma-Aldrich 98%) subsequently added under stirring until
dissolved and the mixture aged for 5 h. EISA was initiated by slow
ethanol removal upon heating at 60 °C under static conditions. After
96 h, the resulting yellow solid was ground to a fine powder, and
then heated at 0.4 °C min−1 under flowing O2 (50 cm3 min−1) to
600 °C for 3 h.
2.2. Pd impregnation
Palladium incorporation was achieved via incipient-wetness im-
pregnation. Mesoporous alumina (1.5 g) was saturated with aqueous
tetraammine palladium(II) nitrate solution (1.5 cm3, with Pd concen-
trations adjusted to achieve nominal loadings spanning 0.05–5 wt%
Sigma-Aldrich 10 wt%) at room temperature. The resulting slurries
were stirred for 18 h before heating to 50 °C. Agitation was stopped
after approximately 5 h, and the residual solids left to dry for 24 h
at 50 °C. The dry powders were subsequently heated in static air at
1 °C min−1 to 500 °C for 2 h, prior to heating at 10 °C min−1 to
400 °C for 2 h under flowing H2 (10 cm3 min−1). Samples were
then cooled under X to room temperature and stored in air. A nominal
3. Results and discussion
3.1. Catalyst characterisation
The successful synthesis of alumina possessing a p6mm 2D hexag-
onally packed, cylindrical pore architecture (meso-Al2O3) with a high
degree of mesopore ordering is evidenced in Fig. 1. Low-angle XRD
confirmed a p6mm mesostructure, which was readily visualised by
TEM (see also Fig. S1). Textural properties were in good agreement
1 wt% Pd on commercial Degussa C γ-alumina (surface area 180 m2
g
−1) was also prepared following the same protocol.
Please cite this article as: C.M.A. Parlett, et al., Selective oxidation of allylic alcohols over highly ordered Pd/meso-Al2O3 catalysts, Catalysis Com-
Parlett, Christopher M.A.
