The Comparative Effect of Particle Size and Support Acidity on Hydrogenation of Aromatic Ketones

Article abstract of DOI:10.1002/cctc.201900993

A comparative study was reported for both the effects of shape-confined cubic Pd particle size (8, 13, and 21 nm) and surface property of most commonly used supports (SiO₂, Al₂O₃, and silica-alumina) on catalytic performance in the chemoselective hydrogenation of three model bio-oil chemicals (benzaldehyde, acetophenone, and butyrophenone). The results showed that the size of Pd particles could be more associated with the hydrogenation reaction than acidities of the supports. Smaller size of Pd particles, regardless of the type of the support, provided the higher catalytic performance. XPS data showed that the electronic properties of Pd particles were similar, therefore, the possible reasons were the higher fraction of Pd atoms on corner in smaller particles, the lower accessibility of hydrogen atom to reactant on bigger particles, and the more low-coordinated sites in the small-size particles due to the short-range ordering. In addition, Pd/SA catalysts (Br?nsted acid sites on the support) showed the highest conversion and TOF compared to Pd/Al₂O₃ and Pd/SiO₂ catalysts. This might be due to the enhanced the diffusion rates of the chemicals on the surface of the catalysts although they could not induce the ionic effect from the metal surface. Pd/SiO₂ catalysts performed better than Pd/Al₂O₃ catalysts (Lewis acid sites on the support). The flexible SiOH groups on surface made the easy interaction with the metal particles and promote the reaction.

Full text of DOI:10.1002/cctc.201900993

Accepted Article
Title: The comparative effect of particle size and support acidity on
hydrogenation of aromatic ketones
Authors: Jun Huang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201900993
Link to VoR: http://dx.doi.org/10.1002/cctc.201900993
A Journal of
www.chemcatchem.org

10.1002/cctc.201900993
ChemCatChem
The comparative effect of particle size and support
acidity on hydrogenation of aromatic ketones
Dr. Kyung Duk Kim†,Dr. Zichun Wang†, Yongwen Tao†, Dr. Huajuan Ling†, Yuan Yuan‡, Dr.
Cuifeng Zhou†, Prof. Dr. Zongwen Liu†, Prof. Dr. Marianne Gaborieau⊥, Prof. Dr. Jun
Huang†,*, and Prof. Dr. Aibing Yu§,*
† Laboratory for Catalysis Engineering, School of Chemical and Biomolecular Engineering, The
University of Sydney, NSW 2006, Australia
‡ School of Chemistry, University of New South Wales, Sydney, NSW 2052, Australia
⊥ Molecular Medicine Research Group, School of Science and Health, Australian Centre for
Research of Separation Science (ACROSS), Western Sydney University, Parramatta NSW 2150,
Australia
§ Laboratory for Simulation and Modelling of Particulate Systems, Department of Chemical
Engineering, Monash University, Clayton, VIC 3800, Australia
This article is protected by copyright. All rights reserved.
1

10.1002/cctc.201900993
ChemCatChem
ABSTRACT A comparative study was reported for both the effects of shape-confined cubic Pd
particle size (8, 13, and 21 nm) and surface property of most commonly used supports (SiO2, Al2O3,
and silica-alumina) on catalytic performance in the chemoselective hydrogenation of three model
bio-oil chemicals (benzaldehyde, acetophenone, and butyrophenone). The results showed that the
size of Pd particles could be more associated with the hydrogenation reaction than acidities of the
supports. Smaller size of Pd particles, regardless of the type of the support, provided the higher
catalytic performance. XPS data showed that the electronic properties of Pd particles were similar,
therefore, the possible reasons were the higher fraction of Pd atoms on corner in smaller particles,
the lower accessibility of hydrogen atom to reactant on bigger particles, and the more low-
coordinated sites in the small-size particles due to the short-range ordering. In addition, Pd/SA
catalysts (Brønsted acid sites on the support) showed the highest conversion and TOF compared
to Pd/Al2O3 and Pd/SiO2 catalysts. This might be due to the enhanced the diffusion rates of the
chemicals on the surface of the catalysts although they could not induce the ionic effect from the
metal surface. Pd/SiO2 catalysts performed better than Pd/Al2O3 catalysts (Lewis acid sites on the
support). The flexible SiOH groups on surface made the easy interaction with the metal particles
and promote the reaction.
This article is protected by copyright. All rights reserved.
2

10.1002/cctc.201900993
ChemCatChem
■ INTRODUCTION
Biomass derived ketones and aldehydes can be further upgraded by the chemoselective
hydrogenation, to qualified fuels or valuable unsaturated alcohols as feedstocks for drugs and
fragrance.[1, 2] The challenge is to design new catalysts for reaching the desired chemoselectivity
and avoiding thermodynamically favoured by-products.[3, 4] Recently, approaches to improve the
efficiency of supported metal nanocatalysts have been intensively investigated. One strategy for
enhancing catalytic activity or selectivity is to control the features of metal nanoparticles.[5] The
hydrogenation activity and selectivity can be strongly influenced by metal particle size and
dispersion. Often, reducing the particle size can introduce more unsaturated surface sites with
higher surface free energy of metal species, which can promote surface reactions, for instance, the
acrolein hydrogenation activity increased with decreasing Au particle size from 8 to 4 nm on
Au/ZrO2.[6] However, the acrolein hydrogenation can be enhanced with increasing the Ag particle
size on Ag/SiO2 from 1 to 9 nm.[7] It has also been observed that hydrogenation of 2-cyclohexenone
over Rh/SiO2 is size insensitive.[8] Obviously, the particle size of metal catalysts should definitely
influence their catalytic performance. Therefore, understanding the size effect of metal
nanoparticles is a key to tune their catalytic performance in target hydrogenation reactions.
The performance metal nanocatalysts can strongly influence by the nature of metals and supports.
Normally, the acidic supports can inhibit metal aggregation, provide a large surface to volume
ratio as compared to the bulk equivalents, and enhance the electron deficiency of the active metal
via metal-support interaction.[9-13] The metal-support interaction can be generated through the
acidic sites withdrawing electrons from the metal surface, which results in ionic effects and tunes
the surface electronic properties of metal particles.[3, 9, 14] With increase the support acidity, the
XPS binding energy and Fermi level of Pd supported on zeolite LTL increases, leading to increase
This article is protected by copyright. All rights reserved.
3

10.1002/cctc.201900993
ChemCatChem
the difference in energy between the Pd–H antibonding orbital, which enhance the activity of Pd
catalysts in neopentane hydrogenolysis reaction.[15] Moreover, extensive works focus on the
preparation of Pd particle with uniform size to control their catalytic performance.[16-18] Recently,
uniform Pd particles (1-2 nm) supported on various metal oxides (SiO2, Al2O3 and silica-alumina)
having different acidities under control are prepared by double flame-spray pyrolysis techniques.[3,
19] In these works, the Pd particles exhibit similar surface electronic properties independent from
support acidity by decoupling the interaction between metal and support. This is obviously
different from the metal particle size and electron properties varies from support
composition/acidity prepared by classical methods, particularly on non-porous metal oxide
support.[20, 21] As the supported metal catalysts prepared via impregnation techniques, it is difficult
to control the properties of the final catalysts due to the strong interaction between metal and
supports. Therefore, the evaluation of the metal catalyst activity based on the particle sizes of metal
nanocatalysts and the properties of the supports individually is still open, however, which is
necessary to understand the effects of particle size and supports on hydrogenation reactions, such
as in the hydrogenation of ketones and aldehydes.
Chemoselective hydrogenation of carbonyl compounds, such as acetophenone,[3, 14, 19, 22]
benzaldehyde,[23, 24] and butyrophenone are important industrial chemical process in the production
25]
of gasoline-oil and flavours, fragrances, agrochemicals, and pharmaceuticals.[23,
The
chemoselective hydrogenation reaction of the carbonyl groups, other than phenyl groups, is
important to reduce the consumption of the expensive hydrogen. The chemoselective
hydrogenations of these compounds are sensitive to the surface properties of metal particles and
support acidity,[9, 10] which can be utilized to evaluate the effect of metal properties and the metal-
support interaction.
This article is protected by copyright. All rights reserved.
4

10.1002/cctc.201900993
ChemCatChem
In this work, uniform cubic Pd nanoparticles with different particle sizes (8, 13, and 21 nm) but
the same geometrical properties were synthesized by colloidal-metal nanoparticle preparation
technique and loaded on SiO2, Al2O3, and silica-alumina supports. Therefore, the morphology and
metal loading are performed in a control manner. XPS analysis demonstrates Pd nanoparticles
showing similar electronic properties over different supports. Therefore, the effects of Pd particle
size and the support acidity on the catalytic performance of the supported Pd catalysts can be
investigated in the hydrogenation reaction, such as in the chemoselective hydrogenation of three
bio-oil model compounds (benzaldehyde, acetophenone, and butyrophenone).
■ EXPERIMENTAL
Chemicals. All chemicals employed for material synthesis and catalytic reaction, including
poly(vinyl pyrrolidone) (PVP), L-ascorbic acid, potassium bromide (KBr), sodium
tetrachloropalladate (II) (Na2PdCl4, 99.99% pure on metals basis), acetophenone (Aph, 99%),
benzaldehyde (BA, 99 %), and butyrophenone (Bph, 99 %), etc. were purchased from Sigma-
Aldrich.
Preparation of Pd/SA catalysts. Size-controlled Pd nano-particles were prepared by modifying
a previously reported method.[26-32] PVP (Mw = 55000, 105 mg) (colloidal stabilizer), L-ascorbic
acid (60 mg) (reductant) and KBr (75 mg, 300 mg and 600 mg for Pd particle size of 8 nm, 13nm
and 21 nm, respectively) (capping agent), were all dissolved in deionized water (8 ml) in a 25 ml
o
round-bottom flask at room temperature. The mixture was heated at 80 C for 3 hours under
magnetic stirring, resulting in a dark brown solution. For preparing Pd particle with a size of 13
and 21 nm, an additional sodium tetrachloropalladate (II) (Na2PdCl4) solution (57 mg) water
solution (3 ml) was added into the mixture after heating for 10 minutes. Then the mixture was
This article is protected by copyright. All rights reserved.
5

10.1002/cctc.201900993
ChemCatChem
cooled down, washed 3 times with deionized water and the black products (Pd nanoparticles) were
collected by centrifugation at 15000 rpm.
The support materials (SiO2, Al2O3, or, Silica-alumina) were prepared according to those
reported in literatures.[33, 34] Pd nanoparticles was pre-dissolved in ethanol (1 mg/ml) and sonicated
for 3 h at room temperature. Then a desired amount of the solution and support (2 wt % Pd) was
mixed, and stirred overnight under room temperature. Then, the mixture was sonicated for 30 min,
o
and the product was separated by centrifugation (3000 rpm) and dried in an oven at 100 C
overnight.
Catalyst characterizations. BET measurements. The Brunauer-Emmett-Teller (BET) was
performed by an Autosorb IQ-C system to determine the BET specific surface areas from the
adsorption data obtained at P/P0 between 0.1 and 0.3. Before measurements, the samples were
degassed at 200 °C under vacuum to remove adsorbents from the surface. The specific surface
areas were determined at -200 oC by BET measurement.
X-ray Diffraction. SIEMENS X-Ray Diffractometer (XRD-D5000) with Cu Kα radiation was
operated at 40 kV and 30 mA in the scan range of 1 to 70 degrees with continuous scanning mode
at a rate of 2° min-1 (0.1542 nm wavelength).
Transmission Electron Microcopy. TEM images were obtained by using a Jeol 2200FS (200 kV).
The specimens were mounted on a carbon-coated copper grid by drying a droplet of a suspension
of the ground sample in ethanol.
CO chemisorption. Pd dispersion was determined by CO-pulse chemisorption on a
Micromeritics Autochem II 2920 unit. Off-gas was analysed via a mass spectrometer to derive the
amount of chemisorbed CO.
This article is protected by copyright. All rights reserved.
6

10.1002/cctc.201900993
ChemCatChem
X-ray photoelectron spectroscopy (XPS). The supported Pd catalysts wre analyzed by XPS
(ESCALAB250Xi spectrometer, Thermo Scientific, UK), which was equipped with a mono-
chromated Al K (1486. 68 eV) X-ray source and operated at 15.2 kV and 164 W (10.8 mA). The
catalyst powder was measured in the vacuum chamber (2 × 10-9 mbar) and the photoelectron take-
off angle was 90o with respect to the surface of the sample.
27
Solid state NMR spectroscopy. The Al MAS (magic-angle spinning) NMR investigation of
catalysts was carried out on a Bruker DPX 200 WB spectrometer at resonance frequency of 52
MHz with a sample spinning rate of 8 kHz using 4 mm MAS rotors. Before measurements, all the
samples were exposed to a saturated vapour of a Ca(NO3)2 solution at ambient temperature
overnight in a desiccator for full hydration. Power levels and chemical shift scales were calibrated
with respect to aqueous aluminium nitrate (with a chemical shift of 0 ppm). Spectra were recorded
after single–pulse π/6 excitations with a repetition time of 0.5s for studying 27Al nuclei.
TPD characterization. Temperature programmed desorption of ammonia (NH3-TPD) was
performed on a Chem BET TPR/TPD Chemisorption Analyzer, CBT-1, manufactured by
QuantaChrome instruments. Prior to TPD, the sample (30 mg) was loaded and degassed at 520 oC
for 1 h under a He flow rate of 120 ml/min to completely remove molecules adsorbed on catalyst
o
surface. Then the sample was saturated with NH3 at 100 C, and flushed with a He flow (120
ml/min), then desorbed NH3 with increasing temperature from 50 oC up to 1000 oC at a rate of 10
oC /min. The concentration of desorbed ammonia was analyzed by a TCD detector connected to
the TPD system.
Chemoselective hydrogenation of bio-oil model compounds. Pd/SA catalysts were put in an
autoclave which was purged three times with nitrogen. Acetophenone (30mg) was dissolved in
purified water (6 ml) and then added to the autoclave. The reaction was carried out at a hydrogen
This article is protected by copyright. All rights reserved.
7

10.1002/cctc.201900993
ChemCatChem
pressure of 3 bar at 65 oC under magnetic stirring for 3 h with a Pd/Aph molar ratio of 1:62.5. The
reaction product(s) were collected at intervals of 10, 20, 30, 45, 60, 90, 120 and 180 min.
Chemoselective hydrogenation of BA was carried out in a similar manner, in which, BA (30 mg)
o
replaced Aph and the reaction was performed at 25 C with a Pd/BA molar ratio of 1:75. For
chemoselective hydrogenation of Bph, Bph (30 mg) was dissolved in 2-propanol (6 ml). Then the
reaction was carried out in autoclave at 65 oC with 3 bar hydrogen pressure with a Pd/Bph molar
ratio of 1:50. The conversion and composition of the product mixtures were analyzed using a
Shimadzu GCMS-QP2010 Ultra with a RTX-5MS capillary column (30 m × 0.25 mm × 0.25 µm)
and quantified by Shimadzu GC-FID equipped with a RTX-5 capillary column (30 m × 0.32 mm
× 3 µm). Bicyclohexyl (0.1 mmol) was used as an internal standard. The selectivity to specific
product(s) i (Si) was calculated as Si (%) = (ci)/[(cAPh)0-(cAPh)], where (ci) is the molar concentration
of the products, such as 1-phenylehtanol (PhE), and (cAPh)0 and (cAPh) correspond to the molar
concentration of acetophenone before and after the reaction, respectively. TOF was calcualted as
(moles of substrate converted)/(moles of Pd center x dispersion) at specific time (s) of the
conversion of about 10 % of reactant.
■ RESULTS AND DISCUSSION
Characterization of catalysts. The size and shape of as-synthesized Pd particles were studied
by TEM. The synthesis is based on a water-based process that involves addition of Na2PdCl4 into
o
an mixture solution of L-ascorbic acid (reductant) at 80 C and poly(vinyl pyrrolidone) (PVP)
working as a stabilizer and Br- ions serving as a capping agent. Control over the size of the Pd
cubes was achieved by adjusting the rate of reduction of Pd2+ ions with Br- ions. As shown in
Supporting Information Figure S1 and S2, the average Pd particle sizes (8, 13 and 21 nm) increased
with increasing the amount of the reducing agent (KBr: 75, 300, and 600 mg), respectively, in
This article is protected by copyright. All rights reserved.
8

10.1002/cctc.201900993
ChemCatChem
good consistent with previous literatures.[30-32] All Pd particles showed the same cubic morphology
(Figure S1). These Pd particles can sustain their morphology even after loading on different
support materials (Figure 1 and S3-4). It demonstrates the Pd particles are stable under current
conditions.
30
30
30
25
25
25
20
20
20
15
15
15
10
10
10
5
5
0
5
0
0
9
10
11
12
13
14
15
16
6
7
8
9
10
11
18
19
20
21
22
23
24
Particle size (nm)
Particle size (nm)
Particle size (nm)
Figure 1. TEM and HRTEM images of particle distributions of silica-alumina supported samples,
and Pd particle distributions: (a) TEM of Pd/SA-8, (b) TEM of Pd/SA-13, (c) TEM of Pd/SA-21
Pd particle size distribution of Pd/ SA-8, (d) HRTEM of Pd/ SA -8, (e) HRTEM of Pd/SA-13, (f)
This article is protected by copyright. All rights reserved.
9

10.1002/cctc.201900993
ChemCatChem
HRTEM of Pd/SA-21 Pd particle size distribution of Pd/SA- 13, (g) TEM of Pd/SA-8, (h) Pd
particle size distribution of Pd/SA- 13, and (i) Pd particle size distribution of Pd/SA- 21.
BET surface areas of Pd catalysts on different supports were determined by nitrogen adsorption-
desorption isotherm and summarized in Table 1. The surface areas of Pd/SA, Pd/SiO2 and
Pd/Al2O3 were 171 to 167 m2/g, 309 to 297 m2/g, and 119 to 112 m2/g, respectively. Compared to
the parent support materials,[33, 34] Pd loading and their particle size have very slight effect on the
surface area of Pd supported catalysts. The surface areas increased with increasing the content of
silica, in a good agreement that silica often exhibits a larger surface areas than that of γ-Al2O3.[19,
35]
Table 1. Specific surface area, particle size, dispersion, and Turnover Frequencies (TOF) for the
Chemoselective Hydrogenation of acetophenone, benzaldehyde, and butyrophene over Pd
supported catalysts.
ABET
Sample
PPd (nm)b DPd (%)c TOFAph (s-1)d TOFBA (s-1)d TOFBph (s-1)d
(m2g-1)a
Pd/SA-8
171
8.0
17.1
10.5
5.5
2.2×10-2
1.6×10-2
1.0×10-2
1.4×10-2
1.1×10-2
0.9×10-2
1.1×10-2
1.0×10-2
0.7×10-2
5.2×10-2
4.5×10-2
3.2×10-2
4.2×10-2
3.3×10-2
2.9×10-2
3.2×10-2
2.1×10-2
1.3×10-2
2.1×10-2
1.6×10-2
0.8×10-2
1.9×10-2
1.7×10-2
0.8×10-2
1.2×10-2
1.0×10-2
0.5×10-2
Pd/SA-13
Pd/SA-21
Pd/SiO2-8
Pd/SiO2-13
Pd/SiO2-21
Pd/Al2O3-8
168
12.6
21.6
8.5
167
309
16.5
12.5
5.1
300
13.2
21.2
8.4
297
119
16.9
11.9
5.9
Pd/Al2O3-13 117
Pd/Al2O3-21 112
12.9
21.8
a Surface area (ABET) was determined by BET analysis. b Pd particle size (PPd)was obtained from
corresponding TEM images. c Pd dispersion (DPd) was determined by CO chemisorption. d TOFs
This article is protected by copyright. All rights reserved.
10

10.1002/cctc.201900993
ChemCatChem
for acetophenone (TOFAph), benzaldehyde (TOFBA) and butyrophenone (TOFBph) over Pd
catalysts were calculated at the conversion of about 10 % of reactant.
The effect of Pd loading on the support materials was investigated by XRD. Both silica and
silica-alumina support were highly amorphous in nature. Only a very broad peak at 2θ = 25o
corresponding to amorphous silica can be observed in XRD patterns (Figure 2).[14] The weak
diffraction peaks at 42o and 48o were assigned to Pd (111) and Pd (200) on Pd nanocatalysts,
respectively. Pd/Al2O3 catalysts exhibit similar diffraction patterns to the Al2O3 support reported
in literature.[19] No obvious change of crystal structure can be observed among Pd particles with
different size on the all supports. Since XRD is less sensitive to small particles such as Pd particles
under current study, the dispersion of surface Pd atoms was obtained by CO chemisorption
experiments and summarized in Table 1, which decreased with increasing the Pd particle size but
independent from the support surface areas.
Pd (111)
Pd/SA-8
Pd (200)
Pd/SA-13
Pd/SA-21
Pd/SiO2-8
Pd/SiO2-13
Pd/SiO2-21
Pd/Al2O3-8
Pd/Al2O3-13
Pd/Al2O3-21
10
20
30
40
50
60
70
2 Theta (degree)
Figure 2. XRD patterns of different types of supported Pd-nanoparticle catalysts.
This article is protected by copyright. All rights reserved.
11

10.1002/cctc.201900993
ChemCatChem
The effect of Pd particles on the local structure of support was investigate by 27Al MAS NMR
spectroscopy. As shown in Figure 3, a strong signal of octahedral coordinated Al (AlVI) at -3 ppm
and a weak signal of tetrahedral coordinated Al (AlIV) at 51.8 ppm have been observed for all
27
Pd/Al2O3 catalysts, typically for the Al NMR spectra of γ-Al2O3.[3, 35] Compared to Pd/Al2O3
catalysts, the content of AlIV (signal at 51.8 ppm) was significantly enhanced in Pd/SA catalysts.
27
The Al MAS NMR spectra of Pd/SA is similar to that reported for amorphous silica-alumina
27
supports in literature.[33] Clearly, the Al MAS NMR spectra of Pd/Al2O3 and Pd/SA catalysts
demonstrated that the loading process and various Pd particle sizes cannot strongly influence the
support structure and coordination. Often, AlVI species in alumina and silica-alumina are
associated to the formation of surface Lewis acid sites (LAS). The enhanced content of AlIV in
silica-alumina has been widely accepted to the incorporation of Al into the silica network, which
can interact and enhance the acid strength of neighboring SiOH groups, acting as surface Brønsted
acid sites.[33, 36, 37]
-3.0
51.8
(a) Pd/Al2O3-8
(b) Pd/Al2O3-13
(c) Pd/Al2O3-21
(d) Pd/SA-8
(e) Pd/SA-13
200
150
100
50
0
-50
-100 -150 -200
d
27Al / ppm
This article is protected by copyright. All rights reserved.
12

10.1002/cctc.201900993
ChemCatChem
Figure 3. 27Al MAS NMR spectra of Pd catalysts on alumina (a-c) and silica-alumina (d and e)
supports.
Since the metal-support interaction induced by acidic support (e.g. alumina and silica-alumina)
can strongly influence the catalytic performance of Pd catalysts,[3, 9, 19] the surface acidity of
supported Pd catalysts are characterized by the temperature programmed desorption of ammonia
(NH3-TPD) experiments. As shown in Figure 4 and Figure S5, nearly no desorption peaks can be
observed for Pd/SiO2 catalysts since amorphous SiO2 is a neutral support. Broad strong peaks were
observed for both Pd/SA and Pd/Al2O3 catalysts with desorption maximum peak (Tm) cantered at
o
o
240 C and 205 C, respectively, indicating a significant amount of acid sites generated on the
o
supports. γ-Al2O3 was a typical Lewis acidic catalyst,[38] and thus, the peak at 205 C observed
with Pd/Al2O3 catalysts has been assigned to ammonia adsorbed on LAS on alumina support.
o
Alternatively, the peak at 240 C was assigned to BAS generated on silica-alumina, which is
generated by introducing Al atoms into the silica network, and can only be formed on amorphous
silica-alumina among these three supports. Often, the relative strength of the acid sites can be
scaled by their Tm, for instance the higher Tm the higher acid strength. Therefore, the acid sites
on Pd/SA showed acid sites seems stronger than those on Pd/Al2O3.
This article is protected by copyright. All rights reserved.
13

10.1002/cctc.201900993
ChemCatChem
Tm = 240 oC
Pd/SA-8
Tm = 205 oC
Pd/Al2O3-8
Pd/SiO2-8
Temperature (oC)
Figure 4. NH3-TPD profiles of Pd catalysts with a particle size of 8 nm on various supports.
The interaction between acidic supports and metal nanocatalysts can introduce an ionic effect
on active metal species.[15, 39] This effect can strongly influence the surface electronic properties
of metal catalysts, which is crucial to the chemoselectivity of supported metal nanocatalysts in
hydrogenation reactions.[1, 9, 14, 19] Often, the surface electronic properties of the Pd nanoparticles
can be determined by XPS.[15, 19] As shown in Figure 5, the binding energy of Pd particles on
different supports are all similarly, exhibiting two signals at 334.8 eV for Pd 3d5/2 and 340.2 eV
for Pd 3d3/2. It indicates that the electronic properties of Pd particles are similar, independent from
the support properties. Normally, surface acid sites can withdraw electrons from the metal particles
via ionic effect. A higher acidity results in a higher binding energy.[15] Our recent work shows that
the strength of acid sites is the key to tune the electronic properties of Pd particles on acidic
support.[14] In this work, Pd particles exhibit similar electronic properties over support with
different acidic properties as indicated by NH3-TPD. The similar binding energy may indicate the
metal-support interaction is very weak in supported Pd nanocatalysts prepared in this work, and
consequently, the acid sites on support are hardly to withdraw electron density from metal
This article is protected by copyright. All rights reserved.
14

10.1002/cctc.201900993
ChemCatChem
component. This is similar to that observed in Pd/SA prepared by double flame-spray pyrolysis
technique.[19] In that work, the electron properties of Pd particles is almost the same over support
materials with various strength and density of acid sites, which has been attributed to minimize
the ionic effect through decoupling the formation of Pd particles and supports, one could efficiently
from supports on Pd catalysts.
Pd3d5/2
Pd3d3/2
Pd/SiO2-8
Pd/Al2O3-8
Pd/SA-8
330 332 334 336 338 340 342 344 346
Binding energy (eV)
Figure 5. Binding energy of Pd3d5/2 and Pd3d3/2 lines of Pd catalysts with a particle size of 8 nm
on various supports
Chemoselective hydrogenation. The performance of the obtained Pd catalysts was tested in the
hydrogenation of acetonphenone (Aph), benzaldehyde (BA) and butyrophenone (Bph), and the
proposed reaction pathway are described in Scheme S1. It is desired to the chemoselective
hydrogenation to C=O bond to produce unsaturated alcohols, while those produced by
hydrogenation of aromatics are the main byproducts. In hydrogenation reactions, the
chemoselectivity is sensitive to the surface electronic properties of metal catalysts. Generally, the
ionic effect induced by acidic supports leads to different surface electronic properties of metal
catalysts, and thus, provide different chemoselectivity in hydrogenations. In the hydrogenation of
This article is protected by copyright. All rights reserved.
15

10.1002/cctc.201900993
ChemCatChem
BA at 25 oC under 3 bar hydrogen, mainly benzyl alcohol can be detected by GC analysis, which
indicates all the catalysts exhibited an excellent chemoselectivity (100%) to the hydrogenation of
the carbonyl group rather than aromatic rings in BA. This has been attributed to similar surface
electronic properties of Pd nanoparticles, similar as observed in earlier studies.[14, 19] Moreover, it
demonstrates that the various functional groups (neutral for Pd/SiO2, Lewis acid sites for Pd/Al2O3
and Brønsted acid sites for Pd/SA catalyst) and different densities/strengths of acid sites on the
supports, cannot significantly influence the chemoselectivity of Pd catalysts in hydrogenation
reaction.
The conversion of BA as a function of time was shown in Figure 6 and corresponding TOFs at
10% conversion of BA are summarized in Table1. All catalysts are active in the conversion of BA.
Reducing Pd particle size from 21 to 8 nm can significantly enhance the conversion of BA. For
instance, Pd/SA-8 obtained ca. 100% conversion of BA after 3 h reaction, and only 65% and 50%
for Pd/SA-13 and Pd/SA-21, respectively. A similar trend was observed for the TOFs of Pd/SA
(5.2 × 10-2 s-1, 4.5 × 10-2 s-1 and 3.2 × 10-2 s-1 for Pd/SA-8, Pd/SA-13, and Pd/SA-21, respectively).
Pd/SiO2 and Pd/Al2O3 showed similar tendencies as observed for Pd/SA, although they were
characterized with different acidic properties. With respect to the same electronic properties of Pd
catalysts and support properties as revealed by XPS and NH3-TPD analysis, a size effect on the
catalytic performance of Pd particles has been probed directly. In general, small particles can result
in more defects (e.g. Pd atoms on the corner) than large particles, acting as active sites in catalytic
reactions.[40] It has been reported that the low coordinated sites in the small particles due to the
short-range ordering shows a significantly higher catalytic performance compared to those having
larger particle size.[41] Moreover, smaller particle size can inhibit the hydrogen atoms diffuse
deeply into the bulk with large particle size, and thus, hydrogen are readily available for the
This article is protected by copyright. All rights reserved.
16

10.1002/cctc.201900993
ChemCatChem
reaction at the surface.[42] In addition, the product is easy to desorb due to the relatively lower
enthalpy desorption on low-coordinated Pd sites, which may result in an increase in the
conversion.[43] Therefore, the smaller Pd particle size and higher Pd dispersion are mainly response
for the catalytic enhancement.
Pd/SA (8 nm)
Pd/SA (13 nm)
Pd/SA (21 nm)
Pd/SA (8 nm)
Pd/Al2O3 (8nm)
100
80
60
40
20
0
100
80
60
40
20
(a)
(b)
Pd/SiO2 (8 nm)
0
100
Pd/Al2O3 (8 nm)
Pd/Al2O3 (13 nm)
Pd/Al2O3 (21 nm)
100
80
60
40
20
0
Pd/SA (13 nm)
Pd/Al2O3 (13 nm)
80
60
40
20
Pd/SiO2 (13 nm)
0
100
Pd/SiO2 (8 nm)
Pd/SiO2 (13 nm)
Pd/SiO2 (21 nm)
100
80
60
40
20
0
Pd/SA (21 nm)
Pd/Al2O3 (21 nm)
80
60
40
20
0
Pd/SiO2 (21 nm)
0
30
60
90
120 150 180
0
30
60
90
120 150 180
Time (min)
Time (min)
Figure 6. Conversion of benzoaldehyde (BA) over different supported Pd catalysts by (a) Pd
particle sizes and by (b) supports (alumina-silica, SiO2, and Al2O3 supports), as a function of
reaction time at 3 bar of hydrogen and 25 oC.
It should be noted that the TOFs should be similar with Pd particles having the same morphology
(size, dispersion and lattice structure) and electronic properties. However, the TOFs are differ
significantly for the same Pd particles on different supports in the BA hydrogenation. Silica-
alumina supported catalysts exhibited better performance than the other two catalysts, e.g. 5.2 ×
This article is protected by copyright. All rights reserved.
17

10.1002/cctc.201900993
ChemCatChem
10-2 s-1, 4.2 × 10-2 s-1 and 3.2 × 10-2 s-1 for Pd/SA-8, Pd/SiO2-8, and Pd/Al2O3-8, respectively.
Although no obvious ionic effect on Pd particles can be induced by BAS on silica-alumina support,
BAS can enhance the surface diffusion of BA on Pd/SA catalyst. This is similar as reported earlier
that BAS can increase the concentration of BA in the local environment of Pd-sites, thereby
increasing the accessibility of BA to Pd active sites.[19] LAS on the Al2O3 support can also adsorb
BA on Pd/Al2O3 surface. However, LAS are often probed with the strong adsorption and activation
of C=O over BAS.[36, 44] Compared to LAS, reactant adsorbed on the flexible SiOH sites on surface
can easily interact with the metal nanocatalyst on the surface. With respect to the neutral silica
support having larger surface area, it may improve the surface diffusion of BA on Pd/SiO2 than on
Pd/Al2O3, providing a higher TOFs.
Similar chemoselective hydrogenation results were obtained using acetophenone (Aph) and
butyrophenone (Bph) as shown in Figure S6. A 100 % selectivity to C=O hydrogenation was
obtained over all catalysts. The catalysts containing the small-size Pd particles (ca. 8 nm) all
achieved a 100 % conversion of Aph and Bph after 8 hours reaction, while those having large-size
Pd particles (ca. 21 nm) provide less than 50 % conversion under the same conditions. Both of
Aph and Bph are less active than BA due to the strong electron density of carbonyl oxygen. In all
reaction, the size effect is more significant than the effect of support acidity. This can be explained
by Pd particle sizes are the active sites for hydrogen dissociation and hydrogenation reactions,
while the support mainly governs the surface diffusion of reactants and products. Therefore, the
understanding of the adsorption-desorption balance of each species involved and the behaviour of
metal and supports is of great importance.
This article is protected by copyright. All rights reserved.
18

10.1002/cctc.201900993
ChemCatChem
■ Conclusion
In this study, size-controlled Pd particles (8, 13, and 21 nm) were prepared by the colloidal
method and then loaded on silica-alumina, Al2O3, and SiO2 supports which have various acidities
(BAS, LAS, and very weak SiOH, respectively) as observed in NH3-TPD. Pd nanoparticles exhibit
similar surface lattice structure and electronic properties as revealed by TEM, XRD and XPS
analysis. Therefore, these catalysts are applied to investigate the effect of Pd particle size and the
support acidity on their catalytic performance in the chemoselective hydrogenation of three
chemicals (acetophenone, benzaldehyde, and butyrophenone) from bio-oil, combined with
molecular dynamics study.
It has been found the chemoselectivity mainly relied on the Pd surface electronic properties. The
particle size has a significant effect on the catalytic performance of these catalysts. Regardless of
the type of the support (i.e., the type of acidity), catalysts with the smallest (8 nm) Pd particle size
exhibit the highest conversion of the three model compounds compared to the catalysts of the
medium (13 nm) and the largest (21 nm) Pd particle sizes. This has been explained by the smaller
particle size provides more corner Pd atoms, higher surface hydrogen atom to reactant, and more
low-coordinated sites.
Pd/SA catalysts (BSA containing on the support) showed the highest conversion and TOF
compared to Pd/Al2O3 and Pd/SiO2 catalysts. This might be due to the enhanced the diffusion rates
of the chemicals on the surface of the catalysts although they could not induce the ionic effect from
the metal surface. The reactants could be desorbed from support surface and diffused to the metal
particle surface for hydrogenation. Interestingly, Pd/SiO2 catalysts have both a slightly higher
catalytic activity and TOFs than Pd/Al2O3 catalysts (LAS containing on the support). It could be
due to the larger specific surface area on the silica support and the flexible SiOH groups on surface
This article is protected by copyright. All rights reserved.
19

10.1002/cctc.201900993
ChemCatChem
that make the easy interaction possible with the metal particles on the surface. The results of this
study indicate that the catalytic activities were influenced more by the particle size than the acidity
of the support.
ASSOCIATED CONTENT
Supporting Information.
The following files are available free of charge.
TEM and HRTEM images, particle size distribution of supported Pd catalysts, NH3-TPD profiles
for supported Pd catalysts, calculated diffusivity of reactants and product on supported Pd
catalysts, adsorption of acetophenone on various supports, conversion of butyrophenone over
different Pd particle size on various supports.
AUTHOR INFORMATION
Corresponding Author
*jun.huang@sydney.edu.au (J. Huang). *Aibing.Yu@monash.edu.au (A.Yu).
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
ACKNOWLEDGMENT
J.H. K.K. and Z.W. acknowledge the financial supports from Australian Research Council
Discovery Projects (DP150103842). J.H. acknowledges the Australian Research Council
Discovery Projects (DP150103842 and DP180104010), the SOAR Fellowship, and the Sydney
This article is protected by copyright. All rights reserved.
20

10.1002/cctc.201900993
ChemCatChem
Nano Grand Challenge from the University of Sydney for the support of this project. J.H. thanks
Faculty’s Energy & Materials Clusters and MCR scheme and the International Project
Development Funding at the University of Sydney.
KEYWORDS: Chemoselective hydrogenation, Heterogeneous catalyst, Shape confined cubic
Pd catalysts, Silica/alumina supports, Support acidity
REFERENCES
[1]. M. M. Chen, N. Maeda, A. Baiker and J. Huang, ACS. Catal., 2012, 2, 2007.
[2]. G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106, 4044.
[3]. K. D. Kim, S. Pokhrel, Z. C. Wang, H. J. Ling, C. F. Zhou, Z. W. Liu, M. Hunger, L. Madler
and J. Huangt, ACS. Catal., 2016, 6, 2372.
[4]. C. Mohr, H. Hofmeister, J. Radnik and P. Claus, J. Am. Chem. Soc., 2003, 125, 1905.
[5]. A. Corma and P. Serna, Science, 2006, 313, 332.
[6]. C. Mohr, H. Hofmeister and P. Claus, J. Catal., 2003, 213, 86.
[7]. H. J. Wei, C. Gomez, J. J. Liu, N. Guo, T. P. Wu, R. Lobo-Lapidus, C. L. Marshall, J. T.
Miller and R. J. Meyer, J. Catal., 2013, 298, 18.
[8]. E. Ronzon and G. Del Angel, J. Mol. Catal. A. Chem., 1999, 148, 105.
[9]. J. Huang, Y. J. Jiang, N. van Vegten, M. Hunger and A. Baiker, J. Catal., 2011, 281, 352.
[10]. F. Hoxha, B. Schimmoeller, Z. Cakl, A. Urakawa, T. Mallat, S. E. Pratsinis and A. Baiker,
J. Catal., 2010, 271, 115.
[11]. F. Jiang, L. Zeng, S. R. Li, G. Liu, S. P. Wang and J. L. Gong, ACS Catal., 2015, 5, 438.
This article is protected by copyright. All rights reserved.
21

10.1002/cctc.201900993
ChemCatChem
[12]. Y. F. He, J. X. Fan, J. T. Feng, C. Y. Luo, P. F. Yang and D. Q. Li, J. Catal., 2015, 331,
118.
[13]. C. F. Xu, G. X. Chen, Y. Zhao, P. X. Liu, X. P. Duan, L. Gu, G. Fu, Y. Z. Yuan and N. F.
Zheng, Nat. Commun., 2018, 9.
[14]. Z. C. Wang, K. D. Kim, C. F. Zhou, M. M. Chen, N. Maeda, Z. W. Liu, J. Shi, A. Baiker,
M. Hunger and J. Huang, Catal. Sci. Technol., 2015, 5, 2788.
[15]. B. L. Mojet, J. T. Miller, D. E. Ramaker and D. C. Koningsberger, J. Catal., 1999, 186,
373.
[16]. F. K. Shieh, S. C. Wang, C. I. Yen, C. C. Wu, S. Dutta, L. Y. Chou, J. V. Morabito, P. Hu,
M. H. Hsu, K. C. W. Wu and C. K. Tsung, J. Am. Chem. Soc., 2015, 137, 4276.
[17]. G. Darabdhara, P. K. Boruah, P. Borthakur, N. Hussain, M. R. Das, T. Ahamad, S. M.
Alshehri, V. Malgras, K. C. W. Wu and Y. Yamauchi, Nanoscale, 2016, 8, 8276.
[18]. K. Murata, Y. Mahara, J. Ohyama, Y. Yamamoto, S. Arai and A. Satsuma, Angew. Chem.
Int. Edit., 2017, 56, 15993.
[19]. Z. C. Wang, S. Pokhrel, M. M. Chen, M. Hunger, L. Madler and J. Huang, J. Catal., 2013,
302, 10.
[20]. T. Kimura, T. Miyazawa, J. Nishikawa, S. Kado, K. Okumura, T. Miyao, S. Naito, K.
Kunimori and K. Tomishige, Appl. Catal. B. Environ., 2006, 68, 160.
[21]. S. Handjani, E. Marceau, J. Blanchard, J. M. Krafft, M. Che, P. Maki-Arvela, N. Kumar, J.
Warna and D. Y. Murzin, J. Catal., 2011, 282, 228.
[22]. H. W. Chen, C. S. Chen and S. J. Harn, J. Phy. Chem., 1995, 99, 10557.
[23]. F. Pinna, F. Menegazzo, M. Signoretto, P. Canton, G. Fagherazzi and N. Pernicone, Appl.
Catal. A. Gen., 2001, 219, 195.
This article is protected by copyright. All rights reserved.
22

10.1002/cctc.201900993
ChemCatChem
[24]. N. Perret, F. Cardenas-Lizana and M. A. Keane, Catal. Commun., 2011, 16, 159.
[25]. P. Gallezot and D. Richard, Catal. Rev. Sci. Eng., 1998, 40, 81.
[26]. Y. Yuan, Y. V. Kaneti, M. S. Liu, F. Z. Jin, D. F. Kennedy, X. C. Jiang, J. Huang and A. B.
Yu, J. Energy. Chem., 2015, 24, 660.
[27]. Y. Li, E. Boone and M. A. El-Sayed, Langmuir, 2002, 18, 4921.
[28]. Y. E. Wu, S. F. Cai, D. S. Wang, W. He and Y. D. Li, J. Am. Chem. Soc., 2012, 134, 8975.
[29]. J. B. Wu, A. Gross and H. Yang, Nano Lett., 2011, 11, 798.
[30]. V. Tohver, J. E. Smay, A. Braem, P. V. Braun and J. A. Lewis, P. Natl. Acad. Sci., 2001,
98, 8950.
[31]. M. S. Jin, H. Y. Liu, H. Zhang, Z. X. Xie, J. Y. Liu and Y. N. Xia, Nano. Res., 2011, 4, 83.
[32]. T. H. Yang, S. Zhou, K. D. Gilroy, L. Figueroa-Cosme, Y. H. Lee, J. M. Wu and Y. N. Xia,
P. Natl. Acad. Sci., 2017, 114, 13619.
[33]. E. J. M. Hensen, D. G. Poduval, P. C. M. M. Magusin, A. E. Coumans and J. A. R. van
Veen, J. Catal., 2010, 269, 201.
[34]. D. M. Ladd, A. Volosin and D. K. Seo, J. Mater. Chem., 2010, 20, 5923.
[35]. R. Feng, S. T. Liu, P. Bai, K. Qao, Y. H. Wang, H. A. Al-Megren, M. J. Rood and Z. F.
Yan, J. Phys. Chem. C., 2014, 118, 6226.
[36]. J. Huang, N. van Vegten, Y. J. Jiang, M. Hunger and A. Baiker, Angew. Chem. Int. Edit.,
2010, 49, 7776.
[37]. Z. C. Wang, Y. J. Jiang, O. Lafon, J. Trebosc, K. D. Kim, C. Stampfl, A. Baiker, J. P.
Amoureux and J. Huang, Nat. Commun., 2016, 7.
[38]. G. Busca, Chem. Rev., 2007, 107, 5366.
This article is protected by copyright. All rights reserved.
23

10.1002/cctc.201900993
ChemCatChem
[39]. A. Y. Stakheev, Y. Zhang, A. V. Ivanov, G. N. Baeva, D. E. Ramaker and D. C.
Koningsberger, J. Phys. Chem. C., 2007, 111, 3938.
[40]. S. W. Cao, F. Tao, Y. Tang, Y. T. Li and J. G. Yu, Chem. Soc. Rev., 2016, 45, 4747.
[41]. R. Hess, F. Krumeich, T. Mallat and A. Baiker, Catal. Lett., 2004, 92, 141.
[42]. A. Binder, M. Seipenbusch, M. Muhler and G. Kasper, J. Catal., 2009, 268, 150.
[43]. W. Ludwig, A. Savara, R. J. Madix, S. Schauermann and H. J. Freund, J. Phys. Chem. C.,
2012, 116, 3539.
[44]. Z. C. Wang, Y. J. Jiang, M. Hunger, A. Baiker and J. Huang, Chemcatchem, 2014, 6, 2970.
This article is protected by copyright. All rights reserved.
24