Nanoceria-promoted low Pd-Ni catalyst for the synthesis of secondary amines from aliphatic alcohols and ammonia

Article abstract of DOI:10.1039/c8cy01670d

This paper describes the preparation of a series of bimetallic Pd-Ni catalysts supported over nanoceria with very low Ni and Pd loading (<0.5 wt%) for the direct amination of aliphatic alcohols with ammonia. Different metal impregnation methods were used for engineering the interaction between Pd, Ni and nanoceria. The as-prepared catalysts were characterized in detail by combining XRD, H₂-TPR, H₂-TPD, STEM-EDS-SDD and XPS. The sequence of impregnation of both metals and the Pd loading affected to an important extent the catalytic activity by conditioning the crystallite size and the Pd and Ni speciation, as well as the reducibility and reversible H₂ storage properties. By optimizing the preparation protocols, a 0.5wt% Pd-0.5wt% Ni-Pd/CeO₂ formulation prepared by sequential impregnation of the nickel and palladium precursors afforded 80% yield of dioctylamine at almost full conversion [TON = 1160 mmol per mmol (Ni + Pd)_surface] in the direct amination reaction of 1-octanol with ammonia at 180 °C for 2 h. Metal leaching during the reaction could be completely avoided. The high catalytic performance of Pd-Ni induced by nanoceria places this catalyst among the best ever reported catalysts for the synthesis of secondary amines.

Full text of DOI:10.1039/c8cy01670d

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: M. Pera-Titus, L.
Fang, Z. Yan, K. Vits and B. Southward, Catal. Sci. Technol., 2019, DOI: 10.1039/C8CY01670D.
This is an Accepted Manuscript, which has been through the
Volume 6 Number 1 7 January 2016 Pages 1–308
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
Accepted Manuscripts are published online shortly after
www.rsc.org/catalysis
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
ISSN 2044-4753
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
PAPER
Qingzhu Zhang et al.
Catalytic mechanism of C–F bond cleavage: insights from QM/MM
analysis of fluoroacetate dehalogenase
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/catalysis

Page 1 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
DOI: 10.1039/C8CY01670D
Nanoceria-promoted low Pd-Ni Catalyst for the Synthesis of
Secondary Amines from Aliphatic Alcohols and Ammonia
L. Fang,1,* Z. Yan,1 K. Vits,2 B. Southward,2 and M. Pera-Titus1,*
1 Eco-Efficient Products and Processes Laboratory (E2P2L), UMI 3464 CNRS – Solvay, 3066 Jin
Du Road, Xin Zhuang Ind. Zone, 201108 Shanghai, China
2 Solvay (China) Co., Ltd., 3966 Jin Du Rd., 201108 Shanghai, China
*Corresponding author: lin.fang@solvay.com, marc.pera-titus-ext@solvay.com
Abstract
This paper describes the preparation of a series bimetallic Pd-Ni catalysts supported over
nanoceria with very low Ni and Pd loading (<0.5 wt%) for the direct amination of aliphatic alcohols
with ammonia. Different metal impregnation methods were used for preparing the catalysts for
engineering the interaction between Pd, Ni and nanoceria. The as-prepared catalysts were
characterized in detail by combining XRD, H2-TPR, H2-TPD, STEM-EDS-SDD and XPS. The sequence
of impregnation of both metals and the Pd loading affected to an important extent the catalytic
activity by conditioning the crystallite size and the Pd and Ni speciation, as well as the reducibility
and reversible H2 storage properties governed by nanoceria. By optimizing the preparation
protocol, a 0.5wt%Pd-0.5wt%Ni-Pd/CeO2 formulation prepared by the sequential impregnation of
the nickel and palladium precursors afforded 80% yield to dioctylamine at almost full conversion
[TON = 1160 mmol / mmol (Ni+Pd)surface] in the direct amination reaction of 1-octanol with
o
ammonia at 180 C for 2 h. Ni leaching during the reaction could be completely avoided. The high
catalytic performance of Pd-Ni induced by nanoceria places this catalyst among the best ever
reported catalysts for the synthesis of secondary amines.
Keywords: 1-octanol, ammonia, H2 borrowing, nanoceria, palladium, nickel
1

Catalysis Science & Technology
Page 2 of 44
Submitted to Catal. Sci. Technol.
View Article Online
DOI: 10.1039/C8CY01670D
1. Introduction
Synthetic amines are extensively used in the chemical industry as solvents, agrochemicals,
pharmaceuticals, detergents, polymers and fabric softeners.1 Alkylamines constitute a particular
family of amines derived from fatty acids, olefins or alcohols issued either from natural sources
(fats, oils) or from petrochemical raw materials. Historically, alkylamines have been produced
from feedstocks such as nitriles, carboxylic acids, alkyl halides and carbonyl compounds using
processes often comprising hazardous reagents, generating salt residues or consuming H2
stoichiometrically (i.e. reductive amination). As a shift from fossil resources, biomass is
expected to supply a large portfolio of alcohols.2 As a result, the direct amination of alcohols
affording water as the sole byproduct could become a competitive pathway for the synthesis of
amines.
The direct amination of aliphatic alcohols with ammonia results in a distribution of
primary, secondary and tertiary amines.3 To tune the selectivity towards the desired degree of
alkylation, a catalyst is required. Raney Ni,4 as well as Ni5 and Co6 supported over alkaline or
amphoteric oxides (e.g., ,-Al2O3) are known to favor the formation of primary amines at
o
temperatures below 180 C with yields higher than 70% driven by a H2 borrowing mechanism.
However, at higher temperatures and at low NH3 excess with respect to the alcohol (>3 equiv),
these formulations show the formation of kinetic mixtures of primary, secondary and tertiary
amines for short-chain alcohols (<C4) with yields well below 50%.6a,7 In parallel, Cu8 and NiCu9
catalysts enhance the formation of secondary and tertiary amines, but also offering poor yields.
All these formulations usually encompass large metal contents (most often >15 wt%), as well as
heterogeneous particle size distributions and low metal dispersions. Furthermore, these catalysts
are prone to metal leaching upon exposure to ammonia and polar solvents, thereby favoring the
contamination of the amine product by the metal.
2

Page 3 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
Besides non-noble metal formulations, very few examples of catalysts based on noble
DOI: 10.1039/C8CY01670D
metals have been reported. Noble metals such as Ru and Pd usually suffer from lower activity
for alcohol amination under the H2 borrowing mechanism and require promoters and co-
catalysts.5c,10 In particular, Pd-supported catalysts such as Pd/MgO,11 Pd/Fe2O3,12 and Pd/K-
OMS-213 are mostly active for the direct amination of aromatic alcohols and/or amines and
promote the synthesis of highly substituted amines driven by amine-imine coupling on the
catalyst surface.
Ceria is a well-known promoter/stabilizer that can modify the structural or electronic
properties of catalysts.14 Recently, we among other authors have shown that ceria can enhance
the intrinsic activity of supported noble metal catalysts in C-N coupling reactions on activated
alcohols. Tomishige and co-workers earlier reported that ceria could rapidly produce aromatic
imines by direct condensation of aromatic alcohols and aniline under air at room temperature.15
Corma and co-workers reported that ceria-promoted Au nanoparticles (Au/CeO2, 2.5 wt% Au)
could afford the one-pot synthesis of propargylamines from N-benzylamine and benzyl alcohol
with 96% alcohol conversion and 90% yield at mild conditions.16 Some of us also reported a
highly active and selective Pd/CeO2 catalyst based on nanoceria (i.e. CeO2-HS) for the direct
amination of benzyl alcohol with aniline.17 Opposing other supports with variable acid-base and
redox properties, nanoceria was able to enhance the catalytic activity of Pd driven by the low-
temperature reducibility of the support below the reaction temperature (160 oC) combined with a
promoted H2 transfer ascribed to (back)spillover between the Pd nanoparticles and nanoceria.
Herein, we report novel bimetallic Pd-Ni formulations supported over nanoceria based on
very low metal loading (0.5wt%Ni and 0.5 wt%Pd) for the direct amination of aliphatic alcohols
with ammonia. By optimizing the preparation protocol, such formulations are able to overcome the
main drawbacks of current catalysts for the direct amination of aliphatic alcohols with ammonia:
(1) both Ni and Pd combined together show high activity at very low loadings, (2) the amine
selectivity can be modulated by tuning the interaction between Pd, Ni and nanoceria, and (3) Ni
3

Catalysis Science & Technology
Page 4 of 44
Submitted to Catal. Sci. Technol.
View Article Online
leaching during the reaction can be completely avoided. A case example is provideDdOfI:o10r.1t0h39e/Cd8iCrYe0c1t670D
amination of 1-octanol with ammonia, targeting the selective formation of the secondary amine (i.e.
dioctylamine) at the expense of primary and tertiary amines.
2. Experimental
2.1. Materials
High-surface-area CeO2 (CeO2-HS, Solvay HSA5, 198 m2/g, 0.15 cm3/g, 2-5 nm), low-surface-
area CeO2 (CeO2-LS, Sinopharm, 2 m2/g), γ-Al2O3 (Alfa Aesar, 200 m2/g), SiO2 (Evonik Aerosil®200,
200 m2/g) and MgO (Strem Chemicals, 230 m2/g) were used as supports for catalysts synthesis.
Palladium(II) nitrate dihydrate [(Pd(NO3)22H2O, Sigma-Aldrich, 40% Pd basis] and nickel nitrate
hexahydrate [Ni(NO3)2.6H2O, Sinopharm, 20%Ni basis] were used for Pd and Ni impregnation over
the supports, respectively. Ammonia, supplied by Air Liquide (99.8%), and 1-octanol, supplied by
J&K (99.5% purity), were used in the catalytic tests. Octylamine, dioctylamine, trioctylamine and
octanenitrile standards for GC calibration were all purchased from J&K (purity 99.5%). All the
reactants were used without further purification.
2.2. Catalyst preparation
Palladium and nickel were impregnated by Incipient Wetness Impregnation (IWI) over the
o
different supports (i.e. SiO2, MgO, -Al2O3, low-surface CeO2) pre-calcined at 300 C for 2 h, and
o
o
nanoceria (CeO2-HS) pre-calcined either at 300 C or 500 C for 2 h. Three different impregnation
precursor solutions were prepared (data given for 0.5 wt% metal basis): (1) an aqueous solution of
the Pd precursor by dissolving 0.0275 g of the precursor into 0.5 mL of deionized water; (2) an
aqueous solution of the Ni precursor by dissolving 0.0495 g of the precursor into 0.5 mL of
deionized water; and (3) an aqueous solution of combined Pd and Ni precursors by dissolving
0.0275 g of the Pd precursor and 0.0495 g of the Ni precursor into 0.5 mL of deionized water. The
monometallic Pd and Ni catalysts (0.5-2 wt% loading) were prepared by adding slowly solutions
(1) and (2), respectively, to 2.0 g of the given support under stirring. After impregnation for 2 h, the
samples were dried at 80 °C overnight and then calcined at 400 °C for 2 h under air.
4

Page 5 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
In the case of bimetallic Pd-Ni/CeO2-HS catalysts, three impregnation methods wereDuOsI:e1d0.:1039/C8CY01670D
i) First, impregnation of Ni with solution (1) over 2.0 g of pre-calcined CeO2-HS under stirring,
o
o
drying at 80 C for 2 h and calcination at 400 C for 2 h; subsequently, Pd impregnation with
solution (2), drying at 80 oC overnight and calcination at 400 °C for 2 h under air.
ii) First, impregnation of Pd with solution (1) over 2.0 g of pre-calcined CeO2-HS under stirring,
o
o
drying at 80 C for 2 h and calcination at 400 C for 2 h; subsequently, Ni impregnation with
solution (2), drying at 80 °C overnight and calcination at 400 °C for 2 h under air.
iii) Coimpregnation of Pd and Ni with solution (3) over 2.0 g of pre-calcined CeO2-HS under
stirring, drying at 80 oC overnight and calcination at 400 °C for 2 h under air.
For the sake of clarity, the nomenclature xPd-yNi, yNi-xPd and (xPd-yNi) is hereinafter used
referring to nanoceria-supported Pd-Ni bimetallic formulations prepared by methods i, ii and iii,
respectively. In this nomenclature, the terms ‘x’ and ‘y’ refer to the nominal weight fraction of Pd
and Ni in the catalysts, respectively.
2.3. Catalyst characterization
The bulk metal composition in the different samples was quantified by inductively coupled
plasma analysis (ICP) using an Activa (Horiba Jobin-Yvon) Optical Emission Spectrometer. Before
the measurements, the samples were dissolved using a mixture of inorganic acids (H2SO4, HNO3 and
HF).
The phases present in the different catalysts were analyzed by powder X-ray diffraction
(PXRD). The PXRD patterns were recorded on a Rigaku D/Max-2200/PC diffractometer provided
with Cu-Kα radiation (λ = 1.5418 Å) and a beam voltage of 45 kV. The spectra were collected in the
2θ range 8-80° with a step size of 0.02°. The patterns were indexed using the Joint Committee on
Powder Diffraction (JCPDS) database and interpreted using MDI JADE 5.0 software. The Scherrer
equation was used to estimate the average size of the oxide particles from the XRD line broadening.
5

Catalysis Science & Technology
Page 6 of 44
Submitted to Catal. Sci. Technol.
View Article Online
The specific surface area and pore volume of the different catalysts was meDaOsuI: r10e.1d03f9r/oC8mCYN012670D
adsorption/desorption isotherms at 77 K using a Micromeritics ASAP 2010 Surface Area Analyzer.
The surface areas were calculated by the Brunauer-Emmett-Teller (BET) method in the relative
pressure range 0.05 < P/P0 < 0.25, while the pore volumes were measured at P/P0=0.81. The
Barrer-Joyner-Halenda (BJH) method was used for measuring the interparticle pore size
o
distributions.18 Prior to the measurements, the catalysts were degassed at 180 C for 3 h under
vacuum (0.5 mbar).
Thermogravimetric analyses (TGA) were conducted using a TA Instruments SDT 2960
apparatus equipped with a flow gas system. The catalyst was treated from room temperature to
1000 °C using a heating rate of 5 °C.min-1 under a 20% O2/He mixture with a flow rate of 75
mL(STP).min-1. Approximately, 10 mg of sample was heated in an open alumina crucible.
The local composition of the catalysts was inspected by STEM-EELS/EDS using a 200 kV Tecnai
F20 microscope equipped with a FEG electron gun, a STEM unit and an EDAX Optima T60 SDD-EDS
spectrometer. The images were analyzed using EDAX Team microanalysis software. Before the
analysis, the given solid powder was directly dispersed over a holey carbon Cu 400 mesh grid
(Agar, Ref S147-4). The images were denoised using HyperSpy software. The Ni/Pd molar ratios
were measured by multiplying the relative intensities of both elements with the factor 1.59
corresponding to the ratio of emission yields between Pd and Ni.
The surface composition of the Pd-Ni/CeO2-HS catalysts was analyzed by X-ray photoelectron
spectroscopy (XPS) using a Kratos Axis Ultra DLD apparatus equipped with a hemispherical
analyzer and a delay line detector. The spectra were recorded using an Al monochromated X-ray
source (10 kV, 15 mA) with a pass energy of 40 eV (0.1 eV/step) for high resolution spectra, and a
pass energy of 160 eV (1 eV/step) for survey spectrum in hybrid mode and slot lens mode,
respectively. The adventitious C1s binding energy (285.0 eV) was used as an internal reference.
Temperature-programmed reduction and desorption (H2-TPR and H2-TPD, respectively)
were carried out on a Micromeritics AutoChem II 2920 system equipped with a thermal
6

Page 7 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
conductivity detector (TCD) to monitor the changes in gas composition, a CryoCooler for sub-
DOI: 10.1039/C8CY01670D
ambient (-100 °C) measurements, and a cold trap before the detector. The H2-TPR profiles were
measured using a heating rate of 10 K/min under a 10% H2/Ar flow (40 mL(STP)/min). H2-TPD
was used to measure the Pd dispersion and the reversible H2 storage by spillover on the supports.
The following protocols were used:
1. Reduction of the sample at 200 °C under H2 flow (40 mL(STP)/min) and evacuation in the
o
o
temperature range from -80 C to 400 C under Ar flow (20 mL(STP)/min) using a 10
K/min heating ramp.
o
2. Cooling under Ar flow (20 mL(STP)/min) until -80 C followed by H2 chemisorption by
injecting 10% H2/Ar pulses.
o
o
3. Evacuation from -80 C to 400 C under Ar flow (20 mL(STP)/min) using a 10 K/min
heating ramp.
Noteworthy, the evacuation conditions were settled to ensure reversible H2 uptake on Pd/CeO2
samples.19
The following expressions were used for monometallic catalysts (M = Ni, Pd) by implementing the
H2-TPR and H2-TPD data in the temperature range from -80 oC to 400 oC
g
M
MW,M
mol
(
)
M
mol H2
gcat
―80→40oC
M dispersion (%) = 2 Φ
(1)
(2)
(3)
(
)
H2,TPD
xM (
g
M
g
)
cat
g
MW,CeO
mol
(
)
cat
mol H2
gcat
2
50→200oC
H2,TPR
Reducibility (red)(%) = 2 Φ
(
)
νCe
o
C
Φ4H0→400
,TPD
2
Reversibility (rev)(%) =
o
C
Φ5H0→200
,TPR
2
50→200oC
H2,TPR
o
o
Φ
where
indicates the amount of H2 uptake measured by H2-TPR from 50 C to 200 C,
50→200oC
H2,TPR
40→400oC
refer to the amount of H2 desorbed from -80 oC to 40 oC and from 40 oC to
Φ
Φ
H2,TPD,i
and
7

Catalysis Science & Technology
Page 8 of 44
Submitted to Catal. Sci. Technol.
o
View Article Online
400 C , respectively, characterizing the metal dispersion and the reversible H2 DsOtIo: 1r0a.1g0e39/oCn8CtYh01e670D
support by spillover, and Ce designates the stoichiometric number of Ce(IV).
In addition to H2-TPD, CO pulse chemisorption was also used for measuring the metal
dispersion. The tests were carried out on a Micromeritics AutoChem 2920 instrument. Briefly, a
known quantity of calcined catalyst (100 mg) was reduced under a 10% H2-Ar flow (40
mL(STP)/min) at 580 oC for 30 min using a heating rate of 10 oC min-1. Afterwards, the sample was
o
cooled down to 0 C and then CO chemisorption was performed by introducing successive pulse
doses of 10% CO-He until no further CO uptake was measured. The stoichiometry factor between
the metal atoms and CO molecules was assumed to be 1.
2.4. Catalytic activity measurements
The activity of the different catalysts was evaluated in the liquid-phase amination reaction of
1-octanol (OL) with NH3 (Scheme 1). The catalytic tests were conducted in a 30-mL stainless steel
autoclave equipped with a pressure gauge and a safety valve. In a typical experiment, the reactor
was charged with OL (1 mL, 6.3 mmol) and 150 mg of pre-reduced catalyst. The reactor was sealed
and evacuated by applying vacuum followed by charging NH3 (7 bar). The reactor was then placed
on a hot plate equipped with a magnetic stirrer for different reaction times and temperatures. At
these conditions, the nominal NH3/OL molar ratio was about 4.6-8.0, while the NH3/OL molar ratio
in the liquid phase was about 0.20.
The reactant (OL) and the expected N-products, i.e. octylamine (OA), dioctylamine (DOA),
trioctylamine (TOA), octanenitrile (ON) and dioctylimine (DOI), were analyzed and quantified using
an Agilent 7890 GC equipped with a HP-5 capillary column with 5 wt% phenyl groups and using
biphenyl as internal standard.
The OL conversion, selectivity, yield to the N-containing products and turnover number
(TON) for DOA formation, as well as the carbon balance, were defined as follows:
Conversion (%) = 1 ― 푛푛
푂퐿
(4)
0
푂퐿
8

Page 9 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
푛푖
View Article Online
Selectivity (i)(%) = 푛
i = OA, DOA, TOA, ON
(5)
DOI: 10.1039/C8CY01670D
0
푂퐿 ― 푛푂퐿
푛푖
Yield (i)(%) = 푛
= OA, DOA, TOA, ON
(6)
0
푂퐿
n0OL YieldDOA
TONDOA
=
 = B (bulk), S (surface)
(7)
(8)
mol (Ni + Pd)훼
Carbon balance (%) = moles of formed products + reactants
moles of reactants
where nOL0 and nOL refer to the initial and final OL mole number and ni indicates the mole number of
the N-containing products formed.
2.5. Thermodynamic calculations
Our system was regarded to consist of 13 reactions, including the direct reaction of OL with
NH3, OA and DOA (R1-R5), DOI hydrogenation (R6), and a set of disproportionation reactions
between the different amines (R7-R9), between OA/DOA and ON (R10-R11), and between OA/DOA
and DOI (Eq R12-R13). The reactions can be classified in three families:
Direct amination reactions:
+
H2O
+
NH2
NH3
OH
OH
OH
R
R
R
R
(R1)
+
+
NH3
N +
R
H2O
2H2
(R2)
(R3)
H2
+
H2O
H2O
+
NH2
NH2
N
+
R
R
R
R
R
+
+
N
H
+
R
R
OH
OH
R
R
(R4)
(R5)
H2O
+
N
N
H
R
R
R
R
R
DOI hydrogenation:
H2
+
N
H
N
R
R
R
(R6)
9

Catalysis Science & Technology
Page 10 of 44
Submitted to Catal. Sci. Technol.
View Article Online
DOI: 10.1039/C8CY01670D
Disproportionation reactions:
NH3
+
+
+
NH2
N
H
R
NH2
NH2
R
R
R
R
R
(R7)
(R8)
N
NH3
N
H
R
+
R
R
R
N
N
H
+
NH2
R
R
R
R
N
H
R
+ R
R
R
2H2
+
R
R
R
(R9)
+
NH2
+
N
N
R
+
R
N
NH3
H
(R10)
+
NH3
+
R
R
N
R
2H2
R
R
N
H
R
R
(R11)
(R12)
H2
+
NH2 +
N
NH3
+
R
N
N
R
N
R
R
R
R
R
+
NH3
+
+
N
R
3H2
R
R
(R13)
2.5.1. Calculation of equilibrium constants
The standard free energy of the R1-R13 reactions was calculated using Aspen Plus® software
(version 8.8). Table S1 lists the values as a function of temperature. In the calculations, the standard
free energy of reaction at 298 K, Gor,298, was computed from the standard free energy of formation
of the reactants (R) and products (P), Gof,298, obtained from the APV88 PURE32 databank, as
follows
푃
푅
∆G표푗,298
=
푖 = 0∆G표푓,푖,298
―
푖 = 0∆G표푓,푖,298
∑
∑
(9)
The standard free energy was corrected to the given temperature, T, using the Kirchoff equation
푃
푅
∆G표푗,푇 = ∆G표푗,298
∑ ∑
+ 푖 = 0퐶푝,푖(푇 ― 298) ― 푖 = 0퐶푝,푖(푇 ― 298)
(10)
Finally, the equilibrium constant, Keq, for each reaction was computed using the expression
표
(
)
∆G푗,푇 = ― RT ln 퐾푒푞
(11)
10

Page 11 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
where R is the gas constant of perfect gases (8.314 Pa.m3.mol-1.K-1).
View Article Online
DOI: 10.1039/C8CY01670D
2.5.2. Calculation of the equilibrium distribution of amines
The equilibrium distribution of amines was calculated from the equilibrium constants of the
different reactions at a given temperature and for various NH3/OL molar ratios in the reactor feed
using the REquil Reactor Block in Aspen Plus® operating at ideal conditions. In these conditions, the
fugacities were equaled to partial pressures, while the activity coefficients of the different liquid
compounds were approached to 1. The reaction system was simulated as a gas-liquid isothermal
flash reactor with the same geometry and operation conditions as our experimental autoclave. The
total initial mole number of OL was set to 6.3 mmol, whereas the total pressure was adjusted to
38.5 bar. Two sets of calculations were implemeted without H2 and with 5 bar H2 pressure.
3. Results and Discussion
3.1. Thermodynamics of OL amination with NH3
In a first step of our study, we studied the thermodynamics of the OL amination reaction with
NH3 using Aspen® software. Figure 1A1,A2 plots the evolution of the equilibrium constant with the
temperature for the R1-R13. Except for reactions R7 and R10, the equilibrium constants decrease
with the temperature, as expected for exothermic reactions. From a thermodynamic perspective,
TOA formation is favored by the direct reaction of OL with DOA (R5) with an equilibrium constant
o
evolving from 1.09x104 to 946 in the temperature range 150-250 C. TOA is also favored by the
disproportionation reactions between DOA and ON (R11), DOI and OA (R12) and DOI and ON (R13)
with equilibrium constants in the range 6.10x105-18, 2.5x104-226 and 3.75x107-332, respectively,
showing all a prominent decline with the temperature. Unlike TOA, DOA formation is only favored
by the direct amination reaction between OA and OL (R4) with an equilibrium constant in the range
1278-178. OA is discouraged compared to DOA with the equilibrium constant varying in the range
31-15 in the temperature range considered (R1). Likewise, ON is thermodynamically discouraged
with an equilibrium constant in the range 8.5-5.3 (R2), whereas DOI is favored (R3) (5.19x105-93).
11

Catalysis Science & Technology
Page 12 of 44
Submitted to Catal. Sci. Technol.
View Article Online
These results point out the facilitated formation of TOA compared to OA and DOAD, OeIs: p10e.1c0i3a9l/lCy8CaYt01a670D
temperature below 200 oC, mainly driven by the set of disproportionation reactions R11-R13.
By implementing the equilibrium constants into the reactor model, we calculated the
equilibrium OL conversion and selectivity to the different N-products as a function of the nominal
feed NH3/OL molar ratio and the reaction temperature. Figure 1B plots the results obtained both in
the absence (B1) and in the presence (B2) of 5 bar H2 pressure. In both cases, irrespective of the
temperature, full OL conversion is predicted for a nominal NH3/OL molar ratio higher than 0.50.
Among the three amines, TOA is predominant with a selectivity decreasing abruptly with the
NH3/OL molar ratio in the range 0.25-1.0 until an asymptotic value. The opposing trend is observed
for the OA and DOA selectivities, the latter displaying larger asymptotic values at a given
temperature. These results point out the possibility of achieving high DOA selectivity in detriment
to TOA at high OL conversion by designing a catalyst kinetically inhibiting the disproportionation
reactions between DOA and ON (R11), DOI and OA (R12) and DOI and ON (R13). We show below
the Pd-Ni formulations over nanoceria are good candidates to meet this requirement.
3.2. Performance of 2Pd/CeO2-HS catalysts over different supports
In a first series of experiments, we explored the performance of Pd-supported catalysts (2
wt%, unreduced) over SiO2, MgO, Al2O3, CeO2-LS and CeO2-HS pre-calcined at 300 oC for 2 h
in the amination reaction of OL with NH3. Figure 2 and Table S2 (entries 1-12) compile the OL
conversion and selectivity to the different N-products for the different catalysts.
2Pd/SiO2, 2Pd/MgO, 2Pd/Al2O3 and 2Pd/CeO2-LS display very low activity with respect
to the bulk ‘B’ metal (TONDOA <10 mmol DOA / mmol PdB) (Table S2, entries 1-8). The
introduction of a H2 pressure exerts a negative effect on the OL conversion for 2Pd/SiO2 and
2Pd/Al2O3. Opposing this behavior, 2Pd/CeO2-HS exhibits a much higher activity with a
TONDOA up to 132 mmol DOA / mmol PdB after pre-reduction and in the presence of 5-bar H2
pressure during the reaction (Table S2, entries 9-12). At these conditions, the OL conversion
reaches a value of 72% with a selectivity of 5.3%, 51% and 13% towards OA, DOA and TOA,
12

Page 13 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
respectively. The DOA yield shows a value of 35%. In addition to amines, ON isDaOlI:s1o0.1g0e39n/Ce8rCaYte01d670D
in the absence of H2, but with very low selectivity (<2.0%). The presence of a H2 pressure (5
bar) during the reaction exerts a positive effect on the OL conversion at the expense of TOA, but
affects negatively the carbon balance. These results highlight the positive effect of nanoceria on
enhancing the catalytic activity of Pd for the amination of OL with NH3 most likely by
enhancing DOI hydrogenation on the catalyst surface (R6) and discouraging in turn the
disproportionation reaction R12 and R13 leading to TOA.
3.3. Performance of 2Pd-0.5Ni/CeO2-HS catalysts
Given the synergistic behavior between nanoceria and Pd for OL amination with NH3, we
concentrated our further attention on refining the 2Pd/CeO2-HS formulation for enhancing the
selectivity towards DOA and the carbon balance while discouraging TOA formation. To this aim,
a series of bimetallic catalysts based on 2 wt%Pd and low Ni loading (0.5 wt%) were prepared
o
over nanoceria pre-calcined at 300 C using different impregnation sequences followed by
o
calcination at 400 C. The catalytic tests were further conducted by pre-reducing the catalysts
before the reaction, and by introducing a H2 pressure (5 bar) during the reaction. Figure 3 and
Table S2 (entries 13-15) compile the results obtained. For comparison, the results measured on
0.5Ni/CeO2-HS and 5Ni/CeO2-HS, as well as on the benchmark 20Ni/Al2O3, are also included
(Table S2, entries 16-18).
The incorporation of 0.5 wt%Ni to 2Pd/CeO2-HS results in an increase of the OL
conversion from 72% for 2Pd/CeO2-HS to 81% for 0.5Ni-2Pd/CeO2-HS (impregnation
sequence: 1st 2Pd, 2nd 0.5Ni) at the expense of the carbon balance, decreasing from 78% for
2Pd/CeO2-HS to 56% for 0.5Ni-2Pd/CeO2-HS. In contrast, for coimpregnated (2Pd-
0.5Ni)/CeO2-HS and 2Pd-0.5Ni/CeO2-HS (impregnation sequence: 1st 0.5Ni, 2nd 2Pd), the OL
conversion keeps almost unchanged with an improved carbon balance, increasing from 78% for
2Pd/CeO2-HS to 94% and 92% for (2Pd-0.5Ni)/CeO2-HS and 2Pd-0.5Ni/CeO2-HS, respectively.
13

Catalysis Science & Technology
Page 14 of 44
Submitted to Catal. Sci. Technol.
View Article Online
Furthermore, for the latter catalysts, a pronounced enhancement of the DOA selectivity to more
DOI: 10.1039/C8CY01670D
than 60% is observed in detriment of TOA, encompassing a DOA yield higher than 40% and a
TONDOA of 71 mmol DOA / mmol (Ni+Pd)B. For comparison, a control 0.5Ni/CeO2-HS catalyst
displays no activity, whereas a 5Ni/CeO2-HS control catalyst displays preferential OA formation.
Finally, a catalytic test on a benchmark 20Ni/Al2O3 catalyst pre-reduced at 550 oC for 1 h under
H2 flow exhibits a DOA yield of 45%, which is comparable to that of (2Pd-0.5Ni)/CeO2-HS
(47%), but at much lower TONDOA (3.6 mmol DOA / mmol NiB). This body of results points out
a synergistic effect of low-Ni loading on the catalytic properties of Pd over CeO2-HS for DOA
synthesis.
3.4. Performance of 0.5Pd-0.5Ni/CeO2-HS
The positive effect of Ni on modulating the selectivity towards DOA and improving the
carbon balance prompted us to optimize the (2Pd-0.5Ni)/CeO2-HS and 2Pd-0.5Ni/CeO2-HS
formulations by reducing the Pd loading. To this aim, a series of formulations based on 0.5
wt%Pd and 0.5 wt%Ni were prepared using different impregnation protocols (vide supra). The
catalysts were further tested in the amination reaction of OL with NH3. In these tests, the
catalysts were pre-reduced and the reaction was conducted under 5-bar H2 pressure.
The results plotted in Figure 4 and also listed in Table S3 point out a dramatic
enhancement of the OL activity and DOA selectivity in detriment of TOA for (0.5Pd-
0.5Ni)/CeO2-HS and 0.5Pd-0.5Ni/CeO2-HS (impregnation sequence: 1st 0.5Ni, 2nd 0.5Pd)
compared to the parent 0.5Pd/CeO2-HS. As a matter of fact, the DOA selectivity and yield for
0.5Pd-0.5Ni/CeO2-HS are 65% and 52%, respectively. Interestingly, this yield is higher than the
value measured on the benchmark 20Ni/Al2O3 (45%) (Table S2 entry 18). Regardless of the use
or not of a catalyst pre-reduction step, the presence of H2 (5 bar) appears to be crucial for
promoting the OL conversion and discouraging TOA formation for 0.5Pd-0.5Ni/CeO2-HS
(Figure S1, Table S3 entries 5-8). A further decrease of the Pd loading down to 0.1 wt% for this
14

Page 15 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
formulation while keeping the Ni loading at 0.5 wt%, i.e. 0.1Pd-0.5Ni/CeO2-HS, results in a
DOI: 10.1039/C8CY01670D
pronounced reduction of the OL conversion with a concomitant decline of the DOA selectivity.
As a matter of fact, the OL conversion and DOA selectivity are 80% and 65%, respectively, for
0.5Pd-0.5Ni/CeO2-HS, whereas the corresponding values for 0.1Pd-0.5Ni/CeO2-HS are 37% and
38% (Figure S2A, Table S3 entries 3 and 8). A similar decreasing trend of the OL conversion
and DOA selectivity is observed when comparing the coimpregnated (0.1Pd-0.5Ni)/CeO2-HS
and (0.5Pd-0.5Ni)/CeO2-HS catalysts (Figure S2B, Table S3 entries 2 and 4). For comparison,
the control 0.5Pd-0.5Ni/Al2O3 and (0.5Pd-0.5Ni)/Al2O3 catalysts display almost no activity
(Table S4 entries 1 and 2).
With these results in hand, we explored the catalytic trends towards DOA formation for the
Pd-Ni/CeO2-HS formulations prepared using the different Pd and Ni impregnation protocols.
Figure 5 shows a cartography of the different catalysts using a pre-reduction step and a 5-bar H2
pressure during the reaction. As a rule, at constant Ni loading, Pd impregnation over nanoceria
results in an increase of the OL conversion, which becomes more pronounced at higher Pd
loading. As an exception, 0.5Pd-0.5Ni/CeO2-HS exhibits a higher activity compared to 2Pd-
0.5Ni/CeO2-HS (impregnation sequence: 1st 0.5Ni, 2nd 0.5-2Pd), suggesting a synergistic effect
between Ni and Pd in the former catalyst. In terms of selectivity, Pd is detrimental for OA,
whereas it favors DOA formation. Unexpectedly, the TOA selectivity does not scale with the Pd
loading at constant Ni loading (0.5 wt%). Indeed, irrespective of the impregnation protocol, the
formulations containing 0.5 wt%Pd and 0.5wt%Ni [i.e. (0.5Pd-0.5Ni)/CeO2-HS, 0.5Pd-
0.5Ni/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS] display the highest DOA selectivity. Among the
three formulations, in terms of catalytic activity, 0.5Pd-0.5Ni/CeO2-HS exhibits the highest
TONDOA with a value of 244 mol DOA / mol (Ni+Pd)B. Overall, this body of results reinforces
the synergistic effect between Pd and Ni in 0.5Pd-0.5Ni/CeO2-HS for boosting the catalytic
activity towards DOA formation.
15

Catalysis Science & Technology
Page 16 of 44
Submitted to Catal. Sci. Technol.
View Article Online
3.5. Effect of the operation variables on the catalytic properties of 0.5Pd-0.5Ni/CeO2-HDSOI: 10.1039/C8CY01670D
Given the promising catalytic properties of 0.5Pd-0.5Ni/CeO2-HS, we surveyed the effect of
the different operation variables on the catalytic performance of this formulation. In the different
tests, the catalyst was pre-reduced and a H2 pressure (5 bar) was used. In the following lines we
show the effect of the temperature, reaction time and catalyst weight on the OL conversion and
selectivity to the different N-products, as well as the main trends of the DOA selectivity and TONDOA
as a function of the OL conversion.
3.5.1. Effect of temperature
Table S5 lists the effect of temperature on the catalytic performance of 0.5Pd-0.5Ni/CeO2-HS.
The OL conversion is very low at temperatures lower than 150 oC, whereas it becomes dramatically
o
o
promoted above 180 C. Meanwhile, the DOA selectivity exhibits an optimal value at 180 C with a
value up to 65%. In parallel, the carbon balance decreases drastically with the temperature until of
32% at 210 oC. These results point out the convenience of 180 oC for conducting the catalytic tests.
3.5.2. Effect of the catalyst loading
We further explored the effect of the catalyst loading on the catalytic performance of 0.5Pd-
o
0.5Ni/CeO2-HS for OL amination with NH3 at 180 C. Figure S3 plots the experimental trend
obtained, whereas Table S6 lists the complete body of results. An increase of the catalyst loading in
the range 80-350 mg results in an enhancement of the OL conversion from 47% to 99%. In parallel,
the OA selectivity decreases from 29% to 13%, while the DOA selectivity is kept in all cases at a
high value (up to 79%). A catalyst loading higher than 170 mg keeps the OA, DOA and TOA
selectivities almost unchanged.
3.5.3. Effect of the reaction time
Figure 6 plots the effect of the reaction time on the catalytic performance of 0.5Pd-
o
0.5Ni/CeO2-HS at 180 C for a catalyst loading in the range 160-180 mg, while Table S7 lists the
results obtained. The OL conversion shows a dramatic increase from 30% after 30 min to 91% after
2 h, reaching 100% after 4 h. In the meantime, the DOA selectivity only shows a moderate increase
16

Page 17 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
from 66% after 30 min to 80% after 2 h, while the OA selectivity declines monotonously from 37%
DOI: 10.1039/C8CY01670D
to 13%. The TOA selectivity only exhibits a moderate increase from 1.6% after 30 min to 7.2% after
4 h. In terms of catalytic activity, the TONDOA increases dramatically with the reaction time from 68
mol DOA / mol (Ni+Pd)B after 30 min to 248 mol DOA / mol (Ni+Pd)B after 4 h.
Figure 7 plots the evolution of the OA, DOA and TOA selectivities, as well as the TONDOA, as a
function of the OL conversion for 0.5Pd-0.5Ni/CeO2-HS. The OA selectivity decreases monotonously
with the OL conversion, while the DOA and TOA selectivity exhibit a slight increasing trend, the
latter keeping at low values (<10%). An almost linear increasing trend is observed for the TONDOA
with the OL conversion at values below 50%.
3.6. Understanding the structure and catalytic behavior of 0.5Pd-0.5Ni/CeO2-HS
The catalytic results for the OL amination reaction presented above point out an enhanced
activity and selectivity to the secondary amine for 0.5Pd-0.5Ni/CeO2-HS based on the sequential
impregnation of the Ni and Pd precursors over nanoceria. In light of these results, we explored in
more detail the structure of 0.5Pd-0.5Ni/CeO2-HS among the different Pd-Ni/CeO2-HS formulations
prepared in this study.
3.6.1. Textural properties and bulk metal composition
The textural properties of the Pd-Ni/CeO2-HS catalysts were measured by N2
o
adsorption/desorption at -196 C. Table 1 and Table S8 (columns 4-6) list the BET specific surface
area, total pore volume and average pore size of the catalysts based on 0.5 wt%Pd and 2.0 wt%Pd,
respectively. After metal impregnation and calcination, the different formulations exhibit a slightly
lower specific surface area compared to the value measured on the pre-calcined CeO2-HS (198
m2/g). In particular, (2Pd-0.5Ni)/CeO2-HS prepared by coimpregnation displays the lowest specific
surface area (120 m2/g), whereas the largest areas are observed for 0.5Pd/CeO2-HS (198 m2/g),
2Pd-0.5Ni/CeO2-HS (194 m2/g) and 0.5Ni-2Pd/CeO2-HS (184 m2/g). All the bimetallic catalysts
with 0.5 wt%Pd display slightly lower specific surface areas with a value of 155 m2/g for 0.5Pd-
0.5Ni/CeO2-HS. In parallel, the total pore volume and average pore size fall into the range 0.11-0.17
17

Catalysis Science & Technology
Page 18 of 44
Submitted to Catal. Sci. Technol.
3
View Article Online
cm /g and 3.1-3.5 nm, respectively, for the different formulations with no clear distiDnOcIt:i1o0.n10b39e/Ctw8CeYe01n670D
the formulations.
3.6.2. Structural properties
The XRD patterns of the different catalysts show in all cases the presence of a pure CeO2
phase (Figure S4). The average particle size of the CeO2 nanoparticles falls in the range 5.8-6.6 nm
as inferred by Scherrer equation applied to the (111) reflection. This size agrees well with the
o
values measured on the pre-calcined CeO2-HS at 300 C by XRD (5.6 nm, XRD) and HR-TEM (7.0
nm) (Table 1 and Table S8 column 3), matching an earlier report.17 The increase of the average
particle size of CeO2 nanoparticles combined with the presence of metal nanoparticles can be
regarded as key drivers for the decrease of the surface area of the catalysts after impregnation and
calcination. It is also worth mentioning that no additional reflections ascribed to either NiOx (2-
theta= 36.0o, 43.5o, JCPDS 47-1049) or PdOy (2-theta= 33.8o, JCPDS 43-1024) are observed, pointing
out a high dispersion of the metal phases over the nanoceria. Note however that, since the PdOy
reflection matches the (200) reflection of ceria, potential superposition of both reflections cannot
be ruled out. Further inspection of 0.5Pd-0.5Ni/CeO2-HS debris by STEM-EDS-SDD confirms a
highly homogeneous Ni and Pd distribution on the ceria surface with intimate interaction between
both metals at the level of spatial resolution of the cartography (Figure S5). A similar distribution
for both metals is observed for (0.5Pd-0.5Ni)/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS (both not shown).
FT-IR spectroscopy was used to visualize the nature of the surface species adsorbed on the
calcined catalysts. Figure S6 collects the spectra measured on 0.5Ni/CeO2-HS, 0.5Pd/CeO2-HS,
0.5Pd-0.5Ni/CeO2-HS, (0.5Pd-0.5Ni)/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS, as well as for the parent
CeO2-HS. In all cases, surface carbonates are visible with characteristic bands centered at 1629-
1634 cm-1, 1505-1521 cm-1 and 1317-1324 cm-1 being indicative of I(C=O) and II(C=O) stretching
vibrations.20 Carbonates can only be desorbed from ceria by calcination at a temperature higher
than 700 oC. These observations suggest that stronger basic sites on the ceria surface are
neutralized by carbonate species even after calcination and that these sites are not expected to
participate in the amination reaction.
18

Page 19 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
DOI: 10.1039/C8CY01670D
3.6.3. Metal surface distribution
Figure 8 and Figures S7, S8 plot the deconvoluted XPS spectra of the Pd3d, Ce3d, O1s, and
Ni2p core levels for the monometallic 0.5Pd/CeO2-HS and 0.5Ni/CeO2-HS and the bimetallic 0.5Pd-
0.5Ni/CeO2-HS, (0.5Pd-0.5Ni)/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS catalysts after calcination,
whereas Tables S9-S12 list the BEs of the deconvoluted bands for the different core levels.
Furthermore, Table 1 (columns 7-12) compares the bulk ‘b’ and surface ‘s’ Ni/Ce, Pd/Ce and Ni/Pd
molar ratios measured by ICP-OES and XPS, respectively.
The deconvoluted Ce3d core-level spectra could be assigned to eight spin-orbit coupling
levels corresponding to the 3d5/2 (v) and 3d3/2 (u) states (Figure S7A). The v components show
binding energies (BE) at 880.4-880.8 eV (v0), 882.2-882.6 eV (v), 884.4 eV (v’), 888.7-888.9 eV (v”)
and 898.2-898.5 eV (v’”), while the u components show BEs at 896.6-896.8 eV eV (u0), 900.8 eV (u),
902.3 eV (u’), 907.2-907.5eV (u”) and 916.6 eV (u’”).21 These components match the values
measured on the deconvoluted XPS spectrum of the metal-free parent. The presence of v’ and u’
components indicates the presence of Ce3+ in the first layers of the ceria nanoparticles and the
relative amount of Ce3+ does not change substantially after metal impregnation. The deconvoluted
XPS spectra of the O1s core level for the different catalysts (Figure S7B) show the presence of two
bands centered at 529.4-529.8 eV (OI) and 531.1-532.0 eV (OII) that can be ascribed to lattice and
weakly bounded oxygen (e.g., hydroxyl groups), respectively.21b,22 The contribution of the former
band is about 63-68% for 0.5Pd/CeO2-HS, 0.5Ni/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS, matching the
value of the parent ceria, whereas it decreases to 56% for (0.5Pd-0.5Ni)/CeO2-HS and 0.5Pd-
0.5Ni/CeO2-HS.
The XPS spectra of the Ni2p core level for the calcined 0.5Ni/CeO2-HS, 0.5Ni-0.5Pd/CeO2-HS,
(0.5Pd-0.5Ni)/CeO2-HS and 0.5Pd-0.5Ni/CeO2-HS exhibit a broad band corresponding to the Ni2p3/2
core level centered at 855.0-856.0 eV (Ni2p3/2) that is indicative of Ni2+ (Figure S8). A broad band is
also present at 861.4-862.1 eV that can be assigned to the shake-up satellite band of NiO.23 The XPS
spectra of the Pd3d core level for the different catalysts display a characteristic spin-orbit doublet
that can be assigned to the Pd3d5/2 and Pd3d3/2 components (Figure 8). Each component can be
19

Catalysis Science & Technology
Page 20 of 44
Submitted to Catal. Sci. Technol.
View Article Online
deconvoluted into three species with BEs: (1) 336.6-337.3 eV, (2) 337.8-338.0 eVDaOnI:d10.(1033)9/3C83C9Y.081-670D
339.2 eV (Pd3d5/2), and (1) 342.0-342.4 eV, (2) 343.1-343.9 eV and (3) 343.8-344.4 eV (Pd3d3/2),
and a spin-orbit energy of 5.2-5.6 eV. Given the high BEs, we can attribute the first species to Pd2+,
whereas the second and third species can be attributed to Pd4+ with variable dispersion level over
ceria. Interestingly, 0.5Pd/CeO2-HS, 0.5Ni-0.5Pd/CeO2-HS and (0.5Pd-0.5Ni)/CeO2-HS show a
Pd2+/Pd4+ relative intensity of 1:6.6, 1:7.8 and 1:11, respectively, being accordingly enriched with
Pd4+. In contrast, 0.5Pd-0.5Ni/CeO2-HS displays a much higher Pd2+ with a Pd2+/Pd4+ relative
intensity 1:1.7 (Table S12).
3.6.4. Bulk and surface metal composition
The (Ni/Ce)b and (Pd/Ce)b bulk ratios measured by ICP-OES for 0.5Pd/CeO2-HS, 0.5Ni/CeO2-
HS, 0.5Ni-0.5Pd/CeO2-HS, (0.5Pd-0.5Ni)/CeO2-HS and 0.5Pd-0.5Ni/CeO2-HS after calcination are all
close to the nominal values (Table 1). The (Ni/Ce)b bulk ratio for Ni-containing catalysts is found in
the narrow range 0.013-0.014, which compares well with the nominal value (0.015). However, a
small yet systematic deficit of Pd is observed for Pd-containing catalysts, encompassing a (Pd/Ce)b
bulk ratio in the range 0.0053-0.0066 compared to the nominal value (0.0081). As a result, the
bimetallic catalysts exhibit a (Ni/Pd)b bulk ratio in the range 2.2-2.4 compared to the nominal value
of 1.8. These values are comparable even if slightly higher to the ratios measured by STEM-EDS-
SDD on sample debris.
Turning now our attention into the (Ni/Ce)S and (Pd/Ce)S surface ratios measured by XPS for
the bimetallic catalysts, both ratios are higher than the corresponding bulk ratios, as expected for a
surface enriched by both metals (Table 1). Notwithstanding this fact, when comparing the surface-
to-bulk (Ni/Ce) and (Pd/Ce) ratios [i.e. (Ni/Ce)S/(Ni/Ce)b and (Pd/Ce)S/(Pd/Ce)b], a higher surface
enrichment is observed for Ni compared to Pd. As a matter of fact, the (Ni/Ce)S/(Ni/Ce)b ratio is
found in the range 30-34, while the corresponding range for the (Pd/Ce)S/(Pd/Ce)b ratio is 3.0-7.0.
This observation can be explained either by a partial encapsulation of Pd in the ceria subsurface as
very small clusters, or by an atomic dispersion of Pd driven by a partial incorporation in the fluorite
20

Page 21 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
framework of ceria (Figure 8). This hypothesis is consistent with the stabilization of high-valent Pd
DOI: 10.1039/C8CY01670D
(i.e. Pd4+) on the ceria surface/subsurface, as inferred from the deconvoluted XPS spectra.
When comparing the surface-to-bulk (Ni/Pd) ratios [i.e. (Ni/Pd)S/(Ni/Pd)b] for the different
catalysts, (0.5Pd-0.5Ni)/CeO2-HS and 0.5Pd-0.5Ni/CeO2-HS display lower ratios than 0.5Ni-
0.5Pd/CeO2-HS (5.5 vs. 10), pointing out a higher surface enrichment by Pd in the former catalysts.
In addition to highly dispersed Pd4+, 0.5Pd-0.5Ni/CeO2-HS shows preferential surface enrichment
by Pd2+ species, suggesting the presence of PdO nanoparticles in strong contact with NiO on the
ceria surface. Overall, this body of results points out that the Pd speciation and location on the ceria
surface/subsurface depends to an important extent on the impregnation sequence of the Ni and Pd
precursors, and that partial Pd2+ enrichment acts as a positive driver for enhancing the catalytic
activity.
3.6.5. Catalyst reducibility
The reducibility of the different catalysts was studied by H2-TPR. Figure 9 and Figure S9A plot
the reduction profiles, whereas Table 2 and Table S13 (columns 3-5) compile the peak temperature,
H2 consumption and reducibility of the catalysts based on 0.5 wt%Pd and 2.0 wt%Pd, respectively.
o
Except for 0.5Ni/CeO2-HS showing a complex reduction pattern in the range 175-270 C, the
reduction profiles of the different catalysts consist of a main reduction band below the reaction
temperature (180 oC) that can be ascribed to both NiOx/PdOx and surface/subsurface ceria
o
o
reduction. Two broader bands are also visible at ca. 380 C and 730 C that are reminiscent of
metal-free surface and bulk ceria reduction, respectively (Figure S9B). The surface reducibility of
the different catalysts as measured from the H2 consumption falls into the range 38-45% for the
catalysts containing 0.5wt%Pd (Table 2, column 5) and into the range 42-54% for the catalysts
containing 2wt%Pd (Table S13, column 5).
The reducibility of the catalysts is strongly affected by the Pd and Ni loading and by the metal
impregnation sequence of both metals over the ceria support. Indeed, a remarkable decline of the
o
o
peak temperature from 149 C to 69 C is observed in the absence of Ni when increasing the Pd
21

Catalysis Science & Technology
Page 22 of 44
Submitted to Catal. Sci. Technol.
View Article Online
loading from 0.5 wt% (0.5Pd/CeO2-HS; Table 2, entry 3) to 2.0 wt% (2Pd/CeO2-HS; Table S13, entry
DOI: 10.1039/C8CY01670D
4). Further incorporation of 0.5 wt%Ni to 2Pd/CeO2-HS results in an increase of the peak
temperature to 99 oC for 2Pd-0.5Ni/CeO2-HS, 113 oC for (2Pd-0.5Ni)/CeO2-HS and 122 / 131 oC for
0.5Ni-2Pd/CeO2-HS (Table S13, entries 1-3). Among the different catalysts, 0.5Pd-0.5Ni/CeO2-HS
o
shows the lowest peak temperature with a value as low as 102 C (Table 2, entry 6). The low
reduction peak observed for the latter catalyst can be attributed to its particular architecture (see
chapter 3.7), conditioning the Ni-Pd-CeO2 interaction, as well as to the presence of Pd(II) species.
Overall, these results point out that a low reduction temperature is necessary for achieving high
catalytic activity for OL amination.
3.6.6. Metal dispersion and average particle size after reduction
H2-TPD was further conducted to assess the metal dispersion and average size of the metal
nanoparticles in 0.5Pd-0.5Ni/CeO2-HS and extensively in the other Pd-Ni/CeO2-HS formulations. In
o
these experiments, the catalysts were in situ reduced under 10%H2/Ar at 180 C for 30 min
followed by cooling down under Ar until -80 oC and further H2 desorption under Ar by heating until
o
o
o
400 C. The low-temperature zone in the H2-TPD profiles (from -80 C to 30 C) was used for
characterizing Pd and Ni species over ceria.17 Figure 10 and Figure S10 (blue bands) plot the
profiles obtained together with the corresponding deconvolution, whereas Table S14 lists the
average temperature and standard deviation for the different bands.
The low-temperature H2-TPD profiles on the pre-reduced 0.5Pd-0.5Ni/CeO2-HS, (0.5Pd-
0.5Ni)/CeO2-HS and 0.5Ni-0.5Pd/CeO2-HS are typically characterized by three/four bands
o
o
o
o
appearing in the temperature range from -65 C to -50 C (band I), from -21 C to -1.0 C (band II)
o
o
and from 16 C to 31 C (band III) (Figure 10, blue bands). Three bands are also visible for 2Pd-
0.5Ni/CeO2-HS and (2Pd-0.5Ni)/CeO2-HS at comparable temperatures (Figure S10, blue bands).
The monometallic catalysts display a variable number of bands depending on the type of metal and
o
loading. As a matter of fact, 0.5Ni/CeO2-HS displays two bands centered at -51 C (not shown) and
o
28 C, whereas two and three bands are observed for 0.5Pd/CeO2-HS and 2Pd/CeO2-HS appearing
at 4.0 oC and 14 oC and -49 oC / -2.0 oC / 24 oC, respectively.
22

Page 23 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
The amount of desorbed H2 for bands I-IV was further used for measuDrOinI:g10.1t0h3e9/Cm8CeYt0a16l70D
dispersion and average particle size of the metal nanoparticles on the ceria support. Table 1
(columns 12, 13) and Table S8 (columns 10, 11) list the results obtained for the catalysts based on
0.5 wt%Pd and 2 wt%Pd, respectively. The choice of the different bands in the H2-TPD profiles for
measuring the metal dispersion was confirmed by CO pulse chemisorption, which showed
systematically higher metal dispersions (see also Table 1 and Table S8). In the case of monometallic
catalysts, the average particle size depends to an important extent on the metal type and loading,
evolving from 3.2 nm (DM = 43%) for 0.5Pd/CeO2-HS to 2.8 nm (DM = 48%) for 2Pd/CeO2-HS and
6.5 nm (DM = 18%) for 0.5Ni/CeO2-HS. The average particle size in the bimetallic catalysts differs
depending on the preparation method. As a matter of fact, all bimetallic catalysts based on 2 wt%Pd
show a higher average particle size compared to that of 2Pd/CeO2-HS, that is 5.6 nm (DM = 26%),
3.9 nm (DM = 36%) and 3.1 nm (DM = 44%) for 0.5Ni-2Pd/CeO2-HS, (2Pd-0.5Ni)/CeO2-HS and 2Pd-
0.5Ni/CeO2-HS, respectively. Conversely, the bimetallic catalysts based on 0.5 wt%Pd exhibit much
higher average particle sizes compared to the value measured on 0.5Pd/CeO2-HS, that is 5.6 nm (DM
= 25%) for (0.5Pd-0.5Ni)/CeO2-HS, 6.6 nm (DM = 21%) for 0.5Pd-0.5Ni/CeO2-HS and 8.3 nm (DM =
18%) for 0.5Ni-0.5Pd/CeO2-HS. Interestingly, the average particle size of 0.5Pd-0.5Ni/CeO2-HS
approaches that of 0.5Ni/CeO2-HS.
Using the metal dispersion values measured for the different catalysts, we recalculated the
TONDOA by dividing the total number of moles of DOA formed by the mole number of surface metal
atoms (Ni+Pd)S measured from the metal dispersion (DM) as inferred by H2-TPD (Table 1 column
12). Among the different catalysts, 0.5Pd-0.5Ni/CeO2-HS exhibits the higher surface TONDOA with a
value of 1160 mol DOA / mol (Ni+Pd)S. Conversely, 0.5Ni-0.5Pd/CeO2-HS and (0.5Pd-0.5Ni)/CeO2-
HS display a lower surface TONDOA with a value of to 521 mol DOA / mol (Ni+Pd)S for the latter
catalyst, approaching the surface TONDOA measured for 0.5Pd/CeO2-HS [536 mol DOA / mol
(Ni+Pd)S]. Overall, these results confirm the much higher activity of 0.5Pd-0.5Ni/CeO2-HS
compared to the other bimetallic formulations.
3.6.7. H2 reversibility
23

Catalysis Science & Technology
Page 24 of 44
Submitted to Catal. Sci. Technol.
View Article Online
The reversible H2 storage capacity of the Pd-Ni/CeO2-HS catalysts was also DinOvI:e10s.t10ig39a/tCe8dCYb01y670D
H2-TPD. Figure 10 and Figure S10 (orange bands) plot the profiles for the different catalysts. On the
o
guidance of a previous study,19 the desorbed H2 amount in the range 40-400 C was used as
descriptor of H2 reversibility due to H2 backspillover from ceria to the metal.
In all cases, a broad band is observed with variable intensity that can be deconvoluted into
several bands (i.e. bands V-X, Table S14 columns 7-12). Table 2 and Table S13 (columns 6-8) list the
average temperature for the two most intense bands together with the desorbed H2 amount and the
H2 reversibility. All the catalysts display H2 desorption bands below the reaction temperature (180
oC) corresponding to bands V-VII, appearing in the temperature range 34-73 oC (band V), 69-134 oC
o
(band VI) and 142-199 C (band VII). Besides, an intense band is observed at higher temperature
o
(>250 C) (i.e. band VIII). For all catalysts, the H2 reversibility falls into the range 21-43% of the
total H2 uptake measured by H2-TPR.
More insight into the H2 reversibility can be gained by exploring the low-temperature H2
o
desorption pattern (range 40-140 C) of the different catalysts corresponding to bands V and VI.
Table 2 and Table S13 (columns 9-11) list the average temperature of bands V and VI together with
the corresponding desorbed H2 amount and the H2 reversibility. Noticeably, 0.5Pd-0.5Ni/CeO2-HS
exhibits the highest H2 desorption amount (175 mol/g) and H2 reversibility (15%) followed by
0.5Pd/CeO2-HS and (0.5Pd-0.5Ni)/CeO2-HS with a H2 desorption amount of 160 mol/g and 89
mol/g, respectively, and a H2 reversibility of 12% and 6.8% of the H2 uptake measured by H2-TPR.
These observations underline a facilitated H2 (back)spillover for 0.5Pd-0.5Ni/CeO2-HS, favoring H2
transfer during the reaction.
3.7. Recyclability and post-catalytic characterization of 0.5Pd-0.5Ni/CeO2-HS
The recyclability of 0.5Pd-0.5Ni/CeO2-HS was surveyed in four consecutive catalytic runs.
After each run, the catalyst was separated by decantation and was used in a subsequent run with
neither calcination nor pre-reduction. Figure 11 plots the results obtained for the amination of OL
with NH3 at 180 oC for 2 h. The OL conversion exhibits a moderate decline after the 1st run, whereas
24

Page 25 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
this becomes sharper after the 3rd run. The OA and DOA selectivities keep almost constant values at
View Article Online
DOI: 10.1039/C8CY01670D
13% and 72-82%, respectively, whereas the TOA selectivity increases from 4% after the 1st run to
14% after the 4th run. This drop in the catalytic activity can be ascribed to both a catalyst loss after
each run, as well as to partial deactivation during the reaction. Interestingly, the calcination and/or
reduction of the catalyst after the first run resulted in an almost complete absence of catalytic
activity in the subsequent run.
To ascertain the contribution of both effects, the TONDOA with respect to both the bulk and
surface metal atoms (Ni+Pd) was computed after each run (Figure 11). The TONDOA computed with
respect to the bulk metal atoms shows a decline from an initial value of 131 mol DOA / mol
(Ni+Pd)B after the 1st run to 70 mol DOA / mol (Ni+Pd)B after the 4th run. However, the TONDOA
computed with respect to the surface metal atoms keeps almost unchanged at a value about 600
mol DOA / mol (Ni+Pd)S. These observations suggest that, in addition to catalyst loss, possible
metal sintering on ceria can occur during the reaction. This hypothesis is supported by the
moderate decrease of the metal (Ni+Pd) dispersion on 0.5Pd-0.5Ni/CeO2-HS during the reaction,
evolving from 21% for the fresh catalyst to 11% after the 4th run (Table 1, entries 6 and 7).
However, no metal leaching was observed after each run as confirmed by ICP-OES analysis on the
final solution. Furthermore, within the limits of the experimental error, the Ni and Pd bulk
composition measured by ICP-OES analysis on the spent catalyst keeps unchanged compared to the
fresh catalyst. In contrast, the (Ni/Ce)S surface ratio measured by XPS on the spent catalyst exhibits
a lower value compared to that on the fresh catalyst (91 vs. 446, Table 1 entries 6 and 7), which can
be explained by a partial encapsulation of Ni by ceria during the reaction. For comparison, the
(Pd/Ce)S surface ratio only decreases slightly after reaction compared to the fresh catalyst (28 vs.
37, Table 1 entries 6 and 7), pointing out a moderate Pd encapsulation by ceria, most likely due to a
preferential positioning of Pd on the surface of Ni nanoparticles.
To rationalize further the catalyst evolution during the reaction, a part of the spent 0.5Pd-
0.5Ni/CeO2-HS catalyst after the 4th run was calcined at 400 oC for 2 h under air flow and subjected
to H2-TPR analysis to assess its reducibility compared to the fresh catalyst (Table 2, entries 6 and
25

Catalysis Science & Technology
Page 26 of 44
Submitted to Catal. Sci. Technol.
View Article Online
7). The spent 0.5Pd-0.5Ni/CeO2-HS shows a similar reduction pattern to that of the fresh catalyst as
DOI: 10.1039/C8CY01670D
o
o
inferred from H2-TPR, but presenting a lower reduction peak (77 C vs. 102 C) and a lower
reducibility (600 mol/g vs. 1140 mol/g). Furthermore, the spent 0.5Pd-0.5Ni/CeO2-HS keeps a
partial H2 reversibility as inferred from H2-TPD (Figure S11). As for the fresh catalyst, the spent
0.5Pd-0.5Ni/CeO2-HS shows a H2-TPD profile being characterized by 10 reduction bands, four of
o
which below 30 C referring to the formation of hydrides on the metal surface, and the additional
o
six bands in the temperature range 30-400 C being attributed to H2 (back)spillover on the ceria
support (Table S14, entry 6).
Overall, this body of results indicates that 0.5Pd-0.5Ni/CeO2-HS keeps a high level of
reducibility combined with a partial H2 reversibility along the reaction, keeping its catalytic activity.
These results reinforce the idea stated above that a change of the metal dispersion, even if
moderate, is at the origin, at least partially, of the observed decline of the OL conversion during the
reaction. Notwithstanding this fact, Pd keeps a very high dispersion level, as inferred from XPS
(Figure S12, Table S12 entry 5). Indeed, the different Pd species on the spent catalyst after
calcination exhibit very high BEs, namely 336.3 / 337.1 eV and 337.9 eV (Pd3d5/2 core level) for
Pd2+ and Pd4+, respectively. This observation reinforces the idea pointed out above on the
preferential location of Pd on the surface of Ni nanoparticles, defining Pd-Ni particles enriched with
Pd compared to the fresh catalyst. Scheme 2 illustrates the morphology of 0.5Pd-0.5Ni/CeO2-HS
and its evolution during the reaction.
At last, we investigated coke deposition on 0.5Pd-0.5Ni/CeO2-HS after the 4th run using TG
analysis. The TG profile of the fresh 0.5Pd-0.5Ni/CeO2-HS (Figure S13, dashed curve) shows a single
o
weight loss of about 6% near 100-120 C due to water removal. In contrast, the TG profile of the
o
spent catalyst (Figure S13, straight curve) displays a sharp weight loss region centered at 200 C
o
followed by a monotonous decline in the temperature range 200-400 C. The weight loss of 1%
o
until 200 C can be attributed to physically adsorbed water, whereas the larger weight loss (10%)
in the region 200-400 oC can be attributed to the combustion of amorphous carbon (C) deposits.24
This observation indicates that coke is not formed in large amounts during the amination of OL
26

Page 27 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
o
View Article Online
with NH3 and that most of the coke can be removed from the catalyst after calcination at 400 C.
DOI: 10.1039/C8CY01670D
This observation is also supported by the comparison of the C-region in the XP spectra (range 283-
292 eV) on the fresh and spent catalysts, showing almost no variation. Furthermore, the analysis of
the N-region in the XP spectrum of the spent catalyst (range 390-410 eV) reveals the formation of
neither surface nor bulk metal nitrides during reaction.
4. Conclusions
Along this paper, we reported the preparation of a series of Pd-Ni cataysts supported over
nanoceria with low Pd and Ni loading (<0.5 wt%) for the liquid-phase amination reaction of 1-
octanol with ammonia. The sequence of impregnation of both metals affected to an important
extent the catalytic activity and selectivity. In particular, a formulation based on a first
o
impregnation of Ni followed by Pd afforded a yield higher than 80% to diocylamine at 180 C at
o
optimized conditions (180 C, 2 h, 350 mg catalyst). The catalyst displayed an high turnover
o
number of 1160 mmol DOA / mmol (Ni+Pd)S at 180 C after 2 h using 170 mg of catalyst. This
enhanced activity was ascribed to the genesis of Ni particles in strong contact with Pd after
reduction, as well as to a high reducibility a low temperature (102 oC) and a reversible H2 storage of
o
the catalysts via (back)spillover over nanoceria, especially below 100 C. This specific formulation
discourages disproportionation reactions between dioctylimine and octylamine or octanenitrile on
the catalyst surface. The selectivity towards dioctylamine and the surface turnover number kept
almost unchanged at 80% and 600 mmol DOA / mmol (Ni+Pd)S, respectively, at 180 oC for 2 h using
350 mg of catalyst after 4 consecutive catalytic runs with negligible metal leaching.
Acknowledgements
This work was supported by the French Agency of Research (ANR) through SHAPes program
(contract 13-CDII-0004-06), the Science and Technology Commission of Shanghai Municipality
(14301-412212-16017), and the Minhang District of Shanghai (2016MH177). The authors would
like to express their gratitude to Dr. Eric Leroy from ICMPE-CMTR (UMR 7182 CNRS) for
conducting the STEM-EDS-SDD analyses.
27

Catalysis Science & Technology
Page 28 of 44
Submitted to Catal. Sci. Technol.
View Article Online
DOI: 10.1039/C8CY01670D
References
1. a) H. A. Wittcoff, B. G. Reuben and J. S. Plotkin, Industrial Organic Chemicals, 2nd ed., Wiley, NY,
2004; b) S. A. Lawrence, Amines: Synthesis, Properties and Applications, Cambridge University
Press, 2004.
2. M. Pera-Titus and F. Shi, ChemSusChem. 2014, 7, 1.
3. a) D. R. Corbin, S. Schwarz and G. C. Sonnichsen, Catal. Today 1997, 37, 71; b) K. S. Hayes, Appl.
Catal. A: Gen. 2001, 221, 187.
4. a) G. Rice and E. J. Koh, J. Org. Chem. 1955, 77, 4052; b) A. Mehta, A. Thaker, V. Londhe and S. R.
Nandan, Appl. Catal. A: Gen. 2014, 478, 241.
5. a) J. H. Cho, J-H. Park, T-S. Chang, J-E. Kim and C-H. Shin, Catal. Lett. 2013, 143, 1319; b) K-I.
Shimizu, K. Kon, W. Onodera, H. Yamazaki and J. N. Kondo, ACS Catal. 2013, 3, 112; c) K-I.
Shimizu, K. Kon, W. Onodera, H. Yamazaki and J. N. Kondo, ACS Catal. 2013, 3, 998; d) K-I.
Shimizu, S. Kanno, K. Kon, S. M. A. H. Siddiki, H. Tanaka and Y. Sakata, Catal. Today 2014, 232,
134; e) K-I. Shimizu, Catal. Sci. Technol. 2015, 5, 1412; f) A. Tomer, Z. Yan, A. Ponchel and M.
Pera-Titus, J. Catal. 2017, 356, 133; g) A. Tomer, F. Wyrwalski, C. Przybylski, J.-F. Paul, E.
Monflier, M. Pera-Titus and A. Ponchel, J. Catal. 2017, 356, 111.
6. a) J. V. M. de Pinillos and R. L. Fowlkes, US 4,314,084, 1982; b) J. H. Cho, J. H. Park, T. S. Chang, G.
Seo and C. H. Shin, Appl. Catal. A: Gen. 2012, 417-418, 313.
7. a) A. W. C. Taylor, P. Davies and P. W. Reynolds, US 2,636,902 A, 1953; b) D. M. Gardner and R.
T. Clark, US 4,255,357, 1981; c) G. Sewell, C. O’Connor and E. van Steen, Appl. Catal. A: Gen.
1995, 125, 99; d) G. S. Sewell, C. T. O’Connor and E. van Steen, J. Catal. 1997, 167, 513; e) G. A.
Vedage, L. A. Emig, H-X. Li and J. N. Armor, US 5,917,092, 1998; f) G. A. Vedage, K. S. Hayes, M.
Leeaphon and J. N. Armor, US 5,932,769, 1999; g) A. K. Rausch, E. van Steen and F. Roessner, J.
Catal. 2008, 253, 111.
8. a) A. Baiker and W. Richarz, Ind. Eng. Chem. Prod. Res. Dev. 1977, 16, 261; b) R. C. Lemon, S.
Depot and R. C. Myerly, US 3,022,349, 1982; c) A. Baiker, W. Caprez and W. L. Holstein, Ind. Eng.
Chem. Prod. Res. Dev. 1983, 22, 217.
9. a) A. Baiker and J. Kijenski, Catal. Rev. Sci. Eng. 1985, 27, 653; b) H. Kimura and H. Taniguchi,
Appl. Catal. A: Gen. 2005, 287, 191; c) Y. Li, Q. Li, L. Zhi and M. Zhang, Catal. Lett. 2011, 141,
1635; d) J. Sun, X. Jin, F. W. Zhang, W. Hu, J. Liu and R. Li, Catal. Commun. 2012, 24, 30; e) X. Cui,
Y. Dai, Y. Q. Deng and F. Shi, Chem. Eur. J. 2013, 19, 3665; f) Q. Li, G. Zhang and S. Peng, J. Surf.
Detergents 2002, 5, 229.
10. a) K. Yamaguchi, J. He, T. Oishi and N. Mizuno, Chem. Eur. J. 2010, 16, 7199; b) D. Ruiz, A. Aho, T.
Saloranta, K. Eränen, J. Wärnå, R. Leino and D. Y. Murzin, Chem. Eng. J. 2017, 307, 739; c) T.
Fukushima, M. Jono and M. Terasaka, US 8,247,611 B2, 2012.
11. A. Corma, T. Rodenas and M. J. Sabater, Chem. Eur. J. 2010, 16, 254.
12. Y. Zhang, X. Qi, X. Cui, F. Shi and Y. Q. Deng, Tetrahedron Lett. 2011, 52, 1334.
13. M. Ousmane, G. Perrussel, Z. Yan, J-M. Clacens, F. De Campo and M. Pera-Titus, J. Catal. 2014,
309, 439.
14. a) A. Trovarelli, C. de Leitenburg, M. Boaro and G. Dolcetti, Catal. Today 1999, 50, 353; b) A.
Trovarelli, Catalysis by Ceria and Related Materials, Imperial College Press, London, 2002; c) L.
Vivier and D. Duprez, ChemSusChem. 2010, 3, 654; d) W. Wang, S. Wang, X. Ma and J. Gong,
Chem. Soc. Rev. 2011, 40, 3703; e) Y. Zhang, F. Hou and Y. W. Tan, Chem. Commun. 2012, 48,
2391; f) J. Paier, C. Penschke and J. Sauer, Chem. Rev. 2013, 113, 3949; g) Y. Obora, ACS Catal.
28

Page 29 of 44
Catalysis Science & Technology
Submitted to Catal. Sci. Technol.
View Article Online
2014, 4, 3972; h) T. Montini, M. Melchionna, M. Monai and P. Fornasiero, Chem.DROeI:v1.02.1003196/C,81C1Y0616,70D
5987.
15. M. Tamura and K. Tomishige, Angew. Chem. Int. Ed 2014, 127, 878.
16. A. Corma, J. Navas and M. J. Sabater, Chem. Eur. J. 2012, 18, 14150.
17. Z. Yan, A. Tomer, G. Perrussel, M. Ousmane, B. Katryniok, F. Dumeignil, A. Ponchel, A. Liebens
and M. Pera-Titus, ChemCatChem. 2016, 8, 3347.
18. S. Brunauer, P. H. Emmett and E. Teller, J. Am. Chem. Soc. 1938, 60, 309.
19. a) A. Bensalem, F. Bozon-Verduraz and V. Perrichon, J. Chem. Soc., Faraday Trans. 1995, 91,
2185; b) A. Norman and V. Perrichon, Phys. Chem. Chem. Phys. 2003, 5, 3557; c) J. M. Gatica, R.
T. Baker, P. Fornasiero, S. Bernal and J. Kaspar, J. Phys. Chem. B 2001, 105, 1191.
20. G. N. Vayssilov, M. Mihaylov, P. St. Petkov, K. I. Hadjiivanov and K. M. Neyman, J. Phys. Chem. C
2011, 115, 23435.
21. a) F. Le Normand, L. Hilaire, K. Kili, G. Krill and G. Maire, J. Phys. Chem. 1988, 92, 2561; b) A.
Kotani, T. Jo and J. C. Parlebas, Adv. Phys. 1988, 37, 37; b) A. Laachir, V. Perrichon, A. Badri, J.
Lamotte, E. Catherine, J. C. Lavalley, J. El Fallah, L. Hilaire, F. le Normand, E. Quemere, G N.
Sauvion and O. Touret, J. Chem. Soc., Faraday Trans. 1991, 87, 1601.
22. A. Bensalem, F. Bozon-Verduraz, M. Delamar and G. Bugli, Appl. Catal. A: Gen. 1995, 121, 81.
23. a) S. A. Krasnikov, A. B. Preobrajenski, T. Chasse and R. Szargan, Thin Solid Films 2003, 428,
201; b) D. Srinivas, C. V. V. Satyanarayana, H. S. Potdar and P. Ratnasamy, Appl. Catal. A: Gen.
2003, 246, 323.
24. H. Guo, H. Lou and X. Zheng, Carbon 2007, 45, 1314.
29