Selective Hydrogenation of Phenol Catalyzed by Palladium on High-Surface-Area Ceria at Room Temperature and Ambient Pressure

Article abstract of DOI:10.1021/cs502000j

(Chemical Presented). Palladium supported on high-surface-area ceria effectively catalyzes the hydrogenation of phenol to cyclohexanone at atmospheric pressure and room temperature. Activation of H₂ at Pd sites and phenol at surface ceria sites was investigated by probing the redox properties of the catalyst and studying the mechanism of phenol adsorption. Temperature-programmed reduction and pulsed chemisorption were used to examine the effects of prereduction temperature on catalyst dispersion and reducibility. A sharp effect of prereduction temperature on catalytic activity was observed. This dependence is rationalized as a result of interactions between palladium and ceria, which under reducing conditions enhance palladium dispersion and create different types of environments around the Pd active sites and of encapsulation of the catalyst caused by support sintering at high temperatures. Temperature-programmed diffuse reflectance infrared Fourier transform spectroscopy revealed that phenol undergoes dissociative adsorption on ceria to yield cerium-bound phenoxy and water. Reduction of the chemisorbed phenoxy species decreases the number of proton-accepting sites on the surface of ceria and prevents further dissociative adsorption. Subsequent phenol binding proceeds through physisorption, which is a less active binding mode for reduction by hydrogen. High activity can be restored upon regeneration of proton acceptor sites via reoxidation/reduction of the catalyst.

Full text of DOI:10.1021/cs502000j

Subscriber access provided by MICHIGAN STATE UNIVERSITY | MSU LIBRARIES
Article
Selective Hydrogenation of Phenol Catalyzed by Palladium on
High Surface Ceria at Room Temperature and Ambient Pressure
Nicholas C Nelson, Juan Sebastián Manzano, Aaron D. Sadow, Steven H. Overbury, and Igor I. Slowing
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/cs502000j • Publication Date (Web): 18 Feb 2015
Downloaded from http://pubs.acs.org on February 24, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
Selective Hydrogenation of Phenol Catalyzed by
Palladium on High Surface Ceria at Room
Temperature and Ambient Pressure
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nicholas C. Nelson,1,2 J. Sebastián Manzano, 1,2 Aaron D. Sadow, 1,2 Steven H. Overbury3 and
Igor I. Slowing1,2,*
1U.S. D.O.E. Ames Laboratory, 2Department of Chemistry, Iowa State University, 3Oak Ridge
National Laboratory
Corresponding author: *Eꢀmail islowing@iastate.edu (I.I.S.)
Abstract
Palladium supported on high surface ceria effectively catalyzes the hydrogenation of phenol to
cyclohexanone at atmospheric pressure and room temperature. Activation of H2 at Pd sites and
phenol at surface ceria sites were investigated by probing the redox properties of the catalyst and
studying the mechanism of phenol adsorption. Temperatureꢀprogrammed reduction and pulsed
chemisorption were used to examine the effects of preꢀreduction temperature on catalyst
dispersion and reducibility. A sharp effect of preꢀreduction temperature on catalytic activity was
observed. This dependence is rationalized as a result of interactions between palladium and ceria,
which under reducing conditions enhance palladium dispersion and create different types of
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 38
1
2
3
4
5
6
7
8
9
environments around the Pd active sites, and of encapsulation of the catalyst caused by support
sintering at high temperatures. Temperature programmed diffuse reflectance infrared Fourier
transform spectroscopy revealed that phenol undergoes dissociative adsorption on ceria to yield
ceriumꢀbound phenoxy and water. Reduction of the chemisorbed phenoxy species decreases the
number of protonꢀaccepting sites on the surface of ceria and prevents further dissociative
adsorption. Subsequent phenol binding proceeds through physisorption, which is a less active
binding mode for reduction by hydrogen. High activity can be restored upon regeneration of
proton acceptor sites via reoxidation/reduction of the catalyst.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: Mesoporous ceria, phenol hydrogenation, metalꢀsupport interactions, dissociative
adsorption, redoxꢀactive support
Introduction
Cyclohexanone is a precursor to caprolactam and adipic acid, which are key intermediates for
the synthesis of nylons and polyamide resins. Cyclohexanone is commercially produced by
oxidation of cyclohexane or hydrogenation of phenol.1 The former route involves high reaction
temperatures (140ꢀ180 °C), pressures (8ꢀ20 bar), and low conversion (< 10 %).1,2 Phenol
hydrogenation is much more efficient and is catalyzed by palladium on carbon (Pd/C) in the
liquid phase and catalyzed by palladium on alumina (Pd/Al2O3) in the vapor phase.3 Of the two
hydrogenation approaches, liquid phase reduction is more selective to the ketone, but operates at
high temperatures (175 °C) and under excess hydrogen pressures (13 bar).1,3 Given the
importance of this process, significant efforts have been devoted to the development of
alternative catalysts that can further improve yields and enable conversion under milder
ACS Paragon Plus Environment
2

Page 3 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
conditions. Several research groups have now achieved remarkable results with conversion and
selectivity over 90 % at either room temperature or 1 bar H2 pressure.4ꢀ16 The most efficient
processes take advantage of the fact that aromatics can be reduced under mild conditions when
the activation of H2 by heterogeneous Pd catalyst is combined with the electrophilic activation of
aromatics by soluble Lewis acids.17 Liu et al. demonstrated the application of this strategy for the
hydrogenation of phenol by combining Pd/Al2O3 with AlCl3.7 Lee and coꢀworkers significantly
improved the process by combining Sc(OTf)3 with Pd/C, achieving full conversion at 20 °C and
1 bar H2.8 The Lewis acid also favors the selectivity to the cyclohexanone by forming an adduct
that is more difficult to reduce to the alcohol. Li, Luque and collaborators achieved similar
results with a fully heterogeneous system using Pd supported on the chromiumꢀbased MOF MILꢀ
101, with the chromium centers in the MOF support proposed as Lewis acidic activators for the
reaction.11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The strategy of using the support as an active component of the reaction rather than an inert,
inactive component merely for catalyst dispersion is particularly attractive from the perspective
of advanced catalyst design. In this regard, redox active supports are especially appealing as they
have the potential of participating in electron transfer processes during catalysis. Cerium dioxide
(ceria) is an interesting support for this purpose because of its reducibility and its interactions
with noble metals.18 These interactions lead to high dispersion of catalytic metals and provide
beneficial electronic effects.19 Ceria and ceriaꢀbased materials have been extensively studied as
structural and electronic promoters to improve activity and selectivity of heterogeneous
catalysts.20 Ceria is the most industrially significant rare earth oxide catalyst mainly due to its use
in threeꢀway catalytic converters (TWC) and fluid catalytic cracking (FCC).19,21 Recently, ceriaꢀ
based materials have been investigated for use in soot removal from diesel engine exhausts,22
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 38
1
2
volatile organic compound (VOC) degradation,23 fuel cell technology,24 waterꢀgas shift
reaction,25,26 preferential CO oxidation (PROX),27 oxidative dehydrogenation,28 and selective
hydrocarbon oxidations.29,30 The success of ceria and ceriaꢀbased materials in catalysis is
oftentimes due to facile Ce4+/Ce3+ redox cycling without disruption of the fluorite lattice
structure.20 Furthermore, the redox properties of ceriaꢀbased materials can be tuned by
incorporation of dopants or deposition of metals, which offer significant opportunities for
modifying their activity and improving their performance.31
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Only a few studies exist for phenol hydrogenation on ceriaꢀbased materials. The groups of
Inagaki and Scire studied the reaction in the vapor phase over palladium supported on high
surface ceria (Pd/CeO2) at 180 °C (80 % conversion, 50 % selectivity) and 160 °C (40 %
conversion, 95 % selectivity) respectively.32,33 The liquid phase process was investigated by Li et
al. using Ceꢀdoped Pd nanospheres (0.43 mol % Ce) with a hollow chamber, obtaining 82 %
cyclohexanone yield under 10 bar H2 at 80 °C.34 In another paper, Li and coꢀworkers used
palladiumꢀceriumꢀboron supported on hydrotalcite to achieve 82 % conversion and 80 %
selectivity within 4 h at 10 bar H2 and 100 °C.35 Using ceria as the support rather than as a minor
component of the catalyst may be a good way to take advantage of its redox properties and its
electronic effects on the metal in the liquid phase process. Furthermore, using high surface area
ceria as a support should result in higher activity than a low surface area counterpart, because
this could lead to increased dispersion of supported metal and larger amounts of reactive species
on the surface.36 Herein, we report the synthesis of palladium supported on high surface ceria
(Pd/CeO2) and its exceptional performance for the selective hydrogenation of phenol to
cyclohexanone in the liquid phase at low temperature and H2 pressure.
ACS Paragon Plus Environment
4

Page 5 of 38
ACS Catalysis
1
2
3
4
Experimental
5
6
7
8
Reagents. Cerium(III) nitrate hexahydrate (Ce(NO3)3•6H2O), palladium(II) acetate
(Pd(O2CCH3)2), phenol (C6H5OH), tetramethylorthosilicate (TMOS), aluminum(III)
isopropoxide (Al(OiPr)3), cerium (IV) oxide (CeO2), 10 wt. % palladium supported on carbon
(Pd/C), and concentrated nitric acid were purchased from Sigma Aldrich. Pluronic P104 and
Pluronic P123 were obtained from BASF. Ce(NO3)3•6H2O was dried under vacuum at room
temperature prior to synthesis. 100 % ethyl alcohol was used for all syntheses. All other
chemicals were used without further purification.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of high surface ceria (CeO2). In a typical synthesis, Ce(NO3)3•6H2O (8.80 g, 20.3
mmol) and nonionic block coꢀpolymer surfactant Pluronic P104 (10.1 g, 1.71 mmol) were
dissolved in ethanol (200 mL) under vigorous stirring (700 rpm) for 3 h. Once thoroughly
homogenized, the solution was cast into a large crystallization dish (diameter 185 mm) and
placed into a preꢀheated 65 °C oven to undergo solvent evaporation. After 12ꢀ24 h, the gel was
placed into a preꢀheated 150 °C oven for an additional 12 h. To remove the remaining P104
surfactant, the yellow powder was calcined in air at 450 °C for 4 h with a ramp rate of 1 °C minꢀ
1. Optimal surface areas were obtained when relative humidity was kept below 60 %.
CAUTION: During thermal treatment at 150 °C combustion occurs within ~12 minutes
producing flames which selfꢀextinguish within seconds after all combustible material (i.e. block
coꢀpolymer) is burned. The thermal treatment step should be conducted in an oven, preferably in
a fume hood to avoid exposure to gaseous decomposition products.
Synthesis of mesoporous silica (SiO2). This material was prepared following our previously
published method.37 Pluronic P104 (7.00 g, 1.19 mmol) was dissolved in aqueous HCl (273.0 g,
1.6 M). After stirring for 1 h at 56 °C, tetramethylorthosilicate (10.6 g, 69.9 mmol) was added
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 38
1
2
3
4
5
6
7
8
9
and stirred (500 rpm) for additional 24 h. The resulting mixture was hydrothermally treated for
24 h at 150 °C in a highꢀpressure reactor. Upon cooling to room temperature, the white solid was
collected by filtration, washed with copious amounts of methanol, and dried in air. The MSN
material was calcined in air at 550 °C for 6 h with a ramp rate of 1.5 °C minꢀ1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of mesoporous alumina (Al2O3).The synthesis of alumina was adapted from Yan and
coꢀworkers.38 Briefly, nonꢀionic block coꢀpolymer surfactant Pluronic P123 (0.92 g, 0.16 mmol)
was dissolved in ethanol (20 mL) at room temperature. After stirring for 0.5 h, concentrated
nitric acid (1.5 mL, 15.8 M) and Al(OiPr)3 (2.04 g, 10.0 mmol) were added into the above
solution with vigorous stirring (800 rpm). The mixture was capped, stirred at room temperature
for 5 h, and then cast into a large crystallization dish. The dish was placed into a preꢀheated 60
°C oven to undergo solvent evaporation for 48 h. The lightꢀyellow solid was calcined in air at
400 °C for 6 h with a ramp rate of 1.5 °C minꢀ1.
Synthesis of Pd/MxOy. All Pd catalysts were prepared by an impregnation method with a 1 wt.
% Pd loading relative to the mass of the support. In a typical synthesis, Pd(O2CCH3)2 (0.0419 g,
0.187 mmol) was dissolved in acetone (1 mL). The support (2.00 g) was placed into a mortar and
impregnated with the Pd solution in 0.20 mL increments. After each impregnation step, the
catalyst was mixed thoroughly with a pestle until seemingly dry. The material was calcined at
350 °C for 2 h with a 2.5 °C minꢀ1 ramp rate and then reduced under flowing hydrogen at 350 °C
for 2 h with a ramp rate of 2.5 °C minꢀ1. CAUTION: Special precaution should be taken when
preparing (i.e. reducing) Pd/Al2O3 as pyrophoric aluminum hydrides may form. The
physicochemical properties of nonꢀceriaꢀbased catalysts are reported in Table S1.
Hydrogenation reactions. All reactions were conducted in batch mode under flowing hydrogen
(~1 bar) using a Schlenk line. In a typical experiment, the catalyst (48 mg, 1.1 wt. % Pd loading)
ACS Paragon Plus Environment
6

Page 7 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
and phenol (or phenol derivative) solution (4 mL, 0.025 M) were added to a 10 mL round bottom
flask equipped with condenser and stir bar. The palladium to substrate ratio was maintained at 5
mole % for all reactions. The flask was placed onto a Schlenk line under a headspace flow of
hydrogen and allowed to stir (800 rpm) for 4 h. The reaction product was collected by
centrifugation. A 50 μL aliquot of the supernatant was added to 1 mL of ethanol and analyzed in
an Agilent GCꢀMS (7890A, 5975C) with a HPꢀ5MS column. The run started at 60 °C for 0
minutes, then ramped to 150 °C at 5 °C minꢀ1 for 0 minutes, then ramped to 300 °C at 20 °C minꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
for 3 minutes. Resorcinol was used as an internal standard. Conversion was defined as mol %
and calculated as moles of converted phenol per mole of starting phenol times 100 %. Selectivity
was defined as mol % and calculated as moles of cyclohexanone per moles of products times 100
%. Yields were defined as mole percent and were calculated as moles of each product multiplied
by conversion times 100 %.
TPR and chemisorption. Hydrogen temperature programmed reduction (H2ꢀTPR) and
hydrogen chemisorption were performed in a Micromeritics AutoChem II. H2 in Ar (H2/Ar)
(10.13 %) was used as the reducing agent or metal dispersion probe. H2ꢀTPR experiments were
performed with a flow rate of 50 mL minꢀ1 and a ramp rate of 10 °C minꢀ1. A cold trap was used
to collect water produced during the reduction. H2ꢀchemisorption analysis was carried out by
reducing samples at specified temperatures, then heating/cooling to 250 °C under H2/Ar,
followed by flowing Ar for 15 minutes at 250 °C to remove surface bound hydrogen from Pd
crystallites. The sample was then cooled under Ar to ꢀ35 °C for hydrogen pulse chemisorption
measurements. The palladium dispersion of the catalysts was calculated based on the following
equation:
S ×M ×V
f
D(%) =
ad ×100
m×W ×V
m
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 38
1
2
3
4
5
6
7
8
9
where Sf = stoichiometry factor (the Pd/H2 molar ratio) = 2; M = the atomic mass of Pd (106.42
g molꢀ1); Vad = the volume of chemisorbed H2 at standard temperature and pressure conditions
(mL); m = the mass of the sample (g); W = the weight fraction of Pd in the sample as determined
by ICPꢀOES (0.011 g Pd per g sample); Vm = the molar volume of H2 (22414 mL molꢀ1) at STP.
The specific surface area of palladium was calculated based on the following equation:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Sf
× N
A
× SAcross ×Vad
Pd Surface Area (m2 gꢀ1 of Pd) =
m×W ×V
m
where Sf = stoichiometry factor (the Pd/H2 molar ratio) = 2; NA = Avogadro’s number (6.023 ×
1023 molꢀ1); SAcross = palladium crossꢀsectional area (7.87 × 10ꢀ20 m2); Vad = the volume of
chemisorbed H2 at standard temperature and pressure conditions (mL); m = the mass of the
sample (g); W = the weight fraction of Pd in the sample as determined by ICPꢀOES (0.011 g Pd
per g sample); Vm = the molar volume of H2 (22414 mL molꢀ1) at STP.
The cubic crystallite size of palladium was calculated based on the following equation:
6× M
Pd Crystallite Size (nm) =
m×W ×dPd × N × SAPd
A
Where M = the atomic mass of Pd (106.42 g molꢀ1); m = the mass of the sample (g); W = the
weight fraction of Pd in the sample as determined by ICPꢀOES (0.011 g Pd per g sample); dPd =
density of palladium (1.202 × 10ꢀ20 g nmꢀ3); NA = Avogadro’s number (6.023 × 1023 molꢀ1); SAPd
= palladium surface area from equation above (nm2 gꢀ1 of Pd). The number six is derived from
assuming cubic geometry.
Powder Xꢀray diffraction (PXRD). Diffraction patterns were collected using Co Kα1, Kα2 split
radiation (45 kV, 40 mA, λavg = 1.7903 Å) on a PANalytical X’Pert PRO diffractometer
equipped with a theta−theta vertical mode goniometer, incident Fe filter, an airꢀcooled
X’Celerator real time multiple strip (RTMS) detector, and spinner stage. The spectra were
ACS Paragon Plus Environment
8

Page 9 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
converted to Cu Kα radiation for comparison to standard patterns. Powder XRD samples were
prepared by placing powders onto a backgroundꢀless polycarbonate sample holder. Crystallite
sizes were calculated using Scherrer equation:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
K
λ
D
(hkl)
=
β
cosθ
where K is the shape factor (0.9) of the average crystallite, λ is the Xꢀray wavelength (0.17903
nm), β is the full width at half maximum (radians), and ϴ is the Bragg angle (radians).
Electron microscopy/Energy Dispersive Xꢀray Spectroscopy. Transmission electron
microscopy (TEM) was conducted using a FEI Technai G2 F20 field emission microscope and
scanning transmission electron microscope (STEM) operating at 200 kV (pointꢀtoꢀpoint
resolution <0.25 nm and a lineꢀtoꢀline resolution of <0.10 nm). TEM samples were prepared by
placing 2−3 drops of dilute ethanol suspensions onto lacey carbonꢀcoated copper grids. The
composition of Pd/CeO2 structures were characterized by elemental mapping and energy
dispersive Xꢀray spectroscopy (EDS) scans in STEM mode.
Surface area and porosimetry. Textural properties of the supports and catalysts were measured
by nitrogen sorption isotherms at ꢀ196 °C in a Micromeritics Tristar analyzer. The surface areas
were calculated by the BrunauerꢀEmmettꢀTeller (BET) method, and the pore size distribution
was calculated by the BarrettꢀJoynerꢀHalenda (BJH) method. Pretreatment of samples for surface
area measurement was done by flowing N2 for 6 h at 100 °C.
ICPꢀOES. Pd loadings were analyzed by a Perkin Elmer Optima 2100 DV Inductively Coupled
PlasmaꢀOptical Emission Spectroscope (ICPꢀOES). Samples (5 mg) were digested for 24 h in
aqueous HF and HCl solution (0.18 v/v % and 5 v/v % respectively).
Diffuse reflectance infrared Fourier transform (DRIFT). Measurements were made on a
Bruker Vertex 80 FTꢀIR spectrometer with OPUS software and apodized spectral resolution of
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 38
1
2
0.2 cmꢀ1. The spectrometer was equipped with a HeNe laser and photovoltaic MCT detector. A
Praying MantisTM diffuse reflectance accessory and high temperature reaction chamber was used
for room temperature and variable temperature measurements, respectively.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diffuse reflectance UV/Visible absorption spectroscopy (DR UV/Vis). Measurements were
made on an Ocean Optics USB2000+ fiber optic spectrometer (bandwidth = 350ꢀ1100 nm)
operating in absorption mode. Samples were prepared by suspending 50 mg of ceria in 0.5 mL of
phenol dissolved in test solvents (33 mg mLꢀ1) overnight. The solid was collected by
centrifugation and analyzed.
Nuclear Magnetic Resonance. Proton nuclear magnetic resonance (1HꢀNMR) spectra were
collected on a Varian VXRꢀ300 equipped with a narrow bore 7.05 tesla/300 MHz magnet and a
1
standard H probe. Deuterated benzene and cyclohexane were used as solvent and internal
standard, respectively.
XꢀRay Photoelectron Spectroscopy (XPS). XPS analysis was done with a PHI 5500 multiꢀ
technique system using a standard Al xꢀray source. The Pd/CeO2 catalyst exposed to air was
prepared by deposition onto doubleꢀsided tape sample holder. The Pd/CeO2 catalyst exposed to
hydrogen was prepared in a glove box and then transferred to the XPS chamber in an airꢀfree
sample cell. Charge correction was accomplished by shifting the O1s peak to 529.0 eV for all
spectra.39
ACS Paragon Plus Environment
10

Page 11 of 38
ACS Catalysis
1
2
3
4
Results and Discussion
5
6
Support Synthesis and Characterization
7
8
9
The ceria support was synthesized by an evaporationꢀinduced selfꢀassembly method (EISA)
commonly used for the synthesis of high surface metal oxides.40ꢀ43 The physical properties of
CeO2 support and Pd/CeO2 catalyst are summarized in Table 1. The PXRD pattern measured was
indexed to the fluorite cubic structure of ceria (space group Fm3m (225), JCPDS 34ꢀ0394)
(Figure 1a). The broad peaks indicated small ceria crystallites, and estimation of the average
crystallite size using the Scherrer equation provided a value of 5.9 nm. Nitrogen sorption
analysis gave a high specific surface area and displayed a hysteresis loop at high partial pressures
suggesting significant textural porosity (Figure 1b). TEM imaging revealed that the material is
comprised of aggregated particles with small voids within each aggregate that likely to account
for the measured pore volume (Figure 1c). HRꢀTEM (Figure 1d) examination indicated the
presence of the most thermodynamically stable (111) planes of the cubic lattice.31 Given the
polycrystalline nature of the material, (200) and (220) planes were also observed, although far
less prevalent than the (111) planes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Table 1. Physical properties of CeO2 support and Pd/CeO2 catalyst.
BET Surface
BJH Pore
CeO2 Crystallite Pd Dispersion Pd Crystallite
Sample
Area (m2 gꢀ
Volume (cm3 gꢀ1) Size (nm)a
(%)b
Size (nm)b
1
47
48
49
50
51
52
53
54
55
56
57
58
59
60
)
CeO2
241
159
0.31
0.27
5.9
7.1
ꢀꢀꢀ
64
ꢀꢀꢀ
Pd/CeO2
1.5 nm
a Obtained by PXRD analysis. b Obtained by H2 chemisorption.
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. a) Wideꢀangle PXRD pattern of the synthesized CeO2 (black) and reference
cerium(IV) oxide (red); b) N2 sorption isotherms of CeO2; c) TEM and d) HRꢀTEM images of
CeO2 showing a dꢀspacing of 0.31 nm corresponding to the (111) planes of CeO2.
Catalyst Synthesis and Characterization
The catalyst was prepared via incipient wetness impregnation. An acetone solution of
palladium (II) acetate was initially added to the ceria material, the mixture was then calcined in
air to produce PdOxHy/CeO2 precatalyst, which was finally reduced by flowing hydrogen at 350
°C to give the Pd/CeO2 catalyst. The PXRD pattern of Pd/CeO2 is nearly identical to CeO2
(Figure 2a). The ceria reflections of the catalyst are slightly narrower than the original support,
suggesting sintering occurred during the additional thermal aging, which is common for ceria
under reducing environments.44 This sintering likely led to the lowered specific surface area and
pore volume of Pd/CeO2 relative to CeO2 (Table 1). Palladium species could not be detected by
XRD analysis, suggesting it was highly dispersed over the support,45 and/or incorporated into the
ACS Paragon Plus Environment
12

Page 13 of 38
ACS Catalysis
1
2
ceria framework.46 The rationalization of highly dispersed palladium is consistent with the low
loading of palladium (1.1 wt. % as determined by ICPꢀOES) and its high affinity for ceria.47 In
support of this idea, subꢀambient pressure H2 chemisorption measurements gave a high Pd
dispersion of 64 % and a calculated cubic crystallite size of 1.5 nm. The catalyst was further
investigated by STEMꢀHAADF imaging (Figure 2b). Palladium particles could not be directly
observed possibly due to high dispersion, small crystallite size, and/or low contrast between Pd
and Ce. However, Energy dispersive Xꢀray spectroscopy (EDS) mapping confirmed the presence
of Pd on the support (Figures 2c, S2). XPS analysis was also used to evaluate the oxidation state
of palladium in the catalyst. The binding energy of Pd 3d5/2 electrons in the catalyst exposed to
ambient air (336.5 eV) is close to that in PdO (336.9 eV) suggesting part of the metal, likely the
atoms at the surface, is easily oxidized in air (Figure 2d, black trace).39 Exposure of the catalyst
to H2 at room temperature (a more accurate description of the catalyst under catalytic conditions)
shifts the binding energy to lower values (335.2 eV) suggesting significant conversion to the
metallic form (Figure 2d, red trace).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. a) Wideꢀangle PXRD patterns of CeO2 support (black) and Pd/CeO2 catalyst (red); b)
STEMꢀHAADF image of Pd/CeO2; c) EDS map of Pd (red image) on a 25 x 25 square section of
the support (Region M), the square labeled ‘Ref’ was taken as a reference for drift correction
between scans; d) xꢀray photoelectron spectra of Pd/CeO2 exposed to air (black) and to hydrogen
atmosphere (red).
H2ꢀTPR experiments were run on CeO2 and the precatalyst (PdOxHy/CeO2) to study the effect
of Pd impregnation on the redox properties of the material (Figure 3). The results are
summarized in Table 2. Upon hydrogen treatment of CeO2, two peaks were observed
corresponding to reduction of surface and bulk ceria at 473 °C (T2) and 791 °C (T3),
respectively.48 Overall, the amount of H2 consumed in the pure CeO2 sample was less than the
total amount consumed by the PdOxHy/CeO2 precatalyst. There was considerable hydrogen
consumption by the PdOxHy/CeO2 precatalyst at 95 °C (T1) which can be attributed, in part, to
the reduction of PdOxHy species.49 Interestingly, the PdOxHy/CeO2 precatalyst consumed less H2
ACS Paragon Plus Environment
14

Page 15 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
at the higher temperatures associated with surface and bulk ceria reduction than the Pdꢀfree CeO2
support. Thus, the presence of Pd facilitated the reduction of surface and bulk CeO2 as indicated
by the lower reduction temperatures compared to the support alone (332 °C for T2 and 703 °C for
T3). The total amount of hydrogen consumed by PdOxHy/CeO2 was 0.40 mmol H2 gꢀ1 greater
than the amount consumed by the CeO2 support (Table 2). Assuming that Pd is in the +2
oxidation state, this is 4 times greater than the theoretical amount of hydrogen that ought to be
consumed (0.10 mmol H2 gꢀ1) to reduce PdO to Pd based off ICP results. If all Pd atoms formed
hydrides this would imply consumption of additional 0.05 mmol H2 gꢀ1, which would leave still
0.25 mmol H2 gꢀ1. This difference suggests hydrogen spilled over onto the support during the
reduction of PdOxHy, which could have been either physisorbed or led to the reduction of ceria.50
A significant decrease in the amount of hydrogen consumed by the PdOxHy/CeO2 catalyst at high
temperatures (Table 2) relative to the CeO2 support suggested that some hydrogen spilled over
leading to the reduction of ceria. These results are consistent with previous hydrogen adsorption
studies for palladiumꢀceria catalysts.50 Hydrogen spillover is common in reducible oxides
serving as supports for noble metals.51
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Reducibility of CeO2 support and PdOxHy/CeO2 precatalyst.
H2 Consumed at
T1 (mmol gꢀ1)
H2 Consumed at
T2 (mmol gꢀ1)
H2 Consumed at
T3 (mmol gꢀ1)
Total H2 Consumed
(mmol gꢀ1)
Sample
CeO2
ꢀꢀꢀ
0.69
0.63
1.1
1.79
2.19
PdOxHy/CeO
0.98
0.58
2
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. H2ꢀTPR profile of CeO2 (black) and PdOxHy/CeO2 (red).
Figure 4. H2ꢀTPR profiles of PdOxHy/CeO2 precatalyst reduced at 150 °C, 250 °C, 350 °C, and
450 °C. To measure the TPR each sample was previously reꢀoxidized at 350 °C. TPR of the
precatalyst is included at the bottom as a reference. Temperatures of selected peaks are indicated
in red at the top.
ACS Paragon Plus Environment
16

Page 17 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
The TPR profile of the PdOxHy/CeO2 precatalyst (Figure 3) suggested that complete reduction
of Pd is achieved at temperatures above 125 °C. The effects of PdOxHy/CeO2 reduction
temperature on the redox properties of the catalyst were further evaluated on the basis of
literature reports of the significant structural changes observed for noble metals supported on
ceria under reducing conditions.47 To this end, four Pd/CeO2 samples were produced by
reduction of PdOxHy/CeO2 precatalyst at 150 °C, 250 °C, 350 °C and 450 °C respectively. Each
sample was then reꢀoxidized (350 °C under O2/He mixture for 1 h, 10 °C minꢀ1) to examine their
reducibility as a function of pretreatment temperature. The TPR analysis revealed significant
effects of the preꢀreduction temperature on the H2 uptake of the catalyst (Figure 4).
PdOxHy/CeO2 preꢀreduced at 150 °C consumed hydrogen near room temperature (28 °C), but
there was also considerable hydrogen uptake at the same temperature as the original precatalyst
(95 °C). Pretreatment of PdOxHy/CeO2 under H2 at 250 °C or 350 °C resulted in only one peak
centered at 28 °C and 14 °C, respectively. PdOxHy/CeO2 preꢀreduced at 450 °C showed three
hydrogen consumption peaks at 8 °C, 25 °C and 51 °C. These changes could be due to
interactions between Pd and CeO2 and structural transformations undergone by CeO2 as
reduction temperature increases, which have been well documented for noble metals supported
on ceria under reducing conditions.47 Thus, increasing reduction temperatures may lead to
incremental spreading of Pd over CeO2 to give smaller particles that are reactive at lower
temperatures, as indicated by the shift in the TPR peak from 95 °C to 28 °C and 14 °C. While
reduction of the precatalyst at 450 °C led to further shifting of signal to 8 °C, additional peaks
appeared at 25 °C and 51 °C. These higher temperature H2 uptake peaks can be attributed to the
metal existing in different types of environments that involve additional interactions with CeO2,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 38
1
2
such as decoration phenomena (i.e. burial of surface palladium by sintered ceria),47,52 alloying of
PdꢀCe,53 and/or electronicꢀsupport effects.31
3
4
5
6
7
8
9
Dispersion measurements were taken on the catalysts after the initial reduction in order to
further evaluate the changes in Pd surface with reduction temperature (Figure S1). Measurements
were performed at ꢀ35 °C to eliminate erroneous results from hydrogen spillover.31,54 The results
revealed that, indeed, Pd dispersion increased as reduction temperature increased from 150 °C to
350 °C. Sanchez and Gazquez52 proposed a model to explain the changes in dispersion of metals
in oxide supports upon reduction at high temperatures. In their model, they proposed the
occupancy of oxygen ion vacancies by metal atoms; as the reduction temperature increases, more
vacancies are formed which lead to migration of Pd crystallites into these high energy
environments resulting in increased dispersion. Part of the hydrogen consumed by PdOxHy/CeO2
preꢀreduced at 150 °C was observed at the same temperature (95 °C) as the original precatalyst
(Figure 4). This suggests that some Pd crystallites remained in the same environment as they
were upon deposition of Pd(O2CCH3)2 precursor and calcination in air, while others relocated to
a different environment. As the preꢀreduction temperature was increased to 250 °C, only one
peak (28 °C) was observed, which coincided with part of the hydrogen uptake temperature for
the catalyst preꢀreduced at 150 °C (Figure 4). These data are consistent with the Pd species
spreading over different sites on the surface of CeO2. Recent studies have shown that surface
oxygen vacancies promote late transition metal dispersion over ceria supports.55,56 Preꢀreduction
at 350 °C resulted in one peak at 14 °C (Figure 4). Interestingly, this preꢀreduction temperature
corresponded to the temperature for reduction of surface ceria (i.e. vacancy formation) in the
PdOxHy/CeO2 precatalyst (Figure 3 (red)). Preꢀreduction of PdOxHy/CeO2 at 450 °C led to a
dramatically lower dispersion than the catalysts preꢀreduced at lower temperatures. The sharp
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
18

Page 19 of 38
ACS Catalysis
1
2
decrease in metal surface area (from 285 m2gꢀ1 to 34 m2gꢀ1) supports the idea that reduction of
the precatalyst at 450 °C led to burial of surface Pd into CeO2 as previously suggested by the
TPR measurements; this catalyst encapsulation is likely the result of support sintering (Figures 4,
S1).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Catalysis
Hydrogenation reactions were carried out in batch mode by flowing hydrogen (~ 1 bar) in the
headspace of a hexane solution of phenol, with a Pd to phenol mole ratio of 0.05. The results are
summarized in Table 3. There was no phenol conversion when CeO2 or PdOxHy/CeO2 was used
as the catalyst, confirming that Pd was crucial for hydrogen activation (Entry 1 and 2). The
Pd/CeO2 catalyst hydrogenates 85.5 % of the phenol with a ketone selectivity of 96.1 % at room
temperature and 1 bar hydrogen in only 4 h (Entry 3). These results show that Pd/CeO2 is
competitive with the best performing catalysts reported for the liquidꢀsolid interfacial
hydrogenation of phenol.4 Catalysts prepared in the same way using other commonly employed
supports (e.g. alumina, silica and carbon) gave significantly lower conversion under the same
reaction conditions. This could be due to the weak interactions between the supports and Pd,
which lead to low catalyst dispersion (Table S1) or due to the inability to shuttle hydrogen from
Pd to nonꢀreducible supports (i.e. spillover),51 which limits their influence on catalytic turnover.
Furthermore, Pd supported on commercial ceria (surface area 55 m2 gꢀ1) was less active than
Pd/CeO2, showing the clear advantages of using high surface area ceria support to promote the
selective hydrogenation of phenol.
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 38
1
2
3
4
5
6
Table 3. Reaction conditions and results for the hydrogenation of phenol in hexane.
7
8
9
Selectivity (%)
C=O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Entry Catalyst
T (°C)
t (h)
Conversion (%)
OH
ꢀꢀꢀ
1
CeO2
25
25
25
35
25
25
25
4
4
4
4
4
4
4
4
0.0
ꢀꢀꢀ
2
PdOxHy/CeO2
Pd/CeO2
Pd/CeO2
Pd/Al2O3
Pd/SiO2
Pd/C
0.0
ꢀꢀꢀ
ꢀꢀꢀ
3
86.2 ± 1.8c
94.1 ± 0.4
7.4 ± 3.4
36.6 ± 3.3
42.7 ± 2.5
96.3 ± 1.4
90.8 ± 3.4
100.0 ± 0.0
100.0 ± 0.0
95.4 ± 2.6
96.2 ± 2.0
3.7 ± 1.4
9.2 ± 3.4
0.0
4
5
6
0.0
7a
8b
4.6 ± 2.6
3.8 ± 2.0
Pd/CeO2
25
62.5 ± 3.6
aCommercial Pd/C. Commercial ceria. Standard deviations calculated from 3 different
b
c
catalytic runs.
The reduction temperature used during the synthesis of Pd/CeO2 catalyst greatly affected its
activity for phenol hydrogenation. The catalysts reduced at 150, 250 and 450 °C gave
significantly lower conversions than the catalyst reduced at 350 °C (Figure 5). The differences
can be attributed to the dispersion and the nature of the active sites (Figures 4, S1). The changes
in TPR profiles relative to catalyst reduction temperature must reflect the changes that occur in
the interactions between the metal and support. Considering that the amount of H2 taken up
during the TPR experiments significantly exceeded the stoichiometric amount required to reduce
the precatalyst, the temperatures of the TPR peaks could be taken as indicators of the capacity of
ACS Paragon Plus Environment
20

Page 21 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
the catalytic sites to activate H2. The catalysts prepared at 250 and 350 °C gave single TPR peaks
at 28 and 14 °C, and the conversion of phenol using the former as a catalyst was 76 % relative to
that using the latter. The catalyst reduced at 150 °C had a TPR peak at a similar temperature (29
°C) as the sample prepared at 250 °C; however, there was a second catalytic site, identified by a
reduction peak at 95 °C, which could be responsible for lowering the conversion of phenol (42 %
relative to the conversion with the catalyst prepared at 350 °C). While the catalyst prepared at
450 °C had a TPR peak at 8 °C, there were also two additional peaks at 25 and 51 °C that
accounted for 75 % of the H2 consumed. If the latter peaks correspond to sites that are, as
suspected, buried within the CeO2 matrix because of high temperature reduction, then they could
be responsible for the decreased conversion (44 % relative to that of the catalyst prepared at 350
°C). Decreased hydrogenation efficiency upon high temperature reduction of metalꢀsupported
ceria catalysts has also been observed for other systems and has been attributed to catalyst
burial,57,58 which could result from support sintering induced by thermal treatment under H2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Relative conversion of phenol over Pd/CeO2 as a function of PdOxHy/CeO2 reduction
temperature.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 38
1
2
3
4
5
6
7
8
9
While H2 adsorption and activation takes place at Pd sites, phenol is expected to adsorb to the
surface of the support. The high catalytic activity of Pd/CeO2 could result from high dispersion
of Pd and H2 spillover onto the reducible support, but also from activation of phenol upon
adsorption to surface sites of CeO2. To better understand how phenol interacts with the CeO2
support, temperature programmed DRIFT studies were conducted. The spectrum of CeO2 after
evacuation under vacuum at 150 °C showed four distinct bands at 3710, 3689, 3660, and 3554
cmꢀ1 (Figure 6a). Lavalley and coworkers assigned the bands to monoꢀcoordinated OH (I)
species (3710 cmꢀ1), doubly bridging OH (II) species (3657 cmꢀ1), triply bridging OH (III)
species (3581 cmꢀ1), and surfaceꢀbound water (3687 cmꢀ1).59ꢀ61 The latter was classified on the
basis of the characteristic δ(OH) mode at 1631 cmꢀ1 and the disappearance of the band at 3687
cmꢀ1 upon evacuation at 150 °C. However, our experimental data showed no δ(OH) mode at
1631 cmꢀ1 after evacuation at 150 °C (Figure 6b) while the band at 3689 cmꢀ1 remained even
after increasing the temperature to 200 °C. The band at 3689 cmꢀ1 is likely due to the hydroxyl
stretch of surfaceꢀbound hydrogen carbonate species.62 The other peaks in Figure 6b correspond
to various binding modes of carbonate from contamination during the synthesis and/or
calcination in air.62 After evacuation at 150 °C, a dilute solution of phenol in hexane was
deposited onto CeO2 at room temperature under airꢀ and waterꢀfree conditions. The difference
spectrum in the hydroxyl region clearly shows that some hydroxyl bands have disappeared
indicated by the negative peaks at 3710, 3689, and 3660 cmꢀ1 (Figure 6c). In addition, there was
an emergence of a band at 1631 cmꢀ1 attributed to newlyꢀformed water (Figure 6d). Evacuation at
150 °C resulted in the removal of the band at 1631 cmꢀ1 (Figure 6d, red). The disappearance of
hydroxyl groups and the appearance of water suggest that phenol undergoes dissociative
chemisorption on ceria to form phenoxy species (Scheme 1). The formation of phenoxy species
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
22

Page 23 of 38
ACS Catalysis
1
2
is also consistent with the basic nature of cerium dioxide.31 The CꢀO stretching frequency of
ceriumꢀbound phenoxy is expected to shift relative to that of phenol,63 and the OꢀH bands of
phenol should not be observed. The peaks in Figure 6d at 1587, 1479, and 1273 cmꢀ1 correspond
to adsorbed phenyl species (Figure S3). For comparison, the intense CꢀO stretching frequency of
phenol occurs at 1259 cmꢀ1 (Figure S4a).64 Figure 6d shows an intense absorption band at 1273
cmꢀ1, which is attributed to the CꢀO stretch of a phenoxy molecule covalently bonded to a cerium
cation. The increased CꢀO vibrational energy relative to phenol is consistent with the formation
of ceriumꢀbound phenoxy owing to the strong interaction expected between cationic cerium and
anionic oxygen.63 This likely results in electron withdrawal from phenoxy towards cerium cation
and shortening (i.e. strengthening) of the CꢀO bond. The various OꢀH bands of phenol are
typically observed at 3623, 3350, 1343, and 1207 cmꢀ1.64 The OꢀH outꢀofꢀplane (1343 cmꢀ1) and
intense inꢀplane (1207 cmꢀ1) bends are absent upon phenol adsorption onto CeO2 (Figure S4a).
Also, Figure 6c and Figure S4b show no indication of the OꢀH stretch at 3623 and 3350 cmꢀ1.
The only increased intensity is observed in the triply bridging hydroxyl region (3552 cmꢀ1) even
after degassing the sample at 150 °C under vacuum (Figure S5), which suggests new triply
bridging hydroxyl groups are formed upon adsorption of phenol (Scheme 1c).65
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. DRIFT spectra of CeO2 recorded at 150 °C under evacuation from (a) 4000ꢀ3000 cmꢀ1
and (b) 1800ꢀ1200 cmꢀ1, (c) DRIFT difference spectrum of CeO2 evacuated at 150 °C and phenol
adsorbed onto CeO2 at room temperature, (d) DRIFT spectra of phenol adsorbed onto CeO2 at
room temperature and under evacuation at 150 °C.
Scheme 1. Possible phenol adsorption mechanisms on CeO2. Phenol undergoes dissociative
adsorption onto cerium cation (a) adjacent to a surface hydroxyl, (b) containing a surface
hydroxyl, and (c) at an oxygen vacancy to produce triplyꢀbridging surface hydroxyl in its
neighborhood.
ACS Paragon Plus Environment
24

Page 25 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
The phenol adsorption analysis suggests that coordinatively unsaturated cerium cations near
hydroxyl groups may be active sites for the reaction. It is not entirely clear, however, whether the
newlyꢀformed water stays bound to the surface or desorbs under the reaction conditions. To
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
address the latter, HꢀNMR spectroscopy studies were conducted in deuterated benzene. The
CeO2 support was degassed at 150 °C under vacuum and placed into a NMR tube containing
1
phenol dissolved in benzeneꢀd6. The HꢀNMR spectrum showed the formation of water and
decrease of phenol upon addition of CeO2 (Figure S6). Introduction of more CeO2 into the same
NMR tube led to increased water signal and decreased phenol signal. Because the signals from
phenol and water result from the liquid phase species (i.e nonꢀsurfaceꢀbound), it can be
concluded that water desorbs from the surface of CeO2 in benzene. However, when CeO2 was
introduced into an NMR tube containing phenol dissolved in hexaneꢀd14 phenol signals were lost
from the spectra but no water was observed in the liquid phase, even though water was detected
on the solid by DRIFT analysis. Thus, the water formed upon phenol adsorption on CeO2
remains bound to the surface of the support when the reaction is performed in hexane.
The production of water upon phenol binding suggested that performing the reaction in water
should affect the binding equilibrium and consequently conversion. Indeed, the catalytic activity
decreased dramatically when the reaction was run in water, requiring up to 12 h to achieve 69 %
conversion at room temperature and 1 bar H2 (Table S2). Inhibition of the reaction was likely
due to adsorption of water on CeO2 competing with phenol, thus limiting formation of phenoxy
species. Dissociative and nonꢀdissociative adsorption of water on ceria has been shown to occur
at room temperature and is even more pronounced on reduced ceria.66 The formation and relative
amounts of phenoxy species could be indirectly determined by the color change of the material,
which results from the charge transfer of the bound phenoxy to cerium cations.67 The color
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 38
1
2
3
4
5
6
7
8
9
change of CeO2 was obvious when in contact with hexane solution of phenol, but was less
apparent in an aqueous solution of phenol (Figure 7a). Diffuse reflectance UVꢀVis measurements
of the solids obtained by subtracting the contribution of CeO2 clearly showed a higher
absorbance (λmax 480 nm) when hexane was used as solvent compared to water, an indication of
higher amounts of phenoxy species present on the former (Figure 7b). These findings suggest
that the increased catalytic activity in hexane relative to water may be partially due to the
increased amount of phenoxy species formed on the ceria surface. The formation of a ceriumꢀ
bound phenoxy group as an intermediate in the reduction can lead to a similar path of activation
as the one exploited by Liu,7 Lee,8 Li and Luque,11 because the direct complexation of oxygen to
cerium cation results in electron withdrawal from the ring. This activation mechanism is also
consistent with the increased vibrational energy of the CꢀO stretch of phenol adsorbed onto CeO2
(Figure 6d).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. (a) CeO2 samples treated with
phenol, and water, (b) DR UVꢀVis difference spectra of CeO2 and CeO2 after removal from
different phenol solutions.
1 hexane solution of phenol, 2 aqueous solution of
3
ACS Paragon Plus Environment
26

Page 27 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
Catalyst recycling experiments performed in hexane and water provided additional insight on
the system (Figure 8). In hexane, the conversion decreased by nearly 50 % after the third cycle,
but then appeared to stabilize during subsequent reaction cycles. In water, the decrease in
catalyst performance was much milder with successive cycles. After an initial decrease of about
10 %, the conversion became steady in the following cycles. ICPꢀOES analysis of both hexane
and water spent solutions had no detectable metal, indicating that leaching was not the cause of
the decreased activity.68 In addition, hot filtration experiments proved the active phase was not
homogenous.69,70 Interestingly, the catalytic activity could be fully recovered for both samples by
regeneration via oxidation followed by reduction of the materials.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Cycling experiments for the hydrogenation of phenol using Pd/CeO2 (a) in hexane at T
= 25 °C, (b) in water at T = 35 °C. Reaction conditions: t = 4 h, PH2 = 1 bar, Pd : Phenol = 5 mol
%.
The catalysts were washed with the solvents used in the reaction, dried and analyzed by
DRIFT to elucidate the types of surface species present after the conversion (Figure 9). The
different possible types of surfaceꢀbound species were independently supported by DRIFT
spectra of phenol, cyclohexanone, and cyclohexanol deposited onto the catalyst (Figure S3). The
bands at 3070, 1596, 1482, 1263, and 1166 cmꢀ1 were assigned to adsorbed phenyl species and at
2929, 2854, 1446, 1400, and 1157 cmꢀ1 to cyclohexyl species. The bands at 3699, 3558 cmꢀ1
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 38
1
2
3
−
correspond to surface hydroxyls, and at 1654, 1544, and 1236 cmꢀ1 to H2O, CO2, and HCO2 on
the catalyst surface, respectively (Figure S7).61,62 Phenyl and cyclohexyl compounds
(cyclohexanone and/or cyclohexanol) were observed on the catalyst for the reaction run in
hexane (black trace), while only cyclohexyl compounds were observed on the catalyst used in
water (red trace). The absence of adsorbed phenyl species for the reaction in water, even if the
reaction is incomplete, supports the idea that water competes for adsorption with phenyl or
phenoxy species, which is consistent with the visual inspection and DR UV/Vis measurements of
CeO2 treated with aqueous phenol (Figure 7). Also, the hydroxyl band at 3699 cmꢀ1 is present on
the catalyst for the reaction in water, but absent for the reaction in hexane. The absence of the
hydroxyl band for the reaction in hexane is consistent with DRIFT analysis of phenol adsorbed
onto CeO2 (Figure 6c) and the presence of the hydroxyl band for the reaction in water is
consistent with the lower amounts of phenoxy species (formed at the expense of hydroxyl
species) on CeO2 in water (Figure 7).
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. DRIFT spectra of Pd/CeO2 catalysts after reaction in hexane (black) and water (red).
The transformation of ceria hydroxyl groups into water during phenol adsorption and the
decreased activity upon cycling suggests that cerium cations near hydroxyls are highly effective
ACS Paragon Plus Environment
28

Page 29 of 38
ACS Catalysis
1
2
3
4
5
6
7
8
9
sites for phenol activation in hexane and were not regenerated in situ. The data put forth suggests
that the fundamental difference between catalyst activity and recyclability in water and hexane is
the way in which phenol adsorbs to the CeO2 support. While dissociative adsorption of phenol
onto the support is significant in hexane, the interaction in water seems to occur mainly through
physisorption, as dissociative binding is limited by competitive adsorption of water. Physisorbed
phenol is expected to be less active for reduction by dissociated H2 than phenoxy coordinated to
cerium cation. The reason being, the cerium centers are better able to activate the phenolic ring
for nucleophilic attack by dissociated H2 because they pull more electron density away from the
ring, than simple hydrogen bonds with hydroxyls/water on the ceria surface. Thus, the higher
activity observed in hexane should be due to the dissociative phenol adsorption and its direct
binding to coordinatively unsaturated Ce cations, which is enabled by the lack of competition
between substrate and solvent (Scheme 2). However, upon phenoxy formation and subsequent
reduction to the ketone or alcohol, the phenoxyꢀforming sites were seemingly not regenerated as
indicated by the decreased activity observed in the recycling experiments. In order for the active
site to be retained, a hydroxyl near a coordinatively unsaturated cerium cation should be present,
which is not the case after one phenol molecule adsorbs dissociatively and is reduced (Scheme
2). The proposed deactivation mechanism is supported by the red shift of the CꢀO band for
phenol adsorbed on the fresh (1273 cmꢀ1) and recycled (1263 cmꢀ1) catalyst (Figure S8). Also,
the CꢀO band red shift on the recycled catalyst indicates that phenol is no longer able to bind
dissociatively to ceria (Figure 9). The proposed solventꢀdependent adsorption mechanism is
consistent with the greater amounts of phenoxy species formed in hexane relative to in water, the
higher catalytic activity in hexane compared to in water, the sharp decrease in activity after the
first catalytic run in hexane, and the stability of catalyst activity after cycling experiments in both
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
29