Enhanced Hydrodeoxygenation of m-Cresol over Bimetallic Pt-Mo Catalysts through an Oxophilic Metal-Induced Tautomerization Pathway

Article abstract of DOI:10.1021/acscatal.6b01131

Supported bimetallic catalysts consisting of a noble metal (e.g., Pt) and an oxophilic metal (e.g., Mo) have received considerable attention for the hydrodeoxygenation of oxygenated aromatic compounds produced from biomass fast pyrolysis. Here, we report that PtMo can catalyze m-cresol deoxygenation via a pathway involving an initial tautomerization step. In contrast, the dominant mechanism on monometallic Pt/Al₂O₃ was found to be sequential Pt-catalyzed ring hydrogenation followed by dehydration on the support. Bimetallic Pt₁₀Mo₁ and Pt₁Mo₁ catalysts were found to produce the completely hydrogenated and deoxygenated product, methylcyclohexane (MCH), with much higher yields than monometallic Pt catalysts with comparable metal loadings and surface areas. Over an inert carbon support, MCH formation was found to be slow over monometallic Pt catalysts, while deoxygenation was significant for PtMo catalysts even in the absence of an acidic support material. Experimental studies of m-cresol deoxygenation together with density functional theory calculations indicated that Mo sites on the PtMo bimetallic surface dramatically lower the barrier for m-cresol tautomerization and subsequent deoxygenation. The accessibility of this pathway arises from the increased interaction between the oxygen of m-cresol and the Mo sites in the Pt surface. This interaction significantly alters the configuration of the precursor and transition states for tautomerization. A suite of catalyst characterization techniques including X-ray absorption spectroscopy (XAS) and temperature-programmed reduction (TPR) indicate that Mo was present in a reduced state on the bimetallic surface under conditions relevant for reaction. Overall, these results suggest that the use of bifunctional metal catalysts can result in new reaction pathways that are unfavorable on monometallic noble metal catalysts.

Full text of DOI:10.1021/acscatal.6b01131

Subscriber access provided by LAURENTIAN UNIV
Article
Enhanced hydrodeoxygenation of m-cresol over bimetallic Pt-Mo
catalysts through an oxophilic metal-induced tautomerization pathway
Allison M Robinson, Glen A. Ferguson, James R Gallagher, Singfoong Cheah,
Gregg T Beckham, Joshua A. Schaidle, Jesse E. Hensley, and J. Will Medlin
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01131 • Publication Date (Web): 26 May 2016
Downloaded from http://pubs.acs.org on June 5, 2016
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 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Enhanced Hydrodeoxygenation of m-Cresol over Bimetallic Pt-Mo Catalysts through an Oxophilic Metal-Induced
Tautomerization Pathway
1
2
3
2
2
Allison Robinson §, Glen Allen Ferguson §, James R. Gallagher , Singfoong Cheah , Gregg T. Beckham , Joshua A.
2
,
2
1,
Schaidle *, Jesse E. Hensley , and J. Will Medlin *
Chemical and Biological Engineering Department, University of Colorado, Boulder, CO 80303
National Bioenergy Center, National Renewable Energy Laboratory, Golden, CO 80401
Chemical Sciences and Engineering Division, Argonne National Laboratory, Argonne, IL 60439
1
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
§Equally contributing authors
*Corresponding Authors: J. Will Medlin (will.medlin@colorado.edu) and Joshua A. Schaidle (joshua.schaidle@nrel.gov)
Abstract
Supported bimetallic catalysts consisting of a noble metal (e.g., Pt) and an oxophilic metal (e.g., Mo) have received
considerable attention for the hydrodeoxygenation of oxygenated aromatic compounds produced from biomass fast
pyrolysis. Here, we report that PtMo can catalyze m-cresol deoxygenation via a pathway involving an initial
tautomerization step. In contrast, the dominant mechanism on monometallic Pt/Al O was found to be sequential Pt-
2
3
catalyzed ring hydrogenation followed by dehydration on the support. Bimetallic Pt Mo and Pt Mo catalysts were
10
1
1
1
found to produce the completely hydrogenated and deoxygenated product, methylcyclohexane (MCH), with much higher
yields than monometallic Pt catalysts with comparable metal loadings and surface areas. Over an inert carbon support,
MCH formation was found to be slow over monometallic Pt catalysts while deoxygenation was significant for PtMo
catalysts even in the absence of an acidic support material. Experimental studies of m-cresol deoxygenation together with
density functional theory calculations indicated that Mo sites on the PtMo bimetallic surface dramatically lower the
barrier for m-cresol tautomerization and subsequent deoxygenation. The accessibility of this pathway arises from the
increased interaction between the oxygen of m-cresol and the Mo sites in the Pt surface. This interaction significantly
alters the configuration of the precursor and transition states for tautomerization. A suite of catalyst characterization
techniques including X-ray absorption spectroscopy (XAS) and temperature programmed reduction (TPR) indicate that Mo
was present in a reduced state on the bimetallic surface under conditions relevant for reaction. Overall, these results
suggest that the use of bifunctional metal catalysts can result in new reaction pathways that are unfavorable on
monometallic noble metal catalysts.
Key Words
Platinum, molybdenum, bimetallic, biomass, hydrodeoxygenation, oxophilic promoter, catalytic fast pyrolysis
Introduction
Lignocellulosic biomass is a potentially useful feedstock for production of renewable fuels and chemicals. One method of
1-3
biomass conversion is ex-situ catalytic fast pyrolysis (CFP). Fast pyrolysis breaks down the biomass at high temperatures
4
in the absence of oxygen to produce a complex mixture of hydrocarbons and oxygenates, which includes large quantities
5-7
of lignin-derived oxygenated aromatic species (i.e., phenolics). This complex mixture can then be catalytically upgraded
in the vapor or liquid phase to improve the fuel quality and stability of the resulting bio-oil through hydrogen
8
incorporation and oxygen removal. Phenolic compounds comprise 30-40% of the lignin fraction, and are difficult to
9
deoxygenate because of their high C–O bond strength. Accordingly, selective catalytic deoxygenation of the aromatic
fraction of pyrolysis oil, either in the vapor or condensed phase, is an essential step in thermochemical conversion of
2,3,10,11
biomass into useful fuels and other products.
1
,11-13
A variety of catalysts have been studied for aromatic deoxygenation
significant attention due to their relatively high activity and stability.
, and noble metal catalysts have received
As shown in Scheme 1, several mechanisms for
1
1,14,15
1
5,16
deoxygenation of oxygenated aromatics have been proposed including ring hydrogenation-dehydration (Scheme 1a)
direct deoxygenation (Scheme 1b) , ring hydrogenation-direct C–O scission (Scheme 1c) , and a tautomerization
,
17
14,18
19-21
mechanism (Scheme 1d).
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3-methylcyclohexanone 3-methylcyclohexanol
a.
+
H
2
a.
O
OH
b.
toluene
2
a. - H O
+
H
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
- H O
b. d.
methylcyclohexane
OH
c.
2
m-cresol
d.
2
2
O
H
-
d.
O
Scheme 1: Proposed deoxygenation mechanisms over supported noble metal catalysts: a) ring hydrogenation-
dehydration, b) direct deoxygenation, c) ring hydrogenation-direct C-O scission, and d) tautomerization.
In addition to the active phase used, the preferred pathway depends upon the reaction conditions and the catalyst
support. The most cited mechanism for supported noble metal catalysts on acidic supports is ring hydrogenation on the
1
5,16,21,22
metal followed by dehydration on the support.
metal catalysts using an inert carbon support, demonstrating that aryl-O bond cleavage can occur over the metal active
Deoxygenation has also been observed in some cases over noble
1
8,23,24
phase alone.
Bimetallic catalysts that combine oxophilic metals with noble metals exhibit promising selectivity and activity for
25-32
deoxygenation reactions.
For example, Rh-based catalysts that were modified with Re or Mo were shown to increase
activity and selectivity in C-O bond scission reactions using cyclic ethers. One of the reactants tested, tetrahydrofurfuryl
alcohol, produced 1,5-pentanediol at over 90% selectivity on the bimetallic catalysts vs. less than 60% selectivity on
monometallic Rh. These bimetallic catalysts have been proposed to reduce the barrier for C–O scission through a strong
interaction of the oxophilic metal with the oxygen-containing moiety of the aromatic compound. This mechanism has
3
1
2
0
been proposed for NiFe catalysts used for vapor-phase hydrodeoxygenation (HDO) of m-cresol to toluene and for the
18
deoxygenation of phenol by PdFe catalysts. Studies on Fe-based catalysts indicate that while oxophilic metals can
14,20
themselves deoxygenate m-cresol, they suffer from poisoning due to surface oxidation.
The addition of noble metals
14
to these catalysts increased water desorption and the stability of the reduced Fe.
14,20,33
Previous experimental results have been complemented by computational studies.
One investigation of phenol on
an Fe(110) surface showed water formation as the rate-limiting step, with a high barrier of 1.32 eV for a stepwise
mechanism of single hydrogen ring hydrogenation, C–O bond scission, and intermolecular hydrogen transfer. These
studies focused on isolated surface intermediates at low coverages. However, separate computational studies have
33
3
4
suggested that the Fe(110) surface would have high hydrogen coverage at reaction temperatures and pressures.
Moreover, phenolic species are known to adsorb strongly to late transition metal surfaces, creating a crowded surface
3
5
environment that strongly affects reaction chemistry. A substantial body of previous work for other reactions, such as
olefin hydrogenation, shows that high coverage of co-adsorbed surface species can strongly influence reaction
3
6-38
mechanisms.
Furthermore, prior work has suggested that the effects of high coverage can be modeled by adsorption
39
of a simple adsorbate such as hydrogen.
In this work we used a combined experimental and computational approach to elucidate the role of oxophilic metals in
the deoxygenation of m-cresol, a phenolic model compound. Supported Pt catalysts with various Mo loadings were
evaluated for m-cresol hydrodeoxygenation. These catalysts, as well as model PtMo surfaces, were characterized using a
variety of techniques including X-ray absorption spectroscopy (XAS), chemisorption of probe molecules, and temperature-
programmed reduction (TPR). Complementary density functional calculations were performed to map the reaction
pathways for m-cresol HDO on Pt(111) with Mo atoms included in the lattice; importantly, the model included a high
coverage of hydrogen to investigate the HDO chemistry on the types of crowded surfaces expected under reaction
conditions. As discussed below, the combined studies strongly suggest that addition of Mo to Pt catalysts creates active
sites for m-cresol tautomerization to an unstable alcohol that has a significantly lower barrier to deoxygenation.
2
ACS Paragon Plus Environment

Page 3 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Methods
Catalyst Synthesis
Catalysts were prepared by incipient wetness impregnation of a porous Υ-Al O support (Albemarle MARTOXID AN/I,
2
3
2
1
50 m /g BET surface area, 75-μm particle diameters). Aqueous solutions were prepared of the metal precursors using
chloroplatinic acid hexahydrate (Strem Chemical Inc.) and ammonium molybdate tetrahydrate (Strem Chemical Inc.).
Pt/Al O was prepared with a nominal loading of 5 wt%. PtMo/Al O catalysts were prepared via co-impregnation of the
2
3
2
3
two precursors with a 10:1 and 1:1 atomic ratio of Pt:Mo, containing the same amount of Pt with Mo added to create the
desired composition at increased total metal loading. The catalysts were dried in air at 373 K then calcined for 2 hours at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
23 K in 50 sccm He and 30 sccm O . Carbon supported catalysts were prepared using the same metal precursors and
2
deposited on a carbon black powdered support (Vulcan XC 72R). After preparation the catalyst was dried in air at 393 K
for 10 h. Elemental analysis was performed to determine actual weight loading of each metal using inductively coupled
plasma atomic emission spectroscopy (ICP-AES).
Catalyst Characterization
CO pulse chemisorption was performed on a Micromeritics ChemiSorb 2720 to characterize active metal surface area.
Approximately 100 mg of catalyst was reduced in-situ at 673 K for 2 h in 40 sccm H and then purged in 20 sccm Ar for 15
2
h without exposure to air. This temperature is lower than that used for the reactor studies due to the upper limit of the
system used. However, temperature-programmed reduction (TPR) results presented here show no additional reduction
features from 673-723 K. CO chemisorption was performed at 323 K using 1 mL injections of 20% CO in Ar. This
40
temperature was chosen to ensure that any CO dosed would not adsorb onto the alumina surface while remaining low
41
42
43
enough to ensure no desorption would be expected from platinum , molybdenum , or Pt-Mo sites. These injections
were repeated until the intensity of the CO desorption peak remained constant within ± 0.5%, indicating that no
additional CO was adsorbing.
Particle sizes were also measured using transmission electron microscopy (TEM) performed on an FEI Tecnai G2 ST20 TEM
operating at 200 kV. Particle size estimates were obtained using ImageJ software to analyze images, sampling at least 100
particles for each material.
TPR experiments were performed in a U-shaped quartz tube at atmospheric pressure. Approximately 100 mg of catalyst
was held in the tube with quartz wool and pretreated at 723 K for 1 h in 40 sccm of 1% O in He. The sample was then
2
cooled to room temperature in He. 0.2 sccm H and 20 sccm He were then flowed over the sample as the temperature
2
was ramped linearly at 10 K/min to 1073 K with outlet gas monitored by a quadrupole residual gas analyzer (Stanford
Research Systems model RGA 100).
Acid site characterization was performed using ammonia temperature-programmed desorption (TPD) in a reactor
described in detail elsewhere. Approximately 12 mg of catalyst was loaded into a quartz tube (6.35 mm inner diameter)
and reduced in-situ at 723 K for 1 hour in 20 sccm H , 40 sccm He. The tube was then cooled under He flow to 373 K and
2
three 1 mL injections of NH (Airgas, anhydrous, 99.99%) were performed to ensure saturation of the catalyst. Then three
3
2
00 μL injections of NH were performed for use as a calibration for quantification of the resulting TPD. Desorption of
3
ammonia was measured using a Pfeiffer mass spectrometer to track mass fragment m/z=15. The area under this peak was
used to quantify acid site density by comparing it to the area of the calibration injections.
In-situ XAS was performed at Argonne National Laboratory at the Advanced Photon Source (APS) on beamline 10-BM,
operated by the Materials Research Collaborative Access Team (MRCAT). Experiments were performed in a continuous
flow reactor consisting of a quartz tube (2.54 cm OD, 25.4 cm length) sealed at each end with Kapton windows held in
place by Ultra-Torr fittings with ball valves welded to each to allow gas flow through the tube. Catalysts were pressed into
a cylindrical sample holder containing six wells to produce a self-supporting wafer. A K-type thermocouple (Omega) was
placed against the catalyst sample holder to measure temperature. The sample loading was determined by calculating the
weight of catalyst required to give an absorbance (μx) of approximately 1.0 when possible. Spectra were collected in
transmission mode using ionization chamber detectors that were optimized for maximum current while maintaining a
10
-1
linear response range (ca. 10 photons detected s ). To calibrate the energies, a third detector in series with the two
ionization chambers was used to collect a foil reference spectrum simultaneously with each measurement. Before
collecting the spectra, the sample was reduced in-situ at 723 K in hydrogen and then allowed to cool to room
temperature. This reduction temperature was chosen based on literature precedent for successfully reducing PtMo
3
1
catalysts. TPR performed on these catalysts was also used to confirm that the chosen reduction temperatures were
higher than the observed hydrogen uptake features.
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Standard methods were employed to normalize and calibrate the energy of the X-ray absorption near-edge structure
XANES) spectra. The maximum of the first peak was used to calculate edge energies by taking the first derivative of the
(
XANES spectrum while the energy of pre-edge features was determined by the maximum in the peak intensity. Linear
combination fitting of the XANES was performed where possible to estimate the fraction of metal in different oxidation
4
4
states. Standard data reduction techniques were used to fit the data using the software program WinXAS 3.1. The
extended X-ray absorption fine structure (EXAFS) parameters were determined by a least square fit in k or R-space (as
2
labeled in the table of fits) of the k -weighted Fourier Transform (FT) data. Where possible, experimental phase shift and
backscattering amplitude were determined using standards of known structure; Pt-Pt scattering was determined using Pt
foil (12 Pt-Pt at 2.77 Å) and Mo-O scattering was determined using Na MoO (4 Mo-O at 1.77 Å). Theoretical phase and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
45
amplitude functions for Pt-Mo and Mo-Pt scattering were calculated with FEFF6 using a two atom calculation, calibrated
2
to Pt foil and Mo foil respectively. The mean-square disorder refined in WinXAS (Δσ ) reflects the difference in disorder of
the nanoparticles relative to the references.
Upgrading Experiments
Approximately 10 to 50 mg of catalyst, that was pressed and sieved to an agglomerated material of 0.125-0.15 mm
diameter, was diluted with approximately 1.5 g silicon carbide of similar size to ensure isothermal operation. The mixture
was loaded into a 13 mm ID stainless steel reactor tube with the catalyst bed held in place by quartz chips. The catalyst
was reduced in-situ at 723 K for 2 h in 95% H (balance Ar). The reactor was then cooled and purged at room temperature
2
for at least 12 h in N without exposure to air. Experiments were then performed at 0.5 MPa and 523 K with 12:1 molar
2
ratio of H :oxygenate. These reaction conditions were chosen to study HDO at relatively low temperature and pressure; it
2
would be preferable to perform ex-situ CFP without additional heating or pressurization of the vapor. The hydrogen fed
was pre-blended with an Ar internal standard. Oxygenates were fed as a liquid using an HPLC pump to a heated zone
upstream of the reactor where they were vaporized prior to reaching the inlet. The weight-hourly space velocity (WHSV)
was varied to obtain data points at varying conversion. Products from the reactor outlet were monitored using an online
gas chromatography/mass spectrometry system equipped with flame ionization and thermal conductivity detectors
(Agilent). Concentrations of each compound were identified and quantified using standards prepared with known
concentrations of the relevant compounds. These standards were prepared at a minimum of three concentrations
spanning the range of concentrations expected under reaction conditions. This allowed for identification of retention
times as well as quantification of concentration. Product identification was further confirmed using the system’s built-in
mass spectrometer. All data had a mass balance closure within ± 5%. Operation under rate-limited conditions was
confirmed by changing the space velocity both by altering flow rates with the same catalyst mass and by loading a
different mass of catalyst at the same reactant flows and noting that the results were the same. Internal mass transfer
limitations were ruled out using the Weisz-Prater parameter, calculated to be approximately 0.02 (see calculations in the
46,47
Supporting Information).
Conversion was calculated by dividing the moles of carbon-containing primary products
formed by the total moles of these products and the remaining reactant measured in the outlet. For example, for m-
cresol (C) upgrading, the products detected were methylcyclohexane (R), 3-methylcyclohexanol (OH), and 3-
methylcyclohexanone (O). The conversion (X) was determined using equation 1
ꢂ ꢄꢂꢅꢆꢄꢂꢅ
ꢃ
ꢀꢁ
=
(equation 1)
ꢂ ꢄꢂ ꢄꢂꢅꢆꢄꢂꢅ
ꢇ
ꢃ
where N is the moles of compound i. Conversions were determined at stead state, after approximately 2 hours on stream
i
at the given condition. Rates were computed by multiplying the measured mol% of a given compound in the outlet and
the total molar flow rate through the system. The water-free product selectivities reported were determined on a molar
basis by dividing the moles of a given species by the total moles of products formed. Weight hourly space velocity (WHSV)
was calculated by dividing the mass flow rate of the reactant by the total mass of catalyst loaded.
A separate, atmospheric pressure tubular packed bed flow reactor optimized for measuring kinetics at low conversion was
used for some experiments. Reactants were introduced by bubbling helium through the liquid reactant (either m-cresol or
3
-methylcyclohexanol) that was immersed in a water bath at 333 K. The saturated He was then mixed with H and diluent
2
He before flowing over the catalyst bed. The reactor effluent was analyzed using an Agilent Technologies 7890A gas
chromatograph with a 30 m × 0.320 mm Agilent HP-5 (5% phenyl)-methylpolysiloxane capillary column and flame
ionization detector.
Density Functional Theory (DFT)
All calculations used periodic DFT as implemented in the Vienna Ab initio Simulation Package 5.3.3 (VASP).
electron interactions were described using the projector augmented wave (PAW) potentials
48-51
The ion-
52,53
with an energy cutoff of
4
ACS Paragon Plus Environment

Page 5 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
400 eV for the plane wave basis set. The generalized gradient corrected Perdew-Burke-Ernzerhof (PBE) functional was
used for all periodic calculations with Monkhorst-Pack 5×5×1 k-point sampling. Optimizations were carried out until the
forces converged to 0.05 eV/Å for geometry optimizations and 0.08 eV/Å for nudged elastic band calculations using the
quasi-Newton method for geometry optimization. The van der Waals (vdW) forces were calculated using the method of
54,55
Tkatchenko and Scheffler as implemented in VASP.
To understand the reactivity of m-cresol over pure Pt(111) and PtMo(111), models were constructed for each surface.
Slab models were formed from the optimized platinum unit cell. The molybdenum was added to a Pt(111) surface by
substituting a Mo atom for one of the Pt atoms. The slabs were otherwise unchanged. The slab geometries were
generated using Atomic Simulation Environment 3 (ASE). The optimized lattice vectors for the slab were (11.247, –5.624,
0.000 × 0.000, 9.741, 0.000 × 0.000, 0.000, 30.000). The final vector allows for sufficient distance between slabs to avoid
unphysical interactions between slabs. The surface models are five-layers thick with a stoichiometry of Pt80 and Pt Mo .
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
79
1
The upper two layers were relaxed while the lower layers were fixed at their ideal crystal positions. The Mo atoms were
included in the surface layer.
Previous ab initio thermodynamics studies indicate the Pt(111) surface would have a high coverage of hydrogen at the
56,57
experimental temperature and pressure.
Previous work with Pt and PtRe supported catalysts used for glycerol
hydrogenolysis at 548 K reported that the reaction order with respect to hydrogen was zero, further suggesting that the
surface is crowded under the experimental conditions used here. For these reasons, all of the calculations were started
using a surface with complete hydrogen coverage or a single surface vacancy in the hydrogen monolayer. For the models
including Mo, the vacancy was over the Mo site, as the hydrogen vacancy formation energy is zero at the Mo site. The full
models including the reacting m-cresol are shown in Figure 1. Barriers were determined using nudged elastic band
calculations with 7-11 images in addition to the reactants and products. All energies were referenced to the preceding
step with hydrogen adsorption energies neglected in the reference and at each step in the reaction. This model of a PtMo
surface is not meant to replicate experimental results, but rather to provide an understanding of fundamental
interactions driving reactivity by understanding the effect of Mo atoms that are interacting with Pt atoms.
z
y
Pla num
x
Molybdenum
Oxygen
Carbon
Hydrogen
z
y
x
Pt80
79
Pt Mo1
Figure 1: The models used for DFT calculations of the reaction network.
Results
M-cresol HDO
The vapor phase upgrading of m-cresol was studied over Pt/Al O and PtMo/Al O catalysts. At 523 K, 0.5 MPa, and 12:1
2
3
2
3
H :m-cresol, the major products formed were methylcyclohexane (MCH), 3-methylcyclohexanol, and 3-
2
methylcyclohexanone over all materials, shown in Scheme 1. Benzene and toluene were observed in trace amounts (less
58
than 0.5 mol%). No transmethylation reactions were observed, consistent with a prior report. At approximately 10%
-1
-1
conversion the rate of product formation in units of mol-min -g catalyst was 0.006 over Pt, 0.004 Pt Mo , and 0.003
10
1
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
over Pt Mo . Using the CO chemisorption data presented in the characterization section below to estimate the moles of
1
1
-
1
exposed sites, these rates correspond to turnover frequencies (TOF) of 0.76, 0.65, and 0.68 s over Pt, Pt Mo , and
10
1
Pt Mo respectively. Activity trends are further analyzed below.
1
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2: Selectivities at 523 K, 0.5 MPa for 3-methylcyclohexanone (
methylcyclohexane ( ) production from m-cresol HDO over alumina-supported A. Pt, B. Pt Mo , and C. Pt Mo
1
), 3-methylcyclohexanol (
+), and
1
0
1
1
The product selectivities measured as a function of conversion are reported in Figure 2. This selectivity data can be seen
as a function of space velocity instead of conversion in the Supporting Information (Figure S1). The selectivity to 3-
methylcyclohexanone was highest over monometallic Pt, whereas the bimetallics had a greater selectivity towards 3-
methylcyclohexanol and MCH, especially Pt Mo . In addition to higher deoxygenation selectivity, the bimetallic catalysts
1
1
showed higher activity for MCH production. Over Pt, Pt Mo , and Pt Mo respectively the rates of MCH production at
10
1
1
1
-1
-1
approximately 20% conversion were 0.4, 20, and 10 mmol-min -g catalyst . The alcohol:ketone ratio observed
experimentally over the metal/alumina catalysts was lower than the equilibrium value of ~3.5, especially at lower
conversions, consistent with there being a kinetically relevant barrier for ketone hydrogenation during these experiments.
These ratios can be seen in the Supporting Information (Table S1). On the monometallic Pt/Al O catalyst, the selectivity
2
3
toward 3-methylcyclohexanone approached 100% at low conversion, suggesting that the other reaction products
observed at higher conversion were formed in series from this primary product, as shown in Scheme 1a. This is consistent
with previous work indicating that HDO on Pt/Al O catalysts involves sequential hydrogenation and dehydration steps,
2
3
16,22,59,60 61, 62
with dehydration being attributed to acid sites on the catalyst support.
To confirm that dehydration was
occurring over the support in the experiments reported here, the intermediate product 3-methylcyclohexanol was fed
over pure alumina under the same reaction conditions (523 K, 0.5 MPa total pressure, 12:1 H :alcohol, and the same
2
range of space velocities). For 3-methylcyclohexanol conversions of 25-70%, the product stream was 96-99 mol%
methylcyclohexene, the alkene product formed from dehydration of the saturated alcohol. The balance was 3-
methylcyclohexanone, formed through alcohol dehydrogenation. A plot of these data can be found in the Supporting
Information (Figure S2).
Another test for the role of the support in HDO of cresol is to employ a support with a low acid site concentration. We
therefore tested Pt/C for reaction of cresol at the same conditions. As the carbon supports have low activity for
6
ACS Paragon Plus Environment

Page 7 of 21
ACS Catalysis
1
9
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
dehydration , the reactivity should result from the metal sites. Selectivity toward MCH was observed to be very low over
the Pt/C catalysts, even at high conversion. Thus, the results for the pure Pt/Al O catalyst can largely be understood in
terms of a sequence of hydrogenation reactions on Pt and deoxygenation reactions on the support, though we note that
behavior at high conversions in Figure 2 appears to be complex.
2
3
Mo-containing catalysts exhibited much higher selectivities and rates for MCH formation, i.e. Mo served to provide
additional deoxygenation activity that augmented the activity of acid sites on the support. As shown in Figure 3, a
Pt Mo /C catalyst demonstrated significant MCH production without requiring acidic sites on the support. Using an Al O
1
1
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
support for Pt Mo resulted in even higher MCH selectivities. Thus, one possible explanation for the enhanced MCH rate
1 1
and selectivity over Mo-containing catalysts is that Mo sites simply provide additional acid sites, acting to augment those
provided by the support. To examine this possibility, ammonia TPD experiments were performed over Pt/Al O and
2
3
Pt Mo /Al O , showing less acidity on the bimetallic catalyst. The acid site density over Pt was measured to be 391 ± 38
1
1
2
3
μmol/g catalyst and over Pt Mo was 214 ± 43 μmol/g catalyst (error bars represent the standard deviation based on the
1
1
average of three separate samples). We note that this measurement does not distinguish between types of acid sites,
and the drop in overall measured acid site density with Mo content is not reflected in the near-constant rate of MCH
formation from 3-methylcyclohexanol discussed above. However, these results do suggest that addition of Mo does not
result in a large increase in acid site concentration. Furthermore, it is important to note that the selectivity as a function
of conversion profiles for the bimetallic catalysts differ significantly from their monometallic counterpart. In particular,
the selectivity to MCH was approximately 40% on the Pt Mo /Al O catalyst even at <10% conversion. This high selectivity
1
1
2
3
at low conversion indicates a possible parallel path for formation of MCH (i.e., one that does not involve initial
methylcyclohexanone formation), as in the mechanism shown in Scheme 1d. We therefore hypothesize that there are
two mechanisms responsible for MCH formation on PtMo/Al O catalysts. Because the catalyst still contains active Pt
2
3
hydrogenation sites and acid sites on the support, the sequential hydrogenation/dehydration mechanism is still possible.
In addition, however, we propose that the presence of Mo sites enables an additional pathway that does not require
initial saturated ketone formation.
We conducted various control experiments to test this hypothesis. For example, we tested the Pt Mo /C catalyst in a
1
1
separate flow reactor optimized for measuring kinetics at low conversion. Note that this low-conversion reactor operated
at much lower total pressure and reactant concentration, leading to quantitative differences in activity and selectivity;
however, as discussed below, the trends in selectivity were preserved. At a conversion of less than 2%, the MCH
selectivity was measured to be 22%. This result suggests that a portion of the deoxygenation activity originates from a
reaction that is parallel to the production of methylcyclohexanone. However, note from Figure 3 that selectivity does
improve with conversion, even in the absence of an active support. This may indicate a series component to the
production of MCH on the bimetallic, or alternatively to kinetics for competing reactions that vary with gas composition.
We also measured the rates of 3-methylcyclohexanol dehydration over the three catalysts (Table 1) to focus on the
dehydration step. In contrast to the results with m-cresol, there was little change in the deoxygenation rate. This suggests
that MCH production from 3-methylcyclohexanol occurs mainly via reaction on acidic sites on the alumina, which are
assumed to have approximately equal abundance on all the catalysts due to the similar surface areas of each metal (Table
2
). Thus, Mo appears to affect a different step in the overall reaction mechanism for cresol.
Addition of Mo to the Pt catalysts affected not only the steady state catalyst performance, but also the behavior as a
function of time on stream. Figure S3 shows time on stream data for low conversion of m-cresol over each catalyst. Note
that with increased Mo content, a longer time was required to achieve a nearly steady reaction rate. The evolution of the
PtMo structure with exposure to reaction conditions appears to be complex, and is the subject of ongoing investigations
in our laboratories.
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3: Selectivities at 523 K and 0.5 MPa for MCH production from m-cresol HDO over carbon supported Pt and Pt Mo .
1
1
Table 1: Rate of MCH production from the reaction of 12:1 H : 3-methylcyclohexanol at 523 K, 0.1 MPa
2
-
1
-1 a
Catalyst
Pt/Al O
rate (μmol alkane s g cat )
37.2± 0.4
Conversion (%)
20
25
17
2
3
Pt Mo /Al O
3
38.0 ± 0.4
10
1
2
Pt Mo /Al O
3
37.9 ± 0.4
1
1
2
a
Error ranges indicate the standard deviation in 3-5 repeated measurements of product concentration
Reaction network
DFT calculations were performed on models of the hydrogen covered Pt and PtMo surfaces (Figure 1) to evaluate
mechanisms by which PtMo catalyzes m-cresol deoxygenation. Mechanisms that have previously been proposed for HDO
over a variety of catalysts (Scheme 1) were investigated to find the lowest barrier pathway over the surfaces studied here.
The models used in these studies are intended to elucidate trends in the reactivity rather than to be exact representations
of the experimental catalysts. There are several possible mechanisms of m-cresol deoxygenation over the Pt surface, as
shown in Scheme 1. The two pathways studied here (ring hydrogenation and tautomerization followed by deoxygenation)
were investigated to determine if the inclusion of Mo in the Pt(111) surface could alter the deoxygenation pathway. The
reaction network including these two mechanisms along with calculated reaction energies and barriers are presented in
61
Figure 4. The calculations performed over Pt(111) without Mo inclusion are also reported in a previous study. The
computed ring hydrogenation mechanism is shown in Figure 4A.
8
ACS Paragon Plus Environment

Page 9 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
A.
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
B.
Pla num
Molybdenum
Oxygen
Carbon
Hydrogen
Figure 4. The reaction network calculated for direct ring hydrogenation on H/Pt(111) shown in pathway A and
tautomerization shown over both monometallic and bimetallic surfaces in pathway B. All values are in eV. In pathway B
the values in italics are reaction energies and barriers over the Mo site in the H/Pt(111) surface and the values in plain
text are the corresponding value on an unmodified H/Pt(111) surface. Important minima are represented by ball and stick
models with the designation corresponding to the reaction networks. Reaction barriers are denoted in brackets with a ‡
and other reported values are reaction energies.
The ring hydrogenation pathway was examined starting from a surface covered with a complete hydrogen monolayer.
The first step in the pathway is hydrogenation of the aromatic carbon-carbon bond, shown in step R to RH , which has a
barrier of 0.48 eV. This step is endothermic by 0.47 eV and occurs as a simultaneous transfer of two H atoms. This is in
contrast to the hydrogenation occurring as two single hydrogen transfer steps as has been shown over the surface
1
1
1
9,33,62-64
without hydrogen coverage.
The change in reactivity with hydrogen coverage has been previously observed for
6
5
furfural reacting on the Pd(111) surface. In this work, the structure resulting from a single hydrogen transfer was not a
minimum on the potential energy surface. Geometry optimization starting from the single hydrogen-transfer structure
results in spontaneous transfer of the second hydrogen to form the final product, or else the optimization results in the
reactant. The next step in the reaction is the conversion from RH to RH , which is the highest energy point in the reaction
pathway. The following step is a keto-enol tautomerization to form TH , which can subsequently be hydrogenated to form
the alcohol. Direct removal of the hydroxyl group from structure TH is significantly endothermic by 1.14 and has a barrier
of 1.23 eV. As the energies of TH and TH are equivalent, a distribution of these species would be predicted as desorption
products. While deoxygenation is unlikely via pathway A on the Pt(111) surface, deoxygenation over the acidic alumina
support to form the aromatic product labeled P is most likely the major route for deoxygenation. This may then be
followed by hydrogenation to P to produce the saturated alkane. Figure 4A only shows direct ring hydrogenation for
1
2
3
4
3
4
1
3
Pt(111); adsorption of H atoms on Mo sites was found to be unfavorable as these sites strongly prefer to interact with the
oxygen atom. Therefore, direct ring hydrogenation on PtMo is expected to occur over Pt sites in a similar fashion to that
on the unmodified H/Pt(111) surface, as shown in Figure 4A.
21
The mechanism shown in Figure 4B, first proposed by Nie and Resasco , begins with tautomerization of the m-cresol into
the keto-intermediate followed by deoxygenation. Ring hydrogenation is also possible but is considered elsewhere. In
this mechanism, the reaction is a concerted H-transfer, as shown in Figure 5. One H transfers from the surface to the m-
cresol ring and the second H transfers from the m-cresol hydroxyl group to the surface.
61
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. The geometry of the transition state with bond distances indicated for the tautomerization step R ’–T over the
Pt(111) surface A. and the Pt(111) surface with Mo substitution B.
1
1
The energetics for this tautomerization reaction, shown by the step from R ‘ to T in Figure 4B, indicate a significant
1
1
difference between the Pt and PtMo surfaces. The difference of 0.26 eV for the barrier and 0.49 eV for the reaction
energy indicates a significant effect of the Mo on this reaction step. This energetic change is due to the stronger O-Mo
interaction in the transition and final states. In the transition state structure for Pt(111) in Figure 5A, the oxygen does not
directly interact with the surface, while in the transition state for the bimetallic surface (Figure 5B), the oxygen directly
interacts with the surface Mo. The Mo inclusion significantly reduces the oxygen-surface distance from 3.05 Å to 2.14 Å.
Thus if Mo is present, surface-assisted tautomerization would become more favorable. This additional deoxygenation
pathway agrees with experimental results showing increased deoxygenation rates over bimetallic catalysts (Figure S3). At
less than 5% conversion, the steady state rates of alkane production from m-cresol HDO over Pt/Al O , Pt Mo /Al O , and
2
3
10
1
2
3
-6
-5
-5
-1
-2
Pt Mo /Al O were approximately 8 x 10 , 1.5 x 10 , and 3.5 x 10 mol-min -m catalyst respectively.
1
1
2
3
After tautomerization, deoxygenation occurs in step T to P , breaking the aryl–O bond. The barrier to this step is
2
1
significantly lower (0.70 eV vs. 1.27 eV) with Mo included in the Pt(111) lattice. Toluene was not observed as a product;
however, thermodynamic equilibrium under the conditions in this study significantly favors MCH. The equilibrium ratio of
MCH to toluene was calculated to be 320. This is consistent with previous studies of aromatic hydrogenation, which has
been shown to be fast over Pt catalysts at comparable temperatures. For example, toluene hydrogenation over 10 wt%
-
1 66
Pt/Al O at 573 K had a TOF of 139.8 s . Previous reports of toluene production from m-cresol HDO at high H :m-cresol
2
3
2
19,21
ratios were for reactions performed at 573 K.
This reaction temperature increase of 50 K dramatically impacts the
calculated thermodynamic equilibrium ratio of MCH to toluene, lowering it from 320 to 4. If the overall mechanism is
considered, the minimum energy pathway is ring hydrogenation over the Pt(111) surface with an additional mechanism of
tautomerization followed by deoxygenation possible over the PtMo surface. This agrees with the experimental results in
Figure 2, which suggest that the ketone is a primary product over both Pt and PtMo surfaces due to the lowest energy
pathways beginning with a hydrogenation or hydrogen transfer step.
The final step probed by DFT calculations is the formation of water from a surface-adsorbed hydroxyl group. As Mo sites
have stronger interactions with oxygen, it is possible that water formation could become rate limiting. To evaluate this
hypothesis, the barrier to water formation was calculated over the Pt(111) and PtMo surfaces. Over the PtMo surface, the
barrier to water formation was 0.35 eV and was exothermic by 0.31 eV, while the barrier for the Pt surface was 0.11 eV
and was exothermic by 0.78 eV. Based on these results, this reaction step is not rate limiting.
Catalyst Characterization
To correlate catalyst performance with structure, including the simulated structures in the previous section,
characterization efforts were focused on determining the state and distribution of the Mo on the bimetallic catalyst. TPR
was used to characterize carbon- and alumina-supported Pt, Mo, and PtMo samples, with the results shown in Figure 6.
The monometallic Pt/Al O catalyst showed two hydrogen uptake features, one at 490 K and a higher temperature peak at
2
3
650 K. The reduction at 490 K is consistent with previous TPR results for Pt/Al O catalysts, and a smaller feature at higher
2
3
1
0
ACS Paragon Plus Environment

Page 11 of 21
ACS Catalysis
1
15,67,68
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
temperature is also frequently reported.
This latter feature is often attributed to a strong interaction with the
support, but may also be the reduction of under-coordinated sites such as steps and edges on some of the particles. TPR
performed on Pt/C catalysts also demonstrated more than one reduction feature for catalysts with low weight loadings
6
9,70
(
<1 wt%).
Previous studies reporting a series of TPR profiles of Pt/Al O catalysts with different metal loadings showed
2 3
6
8
peaks around 490 K and 650 K with lower wt% Pt. For the lowest metal loadings, these two peaks were comparable in
size. However, as the metal loading was increased (i.e., decreasing the relative concentration of defect sites), the lower
temperature peak dominated. At the highest metal loadings, the high temperature reduction feature was no longer
observed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6: TPR profiles for Pt, Mo, and PtMo catalysts. Samples are supported on alumina unless otherwise noted. All
signals are normalized to mass of catalyst loaded and mass spectrometer pressure.
In the present study, we observed that as Mo was added to the Pt catalyst, the high-temperature TPR feature
disappeared. This peak was not detected for Mo loadings with a Pt:Mo atomic ratio less than 50:1. Thus, addition of Mo
(which forms strong metal-oxygen bonds) to Pt actually decreases the fraction of sites that require high temperature to
reduce. This may be attributable to Mo increasing the average particle size of the catalyst, thus decreasing the fraction of
small particles with large concentrations of step and edge sites. Addition of Mo has previously been shown to decrease
71
the dispersion of supported Pt catalysts. TEM images (Figure S4) show slightly larger average particle sizes for the PtMo
bimetallic catalysts. Note that monometallic Mo/Al O exhibited reduction features above 800 K that were not observed
2
3
in the PtMo catalysts. Mo sites on the PtMo catalysts are likely reduced through a spillover effect, where Pt is responsible
for activating H that then migrates to Mo sites. This spillover effect has often been suggested for bimetallic catalysts
2
7
2
containing a hydrogenation metal, including previous work with PtMo. The Pt Mo and Pt Mo bimetallic catalysts used
10
1
1
1
here also did not show reduction peaks around 850-870 K as seen on monometallic Mo, suggesting there is sufficient
mixing of the two metals to prevent large domains that act as bulk Mo or Mo oxides. This is expected to be advantageous
for catalytic activity because previous surface science investigations of Mo-modified Pt(111) have shown that reduced Mo
43
has a significant effect on surface reactivity, whereas oxidized Mo mainly serves to block active sites.
CO chemisorption was used to estimate the active surface area of each catalyst. The results can be seen in Table 2. As the
Mo loading increased, the measured surface area of the catalyst decreased. Previous CO-TPD studies on pure alumina
show desorption temperatures less than 200 K, so that CO adsorption is expected to measure only the exposed metal
4
0
sites. While our previous work with PtMo surfaces under ultra-high vacuum (UHV) conditions showed decreased CO
4
3
adsorption energy on reduced PtMo bimetallics, the expected desorption temperatures were still higher than 425 K.
This suggests the decrease in CO uptake was due to a decreased active surface area. This could be due to the increased
1
6,73,74
total metal loading as catalysts with higher weight percent metal tend to result in larger particle sizes.
These TPR
and CO chemisorption results suggest that the PtMo catalysts had fewer exposed active sites than the monometallic Pt
sample.
1
1
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2: TEM estimated particle sizes and CO chemisorption results for estimating dispersion and surface area of Pt and
PtMo catalysts, reported with a 95% confidence interval
Catalyst
Weight % Pt Weight
%
Mo Dispersion (%)
CO uptake
(μmol CO/ g area
catalyst)
125 ± 6
Active surface Particle
a
a
(from ICP)
(from ICP)
diameter
(nm)
2
b
m /g catalyst
Pt/Al O
5.77
5.02
5.07
-
48.7 ± 2.5
33.5 ± 1.3
16.5 ± 0.7
6.0 ± 0.3
4.5 ± 0.2
3.9 ± 0.1
1.25
±
±
±
2
3
0
.05
1.31
.06
1.60
.06
Pt Mo /Al O
3
0.300
1.77
95 ± 4
82 ± 6
10
1
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
Pt Mo /Al O
1
1
2
3
0
a
b
9
5% confidence interval from replicate ICP results indicated an error less than 0.01%
Measured from TEM images. 95% confidence intervals obtained from sampling >120 particles.
X-ray absorption spectroscopy (XAS) was performed to further investigate the interaction between Pt and Mo in these
catalysts. Experiments were performed on Pt/Al O , Pt Mo /Al O , and Pt Mo /Al O . The Pt L -edge XANES spectra and
2
3
10
1
2
3
1
1
2
3
3
the Fourier-Transform of the Pt L -edge EXAFS spectra can be seen in Figure 7. Reference spectra can be found in
3
Supporting Information (Figure S5).
Figure 7: Pt L -edge XANES spectra of Pt and PtMo catalysts (A) and magnitude of the Fourier-Transform of the Pt L -edge
3
3
EXAFS spectra of Pt and PtMo catalysts (B). All spectra taken at room temperature after pretreatment at 723 K in H2.
The shape and height of the white line of the XANES at the Pt edge is similar to Pt foil, indicating that all catalysts
contained metallic Pt after reduction. Interestingly, the Pt Mo white line intensity was similar to the Pt foil, while
1
1
Pt Mo was similar to Pt/Al O . This is potentially due to the larger particle size of Pt Mo ; the effect of particle size on
1
0
10
2
3
1
1
75
white line intensity was previously modeled using the assumption that charge transfer was negligible. Though alloying
could also affect the d-band density, which in turn affects the XAS white line intensity, this effect may be negligible in this
case. The Fourier-Transform of the Pt L -edge EXAFS spectra is also shown in Figure 7. Note the much higher amplitude of
3
the foil (with 12 well-ordered Pt first neighbors and higher shell Pt neighbors). The Pt Mo /Al O sample had larger higher
1
1
2
3
1
2
ACS Paragon Plus Environment

Page 13 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
shells than the Pt Mo /Al O sample, suggesting a catalyst with larger, more well-ordered metallic particles. A summary
10 1 2 3
of the Pt L -edge XANES and EXAFS from R-space fitting can be found in Table 3.
3
2
-1
Table 3. Summary of Pt L -edge XANES and EXAFS results from R-space fitting. Fitting range; k : Δk = 2.7-12.1 Å ; ΔR =
3
2
1
.70-3.10 Å. CN = coordination number; R = bond distance; Δσ = mean-square disorder in the distribution of interatomic
distances; ΔE = energy offset. Samples were pre-treated in flowing H at 723 K and measured in H at RT. The estimated
0
2
2
7
6
errors are: CN, ± 1 and R, ± 0.02 Å, within those typical for EXAFS fitting.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
Scattering
path
Δσ
ΔE0
(eV)
CN
R (Å)
2
Sample
Pt foil
Edge energy (eV)
11564.1
(Å )
Pt-Pt
Pt-Pt
Pt-Pt
Pt-Pt
Pt-Mo
12.0
7.4
6.7
8.0
0.8
2.77
2.71
2.70
2.75
2.73
N/A
N/A
–4.5
–4.0
–1.2
10.8
Pt/Al O3
11564.3
0.005
0.005
0.002
0.002
2
Pt Mo /Al O
3
11564.4
10
1
2
Pt Mo /Al O
3
11564.3
1
1
2
The Pt edge EXAFS presented in Figure 7 shows significant differences between Pt and PtMo samples. The low Pt-Pt
coordination numbers found in the monometallic platinum and the Pt Mo catalysts (7.4 and 6.7 respectively, as
10
1
compared to 12 for a Pt foil) indicates that very small Pt particles are present. This is further supported by the contraction
of the bond distances from 2.77 Å in the foil to only 2.71 Å for monometallic Pt and 2.70 Å for Pt Mo . Using previously
10
1
77
established correlations between dispersion and coordination number , the estimated particle size for these two samples
is 1-2 nm. The Pt Mo catalyst is estimated to have slightly larger particles, approximately 3 nm, due to the higher
1
1
coordination number (8.0) and longer Pt-Pt bond distances (2.75 Å). This is in agreement with the trend shown from CO
chemisorption results reported in Table 2, which also indicated slightly larger particle sizes with increasing Mo loading. It
is also consistent with the particle sizes measured using TEM (see Supporting Information Figure S4). These particle size
trends from the EXAFS results further support the hypothesis that increased particle size could contribute to the
disappearance of the high temperature reduction feature on bimetallic catalysts as seen in TPR. The relatively high Pt-Pt
coordination number and low Pt-Mo coordination number indicates that there may be some Mo present as an oxide on
78,79
the surface and/or that there may be a higher concentration of Mo near the surface of metal nanoparticles.
However,
it should also be noted that uncertainties in the coordination numbers limit the detail with which a precise bimetallic
structure can be identified.
Only Pt-Pt scattering could be fit on the Pt Mo sample due to the low concentration of Mo. However, the Pt Mo had a
10
1
1
1
stronger signal intensity and Pt-Mo scattering was also detected, indicating formation of bimetallic particles. The Pt-Pt
scattering was 10 times larger than Pt-Mo, indicating the particles were Pt rich or that Mo segregated within the particles.
Since there were only small Pt-Mo contributions, the higher shells were fairly unchanged relative to the foil due to the
small perturbation to the structure.
The Mo K-edge was also studied and analyzed; the Mo K-edge XANES spectra and Fourier Transform of the EXAFS data
can be seen in Figure 8. Reference spectra can be found in Supporting Information (Figure S5). No Mo-Mo scattering was
observed (Table 4), indicating that Mo is most likely not forming aggregates within the bimetallics. These nanoparticles
are expected to be a kinetically trapped bimetallic structure rather than a true bulk alloy. PtMo alloys do exist, but require
8
0
annealing to over 1073 K.
1
3
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Mo K-edge XANES spectra of PtMo catalysts (left) and magnitude of the Fourier-Transform (non-phase shift
corrected) of the Mo K-edge EXAFS spectra of PtMo catalysts collected at room temperature after pretreatment at 723 K
in H2.
The XANES spectra in Figure 8 suggest that both Pt Mo /Al O and Pt Mo /Al O contain a mixture of reduced and
10
1
2
3
1
1
2
3
oxidized Mo. The Fourier-Transform of the Mo K-edge EXAFS spectra shows a much larger Mo-Pt correlation in the
Pt Mo /Al O sample compared to the Pt Mo /Al O sample. A summary of the XANES and EXAFS results at the Mo K-
10
1
2
3
1
1
2
3
edge can be found in Table 4. For comparison, the edge energies for Mo foil, MoO and MoO are 20000.0, 20012.2, and
2
3
2
0015.8 eV respectively.
2
-1
Table 4. Summary of Mo K-edge XANES and EXAFS results from R-space fitting. Fitting range; k : Δk = 3.0-10.7 Å ; ΔR =
2
1
.40-3.10 Å. CN = coordination number; R = bond distance; Δσ = mean-square disorder in the distribution of interatomic
distances; ΔE = energy offset. Samples were pre-treated in flowing H at 723 K and measured in H at room temperature.
0
2
2
76
The estimated errors are: CN, ± 1 and R, ± 0.02 Å, within those typical for EXAFS fitting.
Edge
(eV)
energy
2
2
CN
R (Å)
Δσ (Å )
ΔE (eV)
0
Sample
Scattering path
Mo-O
Na2MoO3∙2H2O
20016.0
4
1.77
2.10
2.72
2.06
2.73
N/A
N/A
9.5
2.1
3.6
0.2
Mo-O
0.9
4.7
1.5
1.3
0.001
0.003
0.001
0.003
Pt Mo /Al O
3
20001.8
10
1
2
Mo-Pt
Mo-O
Pt Mo /Al O
3
20003.0
1
1
2
Mo-Pt
Mo–O bonds were observed in the bimetallic samples, indicating some of the Mo present is oxidized. The poorly defined
+
6
feature at 20,006 eV in both PtMo catalysts suggest there is Mo in the catalysts. A linear combination fitting of the Mo
81
K-edge XANES conducted with the Demeter suite of software using metallic Mo, MoO , and MoO reference compounds
2
3
show that both catalysts are best fit by a combination of metallic Mo and MoO . However, because of the large
3
1
4
ACS Paragon Plus Environment

Page 15 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
uncertainty in the fitting results, it was not possible to quantify the relative concentrations (percentages) of each
oxidation state with the data. From the edge position, we can infer that the metallic Mo fraction is higher in the
Pt Mo /Al O sample than the Pt Mo /Al O sample. This qualitative conclusion of higher metallic Mo fraction is
10
1
2
3
1
1
2
3
consistent with the EXAFS finding of larger Mo-Pt correlation in the Pt Mo /Al O sample, where each Mo had 4.7 Pt
10
1
2
3
neighbors (compared to 1.3 in the Pt Mo / Al O sample). As the Mo content increased, the fraction of Mo in MoO
3
1
1
2
3
increased. However, considering the much larger (10x) number of total Mo atoms in Pt Mo /Al O , the total number of
1
1
2
3
metallic Mo sites was higher in this catalyst as well.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6+
In summary, the XAS results are consistent with Mo existing in both metallic and Mo oxidation states. These results are
consistent with the presence of two types of Mo on the Al O surface: Mo–Pt bimetallic nanoparticles and MoO clusters.
2
3
3
This interpretation of the XAS data is consistent with a previous study of a carbon supported PtMo catalyst used for
8
2
aqueous phase glycerol reforming. The lack of Mo–Mo correlation indicates Mo was not segregated or forming the core
of the bimetallic particles; rather, it was dispersed in the Pt-rich particles, hypothetically contributing as an oxophilic
metal site surrounded by Pt. Existing phase diagrams for PtMo systems show formation of a single-phase alloy at very high
80
83
temperatures (800-2000°C) , while below 800°C a thermodynamic alloy does not exist. The catalysts used in this study
were not annealed above 450°C. The absence of such bulk alloys at low temperatures suggests relatively unfavorable
interactions between Pt and Mo. Despite this, the results of our characterization work suggest the two metals are in close
enough contact to have a synergistic effect. While this interaction may cause electronic effects on the Pt, our DFT results
indicate that it is most likely the proximity of these two distinct sites that results in the altered reaction pathway. The d-
band effects on the Pt sites that may cause the aromatic ring to bind less strongly still do not explain the promotion of a
15
tautomerization pathway; other studies have suggested distinct sites are responsible for tautomerization.
Discussion
M-cresol deoxygenation on Pt has been previously studied on supported catalysts, with products and proposed
1
5-17,21,61,84,85
mechanisms depending on the support material and reaction conditions.
affinity for hydrogenation, it is generally accepted that under HDO conditions, the aromatic ring must be saturated before
Over metals such as Pt with a high
16,20
C–O scission can occur.
A study of phenol deoxygenation over Pt/Al O , at hydrogen pressure (0.3 MPa total, 3:1
2
3
86
H :phenol) comparable to the present study, reported production of primarily cyclohexanone and cyclohexanol.
Previous calculations and microkinetic modeling of guaiacol deoxygenation over the Pt(111) surface by Lu et al.
2
64
indicated that direct deoxygenation without initial ring hydrogenation is unfavorable. They concluded that reaction of
catechol and phenol over the Pt catalysts must occur at corner, edge, or step sites over the Pt(111) surface. Other
87
previous studies have also concluded that deoxygenation over noble metal catalysts may occur at step sites. Our results
on Pt/Al O and Pt/C support the idea of bifunctional catalysis, with hydrogenation occurring on Pt and deoxygenation on
2
3
the support. The small amounts of deoxygenation products observed on Pt/C may be due to reaction on
undercoordinated Pt sites.
The primary focus of this work is to elucidate the mechanism by which Mo can improve the deoxygenation activity and
selectivity of supported Pt catalysts. Characterization results indicate a complex structure for the catalyst: in particular,
both reduced Mo (present as a bimetallic with Pt) and oxidized Mo were observed to be present on the surface after in
situ reduction. Either or both of these components may contribute to the observation of improved deoxygenation rates in
88-90
Figures 2 and 3. One possibility is that MoO species adds additional Brønsted acid sites.
Previous studies indicate that
x
31
oxidized modifiers like Mo and Re may introduce a possible route for dehydration. In this study, the results in Table 1
demonstrate that addition of Mo did not affect the rate of 3-methylcyclohexanol dehydration, such that essentially all the
dehydration activity could be attributed to the Al O support. Therefore, it is unlikely that the deoxygenation observed in
2
3
Figure 3 for the PtMo/C catalyst results from Brønsted acid sites on MoO that function in analogy to the support, but
x
instead is the result of an additional pathway over the bimetallic surface. The differing trends in selectivity as a function of
conversion also support this idea. However, some role for MoO species in promoting deoxygenation cannot be ruled out,
x
and we note that chemistry over MoO -containing surfaces was not considered with DFT models in this study.
x
The DFT calculations reported above support the hypothesis that PtMo bimetallic surfaces can provide an additional route
to deoxygenation. The oxophilicity of Mo may suggest a scheme in which Mo directly extracts an O atom, but here we
report the effect appears to be more complex. First, Mo favors geometries of aromatic surface species with a more
upright orientation through favorable Mo-O interactions at the surface (Figures 4 and 5). This upright orientation is
associated with lower barriers for the H transfer reactions required for tautomerization. The barrier to the subsequent C-
O scission reaction is also reduced, which in this case can be attributed to the oxophilicity of Mo, which stabilizes
adsorbed OH in the final state. The Mo in the bimetallic surface does not necessarily supply a site that is widely active for
1
5
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
deoxygenation, regardless of the reactant: calculation of 3-methylcyclohexanol deoxygenation over the PtMo surface was
endothermic by 0.85 eV, supporting the hypothesis that direct C–O scission of the alcohol is unlikely, even on the
bimetallic catalyst surface. Similar observations of direct deoxygenation have been seen recently on SiO supported
2
materials. On Pt/SiO , the primary product formed was 3-methylcyclohexanone, resulting from a hydrogenation pathway;
2
1
9
meanwhile, over a more oxophilic Ru/SiO catalyst, HDO became more prevalent. We note that the highest energy
2
barrier to tautomerization on the PtMo surface was calculated to be similar to the barrier for hydrogenation on the Pt
surface (Figure 4), consistent with the overall similar conversion rates of cresol (but with very different selectivities)
observed experimentally for the two catalysts.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Thus, the results presented here suggest that use of noble metal – oxophilic metal combinations can influence selectivity
in reactions of oxygenates in unexpected ways. That is, the strong interaction between Mo and the oxygenated functional
group can stabilize adsorbate geometries that favor elementary steps that involve C-OH dissociation. An orientation effect
altering the barrier for tautomerization has been observed previously for differences in phenol interactions between Pt
63
20
and Pd at low hydrogen pressures. Calculations by Nie et al. for NiFe catalysts demonstrated that the Fe component
strengthened interaction of alcohol functional groups with the surface, changing the adsorbate configuration; this
observations was connected with increased deoxygenation selectivity on Fe-containing catalysts.
33
A previous examination of tautomerization over Fe(110) was performed by Hensley et al. This work found barriers of
2.03 and 2.43 eV depending on orientation relative to the surface, indicating that tautomerization was not a likely
pathway. In both reactions, the tautomerization did not include intramolecular hydrogen transfer to the surface. Yoon et
9
1
33
al. also explored the tautomerization mechanism with barriers comparable to those of Hensley et al. for reactions in
19
ultra-high vacuum conditions. Work by Tan et al. showed that tautomerization may occur over the bare Pt(111) surface,
though the barrier for subsequent deoxygenation is higher in this mechanism than if surface hydrogen is included. Thus
the participation of the surface in tautomerization and the inclusion of hydrogen are both important to understanding the
mechanism of deoxygenation.
Conclusions
Addition of Mo to supported Pt catalysts increased selectivity to deoxygenated products by creating sites that introduce a
new preferred reaction pathway by changing the kinetic barriers. M-cresol was proposed to follow a ring hydrogenation
pathway over Pt, but a tautomerization-deoxygenation pathway over PtMo catalysts, due to oxophilic sites that increase
the interaction of the surface with oxygen-containing functional groups. DFT calculations suggest a bifunctional PtMo
surface, which contains sites for hydrogen activation and sites for facile tautomerization of m-cresol creates a lower
energy barrier for deoxygenation. This lowered barrier may render deoxygenation reactions on the metal surface
kinetically relevant, whereas a lack of bifunctionality (Pt) requires a separate acid function (from the catalyst support) for
C—O bond cleavage.
Acknowledgements
This work was supported by the Department of Energy BioEnergy Technologies Office under Contract no. DE-AC36-08-
GO28308. Computer time was provided by the Texas Advanced Computing Center under the National Science Foundation
Extreme Science and Engineering Discovery Environment Grant MCB-090159 and by the National Renewable Energy
Laboratory Computational Sciences Center. A.M.R. acknowledges support from the National Science Foundation for
funding this research (Award CHE-1149752) and partial support from the Department of Education Graduate
Assistantships in Areas of National Need (GAANN). The authors gratefully acknowledge Dr. Susan Habas for TEM imaging.
X-ray absorbance spectroscopy experiments were conducted at the Sector 10 beamline, which is operated by the
Materials Research Collaborative Access Team. This research used resources of the Advanced Photon Source, a DOE Office
of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-
AC02-06CH11357. We gratefully acknowledge Dr. Jeffrey T. Miller (Purdue University) for his help and insight related to
XAS data analysis. We also thank Dr. Vassili Vorotnikov for insightful discussions concerning the reaction pathways.
Supporting Information Available: Details on calculations to determine the influence of mass transfer effects, selectivities
to methylcyclohexene and 3-methylcyclohexanone from reacting 3-methylcyclohexanol over an Al O blank as a function
2
3
of conversion, TEM images of the catalysts used in this study, and Pt L -edge XANES spectra of Pt references and Mo K-
3
edge XANES spectra of Mo references in different oxidation states.
1
6
ACS Paragon Plus Environment

Page 17 of 21
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
This
material
is
available
free
of
charge
via
the
Internet
at
http://pubs.acs.org.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table of Contents Graphic
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
References
(
1)
2)
Ruddy, D. A.; Schaidle, J. A.; Ferrell III, J. R.; Wang, J.; Moens, L.; Hensley, J. E. Green Chem.
014, 16, 454-490.
2
(
Dutta, A. S., A.; Tan, E.; Humbird, D.; Snowden-Swan, L. J.; Meyer, P.; Ross, J.; Sexton, D.;
Yap, R.; Lukas, J. “Process Design and Economics for the Conversion of Lignocellulosic
Biomass to Hydrocarbon Fuels,” National Renewable Energy Laboratory, 2015.
Dutta, A.; Schaidle, J. A.; Humbird, D.; Baddour, F. G.; Sahir, A. Top. Catal. 2016, 59, 2-18.
Bridgwater, A. V. Biomass Bioenergy 2012, 38, 68-94.
(3)
(4)
(5)
(6)
Sipilä, K.; Kuoppala, E.; Fagernäs, L.; Oasmaa, A. Biomass Bioenergy 1998, 14, 103-113.
Pham, T. N.; Shi, D.; Resasco, D. E. Appl. Catal., B 2014, 145, 10-23.
(7)
Greenhalf, C.; Nowakowski, D.; Harms, A.; Titiloye, J.; Bridgwater, A. Fuel 2013, 108, 216-
2
30.
(
8)
9)
Ingram, L.; Mohan, D.; Bricka, M.; Steele, P.; Strobel, D.; Crocker, D.; Mitchell, B.;
Mohammad, J.; Cantrell, K.; Pittman Jr, C. U. Energy Fuels 2007, 22, 614-625.
Furimsky, E. Appl. Catal., A 2000, 199, 147-190.
(
(10) Lu, Q.; Li, W.-Z.; Zhu, X.-F. Energy Convers. Manage. 2009, 50, 1376-1383.
(
11) Bu, Q.; Lei, H.; Zacher, A. H.; Wang, L.; Ren, S.; Liang, J.; Wei, Y.; Liu, Y.; Tang, J.; Zhang, Q.
Bioresour. Technol. 2012, 124, 470-477.
(
(
12) Lin, Y.-C.; Li, C.-L.; Wan, H.-P.; Lee, H.-T.; Liu, C.-F. Energy Fuels 2011, 25, 890-896.
13) Bui, V. N.; Laurenti, D.; Afanasiev, P.; Geantet, C. Appl. Catal., B 2011, 101, 239-245.
(14) Hong, Y.; Zhang, H.; Sun, J.; Ayman, K. M.; Hensley, A. J.; Gu, M.; Engelhard, M. H.; McEwen,
J.-S.; Wang, Y. ACS Catal. 2014, 4, 3335-3345.
15) Do, P. T.; Foster, A. J.; Chen, J.; Lobo, R. F. Green Chem. 2012, 14, 1388-1397.
16) Foster, A. J.; Do, P. T.; Lobo, R. F. Top. Catal. 2012, 55, 118-128.
(
(
(17) Zanuttini, M.; Lago, C.; Querini, C.; Peralta, M. Catal. Today 2013, 213, 9-17.
(18) Sun, J.; Karim, A. M.; Zhang, H.; Kovarik, L.; Li, X. S.; Hensley, A. J.; McEwen, J.-S.; Wang, Y. J.
Catal. 2013, 306, 47-57.
(19) Tan, Q.; Wang, G.; Nie, L.; Dinse, A.; Buda, C.; Shabaker, J.; Resasco, D. E. ACS Catal. 2015, 5,
6271-6283.
(20) Nie, L.; de Souza, P. M.; Noronha, F. B.; An, W.; Sooknoi, T.; Resasco, D. E. J. Mol. Catal. A:
Chem. 2014, 388, 47-55.
(21) Nie, L.; Resasco, D. E. J. Catal. 2014, 317, 22-29.
1
8
ACS Paragon Plus Environment

Page 19 of 21
ACS Catalysis
1
2
(
22) Zhao, C.; Kou, Y.; Lemonidou, A. A.; Li, X.; Lercher, J. A. Angew. Chem. 2009, 121, 4047-
050.
23) Fu, J.; Hakim, S. H.; Shanks, B. H. Top. Catal. 2012, 55, 140-147.
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
(
(24) Yang, J.; Williams, C. L.; Ramasubramaniam, A.; Dauenhauer, P. J. Green Chem. 2014, 16,
675-682.
(
25) Simonetti, D.; Kunkes, E.; Dumesic, J. J. Catal. 2007, 247, 298-306.
(26) Shimao, A.; Koso, S.; Ueda, N.; Shinmi, Y.; Furikado, I.; Tomishige, K. Chem. Lett. 2009, 38,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
540-541.
(27) Daniel, O. M.; DeLaRiva, A.; Kunkes, E. L.; Datye, A. K.; Dumesic, J. A.; Davis, R. J.
ChemCatChem 2010, 2, 1107-1114.
28) Zhang, L.; Karim, A. M.; Engelhard, M. H.; Wei, Z.; King, D. L.; Wang, Y. J. Catal. 2012, 287,
(
(
(
(
37-43.
29) Dietrich, P. J.; Wu, T.; Sumer, A.; Dumesic, J. A.; Jellinek, J.; Delgass, W. N.; Ribeiro, F. H.;
Miller, J. T. Top. Catal. 2013, 56, 1814-1828.
30) Kunkes, E. L.; Simonetti, D. A.; Dumesic, J. A.; Pyrz, W. D.; Murillo, L. E.; Chen, J. G.; Buttrey,
D. J. J. Catal. 2008, 260, 164-177.
31) Chia, M.; Pagán-Torres, Y. J.; Hibbitts, D.; Tan, Q.; Pham, H. N.; Datye, A. K.; Neurock, M.;
Davis, R. J.; Dumesic, J. A. J. Am. Chem. Soc. 2011, 133, 12675-12689.
(32) Greeley, J.; Mavrikakis, M. Catal. Today 2006, 111, 52-58.
(33) Hensley, A. J. R.; Wang, Y.; McEwen, J.-S. ACS Catal. 2015, 5, 523-536.
(
34) Wang, T.; Wang, S.; Luo, Q.; Li, Y.-W.; Wang, J.; Beller, M.; Jiao, H. J. Phys. Chem. C 2014, 118,
4181-4188.
(
35) Pang, S. H.; Román, A. M.; Medlin, J. W. J. Phys. Chem. C 2012, 116, 13654-13660.
(36) Marshall, S. T.; Medlin, J. W. Surf. Sci. Rep. 2011, 66, 173-184.
(37) Cremer, P. S.; Su, X.; Shen, Y. R.; Somorjai, G. A. J. Am. Chem. Soc. 1996, 118, 2942-2949.
(38) Neurock, M.; Pallassana, V.; van Santen, R. A. J. Am. Chem. Soc. 2000, 122, 1150-1153.
(39) Burkholder, L.; Stacchiola, D.; Tysoe, W. Surf. Rev. Lett. 2003, 10, 909-916.
(40) Diaz, A. L.; Bussell, M. E. J. Phys. Chem. 1993, 97, 470-477.
(41) Ertl, G.; Neumann, M.; Streit, K. Surf. Sci. 1977, 64, 393-410.
(
42) Ko, E.; Madix, R. Surf. Sci. 1981, 109, 221-238.
(43) Robinson, A. M.; Montemore, M. M.; Tenney, S. A.; Sutter, P.; Medlin, J. W. J. Phys. Chem. C
2013, 117, 26716-26724.
(
44) Ressler, T. J. Synchrot. Radiat. 1998, 5, 118–122.
(45) Rehr, J.; Mustre de Leon, J.; Zabinsky, S.; Albers, R. J. Am. Chem. Soc. 1991, 113, 5135-5140.
(
46) Fogler, H. S. Elements of Chemical Reaction Engineering; 4th ed.; Acrivos, A. D., J.; Fogler,
H.S.; Hanratty, T.J.; Prausnitz, J.M.; Scriven, L.E., Eds.; Prentice Hall: New Jersey, 2005; 813-
866.
(47) Welty, J. R.; Wicks, C. E.; Wilson, R. E.; Rorrer, G. Fundamentals of Momentum, Heat, and
Mass Transfer; 4th ed.; Anderson, W., Eds.; John Wiley & Sons: New York, 2001; 443.
(
(
(
48) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169-11186.
49) Kresse, G.; Furthmuller, J. Comput. Mater. Sci. 1996, 6, 15-50.
50) Kresse, G.; Hafner, J. Phys. Rev. B 1993, 47, 558-561.
(51) Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251-14269.
(52) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758-1775.
(
(
(
53) Blochl, P. E. Phys. Rev. B 1994, 50, 17953-17979.
54) Tkatchenko, A.; DiStasio, R. A., Jr.; Car, R.; Scheffler, M. Phys. Rev. Lett. 2012, 108, 236402.
55) Tkatchenko, A.; Scheffler, M. Phys. Rev. Lett. 2009, 102, 073005.
1
9
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
56) Chizallet, C.; Raybaud, P. Catal. Sci. Technol. 2014, 4, 2797-2813.
(57) Mager-Maury, C.; Bonnard, G.; Chizallet, C.; Sautet, P.; Raybaud, P. ChemCatChem 2011, 3,
00-207.
2
(58) Runnebaum, R. C.; Nimmanwudipong, T.; Block, D. E.; Gates, B. C. Catal. Sci. Technol. 2012,
2, 113-118.
(59) Shin, E.-J.; Keane, M. A. Ind. Eng. Chem. Res. 2000, 39, 883-892.
(60) Hong, D.-Y.; Miller, S. J.; Agrawal, P. K.; Jones, C. W. Chem. Commun. 2010, 46, 1038-1040.
(61) Griffin , M. B.; Ferguson, G. A.; Ruddy , D. A.; Biddy , M. J.; Beckham , G. T.; Schaidle, J. A. ACS
Catal. 2016, 6, 2715–2727.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(62) Hensley, A. J. R.; Hong, Y.; Zhang, R.; Zhang, H.; Sun, J.; Wang, Y.; McEwen, J.-S. ACS Catal.
014, 4, 3381-3392.
2
(
(
63) Li, G.; Han, J.; Wang, H.; Zhu, X.; Ge, Q. ACS Catal. 2015, 5, 2009-2016.
64) Lu, J. M.; Behtash, S.; Mamun, O.; Heyden, A. ACS Catal. 2015, 5, 2423-2435.
(65) Wang, S.; Vorotnikov, V.; Vlachos, D. G. ACS Catal. 2015, 5, 104-112.
(66) Tétényi, P.; Galsán, V. Appl. Catal., A 2002, 229, 181-192.
(67) Tanksale, A.; Beltramini, J. N.; Lu, G. M. Renewable Sustainable Energy Rev. 2010, 14, 166-
182.
(68) Hwang, C.-P.; Yeh, C.-T. J. Mol. Catal. A: Chem. 1996, 112, 295-302.
(69) Lim, D.-H.; Lee, W.-D.; Choi, D.-H.; Lee, H.-I. Appl. Catal., B 2010, 94, 85-96.
(
70) Sepúlveda-Escribano, A.; Coloma, F.; Rodrıguez-Reinoso, F. Appl. Catal., A 1998, 173, 247-
57.
71) Williams, W. D.; Bollmann, L.; Miller, J. T.; Delgass, W. N.; Ribeiro, F. H. Appl. Catal., B 2012,
25, 206-214.
2
(
1
(72) Zosimova, P. A.; Smirnov, A. V.; Nesterenko, S. N.; Yuschenko, V. V.; Sinkler, W.; Kocal, J.;
Holmgren, J.; Ivanova, I. I. J. Phys. Chem. C 2007, 111, 14790-14798.
(
(
73) Perez, J.; Gonzalez, E.; Ticianelli, E. Electrochim. Acta 1998, 44, 1329-1339.
74) Schalow, T.; Brandt, B.; Starr, D. E.; Laurin, M.; Shaikhutdinov, S. K.; Schauermann, S.;
Libuda, J.; Freund, H. J. Angew. Chem. Int. Ed. 2006, 45, 3693-3697.
(75) Bazin, D.; Sayers, D.; Rehr, J.; Mottet, C. The Journal of Physical Chemistry B 1997, 101,
332-5336.
76) Penner-Hahn, J. E. In
Encyclopedia of Life Sciences; John Wiley
5
(
&
Sons, Ltd: Chicester, 2001,
DOI:10.1038/npg.els.0002984 10.1038/npg.els.0002984.
(77) Miller, J. T.; Kropf, A. J.; Zha, Y.; Regalbuto, J. R.; Delannoy, L.; Louis, C.; Bus, E.; van
Bokhoven, J. A. J. Catal. 2006, 240, 222–234.
(78) Ravel, B.; Kelly, S. X-Ray Absorption Fine Structure--XAFS 13 2007, 882, 150-152.
(79) Nelson, R. C.; Miller, J. T. Catal. Sci. Technol. 2012, 2, 461-470.
(80) Ocken, H.; Van Vucht, J. J. Less-Common Met. 1968, 15, 193-199.
(81) Ravel, B.; Newville, M. J. Synchrot. Radiat. 2005, 12, 537–541.
(82) Dietrich, P. J.; Lobo-Lapidus, R. J.; Wu, T.; Sumer, A.; Akatay, M. C.; Fingland, B. R.; Guo, N.;
Dumesic, J. A.; Marshall, C. L.; Stach, E. Top. Catal. 2012, 55, 53-69.
(
83) Rooksby, H.; Lewis, B. J. Less-Common Met. 1964, 6, 451-460.
(84) Zanuttini, M.; Dalla Costa, B.; Querini, C.; Peralta, M. Appl. Catal., A 2014, 482, 352-361.
(85) Zhu, X.; Nie, L.; Lobban, L. L.; Mallinson, R. G.; Resasco, D. E. Energy Fuels 2014, 28, 4104-
4111.
(
(
86) Talukdar, A.; Bhattacharyya, K.; Sivasanker, S. Appl. Catal., A 1993, 96, 229-239.
87) Pang, S. H.; Schoenbaum, C. A.; Schwartz, D. K.; Medlin, J. W. Nat. Commun. 2013, 4, 1-6.
2
0
ACS Paragon Plus Environment

Page 21 of 21
ACS Catalysis
1
2
(88) Peng, Y.; Qu, R.; Zhang, X.; Li, J. Chem. Commun. 2013, 49, 6215-6217.
(89) Han, Y.; Lu, C.; Xu, D.; Zhang, Y.; Hu, Y.; Huang, H. Appl. Catal., A 2011, 396, 8-13.
(90) Papp, J.; Soled, S.; Dwight, K.; Wold, A. Chem. Mater. 1994, 6, 496-500.
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(91) Yoon, Y.; Rousseau, R.; Weber, R. S.; Mei, D.; Lercher, J. A. J. Am. Chem. Soc. 2014, 136,
10287-10298.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
ACS Paragon Plus Environment