Dynamics of CrO3-Fe2O3 Catalysts during the High-Temperature Water-Gas Shift Reaction: Molecular Structures and Reactivity

Article abstract of DOI:10.1021/acscatal.6b01281

A series of supported CrO₃/Fe₂O₃ catalysts were investigated for the high-temperature water-gas shift (WGS) and reverse-WGS reactions and extensively characterized using in situ and operando IR, Raman, and XAS spectroscopy during the high-temperature WGS/RWGS reactions. The in situ spectroscopy examinations reveal that the initial oxidized catalysts contain surface dioxo (O=)₂Cr⁶⁺O₂ species and a bulk Fe₂O₃ phase containing some Cr³⁺ substituted into the iron oxide bulk lattice. Operando spectroscopy studies during the high-temperature WGS/RWGS reactions show that the catalyst transforms during the reaction. The crystalline Fe₂O₃ bulk phase becomes Fe₃O₄,and surface dioxo (O=)₂Cr⁶⁺O₂ species are reduced and mostly dissolve into the iron oxide bulk lattice. Consequently, the chromium-iron oxide catalyst surface is dominated by FeO_x sites, but some minor reduced surface chromia sites are also retained. The Fe_{3 - x}Cr_xO₄ solid solution stabilizes the iron oxide phase from reducing to metallic Fe⁰ and imparts an enhanced surface area to the catalyst. Isotopic exchange studies with C¹⁶O₂/H₂ → C¹⁸O₂/H₂ isotopic switch directly show that the RWGS reaction proceeds via the redox mechanism and only O? sites from the surface region of the chromium-iron oxide catalysts are involved in the RWGS reaction. The number of redox O? sites was quantitatively determined with the isotope exchange measurements under appropriate WGS conditions and demonstrated that previous methods have undercounted the number of sites by nearly 1 order of magnitude. The TOF values suggest that only the redox O? sites affiliated with iron oxide are catalytic active sites for WGS/RWGS, though a carbonate oxygen exchange mechanism was demonstrated to exist, and that chromia is only a textural promoter that increases the number of catalytic active sites without any chemical promotion effect.

Full text of DOI:10.1021/acscatal.6b01281

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
3
2
3
The Dynamics of CrO-FeO Catalysts during the High Temperature
Water-Gas Shift Reaction: Molecular Structures and Reactivity
Christopher J Keturakis, Minghui Zhu, Emma K. Gibson, Marco
Daturi, Franklin (Feng) Tao, Anatoly I Frenkel, and Israel E. Wachs
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b01281 • Publication Date (Web): 13 Jun 2016
Downloaded from http://pubs.acs.org on June 14, 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 25
ACS Catalysis
1
2
3
4
5
6
7
The Dynamics of CrO ꢀFe O Catalysts during the High Temperature
3
2
3
WaterꢀGas Shift Reaction: Molecular Structures and Reactivity
a†*
a
b‡
b
c
Christopher J. Keturakis , Minghui Zhu , Emma K. Gibson , Marco Daturi , Franklin Tao , Anatoly I.
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
d
a*
Frenkel , and Israel E. Wachs
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
a
Operando Molecular Spectroscopy and Catalysis Laboratory, Chemical Engineering Department,
Lehigh University, Bethlehem, PA 18015 USA
b
Laboratoire Catalyse et Spectrochimie, ENSICAEN, Université de Caen, CNRS, 6 Bd Maréchal Juin,
Fꢀ14050 Caen Cedex, France
c
Department of Chemical & Petroleum Engineering, University of Kansas, Lawrence, KS,
66047, USA
d
Department of Physics, Yeshiva University, New York, NY 10016, USA
†
‡
Current Address: Cummins Emission Solutions, Stoughton, WI 53589, USA
Current Address: Department of Chemistry, University College London, London WC1E 6BT, UK
Corresponding Authors*
Email: cjk@cummins.com (C.J.K.)
Email: iew0@lehigh.edu (I.E.W.)
Keywords: waterꢀgas shift, operando, in situ, spectroscopy, metal oxide, isotope exchange, surface
Abstract
A series of supported CrO /Fe O catalysts were investigated for the high temperature waterꢀgas shift
(WGS) and reverseꢀWGS reactions and extensively characterized using in situ and operando IR, Raman
and XAS spectroscopy during the high temperature WGS/RWGS reactions. The in situ spectroscopy
3
2
3
+
6
examinations reveal that the initial oxidized catalysts contain surface dioxo (O=) Cr O species and a
2
2
3+
bulk Fe O3 phase containing some Cr
substituted into the iron oxide bulk lattice.
Operando spectroscopy studies during the high temperature WGS/RWGS reactions show that the
catalyst transforms during reaction. The crystalline Fe O bulk phase becomes Fe O and surface dioxo
2
2
3
3
4
+
6
(O=) Cr O species are reduced and mostly dissolve into the iron oxide bulk lattice. Consequently, the
chromiumꢀiron oxide catalyst surface is dominated by FeO sites, but some minor reduced surface
chromia sites are also retained. The Fe Cr O solid solution stabilizes the iron oxide phase from
reducing to metallic Fe and imparts an enhanced surface area to the catalyst. Isotopic exchange studies
2
2
x
3
ꢀx
x
4
o
16
18
with C O /H → C O /H isotopic switch directly show that the RWGS reaction proceeds via the
2
2
2
2
redox mechanism and only O* from the surface region of the chromiumꢀiron oxide catalysts are
involved in the RWGS reaction. The number of redox O* sites were quantitatively determined with the
isotope exchange measurements under appropriate WGS conditions and demonstrated that previous
methods have undercounted the number of sites by nearly an order of magnitude. The TOF values
suggest that only the redox O* sites affiliated with iron oxide are catalytic active sites for WGS/RWGS,
though a carbonate oxygen exchange mechanism was demonstrated to exist, and that chromia is only a
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 25
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
textural promoter that increases the number of catalytic active sites without any chemical promotion
effect.
1
. Introduction
Mixtures of carbon monoxide and hydrogen, “syngas” or “waterꢀgas”, are encountered in many
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
crucial industrial processes including the manufacturing of ammonia (HaberꢀBosch Process), methanol,
hydrogen (Steam Methane Reforming (SMR), WaterꢀGas Shift (WGS)), and hydrocarbons (Fischerꢀ
1ꢀ4
Tropsch (FT)). Waterꢀgas was used as a source of hydrogen for the HaberꢀBosch process by
converting H O to H and CO to CO since CO was a catalyst poison and needed to be removed from
2
2
2
the waterꢀgas stream. In 1914, Bosch and Wild implemented an ironꢀchromium oxide catalyst that
o
5
converted CO into CO , an easily separable chemical, at 400ꢀ500 C. This first popularized the use of
2
the WGS reaction, Eq. 1, though the reaction was discovered more than a century earlier. Nearly the
same ironꢀchromium oxide based catalyst is still used today industrially, but with the addition of small
amounts of copper.
H O + CO ↔ H + CO
2
(1)
2
2
Catalysts for the WGS reaction are divided into several groups. Iron and chromium oxide based
o
catalysts are used for the high temperature WGS (HTS) at 350ꢀ450 C and copperꢀzincꢀaluminum oxide
o
based catalysts are used for the low temperature WGS LTS) at 190ꢀ250 C. Additionally, there has been
research into medium temperature shift (MTS) catalysts and sulfur tolerant “sour gas” shift catalysts.
2
,6
There have been many recent advancements in our understanding of the LTS catalysts, but HTS
catalysts remain poorly understood and nearly unchanged in composition for industrial use. The waterꢀ
gas shift reaction has been studied for over a century and several good reviews of the literature already
4
exist. Newsome reviewed the catalysis literature up to 1980, Rhodes et al. reviewed the literature up to
1
7
1
995, Ladebeck and Wagner provided a review up to 2003 with an emphasis on fuel cell applications,
2
Ratnasamy and Wagner reviewed recent developments up to 2009, and Zhu and Wachs focused on
ironꢀbased catalyst research up to 2016. In recent years, there has also been an interest in developing
8
Crꢀfree iron oxide based HTS catalysts because of the toxicity of hexavalent chromium, but successful
9
substitutes have still not be achieved. This literature review will primarily focus on new
characterization studies performed in the last two decades that provide new insights to the HTS
chromiumꢀiron oxide catalysts, but will also explore older, relevant publications if they contain
additional insights.
1
.1 Ambient and Ex Situ Characterization
It is well known that ambient and ex situ or post reaction characterization suffers from some
fundamental problems, such as sample hydration, oxidation and hydrocarbon adsorption, which limit the
usefulness of such data. It is now also well established that catalyst surfaces are dynamic under reaction
conditions, a trait that cannot be fully appreciated and understood without in situ and operando
10
spectroscopy analysis. Ambient and ex situ characterization of the ironꢀchromium oxide HTS catalyst
have been performed with many techniques including Xꢀray Diffraction (XRD), IR Spectroscopy,
Electron Microscopy, Xꢀray Absorption Spectroscopy (XAS), Xꢀray Photoelectron Spectroscopy (XPS),
1
1ꢀ35
and Mössbauer Spectroscopy.
Results of these studies have reached the following conclusions: 1)
the fresh catalyst is a Fe O phase, typically either alpha or gamma depending on synthesis and
2
3
3+
6+
3+
calcination conditions, 2) both Cr and Cr exist in the fresh catalyst, 3) Cr is substituted
preferentially into the iron oxide lattice octahedral sites (O ), but not tetrahedral sites (T ), 4) there is
h
d
ACS Paragon Plus Environment

Page 3 of 25
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
some surface segregation of chromium, 5) the reduced catalyst contains Fe O (magnetite), and 6)
3 4
discrete Cr O particles are present above ~14 wt% Cr O /Fe O .
2
3
2
3
2
3
1
.2 In Situ and Operando Spectroscopy Characterization
There have been extensive studies on iron oxides, especially in situ IR studies, due to the additional
3
6ꢀ47
47
interest from the FischerꢀTropsch community.
Boudjemaa et al. performed in situ IR spectroscopy
o
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
on unpromoted Fe O during the WGS at 450 C and did not detect any surface reaction intermediates
2
3
at this high temperature. Relevant surface intermediates such as formates and carbonates, however, are
36ꢀ42
not expected to exist at this high temperature.
Busca and Lorenzelli examined the IR spectrum of a
6
+
48
dehydrated FeCrO catalyst and noted that a Cr =O band is present, but could not identify its origin.
3
In situ Mössbauer spectroscopy of iron oxide catalysts under gas mixtures with different oxyreduction
potentials revealed that nonstoichiometric Fe O was formed and that the stoichiometry depended on the
3
4
4
9,50
oxyreduction potential of the gas mixture.
found to correlate with catalyst activity up to 350 C. In situ XRD measurements by Zanchet et al.
Furthermore, the oxygen vacancy in the iron oxide was
o
o
51
confirmed the transformation of Fe O to Fe O during the HTS at 350 C and above. Kendelwicz et al.
2
3
3
4
performed in situ Ambient PressureꢀXPS (APꢀXPS) studies of room temperature water adsorption
ꢀ
9
(partial pressures of 10 to 2 torr) on Fe O and combined the results with density functional theory
(
3
4
5
2
−4
−5
DFT) calculations. It was found that at the lower partial pressures, ≤ 10 − 10 Torr, water does not
dissociatively adsorb on the surface, except on defect sites, and progressive dissociation into surface
−4
−2
hydroxyl species takes place between 10 and 10 Torr.
Only one operando spectroscopy characterization experiment has been performed on the HTS
catalysts of iron oxide and chromiumꢀiron oxides. The rather unique study by Patlolla et al. combines
three spectroscopic techniques (XAS, XRD, and Raman spectroscopy) with simultaneous online mass
5
3
spectrometry (MS) for product analysis. The quality of the Raman spectroscopy probe was insufficient
for collection of data at high temperatures, thus, only allowing the Raman data to be collected in situ
before the HTS reaction and at room temperature ex-situ after cooling in the WGS mixture. In situ
2
ꢀ
Raman spectroscopy indicated the presence of hydrated CrO4 oxoanions before reaction and
demonstrated their disappearance, interpreted as reduction, after reaction. Before reaction, both catalysts
were a mixture of γ/αꢀFe O at elevated temperatures and only γꢀFe O after cooling to room
2
3
2
3
temperature. The corresponding operando XRD measurements confirmed that the catalysts were γꢀ
Fe O before reaction, Fe O during the reaction, and returned to γꢀFe O upon cooling to room
2
3
3
4
2
3
3
+
temperature. The operando XAS FeꢀK edge data demonstrated that the catalysts were in the Fe
oxidation state before reaction and were partially reduced during the HTS reaction, which is consistently
with Fe O formation.
3
4
1
.3 Catalytic Active Site and Reaction Mechanism
Research concerning the nature and strength of oxygen bonds of metal oxides and its correlation to
catalytic activity has been a topic of interest for many decades.
5
4ꢀ60
The adsorption and relaxation
kinetics of CO/CO and H /H O gas mixtures indicated that surface oxygen sites and vacancies were the
2
2
2
adsorption sites for the WGS reactants and products and comprised of ~10ꢀ20% of the BET surface area
under the conditions studied (pressures < 40kPa or ~0.4atm and temperatures of ~340ꢀ400 C).
o
12,57,58,61ꢀ65
1
6
Recently, the idea of counting sites was revisited by Zhu and Wachs who used the C O /H ꢀ
2
2
18
C O /H isotope exchange to count the number of participating oxygen sites during reaction at more
2
2
o
industrially relevant conditions (pressures of 1 atm and temperatures of ~330ꢀ400 C) followed by postꢀ
isotope exchange H ꢀTPR to probe the total amount of exchanged oxygen. Their results indicated that
2
the entire surface layer of the catalyst is participating in the reaction and that previous oxygen site
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 25
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
counting methods were significantly undercounting the number of participating sites by almost an order
of magnitude and, consequently, reporting turnover frequencies an order of magnitude or greater.
6
6
The reaction mechanism has been debated for many decades with kinetic equations based on a
regenerative or associative mechanism having some quantitative differences, albeit small differences,
2,4,12,58,61,67
from the overall reaction kinetics.
Furthermore, experimental proof exists for both
mechanisms. Equilibrium and kinetic measurements of individual reactants/products or with CO /CO
2
and H O/H gas mixtures has indicated that the regenerative mechanism is the primary pathway for the
2
2
1
2,55,57,61,62,64,65,67ꢀ69
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
HTꢀWGS reaction.
pathway may also exist due to the discrepancies between model and experimental rates mentioned
Many of these same experiments, however, suggest that a second
12,57,61,70
above.
Research has suggested both surface formate and carbonate species, with experimental
5
7,70ꢀ76
evidence for carbonates coming from isotope exchanges.
Paradigm shifts in the fundamental understanding of heterogeneous catalysts have occurred since
research on the chromiumꢀiron oxide catalyst began over a century ago and it is now accepted that
catalysts are dynamic during reaction and heterogeneous catalytic reactions only take place on the
surface of solid catalysts. The absence of extensive fundamental in situ and operando spectroscopic
studies of the surface of the Cr O /Fe O catalyst under appropriate HTꢀWGS conditions in the catalysis
2
3
2
3
literature has hindered the development of molecular level insights about the catalytic active sites,
surface reaction intermediates and the reaction mechanism. To address the surface properties of
chromiaꢀiron oxide catalysts during HTꢀWGS, a series of supported CrO /Fe O catalysts were prepared
3
2
3
and extensively characterized using operando Raman, IR, XAS and in situ APꢀXPS spectroscopy.
Potential surface reaction intermediates were examined with in situ IR spectroscopy during adsorption
of probe molecules (CO, CO , HCOOH, and CH OH) and also monitored during the WGS reaction with
2
3
operando IR spectroscopy. The redox characteristics of the chromiumꢀiron oxide catalysts were
determined by redox studies during in situ Raman and operando IR spectroscopy. The new fundamental
insights allow development of structureꢀreactivity relationships for the high temperature WGS by
supported CrO /Fe O catalysts under realistic WGS conditions.
3
2
3
2
. Experimental
2.1 Catalyst Synthesis and Preparation
The supported CrO /Fe O catalysts were prepared by incipient wetness impregnation of aqueous
3
2
3
solutions of chromium (III) nitrate (Cr(NO ) •9H O, Alfa Aesar, 98.5%) and distilled water on an iron
3
3
2
oxide support (γꢀFe O , Alfa Aesar, 99+%), using an incipientꢀwetness point of ~0.4 mL/g of Fe O
2
3
2
3
under ambient conditions. Samples were prepared with chromium oxide loadings of 1, 2, 3, 5, 7, and 9
wt% CrO . The samples were allowed to dry overnight under ambient conditions, followed by a second
3
o
drying step exposed to ambient air at 100 C for 4 hours in a programmable furnace (Thermolyne, Model
8000). Finally, the samples were subjected to calcination by ramping the temperature at 5 C/min under
flowing air (Airgas, Zero grade) to 350 C for 2 hours. The final synthesized catalysts are denoted as x%
CrO /Fe O , where x = wt% of chromium oxide.
o
4
o
3
2
3
2
.2 BET Specific Surface Area
The BET surface area of the catalyst samples was measured by nitrogen adsorptionꢀdesorption in
o
flowing N at ꢀ196 C with a Quantasorb surface area analyzer (Quantachrome Corporation, Model OSꢀ
2
9
). A sample quantity of ~0.3 g was typically employed for the measurement, and the sample was
o
outgassed at 250 C before N adsorption (Quantachrome Corporation, Model QTꢀ3).
2
2
.3 In Situ FTIR Spectroscopy and Probe Molecules/Reactions
All FTIR studies on supported catalysts in §2.3 were performed at the University of Caen, France.
2
For FTIR spectroscopy studies, the powdered catalyst samples were pressed into disks of ∼10 mg/cm
and activated in situ in the IR quartz cell equipped with KBr windows and attached to a high vacuum
ACS Paragon Plus Environment

Page 5 of 25
ACS Catalysis
1
7
7
o
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
system. Activation consisted of heating the catalyst at a rate of 10 C/min under 13 kPa of O from
2
o
o
room temperature (RT) to 350 C. The sample was held at 350 C for 45ꢀ60 mins followed by cell
evacuation at the same temperature for 30 mins. The sample disk was moved from the furnace part of
the cell to the optical section. Spectra were recorded at RT with a Nicolet Magna 550 FTIR spectrometer
using 4 cm resolution and 128 scans. They were treated by the Nicolet OMNIC™ software. For the
mixed metal oxide catalysts, Diffuse Reflectance FTIR (DRIFTS) was performed at Lehigh University
in a Thermo Nicolet 8700 spectrometer equipped with a commercial Harrick High Temperature
Reaction Chamber and Praying Mantis DRIFTS mirrors. The standard activation procedure is described
below in §2.4.
−
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
2
.3.1 Carbon Monoxide and Carbon Dioxide Probe Molecules
After catalyst activation in oxidizing conditions (10% O /Ar (Airgas, certified, 10.00% O /Ar balance)),
liquid nitrogen was added to a jacket surrounding the reaction cell, except for optical windows, in order
to cool the sample to ꢀ196 C (100K). For CO adsorption, the jacket was, instead, filled with solid CO
to reach ꢀ78 C (195K). A glass bulb containing CO or CO gas was attached to the vacuum system and
incrementally dosed based on pressure gauges and a calibrated volume. The gas was introduced so the
total partial pressure of CO or CO in the cell was increased incrementally from 1 Torr to 100 Torr, with
an IR spectrum taken at each increment. After 100 Torr was reached, an equilibrium partial pressure of
CO or CO between the glass bulb and cell was achieved, followed by the collection of another IR
spectrum. After CO or CO equilibration, the cell was evacuated to see if any species remained on the
surface. If species remained on the surface of the catalyst then the temperature was increased, first to
room temperature by allowing the liquid nitrogen or solid CO to evaporate, then at 10 C/min until the
2
2
o
2
2
o
2
2
2
2
o
2
o
surface species desorbed. Spectra were collected every 25 or 50 C.
2
.3.2 Formic Acid and Methanol Probe Molecules
After catalyst activation in oxidizing conditions (10% O /Ar (Airgas, certified, 10.00% O /Ar
balance)), a glass bulb containing HCOOH or CH OH was attached to the vacuum system and
incrementally dosed based on pressure gauges. The gas was introduced so the total partial pressure of
HCOOH or CH OH in the cell was increased incrementally from 1 Torr to 100 Torr, with an IR
spectrum taken at each increment. After 100 Torr was reached, an equilibrium partial pressure of
HCOOH or CH OH was allowed into the cell, followed by the collection of another IR spectrum. After
HCOOH or CH OH equilibration, the cell was evacuated to eliminate any physisorbed species on the
2
2
3
3
3
3
o
surface. The catalyst temperature was ramped at 10 C/min until the chemisorbed species desorbed.
Spectra were collected every 25 or 50 C.
o
2
.4 In Situ Raman Spectroscopy
The Raman spectra of the chromiunꢀiron oxide catalysts were obtained with a high resolution,
dispersive Raman spectrometer system (HoribaꢀJobin Yvon LabRam HR) equipped with three laser
excitations (532, 442, and 325 nm). The visible laser at 442 nm (violet) and the UV laser at 325 nm (not
visible) were generated by a HeꢀCd laser (Kimmon, model IK5751IꢀG). The lasers were focused on the
samples with a confocal microscope equipped with a 50X long working distance objective (Olympus
BXꢀ30ꢀ LWD) for the visible lasers and 15X objective (OFR LMUꢀ 15XꢀNUV) for the UV laser. The
LabRam HR spectrometer was optimized for the best spectral resolution by employing a 900 groves/mm
grating (HoribaꢀJobin Yvon 51093140HR) for the visible lasers and a 2400 grooves/mm grating
HoribaꢀJobin Yvon HR) for the UV laser. The resolution for both gratings is ~1 cm . The Rayleigh
scattered light was rejected with holographic notch filters (Kaiser Super Notch). The notch filter window
cutoffs were ~100 cm ~300 cm with the visible lasers and UV laser, respectively. The scattered light,
after removing the Rayleigh scattering, was directed into a UVꢀsensitive liquid N cooled CCD detector
ꢀ1
(
ꢀ1
ꢀ1
2
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 25
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
(HoribaꢀJobin Yvon CCDꢀ3000V). The calibration of each laser line was performed with an Hg lamp by
adjusting the groove gratings to match the zero position and minimize the error of the linearity across
the full Raman spectrum range. The Hg lines chosen to represent the 532, 442, and 325 nm lasers were
5
46.07, 441.6, and 365.02 nm, respectively. Additionally, wavenumber calibration of the Raman
ꢀ1
spectrograph was checked using the silicon line at 520.7 cm .
The catalyst samples, typically consisting of between 5 and 10 mg of loose powder, were placed in an
environmentally controlled highꢀtemperature cell reactor (Harrick High Temperature Reaction
Chamber) containing a quartz window and Oꢀring seals that were cooled with flowing cooling water.
The sample temperature was controlled by a temperature controller (Harrick ATC/low voltage
temperature control unit), providing linear heating rates of over 50°C/min through a Kꢀtype
thermocouple. The catalyst bed, however, was fitted with a second Kꢀtype thermocouple for cascade
control. Typical reactor cell conditions were room temperature (RT) ꢀ 400°C, 10ꢀ20°C/min heating and
cooling rates, atmospheric pressure, and ~40 sccm gas flowrates metered by mass flow controllers
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
(Brooks, Model 5850E series).
The protocol for obtaining in situ Raman spectra under an oxidizing (O /Ar) environment was as
2
follows. The sample was initially heated at a rate of 10 °C/min in the in situ cell to 350°C and held for
5ꢀ60 min under flowing 10% O /Ar (Airgas, certified, 10.00% O /Ar balance). For the acquisition of
4
2
2
the Raman spectra, only the laser angles parallel to the incident beam were allowed to hit the catalyst
sample, where the accumulation was collected at 10 s/scan for 7 scans with a 200 micrometer size hole.
78
The Raman spectra were collected with the 442 nm laser, due to known CrO resonance enhancement,
x
at 350 °C and also at RT after dehydration under the oxidizing conditions.
2
.5 Operando Raman Spectroscopy and In Situ Redox Cycles
The reverseꢀWGS (RWGS) was performed using the same Harrick High Temperature Reaction
Chamber and all Raman spectra were collected with a 442 nm laser. The gaseous outlet from the flowꢀ
o
through fixedꢀbed reactor was heated to 120 C and connected to an online mass spectrometer (Varian,
1
200L quadrupole). The online MS spectra were collected every 0.5s and m/e values 15ꢀ75
simultaneously monitored. The m/z values used to detect the reactants and products were m/z = 28 (CO),
m/z = 44 (CO ), and m/z = 18 (H O), with the CO signal corrected for CO cracking in the MS. The
2
2
2
Raman spectra were collected under reaction conditions with the simultaneous MS analysis of the
reaction product stream to constitute the operando spectroscopy experimental mode.
The catalysts underwent the same dehydration procedure as described in section 2.3. After
o
o
dehydration, the catalysts were heated to 400 C at 10 C/min, still under the O /Ar environment. At
2
o
400 C and after a short inert flush (15 mins), a reverseꢀWGS feed of X% H /Y% CO /Ar flowing at 42
2
2
cc/min total was introduced into the reactor cell. Experiments were repeated with the same total flow
rate but with H :CO ratios ranging from 0.5 to 4. Spectra accumulation consisted of 120 s/scan for 5
2
2
th
scans. Spectra were collected every 5 minutes for the first 10 minutes, then once every 20 subsequent
minute. The samples were run for a minimum of 60 minutes.
To determine the redox capabilities of the samples, the catalyst were allowed to reach steady state
o
during RWGS (typically 60ꢀ120 mins) before an inert flush was performed for 5 mins at 400 C. After
the flush, a mixture of either 33% CO /He or 2.5% H O/He (bubbled) was sent through the reactor for
2
2
3
0 mins. Another inert flush was performed for 5 mins before finally introducing a 10% O /Ar gas
2
o
mixture into the reactor at 400 C for another 60 mins to perform the final reꢀoxidation. Raman spectra
were collected after each step.
2
.6 Operando Xꢀray Absorption Spectroscopy (XAS)
The Xꢀray Absorption Spectroscopy (XAS) experiments were performed at Brookhaven National
Laboratory (BNL) National Synchrotron Light Source (NSLS) beamline X19A. The Cr and Fe Kꢀedge
ACS Paragon Plus Environment

Page 7 of 25
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
XAS data were obtained in fluorescence mode, using a PIPS detector. The monochromator Si(111)
crystals were detuned 30% to minimize harmonics. The powder samples were loaded in a 1.0 mm o.d.
(
0.9 mm i.d.) quartz capillary, known also as Clausen cell. An Omega thermocouple was inserted into
the capillary and placed adjacent to and contacting the catalyst bed. The catalyst sample was heated by
using a resistive heater placed under the catalyst bed. Temperature was controlled using an Eurotherm
temperature controller, and gas flow into the cell was controlled by Brooks mass flow controllers. The
catalyst was activated by flowing 10% O /He at 30 cc/min flow rate and ramping the temperature to
3
the temperature was raised to 400°C and the WGS reaction mixture introduced into the cell by directing
CO mixed with helium through a water bubbler before entering the reactor (resulting in ~3% H O and
H O:CO ratio of 2:1). The data were then collected at 400 C during the WGS reaction mixture flow.
Simultaneous online product analysis was done using a 0ꢀ100 amu quadrupole mass spectrometer (SRS).
The WGS reaction was performed for ~90 mins or until the reaction appeared to reach steady state
conversion. After reaction, the catalyst was typically cooled to 350 C under flowing 10% O /He and
held for ~ 30 mins before cooling to room temperature under the same atmosphere. Data were collected
at 350 C and at room temperature. The Fe Kꢀedge EXAFS data were corrected for selfꢀabsorption by
comparing the data measured in the Clausen cell in ambient conditions and using tape samples measured
in transmission mode and free from selfꢀabsorption distortion. This comparison allowed us to obtain the
scaling factor that was later applied to all operando XAS spectra to correct for selfꢀabsorption effects.
2
o
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
50°C at 10 C/min heating rate, followed by annealing at 350°C for 45ꢀ60 mins. For the WGS reaction,
2
o
2
o
2
o
79
2
.7 In situ Ambient Pressure Xꢀray Photoelectron Spectroscopy (APꢀXPS)
APꢀXPS studies were performed by Franklin Tao’s group by using labꢀbased ambient pressure Xꢀray
8
0ꢀ82
photoelectron spectrometer.
Analysis GmbH was used as the Xꢀray source. Energy analyzer is Phobios 150. The sample was
introduced to the reaction cell of this system. The CO and water vapor were mixed before introduction
to the reaction cell. The introduction of H O (1 Torr) and CO (0.5 Torr) was confirmed with mass
spectrometer installed on lens 1. Photoemission features of Fe 2p and Cr 2p were collected during WGS
Monochromator Al K α (Microfocus 600) made by Specs Surface Nano
2
reaction as a function of temperature. Catalytic activity was confirmed by formation of H and CO2
reaction products with the online mass spectrometer.
2
1
6
18
2.8 C O /C O Isotope Exchange during RWGS
2
2
1
6
18
The C O /C O isotope exchange experiments were carried out in the Altamira Instruments system
2
2
(AMI 200) reactor connected to Dymaxion Dycor mass spectrometer (DME200MS). About 20mg of
catalyst was loaded into a quartz Uꢀtube, and was dehydrated with 10% O /Ar at 400ºC. After
dehydration, the ironꢀchromium oxide catalyst was first allowed to equilibrate under C O /H reverse
WGS reaction conditions (10ml/min C O , 10ml/min H ) at 330ºC for 1 hour, then the catalyst was
flushed by inert gas (20ml/min He) for 10 min followed by switch to isotopic reverse WGS reaction
conditions (10ml/min C O , 10ml/min H ). The C O /C O isotope exchange was monitored with the
online mass spectrometer by recording the evolution of H O, H O, C O , C O O and C O
species (m/z=17, 19, 44, 46, 48). All mass spec signals were normalized to the same maximum and
minimum intensity to observe their transient behavior.
2
1
6
2
2
16
2
2
18
16
18
2
2
2
2
16
18
2
16
16 18
18
2
2
2
3
3
. Results
.1 Characterization before WGS/RWGS Reactions
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 25
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
A suite of in situ spectroscopy techniques, IR, Raman, XANES, and XPS were used to characterize
the molecular structure of the supported CrO /Fe O catalyst before reaction. In all experiments, the
3
2
3
o
catalyst was pretreated at 350 C in an oxidizing environment, which is considered to be the before
reaction state. All techniques (Supplemental information, §S2.1 and Figures S1ꢀS4) indicate that the
initial iron oxide phase is Fe O (Raman: γꢀFe O bands present, IR: Fe O bands present, XANES: Fe Kꢀ
2
3
2
3
2
3
edge spectra match Fe O reference compound, XPS: Fe 2p region exhibits characteristic peaks,
2
3
3
+
shoulders, and satellite peaks of Fe from Fe O ). Additionally, in situ IR spectra (Figure S1) revealed
2
3
ꢀ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
strong bands at 1008 cm along with a shoulder at 993 cm . The corresponding in situ Raman
spectroscopy (Figure S2) exhibits a bans at 997 cm . The band at ~1008 cm is assigned to
v (Cr(=O) ) and the band at ~993ꢀ997 cm is assigned to v (Cr(=O) ) vibrations of a surface dioxo
ꢀ1
ꢀ1
ꢀ1
as
2
s
2
7
8,83ꢀ85
(
O=) CrO species.
The Crꢀfree Fe O sample exhibits many surface hydroxyl vibrations and the
2
2
2 3
titration of these hydroxyls with increasing chromia loading reveals that the surface (O=) CrO species
2
2
anchor by reacting with surface FeꢀOH species. The surface hydroxyl titration suggests that ~3%
CrO /Fe O is the surface CrO monolayer coverage, but the reappearance of some surface hydroxyls
3
2
3
x
upon further addition of chromia above monolayer coverage suggests that the iron oxide surface is reꢀ
exposed (either from clustering of chromia on the surface or its dissolution into the bulk). The in situ
ꢀ1
Raman spectrum (Figure S2) also possess a band at 842 cm from the v(CrꢀOꢀFe) bridging bond
7
8
ꢀ1
between the surface dioxo species and Fe O support. A small, unlabeled band at 571 cm is present
2
3
for the 9% CrO /Fe O catalyst, and may indicate the presence of a minor amount of Cr O
3
2
3
2
3
ꢀ1 78
nanoparticles (expected sharp band at 550 cm ). In situ XANES Cr Kꢀedge spectra (Figure S3) reveal
6+
that the monolayer catalyst (3% CrO /Fe O ) almost exclusively possesses Cr O from the intense
3
2
3
4
6
+
3+
XANES preꢀedge peak, while the 9% CrO /Fe O catalyst possesses a mixture of Cr O and Cr O
3
2
3
4
6
based on reference compounds (see supporting information Figure S5 for reference compounds). The Cr
6
+
2p region of the in situ XPS spectra (Figure S4) exhibits a peak at ~578.7 ± 0.2 eV from Cr 2p
3
/2
8
6,87
3+
87
electrons
and two major Cr 2p peaks at 576.6 and 575.6 eV similar to that of Cr O . All in situ
3/2 2 3
spectra and a detailed analysis of each band/peak is given in the supplemental section S2.1 and Figures
S1ꢀ S4.
3
3
.2 Probe Molecules and Reactions
.2.1 In Situ IR for Adsorption of CO , CO, and HCOOH
2
Adsorption of the WGS reactants/products CO and CO was investigated at low temperatures in an
2
IR spectroscopy system while adsorption of formic acid (HCOOH) was investigated at elevated
temperatures and the results are given in Figure S6, S7, and S8 of the supplementary information,
respectively. The adsorption of CO on the Fe O support (Figure S6) initially exhibits bands
2
2
3
characteristic of carboxylates, bicarbonates, bidentate carbonates, and bridged carbonates, but as the
catalyst warms to room temperature bands from a minor amount of surface formates (HCOO*) appear.
The supported 3% CrO /Fe O catalyst does not exhibit any major bands from carbonates and as the
3
2
3
catalyst warms strong bands from surface formates appear which interact with surface dioxo (O=) CrO
2
2
species. The IR spectra for carbon monoxide adsorption are presented in Figure S7 and reveal that CO
primarily bonds weakly with surface hydroxyls indicating the presence of weak Brønsted sites (slightly
9
3,94
more acidic than on silica).
Additionally, a small amount of CO reactively adsorbs forming
carboxylate and carbonate species, as seen for CO adsorption, but the coordination of each is unknown
2
due to poor definition of the IR bands. Corresponding in situ IR spectra for CO and CO adsorption on
2
the bulk Cr O *Fe O mixed metal oxide catalysts were not undertaken.
2
3
2
3
Given that surface formate is the most proposed surface reaction intermediate during the WGS
1
,2,4
reaction,
in situ adsorption of HCOOH was investigated in a temperature programmed IR
ACS Paragon Plus Environment

Page 9 of 25
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
spectroscopy system to determine the coordination and thermal stability of surface formate species and
the spectra are shown in Figure S8 (difference spectra). The Fe O catalyst exhibits several adsorbed
2
3
o
species at 100 C: physisorbed HCOOH, two bidentate formates (bidentateꢀI and ꢀII), and a monodentate
38,95ꢀ98
o
formate with bands summarized in Table S1.
All surface formates decompose by 250 C. The
supported 3% CrO /Fe O (monolayer coverage) catalyst exhibits only one major surface bidentate
3
2
3
formate species and a minor amount of surface monodentate formate. The bidentate formate is thermally
o
o
stable on the catalyst surface until 325 C, while the monodentate formate desorbs by 250 C. The
supported 9% CrO /Fe O catalyst exhibits some bands from all four formates observed on the Fe O
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
3
2
3
2
o
catalyst. The bidentateꢀI and ꢀII formates decompose by 300 C, while the monodentate formate
decomposes by 250 C. The negative band at 1008ꢀ1014 cm on all Cr containing catalysts indicates that
adsorbed formates interact with the surface (O=) CrO species. Catalysts containing the Cr oxide
o
ꢀ1
2
2
promoter exhibited higher thermal stability of surface formates than pure Fe O and also preferred a
2
3
bidentate formate coordination, which is the major species for the monolayer catalyst.
3
.3 Characterization during and after the WGS/RWGS Reactions
3.3.1 BET Surface Area
The BET surface areas of the freshly calcined and used catalysts are indicated in Table 1. The
addition of chromia to the catalysts only slightly decreases the initial surface area before reaction from
2
~
88 to 82ꢀ84 m /g. After the RWGS reaction, the surface area of the catalysts dramatically decreases.
While the surface area of the Crꢀfree Fe O catalyst decreases by an order of magnitude, the addition of
2
3
chromia to iron oxide stabilizes the surface area of the supported CrO /Fe O catalysts (factors of 1.36x
3
2
3
and 2.44x for 3% and 9% CrO /Fe O , respectively).
3
2
3
3
.3.2 Operando Raman Spectroscopy
To investigate the phase changes of the catalysts during reaction, Raman spectroscopy measurements
o
were performed at 400 C with simultaneous gas phase monitoring using an online MS (operando
spectroscopy methodology). The results during the RWGS with H :CO ratios ranging from 0.5 to 4 are
2
2
presented in Figures S9, S10, and S11 of the supplementary information for Fe O , 3% CrO /Fe O and
2
3
3
2
3,
9
% CrO /Fe O catalysts, respectively. The operando Raman spectra of the Fe O catalyst (Figure S9)
3 2 3 2 3
indicate that the catalyst transitions from a Fe O phase to Fe O during the reaction, at all H :CO ratios
2
3
3
4
2
2
tested, while the corresponding MS signal shows steadyꢀstate production of CO and H O after 60
2
minutes, confirming the equilibrated state of the catalyst. The operando Raman spectra of the supported
3
% and 9% CrO /Fe O catalysts are shown in Figures S10 and S11, and indicate the same bulk iron
3 2 3
oxide transformation. Furthermore, Raman bands for the surface (O=) CrO species are initially present
2
2
ꢀ1
at 835 [v(CrꢀOꢀFe)] and 996 cm [v (Cr(=O) )], but disappear during the RWGS reaction suggesting
s
2
that the surface chromia sites have altered and most likely became reduced. The same results are also
observed for all examined H :CO ratios. These findings reveal that Fe O is also the main bulk iron
2
2
3
4
Table 1: BET surface areas of the supported CrO /Fe O catalysts before and after the reverseꢀ
3
2
3
WGS (RWGS) reaction.
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 25
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
oxide phase present for all chromiumꢀiron oxide catalysts during the RWGS reaction. A more detailed
analysis of the data is given in the supplemental section S2.3.
3
.3.3 Operando Xꢀray Absorption Spectroscopy (XAS)
The dynamics of the Fe and Cr bonding environments (FeꢀX and CrꢀX) and oxidation states during
o
the WGS reaction were probed by XAS studies at 400 C (labeled WGS S.S.) with simultaneous
monitoring of the gas phase using a RGA mass spec (operando spectroscopy methodology). The
spectral results during WGS (H O:CO = 2) for Fe O (Figure S12 in supplementary information),
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
2
3
supported 3% CrO /Fe O (Figure 1), and supported 9% CrO /Fe O (Figure S13 in supplementary
3
2
3
3
2
3
information) are presented in the respective figures.
The Fe KꢀEdge XANES spectra of iron oxide before, during, and after the WGS reaction are
presented in Figure S12A of the supplementary information. The XANES spectrum before reaction
o
(
ambient and 350 C Ox) is consistent with Fe O bulk phase (α, γ, or their mixture) from the preꢀedge
2
3
92,99
feature and edge position (reference compounds given in Figure S14 of supplementary information).
After ~70 mins. of the WGS reaction (WGS S.S.), the XANES edge position significantly shifts to lower
energy and the sharp preꢀedge feature is absent. These XANES features are consistent with a zeroꢀvalent
o
metallic iron as the dominant iron phase. Upon subsequent reꢀoxidation (350 C ReOx), the XANES
edge position is not fully recovered indicating an irreversible change to the bulk iron phase and a lower
3+
Cr
Cr K-Edge
A
Fe K-Edge
B
1.5
o
3
50 C ReOx
1.0
0.5
0.0
Ambient
Ambient
WGS S.S.
1.0
6
+
Cr
o
3
50 C Ox
o
3
50 C Ox
WGS S.S.
o
3
50 C ReOx
0.5
0.0
3
% CrO /Fe O
3% CrO /Fe O
3
2
3
3
2
3
5980
6000
6020
6040
6060
6080
7100
7110
7120
7130
Energy (eV)
7140
7150
7160
Energy (eV)
-7
1
.0
.9
0.8
.7
.6
.5
.4
0.3
.2
.1
.0
5.0x10
4.0x10
3.0x10
2.0x10
1.0x10
C
D
0
Fe R-Space
Mass Spec
-
7
3
% CrO /Fe O
3 2 3
0
o
CO
350 C ReOx
-7
0
0
WGS S.S.
-7
0
CO2
o
3
50 C Ox
H2
-7
0
H O
0
2
Ambient
0
0.0
0
1
2
3
4
5
6
7
8
0
10
20
30
40
50
60
70
80
Å
R (
)
Time (min)
Figure 1: Operando XANES/EXAFS spectra of 3% CrO /Fe O at room temperature (ambient), before WGS
3
2
3
o
o
o
(350 C Ox), during WGS at 400 C (WGS S.S.), and reꢀoxidized after WGS (350 C ReOx). A) XANES Cr Kꢀedge
spectra, B) XANES Fe Kꢀedge spectra C) EXAFS Fe RꢀSpace spectra, D) RGA data.
ACS Paragon Plus Environment

Page 11 of 25
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
average oxidation state. The EXAFS data before reaction, Figure S12B of supplementary information,
similarly suggests the presence of a Fe O bulk phase. The spectrum during reaction indicates a large
2
3
fraction of zeroꢀvalent metallic iron exists, due to the appearance of a large FeꢀFe peak at ~1.8 Å, and a
small fraction of an iron oxide phase, due to the low R peak at ~1.1 Å in the FeꢀO bond region. The lack
of high R peaks in the catalyst spectrum (high R peaks observed with the Fe reference compound)
0
indicates an amorphous phase without long range order. After reaction, the EXAFS spectrum is not
oct
oct
completely recovered upon reꢀoxidation, exhibiting a shift in the Fe ꢀFe peak (oct = octahedral
coordination) and a broad shoulder between 1.5ꢀ2.0 Å. The corresponding MS signal (Figure S12C of
supplementary information) shows that steady state production of CO and H begins at ~70 minutes,
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
2
confirming that the WGS is occurring and that the XAS measurements were taken under steady state
reaction conditions.
The Cr and Fe KꢀEdge XANES spectra for the supported 3% CrO /Fe O catalyst before, during and
3
2
3
after the WGS reaction are presented in Figures 1A & 1B. Before reaction, the Cr KꢀEdge XANES preꢀ
6
+
edge peak indicates that the catalyst possesses predominantly Cr , which decreases in intensity upon
o
o
heating to 350 C for pretreatment (350 C Ox). During reaction (WGS S.S.), the sharp preꢀedge peak is
6
+
3+
lost, the edge shifts to lower energies, indicating the reduction of Cr to Cr (near edge peak at ~6011
eV from Cr ). After reꢀoxidation, the catalyst contains an even larger fraction of Cr . A small bump at
the Cr preꢀedge peak position during and after reaction may indicate that a small amount of Cr is
present. The Fe KꢀEdge XANES spectra before reaction (Figure 1B, ambient and 350 C Ox) are
3
+
3+
6
+
6+
o
consistent with an Fe O bulk phase, while the spectrum during reaction shifts to lower energies by ~1.5
2
3
eV, which is consistent with an Fe O phase based on reference compounds (Figure S10 of
3
4
supplementary information). The Fe Kꢀedge position is recovered upon a postꢀreaction reꢀoxidation
treatment indicating that the phase change is reversible. The EXAFS data before reaction, in Figure 1C,
similarly suggests an Fe O phase, while the spectrum during reaction shows an increase in the ~3.2 Å
2
3
oct
tetr
shoulder from Fe ꢀFe bonds (tetr = tetrahedral coordination), consistent with the cubic spinel
structure of an Fe O phase. Postꢀreaction reꢀoxidation treatment recovers the original intensity of the
3
4
Fe O peaks. The corresponding mass spec data in Figure 1D indicates that steady state production of
2
3
CO and H begins at ~40 mins, confirming that the XAS measurements were taken under steady state
2
2
WGS reaction conditions. Results for the supported 9% CrO /Fe O catalyst, given in Figure S13 are
3
2
3
qualitatively the same and provide the same conclusions.
3
.3.5 In situ Ambient Pressure Xꢀray Photoelectron Spectroscopy (APꢀXPS)
Evolution of the oxidation states of the catalyst atoms in the surface region (1ꢀ3 nm) during reaction
o
were probed by in situ APꢀXPS experiments performed at 400 C. Results from a WGS feed with a
H O:CO = 2 are given in Figure 2 and Figure S15 (supplementary information) for supported 3%
2
CrO /Fe O and 9% CrO /Fe O catalysts, respectively. Initially, the Fe 2p region in Figure 2 (left, O
3
2
3
3
2
3
2
o
treatment at 350 C) is indicative of a Fe O phase, as previously discussed in §3.1.5. During reaction,
2
3
3
+
the 719 eV satellite peak vanishes and the large Fe 2p3/2 peak at ~710.8 eV broadens. Likely, the
broadening of the photoemission feature at ~710.8 eV results from the appearance of a shoulder at 709ꢀ
2
+
7
08 eV during reaction, which indicates the presence of Fe in a Fe O or FeO phase, which is typically
3 4
87
o
located at 708.4 eV. As shown in Figure 2, the photoemission feature of Cr 2p region (right, O 350 C)
reveals a contribution from Cr (578.9 eV), with a fraction of 0.44, and Cr (576.6 and 575.6 eV), with
a fraction of 0.56, during pretreatment at 350 C in O before reaction, as previously discussed in §3.1.5.
During reaction, the fraction of Cr peak at 578.9 eV decreases as low as 0.27. In addition, the position
of the main Cr peaks shifts to 577.4/576.4 eV from 0ꢀ45 mins and then to 577.0/575.9 eV at 90
minutes. These peak shifts suggest that the local chemical environment of Cr changes from Cr solely
coordinating with oxygen atoms to Cr binding to lattice oxygen atoms and surface OH groups at 45
minutes, and then to Cr in a chemical environment similar to FeCr O at 90 minutes. Results for the
2
6+
3+
o
2
6
+
3
+
3
+
3+
3+
3+
87
2
4
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 25
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
% CrO /Fe O
3% CrO /Fe O
3 2 3
3
2
3
700 a.u.
Fe 2p
Cr 2p
100 a.u.
7
08.4
719.0
o
CO + H O 400 C
90 mins
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
4
0
5 mins
mins
90 mins
4
0
5 mins
mins
90 min
08.4
7
0
min
O
3
5
0
C
o
o
7
11 710 709 708 707 706
O 350 C
O 350 C
2
2
7
24
719
714
709
704
582581 580579 578577 576575 574
Binding Energy (eV)
Binding Energy (eV)
Figure 2: In situ NAPꢀXPS Fe 2p (left) and Cr 2p (right) spectra of 3% CrO /Fe O during the WGS reaction at
00 C, 1.0 torr H O and 0.5 torr CO. Fe 2p Inset: Zoomed region containing 708.4 eV shoulder.
3
2
3
o
4
2
supported 9% CrO /Fe O catalyst, given in supplementary info Figure S15, are similar to the supported
% CrO /Fe O catalyst. The catalysts, however, differ in the extent of reduction of Fe O to the Fe O
or FeO phase (reduction of the 3% CrO /Fe O catalyst is much less than for the 9% CrO /Fe O
catalyst, which has a much smaller 708.4 eV shoulder). In addition, in the atomic fraction of Cr
decreases to 0.21 during reaction for the supported 9% CrO /Fe O catalyst.
3
2
3
3
3 2 3 2 3 3
4
3
2
3
3
2
3
6
+
3
2
3
3
.3.5 In situ Raman Spectroscopy during Redox Cycles
After activating the supported 9% CrO /Fe O catalyst in the RWGS reaction conditions, reꢀoxidation of
the catalyst with H O or CO gases was investigated and the results are given in Figure 3. During the
RWGS reaction, the catalyst transforms into the Fe O bulk phase (Raman bands at 279, 493, and 640
cm ) and the surface chromia species (Raman band at ~997 cm ) become reduced. After the RWGS, a
feed of 2.5% H O/He or 33% CO /He was fed into the reactor to determine if either molecule can reꢀ
oxidize the bulk iron oxide phase and surface dioxo (O=) Cr O species. Under the H O feed at 400 C
Figure 3A), a band from αꢀFe O (215 cm ) appears indicating that the bulk iron oxide phase was
partially reꢀoxidized, but the Raman band for the surface dioxo (O=) Cr O species (993 cm ),
however, is not present. Upon exposure of the catalyst to CO after the RWGS reaction at 400 C (Figure
3B), the surface dioxo (O=) Cr O species (997 cm ) is again not present and bulk iron oxide bands
from both Fe O (645 cm ) and αꢀFe O (214 cm ) are present. The surface dioxo (O=) Cr O species
997 cm ) are only partially recovered upon exposure to a 10% O /Ar feed. Although H O and CO are
sufficiently oxidizing for oxidation of the bulk Fe O phase to Fe O , both H O and CO are not
sufficiently strong oxidizing agents for the reꢀoxidation of Cr present in WGS activated chromiumꢀ
iron oxide catalysts back to Cr .
3
2
3
2
2
3
4
ꢀ
1
ꢀ1
2
2
6
+
o
2
2
2
ꢀ1
(
2 3
6+
ꢀ1
2
2
o
2
6
+
ꢀ1
2
2
ꢀ1
ꢀ1
6+
3
4
2
3
2
2
ꢀ1
(
2 2 2
3
4
2
3
2
2
+
3
+
6
1
6
18
3
.3.6 C O /C O Isotope Exchange during RWGS
2 2
1
6
18
The C O /C O isotope switching was performed with bulk Fe O and supported 3%, and 9%
2
2
2
3
CrO /Fe O catalysts during the RWGS reaction to gain insights into the oxygen exchange process
3
2
3
ACS Paragon Plus Environment

Page 13 of 25
ACS Catalysis
1
2
3
4
5
6
7
8
9
o
o
3
38
4
00 C
400 C
3
38
2
15
214
833
833
6
39
640
200 a.u.
4
93
684
200 a.u.
493
684
279
996
279
997
Before
RWGS
Before
RWGS
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
RWGS
RWGS
H O
2
3
90
CO2
3
90
5
90
5
90
A
O2
B
O2
200
400
600
800
1000
1200
200
400
600
800
1000
1200
-
1
-1
Raman Shift (cm )
Raman Shift (cm )
Figure 3: In situ Raman spectroscopy before, during, and after the RWGS reaction with a supported 9%
CrO /Fe O catalyst. A) 2.5% H O/Ar oxidizing treatment after the RWGS reaction, B) 33% CO /Ar oxidizing
3
2
3
2
2
treatment after the RWGS reaction.
taking place during the reaction and the results are presented in Figure 4. The catalysts were initially
equilibrated under a typical RWGS feed, flushed with helium, and then exposed to an isotopic H +C O
RWGS feed. For the bulk Crꢀfree iron oxide catalyst, after the C O /C O isotope switch, the first
observable products were C O and H O, closely followed by the mixed oxygen C O O isotope
note that the transient CO isotope signals are not reliable because of CO cracking in the MS). The
isotopic product H O is formed last. The transient production and decay of C O and C O O are
similar while only the production of the water H O isotope is similar. The complete decays of the
C O and C O O isotopes are rapid, < 2 mins after exchange, compared to H O, ~5 mins after
18
2
2
1
6
18
2
2
1
6
16
16 18
2
2
(
2
1
8
16
16 18
2
2
1
6
2
1
6
16 18
16
2
2
exchange.
1
6
18
16
The C O /C O isotope exchange experiment during RWGS reveals that the amount of O being
supplied by the lattice of the ironꢀchromium oxide catalyst (C O O, H O, C O as well as
undetectable C O) is finite and small since most of the oxygen exchange occurs within ~2 mins. The
2
2
16 18
16
16
2
2
1
6
1
6
16
18
number of O* atoms participating during the C O /C O isotope exchange experiment were
2
2
1
6
16
18 16
quantified by integration of the signals of Oꢀcontaining products (sum of 2*C O , C O O and
2
1
6
H O). The RWGS activity and turnover frequencies based on a surface O* as the most abundant
reactive intermediate (MARI) are presented in Table 2. The overall catalyst activity (H O mol/(gꢀs)) is
virtually the same with or without the addition of chromia to iron oxide, however, the number of
participating O atoms (Ns = O atoms/g) increases by 2.4. The corresponding TOF values
(TOF=activity/Ns) indicate that the specific TOF value for chromiumꢀiron oxide catalyst is the same as
the Crꢀfree iron oxide catalyst up to the initial monolayer coverage (3% CrO /Fe O ), after which
additional chromium causes a decrease in the TOF by a factor of ~2.3. Thus, Cr is a textural promoter
that increases the number of catalytic active sites, but does not chemically promote the WGS reaction by
iron oxide. The slightly lower TOF for the 9% CrO /Fe O catalyst may indicate a slight retardation of
2
2
1
6
3
2
3
3
2
3
the specific activity in the presence of significant chromia.
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 25
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
Fe O
H2 + C O2
3% CrO /Fe O
9% CrO /Fe O
3 2 3
18
2
1
3
3
2
3
8
18
He
He
He
He
H
2
+ C
O
2
He
He
H2 + C O2
18
18
H2
O
H O
2
18
H2
H2
O
H2
H2
18
18
C
O2
18
C O2
C
O2
16
H2
O
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
16
H2
O
1
C
6
18
O
O
1
6
H2
O
8
16
C
O2
16
C
O2
1
6
18
O
C
O
16
C
O2
16
18
O O
C
4
5
6
7
8
9
10 11 12 13 14 15
4
5
6
7
9
10 11 12 13 14 15
4
5
6
7
8
9
10 11 12 13 14 15
Time (min)
Time (min)
Time (min)
1
8
18
18
H2 + C O2
He
He
H2 + C O2
He
H
2
+ C
O
2
He
He
1
8
1
8
H2
18
O
H2
O
18
C
O2
18
C
O2
C
O2
He
H2
H2
H2
16
H2
O
1
8
O
H2
1
6
O
H2
1
6
H2
O
16 18
O O
C
1
6
O
1
6
18
C
2
16 18
O
16
C
O2
1
C
6
C
O
O
C
O
O2
4
.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50 6.75 4.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50 6.75 4.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50 6.75
Time (min)
Time (min)
Time (min)
16
18
Figure 4: RWGS C O /C O isotope exchange with Fe O and x% CrO /Fe O supported catalysts. Top Row)
2
2
2
3
3
2
3
Normalized MS signals, Bottom Row) Zoomed in region.
4
4
4
. Discussion
.1 Catalyst Structures in Initial Oxidized Catalyst
.1.1 Bulk Structures. The bulk iron oxide phase is initially present as Fe O for all catalysts. For the
supported CrO /Fe O catalysts, the iron oxide phase was present only as γꢀFe O (see Figure S2),
however, for pure iron oxide the phase can be a mixture of α- and γꢀFe O . The absence of discrete
2
3
3
2
3
2
3
2
3
Cr O nanoparticles for most chromium oxide loadings greater than monolayer coverage for the
2
3
supported CrO /Fe O catalysts (see Figure S2) suggests Cr dissolution into the Fe O support. This is
3
2
3
2
3
3+
further supported by the significant amount of Cr present in the in situ XANES and XPS (Figures S3
and S4) of the supported 9% CrO /Fe O catalyst, but not of the in situ XANES for the monolayer
3
2
3
supported 3% CrO /Fe O catalyst. In addition, just heating chromiumꢀiron oxide catalysts under
3
2
3
+
3
oxidizing conditions increases the concentration of Cr in the bulk Fe O lattice (see Figure 1).
2
3
3+
2+
3+
Dissolution of chromium oxide into the iron oxide bulk lattice as Cr is known to displace Fe /Fe
1
6
Table 2. RWGS activity, number of sites [Ns: n( O)], density of sites, and turnover frequencies (TOFs).
1
6
RWGS Activity
Density of Ns
TOF
Ns: n( O)
Catalyst
Fe O
ꢀ
6
16
2
ꢀ3 ꢀ1
ꢀ
3
(
× 10 mol/s·g)
( O atoms/nm )
(× 10 s )
(
× 10 mol/g)
5.5
6.0
5.6
1.0
1.1
2.4
7
8
5.5
5.4
2.3
2
3
3
% CrO /Fe O
3 2
3
3
9% CrO /Fe O
17
3
2
ACS Paragon Plus Environment

Page 15 of 25
ACS Catalysis
1
O
O
O
O
O
O
O
O
O
O
O
O
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
Cr6+
Cr6+
Cr6+
Cr6+
O
O
O
O
WGS
Fe O3
Before Reaction
Fe O
3 4
During Reaction
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
Scheme 1: Diagram of the surface and bulk molecule structures of the supported CrO /Fe O catalyst
3
2
3
before and during the WGS reaction.
1
1,13ꢀ17
from octahedral sites and expose the iron oxide surface.
A diagram of the bulk structure of the
initial oxidized chromiumꢀiron oxide catalyst is given in Scheme 1.
4
.1.2 Surface Structures. The initial oxidized chromiumꢀiron oxide catalysts possess a surface chromia
phase primarily present as dioxo (O=) Cr O surface species, but a small amount of surface monoꢀoxo
O=CrO species are also present (see Figure S1). The separation of ~14 cm between the symmetric and
6
+
2
2
ꢀ
1
4
asymmetric vibrations of O=Cr=O observed in the Raman and IR spectra, respectively, also matches the
1
1,13ꢀ17,100,101
vibrational rules for dioxo metal oxides.
The surface chromia species on Fe O are
2 3
metastable since their total concentration decreases at elevated temperatures in oxidizing environments
see Figure 1). There appears to be an equilibrium ratio of Cr /Cr in the surface region as indicated by
nearly identical atomic fractions observed in the in situ XPS (Figure S4). The surface chromia species
anchor to the surface of Fe O by titrating the iron oxide surface hydroxyls (see Figure S1). A diagram
of the major surface chromia species on iron oxide before reaction is given in Scheme 1.
.2 Catalyst Structures during the WGS/RWGS Reactions
4.2.1 Bulk Structures. During the WGS/RWGS reactions, the thermodynamically stable Fe O bulk
phase is present in all chromiumꢀiron oxide catalysts (see Figures 1, S9ꢀS13, S15), even under very
reducing conditions (Raman, H :CO = 4 for RWGS). For the Crꢀfree iron oxide catalyst, the metallic
Fe bulk phase predominates (EXAFS, see Figure S12). This overꢀreduction of iron oxide is wellꢀknown
6
+
3+
(
2
3
4
3
4
2
2
0
1
,2,4
+3
in the literature, indicating the need for very careful catalyst activation. The presence of Cr in the
bulk iron oxide lattice retards the formation of metallic Fe (EXAFS, see Figures 1, S12 and S13). The
dissolution of Cr into the iron oxide bulk lattice at elevated temperatures and WGS/RWGS reaction
0
+
3
+
3
conditions is indicated by the increased Cr signal (EXAFS, see Figure 1). Reꢀoxidation with O after
2
+6
RWGS only partially reoxidizes the chromia to surface Cr (EXAFS and Raman, see Figures 1, 3, S13),
+
3
11,13ꢀ17
+3
and is consistent with the trapping of Cr in the iron oxide bulk lattice.
oxide bulk lattice during the WGS/RWGS reactions are responsible for stabilization of the Fe O phase
(minimization of metallic Fe phase) and the enhanced BET surface area (see Table 1). A diagram of the
bulk structures of the chromiumꢀiron oxide catalysts during the WGS/RWGS reactions is given in
Scheme 1.
The Cr sites in the iron
3
4
0
4.2.2 Surface Structures. The surface of chromiumꢀiron oxide catalysts during WGS/RWGS contains
both exposed Cr and Fe sites. Although most of the chromia is dissolved as Cr in the Fe O bulk lattice
during WGS/RWGS, a small amount of Cr also remains on the surface of the catalyst (see Figure 2).
The Raman band for the dioxo surface Cr species (997 cm ) reappears, but with a low intensity, after
+
3
3
4
+
6
+6
ꢀ1
+
3
the RWGS reaction by reꢀoxidation with O indicating the presence of some reduced Cr species either
2
on the topmost surface or in the surface region (several nm depth) of the chromiumꢀiron oxide catalyst
during the WGS/RWGS reactions. The minor amount of surface chromia species on the chromiumꢀiron
oxide catalyst during the WGS/RWGS reactions suggests that the chromiumꢀiron oxide surface is
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 25
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
+
2
+3
dominated by the Fe O surface features of Fe and Fe in octahedral and tetrahedral coordinations (in
3
4
situ XPS, Figures 2 and S15). A diagram of the surface structure of the chromiumꢀiron oxide catalyst
during the WGS/RWGS reactions is depicted in Scheme 1. For pure iron oxide, it is hypothesized that
the fraction of α- and γꢀFe O phases before reaction will influence the reactivity of the catalyst as either
2
3
phase may promote specific stoichiometries on the catalyst surface.
4.3 Catalytic Active Sites during the WGS/RWGS Reactions
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
Although the surface of the chromiumꢀiron oxide catalyst during the WGS/RWGS reactions contains
mostly Fe and a minor amount of Cr /Cr , only the surface FeO sites are involved in the redox cycle.
The inability to reꢀoxidize the surface Cr sites with both H O and CO , the oxidizing agents during the
WGS/RWGS reactions, suggests that surface Cr sites can’t participate in the WGS/RWGS reactions
see Figure 3) while Cr is a spectator and may even retard the catalytic activity slightly, as indicated by
decreases in the TOF upon chromium addition (Table 2). This leaves only the surface FeO as being the
catalytic active sites for the WGS/RWGS reactions since Fe is reoxidized by both H O and CO (see
Figure 3) and Crꢀfree iron oxide is also able to perform the WGS/RWGS reactions.
The HTꢀWGS catalysis literature is conflicted about the role of chromium in the chromiumꢀiron
oxide catalysts. Traditionally, it was concluded that the only function of chromium is as a textural
promoter to increase the surface area of iron oxide, but the promotion mechanism was not
Multiple mechanisms, however, have been proposed for Cr stabilization of iron oxide
and its catalytic role: chromia forms discrete Cr O particles on Fe O preventing agglomeration of iron
+6
+3
x
+
3
2
2
+
3
+
6
(
x
2
2
1,2,4,7,14,21ꢀ25
known.
2
3
3
4
73,75
+3
oxide,
enrichment of Cr in the surface region is more thermodynamically stable than iron oxide
and reduces ion diffusion and sintering effects, dissolved Cr occupies octahedral sites in the bulk
2
3
+3
1
02
Fe O lattice that prevent sintering by forcing bulk FeO sites to occupy bulk FeO sites, the chromia
3
4
6
4
+3
+6
+2
+3
19
redox cycle of Cr ꢀ Cr promotes the Fe ꢀ Fe redox cycle of the Fe O phase. The proposal
3
4
+3
+6
+2
+3
that the Cr ꢀ Cr redox cycle promotes the Fe ꢀ Fe redox cycle of the Fe O phase is not
supported by the current findings since the chromia redox cycle doesn’t operate during the WGS/RWGS
reactions (see Figure 3). The current Raman studies also demonstrate that discrete Cr O particles are
3
4
2
3
+3
not present on Fe O because of facile dissolution of Cr in the iron oxide bulk lattice (see Figures S2,
3
4
S9ꢀS11), and negate this antiꢀagglomeration proposal. Analogously, a significant surface enriched
chromia phase is not present during reaction and, thus, surface chromia probably can’t be responsible for
preventing sintering of iron oxide. The only stabilization mechanism consistent with the experimental
findings is that dissolved Cr occupies octahedral sites in the bulk Fe O lattice which prevent sintering
by forcing bulk FeO sites to occupy bulk FeO sites. The current operando EXAFS findings (see
Figures 1, S12, and S13) also reveal that the Cr species dissolved in the iron oxide bulk lattice
suppress formation of metallic Fe that is more prone to sintering than iron oxides.
The isotope switch experiment (Figure 4) provides quantification of the number of exchangeable
active O* atoms participating in the RWGS reaction under realistic reaction conditions and using the
same reaction gas composition as for the in situ/operando spectroscopy experiments. Previous efforts
utilized adsorption and relaxation kinetics of CO/CO and H /H O gas mixtures to quantify the number
+
3
3
4
6
4
+
3
0
2
2
2
of O* atoms and concluded that they comprised of ~10ꢀ20% of the surface area under the conditions
studied (pressures < 40kPa or ~0.4atm and temperatures of ~340ꢀ400 C).
demonstrated that the Fe O (111) surface possess ¾ monolayer of oxygen atoms (14.2 atoms/nm ) and
o
12,57,58,61ꢀ65
It has been
2
3
4
2
¼
monolayer of iron atoms that can be saturated with hydroxyls (4.7 atoms/nm ), which gives an overall
2
103,104
2
density of 18.9 oxygen atoms/nm .
Results of the current isotope exchange give an experimental
1
6
value ranging from 7ꢀ17 O atoms/nm , indicating that nearly the entire surface layer can participate in
the HTꢀWGS reaction and that previous methods have significantly undercounted the number of sites by
nearly an order of magnitude. It is likely that undercounting in the literature is due to the use of high
vacuum systems that can alter the catalyst surface with multiple evacuations, low pressures during
ACS Paragon Plus Environment

Page 17 of 25
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
counting steps (< 40kPa), and the use of only half of the WGS reactants/products (CO/CO or H /H O
2 2 2
mixtures) during counting, which does not simulate the oxyreduction potential of the full WGS mixture.
4
.4 Surface Reaction Intermediates during the WGS/RWGS Reactions
There have been many surface reaction intermediates proposed for the WGS reaction with surface
carbonates and formates being the most popular. Although surface intermediates have been observed for
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
the low temperature WGS (CuꢀZnꢀAl) catalysts, there has been no direct evidence for surface reaction
1,2,4
intermediates during the high temperature WGS/RWGS reactions,
only some indirect evidence of
1
8
18
57
carbonates from isotopic ( O and C O ) temperatureꢀprogrammed desorption experiments. This is
2
2
related to the short lifetime and low concentrations of the surface intermediates during high temperature
1
8
18
WGS/RWGS that prevents their observation. For example, the efficient exchange of O from C O
2
1
6
18 16
with the chromiumꢀiron oxide catalyst to form C O and C O O, as observed during isotope exchange
2
(Figure 4), suggests that some exchange proceeds via surface carboxylate/carbonate intermediates, but
these intermediates can only be detected spectroscopically at subꢀambient temperatures where they have
a long lifetime (see Figure S6 and S7). Another possible reason for the lack of detection of surface
intermediates during high temperature WGS/RWGS is the loss of transparency of the catalysts with
respect to IR. Similarly, surface formate intermediates were observed on the iron and chromiumꢀiron
oxide catalysts upon adsorption of CO (see Figure S6), HCOOH (see Figure S8) and CH OH (see
2
3
Figure S18), but at temperatures lower than those typically encountered for the high temperatures of
o
o
~
400 C WGS/RWGS reactions since surface formates are not stable above 325 C (see Figure S8). Even
o
performing the WGS reaction at lower temperatures of 200ꢀ325 C does not yield the IR spectra of
surface formate intermediates (see Figures S16 and S17). The lack of observation of surface reaction
intermediates during high temperature WGS/RWGS are clouded by the possible loss of the IR signal
from the chromiumꢀiron oxide catalyst’s reduced state. The most abundant reactive intermediate is
actually the surface O* sites on the catalysts as shown by the isotope exchange analysis.
4
.5 WGS/RWGS Reaction Mechanisms
Two reaction mechanisms have been proposed for the high temperature WGS/RWGS reactions:
associative mechanism involving a surface reaction intermediate and a redox mechanism involving
1
,2,4
oxygen exchange between the gases and the iron oxideꢀbased catalyst.
As already indicated, there is
no direct evidence for the associative mechanism since surface intermediates are not detectable on either
iron oxide or chromiumꢀiron oxide catalysts during high temperature WGS/RWGS (see Figures S16 and
S17). The associative mechanism may be operating during high temperature WGS/RWGS, but at
present it hasn’t be directly proven.
There is direct evidence, however, that the redox mechanism is operating during the high temperature
1
6
18
RWGS reaction from transient C O /H ꢁ inert flush ꢁ C O /H experiments. The isotope switch
2
2
2
2
experiments demonstrate that the RWGS reaction is proceeding via the redox mechanism since the
16
16
16 18
16
isotope switch readily produces Oꢀcontaining reaction products (C O , C O O and H O) in the
2
2
18
C O /H reaction environment and the exchange is nearly complete within 2 mins. The short exchange
2
2
time indicates that a surface Marsꢀvan Krevelen (MVK) reaction mechanism is taking place where only
the surface layer is exchanging oxygen atoms, rather than a bulk MVK mechanism where complete
1
6
16 18
exchange only occurs after an extended period of time. The formation of the C O and C O O
2
1
8
products is not only from the RWGS since they can also form when only C O is exchanged with the
equilibrated Oꢀcontaining catalyst. This suggests that the CO exchange also takes place via short lived
surface carboxylates/carbonates, which is a parallel oxygen exchange pathway. Formation of H O
2
16
2
5
7
16
2
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 25
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
1
8
18
before H O in the presence of C O /H is consistent with a redox process taking place during RWGS.
2
2
2
Furthermore, the isotope switch experiment also provides quantification of the number of exchangeable
active O* atoms participating in the RWGS reaction under realistic reaction conditions.
4
.6 StructureꢀActivity Relationships
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
The catalytic active sites during the WGS/RWGS are O* atoms affiliated with iron oxide sites in the
outermost surface layer. This conclusion is supported by the (i) low concentration of surface chromia
sites on the equilibrated chromiumꢀiron oxide catalyst because of significant dissolution of Cr into the
iron oxide bulk lattice (see Figures 1ꢀ3, S2ꢀS4, S9ꢀS13, S15), (ii) ability of Crꢀfree iron oxide to perform
HTꢀWGS/RWGS (see Figures 4, S9, S12), (iii) reꢀoxidation of Fe to Fe with CO and H O (see
Figure 3), (iv) inability to reoxidize Cr back to Cr with CO and H O (see Figure 3), (v) inability to
extensively reꢀoxidize Cr back to surface Cr even with O after WGS/RWGS due to its dissolution in
the iron oxide bulk lattice (see Figures 1, 3, S13, S19ꢀS20), and (vi) lack of increase and even slight
+3
+
2
+3
2
2
+
3
+6
2
2
+
3
+6
2
decrease in TOF for Crꢀpromoted iron oxide catalysts (see Table 2). Thus, the only function of the
chromia promoter is to stabilize the Fe O phase with a higher surface area (textural promoter), which is
in agreement with the current accepted mechanism in the literature.
3
4
5
. Conclusions
A series of supported CrO /Fe O catalysts prepared by incipient wetness impregnation were
investigated for the high temperature WGS reaction as a function of chromia loading and extensively
characterized. Characterization before reaction by in situ IR, Raman, XAS, and APꢀXPS revealed that
the catalyst contains a 2ꢀD surface phase consisting primarily of dioxo (O=) Cr O species and a bulk
γꢀFe O phase with some Cr substituted into the bulk lattice. The adsorption of probe molecules was
monitored using in situ IR spectroscopy and revealed various adsorbed carbonates and formates from
CO and CO adsorption, and adsorbed formates from HCOOH adsorption directly. Surface formate
species were thermodynamically preferred and were demonstrated to be stable up to 250ꢀ325 C on the
3
2
3
6+
2
2
3
+
2
3
2
o
catalyst surface.
Characterization during the high temperature WGS reaction by operando IR, Raman, XAS, and in
situ APꢀXPS revealed that crystalline Fe O phase is the active bulk phase and surface dioxo
3
4
+
6
3+
(O=) Cr O species become reduced during reaction to Cr and migrate into the bulk iron oxide lattice
2
2
+
6
(Fe Cr O ). In situ APꢀXPS data have revealedthat a minor amount of Cr species may remain on the
catalyst surface under reaction conditions. Corresponding in situ Raman and operando IR redox cycles
indicate that the trace surface Cr is a spectator species during the WGS reaction. Isotope exchange
experiments during the RWGS indicate a surface Marsꢀvan Krevelen mechanism is occurring and the
catalytic active sites during the WGS/RWGS are surface O* atoms affiliated with iron oxide sites in the
surface region. The chromia promoter only increases the population of sites and it does not chemically
promote the reaction. Furthermore, quantification of the active O* atoms indicates that previous titration
methods have been undercounting the number of sites by nearly an order of magnitude. These new
insights have led to a modern fundamental understanding of the high temperature WGS catalyst.
3
ꢀx
x
4
+6
Supporting Information Available: Results available include 1) in situ IR and Raman spectroscopy, in
situ XANES/EXAFS, and in situ XPS of the dehydrated catalyst, 2) XANES and EXAFS reference
compounds, 3) in situ IR spectroscopy during adsorption of probe molecules CO, CO , HCOOH, and
CH OH, 4) operando Raman spectroscopy during the waterꢀgas shift reaction, 5) operando
XANES/EXAFS during the waterꢀgas shift reaction, 6) in situ APꢀXPS during the waterꢀgas shift
2
3
ACS Paragon Plus Environment

Page 19 of 25
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
reaction, 7) operando IR spectroscopy during the waterꢀgas shift reaction, and 8) operando IR
spectroscopy during redox cycles. This material is available free of charge via the Internet
at http://pubs.acs.org.
Acknowledgements
The authors gratefully acknowledge help of Relja Vasic and Nebojsa Marinkovic with synchrotron
measurements at Brookhaven National Laboratory. CJK and IEW gratefully acknowledge support by the
NSF IREE Award No. 0609018 and NSF Award No. 1511689. AIF acknowledges support by the U.S.
DOE Grant No. DEꢀFG02ꢀ03ER15476. Beamline X19A of the NSLS is supported in part by the U.S.
DOE Grant No. DEꢀFG02ꢀ05ER15688. NSLS is supported by U.S. DOE Contract No. DEꢀAC02ꢀ
98CH10886. FT acknowledges support by the Chemical Sciences, Geosciences and Biosciences
Division, Office of Basic Energy Sciences, Office of Science, U.S. Department of Energy under Grant
No. DEꢀSC0014561.
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
. Rhodes, C.; Hutchings, G.; Ward, A. Catal. Today 1995, 23, 43ꢀ58.
2. Ratnasamy, C.; Wagner, J. P. Catalysis Reviews 2009, 51, 325ꢀ440.
3
. Brenna, G. New Catalyst for the H2 Production By WaterꢀGas Shift Reaction Processes, Università di
Bologna, Italy, 2010.
4. Newsome, D. S. Catal. Rev. -Sci. Eng. 1980, 21, 275ꢀ318.
5
6
. Bosch, C.; Wild, W. United States Patent 1,115,776, 1914.
. Rodriguez, J. A.; Hanson, J. C.; Stacchiola, D.; Senanayake, S. D. Phys. Chem. Chem. Phys. 2013, 15,
12004ꢀ12025.
7
. Ladebeck, J. R.; Wagner, J. P. In Handbook of Fuel Cells – Fundamentals, Technology and
Applications; Vielstich, W., Lamm, A. and Gasteiger, H. A., Eds.; John Wiley & Sons: Chichester,
2003; Vol. 3, pp 190ꢀ201.
8. Zhu, M.; Wachs, I. E. ACS Catalysis 2016, 6, 722ꢀ732.
9
. Lee, D. ꢀ.; Lee, M. S.; Lee, J. Y.; Kim, S.; Eom, H. ꢀ.; Moon, D. J.; Lee, K. ꢀ. Catal. Today 2013, 210,
ꢀ9.
2
10. Wachs, I. E.; Keturakis, C. J. In Comprehensive Inorganic Chemistry; Schlögl, R., Ed.; Elsevier:
Waltham, MA, 2013; Vol. 7, pp 131ꢀ151.
1
1. Filoti, G.; Nistor, L.; Doca, I.; Brasoveanu, I.; Spanu, V.; Teodorescu, V.; Rotaru, P.; Blejoiu, S. I.
Revue Roumaine de Chimie 1979, 24, 1101ꢀ1105.
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 25
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
12. Kubsh, J. E.; Chen, Y.; Dumesic, J. A. J. Catal. 1981, 71, 192ꢀ200.
1
1
1
3. Kundu, M. L.; Sengupta, A. C.; Maiti, G. C.; Sen, B.; Ghosh, S. K.; Kuznetsov, V. I.; Kustova, G.
N.; Yurchenko, E. N. J. Catal. 1988, 112, 375ꢀ383.
4. Pereira, A.; Berrocal, G.; Marchetti, S.; Albornoz, A.; De Souza, A.; Rangel, M. J. Mol. Catal. A:
Chem. 2008, 281, 66ꢀ72.
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
5. Reddy, G. K.; Gunasekera, K.; Boolchand, P.; Dong, J.; Smirniotis, P. G. J. Phys. Chem. C 2011,
115, 7586ꢀ7595.
16. Reddy, G. K.; Gunasekera, K.; Boolchand, P.; Smirniotis, P. G. J. Phys. Chem. C 2011, 115, 920ꢀ
30.
9
1
7. Reddy, G. K.; Boolchand, P.; Smirniotis, P. G. J. Phys. Chem. C 2012, 116, 11019ꢀ11031.
18. Keiski, R. L.; Salmi, T. Appl. Catal. A: General 1992, 87, 185ꢀ203.
1
9. Natesakhawat, S.; Wang, X.; Zhang, L.; Ozkan, U. S. J. Mol. Catal. A: Chemical 2006, 260, 82ꢀ94.
0. Lee, J. Y.; Lee, D. ꢀ.; Hong, Y. ꢀ.; Lee, K. ꢀ. Int. J. of Hydrogen Energy 2011, 36, 8173ꢀ8180.
2
21. Martis, V.; Oldman, R.; Anderson, R.; Fowles, M.; Hyde, T.; Smith, R.; Nikitenko, S.; Bras, W.;
Sankar, G. Phys. Chem. Chem. Phys. , 2013, 15, 168 2013, 15, 168ꢀ175.
2
2
2. Quadro, E. B.; Dias, M. d. L. R.; Maria, A.; Amorim, M.; Rangel, M. d. C. J. Braz. Chem. Soc.
999, 10, 51ꢀ59.
1
3. Edwards, M. A.; Whittle, D. M.; Rhodes, C.; Ward, A. M.; Rohan, D.; Shannon, M. D.; Hutchings,
G. J.; Kiely, C. J. Phys. Chem. Chem. Phys. 2002, 4, 3902ꢀ3908.
24. Martos, C.; Dufour, J.; Ruiz, A. Int. J. of Hydrogen Energy 2009, 34, 4475ꢀ4481.
25. Khaleel, A.; Shehadi, I.; AlꢀShamisi, M. Coll. Surf. A: Phys. Eng. Aspects 2010, 355, 75ꢀ82.
26. Domka, F.; Basinska, A.; Przystajko, W.; Fiedorow, R. Surf. Technol. 1984, 21, 101ꢀ108.
27. Tsokov, P.; Blaskov, V.; Klissurski, D.; Tsolovski, I. J. Mat. Sci. 1993, 28, 184ꢀ188.
2
8. Rangel, M. d. C.; Sassaki, R. M.; Galembeck, F. Catal. Lett. 1995, 33, 237ꢀ254.
9. Chen, L. S.; Lu, G. L. J. Mat. Sci. 1999, 34, 4193ꢀ4197.
2
30. Kumar, P.; Idem, R. Energy & Fuels 2007, 21, 522ꢀ529.
3
1. Scariot, M.; Francisco, M. S. P.; Jordao, M. H.; Zanchet, D.; Logli, M. A.; Vicentini, V. P. Catal.
Today 2008, 133-135, 174ꢀ180.
ACS Paragon Plus Environment

Page 21 of 25
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
32. Boudjemaa, A.; Auroux, A.; Boumaza, S.; Trari, M.; Cherifi, O.; Bouarab, R. React. Kinet. Catal.
Lett. 2009, 98, 319ꢀ325.
3
3
3
3
3. Reddy, G. K.; Boolchand, P.; Smirniotis, P. G. J. Catal. 2011, 282, 258ꢀ269.
4. Reddy, G. K.; Smirniotis, P. G. Catal. Lett. 2011, 141, 27ꢀ32.
5. Meshkani, F.; Rezaei, M. J. Ind. Eng. Chem. 2014, 20, 3297ꢀ3302.
6. Busca, G.; Lorenzelli, V. Mat. Chem. 1980, 5, 213ꢀ224.
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
37. Busca, G.; Lorenzelli, V. React. Kinet. Catal. Lett. 1980, 15, 273ꢀ278.
3
3
4
8. Busca, G.; Lorenzelli, V. J. Catal. 1980, 66, 155ꢀ161.
9. Busca, G.; Lorenzelli, V. Mat. Chem. 1982, 7, 89ꢀ126.
0. Lorenzelli, V.; Busca, G.; Sheppard, N.; AlꢀMashta, F. Journal of Molecular Structure 1982, 80,
181ꢀ186.
4
1. Ramis, G.; Busca, G.; Lorenzelli, V. Mat. Chem. Phys. 1991, 29, 425ꢀ435.
42. Busca, G.; Lorenzelli, V.; Ramis, G.; Willey, R. J. Langmuir 1993, 9, 1492ꢀ1499.
4
4
3. de Faria, D.; Silva, S.; de Oliveira, M. J. Raman Spectrosc. 1997, 28, 873ꢀ878.
4. Cornell, R. M.; Schwertmann, U. The Iron Oxides: Structure, Properties, Reactions, Occurences and
Uses; WileyꢀVCH Verlag GmbH & Co. KGaA: Weinheim, 2003; , pp 703.
4
4
4
5. Baltrusaitis, J.; Jensen, J. H.; Grassian, V. H. J. Phys. Chem. B 2006, 110, 12005ꢀ12016.
6. de Smit, E.; Weckhuysen, B. M. Chem. Soc. Rev. 2008, 37, 2758ꢀ2781.
7. Boudjemaa, A.; Daniel, C.; Mirodatos, C.; Trari, M.; Auroux, A.; Bouarab, R. C. R. Chimie 2011,
14, 534ꢀ538.
48. Busca, G.; Lorenzelli, V. J. Chem. Soc. , Faraday Trans. 1992, 88, 2783ꢀ2789.
49. Topsoe, H.; Dumesic, J. A.; Boudart, M. J. Phys. -Paris 1974, 35, 411ꢀ413.
50. CherkezovaꢀZheleva, Z.; Mitov, I. J. Physics: Conf. Series 2010, 217, 012044ꢀ012047.
51. Zanchet, D.; Rodella, C. B.; Lopes, L. J. S.; Logli, M. A.; Vicentini, V. P.; Wen, W.; Hanson, J. C.;
Rodriguez, J. A. AIP Conf. Proc. 2009, 1092, 25ꢀ28.
5
2. Kendelewicz, T.; Kaya, S.; Newberg, J. T.; Bluhm, H.; Mulakaluri, N.; Moritz, W.; Scheffler, M.;
Nilsson, A.; Pentcheva, R.; Brown, G. E. J. Phys. Chem. C 2013, 117, 2719ꢀ2733.
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 25
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
53. Patlolla, A.; Carino, E. V.; Ehrlich, S. N.; Stavitski, E.; Frenkel, A. I. ACS Catal. 2012, 2, 2216ꢀ
2223.
5
4. Boreskov, G. K. Discuss. Faraday Soc. , 1966, 41, 263ꢀ276.
5
5. Muzykantov, V. S.; Popovskii, V. V.; Boreskov, G. K.; Panov, G. I.; Shkrabina, R. A. Sov. Appl.
Mech. 1969, 5, 344ꢀ350.
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
5
5
5
6. Kasatkina, L. A.; Mekipelov, V. N.; Zhivotenko, N. N. Kinet. Katal. 1973, 14, 363ꢀ371.
7. Udovic, T. J.; Dumesic, J. A. J. Catal. 1984, 89, 314ꢀ326.
8. Lund, C. R. F.; Kubsh, J. E.; Dumesic, J. A. In Solid State Chemistry in Catalysis; Grasselli, R. K.,
Ed.; American Chemical Society: Washington, DC, 1985; Vol. 279, pp 313ꢀ338.
59. Rethwisch, D. G.; Dumesic, J. A. Appl. Catal. 1986, 21, 97ꢀ109.
60. Weckhuysen, B. M.; Wachs, I. E. J. Chem. Soc. , Faraday Trans. 1996, 92, 1969ꢀ1973.
61. Kubsh, J. E.; Dumesic, J. A. AIChE J. 1982, 28, 793ꢀ800.
62. Tinkle, M.; Dumesic, J. A. J. Phys Chem. 1984, 88, 4127ꢀ4130.
6
3. Rethwisch, D. G.; Dumesic, J. A. Langmuir 1986, 2, 73ꢀ79.
4. Tinkle, M.; Dumesic, J. A. J. Catal. 1987, 103, 65ꢀ78.
6
65. Rethwisch, D. G.; Dumesic, J. A. In Catalytic Activation of Carbon Dioxide; Ayers, W. M., Ed.;
American Chemical Society: Washington, DC, 1988; Vol. 363, pp 102ꢀ122.
6
6. Zhu, M.; Wachs, I. E. ACS Catalysis 2016, 6, 1764ꢀ1767.
67. Boreskov, G. K. Kinet. Katal. 1970, 11, 374ꢀ382.
6
6
7
7
8. Boreskov, G. K. Kinet. Katal. 1980, 11, 312.
9. Rethwisch, D. G.; Dumesic, J. A. J. Phys Chem. 1986, 90, 1625ꢀ1630.
0. Oki, S.; Mezaki, R. J. Phys. Chem. 1973, 77, 1601ꢀ1605.
1. Armstrong, E. F.; Hilditch, T. P. In In A Study of Catalytic Actions at Solid Surfaces. IV. The
Interaction of Carbon Monoxide and Steam as Conditioned by Iron Oxide and by Copper;
Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and
Physical Character; The Royal Society: London, 1920; Vol. 97, pp 265ꢀ273.
7
2. Mezaki, R.; Oki, S. J. Catal. 1973, 30, 488ꢀ489.
7
3. Chinchen, G. C.; Logan, R. H.; Spencer, M. S. Appl. Catal. 1984, 12, 97ꢀ103.
ACS Paragon Plus Environment

Page 23 of 25
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
74. Chinchen, G. C.; Logan, R. H.; Spencer, M. S. Appl. Catal. 1984, 12, 69ꢀ88.
7
5. Chinchen, G. C.; Logan, R. H.; Spencer, M. S. Appl. Catal. 1984, 12, 89ꢀ96.
6. Botes, F. G. Applied Catalysis A: General 2007, 328, 237ꢀ242.
7
77. Vimont, A.; Daturi, M.; Winfield, J. M. In Functionalized Inorganic Fluorides: Synthesis,
Characterizations & Properties of Nanostructured Solids; Tressaud, E., Ed.; WileyꢀBlackwell:
Chichester, UK, 2010; Vol. Chap. 4, pp 101ꢀ139.
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
7
8. Lee, E.; Wachs, I. E. J. Phys. Chem. C 2007, 111, 14410ꢀ14425.
9. Frenkel, A. I.; Wang, Q.; Marinkovic, N.; Chen, J. G.; Barrio, L.; Si, R.; LopezꢀCamara, A.; Estrella,
A. M.; Rodriguez, J. A.; Hanson, J. C. J. Phys. Chem. C 2011, 115, 17884–17890.
8
8
0. Tao, F. Chem. Commun. 2012, 48, 3812ꢀ3814.
1. Wen, C.; Zhu, Y.; Ye, Y.; Zhang, S.; Cheng, F.; Liu, Y.; Wang, P.; Tao, F. ACS Nano 2012, 6, 9305ꢀ
9313.
8
2. Zhang, S.; Nguyen, L.; Zhu, Y.; Zhan, S.; Tsung, C. ꢀ.; Tao, F. Acc. Chem. Res. 2013, 46, 1731ꢀ
1739.
8
3. Hardcastle, F. D.; Wachs, I. E. J. Mol. Catal. 1988, 46, 173ꢀ186.
4. CieslakꢀGolonka, M. Coord. Chem. Rev. 1991, 109, 223ꢀ249.
8
85. Vuurman, M. A.; Wachs, I. E.; Stufkens, D. J.; Oskam, A. J. Mol. Catal. 1993, 80, 209ꢀ227.
8
8
6. Kagwade, S. V.; Clayton, C. R.; Halada, G. P. Surf. Interface Anal. 2001, 31, 442ꢀ447.
7. Biesinger, M. C.; Payne, B. P.; Grosvenor, A. P.; Lau, L. W. M.; Gerson, A. R.; Smart, R. S. C.
Appl. Surf. Sci. 2011, 257, 2717ꢀ2730.
8
8
8. Guo, Q.; Moller, P. J. Surface Science Letters 1995, 340, L999ꢀL1004.
9. Lemire, C.; Bertarione, S.; Zecchina, A.; Scarano, D.; Chaka, A.; Shaikhutdinov, S.; Freund, H. ꢀ.
Physical Review Letters 2005, 94, 166101.
9
0. Lutz, H. D.; Moller, H.; Schmidt, M. J. Mol. Struct. 1994, 328, 121ꢀ132.
9
1. Oh, S. J.; Cook, D. C.; Townsend, H. E. Hyperfine Interactions 1998, 112, 59ꢀ66.
92. Frenkel, A. I.; Cross, J. O.; Fanning, D. M.; Robinson, I. K. J. Synchrotron Rad. 1999, 6, 332ꢀ334.
9
9
3. Hadjiivanov, K. I.; Vayssilov, G. N. Adv. Catal. 2002, 47, 307ꢀ511.
4. Vimont, A.; Leclerc, H.; Mauge, F.; Daturi, M.; Lavalley, J. C.; Surble, S.; Serre, C.; Ferey, G. J.
Phys. Chem. C 2007, 111, 383ꢀ388.
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 25
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
95. Busca, G.; Lorenzelli, V. Materials Chemistry 1982, 7, 89.
9
6. Busca, G.; Lamotte, J.; Lavalley, J. C.; Lorenzelli, V. J. Am. Chem. Soc. 1987, 109, 5197ꢀ5202.
7. Hadjiivanov, K. I.; Busca, G. Langmuir 1994, 10, 4534ꢀ4541.
9
98. Glisenti, A. J. Chem. Soc. , Faraday Trans. 1998, 94, 3671ꢀ3676.
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
9
9. Landon, J.; Demeter, E.; Inoglu, N.; Keturakis, C.; Wachs, I. E.; Vasic, R.; Frenkel, A. I.; Kitchin, J.
R. ACS Catal. 2012, 2, 1793ꢀ1801.
100. Stephens, J. S.; Cruickshank, D. W. J. Acta Cryst. 1970, B26, 222ꢀ226.
1
1
01. Busca, G. J. Raman Spectrosc. 2002, 33, 348ꢀ358.
02. Robbins, M.; Wertheim, G. K.; Sherwood, R. C.; Buchanan, D. N. E. J. Phys. Chem. Solids 1971,
32, 717ꢀ729.
1
1
03. Sharp, J. C.; Yao, Y. X.; Campbell, C. T. J. Phys. Chem. C 2013, 117, 24932ꢀ24936.
04. Joseph, Y.; Kuhrs, C.; Ranke, W.; Ritter, M.; Weiss, W. Chem. Phys. Lett. 1999, 314, 195ꢀ202.
ACS Paragon Plus Environment

Page 25 of 25
ACS Catalysis
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
O
O
O
O
O
O
O
O
Cr6+
Cr6+
Cr6+
Cr6+
O
O
O
O
O
O
O
O
WGS
Fe O
Before Reaction
Fe O
2 3
3 4
During Reaction
ACS Paragon Plus Environment