Elucidation of the Reaction Mechanism for Higherature Water Gas Shift over an Industrial-Type Copper-Chromium-Iron Oxide Catalyst

Article abstract of DOI:10.1021/jacs.9b03516

The water gas shift (WGS) reaction is of paramount importance for the chemical industry, as it constitutes, coupled with methane reforming, the main industrial route to produce hydrogen. Copper-chromium-iron oxide-based catalysts have been widely used for the higherature WGS reaction industrially. The WGS reaction mechanism by the CuCrFeO_x catalyst has been debated for years, mainly between a "redox" mechanism involving the participation of atomic oxygen from the catalyst and an "associative" mechanism proceeding via a surface formate-like intermediate. In the present work, advanced in situ characterization techniques (infrared spectroscopy, temperature-programmed surface reaction (TPSR), near-ambient pressure XPS (NAP-XPS), and inelastic neutron scattering (INS)) were applied to determine the nature of the catalyst surface and identify surface intermediate species under WGS reaction conditions. The surface of the CuCrFeO_x catalyst is found to be dynamic and becomes partially reduced under WGS reaction conditions, forming metallic Cu nanoparticles on Fe₃O₄. Neither in situ IR not INS spectroscopy detect the presence of surface formate species during WGS. TPSR experiments demonstrate that the evolution of CO₂ and H₂ from the CO/H₂O reactants follows different kinetics than the evolution of CO₂ and H₂ from HCOOH decomposition (molecule mimicking the associative mechanism). Steady-state isotopic transient kinetic analysis (SSITKA) (CO + H₂¹⁶O → CO + H₂¹⁸O) exhibited significant ¹⁶O/¹⁸O scrambling, characteristic of a redox mechanism. Computed activation energies for elementary steps for the redox and associative mechanism by density functional theory (DFT) simulations indicate that the redox mechanism is favored over the associative mechanism. The combined spectroscopic, computational, and kinetic evidence in the present study finally resolves the WGS reaction mechanism on the industrial-type higherature CuCrFeO_x catalyst that is shown to proceed via the redox mechanism.

Full text of DOI:10.1021/jacs.9b03516

Subscriber access provided by UNIV OF LOUISIANA
Article
Elucidation of the Reaction Mechanism for High-Temperature Water-
Gas Shift over an Industrial-type Copper-Chromium-Iron Oxide Cata-lyst
Felipe Polo-Garzon, Victor Fung, Luan Nguyen, Yu Tang, Franklin Tao, Yongqiang Cheng, Luke L. Daemen,
Anibal J. Ramirez-Cuesta, Guo Shiou Foo, Minghui Zhu, Israel E. Wachs, De-en Jiang, and Zili Wu
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b03516 • Publication Date (Web): 25 Apr 2019
Downloaded from http://pubs.acs.org on April 25, 2019
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 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 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.
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 12
1
2
3
4
5
6
Journal of the American Chemical Society
Elucidation of the Reaction Mechanism for High-Temperature Water-
Gas Shift over an Industrial-type Copper-Chromium-Iron Oxide
Catalyst
7
8
9
Felipe Polo-Garzon,[1] Victor Fung,[1][2] Luan Nguyen,[3] Yu Tang,[3] Franklin Tao,[3] Yongqiang Cheng,[4] Luke L.
Daemen,[4] Anibal J. Ramirez-Cuesta,[4] Guo Shiou Foo,[1]ǂ Minghui Zhu,[5]ǁ Israel E. Wachs,[5] De-en Jiang,[2] Zili
Wu*[1]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[1]Chemical Sciences Division and Center for Nanophase Materials Sciences, Oak Ridge National Laboratory.
Oak Ridge, TN 37831 (USA)
[2]Department of Chemistry, University of California. Riverside, CA 92521 (USA)
[3]Departments of Chemical Engineering and Chemistry, The University of Kansas. Lawrence, KS 66047 (USA)
[4]Neutron Scattering Division, Oak Ridge National Laboratory. Oak Ridge, TN 37831 (USA)
[5]Operando Molecular Spectroscopy & Catalysis Laboratory, Department of Chemical and Biomolecular
Engineering, Lehigh University. Bethlehem, PA 18015 (USA)
KEYWORDS: Water-gas shift, reaction mechanism, iron oxide-based catalyst, NAP-XPS, FTIR, neutron scattering, TPSR,
SSITKA, DFT.
ABSTRACT: The water-gas shift (WGS) reaction is of paramount importance for the chemical industry, as it constitutes, coupled
with methane reforming, the main industrial route to produce hydrogen. Copper-Chromium-Iron oxide-based catalysts have been
widely used for the high-temperature WGS reaction industrially. The WGS reaction mechanism by the CuCrFeOx catalyst has been
debated for years, mainly between a ‘redox’ mechanism involving the participation of atomic oxygen from the catalyst and an
‘associative’ mechanism proceeding via a surface formate-like intermediate. In the present work, advanced in situ characterization
techniques (infrared spectroscopy, Temperature Programmed Surface Reaction (TPSR), near-ambient pressure XPS (NAP-XPS)
and inelastic neutron scattering (INS)) were applied to determine the nature of the catalyst surface and identify surface intermediate
species under WGS reaction conditions. The surface of the CuCrFeOx catalyst is found to be dynamic and becomes partially
reduced under WGS reaction conditions forming metallic Cu nanoparticles on Fe3O4. Both in situ IR and INS spectroscopy do not
detect the presence of surface formate species during WGS. TPSR experiments demonstrate that the evolution of CO2 and H2 from
the CO/H2O reactants follows different kinetics than the evolution of CO2 and H2 from HCOOH decomposition (molecule
mimicking the associative mechanism). Steady-state isotopic transient kinetic analysis (SSITKA) (CO + H216O  CO + H218O)
exhibited significant 16O/18O scrambling, characteristic of a redox mechanism. Computed activation energies for elementary steps
for the redox and associative mechanism by Density Functional Theory (DFT) simulations indicate that the redox mechanism is
favored over the associative mechanism. The combined spectroscopic, computational and kinetic evidence in the present study
finally resolves the WGS reaction mechanism on the industrial-type high temperature CuCrFeOx catalyst that is shown to proceed
via the redox mechanism.
INTRODUCTION
Since its development in 1914 by Bosch and Wild,1 the
industrial water-gas shift catalytic reaction (WGS) has become
a paramount reaction for industrial purification of H2 streams.
Initially, the catalytic WGS reaction was used to remove CO
from the H2 stream for the Haber-Bosch ammonia synthesis,1
and later for tuning the H2/CO ratio of syngas for production of
hydrogen, methanol synthesis, and for Fischer-Tropsch
synthesis of hydrocarbons.2-4 WGS is a reversible and
exothermic reaction and, therefore, it is thermodynamically
favored at low temperatures and kinetically favored at high
temperature.
Thus, the reaction is typically performed in two stages:
high-temperature WGS (HT-WGS) and low-temperature WGS
(LT-WGS) . LT-WGS (~190 – 250 °C) is performed over
copper-zinc-oxide-based catalysts, and HT-WGS (~350-450
°C) is performed over iron-oxide-based catalysts. Improving
the performance of WGS catalysts is of upmost importance for
hydrogen-based energy sustainability, as aimed by the United
States Government.5 Now, even more, since the dramatic
increase in the production of shale gas (mostly methane) has
revitalized industrial processes that involve utilization of
methane, such as steam methane reforming (SMR)-SMR is
currently the main process for the production of H2.6
CO + H2O ↔ CO2 + H2
ΔH = -41 kJ/mol (1)
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 12
The present work focuses on studying the HT-WGS
CuCrFeOx catalyst contains 3 wt. % CuO, 8 wt. % Cr2O3, and
89 wt. % Fe2O3.
Argon and 5% 16O2/He were purchased form Airgas. 2%
CO/Ar/He and 10% CO/He were purchased from Air Liquide.
H218O (97 atom % 18O) and formic acid (≥96%) were purchased
from Sigma-Aldrich.
1
2
3
4
catalytic reaction while presenting some comparisons to the
reverse HT-WGS (CO2 + H2 ↔ CO + H2O). For simplicity,
from this point on, HT-WGS will be referred to as WGS, and
reverse HT-WGS will be referred to as RWGS.
Rational design and optimization of catalysts, particularly
those with industrial relevance, demand characterization of the
catalysts under reaction conditions given the dynamics of
catalyts with reaction environments, with the purpose of
comprehending the underlying reaction mechanism. Such
understanding allows targeting the rate-determining-step
(RDS) to enhance reaction rates. Many authors have studied the
reaction mechanism for WGS. In some catalysts an
“associative” mechanism has been proposed, where CO and
H2O come together to form an “associated” reaction
intermediate as shown via IR (e.g., surface formate).7 In other
cases, a redox mechanism has been suggested, where CO is
oxidized by surface-oxygen and H2O subsequently fills an
oxygen vacancy and produces H2.7-9 The redox mechanism is
the most commonly claimed in current reports for WGS over
Fe oxide-based catalysts.10 Recent publications have shown that
for the CuCrFeOx catalyst, chromium oxide is a textural
promoter that increases the Fe3O4 surface area and reduces
sintering. Copper, on the other hand, increases the catalytic
activity over a wide range of temperature, lowers activation
barriers and, thus, increases turnover frequencies (TOF),
making copper a chemical promoter.6, 11-15 TPSR studies over
Cr-Fe oxide catalyst suggested a redox mechanism for the WGS
reaction.13 Although CuCrFeOx is an industrially relevant
catalyst, the WGS reaction mechanism over this catalyst
remains unknown, which is the focus of the present study.
In the present paper, the surface intermediates present on
a CuCrFeOx catalyst are characterized under WGS reaction
conditions. Kinetic measurements, near-ambient pressure X-
ray photoelectron spectroscopy (NAP-XPS), infrared (IR)
spectroscopy, inelastic neutron scattering (INS) and density
functional theory (DFT) calculations are coupled to provide
molecular level insights into the reaction mechanism. IR
spectroscopy shows the surface formate intermediate is present
only on the transient surface, but is not present on the catalyst
during steady-state reaction. The absence of surface formate
species is supported by INS measurements on the spent catalyst.
INS provides evidence for the existence of hydrides and
hydroxyls on the surface. These surface intermediates were
considered when describing the reaction pathway using DFT.
DFT calculations showed the redox mechanism is energetically
favored over the associative mechanism. Temperature-
programmed surface reaction (TPSR) experiments support this
conclusion. Steady-state isotopic transient kinetic analysis
(SSITKA) showed extensive scrambling of isotopic oxygen
atoms, characteristic of a redox mechanism.
5
6
7
8
9
Near-ambient Pressure X-ray photoelectron spectroscopy
(NAP-XPS). In situ NAP-XPS was performed on a lab-based
system. Detailed information about the system can be found in
a previous publication.17 Briefly, the NAP-XPS system is
equipped with an Al K-alpha X-ray source (SPeCS XR-MF).
Photoelectrons collection was done using a differentially
pumped photoelectron energy analyzer (Scienta-Omicron
R4000 HIPP-2) with a 0.8 mm pre-lense cone aperture.
Resolution of the analyzer was set at 100 eV pass energy.
Heating of the sample was done using an IR laser irradiating the
back of the sample holder. Sample temperature was monitored
using a K-type thermocouple spot-welded on the sample holder.
Reaction gases were introduced into the reaction cell using all-
metal leak valves.
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
Infrared (IR) spectroscopy. IR spectroscopy measurements
were performed using a Thermo Nicolet Nexus 670 FTIR
spectrometer with an MCT detector. Each spectrum was
recorded with 32 scans at a resolution of 4 cm-1. The sample
was loaded into a crucible, and the crucible was placed inside a
Diffuse Reflectance Infrared Fourier Transform Spectroscopy
(DRIFTS) cell (Pike Technologies). The sample was initially
dehydrated in situ under 0.8 % O2/(Ar+He) at 440 °C for 2 h,
and in some cases, activation under WGS reaction conditions
was performed. The reactant mixture for the WGS reaction was
delivered to the catalyst sample using a saturator to bubble 12
mL/min of Ar through liquid water, and flowing 18 mL/min of
2% CO/(Ar+He) to deliver a CO/H2O ratio of 1. After
activation, background spectra were collected in Ar at different
temperatures. Later, the reactant mixture (H2O + CO) was
reintroduced to the cell, and spectra were collected at different
temperatures. Temperatures were increased stepwise, without
holding at each temperature any longer than 15 min. DRIFTS
spectra for formic acid desorption were collected in a similar
fashion, but bubbling liquid formic acid in 30 mL of Ar instead.
Inelastic neutron scattering (INS). INS spectra were collected
at the VISION beamline of the Spallation Neutron Source
(SNS) at Oak Ridge National Laboratory. The catalyst sample,
CuCrFeOx, was first exposed to O2 flow at 400 °C for 1h to
remove any surface contaminations and possible water species
from exposure to air. It was then transferred to a Swagelok steel
cell for gas loading experiments in a He glovebox. The cell was
then loaded in a closed cycle refrigerator (CCR) and cooled
down to the base temperature (5 K) for background
measurement. For WGS, 1 to 1 ratio of CO and H2O was loaded
into the cell, and for RWGS, 1 to 1 ratio of H2 and CO2 was
loaded. The reactions were conducted at 350 °C in the closed
cell. After reaction for 2 hours, in some cases the sample was
cooled down to 100-110 °C and evacuated for 15 min to remove
weakly-bonded and gas phase moisture, while surface
intermediates remained. The sample was quenched in liquid
nitrogen and then cooled to 5 K before collection of any INS
spectrum.
EXPERIMENTAL DETAILS
Materials. The catalyst sample (CuCrFeOx) was synthesized as
described previously.16 Briefly, the catalyst was synthesized
using the ammonia assisted coprecipitation method. The
precursors (iron nitrate, chromium nitrate and copper nitrate)
were mixed and dissolved in deionized water. Aqueous
ammonia was added to the solution dropwise to adjust the pH
to 8.5. The dark brown precipitate was aged overnight and
filtered off. The filtered precipitate was dried at 80 °C for 12 h
and calcined at 400 °C for 3 h in stagnant air. The fresh
Reference INS spectra for water and formic acid were
collected separately to aid the spectral interpretation. In
addition, a spectrum was collected after exposing the CuCrFeOx
sample to H2 at room temperature and 3 bar for 30min. This
2
ACS Paragon Plus Environment

Page 3 of 12
Journal of the American Chemical Society
spectrum constitutes the reference for the identification of only control the temperature inside the bed. Inside the furnace,
1
2
3
4
5
6
7
8
hydride species. All gases were purchased from Airgas.
a U-tube quartz reactor was placed containing 20 mg of the
catalyst. The catalyst was held in place using quartz wool (see
Scheme S1 for the SSITKA setup).
Density functional theory (DFT) calculations. All periodic
DFT calculations were performed with the Vienna ab initio
Simulation Package (VASP).18-19 The Perdew-Burke-Ernzerhof
(PBE)20 functional of generalized-gradient approximation
(GGA) with Hubbard U correction (U=3.8 eV)21 was used for
electron exchange and correlation. The electron-core
interaction was described using the projector-augmented wave
method (PAW).22-23 A Monkhorst–Pack k-point sampling
scheme24 (3x3x1) is used. A plane wave cutoff of 400 eV is used
in all calculations. Transition states were found using the
climbing-image nudged elastic band (CI-NEB) method25. The
structure modeled was spinel Fe3O4 with the plane (111)
exposed at the surface, providing a Fe+2/Fe+3 terminated
surface. The bottom two layers of the simulated slab were fixed.
A vacuum layer of 15 Å was set between slabs to minimize the
interactions of the periodic images.
Before the isotopic switch, the catalyst sample was
pretreated in situ at 410 °C under 50 mL/min of 5% O2/He for
2 h, and under WGS reaction conditions (2 µL/min H2O, 16.5
mL/min 10% CO/He, 33.5 mL/min 2%Ar/He) at 410 °C for at
least 2 h to activate the catalyst and achieve steady-state
conditions. The isotopic switch, H216O ↔ H218O, was possible
using two Nexus 3000 syringe pumps from Chemyx. An
OmniStar mass spectrometer from Pfeiffer Vacuum was used
to analyze the outlet from the reactor. The masses followed
during the switch were: 2 (H2), 4 (He), 17 (H2O), 19 (H218O),
22 (Ne), 28 (CO), 30 (C18O), 40 (Ar), 44 (CO2), 46 (C16O18O),
48 (C18O2). These masses were selected to avoid overlap of
signals as much as possible; when the overlap of signals was
unavoidable, proper decoupling of contributions from each
component was carried out.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Temperature programmed surface reaction (TPSR). TPSR
experiments for WGS and formic acid decomposition were
performed in an Altamira Instruments system (AMI-200). The
catalyst sample (20 mg) was loaded into a U-tube quartz reactor
and held in place using quartz wool. Before each TSPR
experiment, the catalyst sample was initially pretreated in situ
at 440 °C under 50 mL/min of 5% O2/He for 2 h, and
subsequently under WGS reaction conditions (0.66 µL/min of
H2O and 50 mL/min of 1.8% CO/(Ar+He) for a H2O:CO ratio
of ~1) at 415 °C for at least 2 h to activate the catalyst. Liquid
feeds (water or formic acid) were introduced to the reactive
system using a Nexus 3000 syringe pump (Chemyx) at 50 °C
and then the catalyst was heated up to 415°C at a ramping rate
of 5 °C/min. The outlet from the reactor was analyzed using an
OmniStar mass spectrometer from Pfeiffer Vacuum. The
masses followed were: 2 (H2), 18 (H2O), 28 (CO), 29 (formic
acid), and 44 (CO2). The contributions from different
components to a single mass were decoupled and subtracted.
Steady-state isotopic transient kinetic analysis (SSITKA).
SSITKA experiments were performed in a home-built system
equipped with back pressure regulators and digital pressure
readers to ensure the pressure inside the reactor is constant
during the isotopic switch at 1.7 atm. The temperature in the
lines was held at least at 120 °C to avoid condensation of
components. This home-built system was coupled to the
furnace of an AMI-200 unit (Altamira Instruments, Inc.), to
RESULTS AND DISCUSSION
In Situ NAP-XPS
NAP-XPS measurements were conducted to obtain
information on the surface composition and oxidation states of
the catalyst components and surface species under WGS
reaction conditions (Figure 1). The fresh sample was initially
analyzed in UHV at 110 °C; then, it was exposed to a mixture
of CO and H2O (1:1). The temperature was increased stepwise
from 200 °C to 350 °C and spectra were collected at each
temperature. Later, the sample was cooled down in CO+H2O to
110 °C and a spectrum was collected; followed by a spectrum
under vacuum at the same temperature. The sample was held at
each temperature for at least 20 min (2 h at 350 °C) before data
were collected.
As observed in Figure 1e-f, a signal for Cu0 (~932.5 eV)
is detected at the usual reaction temperature (~350 °C) and it
remains after cooling down the system to 110 °C (Figure 1f),
which evidences the formation of metallic copper at the surface.
The formation of metallic Cu nanoparticles at the surface under
reaction conditions, product of surface enrichment from Cu
cations initially dispersed in the bulk, has been previously
observed.16, 26-27 An overlayer of an FeOx-shell on metallic Cu
Hydroxyl and chemisorbed CO
Gas CO
Cu0
Lattice O
Cx
Cu 2p3/2
Fe 2p3/2
C 1s
O 1s
g) Vacuum – 110C
f) WGS, 110 °C
e) WGS, 350 °C
d) WGS, 300 °C
c) WGS, 250 °C
b) WGS, 200 °C
a) UHV, 110 °C
538 536 534 532 530 528
938 936 934 932 930 928
718 716 714 712 710 708 706
294 292 290 288 286 284 282
Binding Energy (eV)
Figure 1 NAP-XPS spectra of CuCrFeOx for Cu2p3/2, Fe2p3/2, O1s, and C1s under a) UHV; and b-f) under WGS reaction (0.5 Torr
of CO and 0.5 Torr of H2O) as the temperature was increased to b) 200 °C, c) 250 °C, d) 300 °C, e) 350 °C, and f) decreased to 110
°C afterwards. After collecting spectra under reaction conditions, a final spectrum was collected under g) vacuum at 110 °C.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 12
nanoparticles has also been previously detected due to the so- The NAP-XPS results clearly show that the catalyst under
1
2
3
4
5
6
7
8
called strong-metal-support-interaction (SMSI) between
reducible oxides and metals.16, 28
reaction conditions is significantly different when compared
with the fresh catalyst (formation of metallic Cu and Fe3O4,
surface carbonaceous deposits and surface intermediates) and,
therefore, any relevant characterization must be performed
under in situ/operando reaction conditions. The Fe3O4 crystal
structure will also be used for the DFT calculations presented
later in this paper.
Partial reduction of the surface iron oxide from Fe2O3 to
Fe3O4 is characterized by the disappearance of a peak at 719 eV
from Fe2O3 and slight broadening of the peak at ~708 eV.16 The
NAP-XPS data collected for Fe 2p3/2 (Figure 1) did not cover
binding energies out to 719 eV and it is hard to distinguish
broadening at 708 eV. However, a previous study shows that
iron in the CuCrFeOx catalyst reduces to Fe3O4 under WGS
reaction conditions.16 Under RWGS at 350 °C, the reduction of
Fe2O3 to Fe3O4 during RWGS at 350 °C was confirmed via
NAP-XPS experiments (Figure S1b). The peak at 719 eV for
the dehydrated Fe2O3 sample (treated under O2 at 400 °C)
disappears as soon as the RWGS mixture is introduced at 350
°C, confirming the transition to Fe3O4. Based on data in the
present paper for the reduction of iron oxide in CuCrFeOx under
RWGS (Figure S1b) and previously reported data on the same
reduction under WGS16, we assume that the sample under study
in the present paper also reduces from Fe2O3 to Fe3O4 under
WGS reaction conditions. Further, the present sample and the
one used in the literature were synthesized following the same
procedure.16 From Figure 1, a weak signal for hydroxyl and/or
chemisorbed CO is observed at~532 eV along with surface
carbonaceous deposits (CxHy, 284.4 eV) and gas phase CO
(~291 eV).29 The surface carbon is considered a spectator since
the relative signal does not change as the reaction temperature
is increased. Notably, surface formate species were not
observed under reaction conditions, with expected binding
energy of 288 - 289 eV in the C 1s region.29
In Situ IR Spectroscopy
Initially, the surface species on the transient CuCrFeOx
catalyst (before reaching steady-state reconstruction), after only
O2 treatment at 440 °C for 2h, were probed via IR spectroscopy
both under WGS (Figure 2a) and during formic acid desorption
(Figure 2b). On the transient catalyst, surface formate species,
characterized by the OCO mode at ~1586 cm-1 and the C-H
vibrational modes around ~3000 cm-1 (Figure 2b),30-33 are
observed under the WGS reactant mixture (Figure 2a) up to 200
°C, although the WGS products are not generated at such low
temperatures as shown later via TPSR experiments. At 300 °C,
the features of surface formate due to formic acid
adsorption/desorption are barely distinguishable (Figure 2b);
therefore, we propose that the peak at 1561 cm-1 under 300 °C
WGS reaction conditions (Figure 2a) is due to surface
carbonates. The evident change in the baseline of the spectra
under WGS at normal operating temperatures (300-360 °C) is
due to the reconstruction of the surface. In the spectra for fomic
acid desorption (Figure 2b), the modes between 1388-1719 cm-
1, other than 1586 cm-1, are assigned to molecularly adsorbed
formic acid.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To compare the surface species under steady-state WGS
reaction conditions, the CuCrFeOx catalyst was pretreated
under O2 at 440 °C for 2 h, followed by exposure to CO+H2O
at 364 °C for 1 h, before recording IR spectra under WGS
reaction conditions (Figure 2c). Assignment of IR modes
corresponding to surface formate species on the steady-state
sample (after pretreatment under WGS at 364 °C) was done
through adsorption (and subsequent desorption) of formic acid
Similar NAP-XPS experiments were performed under
RWGS reaction conditions (Figure S1). Evidence of surface
intermediates in the C 1s region is observed, but the poor signal-
to-noise ratio and the broadness of the peak (288-290 eV)
makes it nearly impossible to identify these species. Some
possible candidates would be surface HCOO, HOCO, CO3 or
CO2-δ species.29
a)
c)
0.03
0.07
WGS - fresh catalyst
WGS - s.s. catalyst
0.01
carbonate
carbonate
0.06
0.05
0.04
360 °C, 30 min
360 °C, 5 min
300 °C, 5 min
cool-down 120 °C, 5min
360 °C, 110min
0.02
360 °C, 5min
Figure 2 In situ DRIFTS study of adsorbed species on a fresh CuCrFeOx catalyst a) under WGS gas mixture and b) after formic acid
desorption. In situ DRIFTS study of adsorbed species on a reconstructed (activated under WGS reaction conditions at 364 °C for 1h)
0.03
200°C, 5 min
120°C, 5 min
25 °C, 10 min
300 °C, 5min
CuCrFeOx catalyst c) under WGS gas mixture and d) after formic acid desorption.
0.01
200 °C, 5min
0.02
130 °C, 5min
0.01
0.005
25 °C, 10min
0.00
3200 3000 2800
0.00
3200 3000 2800
1800 1600 1400 1200
1800 1600 1400 1200
Wavenumber (cm-1)
Wavenumber (cm-1)
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
b)
d)
Formic Acid desorption - s.s. catalyst
Formic acid desorption - fresh catalyst
C-H, formate
0.01
360 °C, 5min
300°C, 5min
200°C, 5min
120°C, 5min
25°C, 10min
380 °C, 5min
300 °C, 5min
200 °C, 5min
120 °C, 5min
25 °C, 10 min
C-H, formate
ACS Paragon Plus Enviro0n.0m1 ent
4
0.00
3200 3000 2800
1800 1600 1400 1200
3200 3000 2800
1800 1600 1400 1200
Wavenumber (cm-1)
Wavenumber (cm-1)

Page 5 of 12
Journal of the American Chemical Society
(see Figure 2d). The presence of surface formate-like species
a)
based on the C-H vibrational mode (around ~2910 cm-1) was
not observed under WGS over the whole temperature range.
The characteristic OCO mode of surface formate species (1582
cm-1) was not clearly evident, and the broad band in this vicinity
is rather presumed to be due to surface carbonates, along with
bands at 1692 and 1717 cm-1.
1
2
3
4
5
6
7
8
1.0
0.8
0.6
0.4
0.2
0.0
H2O
350 °C, 2 h reacted
350 °C, 2 h reacted + 100 °C pump
Hydride reference
Formic acid reference
H2O reference
WGS
hydride
formic
acid
i)
A very important observation is that for the catalyst
conditioned under steady-state WGS reaction (see Figure 2d),
the bands for molecularly adsorbed formic acid (1719, 1392,
1178 cm-1) are more prominent than the band from the OCO
mode for surface formate species (1582 cm-1). This is opposite
to the observation for the fresh catalyst (Figure 2b) and suggests
that formic acid dissociation into surface formate proceeds
easier on the fresh (transient) catalyst surface.
A better IR signal-to-noise ratio was obtained for the
transient catalyst (Figure 2a,b) compared to the steady-state
catalyst (Figure 2c,d), which is probably due to the presence of
surface carbon deposits as indicated by NAP-XPS data in
Figure 1 and reduction of the surface that significantly
decreased sample reflectance.
ii)
iii)
v)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
iv)
-0.2
400 800 1200 1600 2000 2400 2800 3200 3600 4000
Energy Transfer (cm-1)
b)
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
350 °C, 3 h reacted + RT pump
Difference spectrum
350 °C, 30 min reacted + 100 °C pump
Hydride reference
Formic acid reference
formic acid
RWGS
surface
H2O
hydride + OH
hydride
i)
From the in situ IR studies, it is concluded that surface
formate species are present on the transient catalyst up to 200
°C, but there is no conclusive evidence for the presence of
surface formate species at any temperature under the steady-
state WGS reaction. Thus, although surface formate readily
forms on the initial oxidized surface it has difficulty forming on
the surface of the WGS-activated catalyst because of the
distinctly different types of surface formed as presented above
in the NAP-XPS section.
When comparing the fresh and the spent catalyst using
spectroscopic techniques (NAP-XPS and IR), it is evident that
the change in the catalyst structure under reaction conditions
affects the formation of surface species. The extent of
reconstruction of samples with other Cu loadings and particle
size would have an impact on the WGS reaction mechanism
only if the nature of the active sites changes (not their density).
Careful investigation of related CuCrFeOx samples with
different Cu loading, Cu particle size and extent of Cu-FeOx
interaction is certainly of great interest but it lies beyond the
scope of the present work.
iii)
ii)
v)
iv)
400 800 1200 1600 2000 2400 2800 3200 3600 4000
Energy Transfer (cm-1)
Figure 3 INS spectra collected at 5 K a-i) after WGS reaction
(CO:H2O=1) over CuCrFeOx at 350 °C for 2 h, a-ii) after WGS
reaction over CuCrFeOx at 350 °C for 2 h followed by
evacuation at 100 °C; as well as reference spectra for a-iii)
hydride (on CuCrFeOx), a-iv) water and a-v) formic acid. b-i)
after RWGS reaction (CO2:H2=1) over CuCrFeOx at 350 °C for
3 h followed by evacuation at room temperature (RT), b-ii) after
RWGS over CuCrFeOx at 350 °C for 30 min followed by
evacuation at 110 °C, b-iii) Difference spectrum between b-i
and b-ii; as well as reference spectra for b-iv) hydride (on
CuCrFeOx) and b-v) formic acid.
In Situ INS
Further information about the possible presence of
surface formate species during WGS and RWGS over the
CuCrFeOx catalyst was sought using in situ neutron
spectroscopy. This is the first time INS measurements have
been conducted for catalytic WGS system, and it is one of few
reports on neutron spectroscopy for metal oxide catalysts of
industrial relevance.34
As indicated in Figure 3a, the WGS reaction was
performed at 350 °C and INS spectra were collected at 5 K
before and after pumping out the chamber at 100 °C. Reference
spectra for water, formic acid and hydride species were also
collected and shown in Figure 3a. The broad band for the
hydride reference at 770-780 cm-1 is assumed to be mostly due
to surface and/or bulk hydride species since the contribution of
hydroxyl species is minor, given that OH modes on iron oxide
have been reported in the vicinity of 933-951 and 3385-3640
cm-1.35 From Figure 3a, it is readily determined that water is
removed by pumping out the chamber and that hydride species
are present on the catalyst surface. Hydroxyl species are
probably also present on the surface due to the broadness of the
peak centered around ~850 cm-1, which might overlap with
hydroxyl modes (933-951 cm-1). The absence of the
characteristic surface formate INS vibrational mode at 280 cm-
1
(Figure 3a-v) after the WGS reaction (Figure 3a-ii) suggests
that surface formate species are not present on the catalyst after
activating the catalyst under WGS reaction conditions. The INS
formic acid reference spectrum showed good agreement with
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 12
simulation results, thus, confirming the reliability of the
a) Formic acid decomposition. CuCrFeOX
characteristic peak at 280 cm-1 (Figure S2-S3).
1
2
3
4
5
6
7
8
9
INS results after performing the RWGS reaction (Figure
3b) were similar to those recorded for WGS over CuCrFeOx.
Physisorbed water was removed from the chamber after
pumping at 110 °C. The difference spectrum, between the spent
catalyst before and after pumping at 110 °C, reveals the
characteristics of surface water (Figure 3b-iii), which are
similar to the spectrum for water collected as a reference
(Figure 3a-iv). Hydride and hydroxyl species were identified on
the surface of the catalyst after the RWGS. In addition, the 280
cm-1 band of surface formate species is absent, suggesting that
it is not present after activating the catalyst under the RWGS
reaction conditions. INS, however, provided insight into the
presence of surface hydroxyl and hydride intermediates. This is
the first spectroscopic evidence of surface hydride after
activating FeOx-based catalysts under WGS and RWGS
reaction conditions.
DFT simulations showed comparable hydrogen
adsorption energies (from -3.2e v to -3.5 eV) as both hydroxyl
groups on Fe3O4 and as hydrides on the Cu-Fe interface and
undercoordinated Cu edges (Figure S4). Therefore, Cu-H
hydride and O-H species are expected to coexist on the surface
as found with INS. The vibrational modes obtained via DFT
(Figure S5) support the existence of Cu-H and Fe-OH surface
species when compared to the INS spectra for surface species
collected after the WGS reaction at 350 °C (Figure 3a-ii), and
hydride reference (Figure 3a-iii). Bader charge for hydrogen
adsorbed on Cu(111) (Figure S4a, S5) was found to be -0.263,
suggesting it can be considered a hydride as it is slightly
negatively charged. Formation of Fe-H is both energetically
disfavored and its vibrational mode was not detected in the
experimental INS spectra.
1.0
0.8
0.6
0.4
0.2
0.0
.0E-09
4
H2
HCOOH
.0E-09
3
.0E-09
2
CO2
H2O
.0E-09
1
CO
0E+00
0
55 155 255 355
55 155 255 355
Tbed (°C)
Tbed (°C)
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) WGS. CuCrFeOX
1.0
0.8
0.6
0.4
0.2
0.0
.2E-0182
CO
Normalized
H2
.0E-09
9
2 - TPR2
Normalized
CO2
44 - TPR2
.0E-09
6
H2O
H2
.0E-09
3
CO2
0
0E+00
55 155 255 355
55 155 255 355
Tbed (°C)
Tbed (°C)
Figure 4 TPSR (5 °C/min, 20 mg of catalyst) after activation at
440 °C under 50 mL/min of 5% O2/He for 2 h and under WGS
reaction conditions (CO:H2O=1) at 415 °C for at least 2h. a)
Formic acid decomposition over CuCrFeOx (0.45 µL/min of
formic acid, 45 mL/min Ar), and b) WGS over CuCrFeOx.
16
18
SSITKA (H2 O + C16O ↔ H2 O+ C16O)
To study the oxygen-exchange properties of the
CuCrFeOx catalyst system under WGS reaction conditions,
SSITKA experiments were undertaken. The steady-state
reaction conditions under a regular reactant mixture (C16O +
H216O + Ar + He) were achieved after 2 h at 410 °C, and then
an isotopic reactant mixture was introduced (C16O + H218O +
Ne + He). When the isotopic reactant mixture was introduced
(CO + H218O), multiple isotopic product species were
generated: C16O2, C16O18O, and C18O2. Even, the reactants C18O
and H216O were ‘generated’ from the RWGS reaction (Figure
S6). The generation of these species undoubtedly unveils a
dynamic catalyst surface that facilitates the scrambling of
18O/16O atoms. Since the scrambling of oxygen-atoms between
gas phase species and the surface is evident, the surface will
arrive to a steady-state coverage of 18O and 16O species, given
that the surface and bulk before the isotopic switch contains
only 16O atoms.
TPSR
TPSR experiments were also performed over the
CuCrFeOx catalyst, pre-activated under WGS reaction
conditions, to study the kinetic evolution of the products (CO2
and H2). Under the assumption that both CO2 and H2 proceed
from the decomposition of a common surface intermediate (e.g.
HCOO*), the kinetic evolution of CO2 and H2 should have the
same light-off temperature and kinetic behavior.36 Indeed,
TPSR of formic acid decomposition (Figure 4a) shows a
monotonic and similar evolution of CO2 and H2. In contrast, the
evolution of CO2 precedes the evolution of H2 during WGS
(Figure 4b).
TPSR curves from the CuCrFeOx catalyst (Figure 4a,b)
exhibit a light-off temperature for H2 at ~135 °C for formic acid
decomposition, and ~155 °C for WGS. This indicates that the
reaction mechanism for WGS over the CuCrFeOx catalyst does
not proceed through a surface formate intermediate, as this
intermediate decomposes at the lower temperature of ~135 °C.
6
ACS Paragon Plus Environment

Page 7 of 12
Journal of the American Chemical Society
water as the rate-determining-step. In contrast, the redox
pathway is suggested to be dominant on Fe3O4, with the
desorption of CO2 as the rate-determining-step.
C16O18
O
40
44
46
48
17
19
28
30
22
a)
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
-0.1
1
2
3
4
5
6
7
8
C18
O
C18O2
H218O
The presence of metallic Cu nanoparticles on the surface
of CrFeOx has shown to increase the TOF value relative to bare
CrFeOx (Cr has been shown to be a structural promoter,
increasing surface area and stability but not the TOF).13 In
addition, it has been shown that a SMSI effect is present where
H216O
Ne
C16
O
28
FeOx migrates onto the metallic Cu nanoparticles.16,
Therefore, we constructed a model system that captures the
combined Cu-Fe3O4 active sites. A 10-atom Cu cluster was
supported on the (111) surface of a Fe3O4 slab, to model the
interaction of intermediates with the supported-metal system.
This cluster is large enough to exhibit metal-like characteristics
(as shown experimentally via NAP-XPS experiments, Figure 1
and S1), yet computationally affordable together with the oxide
support. Distorted planar Cu(111)-like configuration was the
most energetically preferred configuration of the cluster (Figure
S7).
It was found that CO adsorbs the strongest on the Cu-Fe
interface (-1.60 eV; Figure 6), in agreement with the results
from Xue et al.39 The adsorption of CO on the pristine Fe3O4
(111) surface was found to be -1.11 eV, consistent with the
literature.40 Additionally, CO adsorption on the pure Cu(111)
surface is -0.94 eV,37 which is in agreement with our results for
the supported cluster, -0.8 to -1.0 eV. According to our results,
H2O adsorption is weak, ranging from -0.1 to -0.4 eV for Cu
and Cu-Fe sites (Figure 6), in agreement with Gokhale et al.37
who reported weak H2O adsorption on Cu(111) (-0.18eV). In
contrast, H2O was found to dissociate with a negligible barrier
on Fe3O4, again agreeing with the literature.38 In short, the Cu-
Fe interface is responsible for preferential CO adsorption,
whereas water adsorbs and easily dissociates on the Fe3O4
support.
Ar
CO2
9
89750
89810
1
89870
2
0
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
Relative Time (min)
100%
100
50
45
40
35
30
25
20
15
10
5
b)
H2
H2
C18O2
90%
90
CO2
CO2
CO18
O
80
80%
H218O
70
70%
H2O
60
60%
50
50%
H2O
40
40%
C18O
30
30%
20%
20
CO
CO
10%
10
0
0%
0
CO+H2O
CO+H218O
Feed into
reactor:
CO+H2O
CO+H218
O
H2O conversion
CO conversion
Figure 5 a) Normalized SSITKA results for WGS over an
activated CuCrFeOx sample. Regular reactant mixture: 2
µL/min H2O, 16.5 mL/min 10% CO/He, 33.5 mL/min
2%Ar/He. Isotopic mixture: 2 µL/min H218O, 16.5 mL/min
10% CO/He, 33.5 mL/min 2%Ne/He. Reaction conditions: 21
mg of catalyst, 1.7 atm, 410 °C. b) Concentration of gas phase
species and conversion before and after the switch.
Adsorption site
Adsorbate
Cu
-1.01 eV
Fe3O4
Cu-Fe interface
-0.76 eV
-1.11 eV -1.60 eV
CO
From the normalized responses at 410 °C (Figure 5a), it
is seen that the signals of two of the isotopes of carbon dioxide
(C16O2 and C18O2) plateau out only ~45 s after the isotopic
switch (note the sharp decay/increase of the Ar/Ne signal
reflecting the system response). Along with C16O2 and C18O2,
other species (C18O, H218O, H216O, C16O) require a similar time
to reach a steady concentration. The MS-signal for C16O18O,
however, reaches a steady-state value in a significantly shorter
time after the isotopic switch at 410 °C (~ 20 s). The time it
takes for a certain species to reach a steady production rate is a
convolution of the surface residence time and the time that it
takes for the 16O/18O scramble process to reach steady-state. It
appears that the fast formation of C16O18O reflects the high
WGS reaction rate at 410 °C and the other isotope products
arise from further adsorption and oxygen scrabling. The
SSITKA results point to a single concise conclusion: the intense
scrambling of atoms leads to significant fractions of different
isotopes, including isotopes of the reactants (Figure 5b). This
behavior supports a redox mechanism.
-1.46 eV
-0.09 eV
-0.36 eV
-0.31 eV
H2O
Figure 6 Adsorption configurations for CO and H2O on
different adsorption sites of the simulated Cu/Fe3O4 catalyst.
Color code: Fe: purple, Cu: coral, O: red, C: gray, H: white.
DFT calculations: Redox vs. associative mechanism
DFT studies have been previously performed for WGS on
the Cu(111)37 and Fe3O4 surfaces.38 On Cu(111), the associative
mechanism is suggested to be dominant, with O-H cleavage of
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 12
Next, the energetics in the reaction pathways via the A-i), comparable to 0.61 eV on Cu(111) and lower than the 1.24
1
2
3
4
associative and the redox mechanism were studied. As shown
in Figure 7a, the highest barrier in the associative mechanism is
0.73 eV for the A-i step (formation of HOCO from OH and CO
on the surface), while the highest barrier in the redox
eV barrier on Fe3O4. The O-H bond in the formed HOCO*
intermediate can be readily cleaved by Cu sites in the present
model with a barrier of 0.66 eV, following a facile rotation of
the molecule (Figure 7b, A-ii and A-iii). Next, CO2 desorption
5
6
7
8
Associative
Associative
Redox
a)
R-iii)
R-ii)
0
-1
-2
-3
-4
0
-1
-2
-3
-4
R-i)
A-i)
A-ii)
A-iii)
Re
TS
9
TS
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
TS
TS
TS
H (Enthalpy)
G (Gibbs Free Energy)
H (Enthalpy)
G (Gibbs Free Energy)
)
)
)
)
)
s
*
*
*
)
)
)
)
)
)
)
s
*
S
S
s
s
s
S
S
n
s
s
s
s
s
a
s
O
O
O
T
T
*
O
a
d
d
a
o
T
T
a
d
d
a
a
a
i
C
C
C
g
a
*
t
a
g
(
C
g
a
*
*
a
g
g
g
g
(
(
(
(
+
(
(
(
(
(
(
a
t
O
O
+
)
O
O
C
*
2
2
O
O
2
2
2
2
O
O
o
2
H
H
C
O
O
e
c
O
O
O
C
C
-
r
(
-
O
O
O
O
C
C
-
O
O
-
+
C
C
-
H
C
C
)
C
C
C
C
a
f
r
C
O
-
S
+
+
*
O
e
+
y
+
+
+
*
*
)
*
T
H
H
c
u
s
H
H
c
a
*
H
H
s
(
f
d
+
n
r
+
+
*
O
*
*
a
*
a
c
u
(
C
H
O
s
H
H
O
a
b
O
O
2
v
u
H
s
H
(
O
*
O
Redox Mechanism
Eact=0.31 eV
Associative Mechanism
HO* + CO*  HOCO* + *
b)
Osubsurface* Osurface
*
Eads=-2.81 eV
ΔH=0.20 eV
ΔH=0.32 eV
Eact=0.73 eV
ΔH=0.43 eV
R-i)
A-i)
Osurface* + CO*  CO2* + *
Eact=0.39 eV
HOCO* rotation
ΔH=-0.33 eV
ΔH=-0.03 eV
Eact=0.34 eV
A-ii)
R-ii)
*
HOCO* + *  H* + CO2
H2O* + * OH* + H*
Eact=0.66 eV
Eads=-0.79 eV
ΔH=-1.84 eV
A-iii)
R-iii)
Figure 7 a) Potential energy diagram for WGS over the simulated Cu/Fe3O4. b) Structures of elementary steps involved in the
associative (A-i through A-iii) and redox (R-i through R-iii) pathways. Labels (A-i, R-i, and so on) in a) are correlated to the
structures in b). The step OH*+H*  H*+H* in the redox mechanism involves O-H dissociation and filling an oxygen-vacancy.
Gibbs free energy was calculated at 400 °C. Color code: Fe: purple, Cu: coral, O: red, C: gray, H: white, Diffusing-O: green
mechanism is 0.39 eV for the R-ii step (formation of CO2 from
O and CO on the surface). Although the enthalpy for CO2
desorption in the redox pathway presents a high barrier (1.31
eV), the energy required for desorption reduces considerably
(Gibbs free energy) after the entropy contribution at 400 °C is
included. The Gibbs free energy at 400°C for other
desorption/adsorption processes were also calculated and the
results do not change the conclusion that the redox mechanism
is preferred based on the DFT energetics.
In the redox mechanism, lattice oxygen can easily diffuse
from the sublayer to the top-layer of the catalyst (Eact = 0.31 eV,
Figure 7b, R-i), and then undergo reaction with adsorbed CO to
form adsorbed CO2 (Eact =0.39 eV, Figure 7b, R-ii), which
subsequently desorbs. Following CO2 desorption, the oxygen-
vacancy can be easily refilled by H2O. First, H2O adsorbs on the
Fe-Fe bridge, and later dissociates with a negligible barrier
Figure 7b, R-iii). Dissociation of the surface hydroxyl species
is also facile (Eact =0.16 eV, not shown in Figure 7).
occurs. Although surface formate species were not directly
observed in the most favorable pathway, the transition from the
HOCO* species to the HCOO* species (formate) has a barrier
of ~0.92 eV (Figure S8b).
H2 formation can occur on Cu/Fe3O4 through various
mechanisms. One mechanism is shown in Figure S8a (Eact
0.78 eV), facilitated by facile H migration between the cluster
and the oxide (Figure S8b).
In conclusion, DFT calculations clearly show that the
favored CO adsorption at Cu-Fe sites is crucial for driving the
subsequent steps in both the associative and redox pathways,
with the redox mechanism favored.
=
CONCLUSIONS
The water gas-shift reaction has been investigated over an
industrial-type
CuCrFeOx
catalyst
using
in
situ
In the associative pathway, surface OH* and CO* can
both strongly co-adsorb near the Cu-Fe interface, where the
HO-CO association barrier was found to be 0.73 eV (Figure 7b,
characterization, reaction kinetics and computational tools.
Efforts were devoted to discerning between the redox and
associative mechanism, with the latter characterized by the
8
ACS Paragon Plus Environment

Page 9 of 12
Journal of the American Chemical Society
existence of an active surface ‘formate-like’ intermediate
(HCOO*, HOCO*).
In situ NAP-XPS surface analysis demonstrates that the
All authors have given approval to the final version of the
manuscript.
Acknowledgement
1
2
surface of the CuCrFeOx catalyst is dynamic with the catalyst
undergoing dramatic reconstruction during WGS/RWGS
reaction conditions from an initial oxidized surface to a partially
reduced surface with the existence of metallic Cu nanoparticles
on Fe3O4. Furthermore, in situ DRIFTS experiments reveal that
surface formate intermediates only form on the initial oxidized
surface and not on the reconstructed catalyst, which reflects a
strong depdence of WGS/RWGS on the specific state of the
catalyst surface. The in situ INS spectra after activation under
WGS exhibit the presence of surface Cu-H and hydroxyl
species, supported by assignments from DFT simulations, but
not the vibrations from surface formate species. TPSR
experiments also support a redox mechanism since the kinetic
evolution of CO2 preceeds that of H2 for WGS over CuCrFeOx,
while the evolution of CO2 and H2 is concurrent for surface
formate decomposition that is representative of the associative
mechanism. SSITKA experiments (H216O + C16O ↔ H218O+
C16O) exhibited extensive scrambling of 18O/16O atoms at the
surface that is characteristic of a redox mechanism.
DFT calculations found the redox mechanism to be
energetically favored over the associative mechanism. The slow
step for the associative mechanism is likely the surface HO-
CO* coupling (Eact=0.73 eV), whereas the slow step for the
redox mechanism is likely the surface CO* oxidation by Fe3O4
lattice O (Eact=0.39 eV). Water dissociation is facile on Fe-O
sites near the Cu cluster with little or no barriers. The DFT
simulations also show that either mechanism proceeds easily at
the Cu-Fe3O4 interface, the catalytic active site for
WGS/RWGS.
This research is sponsored by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, Chemical
Sciences, Geosciences, and Biosciences Division. Part of the
work including FTIR and kinetic measurement was conducted
at the Center for Nanophase Materials Sciences, which is a
DOE Office of Science User Facility. This research used
resources of the National Energy Research Scientific
Computing Center, a DOE Office of Science User Facility
supported by the Office of Science of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231. Neutron
scattering measurements were performed in the Spallation
Neutron Source facility, which is a DOE Office of Science
User Facility. The catalysts were synthesized and initially
characterized at Lehigh University with financial support from
National Science Foundation Grant CBET – 1511689.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Notice: This manuscript has been authored by UT-Battelle,
LLC under Contract No. DE-AC05-00OR22725 with the U.S.
Department of Energy. The United States Government retains
and the publisher, by accepting the article for publication,
acknowledges that the United States Government retains a non-
exclusive, paid-up, irrevocable, world-wide license to publish
or reproduce the published form of this manuscript, or allow
others to do so, for United States Government purposes. The
Department of Energy will provide public access to these
results of federally sponsored research in accordance with the
DOE Public Access Plan (http://energy.gov/downloads/doe-
public-access-plan).
In summary, a combination of in situ spectroscopic,
kinetic and computational studies was utilized to shed light
upon the WGS reaction mechanism, redox vs. associative, over
the industrial-type CuCrFeOx catalyst. All the experimental
studies and also the DFT calculations support the redox
mechanism as the dominant reaction pathway on the activated
CuCrFeOx catalyst without any supporting evidence found for
the associative mechanism.
Notes
The authors declare no competing financial interest.
REFERENCES
1.
Bosch, C.; Wild, W. US Patent
1115776A, 1914.
2.
Ratnasamy, C.; Wagner, J. P., Water Gas
Shift Catalysis. Catal. Rev. 2009, 51 (3), 325-
440.
3.
Reaction. Catal. Rev. 1980, 21 (2), 275-318.
4. Rhodes, C.; Hutchings, G. J.; Ward, A.
Newsome, D. S., The Water-Gas Shift
ASSOCIATED CONTENT
Supporting Information.
SSITKA-reactor scheme (Scheme S1).
NAP-XPS for RWGS (Figure S1).
INS spectra for formic acid (Figure S2).
Simulated vibrational modes of formic acid (Figure S3).
DFT calculations (Figure S4-S5, S7-S8).
SSITKA (Figure S6).
M., Water-gas shift reaction: finding the
mechanistic boundary. Catal. Today 1995, 23 (1),
43-58.
5.
Energy, U. S. D. o., The Department of
Energy Hydrogen and Fuel Cells Program Plan,
DOE/EE-0651. U.S. Government Printing Office:
Washingtion, DC 2011.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
6.
Zhu, M.; Wachs, I. E., Iron-Based
* wuz1@ornl.gov
Catalysts for the High-Temperature Water–Gas
Shift (HT-WGS) Reaction: A Review. ACS
Catal. 2016, 6 (2), 722-732.
Present Addresses
ǂW. L. Gore and Associates. Newark, DE 19711 (USA)
ǁ
State Key Laboratory of Chemical Engineering, East China
University of Science and Technology, Shanghai 200237, P. R.
China.
7.
Boudjemaa, A.; Daniel, C.; Mirodatos,
C.; Trari, M.; Auroux, A.; Bouarab, R., In situ
DRIFTS studies of high-temperature water-gas
Author Contributions
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 12
shift reaction on chromium-free iron oxide
catalysts. Comptes Rendus Chimie 2011, 14 (6),
534-538.
18.
Kresse, G.; Furthmüller, J., Efficiency of
1
2
3
4
ab-initio total energy calculations for metals and
semiconductors using a plane-wave basis set.
Comput. Mater. Sci. 1996, 6 (1), 15-50.
8.
Smith, R. J. B.; Loganathan, M.; Shantha,
M. S., A Review of the Water Gas Shift Reaction
Kinetics. Int. J. Chem. React. Eng. 2010, 8, 1-32.
19.
Kresse, G.; Furthmuller, J., Efficient
5
6
7
8
iterative schemes for ab initio total-energy
calculations using a plane-wave basis set. Phys.
Rev. B 1996, 54 (16), 11169-11186.
9.
Khan, A.; Smirniotis, P. G., Relationship
between temperature-programmed reduction
profile and activity of modified ferrite-based
catalysts for WGS reaction. J. Mol. Catal. A:
Chem. 2008, 280 (1), 43-51.
9
20.
Perdew, J. P.; Burke, K.; Ernzerhof, M.,
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
Generalized Gradient Approximation Made
Simple. Phys. Rev. Lett. 1996, 77 (18), 3865-
3868.
10.
Keturakis, C. J.; Zhu, M.; Gibson, E. K.;
Daturi, M.; Tao, F.; Frenkel, A. I.; Wachs, I. E.,
Dynamics of CrO3–Fe2O3 Catalysts during the
High-Temperature Water-Gas Shift Reaction:
Molecular Structures and Reactivity. ACS Catal.
2016, 6 (7), 4786-4798.
21.
Yu, X.; Huo, C.-F.; Li, Y.-W.; Wang, J.;
Jiao, H., Fe3O4 surface electronic structures and
stability from GGA+U. Surf. Sci. 2012, 606 (9),
872-879.
22.
Kresse, G.; Joubert, D., From ultrasoft
11.
Andreev, A.; Idakiev, V.; Mihajlova, D.;
pseudopotentials to the projector augmented-
wave method. Phys. Rev. B 1999, 59 (3), 1758-
1775.
Shopov, D., Iron-based catalysts for the water—
gas shift reaction promoted by first-row transition
metal oxides. Appl. Catal. 1986, 22 (2), 385-387.
23.
Blochl, P. E., PROJECTOR
12.
Idakiev, V.; Mihajlova, D.; Kunev, B.;
AUGMENTED-WAVE METHOD. Phys. Rev. B
1994, 50 (24), 17953-17979.
24.
points for Brillouin-zone integrations. Phys. Rev.
B 1976, 13 (12), 5188-5192.
25.
H., A climbing image nudged elastic band
method for finding saddle points and minimum
energy paths. J. Chem. Phys. 2000, 113 (22),
9901-9904.
Andreev, A., Effect of copper oxide on the
catalytic activity of iron-chromia catalyst for
water gas shift reaction. React. Kinet. Catal. Lett.
1987, 33 (1), 119-124.
Monkhorst, H. J.; Pack, J. D., Special
13.
Zhu, M.; Wachs, I. E., Determining
Henkelman, G.; Uberuaga, B. P.; Jónsson,
Number of Active Sites and TOF for the High-
Temperature Water Gas Shift Reaction by Iron
Oxide-Based Catalysts. ACS Catal. 2016, 6 (3),
1764-1767.
14.
Rhodes, C.; Peter Williams, B.; King, F.;
26.
Kappen, P.; Grunwaldt, J.-D.;
Hutchings, G. J., Promotion of Fe3O4/Cr2O3 high
temperature water gas shift catalyst. Catal.
Commun. 2002, 3 (8), 381-384.
Hammershøi, B. S.; Tröger, L.; Clausen, B. S.,
The State of Cu Promoter Atoms in High-
Temperature Shift Catalysts—An in Situ
Fluorescence XAFS Study. J. Catal. 2001, 198
(1), 56-65.
15.
Rhodes, C.; Hutchings, G. J., Studies of
the role of the copper promoter in the iron
oxide/chromia high temperature water gas shift
catalyst. Phys. Chem. Chem. Phys. 2003, 5 (12),
2719-2723.
27.
Grunwaldt, J.-D.; Kappen, P.;
Hammershoi, B. S.; Troger, L.; Clausen, B. S.,
Fluorescence EXAFS for the in situ study on the
state of promotors in catalysis. J. Synchrotron
Radiat. 2001, 8 (2), 572-574.
16.
Zhu, M.; Rocha, T. C. R.; Lunkenbein,
T.; Knop-Gericke, A.; Schlögl, R.; Wachs, I. E.,
Promotion Mechanisms of Iron Oxide-Based
High Temperature Water–Gas Shift Catalysts by
Chromium and Copper. ACS Catal. 2016, 6 (7),
4455-4464.
28.
Tauster, S. J., Strong metal-support
interactions. Acc. Chem. Res. 1987, 20 (11), 389-
394.
29.
Feria, L.; Kundu, S.; Baber, A. E.; Graciani, J.;
Vidal, A. B.; Agnoli, S.; Evans, J.; Chang, R.;
Axnanda, S.; Liu, Z.; Sanz, J. F.; Liu, P.;
Rodriguez, J. A.; Stacchiola, D. J., Importance of
the Metal–Oxide Interface in Catalysis: In Situ
Mudiyanselage, K.; Senanayake, S. D.;
17.
Nguyen, L.; Tao, F., Development of a
reaction cell for in-situ/operando studies of
surface of a catalyst under a reaction condition
and during catalysis. Rev. Sci. Instrum. 2016, 87
(6), 064101.
10
ACS Paragon Plus Environment

Page 11 of 12
Journal of the American Chemical Society
Studies of the Water–Gas Shift Reaction by
of the Cu-modified Fe3O4 catalysts. Phys. Lett. A
2015, 379 (6), 607-612.
1
2
3
4
5
6
7
8
Ambient-Pressure X-ray Photoelectron
Spectroscopy. Angew. Chem. Int. Ed. 2013, 52
(19), 5101-5105.
40.
Li, X.; Paier, J.; Sauer, J.; Mirabella, F.;
Zaki, E.; Ivars-Barceló, F.; Shaikhutdinov, S.;
Freund, H. J., Surface Termination of Fe3O4(111)
Films Studied by CO Adsorption Revisited. J.
Phys. Chem. B 2018, 122 (2), 527-533.
30.
Jacobs, G.; Davis, B. H., In situ DRIFTS
investigation of the steam reforming of methanol
over Pt/ceria. Applied Catalysis A: General 2005,
285 (1), 43-49.
9
31.
Wu, Z.; Li, M.; Overbury, S. H., On the
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
structure dependence of CO oxidation over CeO2
nanocrystals with well-defined surface planes. J.
Catal. 2012, 285 (1), 61-73.
32.
Wu, Z.; Li, M.; Mullins, D. R.; Overbury,
S. H., Probing the Surface Sites of CeO2
Nanocrystals with Well-Defined Surface Planes
via Methanol Adsorption and Desorption. ACS
Catalysis 2012, 2 (11), 2224-2234.
33.
Busca, G.; Lamotte, J.; Lavalley, J. C.;
Lorenzelli, V., FT-IR study of the adsorption and
transformation of formaldehyde on oxide
surfaces. J. Am. Chem. Soc. 1987, 109 (17),
5197-5202.
34.
Polo-Garzon, F.; Luo, S.; Cheng, Y.;
Page, K. L.; Ramirez-Cuesta, A. J.; Britt, P. F.;
Wu, Z., Neutron Scattering Investigations of
Hydride Species in Heterogeneous Catalysis.
ChemSusChem 2019, 12, 93-103.
35.
Hamilton, N. G.; Warringham, R.;
Silverwood, I. P.; Kapitán, J.; Hecht, L.; Webb,
P. B.; Tooze, R. P.; Zhou, W.; Frost, C. D.;
Parker, S. F.; Lennon, D., The application of
inelastic neutron scattering to investigate CO
hydrogenation over an iron Fischer–Tropsch
synthesis catalyst. J. Catal. 2014, 312, 221-231.
36.
Zhu, M.; Wachs, I. E., Resolving the
Reaction Mechanism for H2 Formation from
High-Temperature Water–Gas Shift by
Chromium–Iron Oxide Catalysts. ACS Catal.
2016, 6 (5), 2827-2830.
37.
Gokhale, A. A.; Dumesic, J. A.;
Mavrikakis, M., On the Mechanism of Low-
Temperature Water Gas Shift Reaction on
Copper. J. Am. Chem. Soc. 2008, 130 (4), 1402-
1414.
38.
Huang, L.; Han, B.; Zhang, Q.; Fan, M.;
Cheng, H., Mechanistic Study on Water Gas
Shift Reaction on the Fe3O4 (111) Reconstructed
Surface. J. Phys. Chem. C 2015, 119 (52), 28934-
28945.
39.
Xue, P.; Fu, Z.; Yang, Z., The density
functional theory studies on the promoting effect
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 12
TABLE OF CONTENTS
1
2
3
4
5
H2
CO2
HO* + OC*  HOCO* + *
Eact=0.73 eV
ΔH=0.43 eV
H2O
CO
6
Osurface* + CO*  CO2* + *
Eact=0.39 eV
7
8
ΔH=0.32 eV
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cu
Fe
O
C
H
12
ACS Paragon Plus Environment