Single-Site vs. Cluster Catalysis in High Temperature Oxidations

Article abstract of DOI:10.1002/anie.202102339

The behavior of single Pt atoms and small Pt clusters was investigated for high-temperature oxidations. The high stability of these molecular sites in CHA is a key to intrinsic structure–performance descriptions of elemental steps such as O₂ dissociation, and subsequent oxidation catalysis. Subtle changes in the atomic structure of Pt are responsible for drastic changes in performance driven by specific gas/metal/support interactions. Whereas single Pt atoms and Pt clusters (> ca. 1 nm) are unable to activate, scramble, and desorb two O₂ molecules at moderate T (200 °C), clusters <1 nm do so catalytically, but undergo oxidative fragmentation. Oxidation of alkanes at high T is attributed to stable single Pt atoms, and the C-H cleavage is inferred to be rate-determining and less sensitive to changes in metal nuclearity compared to its effect on O₂ scrambling. In contrast, when combustion involves CO, catalysis is dominated by metal clusters, not single Pt atoms.

Full text of DOI:10.1002/anie.202102339

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Single-Site vs Cluster Catalysis in High Temperature Oxidations
Authors: Pedro Serna, Aida Rodríguez-Fernández, Sara Yacob,
Christine Kliewer, Manuel Moliner, and Avelino Corma
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202102339
Link to VoR: https://doi.org/10.1002/anie.202102339

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
Single-Site vs Cluster Catalysis in High Temperature Oxidations
Pedro Serna,a* Aida Rodríguez-Fernández,b Sara Yacob,a Christine Kliewer,a Manuel Moliner,b*
Avelino Cormab*
a Dr. P. Serna, Dr. S. Yacob, Dr. C. Kliewer
ExxonMobil Research and Engineering Co., Corporate Strategic Research, Annandale, New Jersey 08801, United States
b A. Rodríguez-Fernández, Dr. M. Moliner, Prof. A. Corma
Instituto de Tecnología Química, Universitat Politècnica de València - Consejo Superior de Investigaciones Cientificas (UPV-CSIC), Av. de los Naranjos,
s/n, 46022 Valencia, Spain
E-mails: pedro.m.serna-merino@exxonmobil.com, mmoliner@itq.upv.es, acorma@itq.upv.es
Supporting information for this article is given via a link at the end of the document.
Abstract: The behavior of single Pt atoms and small Pt clusters for
high temperature oxidation reactions has been investigated using
CHA zeolite. Subtle changes in the atomic structure of the active sites
are responsible for drastic changes in performance driven by specific
interactions between the metal, the gas-phase components, and the
support ― in our case, a well-defined, non-reducible macro-ligand.
The high-stability of these molecular sites in CHA is a key to intrinsic
structure-performance descriptions of elemental steps such as O2
dissociation, and subsequent oxidation catalysis of industrial interest,
such as the combustion of methane, propane, and CO. Whereas
single Pt-atoms and Pt nanoparticles larger than ~ 1 nm are unable to
activate, scramble and desorb two O2 molecules in isotopic exchange
experiments at moderate T (i.e. 200⁰C), clusters < 1 nm can do so
catalytically, but are unstable against oxidative fragmentation under
reaction conditions, leading to catalyst deactivation. For oxidation
processes that require the activation of rather inert molecules at high
T, like alkanes, catalytic combustion is attributed to stable single Pt
atoms when the support is CHA, generated in situ in the O2 stream.
The alkane combustion activity is facilitated when the reacting alkane
includes weaker C-H bonds (e.g. propane vs methane), and the C-H
cleavage is inferred to be rate-determining, which is less sensitive to
changes in the active site size compared to the effect on the O2
scrambling. When the combustion involves a strong reductant such
as CO, in contrast, the catalysis is dominated by metal clusters and
nanoparticles, and not single Pt atoms, as the latter do not form stable
mononuclear carbonyl species in the zeolite. The great stability of
single Pt atoms for methane combustion, and of small Pt clusters for
CO oxidation, including upon regeneration and/or steaming, open new
opportunities for the system in a variety of high temperature oxidation
scenarios.
clusters of a few atoms.[4–6] Those active sites have drawn
significant attention by the scientific community over the last
decade,[1,5,7–9] since they can offer opportunities that more
conventional metal ensembles like nanoparticles or bulk phases
do not.[10,11] To help to stabilize the metal, a second more robust
sub-structure, called the support, is typically introduced to serve
as a solid dispersion media and an active site diluent, while
introducing mechanical strength and favorable mass and heat
transfer.[5,6]
Keeping metal active sites stable is particularly challenging in
processes that require high temperature and/or pressure, and go
through alternating oxidative and reductive conditions (e.g.
reaction vs regeneration), which favor metal agglomeration.
Moreover, catalysts may destabilize at different moments of their
lifetime for different reasons, corresponding to different
mechanisms of activity loss under variable conditions. This
requires researchers to address multiple facets of the problem
through a single material. For example, the coalescence of noble
metals by Brownian motion typically dominates in H2 streams (e.g.
during pre-activation steps or in hydrogenation/dehydrogenation
reactions),
whereas
Ostwald-ripening
via
single-atom
intermediates is usually favored in O2 (e.g. during coke burn out
operations);[3,12] and for temperatures greater than 400°C,
volatile MeOx species may also form, which must be trapped
before they start contributing to the sintering problem by
incorporation to other distant metal particles.[3,13]
Among the supports, zeolites can be adequate thanks to their high
stability, high surface area, porous nature, and the possibility to
build multifunctional catalysis.[14–20] Moreover, zeolites are well-
defined crystalline, tunable supports that can play analog roles to
ligands in organometallic and biological systems, while helping to
stabilize isolated metal atoms and clusters.[19,21–25] Several recent
publications have emphasized the dynamic nature of molecular
metal sites inside zeolites using operando characterization
techniques, showing that they may participate in complex cyclic
processes of cluster formation and cluster re-dispersion in contact
with the gas phase components.[18,26–28]
Recently, we have described the stabilization of several metals
(Pt, Pd, and bimetallic PtPd) within the cavities of highly siliceous
CHA zeolite, under severe redox stress (i.e. at temperatures as
high as 650°C, in both H2 and O2, and steam).[21,29] These
materials offer three important advantages for their use as
heterogeneous catalysts: (1) high hydrothermal stability; (2)
shape-selectivity, to discriminate molecules smaller and larger
than ~4 Å; and (3) fine-tuning of the metal speciation, from single
atoms to small clusters (i.e. 1 nm). The versatility of the metal-
Introduction
Highly-dispersed transition metal catalysts are extensively used
in chemical and petrochemical processes.[1] Their catalytic
properties are determined by the exact nature of the active
species, which often varies over time in response to temperature
and interaction with fluid phase components.[1,2] Noble metal
catalysts, in particular, tend to undergo permanent deactivation
by irreversible sintering when exposed to severe redox and
hydrothermal stress.[1,3] Many research efforts have been devoted
in recent years to improve the stability of well-dispersed transition
metals under relevant industrial conditions, which is most
problematic when the actives sites are isolated metal atoms and
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
CHA system is unusual and allows to perform remarkably well in
different industrial scenarios.
observation that highly siliceous CHA can stabilize small Pt
clusters of 0.8-1.5 nm in reducing H2 streams (Pt/CHAH2-400C
,
In this paper, we will show how to take advantage of the properties
of highly-siliceous Pt-CHA as an efficient and robust
heterogeneous oxidation catalyst. High temperature oxidation by
single-site Pt and by few-atom Pt clusters remains to a large
extent elusive due to the difficulty to sustain a significant
population of these active sites under such a harsh chemical
environment, unless the support also brings redox characteristics
(i.e. on reducible supports), which complicates the
disentanglement of the exact role of the supported metal vs the
support itself. Thus, we have evaluated here the intrinsic behavior
of single Pt atoms and small Pt clusters as oxidation species using
a non-reducible support like the CHA zeolite, in where we can
stabilize metal clusters and metal atoms.[21,29] Among the
reactions investigated here, we include the elemental O2
dissociation via 16O2/18O2 isotopic exchange, and reactions with
relevant industrial substrates such as methane, propane, and CO.
Our work stresses the importance to build structure-performance
relationships that account for subtle changes in the active sites
under reaction conditions. Understanding these dynamic
processes for high-temperature oxidation catalysts is necessary
to identify the true active species, a step forward in the
development of accurate descriptors and predictive tools for
catalyst optimization.
clusters comprising 10 to 150 atoms), but forms single Pt atoms
in O2 that are stable at temperatures as high as 650°C (Pt/CHAO2-
650C).[21,29]
In a first series of experiments, we have tested the performance
of Pt/CHA for the dissociation of O2 using 16O2/18O2 isotopic
exchange experiments. A mass-spectrometer was used to track
the evolution of the mz+ = 34 fragment, characteristic of scrambled
16O18O that can be used as a marker for O2 dissociation of two
molecules (and subsequent desorption of the exchanged product).
In these flow-through experiments, the reduced Pt/CHAH2-400C
catalyst delivered significant amounts of mz+ = 34 at 200°C (Figure
1), whereas the oxidized Pt/CHAO2-650C sample barely produced
scrambled O2 above the blank signal (Figure 1). The first
appearance of O2 scrambling with the oxidized catalyst was not
observed until the temperature reached ~340°C, and it was
+
necessary to reach ~400°C before the intensity of mz = 34
compared to that of the pre-reduced Pt/CHAH2-400C
. We
hypothesize that the absence of neighboring metal sites in the O2-
treated Pt/CHA is responsible for the lower activity to split the O–
O bond, in agreement with other recent works on Pd,[38] and Pt.[39]
1.0e-12
Pt-CHAO2-650C at 400°C
Results and Discussion
5.0e-13
Single Site vs Small cluster catalysis – O2 Dissociation
Experiments
Pt-CHAH2-400C at 200°C
Previous work by Gates et al. illustrates the effect that a single
metal-metal bond can exert on (hydrogenation) catalysis.[23,30–32]
A single Rh-Rh bond in MgO-supported Rh dimers, for example,
is responsible for 60 times activity boost in ethylene
hydrogenation when compared to single Rh atoms, due to the
improvement in the rate of H2 dissociation.[32] Interestingly, the
effect is much less pronounced when the metal sits on an
electron-withdrawing support such as HY zeolite, because then
the resultant electron-deficient single Rh atoms activate H2 more
effectively, making the role of the metal-metal bond secondary.[31]
However, for reactions that take place at high temperatures,
single noble metal atoms are generally unstable (and, thus, more
difficult to evaluate) as the metal/support interaction weakens
sufficiently to allow surface metal migration. Nevertheless, at
moderate temperatures, usually below 100°C in H2, mobility can
be limited by judicious choice of the gas phase reactants to form
stable complexes during the catalysis. This is the case, for
example, of supported Rh, Ir and Ru complexes on HY or MgO
for alkene hydrogenation or dimerization reactions, where stable
M(C2H4)2 species can be formed.[33] For higher temperatures,
researchers must typically introduce more strongly interacting
supports such as TiO2 or CeO2 to avoid that the molecular species
sinter.[3,34] These reducible supports, however, often bring
additional effects like co-activation of the reactants at the metal-
support interface,[35,36] or the supply of new redox active sites[37]
that obscure the evaluation and understanding of the intrinsic
properties of the supported metal species. Nevertheless, we can
evaluate nuclearity effects in Pt catalysts (single site vs cluster)
for high temperature oxidation catalysis using CHA zeolite as a
non-reducible support. This work builds on our previous
Pt-CHAO2-650C at 200°C
Pt-SiO2 H2-400C at 200°C
0.0e+00
0
0.2
0.4
0.6
0.8
Time (h)
+
Figure 1. TOS mass spectrometry signals of mz = 34 during flow-through
16O2/18O2 exchange experiments using (A) CHA-supported single Pt atoms at
200°C (orange diamonds); (B) CHA-supported single Pt atoms at 400°C (green
crosses); (C) CHA-supported 0.8-1.5 nm Pt clusters at 200°C (blue circles); (D)
SiO2-supported 1.5-2 nm Pt nanoparticles at 200°C (black lines).
However, the activity of the oxidized Pt/CHAO2-650C sample for
16O2/18O2 isotopic exchange remained constant over >1.5 h during
the exchange experiment at 400°C (Figure S1), whereas the
activity of the reduced version at 200°C continued to drop over
+
the same period until the mz =34 signal became almost
indistinguishable from zero (Pt/CHAO2—650C at 200°C in Figure 1).
This result shares some similarities with the work recently
reported by the Cargnello’s group on Pd/Al2O3, where it was
inferred that PdO nanoparticles deactivate during methane
combustion experiments due to a re-dispersion of the metal oxide
phase into (inactive) single Pd atoms.[38] We observe a similar
deactivation pattern with the reduced Pt/CHAH2-400C in the isotopic
O2 stream at 200°C, but a more detailed analysis of EXAFS data
taken during this process reveals that a significant fraction of the
Pt remains as small clusters in these conditions, with a drop in the
average Pt-Pt coordination number from ~ 7 to ~ 4 (Table S1).[29]
The decrease in the number of Pt-Pt bonds is accompanied by a
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
growth of Pt-O bonds without the appearance of new long-
distance heavy backscatters signals, which allows us to rule out
the formation of Pt-O-Pt ensembles. STEM images of the reduced
Pt/CHAH2-400C catalyst after 2 h in O2 at 200°C further confirm the
presence of nanoparticles in the 1-1.5 nm range (Figure 2B).
Indeed, to accomplish complete re-dispersion of these 1-1.5 nm
Pt clusters, temperatures greater than 450°C must be employed
(Table S1).[29] By combination of EXAFS and time-resolved
HAADF-STEM images, we also showed that only the smallest Pt
clusters (below ~1 nm) re-disperse into single Pt atoms when the
conditions are relatively mild.[29] Thus, the deactivation
Table 1. 16O2/18O2 isotopic exchange experimentsa using various Pt-containing
catalysts.
Reaction Temperature
mz+=34 Signal
(a.u.)c
Entry
Sampleb
(°C)
1
2
3
4
5
6
7
Pt/CHAH2-400C
Pt/CHAH2-400C
Pt/CHAO2-650C
Pt/CHAO2-650C
Pt/SiO2 H2-400C
Pt/SiO2 H2-400C
K-Pt-MFIH2-400C
200
400
200
400
200
400
200
100-8d
95e
7.2e
97e
mechanism previously proposed for alumina-supported PdO
[38]
nanoparticles in methane combustion experiments
is
3.4e
insufficient to explain the complete loss of catalytic activity of the
reduced Pt/CHA sample in our O2 scrambling experiments. We
note, however, that the temperature used by the Cargnello’s
group in their methane combustion work (775°C) is sufficiently
high to cause full re-dispersion of the metal according to our
previous data for Pd,[21] which supports their conclusions.
1708e
1475-0d
a Feed composition: 1%16O2, 1%18O2, 98% N2
b WHSV = 1.5 h-1 for 0.4 wt % Pt/CHA; 3 h-1 for 0.8 wt % Pt/SiO2; and 4 h-1 for
0.4 wt % K-Pt-MFI
c Mass spectrometer signal after background subtraction
d The first number corresponds to the signal at zero TOS, and the second after
60 min
e
Mass-spectrometer signals reported 30 min after reaching the temperature
setpoint, for which the mz+=34 signal was shown virtually steady; experiments
at 400⁰C were performed following O2-scrambling at 200⁰C without any other
thermal treatment in between, typically, using a 5⁰C/min temperature ramp.
To distinguish between these possibilities, we synthesized a third
catalyst, consisting of ~1.5 to 3 nm Pt nanoparticles supported on
SiO2 (Figure S2). This system is an ideal benchmark because 1)
it lacks subnanometric metal clusters and single atoms due to the
weakly interacting nature of the silica support; 2) it avoids large Pt
nanoparticles when properly prepared;[40] and 3) it is stable in O2
up to approximately 400°C, as demonstrated by H2 chemisorption
experiments. Interestingly, when the reduced Pt/SiO2 catalyst was
exposed to the 16O2/18O2 mixture at 200°C, as in the experiments
conducted with the reduced Pt/CHAH2-400C catalyst, no isotopic
exchange was observed at all (Pt/SiO2 H2-400C at 200°C in Figure
+
1). The temperature had to be raised to 300°C to detect the mz
= 34 fragment above blank levels. At 400°C, a constant, intense
signal was obtained, qualitatively superior to that of single Pt
atoms in CHA (entry 6 in Table 1). The lack of any activity or
deactivation profile at 200°C constitutes a strong proof that
clusters >1 nm cannot close a single turnover during O2 activation,
and that the deactivation profile of the reduced Pt/CHAH2-400C in
Figure 1 is not the result of a gradual surface passivation. Abrupt
deactivation is, indeed, a more likely scenario for these systems
following O2 dissociation. O2 dissociation is anticipated to occur
on Pt surfaces with a rather low-energy barrier.[41] We infer, thus,
that the lack of 16O2/18O2 exchange activity on the Pt/SiO2 likely
responds to the formation of oxygen adatoms too strongly
bounded to the catalyst surface after the first O2 dissociation
event. These inferences hold reasonable for the Pt/CHAO2-650C
sample, based on CO oxidation literature that suggest that single
Pt atoms can dissociate O2,[42,43] with a notable difference that the
activation of two oxidizing molecules in 16O2/18O2 scramble
experiments is probably hindered when each site only brings one
metal center. In contrast, the observation that Pt/CHAH2-400C is
active, but unstable over a rather long period of time (>1h) under
otherwise identical reaction conditions, unambiguously
demonstrates that there is a second, more gradual, deactivation
mechanism due to sites that are present in the Pt/CHA catalyst,
Figure 2. TEM image of a Pt/CHA sample after: (A) reduction in H2 at 400⁰C for
2 h (Pt/CHAH2-400C); (B) reduction in H2 at 400⁰C for 2 h followed by reaction with
50 sccm of 2%O2 in N2 at 200⁰C for 1.5 h; (C) reduction in H2 at 400⁰C for 2 h
followed by reaction with 50 sccm of 2%O2 in N2 at 500⁰C for 1.5 h; (D) reduction
in H2 at 400⁰C for 2 h followed by reaction with 40 sccm of N2, 100 sccm of
10%O2 in N2, and 3 sccm of 100% CH4 at 650⁰C for 24 h.
At this point, we postulated two possibilities to understand the
deactivation kinetics of the reduced Pt/CHAH2-400C in the 16O2/18O2
stream (Figure 1): a) at 200°C, the 16O2/18O2 scramble only
happens on Pt clusters smaller than approximately 1 nm, which
are unstable and gradually form inactive single Pt atoms (in this
scenario, any co-existing Pt clusters > 1 nm is postulated inactive
to catalyze 16O2/18O2 scrambling); or b) two independent
deactivation phenomena are taking place―one that turns small,
active Pt clusters into inactive Pt single atoms and, additionally,
one by which the slightly larger Pt clusters (i.e. > 1nm) are initially
active but somehow passivate over time in the presence of the O2.
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
but not in Pt/SiO2. This leads us to postulate that subnanometric
Pt clusters in Pt/CHA are the only contributing species to the
16O2/18O2 exchange activity in Figure 1, with both single Pt atoms
and larger Pt nanoparticles acting as mere spectators (at least
when the temperature is moderate, i.e. 200°C). We find
remarkable that single Pt atoms and >1 nm Pt clusters show a
similar onset reaction temperature (300-400°C) despite the
obvious differences in their nature/structure, while clusters <1 nm,
structurally in between, are much more active.
allow to rule out the formation of water concomitantly to the
release of 16O18O (Figure S3). Although the mechanism of metal
migration and stabilization inside CHA is not fully understood,
recent investigations show via Density Functional Theory that the
preference of Pt single atoms to grow inside zeolites can be
kinetically hindered by the proper density of suitable rings, Si:Al
ratio, and the loading of Pt atoms.[47]
Single Site vs Small cluster catalysis – Methane and propane
oxidation experiments
Why/how the smallest metal clusters are able to scramble
16O2/18O2 at a rather low temperature remains unclear, but we
were able to demonstrate that the scrambling process is not
stoichiometric, but catalytic (i.e. that more than one turnover per
Pt atom is attainable). While so far we have been unable to
sustain catalytic 16O2/18O2 scrambling activity at a temperature as
low as 200°C, we managed to significantly increase the amount
of 16O18O produced during our exchange experiments by
increasing the amount of initial subnanometric Pt active sites in
the catalyst. This objective was accomplished by rational
selection of K-Pt-MFI as the O2-scrambling catalyst, which we
recently showed contains most of the Pt in the form of <0.8 nm
clusters (Fig S2).[19] Then, a notably higher initial concentration of
A catalyst initially containing single Pt atoms in CHA (i.e.
Pt/CHAO2-650C) was contacted with a mixture of O2/CH4/N2 at
increasing temperatures. First evidences of CO2 formation were
observed in the 435 to 475°C temperature range (Figure 3),
thereby slightly higher than the temperature at which O2 isotopic
exchange was observed with this catalyst (~340°C). These results
suggest that the combustion of methane is not limited by the O2
dissociation step but, likely, the C–H cleavage. The intrinsic
behavior of single Pt atoms in oxidation catalysis has remained
elusive due to the difficulty to stabilize and investigate these
isolated species at high temperatures on supports that lack redox
functionality (it has been otherwise reported for reducible
supports, which help to stabilize the single metal atoms, but also
have a direct contribution to the catalysis).[48] To gather more
information about the alkane activation process, we performed
additional combustion experiments with the Pt/CHAO2-650C sample
using propane as the reacting hydrocarbon, which includes
weaker C–H bonds compared to methane. Propane reacted in the
presence of O2 and the single-atom Pt species at 200-300°C
(entry 3 in Table 2), with almost quantitative conversion of the
limiting reagent at 300°C. These temperatures are notably lower
than those needed to combust methane in Figure 3 (435°C), and
lower than the temperature needed to scramble oxygen in the
16O2/18O2 exchange experiments (~340°C). These data is indeed
consistent with a situation where the C–H cleavage in CH4 is the
rate determining step.
+
mz = 34 was observed in our experiment, despite a decay in
activity over to practically baseline levels in a 1 h period,as in the
Pt/CHA case (Table 1, Entry 7). We postulate that oxidative
fragmentation is also responsible for the loss of activity over time
in this case (currently under investigation) but, in any case, we
can determine that 0.058 mmol of O2 were scrambled over 0.012
mmol of Pt in the reactor at the end of the process, which is
unambiguously indicative of catalysis. On the other hand, the data
demonstrate that catalytic 16O2/18O2 scrambling also happens on
the Pt/CHAO2-650C sample, characterized by the lack of Pt-Pt
bonds according to EXAFS (within the experimental error),
+
provided that the temperature is high enough (stable mz = 34
signal at 400°C, Figure 1, and Figure S1).
We see plausible that the Al sites of the zeolite, or any other
source of negative charge from the zeolite framework, like defects
generated when alkali elements where added to the synthesis
gel,[19] may be playing some role in these processes, especially
when the reaction temperature is high (e.g. > 400°C). As we and
others have suggested,[21,44] the Al sites in CHA are the most likely
grafting points to stabilize single Pt or Pd atoms. This observation
is rather common in instances where the extra-framework species
are cationic, for example in Cu dimers and trimers,[27,45] and single
atoms of noble metals under specific synthesis and reaction
conditions.[5] Before reaching the Al site, PtOx species generated
by oxidation of zero-valent clusters in the starting Pt/CHAH2-400C
sample would be travelling through either the gas-phase, if the
temperature is high enough,[21] or through charge-neutral regions
of the zeolite. In any of these two scenarios, the landing of PtOx
onto a surface site that is negatively charged at the Al position
would be expected to cause weakening of the Pt-O bonds, and
the promotion of O2 desorption (or 16O18O) upon recombination.
We think that this process probably requires the participation of a
pair of adjacent Al sites of the zeolite framework, similar to what
has been reported for other transition metals like Cu,[27,28,46] which
is consistent with our previous observation that Pt in CHA sinters
in O2 at high temperature, instead of stabilizing as single atoms,
when the Si/Al ratio is too high.[21] Interestingly, whether the single
Pt sites end grafted to Al pairs previously counterbalanced by Na+
or H+, or both, remains unclear, but our mass spectrometer data
On the other hand, the pre-reduced Pt/CHAH2-400C, with small 0.8-
1.5 nm metal clusters, combusts propane in a similar temperature
range (Table 2). We acknowledge that a fraction of the initial Pt-
Pt bonds in this sample may breakup under the oxidizing propane
combustion environment, but we expect the majority of the metal
to remain as small clusters/nanoparticles taking into account that
a) the T in this experiment is rather low; and b) the propane/O2
ratio is rather high (1:1). As a reference, we previously reported
that the Pt-Pt coordination number of Pt/CHAH2-400C decays from
~7 to ~4 at 200°C when the gas contains diluted O2 and no
hydrocarbon (i.e. no reductant) is present (Table S1).[29]
Consistent with these observations, Pt/SiO2, consisting of ~2 nm
Pt nanoparticles, also starts reacting propane at ~200°C and
consumes the O2 almost quantitatively at 300°C (entry 8 in Table
2). This series of experiments is important, because it indicates
that despite notable differences in the way O2 is activated by
single Pt atoms and Pt clusters (Figure 1 and Table 1), the active
site nuclearity less of an effect on the C–H activation, which is the
key elemental step in these oxidation processes, as rate-
determining. The data provided in this paper does not allow,
however, to rule out some degree of metal nuclearity effects in the
C-H activation process, which have been otherwise extensively
reported for the activation of alkanes like methane at high
temperatures.[49–51] The fact that propane reacts below the
temperature at which the 16O2/18O2 scrambling occurs suggests,
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
moreover, that O2 dissociation does indeed take place in all these
cases, but it is likely followed by irreversible formation of rather
strong Pt–O bonds that are only broken at low temperatures (i.e.
< 250°C) if a relatively strong reductant is present.
temperature, except for the highest T, which was held for 24 h
(not shown). After 24 h at 650°C, the experiment was run
backwards to collect data at decreasing temperatures from 650°C
to 435°C. Figure 3B shows the approximate methane conversion
at each temperature, both during the ramp up and ramp down
experiments.
Table 2. Combustion of methane and propane with Pt-catalysts
React.
Pt loading
(wt%)
Conv.b
(%)
Entry
Sampleb
Substratea mg Cat Temp.
(°C)
1
2
3
4
5
6
7
8
Pt/CHAH2-400C
Pt/CHAH2-400C
Pt/CHAO2-650C
Pt/CHAO2-650C
Pt/CHAH2-400C
Pt/CHAH2-400C
Pt/SiO2 H2-400C
Pt/SiO2 H2-400C
0.4
0.4
0.2
0.2
0.2
0.2
0.8
0.8
Meth.
Meth.
Prop.
Prop.
Prop.
Prop.
Prop.
Prop.
71
71
300
500
200
300
200
300
200
300
0
12
9.4
9.4
10
0.7c
15.8c
0.9c
18.0c
0.8c
17.9c
10
Figure 3. CH4 combustion using Pt/CHA and Pd/CHA pretreated in O2 at 650⁰C:
(A) mass spectrometry signal for fragment 44 (CO2) as the temperature is raised
in the 435-650°C range stepwise (soak time was 2 h for all steps, excepts for
the highest temperature, which was 24 h in the case of the Pt/CHAO2-650C, and
6 h in the case of Pd/CHAO2-650C); (B) and (C) are the corresponding methane
conversion data at the end of each step, as the temperature was ramped up
(red curve) and then ramped down (green curve) after the 650⁰C aging step (24
h for Pt, and 6 h for Pd); Reaction conditions: 71 mg of 0.4 wt % Pt/CHA or 142
mg of 0.2 wt % Pd/CHA; Feed = 40 sccm N2, 100 sccm 10%O2 in N2, 3 sccm
100%CH4
4.8
4.8
a Feed composition: For methane: 40 sccm N2, 100 sccm 10%O2 in N2, 3 sccm
100%CH4; for propane: 0.5 sccm Ar, 5 sccm O2, 5 sccm propane, 89.5 sccm He
b Hydrocarbon conversion
c The conversion of the limiting reagent (O2) is ~5 times the conversion of the
hydrocarbon
The stability of Pt/CHAO2-650C after 24 h at 650°C is remarkable
(Figure 3B), in a stream that contains O2, CH4, CO2 and H2O (in
variable amounts at the inlet and outlet of the reactor,
corresponding to changes in conversion from 0 to 100 %,
respectively). The rate of CO2 formation at each temperature
appears to be fairly constant over the 2 h period investigated, and
the activity reported on the ramp up and ramp down are virtually
identical. This scenario contrasts with that observed on a Pd/CHA
counterpart, more active when fresh as evidenced by a lower
onset reaction temperature, but manifesting complex activation
and deactivation processes on stream (activity continuously going
up at constant T for T < 500°C, but continuously going down for T
> 500°C, Figures 3A and 3C). This result is independent of the
treatment used to activate the Pd/CHA catalyst, to initiate the
reaction with either 1-2 nm particles (reductive pre-treatment) or
single Pd atoms (oxidizing pre-treatment). The non-steady
phenomena at each temperature step complicates drawing
straightforward structure/performance correlations. Our STEM
data on the spent Pd/CHA catalysts show that the majority of the
CHA crystals lack metal structures visible under the electron
beam, although some areas show rather large Pd aggregates
(Figure S4). We hypothesize that the deactivation observed at the
highest temperatures (>550°C) responds to the fragmentation of
active PdOx clusters into less active single Pd atoms, similar to
the results reported by the Cargnello group on Pd/Al2O3.[38] This
re-dispersion phenomenon probably outcompetes other
activation events (unresolved, but likely related to formation of
PdO particles) that may occur at lower T, evidenced by the
increasing conversion at constant T (T<500°C, red curve, Figure
3A).
Remarkably, the differences in reactivity between Pt/CHAO2-650C
and Pt/CHAH2-400C, manifested in both O2 scrambling and propane
combustion experiments, are no longer observed when the
reacting hydrocarbon is methane and the temperature must be
raised > 500°C to achieve some meaningful conversion (e.g. 10 %
of the limiting reagent, Table 2). Under the harsher oxidizing
conditions for methane combustion (higher
T and higher
O2/hydrocarbon ratio of 3:1), one should expect that both
Pt/CHAO2-650C and Pt/CHAH2-400C mainly consist of single Pt atoms,
considering that oxidative fragmentation of Pt clusters in CHA is
virtually completed at 500°C in diluted O2, as evidenced by
EXAFS in Table S1. This inference was further supported by
comparison of STEM images of a fresh and a spent Pt/CHAH2-400C
catalyst after the CH4 combustion experiment, which shows the
disappearance of any feature attributable to Pt-Pt nanoparticles
or clusters (Figure 2D).
The methane combustion activity of single Pt atoms in CHA is
lower than that of the best oxidation catalysts reported in the
literature, such as supported PdO nanoparticles.[39] The
differences between Pt and Pd as a combustion catalyst is widely
discussed in previous reports,[52–54] and here we put the accent in
the performance of single metal atoms to activate both O2 and the
hydrocarbon. The Pt/CHA system is particularly suitable for this
purpose because it is structurally simpler and more stable at high
temperature in O2 compared to other catalysts, as discussed next.
Figure 3A shows the evolution of CO2 measured by mass
spectrometry at the outlet of the reactor, as the T is ramped
stepwise from 435°C to 650°C in an O2-rich O2/CH4 stream. At
each step, the soak time was 2 h before ramping to the next
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
Single Site vs Small cluster catalysis – CO oxidation
Pt/CHAH2-400C sample holds 1-2 nm particles under otherwise
identical conditions (Figure 5).
Our data above show that, in CHA zeolite, O2 can re-disperse < 1
nm Pt clusters and 1-2 nm Pt clusters at 200°C and ~ 500°C,
respectively, during alkane combustion experiments.
Next, we show that this situation is notably different when Pt/CHA
is used for the oxidation of an even stronger reductant such as
CO, which is another very relevant substrate in petrochemical and
automotive industries. The results we present here show once
more how important is to understand the influence of gas/solid
interactions to shape the catalyst structure under working
conditions.[26]
Figure 5. STEM images of Pt/CHAO2-650C (a), Pt/CHAH2-400C (b) and Pt/Al2O3 (c)
catalysts after CO oxidation, and STEM images of Pt/CHAH2-400C (d) and
Pt/Al2O3 (e) after ageing treatments and CO oxidation reaction. The ageing
treatments were: air at 750°C for 13 h followed by H2 at 400°C for 2 h for the
Pt/CHAH2-400C sample, and air at 750°C for 13 hours for the Pt/Al2O3 sample.
The role of CO as a metal structure modifier is not easy to
anticipate. CO can be regarded as both an effective reductant [55–
57] and an effective oxidant,[16,31] and literature often invokes either
argument to explain the experimental observations. CO has been
proved to cause sintering or re-dispersion of different supported
metal species depending on a number of factors that can be hard
to rationalize. On the one hand, CO adds oxidatively to metal
surfaces and complexes in the absence of any other chemical
event, as other ligands like hydrogen or alkenes do.[58] On small
zero-valent metal clusters of Rh and Ir, CO can even cause
oxidative fragmentation to single (cationic) gem-dicarbonyl
complexes on supports like DAY zeolite, MgO or Al2O3.[16,31,59] In
these processes, reaction of M-CO fragments with –OH groups of
the support facilitate oxidation of the metal and the adsorption of
a second CO molecule.[59] On the other hand, CO can facilitate
reduction of the active sites, which usually occurs if any of the
following situations apply: a) the metal surface is covered by labile
(reactive) oxygen, which is eliminated as CO2 upon reaction with
the CO; or b) the oxidative addition of CO to the metal results in
unstable M-CO species that continue to migrate onto the support
until a new more stable metallic cluster or nanoparticle is formed.
The latter can be avoided, for example, by using a support able
to trap the migrating carbonyl fragments, as in the case of Rh/DAY,
where gem-dicarbonyl Rh(CO)2 species become ultra-stabilized
at the Al sites of the zeolite.[60] It is worth noting that although H2
also adds oxidatively to metals, the result of treating single metal
atoms of, for example, Rh with H2 is typically a site with more
reduced metal centers, because hydride-containing Rh single
sites are relatively unstable and bond too weakly to the support to
restrict their surface migration and subsequent particle growth. In
these cases, Brownian motion usually stops at a given T when the
particle reaches a threshold size that provides sufficient metal-
support bonding. The presence of additional ligands bonded to
the metal further complicates these types of analyses.[60]
Figure 4. CO conversion at different T using Pt/CHAO2-650C, Pt/CHAH2-400C and
Pt/Al2O3 catalysts in their fresh forms (filled) and after being aged (empty). The
ageing treatments were: air at 750°C for 13 h followed by H2 at 400°C for 2 h
for the Pt/CHAH2-400C sample, and air at 750°C for 13 h for Pt/Al2O3.
The CO oxidation activity of Pt/CHA was tested with the catalyst
pre-treated in O2 at 650°C (Pt/CHAO2-650C) or H2 at 400°C
(Pt/CHAH2-400C). For comparison purposes, we also tested a
reference sample, in this case Pt/Al2O3 because it is the most
universal industrial benchmark. The Pt/Al2O3 catalyst was made
by incipient wetness impregnation of Pt(NH3)(NO3)2 on γ-alumina
to deliver well-dispersed 2nm Pt particles upon activation in H2
at 400°C.
Figure 4 shows the light off curves for the combustion of CO using
these three catalysts under O2-rich reaction conditions
(O2/CO=6.3). The onset CO oxidation temperature for the fresh
catalysts (solid lines) follows the order Pt/Al2O3 > Pt/CHAO2-650C
>
Pt/CHAH2-400C. Full conversion of the CO occurs in the 230 to
250°C range for all the catalysts. The T required to combust CO
is notably lower than that needed to activate CH4, consistent with
the more inert nature of the latter (stronger bonds). However, we
found surprising that Pt/CHAO2-650C (with single Pt atoms) and
Pt/CHAH2-400C (with 0.8-1.5 nm particles), behaved so similar. To
better understand these results, we characterized the spent
catalysts using STEM, and the presence of metal nanoparticles in
all cases is observed (Figure 5). We conclude, thus, that CO
promotes the consolidation of Pt-Pt bonds, through a process that
outcompetes the effect of O2 as a re-dispersing agent. Said that,
we still note some remarkable differences between the various
samples depending on the activation treatment. For example,
while Pt/CHAO2-650C, initially containing single Pt atoms, sinters in
the CO stream to yield relatively larger particles (some of them >4
nm, and clearly outside the CHA crystals), the pre-reduced
Our work here evidences that single Pt atoms in Pt/CHAO2-650C do
not form stable small carbonyl fragments or complexes in CO, but
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
rather large nanoparticles outside the zeolite crystals (Figure 5).
In this aggregation process, collisions inside the zeolite between
carbonyl-carrying Pt ensembles and –OH groups of the support
are likely to occur, but we postulate their product to be rather
unstable or, otherwise, no metal would accumulate at the external
surface of the zeolite. Our data indicates that this scenario is
avoided, however, if sufficiently large Pt ensembles are
intentionally generated inside the zeolite prior to exposure to the
CO stream, as in Pt/CHAH2-400C, which remain virtually unchanged
after the interaction with CO (Figure 5). These 1-2 nm Pt clusters
inside CHA resist, thus, any oxidative (re-dispersing) attack by the
CO under the conditions investigated.
On the other hand, many CO oxidation applications require that
the metal stays stable after exposed to particularly extreme
conditions, for example in diesel engines.[61–63] To simulate similar
conditions, we tested the light off performance of the benchmark
Pt/Al2O3 and the Pt/CHAH2-400C after they had been subjected to
an extra aging step in air at 750°C for 13 h before the catalytic
test. Due to the severity of this treatment, the onset reaction
temperature of the benchmark Pt/Al2O3 sample shifted from
210°C to >270°C (Figure 4). This shift was accompanied by clear
symptoms of metal sintering in the corresponding STEM images,
showing the formation of Pt agglomerates larger than 20 nm
(Figure 5). This result is consistent with previous literature.[12,13]
The aggregation is ascribed to the occurrence of Ostwald ripening
of volatile PtOx arising from the small Pt nanoparticles that are
trapped by larger (more stable) ones, resulting in an overall
particle growth.[12,64] As we previously reported,[21,29] the CHA
framework is an excellent atom-trapping support under identical
conditions, eliminating this hard-to-prevent sintering problem.
Accordingly, the performance of the aged Pt/CHA catalyst,
activated in H2 prior to the CO oxidation reaction, is nearly
identical to the one observed with the unaged catalyst, as the
particle size distribution inferred from the STEM images remains
virtually the same (see Figure 5). This catalyst, thus, is not only
stable, but more active than the Pt/Al2O3 benchmark after a single
aging step, which is also an important result from a more practical
standpoint.
single metal atoms are formed in either case, as demonstrated by
in situ EXAFS spectroscopy and ex situ STEM microscopy. First
evidences of CO2 formation on single Pt atoms occur at a T higher
than that required for isotopic O2 scrambling when the substrate
is methane, but not when it is propane, suggesting that these
combustion processes by single atoms are not limited by O2
dissociation, but C–H activation. And, interestingly, our
experiments suggest that the O2 scrambling activity is more
sensitive to changes in the active site nuclearity than the
activation of C–H bonds in propane, as combustion of the later
occurs at fairly low temperatures on both single Pt atoms and
small Pt nanoparticles. Pt in CHA, on the other hand, stabilizes as
small metal clusters during processes that involve the presence
of CO, and leads to high activity and improved stability compared
to reference Pt/Al2O3 formulations.
Experimental Section
Experimental information on synthesis, characterization and
testing of catalysts are given in the SI.
Acknowledgements
This work has been supported by the Spanish Government
through
SEV-2016-0683
and
RTI2018-101033-B-I00
(MCIU/AEI/FEDER, UE). The Electron Microscopy Service of the
UPV is acknowledged for their help in sample characterization.
A.R.F. acknowledges the Spanish Government-MINECO for a
FPU scholarship (FPU2017/01521). We thank Isabel Millet, Tom
Green and Michael Caulfield for technical assistance and to
Randall Meyer for his comments and additional discussions when
writing this paper. We appreciate the support of ExxonMobil
Research and Engineering for our efforts in fundamental catalytic
research.
Keywords: Zeolite • single-site • metal cluster • oxidation
catalyst • structure-activity relationship
Conclusions
References
The dynamic behavior of single Pt atoms and small Pt clusters
trapped within a non-reducible support such as CHA zeolite has
been studied during oxidation reactions to elucidate structure-
performance relationships following subtle changes in the active
sites under working conditions. Our findings demonstrate that
clusters <1 nm are more active than single atoms and >1 nm
clusters to dissociate, recombine and desorb O2; and that the
exact nature of the working species depends on complex
interactions between the gas-phase components, the metal
species and the support that can be largely understood thanks to
the excellent properties of CHA to trap and stabilize rather uniform,
small metal species under the harsh reaction conditions
investigated. Oxidative fragmentation of most active < 1 nm
clusters into single Pt atoms is shown responsible for the gradual
loss of isotopic 16O2/18O2 exchange activity at moderate T (i.e.
200⁰C). In the case of methane combustion, negligible catalytic
differences were observed for samples initially containing either
single Pt or small Pt clusters, as temperatures above 500°C are
needed to achieve any meaningful CH4 conversion, for which
[1]
[2]
L. Liu, A. Corma, Chem. Rev. 2018, 118, 4981–5079.
L. Liu, A. Corma, Nat. Rev. Mater. 2020, DOI 10.1038/s41578-020-
00250-3.
[3]
J. Jones, H. Xiong, A. T. DeLaRiva, E. J. Peterson, H. Pham, S. R.
Challa, G. Qi, S. Oh, M. H. Wiebenga, X. I. Pereira Hernandez, Y.
Wang, A. K. Datye, Science 2016, 353, 150–154.
X. Cui, W. Li, P. Ryabchuk, K. Junge, M. Beller, Nat. Catal. 2018, 1,
385–397.
[4]
[5]
B. C. Gates, M. Flytzani-Stephanopoulos, D. A. Dixon, A. Katz,
Catal. Sci. Technol. 2017, 7, 4259–4275.
[6]
[7]
B. C. Gates, Trends Chem. 2019, 1, 99–110.
K. Narsimhan, K. Iyoki, K. Dinh, Y. Román-Leshkov, ACS Cent. Sci.
2016, 2, 424–429.
[8]
[9]
[10]
A. Wang, J. Li, T. Zhang, Nat. Rev. Chem. 2018, 2, 65–81.
J. D. A. Pelletier, J.-M. Basset, Acc. Chem. Res. 2016, 49, 664–677.
S. Vajda, M. J. Pellin, J. P. Greeley, C. L. Marshall, L. A. Curtiss, G.
A. Ballentine, J. W. Elam, S. Catillon-Mucherie, P. C. Redfern, F.
Mehmood, P. Zapol, Nat. Mater. 2009, 8, 213–216.
G. Kwon, G. A. Ferguson, C. J. Heard, E. C. Tyo, C. Yin, J.
DeBartolo, S. Seifert, R. E. Winans, A. J. Kropf, J. Greeley, R. L.
Johnston, L. A. Curtiss, M. J. Pellin, S. Vajda, ACS Nano 2013, 7,
5808–5817.
[11]
[12]
T. W. Hansen, A. T. DeLaRiva, S. R. Challa, A. K. Datye, Acc.
Chem. Res. 2013, 46, 1720–1730.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
[13]
[14]
[15]
[16]
S. B. Simonsen, I. Chorkendorff, S. Dahl, M. Skoglundh, J.
9, 5445–5461.
Sehested, S. Helveg, J. Am. Chem. Soc. 2010, 132, 7968–7975.
N. Wang, Q. Sun, R. Bai, X. Li, G. Guo, J. Yu, J. Am. Chem. Soc.
2016, 138, 7484–7487.
L. Liu, U. Díaz, R. Arenal, G. Agostini, P. Concepción, A. Corma,
Nat. Mater. 2017, 16, 132–138.
J. Zhang, L. Wang, B. Zhang, H. Zhao, U. Kolb, Y. Zhu, L. Liu, Y.
Han, G. Wang, C. Wang, D. S. Su, B. C. Gates, F.-S. Xiao, Nat.
Catal. 2018, 1, 540–546.
[54]
[55]
[56]
[57]
W. Qi, J. Ran, R. Wang, X. Du, J. Shi, J. Niu, P. Zhang, M. Ran,
RSC Adv. 2016, 6, 109834–109845.
Y. Kang, X. Ye, C. B. Murray, Angew. Chemie 2010, 122, 6292–
6295.
G. Chen, Y. Tan, B. Wu, G. Fu, N. Zheng, Chem. Commun. 2012,
48, 2758.
Y. Dai, X. Mu, Y. Tan, K. Lin, Z. Yang, N. Zheng, G. Fu, J. Am.
Chem. Soc. 2012, 134, 7073–7080.
[17]
S. Goel, S. I. Zones, E. Iglesia, J. Am. Chem. Soc. 2014, 136,
15280–15290.
[58]
[59]
J. A. Labinger, Organometallics 2015, 34, 4784–4795.
A. Suzuki, Y. Inada, A. Yamaguchi, T. Chihara, M. Yuasa, M.
Nomura, Y. Iwasawa, Angew. Chemie Int. Ed. 2003, 42, 4795–
4799.
P. Serna, D. Yardimci, J. D. Kistler, B. C. Gates, Phys. Chem.
Chem. Phys. 2014, 16, 1262–1270.
A. S. Ivanova, E. M. Slavinskaya, R. V. Gulyaev, V. I. Zaikovskii, О.
А. Stonkus, I. G. Danilova, L. M. Plyasova, I. A. Polukhina, A. I.
Boronin, Appl. Catal. B Environ. 2010, 97, 57–71.
Y. Xie, E. Rodrigues, N. Furtado, A. Matynia, T. Arlt, P. Rodatz, P.
Da Costa, Top. Catal. 2016, 59, 1039–1043.
M. H. Wiebenga, C. H. Kim, S. J. Schmieg, S. H. Oh, D. B. Brown,
D. H. Kim, J.-H. Lee, C. H. F. Peden, Catal. Today 2012, 184, 197–
204.
[18]
[19]
P. Serna, B. C. Gates, J. Am. Chem. Soc. 2011, 133, 4714–4717.
L. Liu, M. Lopez-Haro, C. W. Lopes, C. Li, P. Concepcion, L.
Simonelli, J. J. Calvino, A. Corma, Nat. Mater. 2019, 18, 866–873.
K. H. Rasmussen, J. Mielby, S. Kegnæs, ChemCatChem 2018, 10,
4380–4385.
M. Moliner, J. Gabay, C. Kliewer, P. Serna, A. Corma, ACS Catal.
2018, 8, 9520–9528.
P. W. Kletnieks, A. J. Liang, R. Craciun, J. O. Ehresmann, D. M.
Marcus, V. A. Bhirud, M. M. Klaric, M. J. Hayman, D. R. Guenther,
O. P. Bagatchenko, D. A. Dixon, B. C. Gates, J. F. Haw, Chem. - A
Eur. J. 2007, 13, 7294–7304.
P. Serna, B. C. Gates, Acc. Chem. Res. 2014, 47, 2612–2620.
Q. Sun, N. Wang, Q. Bing, R. Si, J. Liu, R. Bai, P. Zhang, M. Jia, J.
Yu, Chem 2017, 3, 477–493.
L. Liu, M. Lopez‐Haro, D. M. Meira, P. Concepcion, J. J. Calvino, A.
Corma, Angew. Chemie Int. Ed. 2020, 59, 15695–15702.
L. Liu, D. N. Zakharov, R. Arenal, P. Concepcion, E. A. Stach, A.
Corma, Nat. Commun. 2018, 9, 574.
C. Paolucci, I. Khurana, A. A. Parekh, S. Li, A. J. Shih, H. Li, J. R. Di
Iorio, J. D. Albarracin-Caballero, A. Yezerets, J. T. Miller, W. N.
Delgass, F. H. Ribeiro, W. F. Schneider, R. Gounder, Science 2017,
357, 898–903.
[60]
[61]
[20]
[21]
[22]
[62]
[63]
[23]
[24]
[64]
T. R. Johns, R. S. Goeke, V. Ashbacher, P. C. Thüne, J. W.
Niemantsverdriet, B. Kiefer, C. H. Kim, M. P. Balogh, A. K. Datye, J.
Catal. 2015, 328, 151–164.
[25]
[26]
[27]
[28]
[29]
K. T. Dinh, M. M. Sullivan, K. Narsimhan, P. Serna, R. J. Meyer, M.
Dincă, Y. Román-Leshkov, J. Am. Chem. Soc. 2019, 141, 11641–
11650.
M. Moliner, J. E. Gabay, C. E. Kliewer, R. T. Carr, J. Guzman, G. L.
Casty, P. Serna, A. Corma, J. Am. Chem. Soc. 2016, 138, 15743–
15750.
[30]
[31]
[32]
[33]
[34]
C. Aydin, J. Lu, N. D. Browning, B. C. Gates, Angew. Chemie Int.
Ed. 2012, 51, 5929–5934.
J. Lu, P. Serna, C. Aydin, N. D. Browning, B. C. Gates, J. Am.
Chem. Soc. 2011, 133, 16186–16195.
D. Yardimci, P. Serna, B. C. Gates, Chem. - A Eur. J. 2013, 19,
1235–1245.
P. Serna, B. C. Gates, Angew. Chemie Int. Ed. 2011, 50, 5528–
5531.
Y. Tang, C. Asokan, M. Xu, G. W. Graham, X. Pan, P. Christopher,
J. Li, P. Sautet, Nat. Commun. 2019, 10, 4488.
A. Corma, P. Serna, Science 2006, 313, 332–334.
A. Corma, P. Serna, P. Concepción, J. J. Calvino, J. Am. Chem.
Soc. 2008, 130, 8748–8753.
[35]
[36]
[37]
[38]
S. Carrettin, P. Concepción, A. Corma, J. M. López Nieto, V. F.
Puntes, Angew. Chemie Int. Ed. 2004, 43, 2538–2540.
E. D. Goodman, A. C. Johnston-Peck, E. M. Dietze, C. J. Wrasman,
A. S. Hoffman, F. Abild-Pedersen, S. R. Bare, P. N. Plessow, M.
Cargnello, Nat. Catal. 2019, 2, 748–755.
[39]
B. Nan, X.-C. Hu, X. Wang, C.-J. Jia, C. Ma, M.-X. Li, R. Si, J. Phys.
Chem. C 2017, 121, 25805–25817.
[40]
[41]
S. Soled, Science 2015, 350, 1171–1172.
P. D. Nolan, B. R. Lutz, P. L. Tanaka, J. E. Davis, C. B. Mullins, J.
Chem. Phys. 1999, 111, 3696–3704.
[42]
[43]
[44]
[45]
B. Qiao, A. Wang, X. Yang, L. F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li,
T. Zhang, Nat. Chem. 2011, 3, 634–641.
T. Kropp, Z. Lu, Z. Li, Y.-H. C. Chin, M. Mavrikakis, ACS Catal.
2019, 9, 1595–1604.
J. Van der Mynsbrugge, M. Head-Gordon, A. T. Bell, J. Mater.
Chem. A 2021, DOI 10.1039/D0TA11254B.
S. Grundner, M. A. C. Markovits, G. Li, M. Tromp, E. A. Pidko, E. J.
M. Hensen, A. Jentys, M. Sanchez-Sanchez, J. A. Lercher, Nat.
Commun. 2015, 6, 7546.
[46]
[47]
J. R. Di Iorio, R. Gounder, Chem. Mater. 2016, 28, 2236–2247.
D. Hou, L. Grajciar, P. Nachtigall, C. J. Heard, ACS Catal. 2020, 10,
11057–11068.
[48]
J. Guzman, S. Carrettin, A. Corma, J. Am. Chem. Soc. 2005, 127,
3286–3287.
[49]
[50]
J. Wei, E. Iglesia, J. Catal. 2004, 224, 370–383.
G. Jones, J. Jakobsen, S. Shim, J. Kleis, M. Andersson, J.
Rossmeisl, F. Abildpedersen, T. Bligaard, S. Helveg, B. Hinnemann,
J. Catal. 2008, 259, 147–160.
[51]
[52]
Y. Wang, L. Xiao, Y. Qi, J. Yang, Y.-A. Zhu, D. Chen, J. Phys.
Chem. C 2020, 124, 2501–2512.
E. D. Goodman, S. Dai, A.-C. Yang, C. J. Wrasman, A. Gallo, S. R.
Bare, A. S. Hoffman, T. F. Jaramillo, G. W. Graham, X. Pan, M.
Cargnello, ACS Catal. 2017, 7, 4372–4380.
[53]
H. Nie, J. Y. Howe, P. T. Lachkov, Y.-H. C. Chin, ACS Catal. 2019,
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202102339
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Metal active sites speciation on Pt-CHA is discussed for different high-temperature oxidation reactions, including 16O2/18O2isotopic
exchange experiments, methane combustion and CO oxidation, unraveling structure-performance relationships built on subtle
changes in the active sites under working conditions
9
This article is protected by copyright. All rights reserved.