Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt Catalysts

Full text of DOI:10.1016/j.chempr.2019.12.029

Article
Insight into Single-Atom-Induced
Unconventional Size Dependence over CeO2-
Supported Pt Catalysts
Chunpeng Wang, Shanjun Mao,
Zhe Wang, ..., Bingbao Mei,
Zheng Jiang, Yong Wang
maoshanjun@zju.edu.cn (S.M.)
chemwy@zju.edu.cn (Y.W.)
HIGHLIGHTS
An unexpected volcano curve was
first reported in the
hydrogenation of p-CNB
Relationship between selectivity
and the quantified steric
hindrance of Pt was built
The steric-hindrance effect was
clearly disclosed from the
electronic level
Atomically precise catalysis specially concentrates on the interaction between the
central active atom and the coordinating environment nearby, as well as the
resulting catalytic performance. The neighboring atoms can have a great impact
on the frontier orbitals of the metal active site and further its catalytic performance.
Here, we provide a thorough example discussing the origin of unexpected
selectivity variation for a structural sensitive reaction on supported catalysts with
various steric hindrances of Pt sites.
Wang et al., Chem 6, 1–14
March 12, 2020 ª 2019 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2019.12.029

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Article
Insight into Single-Atom-Induced
Unconventional Size Dependence
over CeO2-Supported Pt Catalysts
1
1
2
2
3
Chunpeng Wang,1 Shanjun Mao,1, Zhe Wang, Yuzhuo Chen, Wentao Yuan, Yang Ou, Hao Zhang,
*
Yutong Gong,1 Yong Wang,2 Bingbao Mei,3 Zheng Jiang,3 and Yong Wang1,4,
*
SUMMARY
The Bigger Picture
The 100% exposed atomic sites
have made the single-atom
catalyst (SAC) very different from
the nanoparticle ones with unique
selectivity and reactivity.
Identification of the structure-performance relationship at the atomic level is vital
for getting a deep understanding of the size-dependence behavior of metal cata-
lysts. Here, an unconventional size dependence of Pt toward selective hydrogena-
tion of p-chloronitrobenzene has been extensively investigated over CeO2 sup-
port. An upturned volcanic curve toward the selectivity of p-chloroaniline was
found by decreasing the size of Pt from nanoparticles to single atoms. Differences
on predominant orbitals among diverse coordination-environment Pt sites were
identified to be the key factors influencing the modes of interaction with the
C-Cl bond of p-chloroaniline. Specifically, electrostatic repulsion between non-
bonding orbitals of Cl and predominant orbitals of Pt sites with high steric
hindrance were speculated to be responsible for the suppression of dehalogena-
tion and the high selectivity. This strategy to build correlation between the valence
orbitals of active sites and their catalytic behavior could well be expanded to other
structure-sensitive reactions and atomically dispersed catalysts.
However, the deep reasons for
this have not been well
documented in most cases.
Lacking the knowledge for the
specific structure-activity
relationship of the SACs
suppressed their extensive
application in industry. Here, an
unexpected upturned volcano
curve of selectivity to
p-chloroaniline (p-CAN) in the
hydrogenation of
INTRODUCTION
Selective hydrogenation reactions have attracted much attention for many decades
in view of their widespread use in the chemical industry.1,2 These processes have
been intensively studied from both fundamental and practical points of view, and
much has been learned from them.3–5 However, some key fundamental issues asso-
ciated with catalytic hydrogenation remain unsolved. For example, although noble
metal catalysts are active under mild conditions for a variety of heterogeneous hy-
drogenations, they are not always selective toward the preferred products and are
expensive. Especially, the realization of high chemoselectivity is increasingly chal-
lenging as the reducibility of the competing functional groups increases.6,7 Haloani-
lines (HANs), for instance, as the vital organic intermediate widely used in the
production of polymers, pharmaceuticals, herbicides, dyes, etc., are usually ob-
tained through the selective reduction of the corresponding halonitrobenzenes
(HNBs).8 A range of noble-metal-based heterogeneous catalysts such as Pt,9,10
Pd,11,12 Ru,13,14 and Au15,16 have been studied for HNBs hydrogenation. However,
hydrogenolysis of the carbon-halogen bonds always occurs,17,18 especially at high
conversions, resulting in loss of the HANs yield and the extra cost for the product pu-
rification. Hence, to achieve high selectivity of HANs, it is necessary to maximally
limit or, if possible, totally suppress hydrogenolysis of the carbon-halogen bonds.
p-chloronitrobenzene (p-CNB)
was firstly disclosed at atomic
level on the Pt1/CeO2–SAC and
the Pt/CeO2 catalysts with Pt
nanoparticles on CeO2 support.
The electrostatic repulsion
between the non-bonding orbitals
of Cl and the predominant orbitals
of Pt sites with high steric
hindrance, including Pt1/CeO2–
SAC and Pt (111), was speculated
to be the profound reason behind
the size-dependence behavior.
The strategy to build correlation
between valence bands of metal
atoms and catalytic performance
could be expanded to other
heterogeneous reactions as well.
Generally, two approaches are used to control the selectivity in hydrogenation of
HNBs: adding a second substance or engineering the catalytic surface structure of
Chem 6, 1–14, March 12, 2020 ª 2019 Elsevier Inc.
1

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
the catalyst, which can be addressed by using metal nanoparticles (NPs) with specific
sizes and shapes. One effective method of the first-mentioned approach to prevent
the selectivity decrease is adding substances such as amines, sulfur-containing com-
pounds, phosphoric acid derivatives, etc.19 to the reaction medium. However, it
generated new problems associated with isolation and purification of the final prod-
uct. Additionally, much efforts were also made at alloying common noble metals
with a transition metal, but a simple bimetallic composite usually provides a limited
number of active sites, and the high selectivity is achieved at the expense of activ-
ity.20,21 Moreover, catalytic reactions are traditionally characterized as structure sen-
sitive or structure insensitive.22 Dechlorination of HNBs (or HANs) compounds,
which accounts for the undesirable selectivity of HANs, has typically been consid-
ered to belong to the former category. Hence, attempts at inhibiting hydrogenolysis
of the weak carbon-halogen bond were also made in controlling the metal particle
dispersion.23,24 For instance, it was reported that the selectivity of p-chloroaniline
(p-CAN) depended closely on metal particle size, and the higher selectivity to
p-CAN was achieved on large metal particles.25,26 Note that facile adsorption of
reactants and weak binding of products (intermediates) are key requirements
for efficient and selective catalysis. However, these two variables are intimately
linked in some ways that do not allow the optimization of both properties simulta-
neously. In this way, high p-CAN selectivity was achieved only under relative low
p-chloronitrobenzene conversion because of the competitive adsorption between
p-chloronitrobenzene (p-CNB) and p-CAN or at the expense of the low atom
utilization of a noble metal (large metal particles).
Many studies have shown that individual atoms of Pt,27–29 Pd,30,31 or Ru32,33 supported
on oxides or carbons can efficiently catalyze numerous important reactions with high
selectivity. However, the in-depth reasons for the extraordinary performance of these
single-atom catalysts have not been well documented in most cases. In this work,
through a series of controlling experiments and density functional theory (DFT) studies,
the unconventional size dependence toward selectivity in the hydrogenation of p-CNB
over CeO2-supported Pt catalysts was investigated in detail. It was found that the
selectivity toward p-CAN changed drastically with Pt loading and showed an upturned
volcanic curve. DFT calculations suggested that the differences on predominant
orbitals among Pt sites of diverse, coordinated environments had a great influence
on the modes of interaction with the C-Cl bond and was a key factor to determine
the selectivity of p-CAN in p-CNB hydrogenation. The monodispersed Pt active site,
Pt1/CeO2, which was formed by filling a single Pt atom into the vacancies on CeO2,
could be regarded as an alternative of high-steric-hindrance terrace sites to achieve
both high selectivity and atomic economy in the hydrogenation of p-CNB.
1Advanced Materials and Catalysis Group,
Institute of Catalysis, Department of Chemistry,
Zhejiang University, Hangzhou 310028, P.R. China
RESULTS AND DISCUSSION
Characterizations of Pt/CeO2
A series of Pt catalysts over CeO2 support were synthesized with a conventional wet
impregnation method, followed by various characterizations to figure out the disper-
sion and configuration of Pt on the catalysts (Tables S1 and S2), especially for the
samples with Pt loadings below 3 wt %. As shown in the representative high-resolu-
tion HAADF-STEM images of the Pt/CeO2 catalysts with different Pt loadings (Fig-
ures 1A–1C, S1, and S2), the Pt atoms or nanoclusters were exhibited as the bright
dots with higher contrast than the CeO2 lattice. The images of the 0.6 wt % Pt/CeO2
(denoted as 0.6% Pt/CeO2-SAC) indicated that Pt was atomically dispersed over the
CeO2 (110) plane (Figures 1A and S1). These insets in Figures 1A and S1 showed the
intensity profile taken along the lines indicated by the yellow rectangle, highlighting
the positions of Pt1 atoms in the sample. And both single-atom and nanoclusters
2Center of Electron Microscopy and State Key
Laboratory of Silicon Materials, School of
Materials Science and Engineering, Zhejiang
University, Hangzhou 310027, P.R. China
3Shanghai Synchrotron Radiation Facility,
Zhangjiang Lab, Shanghai Advanced Research
Institute, Chinese Academy of Science, Shanghai
201210, P.R. China
4Lead Contact
*Correspondence:
maoshanjun@zju.edu.cn (S.M.),
chemwy@zju.edu.cn (Y.W.)
https://doi.org/10.1016/j.chempr.2019.12.029
2
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
were simultaneously present on the support when the ratio of Pt increased from 0.6
to 2 wt % (Figure 1B). When further increasing the loading to 3 wt %, most of Pt would
aggregate to small nanoclusters (ꢀ1.7 nm, Figures 1C and S2). Furthermore, the
magnitude Pt L3-edge X-ray near-edge structure (XANES) spectra and extended
X-ray absorption fine structure (EXAFS) analyses were performed to elucidate
detailed structures of these catalysts. The white-line intensity, reflecting the oxida-
tion state of Pt, decreased steadily with the rising of Pt loadings (Figure 1D). The
white-line intensity in 0.6% Pt/CeO2-SAC was similar with that of the PtO2 foil,
demonstrating the Pt single atoms in this sample were in oxidized state. Whereas,
the white-line in the 2% Pt/CeO2 sample is between the intensities of Pt foil and
PtO2 and clearly lower than that of 0.6% Pt/CeO2-SAC, suggesting Pt was partially
reduced in this sample. In other words, lower Pt loading results in higher percent-
ages of single-atom structures in the catalysts. Correspondingly, in the Fourier
transform (R space, Figures 1E and S3; Table S2) of the EXAFS data of the 0.6%
Pt/CeO2-SAC sample, Pt is coordinated with O and Ce with coordination numbers
CN = 3.8 G 0.50 and 1.1 G 0.6, respectively. The absence of a Pt-Pt bond confirmed
that Pt was singly dispersed. While in the 2% Pt/CeO2 sample, except for the Pt-O
˚
˚
peak at 1.7 A, a Pt-Pt contribution also occurred at a distance of 2.72 A with
an average CN of 6.2. The much lower CN and Pt-Pt bonding distance in the
˚
2% Pt/CeO2 sample than those in bulk platinum of 12 at a distance of 2.78 A, sug-
gested that the 2% Pt/CeO2 contained Ptx subnanoclusters,34 which is consistent
with Figure 1B mentioned above.
The X-ray photoelectron spectroscopy (XPS) provided further information about
the oxidation state of Pt (Figure S4). The characteristic Pt 4f peaks of the samples
can be clearly identified. Notably, only partially oxidized Ptd+ species were found
in 0.6% Pt/CeO2-SAC. The absence of Pt0 species provided further evidence for
the single-atom structure in this sample. Additionally, it can be easily seen that
the content of Pt0 increased with size, in good agreement with the EXAFS results dis-
cussed above. Further, in-situ infrared spectroscopy of adsorbed CO (Figure 1F)
were carried out to provide more information about the dispersion and oxidation
state of Pt. It can be found that two major regions were detected in the three
samples, a main peak above 2,000 cmÀ1 usually corresponding to linearly adsorbed
CO on Pt atop sites and less intense and broader section below 1,900 cmÀ1 typically
belonging to the bridge adsorption mode.35 Only one peak was observed at
2,088 cmÀ1 for 0.6% Pt/CeO2-SAC, which can be assigned to the linearly bonded
CO on the partially oxidized Pt in the single-atom catalyst.25 The spectra of 2%
Pt/CeO2 not only displayed the features of the 0.6% Pt/CeO2-SAC (Figure 1F inset)
but also showed other peaks around 1,850 cmÀ1 for the bridge adsorption of CO,
demonstrating that the Pt clusters coexisted with individual Pt atoms on this sample.
The peak belonging to bridge-bonded CO increased in a drastic fashion on the 3%
Pt/CeO2 sample, which could be because of the increased Pt particle size and the
vanishing of the single Pt atoms. The dispersion of Pt on 0.6% Pt/CeO2 derived
from CO titration experiments was 95%, confirming the atomically dispersed state
of Pt in this catalyst. As is expected, the dispersion of Pt decreased when raising
the Pt loadings and the calculated average particle size increased from 1.1 to
11.1 nm (Table S3), consistent with the HRTEM results. The XRD patterns also
confirmed this (Figure S5). When the loading of Pt was less than 3 wt %, no significant
diffraction peaks of metallic Pt particles were visible, suggesting the high dispersion
of Pt. The peak intensity attributed to metal Pt appeared when further increasing
the Pt loading. The average particle size of Pt NPs calculated according to
Scherrer formula were all close to the values derived from CO titration experiments
(Table S3). Based on the above discussion, it can be concluded that Pt had
Chem 6, 1–14, March 12, 2020
3

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 1. Structural Properties of Pt/CeO2
(A–C) Representative high-resolution HAADF-STEM images of 0.6% (A), 2% (B), and 3% Pt/CeO2 (C), respectively. Scale bars represent 2 nm.
(D and E) The magnitude Pt L3-edge XANES (D) and FT-EXAFS (E) spectra of 0.6% and 2% Pt/CeO2, respectively.
(F) Infrared spectroscopy of CO adsorption of 0.6%, 2%, and 3% Pt/CeO2, respectively.
successfully been transformed from single-atom to NPs with the rising of Pt loading
on these samples.
Catalytic Performances of Pt/CeO2 for p-CNB Hydrogenation
The hydrogenation of p-CNB was then conducted to evaluate the catalytic perfor-
mance of the Pt/CeO2 catalysts with different Pt particle sizes (Table S4). Surpris-
ingly, an upturned volcanic curve of the selectivity toward p-CAN was observed
(Figure 2A), which has not been reported before. When Pt loading was beyond
3 wt %, the selectivity to p-CAN decreased in a dramatic fashion along with the
decrease of the Pt particle sizes, which is consistent with previous results. However,
the p-CAN selectivity increased reversely when further decreasing the Pt loading,
and it was surprising that both 0.3% Pt/CeO2-SAC and 0.6% Pt/CeO2-SAC with
atomically dispersed Pt showed nearly 100% selectivity to p-CAN under this condi-
tion. Furthermore, 0.6% Pt/CeO2-SAC was much more active than other catalysts
(Table S5). This means that high activity, high selectivity, and theoretical 100%
atom utilization could be achieved by using this single-atom catalyst. In addition,
in order to eliminate the influence of Cl in the precursor, the catalysts, which were
prepared with [Pt(NH3)4](NO3)2 were also tested in the hydrogenation of p-CNB (Fig-
ures S6–S9; Tables S6 and S7). The same results were obtained as that of H2PtCl6,
further confirming that it was the inherent structure of Pt rather than other factors
that controlled the selectivity.
The extraordinarily high selectivity of the 0.6% Pt/CeO2-SAC catalyst was further
evidenced by the evolution of the reactant and products distribution with the
reaction time in Figure 2B. Since the complete conversion of p-CNB was achieved
4
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 2. Catalytic Performance of Pt/CeO2
(A) Selectivity change in p-CNB hydrogenation on serial Pt/CeO2 catalysts with different Pt
loadings.
(B) Distribution of products during the hydrogenation of p-CNB over 0.6% Pt/CeO2-SAC.
Reaction conditions: 15 mL ethanol, 1 MPa H2, 40ꢁC, 0.8 mmol p-CNB, Pt/p-CNB: (A) 31.7 wt & and
(B) 0.48 wt &; AN, aniline; p-CAN, p-chloroaniline; p-CNB, p-chloronitrobenzene; Cl-AZOXY,
Cl-azoxybenzene; Cl-AZO, Cl-azobenzene.
in several minutes when the ratio of Pt and p-CNB was 31.7 wt &, only 0.48 weights
of Pt per thousand of p-CNB was then elaborately added to probe the product dis-
tribution versus the reaction time. As shown in Figure 2B, p-CNB can be transformed
into p-CAN in 32 min and almost no dechlorination product (AN) was detected.
At the same time, no nitrobenzene (NB) was detected during the reaction, indi-
cating that the nitro group was preferentially hydrogenated and subsequently
dehalogenation proceeded. It needs to be emphasized that the selectivity still re-
mained as high as 99.5%, even prolonging the reaction time to 70 min (Figure S10),
revealing the superior performance of 0.6% Pt/CeO2-SAC in the selective hydroge-
nation of p-CNB compared with the catalysts with higher Pt loadings. Additionally,
Figure S11 can confirm that there was no dissolved part contributing to the overall
performance via a homogeneous mechanism, indicating that it is a heterogeneously
catalytic process.
DFT Calculations
Hydrogenation of p-CNB was a pseudo-consecutive reaction, where hydrogenation
of nitro proceeded priorly and was not structure sensitive.26 As a result, selectivity of
p-CAN would depend strongly on the level of the C-Cl hydrogenolysis. Previous re-
ports26,36 showed that the hydrogenolysis of p-CAN to AN was size dependent, and
the performance of 3% Pt/CeO2 also confirmed that Pt NPs of 1.7 nm had a higher
tendency to break the C-Cl bond than those with bigger particle sizes. The
controlled experiments in Figure 2A, however, indicated that both atomic-dispersed
Pt and large Pt NPs contributed to a superior selectivity for hydrogenation of
p-CNB. Such a unique catalytic behavior intrigued us to discover the key factors at
atomic level inducing dechlorination on diverse sites. It is clear that p-CAN must
adsorb on the surface of the catalyst before it can be dechlorinated,25 but there
has been little discussion in the literature about the nature of the C-Cl bond in the
adsorbed p-CAN. Additionally, it is well-known that the distribution of various sites
(terrace, edge, corner, etc.) would vary with the size of the nanocrystal according to
the Wulff construction.37 The fraction of atoms located at the edge and corner sites
increases with particle size and decreases at the expense of terrace sites, and the
low-coordinated sites would be predominant when the particle size is less than
2 nm.38 In this work, diverse models with different CNs of Pt were built to simulate
the catalysts used in the reaction, and the capability of these models to activate
Chem 6, 1–14, March 12, 2020
5

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 3. DFT Calculation Studies on the Adsorption and Dissociation of p-CAN
(A) Adsorption energies (Eads) of Cl on Pt-based catalyst models.
(B) Summary of the optimized structures for the adsorption (3B-a series) and dissociation (3B-d
series) of p-CAN on Pt-based catalysts.
(C and D) Eads (C) and dissociation energies (Edis) (D) of p-CAN versus energy barriers (Ea) of p-CAN
dehalogenation on Pt-based catalysts.
The configurations in B: Pt (111) (a-1, d-1), Pt1/CeO2 (a-2, d-2), Pt4/CeO2 (a-3, d-3), Pt19/CeO2 (a-4,
d-4), Pte (a-5, d-5), Ptc (a-6, d-6), Pt4/C (a-7, d-7); Ce, ; O, ; Pt, ; Cl, ; C, ; N, ; H,
.
the C-Cl bond was then evaluated in detail by first-principles calculations to explore
the key factors influencing the C-Cl bond cleavage (Figures S12–S14; Notes S1
and S2).
First of all, the adsorption of a single chlorine atom on these models was calculated
to probe their abilities to stabilize Cl since the reception of Cl would affect the ther-
modynamics of the dehalogenation of p-CAN (Figure S15). As indicated by the
results (Figure 3A), the adsorption energies (Eads) clearly displayed that a single chlo-
rine atom was preferentially adsorbed on the edge and corner sites rather than
on the Pt1/CeO2-SAC and Pt (111). Further, the adsorption and dissociation of the
C-Cl bond in p-CAN on these model catalysts revealed similar trends to single chlo-
rine atom adsorption (Figures 3B–3D). On the catalysts of low-coordinated sort, the
6
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
C-Cl bond of p-CAN preferentially adsorbed at edge or corner sites with high Eads
(>À0.78 eV). However, the interaction between the C-Cl bond and Pt sites became
very weak or even repulsive on Pt1/CeO2-SAC and Pt (111), respectively. To be spe-
cific, p-CAN was only physically adsorbed on Pt1/CeO2 with an Eads of just À0.23 eV
(Figures 3B and 3C), and the C-Cl bond even tilted away from the Pt (111) surface
(Figure 3B). Note that the Eads of p-CAN on Pt (111) was actually contributed via
the aromatic ring, rather than via the C-Cl bond. Compared with the Eads of
p-CAN, the dissociation energy (Edis) is more straightforward to value the trends
for the dechlorination of p-CAN. Clearly, the dechlorination was more thermody-
namically favorable on these Pt sites, which typically have less coordinated atoms
as seen in Figure 3D.
However, the Brønsted-Evans-Polanyi or scaling relationship for the dechlorination
of C-Cl was not found on these catalysts, which indicated a different type of transi-
tion state of p-CAN on the catalysts with multiple Pt coordination environments.39–42
Then the process of C-Cl cleavage on the three kinds of catalysts was carefully
analyzed. For clarity, the Cl atom in the C-Cl bond would adjust to a stable state
as with the single Cl atom alone before the p-CAN reached the transition state on
Pte (Figure S16). The stabilization of Cl in the C-Cl bond on the Pte site would greatly
stabilize the transition state so that the energy barrier (Ea) for C-Cl cleavage was
reduced (0.37 eV). In contrast, a nucleophilic attack occurred to the C atom of the
C-Cl bond on Pt1/CeO2 and the Cl atom was radicalized before reaching the transi-
tion state (Figure S17) due to the weak interaction between Cl and Pt. The radicali-
zation of Cl would greatly increase the energy of the transition state. As a result, the
Ea for C-Cl cleavage on Pt1/CeO2 was rather high (1.78 eV). The analysis of C-Cl
cleavage on Pt (111) told the same story as Pt1/CeO2 (Figure S18). Actually, the
main energy contribution of the Ea for C-Cl cleavage on Pt (111) was derived from
the Edis of the C-Cl bond.
Based on the above analysis, it was clear that the dechlorination of p-CAN strongly
depended on the coordination environment of Pt. It can be found that the Pt sites of
Pt1/CeO2 and Pt (111), which had high energy barriers for the C-Cl cleavage, were
both confined in a 2D matrix. As for edge and corner sites, the former ones could
be considered as being localized in 1D Pt wires and the corner ones were just like
adatoms on the Pt (111) surface. From a geometric view, these sites had different
steric hindrance, which would be a key factor influencing the catalytic performance.
The cone angle, as shown in Figure 4A, was defined to quantify this steric hindrance,
since it was easy to understand that small cone angles of Pt sites meant a broader
outer space for reactants to attack. Combined with the Ea of p-CAN (Figure 4B), it
can be seen clearly that the C-Cl cleavage would proceed more smoothly on the
Pt sites with less steric hindrance. This was consistent with the experimental data
(Figure 4C). 3% Pt/CeO2 with an average particle size of 1.7 nm, mostly exposing
edge and corner sites38 (cone angle = 88.5ꢁ), showed a relatively low selectivity to
p-CAN. Whereas, 16% Pt/CeO2, mostly exposed terrace Pt sites (cone angle =
180.0ꢁ), and 0.6% Pt/CeO2–SAC (cone angle = 147.4ꢁ) produced negligible hydro-
genolysis products. In addition, considering the complex deposited surface sites of
the CeO2, there may be other coordinated Pt single sites. In order to draw a more
accurate conclusion, other Pt1/CeO2 catalysts with different coordination environ-
ments of Pt were also simulated by DFT calculations and all these sites with a high
steric hindrance showed weak affinity (physical adsorption) for p-CAN and low disso-
ciation energies of the C-Cl bond as illustrated in Figure S19 and Table S8. This
further confirmed the conclusions we have drawn above (Note S3).
Chem 6, 1–14, March 12, 2020
7

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 4. The Steric Hindrance of Pt on the Models
(A) The cone angle of Pt on Pt-based catalyst models.
(B) The energy barriers (Ea) of dehalogenation for p-CAN versus the cone angle of Pt on Pt-based catalyst models.
(C) The selectivity of p-CAN from experiment, the absolute value of dissociation energy (Edis), and Ea of dehalogenation for p-CAN versus the cone
angle of Pt on 3% Pt/CeO2, 0.6% Pt/CeO2-SAC, and 16% Pt/CeO2, respectively. Ce, ; O, ; Pt,
.
DFT Calculation Studies on the Orbital Analysis
In order to unveil how the steric hindrance of Pt influences the C-Cl cleavage, the or-
bitals interactions between p-CAN and Pt active sites of various coordination envi-
ronments were then analyzed systematically. For molecular p-CAN alone, the occu-
pied non-bonding and highest unoccupied C-Cl p* orbitals of Cl are shown in
Figure 5A. The non-bonding p orbitals of Cl, which extend along the xy plane (de-
noted as pxy orbitals of Cl) and occupied by lone electron pairs, give rise to a large
space expansion (Figure S20). In contrast, the non-bonding pz orbitals and C-Cl p*
orbitals are relatively more shrank and in turn wrapped in pxy orbitals, forming a
‘‘core-shell’’ structure with pz and C-Cl p* orbitals as cores and pxy orbitals as shells
(Figure 5A). Since it is necessary to fill electrons into the C-Cl p* orbitals to activate
and dissociate the C-Cl bond,18 the d orbitals of Pt had to interact with the lone
electron pairs on the non-bonding pxy orbitals of Cl before reaching the C-Cl p*
orbitals according to the core-shell electronic structure. Considering the require-
ment of orbital symmetry,43–45 the orbitals of the representative catalysts, i.e.,
Pt (111), Pt1/CeO2, and Pte were then calculated, respectively. As expected, great
distinctions could be visualized among different coordinated Pt sites (Figures S21–
S23). As for Pt (111) terrace, influenced by the adjacent Pt atoms, the shape of the
Pt d orbitals has changed a lot. The most intensive peak shown in projected density
2
2
of states (DOS) was attributed to the orbitals of dz character. However, the dz
orbitals cannot interact effectively with the non-bonding orbitals of Cl due to the
8
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 5. The Orbital Analysis between P-CAN and Different Coordinated Pt Sites
(A) The 0.005 eÀ/A3 isosurfaces of C-Cl p* orbitals (unoccupied) and non-bonding orbitals of Cl in
˚
p-Journal of Cancer Molecules.
3
(B –D) The 0.0005 eÀ/A isosurfaces orbital interaction between Cl in p-CAN and Pt (111) (B),
˚
Pt1/CeO2 (C), and Pte (D) with the energy E0 in the correspondingly projected density of states,
respectively. Projected density of states of Pt1/CeO2 was shown in Figure S24.
(E and F) Projected density of states of d orbitals of Pt, p orbitals of Cl and C in p-CAN adsorbed on
Pt (111) (E) and Pte (F), respectively. (, Non-bonding orbitals of Cl.
unmatched orbital symmetry between them. Furthermore, under the influence of the
adjacent coordinated atoms, two other major orbitals extending along the z axis of
Pt had lost the dxz and dyz characters and thereby could not interact with the non-
bonding pxy orbitals (Figure S21). Thus, C-Cl bond would be repelled from the sur-
face due to electrostatic repulsion as shown in Figure 5B. This could also be
confirmed by the fact that the peak attributed to the non-bonding orbitals of
Cl were broadened (Figure 5E) when p-CAN adsorbed on Pt (111). In the case of
Pt1/CeO2, the occupation of O vacancy made the Pt single atom partially embedded
in CeO2 (110), which resulted in a relatively high cone angle of Pt in the model
(147.4ꢁ) (Figure 4A). The highest occupied orbitals E1 could be assigned to be the
d-p p* orbitals of the Pt-O bonds (Figure S22). As shown in Figure 5C, E1 could
not interact effectively with the non-bonding pxy orbitals of Cl either because of
the incompatible orbital symmetry. This would again lead to the electrostatic
Chem 6, 1–14, March 12, 2020
9

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
repulsion between the non-bonding pxy orbitals of Cl and the predominant orbitals
on Pt1/CeO2, which is similar to the situation on Pt (111) (Figure 5B). Detailed phase
distributions of other chosen orbitals were shown in Figure S22 in the Supplemental
Information, in which they all encountered the same situation. However, the main or-
bitals on the low-coordinated Pt sites with less steric hindrance were quite different
from those mentioned above. Due to the absence of adjacent Pt atoms in the xy
plane, the dxz orbitals of Pte maintained their original characters and became the
most intensive orbitals (Figure S23). The phase of these orbitals matched very well
with that of the non-bonding pxy orbitals of Cl, which turned the electrostatic repul-
sion items on the high-coordinated catalysts into attractive ones and then promoted
the cleavage of C-Cl on Pt via the adsorption of Cl (Figures 5D, 5F, and S25).
Self-Induced Enhancement of Selectivity
Based on the discussion above, it is clear that the selectivity of p-CAN would be
poor on the small particles, which contain more low-coordinated Pt sites with
less steric hindrance responsible for dehalogenation. In principle, the selectivity
of p-CAN should be constant if the ratio of low-coordinated and terrace Pt sites
remained unchanged under certain reaction conditions. However, the selectivity
of p-CAN for 3% Pt/CeO2 rose from 76.4% to 98.8% as the ratio of Pt/p-CNB
decreased from 31.7 to 2.3 wt & (Figure 6A; Table S9). A similar phenomenon
was also observed on 2% Pt/C (Figure 6B; Table S10). Furthermore, the selectivity
continued to rise before the complete conversion of p-CNB at the ratio of
2.3 wt &, whereas it decreased monotonously as reported in many other reactions
when raising the ratio to 31.7 wt & (Figures 6C, 6D, and S26). According to a pre-
vious report,46 the in-situ generated Cl species by dehalogenation could poison
the low-coordinated Pt sites preferentially. It might well be speculated that
this Cl-induced poisoning would actually increase the steric hindrance of surface
Pt, inhibit the dehalogenation, and finally, lead to a rise of the selectivity toward
p-CAN during the conversion of p-CNB.
To gain a deeper understanding of the in-situ passivation by Cl species, the infrared
spectroscopy of adsorbed CO was performed on the fresh and used 3% Pt/CeO2
samples. This method has previously been demonstrated to be effective in identifying
the metal sites. For the fresh 3% Pt/CeO2 sample, the signals of both linear-bond and
bridged-bond CO stretching were observed (Figure 7A). In the region above
2,000 cmÀ1, there were three peaks linked to the linear adsorption of CO molecule
on the surface of the catalyst. The peaks at 2,102 and 2,064 cmÀ1 were attributed
to CO linearly bonded on Ptd+ and metallic Pt (111) facets, respectively, whereas
the shoulder at 2,034 cmÀ1 was associated to CO molecules adsorbed on the low-
coordinated Pt atoms of the small particles.47–49 Additionally, the adsorption of CO
also produced a strong vibration band at 1,936, and 1,849 cmÀ1. According to Bazin
et al.,47 the band at 1,936 cmÀ1 was attributed to bridge-adsorbed CO species bound
to both low-coordinated Pt and Ce at the support surface. As for the peaks at
1,849 cmÀ1, it should be attributed to the bridge-adsorbed CO molecules on the
Pt (111).49,50 However, the spectra of the used catalysts showed many differences
from the fresh ones. As for the used 3% Pt/CeO2 sample, it was collected after
the reaction at the conversion of approximately 70% p-CNB with a 2.3 wt & ratio
of Pt/p-CNB (Figure 7B). The linear-bonded CO stretching at 2,108 and 2,082 cmÀ1
for Ptd+ and metallic Pt (111) facets still existed. While the peaks at 2,034 and
1,936 cmÀ1 almost disappeared, this suggested that these sites were blocked by Cl
produced in the reaction,51 which can explain the high selectivity under this reaction
condition. As for another used 3% Pt/CeO2 sample with the ratio of Pt/p-CNB
increasing to 31.7 wt & (Figure 7C), the peak at 2,034 cmÀ1 for CO adsorbed on
10
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 6. Catalytic Performance of 3% Pt/CeO2 and 2% Pt/C with Different Ratios of Pt/p-CNB
(A and B) Selectivity change in p-CNB hydrogenation on 3% Pt/CeO2 (A) and 2% Pt/C (B) with
different Pt loadings.
(C and D) Conversion and selectivity versus reaction time over 3% Pt/CeO2 catalyst with different
ratio of Pt/p-CNB. Reaction conditions: 15 mL ethanol, 1 MPa H2, 40ꢁC, 1.2 mmol p-CNB; Pt/p-CNB:
(C) 2.3 wt &, (D) 31.7 wt &.
the low-coordinated Pt sites also disappeared, whereas bands at 1,936 cmÀ1 sur-
vived, suggesting that there still existed a part of low-coordinated Pt sites responsible
for poor selectivity toward p-CAN. All the in-situ poisoning led to increase of the steric
hindrance of the exposed Pt site, which should account for the enhancement of selec-
tivity toward p-CAN when increasing the ratio of Pt/p-CNB in the reaction. And this
was further confirmed by catalytic performance with Cl-containing additives (Table
S11; Note S4).
DISCUSSION
In summary, the unconventional size dependence toward selectivity in the hydroge-
nation of p-CNB over CeO2-supported Pt catalysts have been extensively investi-
gated at the atomic level combining both experimental and theoretical analyses.
An inverse volcano curve of the selectivity to p-CAN against the size of Pt active sites
was obtained, in which both atomic-dispersed Pt1/CeO2 and catalysts with large Pt
NPs exhibited desirable selectivity to p-CAN. This means that high activity, high
selectivity, and theoretical 100% atom utilization could be achieved at the same
time by using this single-atom catalyst. Further analysis illustrated that the Pt coor-
dination environment affected the dehalogenation of p-CAN and determined the
final yield. High-coordinated Pt sites with a high steric hindrance were quite inert
to dehalogenation, whereas the low-coordinated ones with less steric hindrance
showed notable activity to the C-Cl bond cleavage. The electrostatic repulsion
between the non-bonding pxy orbitals of Cl and the predominant orbitals on high-
coordinated Pt sites was speculated to be the main electronic-structure factor
derived from the Pt coordination environment for the suppression of the dehaloge-
nation of p-CAN.
Chem 6, 1–14, March 12, 2020
11

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
Figure 7. Infrared Spectroscopy of CO Adsorption with Prolonging Purging Time on Different
Catalysts
(A) The fresh 3% Pt/CeO2 samples.
(B and C) The used 3% Pt/CeO2 samples with Pt/p-CNB ratio of 2.3 wt & (B) and 31.7 wt & (C) in the
reaction.
EXPERIMENTAL PROCEDURES
The Pt/CeO2 catalysts were prepared by impregnation method and the percentage
of Pt in this paper is quality score. In a typical synthesis process of 0.6 wt % Pt/CeO2
(denoted as 0.6% Pt/CeO2-SAC), 240 mL of H2PtCl6 solution (37.5%, 20 mg/mL) as
the precursor was added into a 20 mL crucible containing 10 mL deionized water un-
der magnetic stirring for 10 min. Then 300 mg of the porous nanorods CeO2 was
added into the mixture. The suspension was allowed to be heated to 50ꢁC and
kept under vigorous stirring. After the water was evaporated overnight, the catalyst
was collected and treated under H2 at 200ꢁC for 2 h and then cooled to room
temperature to obtain the final product. The synthesis of other Pt/CeO2 catalysts
with various Pt loading was similar to that of 0.6% Pt/CeO2-SAC but with different
dosages of the Pt precursor (denoted as n% Pt/CeO2, n was the Pt loading). These
catalysts, which denoted as Pt/CeO2–N were prepared in the same way mentioned
above with the precursor of [Pt(NH3)4](NO3)2. The hydrogenation of p-CNB was car-
ried out in a 50 mL stainless steel high-pressure batch reactor. Typically, p-CNB
(0.8 mmol), the catalyst Pt/CeO2, and C2H5OH (15 mL) were loaded into the auto-
clave. First, the reactor was purged three times with pure H2 to remove air. After
that, the reactor was charged with 1 MPa H2 and the reaction mixture was stirred
at 40ꢁC. After the reaction, the reactor was placed into a water bath and cooled to
room temperature. The remaining H2 was carefully vented and the reaction mixture
was centrifuged to obtain the supernatant and the supernatant was followed to be
analyzed by GC.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.chempr.
2019.12.029.
ACKNOWLEDGMENTS
Financial support from the National Key R&D Program of China (2016YFA0202900),
the National Natural Science Foundation of China (21622308, 21908189, and
21872121), and the Key Program supported by the Natural Science Foundation of
Zhejiang Province, China (LZ18B060002) is greatly appreciated. The XAS experi-
ments were conducted in the Shanghai Synchrotron Radiation Facility (SSRF).
AUTHOR CONTRIBUTIONS
C.P.W., S.J.M., and Y.W. designed the research. C.P.W. synthesized the catalysts
and conducted the experimental work. C.P.W. and Y.Z.C. performed the DFT calcu-
lations. Z.W. assisted the experimental work. H.Z., B.B.M., and Z.J. helped with the
12
Chem 6, 1–14, March 12, 2020

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
X-ray absorption fine-structure measurements. W.T.Y, Y.O., and Y.W helped collect
HAADF-STEM images. C.P.W., S.J.M., and Y.W. co-wrote the paper. All authors dis-
cussed the results and commented on the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: August 2, 2019
Revised: September 18, 2019
Accepted: December 26, 2019
Published: January 23, 2020
REFERENCES
1. Cao, Y.L., Mao, S.J., Li, M.M., Chen, Y.Q., and
Wang, Y. (2017). Metal/porous carbon
composites for heterogeneous catalysis: old
catalysts with improved performance
promoted by N-doping. ACS Catal. 7, 8090–
8112.
11. Ca´rdenas-Lizana, F., Hao, Y., Crespo-Quesada,
M., Yuranov, I., Wang, X., Keane, M.A., and
Kiwi-Minsker, L. (2013). Selective gas phase
hydrogenation of p-chloronitrobenzene over
Pd catalysts: role of the support. ACS Catal. 3,
1386–1396.
20. Iihama, S., Furukawa, S., and Komatsu, T.
(2016). Efficient catalytic system for
chemoselective hydrogenation of
halonitrobenzene to haloaniline using PtZn
intermetallic compound. ACS Catal. 6,
742–746.
12. Shi, W., Zhang, B., Lin, Y., Wang, Q., Zhang, Q.,
and Su, D.S. (2016). Enhanced chemoselective
hydrogenation through tuning the interaction
between Pt nanoparticles and carbon
supports: insights from identical location
transmission electron microscopy and X-ray
photoelectron spectroscopy. ACS Catal. 6,
7844–7854.
2. Corma, A., and Serna, P. (2006).
Chemoselective hydrogenation of nitro
compounds with supported gold catalysts.
Science 313, 332–334.
ꢀ
21. Pietrowski, M., Zielinski, M., and
Wojciechowska, M. (2011). High-selectivity
hydrogenation of chloronitrobenzene to
chloroaniline over magnesium fluoride-
supported bimetallic ruthenium-copper
catalysts. ChemCatChem 3, 835–838.
3. Li, X.F., Wang, Z., Mao, S.J., Chen, Y.Q., Tang,
M.H., Li, H., and Wang, Y. (2018). Insight into
the role of additives in catalytic synthesis of
cyclohexylamine from nitrobenzene. Chin. J.
Chem. 36, 1191–1196.
22. Mao, S.J., Zhao, B.W., Wang, Z., Gong, Y.T., Lu,
¨
G., Ma, X., Yu, L., and Wang, Y. (2019). Tuning
the catalytic performance for the semi-
hydrogenation of alkynols by selectively
poisoning the active sites of Pd catalysts.
Green Chem. 21, 4143–4151.
13. Zuo, B.J., Wang, Y., Wang, Q.L., Zhang, J.L.,
Wu, N.Z., Peng, L.D., Gui, L.L., Wang, X.D.,
Wang, R.M., and Yu, D.P. (2004). An efficient
ruthenium catalyst for selective hydrogenation
of ortho-chloronitrobenzene prepared via
assembling ruthenium and tin oxide
4. Kattel, S., Ramırez, P.J., Chen, J.G., Rodriguez,
´
J.A., and Liu, P. (2017). Active sites for CO2
hydrogenation to methanol on Cu/ZnO
catalysts. Science 355, 1296–1299.
23. Tijani, A., Coq, B., and Figueras, F. (1991).
Hydrogenation of para-chloronitrobenzene
over supported ruthenium-based catalysts.
Appl. Catal. 76, 255–266.
nanoparticles. J. Catal. 222, 493–498.
5. Chen, Y., Wang, Z., Mao, S., and Wang, Y.
(2019). Rational design of hydrogenation
catalysts using nitrogen-doped porous carbon.
Chin. J. Catal. 40, 971–979.
14. Pietrowski, M., and Wojciechowska, M. (2009).
An efficient ruthenium-vanadium catalyst for
selective hydrogenation of ortho-
chloronitrobenzene. Catal. Today 142,
211–214.
24. Coq, B., and Figueras, F. (1998). Structure–
activity relationships in catalysis by metals:
some aspects of particle size, bimetallic and
supports effects. Coord. Chem. Rev. 178–180,
1753–1783.
6. Zhang, S., Chang, C.R., Huang, Z.Q., Li, J., Wu,
Z., Ma, Y., Zhang, Z., Wang, Y., and Qu, Y.
(2016). High catalytic activity and
chemoselectivity of sub-nanometric Pd clusters
on porous nanorods of CeO2 for
15. He, D., Shi, H., Wu, Y., and Xu, B.-Q. (2007).
Synthesis of chloroanilines: selective
hydrogenation of the nitro in
25. Coq, B., Tijani, A., and Figue´ ras, F. (1991).
Particle-size effect on the kinetics of p-
chlornitrobenzene hydrogenation over
platinum alumina catalysts. J. Mol. Catal. 68,
331–345.
hydrogenation of nitroarenes. J. Am. Chem.
Soc. 138, 2629–2637.
chloronitrobenzenes over zirconia-supported
gold catalyst. Green Chem. 9, 849–851.
7. Chen, Y., Kong, X., Mao, S., Wang, Z., Gong, Y.,
and Wang, Y. (2019). Study of the role of
alkaline sodium additive in selective
hydrogenation of phenol. Chin. J. Catal. 40,
1516–1524.
16. Boronat, M., Concepcio´ n, P., Corma, A.,
Gonza´lez, S., Illas, F., and Serna, P. (2007). A
molecular mechanism for the chemoselective
hydrogenation of substituted nitroaromatics
with nanoparticles of gold on TiO2 catalysts: A
cooperative effect between gold and the
support. J. Am. Chem. Soc. 129, 16230–16237.
26. Lyu, J., Wang, J., Lu, C., Ma, L., Zhang, Q., He,
X., and Li, X. (2014). Size-dependent
halogenated nitrobenzene hydrogenation
selectivity of Pd nanoparticles. J. Phys. Chem. C
118, 2594–2601.
8. Blaser, H.-U., Steiner, H., and Studer, M. (2009).
Selective catalytic hydrogenation of
functionalized nitroarenes: an update.
ChemCatChem 1, 210–221.
17. Tafesh, A.M., and Weiguny, J. (1996). A review
of the selective catalytic reduction of aromatic
nitro compounds into aromatic amines,
isocyanates, carbamates, and ureas using CO.
Chem. Rev. 96, 2035–2052.
27. Qiao, B.T., Wang, A., Yang, X., Allard, L.F.,
Jiang, Z., Cui, Y., Liu, J., Li, J., and Zhang, T.
(2011). Single-atom catalysis of CO oxidation
using Pt1/FeOx. Nat. Chem. 3, 634–641.
9. Wei, H., Liu, X., Wang, A., Zhang, L., Qiao, B.,
Yang, X., Huang, Y., Miao, S., Liu, J., and Zhang,
T. (2014). FeOx-supported platinum single-
atom and pseudo-single-atom catalysts for
chemoselective hydrogenation of
28. Nie, L., Mei, D., Xiong, H., Peng, B., Ren, Z.,
Hernandez, X.I.P., DeLaRiva, A., Wang, M.,
Engelhard, M.H., Kovarik, L., et al. (2017).
Activation of surface lattice oxygen in single-
atom Pt/CeO2 for low-temperature CO
oxidation. Science 358, 1419–1423.
18. Wang, X., Liang, M., Zhang, J., and Wang, Y.
(2007). Selective hydrogenation of aromatic
chloronitro compounds. Curr. Org. Chem. 11,
299–314.
functionalized nitroarenes. Nat. Commun. 5,
5634.
10. Liang, M., Wang, X., Liu, H., Liu, H., and Wang,
Y. (2008). Excellent catalytic properties over
nanocomposite catalysts for selective
hydrogenation of halonitrobenzenes. J. Catal.
255, 335–342.
19. Pietrowski, M. (2012). Recent developments in
heterogeneous selective hydrogenation of
halogenated nitroaromatic compounds to
halogenated anilines. Curr. Org. Synth. 9,
470–487.
29. Lin, L., Zhou, W., Gao, R., Yao, S., Zhang, X., Xu,
W., Zheng, S., Jiang, Z., Yu, Q., Li, Y.W., et al.
(2017). Low-temperature hydrogen production
Chem 6, 1–14, March 12, 2020
13

Please cite this article in press as: Wang et al., Insight into Single-Atom-Induced Unconventional Size Dependence over CeO2-Supported Pt
Catalysts, Chem (2019), https://doi.org/10.1016/j.chempr.2019.12.029
from water and methanol using Pt/alpha-MoC
catalysts. Nature 544, 80–83.
into chemoselective metal catalysts for the
hydrogenation of substituted nitroaromatics.
J. Am. Chem. Soc. 130, 8748–8753.
45. Van Santen, R.A., Neurock, M., and Shetty, S.G.
(2010). Reactivity theory of transition-metal
surfaces: a BrønstedÀEvansÀPolanyi linear
activation energyÀfree-energy analysis. Chem.
Rev. 110, 2005–2048.
30. Kyriakou, G., Boucher, M.B., Jewell, A.D.,
Lewis, E.A., Lawton, T.J., Baber, A.E., Tierney,
H.L., Flytzani-Stephanopoulos, M., and Sykes,
E.C.H. (2012). Isolated metal atom geometries
as a strategy for selective heterogeneous
hydrogenations. Science 335, 1209–1212.
37. Liu, P., Qin, R., Fu, G., and Zheng, N. (2017).
Surface coordination chemistry of metal
nanomaterials. J. Am. Chem. Soc. 139, 2122–
2131.
46. Gracia, F.J., Miller, J.T., Kropf, A.J., and Wolf,
E.E. (2002). Kinetics, FTIR, and controlled
atmosphere EXAFS study of the effect of
chlorine on Pt-supported catalysts during
oxidation reactions. J. Catal. 209, 341–354.
38. Crespo-Quesada, M., Yarulin, A., Jin, M., Xia,
Y., and Kiwi-Minsker, L. (2011). Structure
sensitivity of alkynol hydrogenation on shape-
and size-controlled palladium nanocrystals:
which sites are most active and selective?
J. Am. Chem. Soc. 133, 12787–12794.
31. Feng, Q., Zhao, S., Wang, Y., Dong, J., Chen,
W., He, D., Wang, D., Yang, J., Zhu, Y., Zhu, H.,
et al. (2017). Isolated single-atom Pd sites in
intermetallic nanostructures: high catalytic
selectivity for semihydrogenation of alkynes.
J. Am. Chem. Soc. 139, 7294–7301.
47. Bazin, P., Saur, O., Lavalley, J.C., Daturi, M., and
Blanchard, G. (2005). FT-IR study of CO
adsorption on Pt/CeO2: characterisation and
structural rearrangement of small Pt particles.
Phys. Chem. Chem. Phys. 7, 187–194.
39. Hammer, B. (2006). Special sites at noble and
late transition metal catalysts. Top. Catal. 37,
3–16.
32. Wang, X., Chen, W., Zhang, L., Yao, T., Liu, W.,
Lin, Y., Ju, H., Dong, J., Zheng, L., Yan, W., et al.
(2017). Uncoordinated amine groups of metal-
organic frameworks to anchor single Ru sites as
chemoselective catalysts toward the
48. Ke, J., Zhu, W., Jiang, Y., Si, R., Wang, Y.-J., Li,
S.-C., Jin, C., Liu, H., Song, W.-G., Yan, C.-H.,
et al. (2015). Strong local coordination
structure effects on subnanometer PtOx
clusters over CeO2 nanowires probed by low-
temperature CO oxidation. ACS Catal. 5,
5164–5173.
40. Nørskov, J.K., Bligaard, T., Logadottir, A.,
Bahn, S., Hansen, L.B., Bollinger, M., Bengaard,
H., Hammer, B., Sljivancanin, Z., Mavrikakis, M.,
et al. (2002). Universality in heterogeneous
catalysis. J. Catal. 209, 275–278.
hydrogenation of quinoline. J. Am. Chem. Soc.
139, 9419–9422.
33. Yang, J., Chen, B., Liu, X., Liu, W., Li, Z., Dong,
J., Chen, W., Yan, W., Yao, T., Duan, X., et al.
(2018). Efficient and robust hydrogen evolution:
phosphorus nitride imide nanotubes as
41. Zaffran, J., Michel, C., Auneau, F., Delbecq, F.,
and Sautet, P. (2014). Linear energy relations as
predictive tools for polyalcohol catalytic
reactivity. ACS Catal. 4, 464–468.
49. Kappers, M.J., and van der Maas, J.H. (1991).
Correlation between CO frequency and Pt
coordination number. A DRIFT study on
supports for anchoring single ruthenium sites.
Angew. Chem. Int. Ed. Engl. 57, 9495–9500.
supported Pt catalysts. Catal. Lett. 10, 365–373.
42. Ferna´ndez, E.M., Moses, P.G., Toftelund, A.,
Hansen, H.A., Martı´nez, J.I., Abild-Pedersen, F.,
Kleis, J., Hinnemann, B., Rossmeisl, J., Bligaard,
T., et al. (2008). Scaling relationships for
adsorption energies on transition metal oxide,
sulfide, and nitride surfaces. Angew. Chem. Int.
Ed. Engl. 47, 4683–4686.
34. Xiao, L., and Wang, L. (2004). Structures of
platinum clusters: planar or spherical? J. Phys.
Chem. A 108, 8605–8614.
50. Aleksandrov, H.A., Neyman, K.M., Hadjiivanov,
K.I., and Vayssilov, G.N. (2016). Can the state of
platinum species be unambiguously
determined by the stretching frequency of an
adsorbed CO probe molecule? Phys. Chem.
Chem. Phys. 18, 22108–22121.
35. Bourane, A., Dulaurent, O., and Bianchi, D.
(2000). Heats of adsorption of linear and
multibound adsorbed CO species on a Pt/
Al2O3 catalyst using in situ infrared
43. Fukui, K.e.a. (1982). Role of frontier obitals in
51. Mare´ cot, P., Fakche, A., Kellali, B., Mabilon, G.,
Prigent, P., and Barbier, J. (1994). Propane and
propene oxidation over platinum and
chemical-reacrions. Science 218, 747–754.
spectroscopy under adsorption equilibrium.
J. Catal. 196, 115–125.
44. Hoffmann, R. (1987). How chemistry and
physics meet in the solid-state. Angew. Chem.
Int. Ed. Engl. 26, 846–878.
36. Corma, A., Serna, P., Concepcio´ n, P., and
Calvino, J.J. (2008). Transforming nonselective
palladium on alumina: effects of chloride and
water. Appl. Catal. B: Environ. 3, 283–294.
14
Chem 6, 1–14, March 12, 2020