Subnanometric Hybrid Pd-M(OH)2, M?= Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts for Hydrogen Generation

Article abstract of DOI:10.1016/j.chempr.2017.07.001

Hybrid multi-metallic nanocatalysts have attracted increasing attention because of the synergistic effect of metal species and considerably improved catalytic performance, but they often suffer from severe sintering and poor stability. Here, we show a facile strategy for preparing subnanometric hybrid bimetallic clusters Pd-M(OH)₂ (M = Ni, Co) within silicalite-1 (S-1) zeolite via a hydrothermal synthesis method. The hybrid bimetallic nanocatalysts exhibit excellent shape-selective catalytic performance and superior thermal stability. The incorporation of secondary Ni(OH)₂ species in S-1 can considerably increase the catalytic activity of the Pd nanoclusters for the dehydrogenation of formic acid (FA) as a result of the electron-enriched Pd surface and bimetallic interfacial effect. Notably, the 0.8Pd0.2Ni(OH)₂@S-1 catalyst affords the highest initial turnover frequency value, up to 5,803 hr^?1 toward complete FA decomposition without any additives at 60°C. The superior catalytic properties and excellent stability of the subnanometric hybrid bimetallic clusters confined in zeolites create new prospects for their practical high-performance catalytic application.

Full text of DOI:10.1016/j.chempr.2017.07.001

Article
Subnanometric Hybrid Pd-M(OH) , M = Ni, Co,
2
Clusters in Zeolites as Highly Efficient
Nanocatalysts for Hydrogen Generation
Qiming Sun, Ning Wang,
Qiming Bing, ..., Peng Zhang,
Mingjun Jia, Jihong Yu
jihong@jlu.edu.cn
HIGHLIGHTS
Synthesis of subnanometric
hybrid bimetallic clusters within
zeolites
The nanocatalysts possessed
excellent activity for hydrogen
generation
The nanocatalysts exhibited
shape-selective catalytic
performance
The nanocatalysts showed
superior thermal stability and
recyclability
Subnanometric hybrid bimetallic palladium-transition metal hydroxide clusters are
encapsulated within zeolites by using metallic ethylenediamine complexes as
precursors under direct hydrothermal conditions. The hybrid bimetallic
nanocatalysts exhibit shape-selective catalytic performance, superior thermal
stability, as well as exceedingly high dehydrogenation efficiency toward complete
formic acid decomposition as a result of the nanoconfinement effect of zeolites as
well as the synergistic effect of hybrid metal species.
Sun et al., Chem 3, 1–17
September 14, 2017 ª 2017 Elsevier Inc.
http://dx.doi.org/10.1016/j.chempr.2017.07.001

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Article
Subnanometric Hybrid Pd-M(OH) , M = Ni, Co,
2
Clusters in Zeolites as Highly Efficient
Nanocatalysts for Hydrogen Generation
Qiming Sun,1,7 Ning Wang,1,7 Qiming Bing, Rui Si, Jingyao Liu, Risheng Bai, Peng Zhang,
2
3
2
1
4
Mingjun Jia, and Jihong Yu1,6,8,
5
*
SUMMARY
The Bigger Picture
Hybrid multi-metallic nanocatalysts have attracted increasing attention
because of the synergistic effect of metal species and considerably improved
catalytic performance, but they often suffer from severe sintering and poor sta-
bility. Here, we show a facile strategy for preparing subnanometric hybrid
Hybrid multi-metallic
nanocatalysts have attracted
increasing attention because of
the synergistic effect of metal
species and considerably
2
bimetallic clusters Pd-M(OH) (M = Ni, Co) within silicalite-1 (S-1) zeolite via a
hydrothermal synthesis method. The hybrid bimetallic nanocatalysts exhibit
excellent shape-selective catalytic performance and superior thermal stability.
improved catalytic activity but
often suffer from severe sintering
and poor stability. Zeolites are
known as ideal supports for
confinement synthesis of metal
nanoparticles. In this work,
2
The incorporation of secondary Ni(OH) species in S-1 can considerably in-
crease the catalytic activity of the Pd nanoclusters for the dehydrogenation
of formic acid (FA) as a result of the electron-enriched Pd surface and bimetallic
interfacial effect. Notably, the 0.8Pd0.2Ni(OH)
2
@S-1 catalyst affords the high-
ꢀ
1
est initial turnover frequency value, up to 5,803 hr
toward complete FA
subnanometric hybrid Pd-M(OH)
(M = Ni, Co) clusters encapsulated
within purely siliceous zeolites
were synthesized via a
2
ꢁ
decomposition without any additives at 60 C. The superior catalytic properties
and excellent stability of the subnanometric hybrid bimetallic clusters confined
in zeolites create new prospects for their practical high-performance catalytic
application.
hydrothermal synthesis method.
The hybrid bimetallic
nanocatalysts exhibit superior
INTRODUCTION
ꢁ
ꢁ
thermal stability at 600 C–700 C
and afford the highest initial
turnover frequency value, up to
Supported metal clusters have been widely used in diverse heterogeneous
catalysis reactions for many important industrial processes, such as petroleum
1
–5
refining, hydrogenation, and conversion of automobile exhaust.
The size of
ꢀ
1
5
,803 hr toward complete
the metal clusters is a critical factor in determining the performance of such cata-
formic acid decomposition
6
7,8
lysts. In general, smaller metal clusters, especially those of subnanometric
ꢁ
without any additives at 60 C. The
hybrid bimetallic nanoclusters
confined in zeolites have potential
practical application in
9
–11
or even single atom
size possess unique and improved catalytic properties
compared with larger metal clusters because of more accessibly exposed active
sites and enhanced charge transfer between the metal species and supports.
Recently, hybrid multi-metal nanoparticles have attracted widespread attention
for their significantly improved catalytic activity, which arises from the superior syn-
dehydrogenation of formic acid as
a viable and effective chemical
hydrogen storage medium for fuel
cells, and also raise more
1
2–16
ergistic effect of their multi-component interfacial active sites.
Notably, intro-
duction of non-noble metal hydroxides to form multi-metallic nanostructures can
not only minimize the use of noble metal but also enhance the catalytic perfor-
possibilities for other important
high-performance catalytic
reactions.
1
2–15
mance.
cient performance for CO oxidation, and Pt/M(OH)
can greatly enhance the activity of the H evolution reaction. However, small
For example, hybrid Pt/FeNi(OH)
x
nanoparticles exhibited highly effi-
1
2
2
(M = Ni, Co) nanoparticles
1
4
2
multi-metal nanoparticles, especially non-noble metal-containing nanoparticles,
tend to suffer from severe aggregation, poor hydrothermal stability, and high
metal leaching during catalytic reaction or thermal treatment because of their
Chem 3, 1–17, September 14, 2017 ª 2017 Elsevier Inc.
1

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
considerable increase in surface free energy at smaller sizes. Generation of
multi-metal nanoparticles, especially non-noble metal hydroxide-containing multi-
metallic nanoparticles with ultrasmall particle size and excellent thermal stability,
is still a great challenge.
Zeolites with ordered micropores (<2 nm), well-defined crystal structure, and
high hydrothermal stability are the ideal supports for confinement synthesis of
1
7–21
metal nanoparticles.
The zeolite matrix can protect nanoparticles against
aggregation, thereby enhancing their catalytic activity and stability. Unfortunately,
metal nanoparticles deposited onto zeolites using conventional ion-exchange
and wetness impregnation methods tend to form non-uniform metal particles
with large particle sizes and exhibit poor thermal stability, resulting in suboptimal
2
2
catalytic activity.
methods for encapsulating series of small noble mono-/bimetal clusters (1–2 nm)
Lately, Iglesia and co-workers developed ligand-stabilized
2
3–27
into microporous cavities of different aluminosilicate zeolites.
However,
introducing Al into zeolites may restrict the catalytic application of these
materials because of the resultant undesirable acidity and decreased hydrothermal
8
stability. Notably, Liu et al. prepared subnanometric Pt clusters within purely sili-
ceous MCM-22 (MWW zeotype) zeolite during the growth of a two-dimensional
zeolite into three dimensions. The catalysts exhibited exceptional high stability
and shape-selective catalytic performance but suffered from 70% weight loss of
2
8
Pt after oxidation-reduction treatments. Very recently, we have developed a
facile method to synthesize ultrasmall Pd clusters encapsulated within purely sili-
2 2 2 2 2 2
ceous silicalite-1 (S-1) zeolite (MFI) by using [Pd(NH CH CH NH ) ]Cl as precursor
2
9
under direct hydrothermal conditions. The as-synthesized catalyst exhibited
excellent catalytic activity for the dehydrogenation of formic acid (FA) in portable
fuel cell applications as well as superior stability as a result of the well-confined
3
0–33
Pd clusters within the silica zeolite matrix. Recently, Dai et al.
successfully
encapsulated some mono-/bimetallic (oxide) nanoparticles inside the hollow zeo-
lites by a top-down dissolution-recrystallization method. However, until now, there
has been no report of ultrasmall non-noble metal hydroxide-containing multi-
metallic nanoclusters in situ confined within purely siliceous zeolites that are
capable of promoting catalytic performance and thermal stability.
1State Key Laboratory of Inorganic Synthesis and
Preparative Chemistry, College of Chemistry, Jilin
University, Changchun 130012, People’s Republic
of China
In this work, we have demonstrated a facile and one-pot synthesis strategy to pre-
2
pare subnanometric hybrid bimetallic clusters Pd-M(OH) (M = Ni, Co) confined
2Laboratory of Theoretical and Computational
Chemistry, Institute of Theoretical Chemistry, Jilin
University, Changchun 130023, People’s Republic
of China
within purely siliceous S-1 zeolite nanocrystals using metallic ethylenediamine
complex mixtures as precursors via a hydrothermal synthesis method. The subnano-
2
metric sizes of the hybrid bimetallic Pd and Ni(OH) clusters were confirmed using
3Shanghai Synchrotron Radiation Facility,
Shanghai Institute of Applied Physics, Chinese
Academy of Sciences, Shanghai 201204, People’s
Republic of China
scanning transmission electron microscopy (STEM) and extended X-ray absorption
fine structure (EXAFS) analyses. The nanocatalysts obtained demonstrated superior
thermal stability under industrial conditions and excellent shape-selective catalytic
4Department of Chemistry, Dalhousie University,
Halifax, NS B3H 4R2, Canada
performance. Most significantly, 0.8Pd0.2Ni(OH)
2
@S-1 afforded exceedingly high
ꢀ1
ꢀ1
initial and total turnover frequency (TOF) values up to 5,803 hr and 1,879 hr
5Key Laboratory of Surface and Interface
Chemistry of Jilin Province, College of Chemistry,
Jilin University, Changchun 130021, People’s
Republic of China
for the complete decomposition of FA; notably these TOF values were achieved
ꢁ
without generating CO, without using any additive at 60 C, making this among
the highest activities for heterogeneously catalyzed FA decomposition under similar
6International Center of Future Science, Jilin
University, Changchun 130012, People’s Republic
of China
3
4–36
conditions.
The superior catalytic performance, excellent thermal stability, and
recyclability of hybrid non-noble metal hydroxide-containing bimetallic clusters
confined in zeolites not only facilitate their practical application in the dehydrogena-
tion of FA as a viable and effective chemical hydrogen storage medium for fuel cells
but also provide opportunities for other important high-performance catalytic
reactions.
7These authors contributed equally
8Lead Contact
*Correspondence: jihong@jlu.edu.cn
http://dx.doi.org/10.1016/j.chempr.2017.07.001
2
Chem 3, 1–17, September 14, 2017

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Figure 1. Schematics of the Preparation of Hybrid Bimetallic Pd-Ni(OH)
2
@S-1 Catalyst
2
+
2+
Pd(en)
2
and Ni(en)
3
complexes are added into the synthetic gel, and the gel is then crystallized in the autoclave under conventional hydrothermal
ꢁ
conditions at 170 C for 4 days. The metal complexes interact with the initial zeolitic gel and are encapsulated within the zeolite framework during the
crystallization process. After calcination in air and reduction by H
zeolite.
2 2
, the subnanometric Pd-Ni(OH) clusters are finally confined inside the nanosized S-1
RESULTS AND DISCUSSION
Synthesis and Characterization of Pd-Ni(OH)
2
@S-1 Samples
clusters encapsulated
@S-1) is illustrated in
@S-1 samples were
The overall synthesis of subnanometric bimetallic Pd-Ni(OH)
2
within nanosized S-1 zeolites (denoted as xPd(1 ꢀ x)Ni(OH)
2
Figure 1. For comparison, monometallic Pd@S-1 and Ni(OH)
2
also prepared under similar conditions. X-ray diffraction analysis of all samples
confirmed the intended MFI zeolite structure, indicating that the introduction of
additional metal species did not interfere with zeolite crystallization (Figure S1).
Inductively coupled plasma atomic emission spectroscopy (ICP-AES) analysis
confirmed that all the bimetallic samples possessed similar total metal loading
amounts (0.63–0.64 wt %). The molar ratios of Pd/Ni in the products were 0.91/
0
.09, 0.83/0.17, and 0.73/0.27, respectively, in accordance with those in the initial
gels (Table S1).
Scanning electron microscopy (SEM) and transmission electron microscopy (TEM)
images show that the metal clusters are uniformly distributed throughout the nano-
sized S-1 zeolite crystals (200–300 nm) (Figures S2 and S3). The visually observed
average bimetallic clusters sizes in all of the samples are ꢂ1.7 nm, similar to
those of the monometallic Pd and Ni(OH)
are much smaller than those formed in the Pd/S-1-im and Pd-Ni(OH)
ꢂ2.9 nm) catalysts prepared with the impregnated method, as well as those formed
in the Pd@S-1/Ni(OH) -im catalysts by impregnating Ni(NO solution into Pd@S-1
2
clusters. Those bimetallic cluster sizes
2
/S-1-im
(
2
3 2
)
catalyst (ꢂ3.3 nm) (Figures S4 and S5). The ultrasmall diameters and monodispersed
size distributions of bimetallic clusters in these zeolites indicate that the in situ
confinement of metal clusters in zeolite matrices can effectively inhibit aggregation
2
of the metal clusters. N
2
adsorption-desorption analyses reveal that about 200 m /g
3
of micropore area and 0.10 cm /g of micropore volume exist in metal-containing
2
samples, ensuring sufficient diffusion for the reactant, with only about 30 m /g
3
and 0.01 cm /g decrease in micropore area and micropore volume compared with
the S-1 zeolite, as a result the partial occupation of metal species in the channels
of zeolites (Figure S6 and Table S1).
Chem 3, 1–17, September 14, 2017
3

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Figure 2. Elemental Mapping and Cs-Corrected STEM Images
(
A and B) STEM image (A) and HAADF-STEM image (B) of 0.8Pd0.2Ni(OH)
corresponding EDX mapping images for Si, Pd, and Ni elements.
C–E) Cs-corrected STEM images of 0.8Pd0.2Ni(OH) @S-1 catalyst viewed along b axis orientation
2
@S-1 catalyst and the
(
2
with low magnification (C), viewed along b axis orientation with high magnification (D), and viewed
along other orientation (E).
(
F–H) Cs-corrected STEM images of Pd@S-1 catalyst viewed along b axis orientation with low
magnification (F), viewed along b axis orientation with high magnification (G), and viewed along
other orientation (H).
Scale bars represent 1 mm (A), 200 nm (B), 20 nm (C, E, F, and H), and 10 nm (D and G).
2
The element distribution of the 0.8Pd0.2Ni(OH) @S-1 sample was determined by
STEM imaging, high-resolution high-angle annular dark-field (HAADF)-STEM imag-
ing, and energy-dispersive X-ray (EDX) spectral mapping in Figures 2A and 2B. The
Pd and Ni elements are uniformly located at nearly identical positions, indicating the
formation of bimetallic clusters. The corresponding EDX spectrum further revealed
the presence of Pd and Ni elements in the sample (Figure S7).
To further investigate the location of metal clusters in the zeolites, Cs-corrected
2
HAADF-STEM was performed on 0.8Pd0.2Ni(OH) @S-1 (Figures 2C–2E) compared
with Pd@S-1 (Figures 2F–2H). The images clearly reveal that the straight channels
along the [010] direction are empty and intact, suggesting that the metal clusters
might be located in the intersectional void spaces between the straight and sinusoi-
2
9
dal channels of the MFI structure, in accordance with our previous work. The
4
Chem 3, 1–17, September 14, 2017

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Figure 3. Characterization of Pd@S-1 and 0.8Pd0.2Ni(OH)
2
@S-1 Catalysts Using XANES, EXAFS, and XPS Measurements
A) Pd K-edge XANES spectra of Pd foil, PdO, Pd@S-1, and 0.8Pd0.2Ni(OH) @S-1 catalysts.
B) Ni K-edge XANES spectra of Ni foil, NiO, Ni(OH) , and 0.8Pd0.2Ni(OH) @S-1 catalysts.
C) Fourier transform of k -weighted EXAFS spectra of Pd@S-1 catalyst at Pd K-edge.
(
(
(
(
(
(
(
2
2
2
3
3
2
D) Fourier transform of k -weighted EXAFS spectra of Ni(OH) @S-1 catalyst at Ni K-edge.
3
E) Fourier transform of k -weighted EXAFS spectra of 0.8Pd0.2Ni(OH)
2
@S-1 catalyst at Pd K-edge.
@S-1 catalyst at Ni K-edge.
3
F) Fourier transform of k -weighted EXAFS spectra of 0.8Pd0.2Ni(OH)
2
2
G) Pd 3d XPS spectra of isolated metal samples after dissolving Pd@S-1 and 0.8Pd0.2Ni(OH) @S-1 catalysts in NaOH solution compared with
commercial Pd/C catalyst.
2 2 2
H) Ni 2p XPS spectra of dissolved 0.8Pd0.2Ni(OH) @S-1 and Ni(OH) @S-1samples compared with the a-Ni(OH) sample.
(
brightness of the metal clusters displaced in Cs-corrected HAADF-STEM images ap-
pears different, indicating the different locations of metal clusters within the S-1
2
9
zeolite. Furthermore, all the metal-containing samples show almost the same Si
magic angle spinning nuclear magnetic resonance (NMR) spectra as that of pure
S-1 zeolite (Figure S8), indicating that the zeolite framework remains intact after
the metal clusters are confined in the micropore of zeolite.
To provide the local structure information of the zeolite-supported metal clusters,
X-ray absorption near-edge structure (XANES) and EXAFS spectra of Pd@S-1,
Ni(OH)
and the structural parameters obtained from EXAFS fitting are summarized in Table
S2. The Ni K-edge XANES spectra of Ni(OH) @S-1 and 0.8Pd0.2Ni(OH) @S-1
strongly resembled that of the Ni(OH) reference spectrum, indicating the possible
existence of Ni(OH) species in these multi-metallic clusters. This is somewhat un-
usual, as Ni(OH) typically transforms into NiO after calcination in air, and further
converts to Ni metal after reduction under a H atmosphere. The formation of
Ni(OH) species instead of Ni metal clusters in S-1 zeolite could be explained by
the ability of the zeolite framework to stabilize ultrasmall Ni(OH) species, prevent-
ing them from undergoing further dehydration and reduction to form Ni metal spe-
cies. However, the Pd K-edge XANES spectra of Pd@S-1 and 0.8Pd0.2Ni(OH) @S-1
2 2
@S-1, and 0.8Pd0.2Ni(OH) @S-1 samples were measured (Figures 3A–3F);
2
2
2
2
2
2
2
2
2
exhibited some differences from both Pd foil and PdO reference spectra. This can be
attributed to the ultrasmall particle sizes and the strong interaction between Pd and
the oxygen atoms of Ni(OH)
increased oxidation state for these Pd atoms.
Pd K-edge absorption to the lower binding energy for 0.8Pd0.2Ni(OH)
2
or even the zeolite framework itself, leading to an
1
5,37
Notably, a small shift of the
@S-1
2
Chem 3, 1–17, September 14, 2017
5

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
can be found compared with Pd@S-1, which demonstrates that the Pd atoms
in 0.8Pd0.2Ni(OH)
2
@S-1 might possess a higher electron density than that in
Pd@S-1. In the Pd K-edge Fourier-transformed EXAFS spectrum of Pd@S-1, signifi-
˚
˚
˚
cant peaks were detected at ꢂ1.97A, ꢂ2.74A, and ꢂ3.43A; these peaks were iden-
tified as Pd–O, first shell Pd–Pd metallic bonds, and second shell Pd–Pd bonds from
3
8
the Pd–O–Pd structure, respectively. It is noteworthy that the Pd–O distance of
˚
0
.8Pd0.2Ni(OH)
2
@S-1 (2.00 G 0.01A) was slightly longer than that of Pd@S-1
1.97 G 0.01A). This also suggests a higher electron density for the Pd atoms in
.8Pd0.2Ni(OH) @S-1 as a result of less charge transfer from Pd atoms to O atoms.
In the Ni K-edge EXAFS spectra of Ni(OH) @S-1 and 0.8Pd0.2Ni(OH) @S-1, peaks of
Ni–O (2.06 G 0.01A) and Ni–Ni (3.06 G 0.01A) bonds were observed; these bond
lengths are consistent with those found in the Ni(OH) reference spectrum (Fig-
ure S9). Together, the above results demonstrate that these bimetallic clusters
comprise a combination of Pd metal and Ni(OH) species. Most significantly, it
was observed that the average coordination numbers of the Pd–Pd metallic bonds
in Pd@S-1 and 0.8Pd0.2Ni(OH) @S-1 are only 2.3 to 3.1 (Table S2), indicating that
the bimetallic Pd-Ni(OH) clusters possess ultrasmall subnanometric sizes.
˚
(
0
2
2
2
˚
˚
2
1
2
2
2
3
9
2
X-ray photoelectron spectroscopy (XPS) analyses were further used to characterize
the elemental composition of bimetallic clusters (Figures 3G and 3H). No XPS signals
were initially observed in the XPS spectrum of 0.8Pd0.2Ni(OH)
the metal clusters were totally encapsulated within the zeolites. However, when
.8Pd0.2Ni(OH) @S-1 was dissolved in NaOH solution, Pd3d5/2 and Pd3d3/2 peaks
were observed at 336.0 eV and 341.3 eV, as well as Ni2p3/2 and Ni2p1/2 peaks at
56.9 eV and 874.5 eV. These peak positions correspond to the typical zero-valent
Pd metal and divalent Ni(OH) phases, respectively, as shown by their resemblance
to peaks obtained from commercial Pd/C catalyst, dissolved Pd@S-1, dissolved
2
@S-1 because
0
2
8
2
4
0
2 2
Ni(OH) @S-1, and as-synthesized a-Ni(OH) references.
Stability Measurement
To investigate the stability of confined bimetallic Pd-Ni(OH)
2
clusters under
commonly used industrial conditions, samples of 0.8Pd0.2Ni(OH)
2
@S-1 were
ꢁ
ꢁ
calcined under various oxidation-reduction atmospheres at 600 C–700 C as well
ꢁ
as in the presence of water vapor at 600 C. As shown in Figures S10 and S11, the
2
particle sizes of bimetallic Pd-Ni(OH) clusters showed only a small increase of
ꢁ
0
.2–0.3 nm after calcination at 700 C for 2 hr under a H
2
2
and N atmosphere, respec-
ꢁ
tively. After treatment under an O
2
atmosphere at 600 C, the average size of the
metal clusters changed just slightly, increasing from 1.7 to 2.2 nm (Figure S12).
ꢁ
Moreover, after aging treatments in water vapor at 600 C, the metal clusters of
0
.8Pd0.2Ni(OH)
2
@S-1 also retained their stability, with a slight increase from 1.7
ꢁ
to 2.2 nm. Even after increasing the flow of water vapor to 18 mL/min at 600 C, a
metal cluster size of ꢂ2.6 nm was still observed (Figure S13). In contrast to the sta-
bility of these bimetallic clusters, the metal particles deposited on the Pd/S-1-im
and 0.8Pd0.2Ni(OH)
2
/S-1-im samples using the impregnation method aggregated
ꢁ
dramatically after calcination at 700 C for 2 hr under a N
2
atmosphere, resulting in
increased particle sizes up to 5.3 and 6.3 nm, respectively. This further confirms
that the confinement effect of zeolites can considerably improve the thermal stability
of the encapsulated metal clusters. Notably, the samples of Pd@S-1 and
0
.8Pd0.2Ni(OH)
2
@S-1 also exhibited exceptional stability after 4-fold cycles of
ꢁ
oxidation-reduction treatment at 650 C, showing just slightly increased particle
sizes (by ꢂ0.4 nm) (Figures S14 and S15). Furthermore, the crystal morphology, crys-
2
tallinity, and metal loading amount of 0.8Pd0.2Ni(OH) were nearly unchanged after
treatment at high temperature (Figures S16–S21 and Table S3).
6
Chem 3, 1–17, September 14, 2017

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Figure 4. Catalytic Activity of Hydrogen Generation from Formic Acid Decomposition
(
2 2 2
A) Volume of the released gas (CO + H ) versus time for the dehydrogenation of 2 M FA solution over Pd/C, Pd@S-1, Ni(OH) @S-1, and Pd-
ꢁ
Ni(OH)
2
@S-1 catalysts at 60 C (nmetal/nFA = 0.012).
(
B and C) Volume of the released gas (CO
2
+ H
2
) versus time (B) and corresponding initial TOF values (C) of H
2
generation for the decomposition of 2 M
FA solution at different temperatures over 0.8Pd0.2Ni(OH)
2
@S-1 catalyst. Inset of (B): Arrhenius plot (ln TOF versus 1/T).
Shape-Selective Catalytic Reaction
As an important high-value-added reaction, the shape-selective hydrogenation of
nitrobenzene and 3-nitrotoluene with different molecular sizes was selected as the
2
mode reaction over the 0.8Pd0.2Ni(OH) @S-1 and Pd/C catalysts. The commercial
Pd/C catalyst showed complete conversion for the hydrogenation of both nitroben-
zene and 3-nitrotoluene, independently of the molecule size of the substrate, which
was confirmed by the UV/visible (UV/vis) absorption and gas chromatography-mass
2
spectrometry (GC-MS) analyses in Figure S22. In the case of the 0.8Pd0.2Ni(OH) @
S-1 catalyst, however, restricted access of the hydrogenation of 3-nitrotoluene into
the micropores of the S-1 zeolite allowed only a negligible amount of conversion,
despite complete conversion of nitrobenzene. These results demonstrate the excel-
lent shape-selective catalytic activity of the zeolite-confined nanocatalysts.
Catalytic Activity of Hydrogen Generation from Formic Acid Decomposition
Catalytic activity of FA decomposition for hydrogen generation without any additive
ꢁ
ꢁ
at 60 C and 25 C was investigated for the hybrid bimetallic Pd-Ni(OH)
lysts compared with Pd@S-1 catalyst, Ni(OH) @S-1, and the commercial Pd/C cata-
lyst (Figures 4A and S23). Compared with the Pd/C and monometallic Pd@S-1
catalysts, the Pd-Ni(OH) @S-1 catalysts exhibited considerably improved activity
for the complete decomposition of FA. Notably, the 0.8Pd0.2Ni(OH) @S-1 catalyst
2
@S-1 cata-
2
2
2
showed the highest catalytic activity; 147 mL of CO-free (<10 ppm) gas could be
ꢁ
released within 2.67 min at 60 C without any additive, reflecting the complete and
2 2
efficient decomposition of FA into H and CO . The formation of released gases
was determined by GC analysis (Figures S24 and S25). In general, the Pd clusters
with smaller particle size are prone to deactivation because of the adsorption
4
1
of poisonous CO intermediates. No detectable amount of CO (<10 ppm) was
found in the gas mixture generated from the decomposition of FA over the
0
.8Pd0.2Ni(OH)
lyst can retain excellent catalytic activity and avoid poisoning in the FA decom-
position reaction. Significantly, the 0.8Pd0.2Ni(OH) @S-1 catalyst afforded an
2 2
@S-1 catalyst, which ensures that the 0.8Pd0.2Ni(OH) @S-1 cata-
2
ꢀ
1
ꢀ1
exceedingly high initial and total TOF values up to 5,803 hr and 1,879 hr at
ꢁ
6
0 C, which are more than a 14-fold and 9-fold improvement compared with the
commercial Pd/C catalyst and monometallic Pd@S-1 catalyst, respectively. To the
best of our knowledge, such TOF values represent the highest activity for heteroge-
neously catalyzed FA decomposition without additives and under similar conditions
4
2–45
and even exceed most results obtained with additives (Table S4).
The initial
Chem 3, 1–17, September 14, 2017
7

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
H
H
2
2
generation rate of the 0.8Pd0.2Ni(OH)
2
@S-1 catalyst was calculated to be 1,418 L
ꢀ1
metalꢀ1
ꢀ1
hr
metalꢀ
g
, corresponding to a theoretical power density of 1,915 W hr
1
g
for energy generation. Assuming an operation efficiency of 60%, 1.0 g of
the 0.8Pd0.2Ni(OH) @S-1 catalyst would be adequate to supply H for 4–14 small
0.5–2.0 Wh) proton exchange membrane fuel cell devices.
2
2
4
6
(
It was observed that the monometallic Ni(OH)
lytic activity, and the physical mixture of Pd@S-1 and Ni(OH)
hibited activity identical to that of the monometallic Pd@S-1 catalyst (Figure S26).
These results indicate that the bimetallic synergistic effect of Pd and Ni(OH) clusters
in zeolite plays an important role in improving the catalytic activity for the generation
of H . Previous work has revealed that a more electron-rich Pd surface favors rate-
determining C–H activation of the Pd-formate intermediate to produce H and
@S-1 catalyst as sug-
2
@S-1 catalyst itself showed no cata-
2
@S-1 catalysts also ex-
2
2
2
1
6
CO
2
.
The electron-rich Pd surface of the 0.8Pd0.2Ni(OH)
2
gested by the EXAFS studies could be beneficial for enhancing its catalytic activity
for FA decomposition.
On the other hand, the Pd@S-1/Ni(OH)
Ni(NO solution into Pd@S-1 catalysts exhibited improved H
tivity relative to Pd@S-1, albeit still much lower than the in situ-synthesized
.8Pd0.2Ni(OH) @S-1 catalyst (Figure S27). The hydrogen generation rates over cat-
alysts synthesized using in situ-synthesized Pd or Pd-Ni(OH) clusters encapsulated
2
-im catalysts synthesized by impregnating
3
)
2
2
generation ac-
0
2
2
within S-1 zeolites were much higher than those of their counterparts synthesized us-
ing the impregnation method (Figure S28). The above results indicate that the ultra-
small Pd-Ni(OH)
generation of H
2
clusters size is also crucial for its superior catalytic activity for the
from FA.
2
The H
perature (Figures 4B, 4C, and S29). The apparent activation energy (E
.8Pd0.2Ni(OH) @S-1 and Pd@S-1 catalysts was 52.4 kJ/mol and 58.9 kJ/mol,
respectively; these values are lower than most of the reported values (Table
2
generation rates were further enhanced with increase in the reaction tem-
a
) of
0
2
3
6,47
S5).
a 2
Moreover, the decreased E value of 0.8Pd0.2Ni(OH) @S-1 compared
with Pd@S-1 further helps to explain the higher catalytic activity of the hybrid bime-
tallic catalyst compared with the monometallic catalyst.
The recycling experiment with the 0.8Pd0.2Ni(OH)
nation showed that the catalytic activity after 10 cycles was identical to that over the
fresh catalyst (Figure S30). Significantly, the 0.8Pd0.2Ni(OH) @S-1 catalyst showed
no decay in activity after being stored under atmospheric conditions for 300 days
Figure S31). Moreover, the morphology, crystallinity, and the metal cluster size of
recycled 0.8Pd0.2Ni(OH) @S-1 catalysts were almost the same as those of the fresh
2
@S-1 catalyst for FA dehydroge-
2
(
2
catalyst, demonstrating the superior recyclability of the 0.8Pd0.2Ni(OH)
lyst during FA decomposition (Figures S32–S35).
2
@S-1 cata-
The base and/or acid sites of zeolite can affect the catalytic activity for FA dehydro-
genation. In our previous work, we found that the additionally introduced basic
sites of zeolite that are beneficial for the cleavage of the O–H bonds in FA molecules
2
9
can enhance the catalytic performance of FA dehydrogenation. To further investi-
gate the influence of acid sites of zeolite on FA decomposition, Al atoms were
introduced into the zeolite framework, and the resultant aluminosilicate zeolites
(
ZSM-5, MFI), denoted as 0.8Pd0.2Ni(OH)
2
@ZSM-5 catalysts (Si/Al = 200 and
1
6
00) with different acidities, were used for the reaction of FA decomposition at
ꢁ
0 C and compared with the 0.8Pd0.2Ni(OH)
2
@S-1 catalyst. Measurement of
8
Chem 3, 1–17, September 14, 2017

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
temperature-programmed desorption of ammonia (NH
with the increase in Al content, the acidity of zeolite was increased (Figure S36).
Compared with the 0.8Pd0.2Ni(OH) @S-1 catalyst, the catalytic activity of the
.8Pd0.2Ni(OH) @ZSM-5 catalysts was remarkably reduced (Figure S37). Notably,
the 0.8Pd0.2Ni(OH) @ZSM-5 (Si/Al = 100) catalyst with the highest acidity exhibited
the worst catalytic performance among the three catalysts. Only 80 mL of gas (about
5% conversion of FA) could be finally released in 20 min, affording a lower initial
3
-TPD) demonstrated that
2
0
2
2
5
ꢀ1
2
TOF of 1,067 hr , which was just about one fifth that of the 0.8Pd0.2Ni(OH) @S-1
ꢀ1
catalyst (TOF = 5,803 hr ). The catalytic results reveal that the acidity of zeolite is
not favorable for the dehydrogenation of FA.
Density Functional Theory Calculations of FA Decomposition
Because of the structural complexity of the bulk structure of this material, it is difficult
2
to obtain structural information on the small Pd or Pd-Ni(OH) clusters encapsulated
within S-1 zeolites by experimental characterization. Theoretical studies based on
first-principle calculations may offer an alternative to explore the metal-support
interaction at the atomic level and provide some electronic information that cannot
be easily obtained in experiments. Because the close-packed (111) facet was found
to be the dominant facet in real supported palladium catalysts, in this work we chose
the Pd(111) surface as the surface model of the Pd@S-1 catalyst. For the transition
metal atom/clusters deposited on oxide/hydroxide supports, the model of a hetero-
geneous catalyst is adopted by a supported metal cluster on the oxide/hydroxide
surfaces, which has been generally accepted for understanding the metal-support
4
8–51
interaction in many theoretical studies.
was modeled by a two-layer Ni(OH) (001) slab with a (5 3 5) unit cell and a three-
layer Pd cluster containing 19 Pd atoms placed on this slab (Figure S38).
Here, a periodic surface of Pd-Ni(OH)
2
2
To account for the effect of the Pd-Ni(OH)
formance, the mechanism of FA decomposition on both Pd(111) and Pd-Ni(OH)
faces was studied by first-principle density functional theory (DFT) calculations, and
the calculated potential energy profiles on both Pd(111) and Pd-Ni(OH) surfaces are
2
interface for the improved catalytic per-
2
sur-
2
presented in Figure 5. More computational details are given in Supplemental Infor-
mation. Based on the calculated results, FA adsorbs on the top site on Pd(111) with
an adsorption energy of ꢀ55.0 kJ/mol, while it prefers to bind at the interface of the
Pd19-Ni(OH)
2
surface, with an adsorption energy of ꢀ77.2 kJ/mol. Seen from Fig-
ure 5A, FA decomposition proceeds via a three-step process on Pd(111) with a
rate-determining energy barrier of 107.2 kJ/mol, which corresponds well with the re-
5
2–54
sults of previous reports.
molecule undergoes two successive dehydrogenation steps (Figure 5B) to form
CO and dissociatively adsorbed hydrogen. It is seen that the first dehydrogenation
2
While on the Pd-Ni(OH) surface, the adsorbed FA
2
step to form HCOO is the rate-determining step for the FA decomposition process
on the bimetallic surface. The rate-determining energy barrier of 58.9 kJ/mol is much
lower than that on the Pd(111) surface. The theoretical calculations show that the FA
decomposition is facilely achieved at the interface of the Pd-Ni(OH)
thus the bimetal Pd and Ni(OH) catalyst exhibits enhanced catalytic activation for
FA decomposition compared with the monometallic Pd catalyst.
2
surface, and
2
Pd-Co(OH)
2
@S-1 Catalysts for FA Decomposition
To further verify the active interfacial effect between Pd and transition metal hydrox-
ide clusters, the Pd-Co(OH)
synthesized in situ by adding [Co(NH
synthesis system. The resulting Pd-Co(OH)
loadings (0.64 wt %), bimetallic cluster size (Figure S39), and crystallinity (Figure S40)
2
@S-1 catalysts (Pd/Co(OH)
2
+
= 9/1 and 8/2) were also
2
2
2 2 2 3
CH CH NH ) ]
complex into the starting
2
@S-1 catalysts showed similar metal
Chem 3, 1–17, September 14, 2017
9

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Figure 5. Potential Energy Diagrams and Optimized Geometries for All Reaction Intermediates and Transition States for FA Decomposition on the Pd
and Pd-Ni(OH)
2
Surfaces
A) Pure Pd(111) surface. TS, transition state.
B) Pd-Ni(OH) surface. Energies (kJ/mol) include zero-point-energy correction. Color scheme: white, H; red, O; gray, C; cyan, Pd; blue, Ni.
(
(
2
compared with the Pd-Ni(OH)
of zero-valent Pd metal and divalent Co(OH)
the Pd-Co(OH) @S-1 catalysts also exhibited a remarkably improved FA decompo-
sition rate relative to Pd@S-1 and were comparable with the Pd-Ni(OH) @S-1 cata-
2
@S-1 catalysts. XPS spectra showed the typical peaks
2
phases (Figure S41). Most significantly,
2
2
lyst (Figure S42). These results further indicate that the synergistic effect between
the interfaces of ultrasmall Pd and transition metal hydroxide clusters is responsible
for enhancing the catalytic activity of these hybrid bimetallic nanocatalysts.
Conclusion
We have demonstrated a facile and one-pot synthesis strategy to prepare subnano-
metric non-noble metal hydroxide-containing hybrid bimetallic clusters encapsu-
lated within nanosized purely siliceous silicalite-1 zeolite using metallic ethylenedi-
amine complex mixtures as precursors via a hydrothermal synthesis method. The
hybrid bimetallic nanocatalysts exhibit excellent shape-selective catalytic perfor-
mance and highly efficient CO-free generation of H
2
via complete FA decomposition
@S-1
without any additives under mild conditions. Significantly, the 0.8Pd0.2Ni(OH)
2
ꢀ1
catalyst affords an exceedingly high initial and total TOF values up to 5,803 hr and
ꢀ1
ꢁ
1
,879 hr at 60 C, which are among the highest values ever reported under similar
conditions. EXAFS studies indicate that the Pd surface of the 0.8Pd0.2Ni(OH) @S-1
2
catalyst is electron-rich compared with the Pd@S-1 catalyst, which is beneficial for
improving the catalytic activity of hybrid bimetallic catalysts. On the other hand,
DFT calculations demonstrate that the interfacial effect between ultrasmall Pd and
transition metal hydroxide clusters can considerably decrease the activation barrier
for FA decomposition, thereby enhancing catalytic activity. The as-synthesized
catalysts exhibit excellent stability after calcination in various industrial oxidation-
ꢁ
ꢁ
reduction conditions at 600 C–700 C and recycling stability during FA decomposi-
tion because the bimetallic hybrid clusters are well confined within the zeolite matrix.
The superb catalytic performance and excellent thermal and recycling stabilities of
these hybrid bimetal nanocatalysts confined in zeolites not only promise their prac-
tical application for fuel cells using FA as an efficient medium for hydrogen storage
10
Chem 3, 1–17, September 14, 2017

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
but also raise more possibilities for their use in other important high-performance
catalytic reactions.
EXPERIMENTAL PROCEDURES
Chemicals and Materials
All chemicals and materials were obtained from commercial sources and used
without further purification unless otherwise specified. Tetrapropylammonium hy-
droxide solution (TPAOH, 25 wt %), cobalt nitrate (Co(NO
zene were purchased from Sinopharm Chemical Reagent Co. Tetraethylorthosilicate
TEOS), nickel nitrate (Ni(NO $6H O), ethylenediamine (NH CH CH NH ), sodium
borohydride (NaBH ), and aniline were purchased from Tianjin Fuchen Chemical Re-
agents Factory. Aluminum sulfate (Al (SO $18H O) was purchased from Guangfu
Chemical Reagent Co. Palladium chloride (PdCl ) was purchased from Shanghai
Chemical Reagent Co. Urea (CO(NH ) was purchased from Xilong Chemical Co.
-Nitrotoluene was purchased from Chengdu Kelong Chemical Reagent Factory.
-Toluidine was purchased from Shanghai Macklin Biochemical Co. Dichlorome-
Cl ) was purchased from Biaoshiqi Science and Technology Development
Co. FA was purchased from Aladdin. Pd/C catalyst was obtained from a commercial
3 2 2
) $6H O), and nitroben-
(
3
)
2
2
2
2
2
2
4
2
4
)
3
2
2
2 2
)
3
3
thane (CH
2
2
2
Pd/C catalyst (loading 5 wt % Pd; Aladdin) followed by reduction in flowing H with
ꢁ
linear heating to 200 C for 2 hr and holding for 2 hr.
Synthesis of [Pd(NH
Co(NH CH CH NH
The solution of [Pd(NH
of PdCl in 10 mL of aqueous solution containing 2 mL of ethylenediamine
under stirring at room temperature until complete dissolution. The solution of
Ni(NH CH CH NH ](NO or [Co(NH CH CH NH ](NO was prepared by
mixing 0.95 g of Ni(NO $6H O or Co(NO $6H O into 10 mL of aqueous solution
2
CH
](NO
CH
2
CH
2
NH
Solutions
CH NH ]Cl
2 2 2 2 2 2 2 3 3 2
) ]Cl , [Ni(NH CH CH NH ) ](NO ) , and
[
2
2
2
2
)
3
3 2
)
2
2
2
2
)
2
2
was prepared by mixing 0.32 g
2
[
2
2
2
2
)
3
3
)
2
2
2
2
2
)
3
3 2
)
3
)
2
2
3
)
2
2
containing 2 mL of ethylenediamine under stirring at room temperature until com-
plete dissolution.
Synthesis of Pd@S-1 and Ni(OH)
2
@S-1 Catalysts
Pd@S-1 catalyst was synthesized with a molar composition of TEOS:TPAOH:
H
2
2 2 2 2 2 2
O:[Pd(NH CH CH NH ) ]Cl of 1:0.4:35:0.0045 under hydrothermal conditions
ꢁ
29
at 170 C for 4 days as described in our previous work. The as-synthesized solid
ꢁ
product was calcined in air at 550 C for 8 hr to remove templates, followed by reduc-
ꢁ
tion in a flowing H
Similarly, the Ni(OH)
procedure as the Pd@S-1 catalyst, except a suitable amount of [Ni(NH
2
atmosphere (linear heating to 400 C for 2 hr and holding for 2 hr).
2
@S-1 catalyst was synthesized with a similar synthesis
CH
solution.
2
2
CH
2
NH
2
)
3
](NO
3
)
2
2 2 2 2 2 2
solution was added instead of [Pd(NH CH CH NH ) ]Cl
ꢁ
After calcination in air at 550 C for 8 hr, the catalyst was reduced in flowing H
2
with
ꢁ
ꢁ
linear heating to 400 C for 2 hr and holding for 2 hr, followed by heating to 500 C for
3
0 min and holding for 1 hr. The molar quantities of Pd and Ni were kept the same
for the Pd@S-1 and Ni(OH) @S-1 catalysts.
2
Synthesis of xPd(1 ꢀ x)Ni(OH)
Co(OH) @S-1 (y = 0.9 and 0.8) Catalysts
xPd(1 ꢀ x)Ni(OH) @S-1 catalysts were synthesized in gels with molar compositions of
TEOS:TPAOH:H
a:b (a:b = 9:1, 8:2, or 7:3) under similar conditions as for the synthesis of the
Pd@S-1 catalyst, except that suitable amounts of both [Pd(NH CH CH NH ]Cl
and [Ni(NH CH CH NH ](NO solutions (0.942 mL and 0.058 mL, 0.88 mL and
.12 mL, 0.81 mL and 0.19 mL) were added to the reaction mixtures, resulting in
2
@S-1 (x = 0.9, 0.8, and 0.7) and yPd(1 ꢀ y)
2
2
2
+
2+
2
O:[Pd(NH
2
CH
2
CH
2
NH
2
)
2
]
: [Ni(NH
2
CH
2
CH
2
NH
2
)
3
]
of 1:0.4:35:
2
2
2
2
)
2
2
2
2
2
2
)
3
3 2
)
0
Chem 3, 1–17, September 14, 2017
11

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
molar ratios nPd/(nPd + nNi) equal to 0.9, 0.8, and 0.7, respectively. The metal content
ꢁ
of ðm + mNiÞ=mSiO2 was kept at a constant 0.008. After calcination in air at 550 C
Pd
ꢁ
for 8 hr, the catalysts were reduced in flowing H
2
with linear heating to 400 C for
ꢁ
2
1
hr and holding for 2 hr, followed by heating to 500 C for 30 min and holding for
hr. Similarly, yPd(1 ꢀ y)Co(OH)
2
@S-1 catalysts were also synthesized with synthetic
procedures similar to that for the Pd@S-1 catalyst, except suitable amounts of
Co(NH CH CH NH ](NO and [Pd(NH CH CH NH ]Cl solutions were added
to the synthesis gels before adding TEOS.
[
2
2
2
2
)
3
3
)
2
2
2
2
2
)
2
2
Synthesis of 0.8Pd0.2Ni(OH)
The synthetic process for the 0.8Pd0.2Ni(OH)
for the 0.8Pd0.2Ni(OH) @S-1 catalyst, except for the introduction of some of
Al (SO into the synthesis gel, and the molar ratios of Si/Al were 200 and 100,
2
@ZSM-5 Catalysts
2
@ZSM-5 catalysts was similar to that
2
2
4 3
)
respectively.
2
Preparation of Pd@S-1/xNi(OH) -im Catalysts
Pd@S-1/xNi(OH)
into Pd@S-1 samples. Typically, 0.7 g of reduced Pd@S-1 catalyst was impregnated
with 0.17 mL of Ni(NO $6H O solution (0.028 M, 0.062 M, and 0.106 M), affording
2 3 2
-im catalysts were prepared by impregnating Ni(NO ) solution
3
)
2
2
molar ratios of nPd/(nPd + nNi) of 0.9, 0.8, and 0.7, respectively. The resulting solids
ꢁ
were dried at 80 C in the oven overnight. Finally, the catalysts were reduced in flow-
ꢁ
ing H
2
with linear heating to 400 C for 2 hr and holding for 2 hr, and then heating to
ꢁ
5
00 C for 30 min and holding for 1 hr.
Preparation of Pd/S-1-im and 0.8Pd0.2Ni(OH)
2
/S-1-im Catalysts
The silicalite-1 zeolite support was synthesized with a molar composition of TEOS:
2
9
TPAOH:H
.8Pd0.2Ni(OH)
nation method with (NH
2
O of 1:0.4:35 as described in our previous work. The Pd/S-1-im and
/S-1-im catalysts were prepared by the incipient wetness impreg-
PdCl and Ni(NO solution. To prepare the Pd/S-1-im
0
2
4
)
2
4
3 2
)
catalyst, 1 g of calcined silicalite-1 zeolite was impregnated with a (NH
4
)
2
PdCl
following
/S-1-im catalyst was
prepared by the same procedure as for the Pd/S-1-im catalyst except that the zeolite
was impregnated with 0.23 mL of mixed solution containing (NH PdCl (0.224 M)
/S-1-im catalyst was reduced in flow-
4
so-
ꢁ
lution. The solid obtained was dried at 80 C and then treated in flowing H
2
the same procedure as used for Pd@S-1. The 0.8Pd0.2Ni(OH)
2
4
)
2
4
and Ni(NO
ing H
3
)
2
(0.056 M). The 0.8Pd0.2Ni(OH)
2
ꢁ
2
with linear heating to 400 C for 2 hr and holding for 2 hr, and then heating
ꢁ
to 500 C for 30 min and holding for 1 hr.
Synthesis of a-Ni(OH) Sample
2
The a-Ni(OH)
CO(NH :2228 H
Typically, 0.291 g of Ni(NO
2
3 2
sample was synthesized with a molar composition of 1.0 Ni(NO ) :2.0
ꢁ
2
)
2
2
O under conventional hydrothermal conditions at 100 C for 20 hr.
3
)
2
$6H
2
O was dissolved in 40 mL of deionized water un-
der continuous stirring at room temperature. Urea (0.12 g) was then added into the
above mixture, followed by continuous stirring for 5 min. Finally, the solution was
then transferred into a 100 mL Teflon-lined stainless steel autoclave, and the crystal-
ꢁ
lization was conducted in a conventional oven at 100 C for 20 hr under static condi-
tions. The resultant solid product was centrifuged, washed with water and ethanol
ꢁ
several times, and then dried at 60 C in the oven overnight. The phase of this
2
a-Ni(OH) was identified using powder X-ray diffraction (PXRD) (Figure S43).
Characterizations
PXRD patterns were recorded on a Rigaku D/Max 2550 diffractometer at 50 kV and
2
00 mA (Cu Ka radiation). The SEM images were measured with a Hitachi SU-8020
12
Chem 3, 1–17, September 14, 2017

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
scanning electron microscope. The TEM images and corresponding EDX spectra were
acquired using an FEI Tecnai G2 F20 electron microscope. Cs-corrected HAADF-
STEM images were taken at 200 kV on a JEM-ARM200F scanning transmission electron
microscope. The metal content in the zeolite samples was analyzed by a PerkinElmer
2
9
Optima 3300 DV ICP-AES instrument. Si MAS NMR measurements were made on a
Varian Infinity Plus 400WB spectrometer with a resonance frequency of 79.5 MHz. N
adsorption/desorption measurements were performed on a Micromeritics ASAP 2020
analyzer at 77.35 K. Before starting the N adsorption measurements, all the samples
2
2
were activated by degassing in situ at about 623 K for 10 hr. XPS was conducted on a
Thermo ESCALAB 250 spectrometer on separated metal species after dissolving
Pd@S-1, 0.8Pd0.2Ni(OH)
NaOH solution. The NH
AutoChem II 2920 chemisorption analyzer. The XAFS spectra measurements at the Ni
K-edge (E = 8333 eV) and Pd K-edge (E = 24,350 eV) were performed at beamline
BL14W1 of the Shanghai Synchrotron Radiation Facility (SSRF) operated at 3.5 GeV un-
2 2 2
@S-1, 0.8Pd0.2Co(OH) @S-1, and Ni(OH) @S-1 samples in
3
-TPD experiments were performed using a Micromeritics
0
0
5
5
der‘‘top-up’’ mode with a constant currentof240mA. TheXAFSdata were recordedin
transmission mode using high-flux ion chambers for the Ni K-edge and under fluores-
cence mode with standard Lytle ion chambers for the Pd K-edge. The energy was cali-
brated according to the absorption edge of a pure Ni or Pd foil, as appropriate. Athena
and Artemis software were used to extract and fit the data. For the XANES part, the
experimental absorption coefficients as a function of energy m(E) were processed by
background subtraction and normalization procedures, and reported as ‘‘normalized
absorption.’’ Based on the normalized XANES profiles, the oxidation state of nickel
and palladium can be determined by comparison with the corresponding standards
2
+
0
2+
0
(
NiO/Ni foil for Ni /Ni and PdO/Pd foil for Pd /Pd ). For the EXAFS data, the Four-
ier-transformed data in R space were analyzed by applying a first shell approximation
for the Ni–O/Pd–O and a metallic Pd model for the Pd–Pd shells. The passive electron
2
0
factors, S
, were determined by fitting the experimental Ni or Pd foil data and fixing
theNi–NiorPd–Pdcoordinationnumber(CN)tobe12,andthenfixedforfurtheranalysis
of the measured samples. The parameters describing the electronic properties (e.g.,
0
correction to the photoelectron energy origin, E ) and local structure environment,
including CN, bond distance (R), and Debye-Waller factor, around the absorbing atoms
3
were allowed to vary during the fitting process. The k weighted fitted ranges for the
ꢀ1
˚
k space were 3–12 (Ni) and 3–11 (Pd) A , and the fitting range for R spaces was selected
˚
to be 1–4 A.
Thermal Stability Tests of Catalysts
The representative 0.8Pd0.2Ni(OH)
2
@S-1 catalyst (0.1 g) was calcined in a
ꢁ
ꢁ
4
0 mL/min N
2
atmosphere at 700 C (linear heating to 700 C for 2 hr and holding
for 2 hr). For comparison, the Pd@S-1/0.2Ni(OH)
and 0.8Pd0.2Ni(OH) /S-1-im catalysts synthesized using the impregnation method
were also treated under the same conditions. The 0.8Pd0.2Ni(OH)
2 2
-im, Pd@S-1/0.3Ni(OH) -im,
2
2
@S-1 catalysts
ꢁ
were also calcined in a flowing O
2
and H
2
atmosphere (40 mL/min) at 500 C–
ꢁ
7
00 C. The high-temperature oxidation-reduction cyclic treatment was performed
2
on the Pd@S-1 and 0.8Pd0.2Ni(OH) @S-1 catalysts. Typically, the catalyst was
ꢁ
first subjected to an oxidation treatment at 650 C in air with linear heating
ꢁ
to 650 C for 2 hr and holding for 2 hr. After the oxidation treatment, the
ꢁ
ꢁ
sample was then reduced at 650 C in H with linear heating to 650 C for 2 hr
2
and holding for 2 hr. The above reduction and oxidation treatment was considered
as one cycle. Subsequently, the catalyst was further applied in three consecutive
oxidation-reduction cycles. The 0.8Pd0.2Ni(OH)
2
@S-1 catalyst was also treated at
00 C and 600 C for 5 hr with water vapor flow of 3 mL/min and 18 mL/min,
respectively.
ꢁ
ꢁ
5
Chem 3, 1–17, September 14, 2017
13

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
Reduction Reaction of Nitrobenzene and 3-Nitrotoluene
First, a mixture of 0.1 mL of NaBH
or 3-nitrotoluene substrate solution (1.46 mM) was added into a 25 mL round-
bottomed flask. Next, 20.1 mg of 0.8Pd0.2Ni(OH) @S-1 catalyst or 2.2 mg of
4
solution (438 mM) and 10 mL of nitrobenzene
2
Pd/C catalyst with the same Pd content was added to the above mixture, and the
molar ratio of nPd/nsubstrate was 0.07. The reduction reaction was performed at
ꢁ
2
5 C under continuous stirring for 5 min. After the reaction, the reactants and prod-
ucts were analyzed by UV/vis measurements. After extraction with dichloromethane,
the reactants and products were further analyzed using a Thermo Fisher Trace ISQ
GC-MS equipped with a TG-5MS column (60 m 3 320 mm 3 25 mm).
Catalytic Activity of Hydrogen Generation from Formic Acid Decomposition
The decomposition reaction of FA was carried out using an apparatus containing a
reaction unit and a gas collecting device. Typically, a mixture of a suitable amount of
catalyst and distilled water (1 mL) was first added in a 25 mL two-necked flask placed
ꢁ
ꢁ
in a water bath with magnetic stirring (600 rpm) at temperatures of 25 C–60 C under
an ambient atmosphere. The volume of released gas was measured using a gas
burette and an electronic balance continuously recording the weight of excurrent
water. FA (0.5 mL, 6.0 M) aqueous solution was injected into the flask using a syringe
to initiate the reaction. The molar ratios of nmetal/nFA were fixed at 0.012 for all the
catalytic reactions. The collected gas products were analyzed using an Agilent GC
6
890N instrument equipped with a thermal conductivity detector and Plot-Q column
(
Agilent J&W GC Columns, HP-PLOT/Q 19095P-Q04, 30 m 3 530 mm 3 40 mm).
Further analyses for CO were performed on a GC-7900 gas chromatograph using
a flame ionization detector (FID)-methanator (detection limit, ꢂ10 ppm).
Durability Test of Catalysts for FA Decomposition
The catalyst isolated after the initial FA decomposition experiment was reused in the
ꢁ
catalytic dehydrogenation of FA (0.5 mL, 6.0 M) at 60 C. These catalysts were iso-
lated and reused for up to 10 consecutive catalytic cycles. Moreover, catalytic dehy-
drogenation of FA was also performed for the 0.8Pd0.2Ni(OH)
under atmospheric conditions for 300 days.
2
@S-1 catalyst stored
2
The Calculation of Turnover Frequency and H Generation Rate (RH2 )
The initial TOF (TOFInitial) and total TOF (TOFTotal) were calculated based on the
quantity of metal atoms (Pd + Ni) in the catalysts when the conversion of FA reached
up to 20% and 100%, respectively. The following equations were used:
ꢀ
PVInitial gas RT
TOFInitial
TOFTotal
=
=
;
(Equation 1)
(Equation 2)
2
n
Metal
t
Initial
ꢀ
PVTotal gas RT
;
2
n
Metal
t
Total
where P = 101.325 kPa; VInitial gas and VTotal gas are the initial and total volume of
released gas (H and CO ), corresponding to conversions of 20% and 100% toward
the conversion of FA, respectively; R = 8.3145 m Pa mol
2
2
3
ꢀ1 ꢀ1
K
; T = 298 K; nMetal is
the total molar number of metal atoms (Pd + Ni) in the catalyst; tInitial and tTotal are the
time of the reaction in hours, corresponding to 20% and 100% conversion toward the
decomposition of FA.
2
2
The H generation rate (RH ) was calculated from the following equation:
V
H2
R
H2
=
;
(Equation 3)
m
Metal
t
14
Chem 3, 1–17, September 14, 2017

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
where VH2 is the initial volume of released gas (H
conversion of 20% toward the conversion of FA; mMetal is the mass of metal atoms
Pd + Ni); t is the reaction time in hours, corresponding to the conversion of 20%
2 2
and CO ), corresponding to the
(
toward the conversion of FA.
DFT Computational Methods
First-principle spin-polarized DFT calculations were performed using the Vienna ab
5
6
57,58
initio Simulation Package. The projector augmented wave pseudopotentials
were used to describe the electron-ion interactions. The exchange-correlation ef-
5
9
fects were treated by the Perdew-Burke-Ernzerhof functional with the generalized
gradient approximation. The energy cutoff for the plane-wave basis sets was set at
4
00 eV in all calculations.
The surface model of Pd-Ni(OH)
tions based on optimized Pd and Ni(OH)
Ni(OH) (001) slab model with two layers was adopted as the support, and a
three-layer Pd cluster containing 19 atoms with the exposed (111) surface facet
was placed on top of the Ni(OH) surface (Figure S38). In the optimization of the
Pd-Ni(OH) slab model, the Pd cluster and the upper layer of Ni(OH) were allowed
to fully relax, and the bottom layer of Ni(OH) was fixed at the optimized bulk geom-
2
was constructed using periodic boundary condi-
2
bulk structures. A (5 3 5) supercell of
2
2
2
2
2
˚
etry. A vacuum region of 20 A was used to remove any interactions between the pe-
riodic slabs. The slab model of the Pd(111) surface was built with a (3 3 3) supercell
and five layers. The upper three layers of Pd(111) were relaxed, and the bottom two
layers were fixed at the bulk geometry in structure optimization. The Brillouin zones
were sampled with Monkhorst-Pack k points of 7 3 7 3 7 for bulk calculations and
3
3 3 3 1 for surface calculations. The geometry optimization was stopped until
˚
the force on each atom was less than 0.02 eV/A. The transition states were located
with the climbing-image nudged elastic band method and were verified by vibra-
tional analysis. The energy barriers were calculated as the energy difference
between the transition state and the reactants. The adsorption energy (Eads) for
the adsorbate was defined as
E
ads = Eadsorbate-surface – Esurface – Eadsorbate
,
(Equation 4)
where Eadsorbate, Esurface, and Eadsorbate-surface represent the energies of the adsor-
bate in gas phase, the corresponding clean surface model, and the surface with
the adsorbate, respectively.
SUPPLEMENTAL INFORMATION
Supplemental Information includes 43 figures and 5 tables and can be found with
this article online at http://dx.doi.org/10.1016/j.chempr.2017.07.001.
AUTHOR CONTRIBUTIONS
J.Y. designed and supervised the project. Q.S. and N.W. contributed equally to this
work, and carried out the synthesis, characterization, and the catalytic measurements.
Q.S. wrote the manuscript, and J.Y. revised the manuscript. Q.B. and J.L. performed
the theoretical calculations. R.S. and P.Z. performed EXAFS measurements of the sam-
ples and analyzed the results. R.B. and M.J. collaborated in characterizing the samples.
ACKNOWLEDGMENTS
The authors thank the Nation Key Research and Development Program of China
(
grant no. 2016YFB0701100), National Natural Science Foundation of China (grant
Chem 3, 1–17, September 14, 2017
15

2
Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH) , M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
nos. 21320102001, 21621001, 21373098, and 21373259) , and the National 111
Project (B17020) for supporting this work. The authors also thank Dr. Paul Duchesne
(
Department of Chemistry, Dalhousie University) for his kind help with revising the
manuscript.
Received: March 14, 2017
Revised: April 19, 2017
Accepted: July 3, 2017
Published: August 31, 2017
REFERENCES AND NOTES
1
. Gates, B.C. (1995). Supported metal clusters:
synthesis, structure, and catalysis. Chem. Rev.
iron-nickel hydroxide–platinum nanoparticles
enhance catalytic oxidation. Science 344,
495–499.
23. Choi, M., Wu, Z., and Iglesia, E. (2010).
Mercaptosilane-assisted synthesis of metal
clusters within zeolites and catalytic
consequences of encapsulation. J. Am. Chem.
Soc. 132, 9129–9137.
95, 511–522.
2
. Flytzani-Stephanopoulos, M., and Gates, B.C.
2012). Atomically dispersed supported metal
catalysts. Annu. Rev. Chem. Biomol. Eng. 3,
45–574.
13. Subbaraman, R., Tripkovic, D., Strmcnik, D.,
Chang, K.-C., Uchimura, M., Paulikas, A.P.,
Stamenkovic, V., and Markovic, N.M. (2011).
(
24. Goel, S., Wu, Z., Zones, S.I., and Iglesia, E.
(2012). Synthesis and catalytic properties of
metal clusters encapsulated within small-pore
5
Enhancing hydrogen evolution activity in water
+
splitting by tailoring Li -Ni(OH)
2
-Pt interfaces.
3
4
5
. Munnik, P., de Jongh, P.E., and de Jong, K.P.
2015). Recent developments in the synthesis of
supported catalysts. Chem. Rev. 115, 6687–
718.
Science 334, 1256–1260.
(
1
SOD, GIS, ANA) zeolites. J. Am. Chem. Soc.
34, 17688–17695.
(
14. Subbaraman, R., Tripkovic, D., Chang, K.-C.,
Strmcnik, D., Paulikas, A.P., Hirunsit, P., Chan,
M., Greeley, J., Stamenkovic, V., and Markovic,
N.M. (2012). Trends in activity for the water
electrolyser reactions on 3d M(Ni,Co,Fe,Mn)
hydr(oxy)oxide catalysts. Nat. Mater. 11,
550–557.
6
25. Goel, S., Zones, S.I., and Iglesia, E. (2014).
Encapsulation of metal clusters within MFI via
interzeolite transformations and direct
hydrothermal syntheses and catalytic
consequences of their confinement. J. Am.
Chem. Soc. 136, 15280–15290.
. Flytzani-Stephanopoulos, M. (2014). Gold
atoms stabilized on various supports catalyze
the water–gas shift reaction. Acc. Chem. Res.
47, 783–792.
. Lu, J., Fu, B., Kung, M.C., Xiao, G., Elam, J.W.,
Kung, H.H., and Stair, P.C. (2012). Coking- and
sintering-resistant palladium catalysts
achieved through atomic layer deposition.
Science 335, 1205–1208.
15. Huang, W., Wang, H., Zhou, J., Wang, J.,
Duchesne, P.N., Muir, D., Zhang, P., Han, N.,
Zhao, F., Zeng, M., et al. (2015). Highly
active and durable methanol oxidation
electrocatalyst based on the synergy of
platinum-nickel hydroxide-graphene. Nat.
Commun. 6, 10035.
2
6. Otto, T., Ramallo-L o´ pez, J.M., Giovanetti, L.J.,
Requejo, F.G., Zones, S.I., and Iglesia, E. (2016).
Synthesis of stable monodisperse AuPd, AuPt,
and PdPt bimetallic clusters encapsulated
within LTA-zeolites. J. Catal. 342, 125–137.
6
. Yang, X.-F., Wang, A., Qiao, B., Li, J., Liu, J.,
and Zhang, T. (2013). Single-atom catalysts: a
new frontier in heterogeneous catalysis. Acc.
Chem. Res. 46, 1740–1748.
27. Otto, T., Zones, S.I., and Iglesia, E. (2016).
Challenges and strategies in the encapsulation
and stabilization of monodisperse Au clusters
within zeolites. J. Catal. 339, 195–208.
16. Yan, J.-M., Wang, Z.-L., Gu, L., Li, S.-J., Wang,
H.-L., Zheng, W.-T., and Jiang, Q. (2015). AuPd-
x
MnO /MOF-graphene: an efficient catalyst for
hydrogen production from formic acid at room
temperature. Adv. Energy Mater 5, 4502–4505.
7
. Vajda, S., Pellin, M.J., Greeley, J.P., Marshall,
C.L., Curtiss, L.A., Ballentine, G.A., Elam, J.W.,
Catillon-Mucherie, S., Redfern, P.C.,
Mehmood, F., et al. (2009). Subnanometre
platinum clusters as highly active and selective
catalysts for the oxidative dehydrogenation of
propane. Nat. Mater. 8, 213–216.
2
8. de Jong, K.P., and Ze ꢀc evi ꢁc , J. (2017). Porous
catalysts: the platinum rush. Nat. Mater. 16,
7
–8.
17. Cꢀ ejka, J., Corma, A., and Zones, S.I. (2010).
Zeolites and Catalysis: Synthesis, Reactions
and Applications (Wiley-VCH Weinheim).
2
9. Wang, N., Sun, Q., Bai, R., Li, X., Guo, G., and
Yu, J. (2016). In situ confinement of ultrasmall
Pd clusters within nanosized silicalite-1 zeolite
for highly efficient catalysis of hydrogen
1
8. Corma, A. (1997). From microporous to
mesoporous molecular sieve materials and
their use in catalysis. Chem. Rev. 97, 2373–2420.
8
. Liu, L., D ı´ az, U., Arenal, R., Agostini, G.,
Concepci o´ n, P., and Corma, A. (2017).
Generation of subnanometric platinum with
high stability during transformation of a 2D
zeolite into 3D. Nat. Mater. 16, 132–138.
generation. J. Am. Chem. Soc. 138, 7484–7487.
3
3
3
3
0. Dai, C., Zhang, A., Liu, M., Gu, L., Guo, X., and
Song, C. (2016). Hollow alveolus-like
nanovesicle assembly with metal-encapsulated
hollow zeolite nanocrystals. ACS Nano 10,
7
19. Farrusseng, D., and Tuel, A. (2016).
Perspectives on zeolite-encapsulated metal
nanoparticles and their applications in
catalysis. New J. Chem. 40, 3933–3949.
9
. Liu, P., Zhao, Y., Qin, R., Mo, S., Chen, G., Gu,
L., Chevrier, D.M., Zhang, P., Guo, Q., Zang, D.,
et al. (2016). Photochemical route for
synthesizing atomically dispersed palladium
catalysts. Science 352, 797–800.
401–7408.
20. Fenwick, O., Couti n˜ o-Gonzalez, E., Grandjean,
D., Baekelant, W., Richard, F., Bonacchi, S., De
Vos, D., Lievens, P., Roeffaers, M., Hofkens, J.,
et al. (2016). Tuning the energetics and
1. Dai, C., Zhang, A., Liu, M., Guo, X., and Song,
C. (2015). Hollow ZSM-5 with silicon-rich
surface, double shells, and functionalized
interior with metallic nanoparticles and carbon
nanotubes. Adv. Funct. Mater. 25, 7479–7487.
1
1
0. Qiao, B., 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
tailoring the optical properties of silver clusters
confined in zeolites. Nat. Mater. 15, 1017–1022.
21. Yang, M., Li, S., Wang, Y., Herron, J.A., Xu, Y.,
Allard, L.F., Lee, S., Huang, J., Mavrikakis, M.,
2. Dai, C., Zhang, S., Zhang, A., Song, C., Shi, C.,
and Guo, X. (2015). Hollow zeolite
encapsulated Ni-Pt bimetals for sintering and
coking resistant dry reforming of methane.
J. Mater. Chem. A. 3, 16461–16468.
Pt
1
/FeO
x
. Nat. Chem. 3, 634–641.
1. Guo, X., Fang, G., Li, G., Ma, H., Fan, H., Yu, L.,
Ma, C., Wu, X., Deng, D., Wei, M., et al. (2014).
Direct, nonoxidative conversion of methane to
ethylene, aromatics, and hydrogen. Science
and Flytzani-Stephanopoulos, M. (2014).
Catalytically active Au-O(OH)
stabilized by alkali ions on zeolites and
mesoporous oxides. Science 346, 1498–1501.
-
x
species
3. Dai, C., Li, X., Zhang, A., Liu, C., Song, C., and
Guo, X. (2015). Pd and Pd-CuO nanoparticles in
hollow silicalite-1 single crystals for enhancing
selectivity and activity for the Suzuki-Miyaura
reaction. RSC Adv. 5, 40297–40302.
344, 616–619.
2
2. Guzman, J., and Gates, B.C. (2003). Supported
molecular catalysts: metal complexes and
clusters on oxides and zeolites. Dalton Trans.
17, 3303–3318.
1
2. Chen, G., Zhao, Y., Fu, G., Duchesne, P.N., Gu,
L., Zheng, Y., Weng, X., Chen, M., Zhang, P.,
Pao, C.-W., et al. (2014). Interfacial effects in
16
Chem 3, 1–17, September 14, 2017

Please cite this article in press as: Sun et al., Subnanometric Hybrid Pd-M(OH)
2
, M = Ni, Co, Clusters in Zeolites as Highly Efficient Nanocatalysts
for Hydrogen Generation, Chem (2017), http://dx.doi.org/10.1016/j.chempr.2017.07.001
3
4. Chen, Y., Zhu, Q.-L., Tsumori, N., and Xu, Q.
2015). Immobilizing highly catalytically active
42. Jiang, K., Xu, K., Zou, S., and Cai, W.-B. (2014).
B-doped Pd catalyst: boosting room-
between Ru cluster and anatase TiO
surface. J. Phys. Chem. C 118, 3514–3522.
2
(101)
(
noble metal nanoparticles on reduced
temperature hydrogen production from formic
graphene oxide: a non-noble metal sacrificial
approach. J. Am. Chem. Soc. 137, 106–109.
acid-formate solutions. J. Am. Chem. Soc. 136, 51. Kacprzak, K.A., Czekaj, I., and Mantzaras, J.
4861–4864.
(2012). DFT studies of oxidation routes for Pd9
clusters supported on g-alumina. Phys. Chem.
Chem. Phys. 14, 10243–10247.
4
3. Song, F.-Z., Zhu, Q.-L., Tsumori, N., and Xu, Q.
3
3
5. Dai, H., Xia, B., Wen, L., Du, C., Su, J., Luo, W.,
and Cheng, G. (2015). Synergistic catalysis of
AgPd@ZIF-8 on dehydrogenation of formic
acid. Appl. Catal. B Environ. 165, 57–62.
(2015). Diamine-alkalized reduced graphene
oxide: immobilization of sub-2 nm palladium
nanoparticles and optimization of catalytic
activity for dehydrogenation of formic acid.
ACS Catal. 5, 5141–5144.
52. Luo, Q., Feng, G., Beller, M., and Jiao, H.J.
(2012). Formic acid dehydrogenation on
Ni(111) and comparison with Pd(111) and
Pt(111). J. Phys. Chem. C 116, 4149–4156.
6. Liu, Q., Yang, X., Huang, Y., Xu, S., Su, X., Pan,
X., Xu, J., Wang, A., Liang, C., Wang, X., et al.
4
4. Zhang, S., Metin, O¨ ., Su, D., and Sun, S. (2013).
5
3. Wang, Y., Qi, Y., Zhang, D., and Liu, C. (2014).
New insight into the decomposition
mechanism of formic acid on Pd(111):
competing formation of CO2 and CO. J. Phys.
Chem. C 118, 2067–2076.
(2015). A Schiff base modified gold catalyst for
Monodisperse AgPd alloy nanoparticles
and their superior catalysis for the
dehydrogenation of formic acid. Angew.
Chem. Int. Ed. 52, 3681–3684.
green and efficient H production from formic
2
acid. Energy Environ. Sci. 8, 3204–3207.
3
7. Yan, H., Cheng, H., Yi, H., Lin, Y., Yao, T., Wang,
C., Li, J., Wei, S., and Lu, J. (2015). Single-atom
Pd1/graphene catalyst achieved by atomic
layer deposition: remarkable performance in
selective hydrogenation of 1,3-butadiene.
J. Am. Chem. Soc. 137, 10484–10487.
4
5. Bi, Q.-Y., Du, X.-L., Liu, Y.-M., Cao, Y., He, H.-Y.,
and Fan, K.-N. (2012). Efficient subnanometric
gold-catalyzed hydrogen generation via formic
acid decomposition under ambient conditions.
J. Am. Chem. Soc. 134, 8926–8933.
5
4. Cho, J., Lee, S., Han, J., Yoon, S.P., Nam, S.W.,
Choi, S.H., Lee, K.Y., and Ham, H.C. (2014).
Importance of ligand effect in selective
hydrogen formation via formic acid
decomposition on the bimetallic Pd/Ag
catalyst from first-principles. J. Phys. Chem. C
3
3
4
8. Moller, K., Koningsberger, D.C., and Bein, T.
46. Tsuchiya, H., and Kobayashi, O. (2004). Mass
production cost of PEM fuel cell by learning
curve. Int. J. Hydrogen Energy 29, 985–990.
118, 22553–22560.
(1989). Stabilization of metal ensembles at
room temperature: palladium clusters in
zeolites. J. Phys. Chem. 93, 6116–6120.
5
5
5
5. Yu, H., Wei, X., Li, J., Gu, S., Zhang, S., Wang, L.,
Ma, J., Li, L., Gao, Q., Si, R., et al. (2015). The
XAFS beamline of SSRF. Nucl. Sci. Tech. 26,
4
7. Zhou, X., Huang, Y., Xing, W., Liu, C., Liao, J.,
and Lu, T. (2008). High-quality hydrogen from
the catalyzed decomposition of formic acid by
Pd-Au/C and Pd-Ag/C. Chem. Commun. 3540–
9. Xu, L., Zhang, Z., and Sachtler, W.M.H. (1992).
Effect of zeolite protons on Pd particle size and
H chemisorption. J. Chem. Soc. Faraday
2
050102.
6. Kresse, G., and Furthm u¨ lle, J. (1996). Efficient
iterative schemes for ab initio total-energy
calculations using a plane-wave basis set. Phys.
Rev. B 54, 11169–11186.
3542.
Trans. 88, 2291–2295.
4
4
8. Shi, X.R., and Sholl, D.S. (2012). Nucleation of
Rh surfaces: a
0. Yan, J., Fan, Z., Sun, W., Ning, G., Wei, T.,
Zhang, Q., Zhang, R., Zhi, L., and Wei, F. (2012).
Advanced asymmetric supercapacitors based
on Ni(OH) /graphene and porous graphene
2
electrodes with high energy density. Adv.
n
(n = 1ꢀ5) clusters on g-Al
2 3
O
density functional theory study. J. Phys. Chem.
7. Blochl, P.E. (1994). Projector augmented-wave
method. Phys. Rev. B 50, 17953–17979.
C 116, 10623–10631.
9. Li, J., Croiset, E., and Sandoval, L.R. (2013).
Funct. Mater. 22, 2632–2641.
58. Kresse, G., and Joubert, D. (1999). From
ultrasoft pseudopotentials to the projector
augmented-wave method. Phys. Rev. B 59,
1758–1775.
Effect of metalꢀsupport interface during CH
and H dissociation on Ni/g-Al : a density
functional theory study. J. Phys. Chem. C 117,
6907–16920.
4
2
2 3
O
4
1. Tedsree, K., Li, T., Jones, S., Chan, C.W.A., Yu,
K.M.K., Bagot, P.A.J., Marquis, E.A., Smith,
G.D.W., and Tsang, S.C.E. (2011). Hydrogen
production from formic acid decomposition at
room temperature using a Ag-Pd core-shell
nanocatalyst. Nat. Nanotech. 6, 302–307.
1
50. Zhang, S.T., Li, C.M., Yan, H., Wei, M., Evans,
D.G., and Duan, X. (2014). Density functional
theory study on the metalꢀsupport interaction
59. Perdew, J.P., Burke, K., and Ernzerhof, M.
(1996). Generalized gradient approximation
made simple. Phys. Rev. Lett. 77, 3865–3868.
Chem 3, 1–17, September 14, 2017
17