Encapsulation of bimetallic metal nanoparticles into robust zirconium-based metal–organic frameworks: Evaluation of the catalytic potential for size-selective hydrogenation

Article abstract of DOI:10.1002/chem.201603984

The realization of metal nanoparticles (NPs) with bimetallic character and distinct composition for specific catalytic applications is an intensively studied field. Due to the synergy between metals, most bimetallic particles exhibit unique properties that are hardly provided by the individual monometallic counterparts. However, as small-sized NPs possess high surface energy, agglomeration during catalytic reactions is favored. Sufficient stabilization can be achieved by confinement of NPs in porous support materials. In this sense, metal–organic frameworks (MOFs) in particular have gained a lot of attention during the last years; however, encapsulation of bimetallic species remains challenging. Herein, the exclusive embedding of preformed core–shell PdPt and RuPt NPs into chemically robust Zr-based MOFs is presented. Microstructural characterization manifests partial retention of the core–shell systems after successful encapsulation without harming the crystallinity of the microporous support. The resulting chemically robust NP@UiO-66 materials exhibit enhanced catalytic activity towards the liquidphase hydrogenation of nitrobenzene, competitive with commercially used Pt on activated carbon, but with superior size-selectivity for sterically varied substrates.

Full text of DOI:10.1002/chem.201603984

A Journal of
Accepted Article
Title: Encapsulation of bimetallic metal nanoparticles into robust Zr-
based metal-organic frameworks: Evaluation of the catalytic
potential for size-selective hydrogenation
Authors: Christoph Rösler, Stefano Dissegna, Victor Lopez Rechac,
Max Kauer, Penghu Guo, Stuart Turner, Kevin Ollegott,
Hirokazu Kobayashi, Tomokazu Yamamoto, Daniel Peeters,
Yuemin Wang, Syo Matsumura, Gustaaf Van Tendeloo,
Hiroshi Kitagawa, Martin Muhler, Francesc X. Llabrés i
Xamena, and Roland A. Fischer
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201603984
Link to VoR: http://dx.doi.org/10.1002/chem.201603984
Supported by

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
Encapsulation of bimetallic metal nanoparticles into robust Zr-
based metal-organic frameworks: Evaluation of the catalytic
potential for size-selective hydrogenation
Christoph Rösler,[a] Stefano Dissegna,[b] Victor L. Rechac,[c] Max Kauer,[a] Penghu Guo,[f] Stuart
Turner,[e] Kevin Ollegott,[f] Hirokazu Kobayashi,[g] Tomokazu Yamamoto,[h] Daniel Peeters,[a] Yuemin
Wang,[d] Syo Matsumura,[h] Gustaaf Van Tendeloo,[e] Hiroshi Kitagawa,[g] Martin Muhler,[f] Francesc X.
Llabrés i Xamena*[c] and Roland A. Fischer*[b]
Abstract: The realization of metal NPs with bimetallic character and
distinct composition for specific catalytic applications is an intensively
studied field. Due to the synergy between metals, most of the
bimetallic particles exhibit unique properties, only hardly provided by
the individual monometallic counterparts. However, as small sized
NPs possess high surface energy, agglomeration during catalytic
reactions is favored. Sufficient stabilization can be achieved by
confinement of NPs in porous support materials. In this sense,
especially MOFs gained a lot of attention during the last years,
however, encapsulation of bimetallic species remains challenging.
Herein, the exclusive embedding of preformed core/shell PdPt and
RuPt NPs into chemically robust Zr-based MOFs is presented.
Microstructural characterization manifests partial retention of the
core/shell systems after successful encapsulation without harming the
crystallinity of the microporous support. The resulting chemically
robust NP@UiO-66 materials exhibit enhanced catalytic activity
towards the liquid-phase hydrogenation of nitrobenzene, competitive
with commercially used Pt on activated carbon, but with simultaneous
superior size-selectivity for sterically varying substrates.
[a]
[b]
C. Rösler, M. Kauer, D. Peeters
Chair of Inorganic Chemistry II
Ruhr-University Bochum
Universitätsstr. 150, 44780 Bochum
S. Dissegna, Prof. Dr. R. A. Fischer
Chair of Inorganic and Metal-organic Chemistry
Technical University Munich
Introduction
Lichtenbergstraße 4, 85748 Garching
E-mail: roland.fischer@tum.de
The utilization of metallic nanoparticles (NP) and reactive species
within porous materials is of great interest, due to a broad field of
applications.[1-3] Especially the synthesis and encapsulation of
bimetallic moieties with controlled size, morphology and
composition inside highly stable metal-organic frameworks
(MOFs) exhibit potential for improved activity and selectivity in
heterogeneous catalysis.[4-6] MOFs are a relatively new class of
crystalline, porous materials with individual and tunable properties
in each structure.[7] Consisting of functional organic entities
connected to metal clusters, multi-dimensional structures with
large inner pore volume and a defined micro- or mesoporosity can
be obtained.[8-9] Hence, MOFs are appropriate candidates for the
stabilization of active NPs together with molecule/substrate
selective properties, due to the individual and regular framework
pore structure.
On the other hand, the well-established bottom-up synthesis of
NPs allows a variety of combinations of different metals.[10] The
most prominent examples for bimetallic NPs are alloys, also
known as solid solutions, with a random distribution of two or more
metals within the NP, and core/shell particles with a defined
interface of the core and shell metal.[11-13] In case of core/shell
systems it is known that the inner metal core significantly
influences the external shell of the second metal and therefore
provides unique synergetic effects and physicochemical
properties.[14] However, so far the exclusive encapsulation of
[c]
[d]
V. L. Rechac, Dr. Francesc X. Llabrés i Xamena
Instituto de Tecnología Química
Universitat Politècnica de València
Consejo Superior de Investigaciones Científicas, Avda. de los
Naranjo, s/n, 46022 Valencia (Spain)
E-mail: fllabres@itq.upv.es
Dr. Yuemin Wang
Institute of Functional Interfaces (IFG)
Karlsruhe Institute of Technology (KIT)
Karlsruhe (Germany)
[e]
[f]
Dr. S. Turner, Prof. Dr. G. Van Tendeloo
EMAT, University of Antwerp
Groenenborgerlaan 171, 2020 Antwerp (Belgium)
P. Guo, K. Ollegott, Prof. Dr. M. Muhler
Laboratory of Industrial Chemistry
Ruhr-University Bochum
Universitätsstr. 150, 44780 Bochum
[g]
[h]
Dr. H. Kobayashi, Prof. Dr. H. Kitagawa
Division of Chemistry, Graduate School of Science
Kyoto University
Kitashirakawa-Oiwakecho, Sakyo-ku, Kyoto 606-8502 (Japan)
T. Yamamoto and Prof. Dr. S. Matsumura
Department of Applied Quantum Physics and Nuclear Engineering,
Graduate School of Engineering,
Kyushu University and the Ultramicroscopy Research Center,
Motooka 744, Nishi-ku, Fukuoka 819-0395 (Japan)
bimetallic NPs with controlled composition into MOFs is
15-19]
challenging and only a few examples are known.[4-5,
The
Supporting information for this article is given via a link at the end of
the document.
synthesis of such systems includes for example the use of
chemical vapor infiltration[4], decomposition of linker-bonded
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
metal complexes[17-18] or the encapsulation of preformed NPs by
a template synthesis.[15, 19] Particularly, the template approach
offers some advantages compared to the other strategies. This
involves i) controlled encapsulation of particles within the
framework with avoidance of undesired particle formation at the
outer surface of the host, suitable for size-selective catalytic
reactions, and ii) control of the exact NP size, morphology and
composition by well-established synthetic routes.[20-21] However,
one important factor is the compatibility of the solvothermal MOF
synthesis and the stability of the NPs at the same time. Recently,
Zhang et al. have established a synthesis for the encapsulation of
Pt NPs into the highly stable UiO-66.[22] This framework belongs
to the group of Zr-based MOFs, offering a three-dimensional pore
system with two types of cavities.[23-24] Moreover, it features high
adjustability of the pore volume and linker functionality.[25-26]
However, to best of our knowledge, no examples for the
encapsulation of bimetallic NPs with controlled composition into
UiO-66-type materials are known so far. One reason for this is the
relatively small pore aperture (~ 6 Å), which excludes a number
of possible organometallic precursors, as well as the high
reactivity of the Lewis-acidic sites and the hydroxy groups, which
are bonded to Zr-cluster inside the framework. Previously, it was
shown that the incorporation of Pd NPs via chemical vapor
infiltration is strongly depending on the infiltration temperature.[24]
Only at low temperatures (-10 °C), sufficient encapsulation of NPs
inside UiO-66 and UiO-67 was possible. However, the applied
conditions allow the infiltration of the metal precursor only within
a long time frame, due to temperature depending deceleration,
and the spatial distribution of the particles was partially
inhomogeneous.
second metal was aspired. It is known that modification of the core
material by a shell can influence core specific properties, such as
reactivity and thermal stability and, consequently, enhances the
overall particle stability.[13]
Scheme 1: Illustration of the template based embedding of metal NPs into MOFs
starting from monometallic NPs. Encapsulation of particles succeeded only in
case of stabilization by a protective layer of a second metal, while the
monometallic particles were leached during the solvothermal synthesis of the
MOF and result in particle free crystals mainly. Color coding: green, secondary
building units of the MOF; grey, organic linkers; blue and yellow, metal atoms of
the NPs.
Herein, the effective core-located encapsulation of surfactant-
stabilized bimetallic PdPt NPs into the Zr-based MOF, UiO-66, by
template synthesis was investigated for the first time. The
resulting materials exhibit substrate specific size-selectivity and
simultaneously improved catalytic activity for the hydrogenation of
different nitrobenzene derivatives, compared to embedded
monometallic counterparts, as also shown by Zhang et al.[29]
Furthermore, the transferability of the concept to other
terephthalate based systems with amino- or nitro-functionality
was investigated and other bimetallic core/shell particles (RuPt)
were successfully incorporated into UiO-66 as well.
Certainly, in case of template synthesis, the solvothermal
conditions of the MOF dictate the type of useable preformed NPs.
For instance, during the synthesis of Zr-based MOFs hydrochloric
acid is produced from commonly used precursors (e.g. ZrCl4),
which in fact dissolves a large number of metal NPs and therefore
prevents an effective encapsulation. One strategy to overcome
this problem is the use of alternative Zr-based precursor (e.g.
zirconium propoxide), for instance for the encapsulation of
hydrochloric acid sensitive Au NPs as shown by Tulig et al.[27]
However, the resulting materials suffer from strong particle growth
and incomplete encapsulation with large numbers of particles at
the outer surface, ineffective for size-selective catalysis.
Results and Discussion
The present absence of examples with bimetallic NPs in Zr-based
MOFs was the motivation for us to exclusively encapsulate
bimetallic nanoparticles into UiO-66 by a protective strategy,
without resignation of the important precursor ZrCl4. As
mentioned above, Pt NPs can be encapsulated into the target
host UiO-66 as their stability is high enough to resist hydrochloric
acid. This fact prompted us to implement a protective, stabilizing
Pt shell around more labile particles, which enables the
incorporation of bimetallic species even under the harsh
solvothermal conditions of the UiO-66 synthesis (Scheme 1).
Initially, the encapsulation of catalytically active Pd NPs was
aimed, because they own promising properties for several
catalytic processes.[28] However, no suitable systems with
exclusively embedded Pd NPs, due to dissolution of small sized
Pd NPs (~ 3 nm) during solvothermal MOF growth, could be
obtained. Hence, the introduction of bimetallic NPs with Pt as the
Initially, the encapsulation of Pd NPs with a size of 3-5 nm was
considered. The particles were produced by reduction of H2PdCl4
in aqueous ethanolic solution in presence of polyvinylpyrrolidone
(PVP). For the synthesis of the metal-loaded UiO-66 structures, a
solvothermal approach with the addition of metal NPs was applied.
Generally, ZrCl4 and acetic acid were mixed in DMF and
sonicated until full dissolution. Afterwards, terephthalic acid was
added and dissolved. Finally, the Pd NPs were dispersed in DMF,
added with subsequent agitation and the mixture was kept in an
oven at 120 °C for 24 h. During the synthesis the original brown
color (Pd NPs) of the mixture disappears, resulting in a colorless,
clear solution. This fact indicates leaching of the used Pd NPs.
The resulting white precipitate was investigated by transmission
electron microscopy (TEM) to elucidate if low quantities of Pd NPs
are still present (Figure S1). The images indicate the formation of
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
well-shaped MOF nanocrystals, however, without the visible
presence of Pd NPs inside the crystallite cores. Only few
agglomerated Pd nanocrystals with broad size distribution were
detected on the outer surface of MOF crystallites, which supports
the observed phenomenon of leaching. Therefore, the
encapsulation process was replaced by PdPt core/shell NPs with
Pt as the shell metal for enhanced stability towards agglomeration
and leaching. PVP functionalized bimetallic nanoparticles as well
as monometallic Pt NPs were produced by established synthetic
strategies. The Pt NPs were synthesized by alcohol based
reduction of the corresponding metal salt H2PtCl6, yielding
particles with a size of approximately 3 nm. Bimetallic core/shell
PdPt NPs were fabricated by a hydrogen-sacrificial protective
approach, as described by Wang et al.[30] A colloidal NP solution
of the core metal Pd was successively coated with Pt atoms under
ambient temperature in hydrogen atmosphere. The introduced H2
dissociates on the Pd NP surface, while in a second step an
aqueous Pt salt solution is slowly added, where Pt ions reduce
predominately at the surface of Pd NPs due to dissociated surface
hydrogen atoms. Following this synthetic approach bimetallic
PdPt NPs with different metal ratios (Pd:Pt, 1:1 and 1:2) were
produced. After purification and drying, the particles were
encapsulated into UiO-66 similar to the previously described
method for the incorporation of Pd NPs. The resulting samples
were denoted as Pd1Pt1@UiO-66 and Pd1Pt2@UiO-66, while
also Pt@UiO-66 with a monometallic species was produced. All
samples retain their dark-brown color during the solvothermal
synthesis. After 24 h, grey (Pt@UiO-66) and brownish-blackish
(PdPt@UiO-66) precipitates with clear, colorless supernatant
were obtained.
(Figure 1).[23] In case of the metal NP containing materials no
additional signals for the corresponding metal NPs in the range of
2 = 38 – 46° were observed. This might be due to the small size
of the particles as well as the low loading quantity. For clarity TEM
measurements were performed.
Figure 2. Bright field TEM images of the metal loaded composites Pt@UiO-66
(a, b), Pd1Pt1@UiO-66 (c, d) and Pd1Pt2@UiO-66 (e, f). c) Magnification of the
area labeled by the white frame shows the presence of core-located, small sized
bimetallic PdPt NPs.
The images indicate successful encapsulation of NPs into the
core of the formed UiO-66 crystals for all samples. Compared to
Pt@UiO-66 (Figure 2a – b), the samples with the bimetallic
species (Figure 2c – f) show higher tendency for coadunate and
intergrown crystals. In case of Pt@UiO-66 and Pd1Pt1@UiO-66
the particles are exclusively encapsulated inside the MOF
crystallites, while Pd1Pt2@UiO-66 indicates partial deposition at
the outer surface. However, this phenomenon was observed
rarely among all the investigated images of different section of the
samples and effects on the size-selective catalysis should be
negligible.
The preformed core Pd NPs exhibit a size of 3.4 ± 0.6 nm (Figure
S2, left). Upon addition of a Pt shell, increase of the particle size
with rising Pt content was expected. The evaluation of the NP
sizes by TEM confirms this assumption with average NPs sizes of
4.2 ± 0.8 nm for Pd1Pt1@UiO-66 and 4.8 ± 0.7 nm for the
d)
c)
b)
a)
composite with
a thicker Pt shell (Pd1Pt2@UiO-66). For
monometallic Pt particles inside UiO-66 an average size of 3.1 ±
0.4 nm was estimated, which is in accordance to literature (Figure
S2, right).[22] Atomic absorption spectroscopy reveals the
expected atomic ratio of Pd and Pt. The obtained results were
also supported by electron-dispersive X-ray (EDX) spectroscopy
measurements as summarized in Table 1.
5
10 15 20 25 30 35 40 45 50
2 Theta [°]
Figure 1. Powder XRD patterns of b) Pt@UiO-66, c) Pd1Pt1@UiO-66 and d)
Pd1Pt2@UiO-66, in comparison to pristine UiO-66 (a).
After washing and solvent exchange with MeOH the samples
were dried in vacuum at room temperature. Initial characterization
was done by powder x-ray diffraction (XRD). The corresponding
patterns of the NP@UiO-66 composites indicate the expected
position of the two intense (111) and (200) reflections of UiO-66
at 2 = 7.4 and 8.5°, analogous to pristine UiO-66 and therefore
verify the intact structure after the encapsulation process
Table 1. Metal ratio between Pd and Pt inside the samples Pd1Pt1@UiO-66 and
Pd1Pt2@UiO-66, determined by elemental analysis and energy-dispersive x-ray
spectroscopy (brackets).
Pd
Pt
Pd1Pt1@UiO-66
Pd1Pt2@UiO-66
55 (54)
36 (38)
45 (46)
64 (62)
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
Interestingly both type of particles with thin and thick shell could
be incorporated without any visible leaching. Clarification of the
NP structure (core/shell or other), with exact location of Pd and Pt
metals, after the encapsulation process was given by annular dark
field scanning transmission electron microscopy (ADF-STEM)
coupled with energy dispersive x-ray (EDX) spectroscopy. The
obtained elemental maps support the presence of core/shell NPs
after the incorporation.
maximum around 1850 cm-1 is characteristic for CO at isolated Pt
in a bridged configuration. Additionally a very weak band at
1945 cm-1 can be assigned to CO adsorbed at a small amount of
Pd that might be diffused through the Pt shell towards the
nanoparticle surface or is not completely covered with Pt.[30] In
contrast, the CO adsorption measurements of Pd1Pt1@UiO-66
with thinner shell reveal a more complex situation of the particle
surface. Similar IR absorption bands corresponding to adsorbed
CO at Zr4+ (2184 cm-1) and electronically modified Zr4+ sites (2153
cm-1) as well as a broad signal at 2068 cm-1 for linearly adsorbed
molecules on Pt were observed. However, the bands for CO
adsorbed at exposed Pd is in this sample more intense compared
to Pd1Pt2@UiO-66, indicating a higher amount of Pd particles at
the nanoparticle surface. The broad low-lying band (1980 to
1820 cm-1) can be assigned to different CO modes on Pd and Pt.
The shoulder at 1850 cm-1 corresponds to bridged CO on Pt, while
the vibrational bands at 1960, 1925 and 1885 cm-1 are
characteristic for bridged-bonded CO–Pd complexes. This
supports our assumption that the sample with higher Pd:Pt ratio
of 1:1 exhibits a thinner Pt shell, compared to the sample with a
Pd:Pt ratio of 1:2, which has a Pt enriched surface. Consequently,
thinner Pt shells result in a higher level of Pd diffusion towards the
nanoparticle surface and therefore alloyed-like surfaces of Pt and
Pd seem to be present. Similar observations were also reported
by Rades et al. for bimetallic Pt-Pd supported zeolites, where the
bimetallic NP surface was composed from both Pd and Pt
atoms.[31]
Figure 3. Left: a) ADF-STEM image of Pd1Pt1@UiO-66. b) - e) Two-dimensional
elemental maps of O (cyan), Pt (red), Pd (green) and Zr (blue). f) Overlay image
of the Pt, Pd and Zr maps. Right: g) High-resolution ADF-STEM image of in UiO-
66 embedded Pd1Pt1 NPs. h) – k) Elemental maps of O (cyan), Pt (red), Pd
(green) and Zr (blue). f) Overlay image of the Pt, Pd and Zr maps.
X-ray photoelectron spectroscopy (XPS) was measured (Figure
S4). The survey XPS spectrum of Pd1Pt1@UiO-66 reveals the
presence of Zr, O and C, without additional signals for Pd and Pt.
The absence of signals for Pd and Pt is most likely due to the
surface sensitivity of XPS, thus confirm the full encapsulation of
Pd/Pt NPs into the host matrix. XPS measurements after
sputtering show the appearance of an additional signal for Pt 4f
(Figure S4c), which was not detected before (Figure S4b). The
ADF-STEM measurements of the encapsulated Pd1Pt1 NPs into
the Zr-based MOF illustrate the exact spatial arrangement of Pd
and Pt inside the particles (Figure 3). The images (Figure 3i–l)
show the presence of Pd (green) preferentially in the NP core
while an enrichment of Pt (red) in the shell of the particles is
displayed. The estimated size of the shell is roughly 1 nm, which
is in accordance to the estimated average NP size (4.2 nm),
compared to pre-synthesized Pd NPs (3.4 nm). However, the
overview maps implicate also, that some Pd core particles are
only partially covered with Pt, as the shell might be very thin and
not clearly detectable by EDX mapping, due to the small size of
particles (Figure 3c–f). Therefore, enrichment of Pd atoms in the
Pt shell cannot be excluded. Besides, STEM-EDX maps illustrate
the embedding of NPs into a Zr and O based (blue, cyan) matrix,
namely UiO-66, which provides evidence for the successful
encapsulation of the particles into the framework.
These observations were supported by low temperature CO
adsorption measurements (UHV-FTIR), comparing the materials
with thinner (Pd1Pt1@UiO-66) and thicker (Pd1Pt2@UiO-66) Pt
shell (Figure S3). Exposing the clean Pd1Pt2@UiO-66 sample to
CO at 100 K leads to the appearance of a major IR band at 2184
cm-1 and a minor IR band at 2153 cm-1, corresponding to CO
adsorbed at parent Zr4+ and electronically modified Zr4+ sites,
respectively. In addition, several bands corresponding to CO
adsorbed at the PdPt core/shell particles can be found. The
intense band at 2061 cm-1 is characteristic for linearly adsorbed
CO at metallic Pt, while the weak and broad band with its
two peaks at 71.3 and 74.6 eV correspond to Pt 4f7/2 and Pt 4f5/2
,
confirming the metallic character of Pt0.[32-33] Signals for Pd could
not be specified because of the overlap of the Pd 3d peak (highest
intensity) around 335 eV with the broad Zr 3p3/2 signal at
333 eV.[34]
Thermogravimetric analysis (TGA) of all samples was performed
(Figure S5). The PdPt@UiO-66 samples, including the pristine
UiO-66, show a decomposition temperature higher than 460 °C.
Nevertheless, the thermal stability of the bimetallic samples is
approximately 40 to 50 °C lower, as compared to the non-loaded
MOF. Destabilization might be attributed to the not ideally grown
MOF structure in the proximate vicinity of the PVP-protected NPs,
as well as PVP chain movement and decomposition. This effect
occurs more clearly in case of the Pt loaded hybrid material.
Decomposition takes place at around 420 °C, while the prior mass
loss (290 to 420 °C) can be explained by the decomposition of
encapsulated PVP. This observation is in accordance to literature
as the stability of NP bounded PVP is decreased.[35] It has to be
mentioned that the Pt NPs were synthesized with higher
metal:surfactant ratio to ensure sizes around 3 nm. Even though
all particles were washed thoroughly to remove the excess of PVP,
it is assumed that slightly higher PVP quantities were present at
the Pt NP. FTIR measurements support this conjecture (Figure 4).
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
a several hundreds of m2/g can be observed, even with small
metal loadings from 0.5 to 2 wt.%. Considering that the loading
amount of a non-porous metal inside a porous system is as low
as 0.5 wt.%, no significant decrease of the surface area should
occur. Therefore, the observed phenomenon can be attributed to
occupation of pores, but more importantly to efficient pore
blocking.[18] Additionally, the precursor decomposition procedure
itself can harm the host. Hence, several sections of the MOF are
not and/or only weakly accessible for guest molecules. In fact this
observation is partially avoided, when the particles are overgrown
by the porous framework, as the second step (precursor
decomposition) is not required and, therefore, the chance for
impairment of the framework structure is minimized.
a)
b)
c)
1653
d)
3644
3600
1500 1200 900 600
Wavenumber [cm-1]
1700 1650 1600
In order to demonstrate the applicability of the concept to other
materials, the encapsulation procedure of the core/shell PdPt NPs
was also applied to two isoreticular Zr-based systems with amino
and nitro functionalized linkers. Therefore, solvothermal synthesis
similar to the non-functionalized UiO-66 was performed, but
terephthalic acid was substituted by 2-aminoterephthalic acid and
2-nitroterephthalic acid, respectively. The resulting materials
PdPt@UiO-66-NH2 and PdPt@UiO-66-NO2 show high
crystallinity after the encapsulation process, as verified by powder
XRD measurements (Figure S7). The reflections of the amino and
nitro derivative of UiO-66 match with those of the pure MOF and,
likewise to the XRD patterns of PdPt@UiO-66, no additional
signals corresponding to the embedded NPs were detected.
Confirmation of the successful incorporation of NPs was provided
by electron microscopy. The TEM images show well grown
partially intergrown UiO-66-NH2 crystals with sharp edges and
core-located NPs (Figure S8 a-b). In case of UiO-66-NO2, similar
results were observed (Figure S8 d-e). Interestingly, the
morphology of the synthesized crystals tends to be more spherical
with no distinct crystal edges and a number of nanocrystals (~ 50
nm) without encapsulated NPs can be observed. Nevertheless,
all PdPt NPs seem to be preferentially imbedded into the crystal
core. The presence of Pd and Pt was confirmed by EDX
measurements (Figure S8 c and f, UiO-66-NH2 and UiO-66-NO2
respectively). Signals for Pd were observed at 2.9 keV (L), while
Pt was detected at 9.4 (L) and 11.1 keV (L). Other detected
elements were C, N, O, Zr, caused by the corresponding UiO-66
derivatives. Therefore, successfully extension to other Zr-
terephthalate based MOFs was demonstrated.
The obtained results demonstrate the possibility for the
incorporation of bimetallic PdPt NPs into UiO-66 under the harsh
solvothermal conditions. Further studies on the encapsulation of
RuPt core/shell NPs were done. Prior to the encapsulation
process into UiO-66, preformed core/shell RuPt NPs were
synthesized by a sequential deposition process of Ru and Pt in
ethylene glycol (EG) according to Alayoglu et al.[36] First Ru(acac)3
(ruthenium acetylacetonate) was reduced at elevated
temperatures in EG. In a subsequent step, PtCl2 was added and
the reaction mixture was allowed to react for another 3 h with
slowly increasing temperature up to reflux. Due to the slow
heating ramp deposition of reduced Pt on the surface of the Ru
seeds was favored. In this way, particles with metal ratios of
Ru1Pt1 and Ru1Pt2 were synthesized. Incorporation of the
particles was done with a similar method than for the PdPt NPs,
Figure 4. Left: FTIR spectra of a) pristine UiO-66 (black), b) Pt@UiO-66 (red),
c) Pd1Pt1@UiO-66 (green) and d) Pd1Pt2@UiO-66 (blue). The yellow bar at
3644 cm-1 indicates the (O-H) stretching of Zr-cluster bonded hydroxy groups.
Right: Enlarged FTIR spectra of the respective samples in the range from 1700
to 1600 cm-1. The dashed line indicates the region of the (C=O) stretching
vibration of PVP.
Pt@UiO-66 shows a weak band at 1653 cm-1, which can be
assigned to the (C=O) stretching vibration of incorporated PVP
(Figure 4, right). Regarding the composites with bimetallic NPs,
the signal is only weakly (Pd1Pt1@UiO-66, green) or not detected
(Pd1Pt2@UiO-66, blue). Moreover, the FTIR measurements verify
the integrity of all materials. The intense signals at 1567 and 1379
cm-1 correspond to the in- and out-of-phase stretching of the
carboxylate groups in UiO-66.[23] Due to the drying procedure at
room temperature the typical (OH) vibration of the Zr bonded OH
groups can be observed at 3644 cm-1 (yellow bar).
The permanent porosity of the samples was confirmed by
nitrogen-sorption measurements (Figure S6). All isotherms show
a type I behavior typical for microporous MOFs, such as UiO-66
with a pore window of 6 Å and cavity sizes of 8 Å and 11 Å for the
tetrahedral and octahedral pore, respectively. Compared to the
reference sample UiO-66, the metal-loaded samples exhibit
slightly decreased Brunauer–Emmett–Teller (BET) surface areas
(Table 2).
Table 2. BET surface area and noble metal loading quantity of the
representative samples UiO-66, Pt@UiO-66, Pd1Pt1@UiO-66, and
Pd1Pt2@UiO-66.
Pt@UiO-
66
Pd1Pt1@UiO-
66
Pd1Pt2@UiO-
66
UiO-66
1135
0
Surface area
[m2/g]
1108
1.1
1039
4.4
1024
5.3
Metal content
[wt.%]
The difference of the measured surface area of all materials is
rather small, compared to pristine UiO-66 (max ~ 100 m2/g), even
though the loading was with up to 5 wt.% of noble metals relatively
high. This shows another advantage of the used template
synthesis method, providing that the excess of stabilizing polymer
was removed from the particle surface. With most of the post-
encapsulation techniques (chemical vapor infiltration, solution
impregnation), also including in-situ incorporation of linker bonded
metals and subsequent decomposition, a significant decrease of
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
resulting in the materials Ru1Pt1@UiO-66 and Ru1Pt2@UiO-66.
Characterization by powder XRD indicates the intact structure of
both metal@MOF composites without detection of embedded
RuPt NPs, in terms of a broad signal around 2 ≈ 40° (Figure S9).
FTIR measurements support the integrity of the structures
compared to pristine UiO-66 and no band for the presence of PVP
at approximately 1650 cm-1 was detected (Figure S10). The
permanent porosity of the samples was verified by N2-sorption
measurements (Figure S11). Both NP@UiO composites exhibit
type I isotherms and the determined BET surface areas were 997
m2g-1 for Ru1Pt1@UiO-66 and 1026 m2g-1 for Ru1Pt2@UiO-66
with a metal loading of 2.3 and 3.1 wt.%, respectively, according
to atomic absorption spectroscopy. Exact calculation of the
atomic ratios shows a deviation from the expected atomic ratios
of Ru and Pt within the composite. The measured ratios of Ru and
Pt of both samples were 38:62 (expected 50:50) and 26:74
(expected 34:66) with a decreased quantity of Ru.
Further evaluation of the materials was done by electron
microscopy. The images verify the encapsulation of NPs into the
core of the UiO-66 crystals (Figure S12). In both cases the
generated crystals show spherical shapes without distinct edges.
Especially the crystals of the composite with a higher Pt quantity
exhibit frayed borders. EDX elemental maps of a loaded crystal
(Ru1Pt1@UiO-66) confirm the presence of Zr, Ru and Pt, with a
lower signal quantity for Ru (Figure S13). However, the exact
particle composition (core/shell or different) could not be
distinguished by elemental mapping. Therefore, high-resolution
ADF-STEM images of the NPs were monitored. The images verify
the formation of core/shell particles with distinct boundary
between the Ru core and Pt shell (Figure S14a – b). Besides RuPt
NPs, pure Pt NPs were partially detected, probably originating
from the alongside reduction of Pt precursor without deposition on
the Ru cores (Figure S14c). The distance between adjacent lattice
fringes was determined to be ~ 0.23 nm, which fits well with the
distance between the (111) planes of fcc Pt. The estimated
particle size after incorporation was found to be 4.2 ± 0.6 and 4.7
± 0.6 nm for Ru1Pt1@UiO-66 and Ru1Pt2@UiO-66, respectively.
These findings are in accordance to the size of the PdPt NP after
encapsulation. Thickening of the Pt shell results in growth of the
NPs and the size of the initial Ru seed NPs (3.3 nm) was found to
be rather similar to those of Pd NPs (3.4 nm).
new highly selective catalysts, including immobilized Au
nanoparticles[41] and Au complexes,[42] which are able to reduce
selectively the nitro group even in the presence of other reducible
functional groups. Therefore, efficient and recyclable catalysts for
the transformation of those nitro compounds to the corresponding
amines are of great interest. The catalytic reactions were
performed at room temperature under hydrogen atmosphere
(p(H2)= 0.2 MPa) in hexane. In order to evaluate size/shape
selective properties of the material, competitive hydrogenation
experiments with two substrates, nitrobenzene (NB) and 3,5-
dimethylnitrobenzene (dmNB), were utilized. The estimated
smallest diameter of NB is approximately 5.8 Å, similar to that of
benzene, while the one from dmNB is about 6.8 Å. Therefore, the
microporous UiO-66 structure with a pore opening of 6 Å allows
basically only the infiltration of the smaller molecule, which could
then subsequently be hydrogenated by the embedded NPs. For
comparison of the size-selective catalytic performance
commercially available Pt NPs on carbon (Pt/C), monometallic Pt
NPs in UiO-66 (Pt@UiO-66) and the parent UiO-66 were also
tested. Unfortunately, testing of a comparable system with
encapsulated Pd NPs was not possible due to the above
mentioned agglomeration and leaching problems during the
solvothermal synthesis of UiO-66. In case of UiO-66 without
encapsulated NPs, no conversion of NB or dmNB was observed
after 24 h. Figure 5 parts a and b show the catalytic results of the
hydrogenation of the pure substrates to the respective anilines,
while parts c and d correspond to the competitive hydrogenation
of equimolar mixtures of both substrates.
The catalytic conversion of NB reveals that the material with
bimetallic NPs (Pd1Pt1@UiO-66) exhibit enhanced activity for the
conversion of NB to aniline with a performance comparable to
Pt/C (Figure 5a). In contrast, the activity of the monometallic
embedded Pt NPs was found to be lower. This fact proofs the
synergetic effects between Pd and Pt atoms in the bimetallic
species. Zhang et al. found similar trends for octahedral core/shell
PdPt NPs during the hydrogenation of 4-chloronitrobenzene.[29] It
should be noted that the material Pd1Pt1@UiO-66 fully reduces
NB quantitatively to aniline within 2 h at room temperature with a
metal:substrate ratio of 1:800, which is superior to the previously
synthesized Pd@UiO-67 system. The Pd@UiO-67 composite
was achieved via incorporation of a Pd-metalated linker (2,2'-
bipyride-5,5'-dicarboxylic acid), where the Pd(II) complex was
subsequently reduced at moderate temperatures to yield small
sized Pd NPs inside the intact framework.[43] Therefore, effective
etching of particles was avoided by post-synthetic formation of the
NPs as it was also the case for other Pd-loaded Zr-MOFs.[24]
Investigation of the molecular size-selective catalytic properties
was realized by use of the sterically more demanding molecule
dmNB. Pt/C exhibits a similar conversion activity for dmNB
compared to NB, while the NP@UiO-66 composites are almost
inactive within the first 3 h (Figure 5b). Comparison of the
catalysts after 1 h indicates a conversion of >99 % for NB and
dmNB in case of Pt/C, featuring surface located Pt NPs. In
contrast, the UiO-66 based hybrid materials show a conversion
as low as 0.2 % (Pd1Pt1@UiO-66) and 0.5 % (Pt@UiO-66). This
strongly supports the complete and successful encapsulation of
the metal NPs within UiO-66 as the diffusion of dmNB into the
MOF is not favored due to the molecule size.
Evaluation of the size-selective hydrogenation potential of
encapsulated PdPt NPs
The catalytic efficiency of the PdPt NP loaded composites
compared to monometallic Pt was tested towards the
hydrogenation of nitrobenzene based compounds. The resulting
amino derivatives represent an important raw material and
intermediate for many fine chemicals, pharmaceuticals, polymers
or dyes.[37] These aromatic amines can be produced from the
corresponding nitroarenes by catalytic hydrogenation, being
various transition metals (viz., Pt,[38] Pd[39] and Ni,[40] either finely
dispersed or adsorbed onto inert supports such as carbon or
alumina) the most common catalysts used since long ago.
Nowadays, however, most efforts are devoted to the design of
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
Figure 5. Time-conversion plot of a) nitrobenzene and b) 3,5-dimethylnitrobenzene in separated reactions. c) Simultaneous conversion of
nitrobenzene (closed symbols) and 3,5-dimethylnitrobenzene (open symbols) in a one-pot reaction. d) Conversion of nitrobenzene (black bars) in
comparison to 3,5-dimethylnitrobenzene (red bars) after 1 h over the catalysts Pt/C, Pt@UiO-66, Pd1Pt1@UiO-66 and Pd1Pt2@UiO-66. The graph
represents the catalytic conversion of both substrates simultaneously. Reaction conditions: 25 °C, p(H2) = 0.2 MPa, hexane.
Independent testing of both substrates separately showed the
potential of those catalysts for shape-selectivity. Additionally, the
catalytic experiments were conducted with both substrates
simultaneously (Figure 5c). The catalytic tests demonstrate the
performance and simultaneous sieving behavior of the NP@UiO
systems. Especially Pd1Pt1@UiO-66 (blue) with an equal molar
metal ratio show enhanced activity towards the conversion of NB,
similar to the observed results from the preliminary studies (Figure
5a), while Pt@UiO-66 (red) is consequently less active.
contrast, Pt/C shows no selectivity for one of the substrates. Only
a slightly enhanced hydrogenation activity for NB can be found.
Pd1Pt1@UiO-66 exhibits
a good activity towards the NB
hydrogenation with ~ 66 % conversion after 1 h of reaction, similar
to Pt/C (74 %), while the reactant specific selectivity is retained
(Figure 5d). This is obviously not the case for Pt@UiO-66 and
Pd1Pt2@UiO-66. Both materials exhibit similar conversion with
approximately 14 % for NB, which in fact confirms that the system
with a thicker Pt shell can be considered as monometallic species
Furthermore, Pd1Pt2@UiO-66 with a thicker shell of Pt was tested. in terms of the activity. Similar results were also obtained for the
The obtained slope (orange) of the NB conversion proceeds
similar to that of Pt@UiO-66 with decreased formation speed to
aniline. This reveals the importance of the shell thickness and
composition in general. It is known from literature that thinner
shells can give higher shell metal dispersion as well as stronger
RuPt NP loaded composites (Figure S15). Both exhibit size-
selective properties with enhanced selectivity for the smaller
molecule NB and the tendency for decreased activity with
increasing Pt shell thickness was also detected.
The observed catalytic activities for hydrogenation of both
substrates under separated and simultaneous conditions were
also confirmed by calculation of turnover frequencies (TOFs)
derived from the kinetic data (Table 3).
It can be observed that both molecules are reduced practically at
the same rate over Pt/C (TOFNB/TOFdmNB = 0.85) in separated
reactions. Under reaction conditions with simultaneous use of
both reactants, NB is preferentially hydrogenated compared to
29]
core/shell interactions, thus resulting in a higher activity.[13,
Therefore, it is assumed that the thick Pt shell eliminates or
strongly reduces the positive electronic effect of the core metal
(Pd) and simultaneously prevents migration of Pd atoms to the
surface as it was observed in case of Pd1Pt1 with equimolar ratio
of both metals. Nevertheless, all materials are size-selective and
almost no conversion of dmNB is observed within the first 3 h. In
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
dmNB (TOFNB/TOFdmNB = 1.83). Nevertheless, Pt/C shows no
appropriate selectivity for any of the molecules. This behavior
completely changes upon the utilization of the NP@MOF based
catalysts. All materials exhibit high reaction rates for NB
(TOFNB/TOFdmNB > 100) without conversion of dmNB, either
separated nor in the mixture. As indicated the TOFs for NB
conversion of the sample with encapsulated bimetallic NPs (691
h-1) is only slightly lower than Pt/C (740 h-1). In contrast
monometallic Pt in UiO-66 exhibits a much lower TOF of 121 h-1,
similar to the value of Pd1Pt2@UiO-66 (119 h-1) with a thicker Pt
shell. At this point it needs to be mentioned that the low
conversion observed for dmNB over the MOF-based catalysts
makes it difficult to provide accurate values for relative TOFs, due
to the error in the calculations. Rather those values should
illustrate the potential of embedded NPs with a shape-selective
support material compared to surface located NPs.
products, such as 3,5-dimethylnitrosobenzene and azoxy
compounds
(azoxybenzene,
dimethylazoxybenzene
and
tetramethylazoxybenzene). It is known from literature that
hydrogenation of nitro-based compounds leads to the formation
of intermediates, such as nitroso and hydroxylamines, which
further react to the corresponding azoxy substances.[44] Du et al.
performed hydrogenation of NB using ethanol with Pt NPs
supported by the mesoporous framework (MIL-101) and detected
similar intermediates.[45] The conversion of the nitro-based
compounds is approximately similar over Pt/C and the MOF
composite in ethanol. Therefore, the MOF is competitive or even
superior to Pt/C in hexane, toluene and ethanol. However, use of
EtOAc as solvent significantly decreased the activity of
Pd1Pt1@UiO-66 and even after 6 h the NB conversion was lower
than 20 %, while Pt/C reaches 100 % after 4 h. In this case, two
facts need to be considered. i) The solubility of H2 is higher in
ethanol and therefore it is assumed that the reaction proceeds
faster as observed with Pt/C.[46] However, this is no explanation
for the significant decrease of activity in case of the MOF
composite. ii) Here a strong influence of the solvent effect may be
inferred. It might be that different solvation of the reactants
hampers the infiltration process of NB into the pores of UiO-66,
while the surface located Pt NPs at carbon are negligible affected.
Albeit reactivity and selectivity are strongly depending on the
solvent, the UiO-66 based samples retain their sieving properties
always compared to Pt/C.
Furthermore, recycling experiments of PdPt@UiO-66 were
performed. The sample showed no significant loss of reactivity
and size-selectivity after three consecutive cycles (Figure S17,
left). Powder XRD measurements confirm the integrity of
crystalline UiO-66 after the catalytic runs (Figure S17, right) and
reflections belonging to metal NPs were not detected. Additionally,
TEM of the hybrid material after catalysis (3 runs) was measured
(Figure S18). The images verify the intactness of the MOF
crystals with core-located PdPt NPs. The formation of PdPt NPs
at the outer surface of the crystallites, caused by metal migration,
is not observed. Agglomeration and growth of NPs was also
excluded, as the average size of the particles after catalysis was
measured to be 4.1 ± 0.7 nm and therefore similar to those of the
material before catalysis (4.2 ± 0.8 nm).
Table 3. Turnover frequencies (TOFs) obtained for NB and dmNB
hydrogenation separately and in an equimolar mixture of both substrates over
the Pt and PdPt containing catalysts.
TOFNB
TOFdmNB
/
TOFNB [h-1]
TOFdmNB [h-1]
[a]
[b]
[a]
[b]
[a]
[b]
1033.
6
Pt/C
880.0
321.6
790.4
-
740.8
121.6
691.2
119.5
404.8
1.1
0.8
1.1
0.85
201
659
1.83
111
864
109
Pt@UiO-66
Pd1Pt1@UiO-66
Pd1Pt2@UiO-66
1.6
1.2
-
[a] TOFs of NB and dmNB in separated reactions. [b] Individual TOFs of NB and
dmNB in an equimolar mixture of both substrates. TOFNB/TOFdmNB illustrates the
NB to dmNB reaction rate ratio. The TOF values were calculated on the basis
of converted substrate per mol of metal per hour at short reaction times
(conversion below 30%), according to metal loadings determined by atomic
absorption spectroscopy.
In fact the utilization of bimetallic NPs in UiO-66 compensates low
diffusion rates through the microporous structure by an activity
enhancement, with simultaneous size-selectivity. Moreover,
partial substitution of Pt by cheaper metals such as Pd decrease
the overall cost for rare noble metals and, therefore, the catalyst
itself.
Evaluation of CO oxidation over RuPt NPs
It is well established that the reaction solvent plays also a major
role for the activity and selectivity of specific catalysts. For the
investigation of these effects upon size-selective hydrogenation
of NB and dmNB over Pt/C and Pd1Pt1@UiO-66 catalysts,
different solvents, such as ethanol, ethyl acetate and toluene,
were tested (Figure S16). Interestingly, the change from hexane
to toluene almost doubled the reaction time needed until 100 %
conversion of NB and dmNB was achieved over Pt/C. In case of
the MOF based composite the decrease of the initial reaction rate
with respect to hexane is much less pronounced, resulting in an
inversion of the activity of both Pt/C and Pd1Pt1@UiO-66 catalysts,
while Pd1Pt1@UiO-66 still maintain the size-specific substrate
selectivity.
In 2008, the group of Eichhorn intensively studied RuPt NPs with
core located Ru and thin Pt layers for the preferential oxidation of
CO in excess hydrogen.[36] Interestingly, the utilized core/shell
systems exhibit the highest catalytic performance for preferential
oxidation of CO among all other utilized particles, including Ru, Pt,
and alloyed particles as well as physical mixtures. The origin of
the activity of RuPt core/shell particles was attributed to changes
in the electronic structure of surface-located Pt species as
evidenced by DFT calculations. At the light of these observed
effects, we decided to test the catalytic performance of our Pt and
Ru1Pt1 NPs embedded in UiO-66 for CO oxidation in presence
and absence of hydrogen (Figure S19)
On the other hand, a change to polar solvents (ethyl acetate
(EtOAc) or ethanol), reveals the formation of several side
The accessibility of the PVP coated encapsulated NPs was
confirmed by the resulting activity, as non-loaded UiO-66 (Figure
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
S19, green line) showed no activity over the whole temperature
range either in H2 or He. Conversion of CO over the NP@UiO-66
The integrity of the materials after the catalytic reaction was
confirmed by powder XRD measurements (Figure S20). No
signals for agglomeration or sintering of NPs were detected,
which, in fact, proofs the stabilization effect by the host matrix
UiO-66.
catalysts reveals
a clear temperature shift towards lower
temperatures for the system with the bimetallic species as
compared to monometallic NPs. In excess hydrogen atmosphere
the onset (light-off) of CO conversion over Ru1Pt1@UiO-66 is
observed at 115 °C, while 50 % conversion is achieved at 130 °C.
In case of Pt@UiO-66, conversion starts approximately at 170 °C,
but only 50 % conversion is obtained until the maximum
measured temperature of 200 °C. The results indicate the higher
catalytic activity of the bimetallic RuPt NPs due to synergetic
Conclusions
In summary, the encapsulation of preformed bimetallic NPs into
the thermally and chemically robust Zr-based UiO-66 and its
derivatives UiO-66-NH2 and UiO-66-NO2 was demonstrated for
the first time. The applied concept combines the utilization of a
stable with a more labile metal in core/shell-like fashion and
subsequent imbedding into Zr-based MOFs without the
abandonment of the important precursor ZrCl4. Due to the
formation of hydrochloric acid during the solvothermal synthesis
and the harsh conditions (120 °C), the encapsulation process is
strongly limited to very stable metals. Pt proved to be promising
for the utilization as protective metal as it indicates successful
encapsulation into UiO-66 without strong leaching behavior. We
faced the problem with monometallic Pd NPs, which show strong
aggregation and almost complete leaching during the
solvothermal synthesis. Therefore, direct encapsulation was
challenging and no appropriate NP@MOF systems with
exclusively embedded NPs were obtained. However, following
the protective synthetic path, enabled the successful
encapsulation of PdPt core/shell NPs into UiO-66 with different
metal compositions. The resulting PdPt@UiO-66 materials show
excellent catalytic performance for the hydrogenation of
nitrobenzenes, compared to their monometallic counterpart
Pt@UiO-66, with simultaneous shape-selectivity for different
sized substrates. The stability of the catalyst was proven by
recycling experiments, with no significant change of the catalyst
(NP size, crystallinity) or the catalytic performance. Moreover, the
multi-metal induced catalytic activity and substrate specific
selectivity of the microporous support could even surpass the
performance of commercially available Pt/C, especially with view
to selectivity. Further, the concept was transferred to other UiO-
66 derivatives and also RuPt core/shell particles were
successfully incorporated into the Zr-based MOF. The
RuPt@UiO-66 materials show high catalytic activity for the
oxidation of CO, compared to Pt@UiO-66. All together the
findings demonstrate the great potential and the beneficial
properties of metal@MOF-based materials with bimetallic species
for catalytic processes. Application of bimetallic particles seems
to balance diffusion limitations of the porous support compared to
surface located Pt NPs (Pt/C) without loss of general shape- and
size-selectivity. As the situation of the particle surfaces was not
completely clear, further studies will concentrate on the exact NP
composition before and after hydrogenation reactions in future.
The manufacture of NP@MOF composites with multi-metallic
NPs remains an attractive research field, especially in view to
environmental and economically relevant reasons, and will be
interesting for several other applications as well.
effects under hydrogen atmosphere with
a difference of
approximately 70 °C (T50) compared to monometallic Pt NPs. For
comparison, further reactions were conducted under exclusion of
hydrogen. Surprisingly, it was found that Ru1Pt1@UiO-66 shows
even higher activity in absence of hydrogen with temperatures
around 120 °C. On the contrary, the temperature of Pt@UiO-66
shifts towards higher values (~ 240 °C). In a precedent study by
Zhang et al. on similar systems (Pt@UiO-66) with preformed NPs,
full CO conversion was already observed at 180 °C.[22] Obviously,
the used conditions (gas flow, gas concentration, heating ramp,
etc.) as well as the equipment strongly influence the final activity
and direct comparison to literature is rather difficult.
Nevertheless, CO conversion with this specific set-up illustrates
the superior properties of the bimetallic NPs, as full conversion
was reached at half of the temperature compared to monometallic
Pt NPs (T = 120 °C). In order to elucidate that the support has
no crucial effect on the CO conversion, as the catalytic findings
for RuPt NPs were very different from those of the published
ones,[36] we reproduced the systems, with Al2O3 support,
according to literature. Utilization of similar conditions as for the
loaded MOFs, shows no significant difference of temperatures in
case of the Al2O3 supported NPs (Table 4). This implies that the
support itself has no direct or only a slight impact on the overall
activity in this particular case. Ru1Pt1@Al2O3 (147 °C) exhibits
slightly higher temperatures towards CO oxidation with respect to
the MOF analog (130 °C), while the observed trend of lower
needed temperatures in helium was similarly present (H2 = 147 °C
vs. He = 127 °C). Pt@Al2O3 exhibits conversion of 50 % at T50
=
269 °C, which is 30 °C higher in comparison to Pt@UiO-66,
probably caused by the larger Pt NPs size (~ 6 nm), which were
prepared by reduction in ethylene glycol instead of methanol
(Pt@UiO-66). Table 4 summarizes the obtained CO oxidation in
H2 and He.
Table 4. Catalytic activity of the UiO-66 and Al2O3 based systems on basis of
the temperature at 50 % CO conversion (T50
) in hydrogen or helium,
respectively. The concentration for the reactive gases was set to 0.25 % for CO
and 0.5 % for O2.
Gas
Temperature (T50) [°C]
Pt@UiO-
66
Ru1Pt1@UiO-
Pt@Al2O3
Ru1Pt1@Al2O3
66
Hydrogen
Helium
~ 200
237
130
-
147
127
118
269
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
dark brown residue was redissolved and washed several times with ethanol.
The particles were collected by centrifugation and dried under vacuum (p =10–3
mbar) for 12 h to yield a dark brown product.
Experimental Section
Materials: All chemicals were purchased from commercial suppliers (Sigma-
Aldrich, Alfa Aesar, Acros Organics and others) and used without further
purification.
PVP stabilized core/shell Pd1Pt1 and Pd1Pt2 NPs
The synthesis was done by a modified procedure of Wang et al.[30] The prepared
Pd NPs (core material) were dispersed in a 150 ml mixture of water-ethanol-
ethylene glycol (1:1:1) in a 500 ml flask which was equipped with a dropping
funnel and a connection to hydrogen (99.999 %). The dropping funnel was
charged with a degassed aqueous solution (100 ml) of K2PtCl4 (83 mg, 0.2
mmol) which was kept under Argon atmosphere. The air of the flask with the Pd
colloid was replaced by a hydrogen atmosphere and the colloid was allowed to
stay under these conditions for 2 h under stirring. Subsequently, the K2PdCl4
solution was added drop wise into the reaction system within 6 h and the
reaction was continued for another 8 h. Afterwards the PdPt NPs were collected
by centrifugation, washed several times with EtOH and dried under vacuum (p
=10–3 mbar) for 12 h. To avoid any oxidation the particles were stored under an
inert gas atmosphere until usage. Similarly, core/shell particles with a thicker Pt
shell (named as Pd1Pt2) were synthesized. But instead of a 100 ml K2PtCl4
solution, 200 ml of a K2PtCl4 (166 mg, 0.4 mmol) solution were used.
Characterization Methods: Powder X-ray diffraction (PXRD) measurements
were recorded in flat mode with an X’Pert PRO MRD PANalytical equipment in
Bragg–Brentano geometry, with a PIXcel position sensitive detector and a CuKα
radiation source (λ = 1.54178 Å) at room temperature. FTIR spectra were
measured with a Bruker Alpha-P FTIR spectrometer equipped with a single-
reflection diamond ATR module under inert gas conditions. Nitrogen sorption
measurements were performed with a 3Flex Surface Characterization Analyzer
instrument from Micromeritics with nitrogen gas (99.9995%) at 77 K. The
specific surface areas were calculated using the sorption data in the relative
pressure range from 0.05 to 0.2 bar. Prior to the measurements, all samples
were evacuated in dynamic vacuum (p = 10–6 mbar) at 100 °C for 6 h. The
elemental composition of the compounds was measured at the
Mikroanalytisches Labor Kolbe in Mülheim an der Ruhr, Germany.
Thermogravimetric analysis was performed with a Netzsch STA 409 PC/PG
instrument (sample weight approximately 5–10 mg) in a temperature range from
303 to 1073 K with a heating ramp of 5 Kmin–1 at atmospheric pressure and
under a stream of nitrogen (99.999%, 300 mLmin–1). X-ray photon spectroscopy
(XPS) was performed with a Physical Electronics PHI 5000 VersaProbe
spectrometer, using a standard Al Ka source (1486.6 eV) powered at 300 W.
The working pressure was lower than 10-7 mbar. Reported binding energy (BE)
were charge effect corrected by assigning a BE of 284.8 eV to the adventitious
C1s signal. Ar+ sputtering was carried out at 3 kV and a 0.5 mA cm-2 beam
PVP stabilized core/shell Ru1Pt1 and Ru1Pt2 NPs
The RuPt NPs were synthesized by a procedure of Alayoglu et al.[36] Ru(acac)3
(80 mg, 0.2 mmol) and PVP (55 mg, 0.5 mmol) are dissolved in 40 ml ethylene
glycol. Subsequently the Ru NPs (core metal) are synthesized by a fast
temperature ramp and subsequent refluxing for 3 h under stirring and Argon
atmosphere. When the colloidal Ru NP solution was cooled down to room
temperature, PtCl2 (54 mg, 0.2 mmol) was added and the mixture was heated
slowly with 1-2 K min-1 to reflux and kept at this conditions for 1.5 h. Afterwards
the Ru1Pt1 NPs were collected by centrifugation, washed several times with
EtOH and dried under vacuum (p =10–3 mbar) for 12 h. To avoid any oxidation
the particles were stored under an inert gas atmosphere until usage. For the
synthesis of core/shell particles with a thicker layer of Pt, the amount of PtCl2
added was doubled.
current density with an argon partial pressure of
5 ×
10-8 mbar. The
measurements were performed at the Faculty of Physics and Astronomy,
Department of Experimental Physics II, Ruhr-University Bochum. Transmission
electron microscopic measurements were carried out with a Tecnai G2 F20 and
a Philips CM20 with an acceleration voltage of 200 kV at the department of
Mechanical Engineering, Ruhr-University Bochum, Germany. Energy-
dispersive X-ray spectroscopy was performed at the Tecnai G2 F20. Annular
dark field scanning transmission electron microscopy (ADF-STEM) and high
resolution ADF-STEM images were recorded using an aberration corrected FEI
Titan ‘Cubed’ 60-300 transmission electron microscope, operated at 200 kV.
The inner collection semi-angle of the ADF-STEM detector was approximately
32 mrad and the probe convergence semi-angle 21 mrad. High-angle annular
dark-field scanning transmission electron microscopy and energy-dispersive X-
ray spectroscopy were performed at the Research Laboratory of High- Voltage
Synthesis of UiO-66 and NP@UiO-66
UiO-66
In a typical synthesis ZrCl4 (612 mg, 2.63 mmol) was dissolved in 600 ml N,N-
Dimethylformamide (DMF) in a teflon-capped 1 l glass flask. Subsequently, 36
ml of acetic acid were added and the mixture was allowed to sonicate for 20 min.
Afterwards, Terepthalic acid (435 mg, 2.62 mmol) was added and the solution
was treated for another 20 min in an ultrasonic bad. Solvothermal synthesis was
done in an oven at 120 °C for 24 h. When the solution was cooled down to room
temperature, the white precipitate was collected and washed three times with
fresh DMF (30 ml). Additionally the washing procedure includes 3 periods of
fresh DMF exchange after 6 h, respectively. Afterwards DMF was exchanged
with MeOH (3 times, 30 ml). Finally, the material was dried in vacuum (p = 10-3
mbar) at room temperature for 24 h and stored inside a glovebox.
Electron Microscope, Kyushu University, Japan, recorded with
ARM200F operating at 200 kV.
a JEM-
Synthesis of metal nanoparticles
PVP stabilized Pt NPs
The Pt NPs were synthesized by a slightly modified procedure of Lu et al.[21]
Therefore, a mixture of PVP (399 mg, 3.6 mmol, Mw=55.000), methanol (540
ml) and an aqueous solution of H2PtCl6 (6 mM, 60 ml) was refluxed for 3 h under
air. Afterwards methanol was removed with help of a rotary evaporator and the
Pt particles in the remaining solution were precipitated with acetone.
Subsequently, the particles were separated by centrifugation at 7000 g for 3
minutes and washed several times with ethanol and ether for the removal of
excess free PVP. Drying of the precipitated particles under reduced pressure
(10-3 mbar) results in a black powder.
UiO-66(NH2) and UiO-66(NO2)
ZrCl4 (51.3 mg, 0.220 mmol) was dissolved in 50 ml DMF in a Teflon-capped
100 ml glass flask. Afterwards, 3 ml acetic acid were added and the mixture was
sonicated for 20 min. Subsequently, 2-Aminoterephthalic acid (39.9 mg, 0.220
mmol) or 2-Nitroterephthalic acid (46.4 mg, 0.220 mmol) was added and the
mixture was allowed to sonicate for another 20 min. The solvothermal synthesis
was done for 24 h at 120 °C. After cooling to room temperature the precipitate
was collected. The washing and drying procedure is similar to normal UiO-66.
PVP stabilized Pd NPs
Pd NPs were synthesized by the reduction of H2PdCl4 with ethanol. A 2 mM
H2PdCl4 solution (100 mL, 0.2 mmol) was mixed with PVP (23 mg, 0.2 mmol)
and H2O (560 mL). Under vigorous stirring, ethanol (160 mL) was added and
the mixture was heated to reflux for 3 h. Afterward the brown solution was
cooled to room temperature, and the solvent was removed by evaporation. The
PdPt@UiO-66
The synthesis of PdPt@UiO-66 is similar to pristine UiO-66, but after the last
sonication step a colloidal dispersion of the particles (30 mg dissolved in 30 ml
DMF) was added. The reaction mixture was agitated for 10 sec and then allowed
to react at 120 °C for 24 h without stirring, resulting in a brown-black precipitate.
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
[9]
H. Furukawa, K. E. Cordova, M. O’Keeffe, O. M. Yaghi, Science 2013,
341, DOI: 10.1126/science.1230444.
RuPt@UiO-66
[10] C.-J. Jia, F. Schuth, Physical Chemistry Chemical Physics 2011, 13,
The synthesis of RuPt@UiO-66 was similar to PdPt@UiO-66, but instead RuPt
NPs were used. The solvothermal reaction with the RuPt NPs results in a grey
powder.
2457-2487.
[11] D. Wang, Y. Li, Advanced Materials 2011, 23, 1044-1060.
[12] H. Kobayashi, K. Kusada, H. Kitagawa, Accounts of Chemical
Research 2015, 48, 1551-1559.
[13] R. Ghosh Chaudhuri, S. Paria, Chemical Reviews 2012, 112, 2373-
Catalytic hydrogenation of nitrobenzenes
2433.
The competitive hydrogenation of nitrobenzene (0.25 mmol) and 3,5-
dimethylnitrobenzene (0.25 mmol) was carried out at ambient temperature
inside 2 mL reactors pressurized at 2 bars H2, with 1 ml of hexane, ethanol ,
ethyl acetate or toluene, respectively. The catalyst amount for each run was
adjusted to 0.06 mol-% metal with respect to both compounds. Liquid reaction
products were analyzed by GC and GC-MS. Catalytic Reusability of the catalyst
was performed in ethanol. After each cycle, the solid catalyst was recovered by
filtration and washed thoroughly with ethanol. After drying in vacuum (10-3 mbar)
overnight the catalyst was reused.
[14] R. Ferrando, J. Jellinek, R. L. Johnston, Chemical Reviews 2008, 108,
845-910.
[15] Y. Huang, Y. Zhang, X. Chen, D. Wu, Z. Yi, R. Cao, Chemical
Communications 2014, 50, 10115-10117.
[16] J. Zhou, P. Wang, C. Wang, Y. T. Goh, Z. Fang, P. B. Messersmith, H.
Duan, ACS Nano 2015.
[17] L. Chen, B. Huang, Q. Xuan, X. Wang, R. Luque, Y. Li, Chemical
Science 2015.
[18] L. Chen, X. Chen, H. Liu, Y. Li, Small 2015, 11, 2642-2648.
[19] C. Rösler, D. Esken, C. Wiktor, H. Kobayashi, T. Yamamoto, S.
Matsumura, H. Kitagawa, R. A. Fischer, European Journal of Inorganic
Chemistry 2014, 2014, 5514-5521.
[20] Y. Liu, Z. Tang, Advanced Materials 2013, 25, 5819-5825.
[21] G. Lu, S. Li, Z. Guo, O. K. Farha, B. G. Hauser, X. Qi, Y. Wang, X.
Wang, S. Han, X. Liu, J. S. DuChene, H. Zhang, Q. Zhang, X. Chen, J.
Ma, S. C. J. Loo, W. D. Wei, Y. Yang, J. T. Hupp, F. Huo, Nature
Chemistry 2012, 4, 310-316.
[22] W. Zhang, G. Lu, C. Cui, Y. Liu, S. Li, W. Yan, C. Xing, Y. R. Chi, Y.
Yang, F. Huo, Advanced Materials 2014, 26, 4056–4060.
[23] J. H. Cavka, S. Jakobsen, U. Olsbye, N. Guillou, C. Lamberti, S.
Bordiga, K. P. Lillerud, Journal of the American Chemical Society 2008,
130, 13850-13851.
[24] I. Luz, C. Rösler, K. Epp, F. X. Llabrés i Xamena, R. A. Fischer,
European Journal of Inorganic Chemistry 2015, 2015, 3904-3912.
[25] M. Kandiah, M. H. Nilsen, S. Usseglio, S. Jakobsen, U. Olsbye, M.
Tilset, C. Larabi, E. A. Quadrelli, F. Bonino, K. P. Lillerud, Chemistry of
Materials 2010, 22, 6632-6640.
[26] Z. Hu, D. Zhao, Dalton transactions 2015, 44, 19018-19040.
[27] K. Tulig, K. S. Walton, RSC Advances 2014, 4, 51080-51083.
[28] I. Saldan, Y. Semenyuk, I. Marchuk, O. Reshetnyak, Journal of
Materials Science 2015, 50, 2337-2354.
[29] P. Zhang, Y. Hu, B. Li, Q. Zhang, C. Zhou, H. Yu, X. Zhang, L. Chen, B.
Eichhorn, S. Zhou, ACS Catalysis 2015, 5, 1335-1343.
[30] Y. Wang, N. Toshima, The Journal of Physical Chemistry B 1997, 101,
5301-5306.
[31] T. Rades, V. Y. Borovkov, V. B. Kazansky, M. Polisset-Thfoin, J.
Fraissard, The Journal of Physical Chemistry 1996, 100, 16238-16241.
[32] G. M. Bancroft, I. Adams, L. L. Coatsworth, C. D. Bennewitz, J. D.
Brown, W. D. Westwood, Analytical Chemistry 1975, 47, 586-588.
[33] H. Liu, L. Chang, L. Chen, Y. Li, J. Mater. Chem. A 2015, 3, 8028-8033.
[34] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. D. Bomben, Handbook of
X-ray Photoelectron Spectroscopy, Perkin Elmer Corporation, Eden
Prairie, Minnesota, USA, 1992.
CO oxidation
A U-shaped quartz reactor was loaded with 20 mg catalyst diluted in 80 mg
silicon carbide and placed in a continuous flow set-up. Prior to CO oxidation, a
reductive pretreatment in hydrogen atmosphere at 200 °C was performed. The
catalytic oxidation of CO was performed in a gas mixture of 0.25 % CO (minimal
value limited by accurarcy of the mass flow controllers used) and 0.5 % O2
diluted in helium or hydrogen with a total flow rate of 100 mL min-1. Prior to CO
oxidation a reductive pretreatment of the catalysts was done. First pure helium
was purged for 15 min through the reactor. Afterwards, the rector was heated
up to the final pretreatment temperature (200 °C) with 1 K min-1 in a flow of 100
mL min-1 in 10 % hydrogen and maintained at this temperature for 60 min, before
the system was allowed to cool down to room temperature. Subsequently the
reactor was purged with He and the desired reaction mixture was flushed
through the reactor and the sample was heated to 200 °C (carrier gas H2) and
300 °C (carrier gas He) with a heating rate of 1 Kmin-1. The effluent gas stream
after the catalytic process was analyzed with a non-dispersive IR-detector
(Hartmann&Braun, Advanced Optima URAS 14) to determine the concentration
of the out coming species. Concentrations of CO and CO2 during the catalytic
CO oxidation were measured by a NDIR detector.
Acknowledgements
This work is supported by the Cluster of Excellence RESOLV
(EXC 1069) funded by the Deutsche Forschungsgemeinschaft
(DFG).
[35] Y. K. Du, P. Yang, Z. G. Mou, N. P. Hua, L. Jiang, Journal of Applied
Polymer Science 2006, 99, 23-26.
[36] S. Alayoglu, A. U. Nilekar, M. Mavrikakis, B. Eichhorn, Nature Materials
2008, 7, 333-338.
[37] F. A. Westerhaus, R. V. Jagadeesh, G. Wienhöfer, M.-M. Pohl, J.
Radnik, A.-E. Surkus, J. Rabeah, K. Junge, H. Junge, M. Nielsen, A.
Brückner, M. Beller, Nature Chemistry 2013, 5, 537-543.
[38] R. Adams, F. L. Cohen, Organic Syntheses 1928, 8, 66.
[39] P. M. G. Bavin, Organic Syntheses 1960, 40, 5.
[40] C. F. H. Allen, J. VanAllan, Organic Syntheses 1942, 22, 9.
[41] A. Corma, P. Serna, Science 2006, 313, 332-334.
[42] A. Corma, C. Gonzalez-Arellano, M. Iglesias, F. Sanchez, Applied
Catalysis A. General 2009, 356, 99-102.
Keywords: Metal-Organic Frameworks
Nanoparticles • Catalysis
•
Nanostructures •
[1]
[2]
H. R. Moon, D.-W. Lim, M. P. Suh, Chemical Society Reviews 2013, 42,
1807-1824.
J. Juan-Alcaniz, J. Gascon, F. Kapteijn, Journal of Materials Chemistry
2012.
[43] L. Chen, H. Chen, R. Luque, Y. Li, Chemical Science 2014, 5, 3708-
3714.
[44] L. Cisneros, P. Serna, A. Corma, Angewandte Chemie International
Edition 2014, 53, 9306-9310.
[45] W. Du, G. Chen, R. Nie, Y. Li, Z. Hou, Catalysis Communications 2013,
41, 56-59.
[46] M. S. Wainwright, T. Ahn, D. L. Trimm, N. W. Cant, Journal of Chemical
& Engineering Data 1987, 32, 22-24.
[3]
[4]
C. R. Kim, T. Uemura, S. Kitagawa, Chemical Society Reviews 2016.
J. Hermannsdörfer, M. Friedrich, N. Miyajima, R. Q. Albuquerque, S.
Kümmel, R. Kempe, Angewandte Chemie 2012, 124, 11640-11644.
Q.-L. Zhu, J. Li, Q. Xu, Journal of the American Chemical Society 2013,
135, 10210-10213.
Z. Li, R. Yu, J. Huang, Y. Shi, D. Zhang, X. Zhong, D. Wang, Y. Wu, Y.
Li, Nature Communications 2015, 6.
W. Lu, Z. Wei, Z.-Y. Gu, T.-F. Liu, J. Park, J. Park, J. Tian, M. Zhang,
Q. Zhang, T. Gentle Iii, M. Bosch, H.-C. Zhou, Chemical Society
Reviews 2014, 43, 5561-5593.
[5]
[6]
[7]
[8]
N. Stock, S. Biswas, Chemical Reviews 2011.
This article is protected by copyright. All rights reserved.

10.1002/chem.201603984
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
Christoph Rösler, Stefano Dissegna,
Victor L. Rechac, Max Kauer, Peng-Hu
Guo, Maria Filippousi, Kevin Ollegott,
Hirokazu Kobayashi, Yuemin Wang,
Hiroshi Kitagawa, Martin Muhler,
Francesc X. Llabrés i Xamena* and
Roland A. Fischer*
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))
Page No. – Page No.
Encapsulation of bimetallic metal
nanoparticles into robust Zr-based
metal-organic frameworks:
Evaluation of the catalytic potential
for size-selective hydrogenation
This article is protected by copyright. All rights reserved.