Organometallic Synthesis of Bimetallic Cobalt-Rhodium Nanoparticles in Supported Ionic Liquid Phases (CoxRh100?x@SILP) as Catalysts for the Selective Hydrogenation of Multifunctional Aromatic Substrates

Article abstract of DOI:10.1002/smll.202006683

The synthesis, characterization, and catalytic properties of bimetallic cobalt-rhodium nanoparticles of defined Co:Rh ratios immobilized in an imidazolium-based supported ionic liquid phase (Co_xRh_100?x@SILP) are described. Following an organometallic approach, precise control of the Co:Rh ratios is accomplished. Electron microscopy and X-ray absorption spectroscopy confirm the formation of small, well-dispersed, and homogeneously alloyed zero-valent bimetallic nanoparticles in all investigated materials. Benzylideneacetone and various bicyclic heteroaromatics are used as chemical probes to investigate the hydrogenation performances of the Co_xRh_100?x@SILP materials. The Co:Rh ratio of the nanoparticles is found to have a critical influence on observed activity and selectivity, with clear synergistic effects arising from the combination of the noble metal and its 3d congener. In particular, the ability of Co_xRh_100?x@SILP catalysts to hydrogenate 6-membered aromatic rings is found to experience a remarkable sharp switch in a narrow composition range between Co₂₅Rh₇₅ (full ring hydrogenation) and Co₃₀Rh₇₀ (no ring hydrogenation).

Full text of DOI:10.1002/smll.202006683

Full PaPer
www.small-journal.com
Organometallic Synthesis of Bimetallic Cobalt-Rhodium
Nanoparticles in Supported Ionic Liquid Phases
(CoxRh100−x@SILP) as Catalysts for the Selective
Hydrogenation of Multifunctional Aromatic Substrates
Simon Rengshausen, Casey Van Stappen, Natalia Levin, Simon Tricard, Kylie L. Luska,
Serena DeBeer, Bruno Chaudret, Alexis Bordet,* and Walter Leitner*
1. Introduction
The synthesis, characterization, and catalytic properties of bimetallic
cobalt-rhodium nanoparticles of defined Co:Rh ratios immobilized in an
imidazolium-based supported ionic liquid phase (CoxRh100−x@SILP) are
described. Following an organometallic approach, precise control of the
Co:Rh ratios is accomplished. Electron microscopy and X-ray absorption
spectroscopy confirm the formation of small, well-dispersed, and
homogeneously alloyed zero-valent bimetallic nanoparticles in all investigated
materials. Benzylideneacetone and various bicyclic heteroaromatics are
used as chemical probes to investigate the hydrogenation performances
of the CoxRh100−x@SILP materials. The Co:Rh ratio of the nanoparticles is
found to have a critical influence on observed activity and selectivity, with
clear synergistic effects arising from the combination of the noble metal
and its 3d congener. In particular, the ability of CoxRh100−x@SILP catalysts to
hydrogenate 6-membered aromatic rings is found to experience a remarkable
sharp switch in a narrow composition range between Co25Rh75 (full ring
hydrogenation) and Co30Rh70 (no ring hydrogenation).
The preparation of bimetallic nanopar-
ticles (NPs) combining noble metals
with their base metal 3d congeners has
attracted significant attention as an effec-
tive strategy to produce materials with
promising optical,[1,2] magnetic,[1,3] and
catalytic[1,4] properties. The potential
benefits are particularly pronounced in
the field of catalysis, where combining
metals to produce core–shell or alloyed
nanostructures was shown to generate
significant synergistic effects, providing
access to activities and selectivities that
are out of reach for the metals taken indi-
vidually.[4] As a result, important technical
applications such as catalytic reforming,[5]
alcohol oxidation[6] and electrocatalysis in
fuel cells[7] are now relying on the use of
bimetallic catalysts. The use of bimetallic
NPs to achieve the controlled activation
S. Rengshausen, Dr. C. Van Stappen, Dr. N. Levin, Prof. S. DeBeer,
of molecular hydrogen has also gained increased attention,
with the goal of developing supported chemoselective cata-
lysts capable of hydrogenating complex molecules possessing
various functional groups.[4b–d] While a wide range of supports
and preparation techniques are available for the synthesis of
monometallic NPs,[8] systematic approaches compatible with
the preparation of bimetallic structures over a wide range
of defined compositions are comparatively scarce.[9] Such
studies would be highly desirable to arrive at a fundamental
understanding of how the composition of bimetallic alloys in
nanoparticles affects the performance of catalyst materials.
In this context, the immobilization of metal NPs on sup-
ported ionic liquid phases (SILP) has emerged as a versatile
molecular approach to producing multifunctional catalytic
systems with tailored reactivity.[10] Recent studies have shown
that SILPs provide suitable matrices for the synthesis and
stabilization of monometallic and bimetallic NPs, producing
catalytic systems possessing excellent properties for hydro-
genation,[11] hydrodeoxygenation,[12] and hydrogenolysis[13]
reactions. For example, bimetallic FeRu NPs immobilized
on an imidazolium-based SILP have shown interesting reac-
Dr. A. Bordet, Prof. W. Leitner
Max Planck Institute for Chemical Energy Conversion
Stiftstrasse 34–36, Mülheim an der Ruhr 45470, Germany
E-mail: alexis.bordet@cec.mpg.de; walter.leitner@cec.mpg.de
S. Rengshausen, Dr. K. L. Luska, Prof. W. Leitner
Institut für Technische und Makromolekulare Chemie
RWTH Aachen University
Worringerweg 2, Aachen 52074, Germany
Dr. S. Tricard, Prof. B. Chaudret
Laboratoire de Physique et Chimie des Nano-Objets
Université de Toulouse
INSA
UPS
LPCNO
CNRS-UMR5215
135 Avenue de Rangueil, Toulouse 31077, France
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/smll.202006683.
© 2020 The Authors. Small published by Wiley-VCH GmbH. This is an
open access article under the terms of the Creative Commons Attribu-
tion-NonCommercial License, which permits use, distribution and repro-
duction in any medium, provided the original work is properly cited and
is not used for commercial purposes.

tivity for the selective hydrogenation of non-aromatic C C
DOI: 10.1002/smll.202006683
2006683 (1 of 10)
Small 2021, 17, 2006683
© 2020 The Authors. Small published by Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Figure 1. Illustration of the approach followed in this study. a) Organometallic synthesis, b) characterization, and c) catalytic application of bimetallic
CoxRh100−x@SILP catalysts.

and C O bonds while showing no activity toward aromatic
2. Results and Discussion
rings.[11c]
In order to further explore the potential of bimetallic
NPs, we have turned to systematically investigating NPs
comprised of Rh and its 3d congener, Co, in defined com-
positions within a SILP matrix. Bimetallic CoRh NPs have
previously been studied for their magnetic,[14] structural[15]
and catalytic properties, e.g., in hydroformylation,[16] Pauson-
Khand reaction,[16] and reductive cyclization.[17] However, their
use in hydrogenation reactions has rarely been explored,[18]
and versatile synthetic methods allowing the preparation
of supported CoxRh100−x NPs with tunable Co:Rh ratios are
lacking.
Herein, we present the synthesis, characterization, and
application of supported bimetallic CoRh NPs. CoxRh100−x
NPs were synthesized following an organometallic approach
involving the co-reduction of the precursors [Co(cod)(cycloocta-
dienyl)] (with cod = 1,5 cyclooctadiene) and [Rh(allyl)3] directly
in presence of a SILP consisting of an imidazolium-based
ionic liquid covalently grafted on silica. This SILP was chosen
due to its proven potential as suitable matrix for the forma-
tion and stabilization of mono- and bimetallic nanoparticles
(Figure 1a).[11c,12c,e] The selected organometallic metal com-
plexes were chosen based on a combination of stability during
handling and ligand lability upon hydrogenation, allowing the
formation of well-defined NPs with a clean surface and tun-
able composition. In-depth characterization of the resulting
CoxRh100−x@SILP materials was achieved using a variety of
techniques including electron microscopy, XRD, and XAS to
collect information regarding NP size, dispersion, oxidation
state and alloying extent (Figure 1b). The catalytic properties of
the CoxRh100−x@SILP materials towards hydrogenation of mul-
tifunctional aromatic substrates were investigated, with ben-
zylideneacetone and bicyclic heteroaromatics used as chem-
ical probes to study the influence of the Co:Rh metal ratio on
activity and selectivity (Figure 1c).
2.1. Synthesis
Preparation of the imidazolium-based SILP was achieved fol-
lowing a previously published procedure.[12a] Briefly, the silane-
functionalized IL [1-butyl-3-(3-triethoxysilylpropyl)imidazolium]
NTf2 was condensed on dehydroxylated SiO2 with a loading of
0.7 mmol·g−1 to produce the corresponding SILP. The prepa-
ration of CoxRh100−x NPs was achieved by adapting the orga-
nometallic approach developed in our group for bimetallic
FeRu@SILP.[11c] This synthesis involved the decomposition of
metal precursors ([Rh(allyl)3] and [Co(cod)(cyclooctadienyl)])
via hydrogenation of the hydrocarbon ligands directly in pres-
ence of the SILP material under mild reaction conditions (total
metal loading = 0.4 mmol·g−1, T = 150 °C, p = 3 bar H2, t = 18 h,
solvent = mesitylene). The syntheses of CoxRh100−x@SILP cata-
lysts were carried out in Fischer-Porter bottles allowing for the
observation of the reaction progress, starting with a brownish-
yellow colored suspension and ending with a black powder and
a clear supernatant after 18 h. This approach provided access to
a wide range of well-defined bimetallic CoxRh100−x@SILP cata-
lysts (See Figure 2a for illustration) with tunable Co:Rh ratios
through the accurate adjustment of the relative quantities of
metal precursors introduced in the reaction mixture (Table S1,
Supporting Information).
2.2. Characterization
BET measurements showed that the CoxRh100−x@SILP cata-
lysts possess a surface area of about 297 m2·g−1. This is sig-
nificantly lower than for the starting dehydroxylated silica
(343 m2·g−1), which is anticipated due to the presence of the
chemisorbed ionic liquid. The theoretical total metal loading
and Co:Rh ratios were confirmed by scanning electron
Small 2021, 17, 2006683
2006683 (2 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Figure 2. a) Illustration of bimetallic CoRh NPs immobilized on SILP (CoxRh100−x@SILP, 100% = 0.4 mmol g−1), b,c) STEM-HAADF image of Co80Rh20@
SILP, d) STEM-HAADF-EDX elemental mapping of Co80Rh20@SILP looking at Co and e) STEM-HAADF-EDX elemental mapping of Co80Rh20@SILP
looking at Rh.
microscopy with energy dispersive X-ray spectroscopy (SEM-
EDX) (for details see Table S1 in the Supporting Informa-
tion). Characterization of the CoxRh100−x@SILP materials by
transmission and scanning transmission electron micros-
copy evidenced the formation of small (2–3 nm) and well-
dispersed NPs on the SILP (example for Co80Rh20@SILP in
Figure 2b, additional details in Table S1 and Figures S1–S6
in the Supporting Information). Elemental mapping using
STEM-HAADF-EDX confirmed the formation of NPs con-
taining both Rh and Co, as exemplified for Co80Rh20@SILP
in Figure 2c–e (See Figure S7 in the Supporting Information
for other examples). Interestingly, the elemental mapping
revealed a higher concentration of sulfur around the NPs
than on the rest of the support, suggesting a strong inter-
action between the ionic liquid (especially the S-containing
anion) and the NPs. Overall, these data confirm the versatility
of the organometallic approach in the controlled formation of
bimetallic NPs on SILP.
To characterize the oxidation state and alloying properties
of the CoxRh100−x@SILP materials, a series of materials with
x = 100, 80, 30, 25, and 0 were characterized using Co and
Rh K-edge X-ray absorption spectroscopy (XAS). Full details
regarding data collection, processing, and analysis are given in
the Supporting Information. The Co K-edge XAS spectra show
minor modulations in the pre-edge (7709–7713 eV) and white-
line (≈7723 eV) regions upon addition of Rh (Figure 3a).
The lack of significant increase in white-line intensity in our
CoxRh100−x@SILP materials relative to the Co° foil is indicative
of the fact that the bulk oxidation state of Co remains zero-
valent throughout the series (Figure S8, Supporting Informa-
tion). The spectral changes anticipated with increased valency
are demonstrated by comparisons of Co100@SILP before and
after O2 exposure, where large increases in the white line
region (≈7723 eV) are accompanied by significant decreases
in pre-edge intensity (7709–7713 eV) (Figure S8, Supporting
Information). Similarly, little modulation of the Rh K-edge is
Figure 3. Comparison of normalized a) Co K-edge and b) Rh K-edge XAS for CoxRh100−x@SILP, x = 100, 80, 30, 25, and 0. The Rh K-edge spectrum of
Co25Rh75@SILP is partially obscured due to its nearly complete overlap with that of Co30Rh70@SILP.
2006683 (3 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH
Small 2021, 17, 2006683

www.advancedsciencenews.com
www.small-journal.com
Figure 4. Comparison of a) Co and b) Rh K-edge EXAFS spectra in R-space (non-phase shift corrected) for CoxRh100−x@SILP, x = 100, 80, 30, 25, and
0. Spectra were generated by Fourier transform of the k-space data from k = 2–11 Å−1. EXAFS data were k3-weighted in the case of Co, and k2-weighted
in the case of Rh.
observed along the CoxRh100−x@SILP series (Figure 3b). Based
on the similar overlap of these spectra with that of the Rh° foil
(Figure S9, Supporting Information), in terms of both edge
position and spectral shape, the bulk oxidation state of Rh
in our CoxRh100−x@SILP materials also remains zero-valent
(Figure S9, Supporting Information). Interestingly, the oxida-
tion state of Rh does not appear oxygen sensitive, as exposure
of the Rh100@SILP to ambient O2 does not result in any major
deviations from the established Rh° foil.
The Fourier-transform (FT) of the Co K-edge extended X-ray
absorption fine structure (EXAFS) region of Co100@SILP
exhibits a single, dominating feature at 2.1 Å, along with several
long-range features at 3.8 and 4.6 Å (Figure 4a). The presence
of a proportionally small amount of Rh to form Co80Rh20@SILP
similarly produces a single main feature, but with dramatically
decreased intensity in the FT (relative to Co100@SILP) and a
decrease in position to 1.9 Å. Furthermore, no clear long-range
features appear.
Increasing Rh content to Co30Rh70@SILP and Co25Rh75@SILP
produces a clear second feature in the FT ≈2.45 Å, indica-
tive of the presence of an additional scatterer. Again, no clear
long-range scatterers are observed for either of these mate-
rials. The decrease in total FT intensity of the Co EXAFS for
all Rh-containing materials relative to Co100@SILP may arise
from the presence of additional scattering pathways that result
in partial phase-cancellation, and/or from increased disorder
within the material. The possibility of increased disorder,
previously observed by Chaudret et al. in the case of Co-rich
CoRh nanocrystals with a polytetrahedral structure,[14] is fur-
ther supported by the disappearance of any clear features in
the FT at higher R (3.5–5 Å), which would otherwise arise
from ordered long-range single scattering or multi-scattering
pathways.
The Rh K-edge FT-EXAFS exhibits similar trends to those
observed for the Co K-edge FT-EXAFS (Figure 4b). The
Rh100@SILP produces a single major feature at 2.4 Å with
a shoulder at 1.9 Å, and two minor long-range features at
4.4 and 4.9 Å. Decreasing Rh content to Co25Rh75@SILP and
Co30Rh70@SILP results in an overall decreased FT intensity
and an increased intensity of the 1.9 Å relative to that at 2.4 Å.
Similar to the Co FT-EXAFS, the long-range features at 4.4 and
4.9 Å are no longer clearly present. Further decreasing Rh
content to Co80Rh20@SILP produces a single clear feature at
2.0 Å, still with a significantly decreased FT intensity relative to
Rh100@SILP.
To further analyze these trends, EXAFS analysis was per-
formed involving simultaneous fitting of the Co and Rh
K-edge spectra for a given material composition (detailed in
the Supporting Information). Over a range of R = 1–3 Å, both
Co100@SILP and Rh100@SILP could be reasonably fit using
single Co-Co and Rh-Rh scattering paths. The Rh K-edge EXAFS
of Co80Rh20@SILP, Co30Rh70@SILP, and Co25Rh75@SILP
could also be adequately modeled using Rh-Rh and Rh-Co
scattering paths. However, additional scatterers were required
to obtain reasonable fits of the Co K-edge EXAFS for all three
bimetallic samples. In Co30Rh70@SILP and Co25Rh75@SILP,
a longer Co-Co scattering path with a distance of ≈3.15 Å was
required in addition to the Co-Rh and shorter Co-Co scattering
paths. In Co80Rh20@SILP, fits performed using solely Co-Co
and Co-Rh scatterers revealed a significant unaccounted for
short scatterer with a maximum at R ≈ 1.7 Å in the non-phase
shifted FT spectrum. Based on the elemental composition of
the SILPs, a variety of Co-X scattering pathways were tested
including X = O, N, C, F, S, and Si. Of these, only a Co-S
scatterer with a distance of 2.16 Å could provide a reasonable
fit. This scatterer may arise from partial degradation of the NTf2
(bis(trifluoromethane)sulfonimide) anion contained in the ionic
liquid phase during synthesis, which could either form Co-S
at the surface of the Co80Rh20@SILP, or independently SILP-
deposited cattierite or linnaeite-like CoxSy clusters. Regardless,
the presence of this possible Co-S scatterer does not appear to
modulate catalytic selectivity, as discussed later.
EXAFS fitting results are summarized in Tables S2 and S3
in the Supporting Information. The Co-Co distance does
not appear significantly modulated when moving between
Co100@SILP and Co80Rh20@SILP (≈2.48 Å), but expands to
≈2.53 Å in Co30Rh70@SILP and Co25Rh75@SILP. Similarly, the
Co-Rh/Rh-Co distance increases from 2.55 Å in Co80Rh20@SILP
to ≈2.60 Å in Co30Rh70@SILP and Co25Rh75@SILP. This is gen-
erally consistent with an expanding lattice, and is supported by
our X-ray powder diffraction measurements (Figure S10, Sup-
porting Information). The fit Rh-Rh distances range between
2.65 and 2.69 Å across the series, appearing slightly contracted
in the Co30Rh70@SILP and Co25Rh75@SILP samples. These
Small 2021, 17, 2006683
2006683 (4 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Figure 5. Hydrogenation of benzylideneacetone (1) using a series of CoxRh100−x@SILP catalysts with different Co:Rh ratios. Reaction conditions:
CoxRh100−x@SILP (20.0 mg, 0.008 mmol [m]), H2 (50 bar), substrate (0.80 mmol, 100 eq.), 150 °C, 16 h, n-heptane (0.5 mL). For all ratios, complete
conversion (>99%) of 1 to a combination 1a/1b/1c/1d was observed. Yields were determined by GC-FID using tetradecane as internal standard.
fit distances are again consistent with an average zero-valent
state of both Co and Rh, and their expansion as a function of
increasing Rh content further supports alloying. Atomic dis-
tributions were calculated to determine the extent of alloying
within the bimetallic nanoparticles. This involves calcula-
tion of the parameter J, described by Hwang, et al.,[19] which
relies on the determined degeneracy of the Co-Co, Co-Rh, and
Rh-Rh scattering paths. Additionally, we have formulated a
disorder-corrected atomic distribution parameter, S, by modi-
fying the formulation of J to include the fit Debye-Waller factors
(σ2). Fitting results generally support a fully alloyed bimetallic
structure in these materials based on the determined values of
J and S (Table S2, Supporting Information).
Starting with monometallic Rh100@SILP, 1 was completely
hydrogenated to give 1d in quantitative yield, which corre-
sponds to the expected reactivity for pure Rh NPs. No change
in product distribution was observed under these conditions
with addition of Co up to Co25Rh75. However, “diluting” Rh con-
tent further to Co30Rh70 resulted in a sharp switch in selectivity,
leading to production of the unsaturated alcohol 1c in high
yield (92%). Excellent selectivity towards the formation of 1c
was conserved across a broad range of Co:Rh ratios, reaching
a maximum (>99%) at Co20Rh80. Further lowering Rh content
led to a progressive decrease in total hydrogenation activity,
resulting in mixtures of 1a (65%) and 1c (35%) for the mono-
metallic Co100@SILP catalyst.
To gain further insight into the catalysts reactivity, the

initial rate for C O hydrogenation of Co80Rh20@SILP,
2.3. Catalysis
Co30Rh70@SILP and Co25Rh75@SILP were determined by
recording time profiles of product formation (Figure 6). To do
so, reduced temperatures of 100 °C were used to slow down
The catalytic properties of the CoxRh100−x@SILP materials
were investigated using first the hydrogenation of benzylide-
neacetone (1) as a model reaction. Substrate 1 possesses an
aromatic ring, a conjugated double bond, and a non-benzylic
ketone, giving the opportunity to probe the reactivity of
CoxRh100−x@SILP for a whole range of prototypical functional
groups. Hydrogenation of 1 can lead to the ketones 4-phenyl-
2-butanone (1a) and 4-cyclohexyl-2-butanone (1b) as well as to
the alcohols 4-phenyl-2-butanol (1c) and 4-cyclohexyl-2-butanol
(1d) (Figure 5; Table S4, Supporting Information). These prod-
ucts are connected through a network of sequential and parallel
hydrogenation pathways. Figure 5 shows the product yields as
a function of the Co:Rh ratio in CoxRh100−x@SILP obtained
under an identical set of reaction conditions (T = 150 °C, p(H2)
= 50 bar, t = 16 h, solvent = heptane).

the reaction. Under these conditions, the C C hydrogena-
tion was extremely fast (completed in the first few minutes)
and the three catalysts considered led only to the observable
formation of 1c as a product from 1a during the first 4 h,
allowing for a direct comparison. The activity for C O hydro-
genation increases with increasing Co content in the order

kCO(Co25Rh75) < kCO(Co30Rh70) << kCO(Co80Rh20). Further
increasing the Co content led to a drop in C O hydrogenation

activity as evidenced by formation of mixtures of 1a and 1c
already for Co85Rh15@SILP (Figure 5). Interestingly, the incor-
poration of Co in Rh NPs also appears to modify the sequence
of hydrogenation of the different functional groups (Table S5,
Supporting Information). Rh100@SILP exhibits a preference


for hydrogenating C C > aromatic > C O, and thus displays
2006683 (5 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH
Small 2021, 17, 2006683

www.advancedsciencenews.com
www.small-journal.com
Supporting Information). Instead, material prepared using
a 80:20 ratio of the Co and Rh precursor complexes dis-
played activity typical for catalysts with very high Rh content
(Co25Rh75 and above), leading to deep hydrogenation and pro-
ducing 1d as main product (Table S6, Supporting Informa-
tion). This activity is indicative of the failure of the controlled
formation of Co80Rh20 NPs on the SiO2 (Table S6, Supporting
Information). In contrast, no evidence for agglomeration
or segregation was obtained by TEM and SEM-EDX anal-
ysis of the CoxRh100−x@SILP materials even after catalysis
(Figures S13–S14 and Tables S7–S8, Supporting Information).
These results highlight the importance of the ionic liquid
layer in the SILP for the controlled preparation and structural
stability of these alloy-type bimetallic particles.
The in-depth characterization of the pristine NPs and cata-
lytic studies clearly indicate that the bimetallic alloy-type NPs
with both metals in the formal oxidation state zero are the cat-
alytically active components across the whole range of Co:Rh
compositions. A possible explanation for the pronounced
activity and selectivity effects might be a dynamic modifica-
tion of the CoRh NPs surface composition under reaction con-
ditions due to the attraction of the oxophilic Co atoms by the
polar substrate, leading to a Co enrichment in the CoRh sur-
face of the alloys. Similar examples of dynamic surface enrich-
ment with 3d metals have been reported for related bimetallic
systems including, for example, Cu-Pd and Co-Pd alloys.[20]
For Co contents below 25%, the surface of the NPs can still be
rich enough in Rh atoms to allow the arrangement of 3–4 adja-
cent Rh atoms necessary to hydrogenate 6-membered aromatic
rings.[21] However, once the Co content is increased above 30%,
the surface of the alloyed CoRh NPs becomes too Co rich when
under dynamic conditions, thus preventing the necessary Rh
ensembles and hence leaving the aromatic ring intact under

Figure6. DeterminationofinitialC OhydrogenationratesforCo80Rh20
@
SILP, Co30Rh70@SILP, and Co25Rh75@SILP. Reaction conditions: Catalyst
(20.0 mg, 0.008 mmol [m]), H2 (50 bar), benzylideneacetone (117 mg,
0.80 mmol, 100 eq.), 100 °C, n-heptane (0.5 mL). Yield determined by
GC-FID using tetradecane as internal standard.
a product pathway of 1 → 1a → 1b → 1d. Meanwhile, the


presence of Co changes the order to C C > C O > aromatic,
resulting in a 1 → 1a → 1c → 1d product pathway for all three
CoxRh100−x@SILP catalysts, albeit with almost complete shut
down of aromatic ring hydrogenation for Co contents above
Co25Rh75. This last point was confirmed through the deter-
mination of the initial rate for aromatic ring hydrogenation
of Co80Rh20@SILP, Co30Rh70@SILP, and Co25Rh75@SILP.
Time profiles were recorded for the conversion of 1c to 1d
under standard conditions, evidencing no or extremely low
activity for Co80Rh20@SILP (karomatic = 0 h−1) and Co30Rh70@
SILP (karomatic = 0.009 h−1) in contrast to Co25Rh75@SILP
(karomatic = 0.104 h−1) (Figure S11, Supporting Information).
As aforementioned, XAS analysis suggested the presence of
a Co-S scattering path in Co80Rh20@SILP. The further lack of
any significant Rh-S scattering indicated that such an inter-
action would arise from either coordination of sulfur with
Co on the surface of the NP, or from independently formed
CoxSy-type clusters deposited on the SILP. To test whether the
presence of a sulfur-containing anion impacts catalytic prop-
erties, Co80Rh20@SILP NPs were prepared additionally using
a sulfur-free SILP (SILP-BPh4) as well as an unmodified SiO2
support. The resulting Co80Rh20@SILP-BPh4 catalyst displays
reactivity largely identical to the standard Co80Rh20@SILP
(Table S6, Supporting Information). This indicates that even
if some form of Co-S interaction occurs when using the NTf2
anion, as suggested by XAS, this does not significantly influ-
ence the reactivity of the bimetallic NPs. More importantly,
however, using unmodified SiO2 as a support rather than
the IL-grafted SILP – produced very different results. Specif-
ically, use of unmodified silica did not allow for generation
of homogeneously dispersed bimetallic CoRh NPs, as evi-
denced by the big aggregates observed in TEM (Figure S12,

hydrogenation conditions. The enhancement of the C O
hydrogenation rate upon further increase of the Co content

can be rationalized by a more efficient activation of the C O
bond on the increasingly oxophilic CoRh surface together with
an increasingly polarized activation of H2.4 At very high Co:Rh
ratios, the noble metal is no longer available to enhance the H2
activation and the reactivity of pure Co prevails. The rigorous
investigation of this hypothesis by characterization of the cata-
lysts under dynamic conditions using operando X-Ray spectros-
copies is underway and will be the subject of future reports.
Based on this rationale, the generality of the selectivity switch
between Co30Rh70@SILP and Co25Rh75@SILP was investigated
by expanding the substrate scope to a variety of probe molecules
also comprising a polar functional group and a C6 aromatic
ring, i.e., bicyclic heteroaromatics. Achieving the selective par-
tial or deep hydrogenation of such compounds in a control-
lable manner is very attractive because of the widespread use
of these motifs as building blocks in the production of various
fine chemicals and pharmaceuticals.[22] Benzofurans, indoles,
and quinolines were chosen as representatives of this group and
the results are summarized in Table 1. Hydrogenation using
Co30Rh70@SILP under standard conditions (50 bar H2, 16 h
reaction time) without further optimization resulted in forma-
tion of the partially saturated products (dihydrobenzofurans,
dihydroindoles, and tetrahydroquinolines, respectively) in good
to excellent yields. The hydrogenation of the C6 aromatic ring
Small 2021, 17, 2006683
2006683 (6 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Table 1. Selective hydrogenation of bicyclic heteroaromatics using CoxRh100−x@SILP.
Substratea)
Products
Co30Rh70
Sa/b [%]
Co25Rh75
T [°C]
X [%]
>99
X [%]
Sa/b [%]
150
>99/0 (89)
>99
>99
0/>99 (90)
180
180
60
98/2
0/>99
>99
97/3
>99
>99
35/47b
150
180
87
26
>99/0 (71)
0/>99 (85)
23/77
97/3
>99
0/>99
180
70
86
35/42c
150
150
>99
>99
>99/0 (93)
>99
0/>99 (88)
83/17
98
0/>99
a)Reaction conditions: Catalyst (20.0 mg, 0.008 mmol [m]), H2 (50 bar), substrate (0.20 mmol, 25 eq.), decalin (0.5 mL), 16 h. Conversion and selectivity determined by
GC-FID using tetradecane as internal standard; b)2b (18%); c)5b (23%). X = conversion, S = selectivity, (isolated yields).
was suppressed with very high selectivity, substrate 6 being the
only exception. In this case, the methyl substituent adjacent to
the NH seems to reduce the selectivity in general, as evidenced
by the somewhat lower selectivity also found for quinaldine 9.
In contrast, use of Co25Rh75@SILP led to complete
hydrogenation of the substrates considered, providing the
reduced by the presence of electron-donating methoxy groups,
specifically for substrates 4 and 7. Interestingly, partial cleavage
of the methoxy group was observed for these substrates, sug-
gesting that the Co25Rh75@SILP material also exhibits signifi-
cant hydrogenolysis activity.
These results are consistent with those described above
for benzylideneacetone, and confirm the existence of a sharp
switch in the catalysts’ ability to hydrogenate 6-membered aro-
matic rings depending on the Co:Rh ratio. Importantly, this
corresponding
octahydrobenzofurans,
octahydroindoles
and decahydroquinolines in particularly excellent yields for
unsubstituted C6 rings. The degree of deep hydrogenation was
2006683 (7 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH
Small 2021, 17, 2006683

www.advancedsciencenews.com
www.small-journal.com
stored under argon. Benzylideneacetone was sublimated prior to use.
Mesitylene, toluene, heptane, DCM and acetone were distilled, dried over
molecular sieves (4 Å) and stored under argon. Decalin was purchased
commercially (from ACROS, 99% anhydrous), flushed with argon and
stored over molecular sieves (4 Å). SiO2 (Merck Grade 10184, pore size
100, 63–200 mm) was dehydroxylated in vacuo at 500 °C for 16 h prior
to use. [Co(cod)(cyclooctadienyl)] was purchased from Nanomeps and
[Rh(allyl)3] was synthesized according to a literature procedure.S1 High-
pressure experiments were performed using in-house engineered 10 and
20 mL stainless steel finger autoclaves equipped with a glass insert.
Analytics: Gas chromatography (GC) was performed on a Thermo
Scientific Chromatograph Trace GC Ultra equipped with a CP-Wax 52 CB
column from Agilent and Rtx-1 from Restek. Conventional bright-field
TEM images were taken using a JEOL JEM 1400 operated at 120 kV.
Scanning transmission electron microscopy with energy dispersive X-ray
spectroscopy (STEM-HAADF/EDS) was performed on a JEOL JEM-
ARM200F Cold FEG operated at 200 kV, and on a Hitachi HF-2000 Cold
FEG operated at 200 kV. The NP size and distribution was determined
from the measurement of >150 spherical particles chosen in arbitrary
areas of enlarged micrographs. Scanning electron microscopy with
energy dispersive X-ray spectroscopy (SEM/EDS) was performed on a
JEOL JSM 7800F operated at 10 kV with a thermally assisted Schottky
electron gun and equipped with a Bruker XFlash 6|60 detector (silicon
drift detector technology), and on a Hitachi S-3500N using a 10 mm²
Si(Li) detector (Oxford Instruments). Samples for SEM/EDS analysis
were prepared by pressing powder of NPs@SILPs on carbon tape.
BET was measured using a QuadraSorb station (7.01). Powder X-ray
diffraction was performed using a X’Pert PRO powder diffractometer
(PANalytical) equipped with a copper X-ray source. The data collection
was performed with the aid of X’Celerator detectors. A detailed section
on the characterization by X-ray absorption spectroscopy is available in
the Supporting Information.
Synthesis of CoxRh100−x@SILP (Example of Co80Rh20@SILP): In a typical
experiment [Rh(allyl)3] (18.1 mg, 0.08 mmol) was dissolved in mesitylene
(4 mL) and combined with a solution of [Co(cod)(cyclooctadienyl)]
(88.4 mg, 0.32 mmol) in mesitylene (4 mL) in a Fischer-Porter bottle
(70 mL). While stirring the solution in a glovebox, the SILP (1.00 g) was
added and the suspension was stirred for 30 min at room temperature
(RT). The Fischer-Porter bottle was placed into a preheated oil bath
(150 °C), stirred for 15 min under argon and subsequently pressurized
with H2 (3 bar). The reaction mixture was then stirred for 18 h, yielding a
black powder (the catalyst) and a clear supernatant. After cooling down,
the mesitylene was decanted and the catalyst washed with toluene (dried
and degassed, 5 mL) before drying in vacuo at RT for 1 h.
Catalytic Reactions: In a typical catalytic experiment, the catalyst
(20 mg, 0.008 mmol metal) and the substrate (0.2–0.8 mmol,
25–100 eq.) were weighed into a glass insert and the solvent (heptane for
benzylideneacetone, decalin for heteroaromatics) (0.5 mL) was added,
as well as a stir bar. The inlet was placed in a high-pressure autoclave
and pressurized with H2 (50 bar). The reaction mixture was stirred at
150 °C for 16 h in a pre-heated aluminum heating cone. At the end of
the reaction, the autoclave was cooled in a water bath, carefully vented
and the reaction mixture was filtered into a GC vial for characterization,
using tetradecane as an internal standard. Conversion and selectivity
were determined by GC-FID.
selectivity switch occurred systematically between the 30:70 and
25:75 ratios for all substrates considered, suggesting this is a
general characteristic of CoxRh100−x@SILP catalysts under these
conditions.
3. Conclusions
In conclusion, we report a synthetic organometallic approach
that is suitable for the preparation of small (2–3 nm) and
well-dispersed bimetallic CoRh nanoparticles on an imida-
zolium-based SILP support. Investigation of the resulting
CoxRh100−x@SILP catalysts using STEM-HAADF-EDX and XAS
confirmed the formation of homogeneous Co-Rh alloys in a
zero-valent state over the entire composition range. The mole-
cular approach to assemble the materials allows precise control
over the Co:Rh ratio which in turn has a great influence on the
catalytic properties of CoxRh100−x@SILP materials for hydro-
genation of substituted aromatics. In particular, while NPs with
high Rh content (>75 at%) possess activity typical of Rh metal
and can fully hydrogenate aromatic substrates, increasing Co
content (30–80 at%) leads to a selectivity switch with nearly
complete suppression of the arene hydrogenation accompa-

nied by an enhancement of the C O hydrogenation activity.
The materials Co30Rh70@SILP and Co25Rh75@SILP were found
highly active, selective and stable for the hydrogenation of ben-
zylideneacetone and various bicyclic heteroaromatics to the
corresponding partially and fully saturated products, respec-
tively. As electron microscopy and XAS do not evidence major
morphological, structural, or electronic differences between
Co30Rh70@SILP and Co25Rh75@SILP catalysts, the sharp selec-
tivity switch observed is tentatively associated with the inter-
ruption of the arrangement of Rh atoms necessary for aromatic
ring hydrogenation above a specific Co content under dynamic
catalytic conditions.
In general, the results presented here substantiate the
opportunities offered by controlled “dilution” of noble metal
nanoparticles with the respective first row base-metal congener.
Provided that suitable support materials allow for the formation
of defined and stable alloy-type materials, synergistic effects
can result in catalytic properties that are not provided by any of
the two metals individually and may be difficult—if not impos-
sible—to be achieved with traditional supported metal catalysts
in general. The combination of molecularly modified and in
particular SILP-type support materials with the organometallic
approach to nanoparticle synthesis provides an excellent plat-
form to explore this opportunity in a wide parameter space.
4. Experimental Section
Supporting Information
Safety Warning: High-pressure experiments with compressed H2 (g)
must be carried out only with appropriate equipment and under rigorous
safety precautions.
Supporting Information is available from the Wiley Online Library or
from the author.
General: If not otherwise stated, the synthesis of ionic liquids
(ILs), supported ionic liquid phases (SILPs) and nanoparticles (NPs)
supported on SILPs (NPs@SILP) were carried out under argon using
standard Schlenk techniques or inside a glovebox. After synthesis, ILs,
SILPs and NPs@SILPs were stored under argon. The substrates used
in the catalytic study were purchased from commercial sources and
Acknowledgements
The authors acknowledge financial support by the Max Planck Society
and by the Deutsche Forschungsgemeinschaft (DFG, German Research
Small 2021, 17, 2006683
2006683 (8 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH

www.advancedsciencenews.com
www.small-journal.com
Foundation) under Germany’s Excellence Strategy – Exzellenzcluster
2186 „The Fuel Science Center“ ID: 390919832. The authors thank
ERC Advanced Grant (MONACAT 2015-694159). Fruitful discussion
on the dynamics of NP catalysts with Prof. Robert Schlögl is gratefully
acknowledged by W.L. C.V.S. would like to thank the IMPRS-RECHARGE
for funding. Use of the Stanford Synchrotron Radiation Lightsource,
SLAC National Accelerator Laboratory, was supported by the U.S.
Department of Energy, Office of Science, Office of Basic Energy Sciences
under Contract No. DE-AC02-76SF00515. For experiments performed
at beamline 9-3 of SSRL, the authors would like to thank Dr. Matthew
Latimer for assistance. The authors acknowledge the Paul Scherrer
Institut, Villigen, Switzerland for provision of synchrotron radiation
beamtime at beamline SuperXAS of the SLS and would like to thank
Dr. Maarten Nachtegaal for assistance. The authors acknowledge DESY
(Hamburg, Germany), a member of the Helmholtz Association HGF,
for the provision of experimental facilities. For parts of this research
carried out at PETRA III, the authors would like to thank Dr. Edmund
Welter for assistance in using beamline P65. The research leading to
these results has received funding from the European Community’s
Seventh Framework Programme (FP7/2007-2013) under grant
agreement no. 312284. Furthermore, authors would like to thank Josef
Vaeßen (ITMC, RWTH Aachen) for BET absorption measurements,
Hannelore Eschmann and Elke Biener (ITMC, RWTH Aachen) as
well as Annika Gurowski, Alina Jakubowski, and Justus Werkmeister
(MPI-CEC) for GC and GC-MS measurements, Norbert Pfänder
(MPI-CEC), Adrian Schülter, Silvia Palm (MPI-Kofo), Simon Cayez
(LPCNO) and Maria Teresa Hungria (Centre de MicroCaractérisation
Raimond Castaing) for help concerning TEM, STEM-HAADF-EDX
and SEM-EDX analyses, and Claudia Weidenthaler for XRD
measurements.
[4] a) D. Wang, Y. Li, Adv. Mater. 2011, 23, 1044; b) M. Sankar,
N. Dimitratos, P. J. Miedziak, P. P. Wells, C. J. Kiely, G. J. Hutchings,
Chem. Soc. Rev. 2012, 41, 8099; c) J. Lee, Y. T. Kim, G. W. Huber,
Green Chem. 2014, 16, 708; d) D. M. Alonso, S. G. Wettstein,
J. A. Dumesic, Chem. Soc. Rev. 2012, 41, 8075.
[5] a) G. A. Somorjai, Introduction to Surface Chemistry and Catalysis,
Wiley, New York 1994, pp. 500–512; b) G. W. Huber, J. W. Shabaker,
J. A. Dumesic, Science 2003, 300, 2075.
[6] D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu,
A. F. Carley, A. A. Herzing, M. Watanabe, C. J. Kiely, D. W. Knight,
G. J. Hutchings, Science 2006, 311, 362.
[7] F. Maroun, F. Ozanam, O. M. Magnussen, R. J. Behm, Science 2001,
293, 1811.
[8] a) L. Zhang, M. Zhou, A. Wang, T. Zhang, Chem. Rev. 2020, 120, 683;
b) A. Sánchez, M. Fang, A. Ahmed, R. A. Sánchez-Delgado, Appl. Catal.,
A 2014, 477, 117; c) J. C. Matsubu, S. Zhang, L. DeRita, N. S. Marinkovic,
J. G. Chen, G. W. Graham, X. Pan, P. Christopher, Nat. Chem. 2017,
9, 120; d) A. Karakulina, A. Gopakumar, Z. Fei, P. J. Dyson, Catal. Sci.
Technol. 2018, 8, 5019; e) C. Deraedt, R. Ye, W. T. Ralston, F. D. Toste,
G. A. Somorjai, J. Am. Chem. Soc. 2017, 139, 18084.
[9] a) A. Wong, Q. Liu, S. Griffin, A. Nicholls, J. R. Regalbuto, Science
2017, 358, 1427; b) K. Ding, D. A. Cullen, L. Zhang, Z. Cao, A. D. Roy,
I. N. Ivanov, D. Cao, Science 2018, 362, 560; c) S. H. Hakim, C. Sener,
A. C. Alba-Rubio, T. M. Gostanian, B. J. O’Neill, F. H. Ribeiro,
J. T. Miller, J. A. Dumesic, J. Catal. 2015, 238, 75.
[10] a) P. Migowski, K. L. Luska, W. Leitner, In Nanocatalysis in Ionic
Liquids (Ed: M. H. G. Prechtl), Wiley VCH, Weinheim 2016;
b) R. Fehrmann, A. Riisager, M. Haumann, Supported Ionic Liquids:
Fundamentals and Applications, Wiley-VCH Verlag GmbH, Weinheim
2014.
Open access funding enabled and organized by Projekt DEAL.
[11] a) M. A. Gelesky, S. S. X. Chiaro, F. A. Pavan, J. H. Z. dos Santos,
J. Dupont, Dalton Trans. 2007, 5549; b) M. Ruta, G. Laurenczy,
P. J. Dyson, K.-L. Minsker, J. Phys. Chem. 2008, 112, 17814;
c) K. L. Luska, A. Bordet, S. Tricard, I. Sinev, W. Grünert,
B. Chaudret, W. Leitner, ACS Catal. 2016, 6, 3719; d) S. El Sayed,
A. Bordet, C. Weidenthaler, W. Hetaba, K. L. Luska, W. Leitner,
ACS Catal. 2020, 10, 2124; e) A. Bordet, C. Welsh, P. Lience,
K. L. Luska, W. Leitner, ACS Catal. 2020, 10, 13904.
[12] a) K. L. Luska, J. Julis, E. Stavitski, D. N. Zakharov, A. Adams,
W. Leitner, Chem. Sci. 2014, 5, 4895; b) K. L. Luska, P. Migowski,
S. El Sayed, W. Leitner, Angew. Chem., Int. Ed. 2015, 54, 15750;
c) L. Offner-Marko, A. Bordet, G. Moos, S. Tricard, S. Rengshausen,
B. Chaudret, K. L. Luska, W. Leitner, Angew. Chem., Int. Ed. 2018,
57, 12721; d) G. Moos, M. Emondts, A. Bordet, W. Leitner, Angew.
Chem., Int. Ed. 2020, 59, 11977; e) L. Goclik, L. Offner-Marko,
A. Bordet, W. Leitner, Chem. Commun. 2020, 56, 9509.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
bimetallic nanoparticles, catalytic hydrogenation, cobalt-rhodium, SILP,
synergistic
Received: October 27, 2020
Revised: December 2, 2020
Published online: December 21, 2020
[13] S. Rengshausen, F. Etscheidt, J. Großkurth, K. L. Luska, A. Bordet,
W. Leitner, Synlett 2019, 30, 405.
[1] K. D. Gilroy, A. Ruditskiy, H.-C. Peng, D. Qin, Y. Xia, Chem. Rev.
2016, 116, 10414.
[14] D. Zitoun, C. Amiens, B. Chaudret, M.-C. Fromen, P. Lecante,
M.-J. Casanove, M. Respaud, J. Phys. Chem. B 2003, 107, 6997.
[15] H. F. J. Van’t Blik, D. C. Koningsberger, R. Prins, J. Catal. 1986, 97,
210.
[16] J. H. Park, Y. K. Chung, Dalton Trans. 2008, 18, 2369.
[17] I. Choi, H. Chung, J. W. Park, Y. K. Chung, Org. Lett. 2016, 18, 5508.
[18] a) F. Benseradj, F. Sadi, M. Chater, C. R. Chim. 2004, 7, 669;
b) X. Zheng, Y. Lin, H. Pan, L. Wu, W. Zhang, L. Cao, J. Zhang,
L. Zheng, T. Yao, Nano Res. 2018, 11, 2357; c) H. Kusama, K. Okabe,
K. Sayama, H. Arakawa, Appl. Organomet. Chem. 2000, 14, 836.
[19] B. J. Hwang, L. S. Sarma, J. M. Chen, C. H. Chen, S. C. Shin,
Q. R. Wang, D. G. Liu, J. F. Lee, M. T. Tang, J. Am. Chem. Soc. 2005,
127, 11140.
[2] a) G. Mattei, P. Mazzoldi, M.L. Post, D. Buso, M. Guglielmi,
A. Martucci, Adv. Mater. 2007, 19, 561; b) D. A. Garfinkel, G. Pakeltis,
N. Tang, I. N. Ivanov, J. D. Fowlkes, D. A. Gilbert, P. D. Rack,
ACS Omega 2020, 5, 19285; c) A. Y. Soloveva, N. K. Eremenko,
I. I. Obraztsova, A. N. Eremenko, S. P. Gubin, Russ. J. Inorg. Chem.
2018, 63, 444; d) S. Thota, Y. Wang, J. Zhao, Mater. Chem. Front.
2018, 2, 1074.
[3] a) S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science 2000,
287, 1989; b) I.-C. Chiang, D.-H. Chen, Adv. Funct. Mater. 2007,
8, 1311; c) K. I. M. da Silva, F. Bernardi, G. Abarca, D. L. Baptista,
M. J. L. Santos, L. F. Barquin, J. Dupont, I. del Pedro, Phys. Chem.
Chem. Phys. 2018, 20, 10247; d) V. Amendola, S. Scaramuzza,
L. Litti, M. Meneghetti, G. Zuccolotto, A. Rosato, E. Nicolato,
P. Marzola, G. Fracasso, C. Anselmi, M. Pinto, M. Colombatti,
Small 2014, 12, 2476.
[20] a) J. S. Bradley, E. W. Hill, B. Chaudret, A. Duteil, Langmuir
1995, 11,
693; b) M. Cokoja, H. Parala, A. Birkner, R. A. Fischer, O. Margeat,
D. Ciuculescu, C. Amiens, B. Chaudret, A. Falqui, P. Lecante, Eur. J.
Inorg. Chem. 2010, 11, 1599; c) C. H. Wu, C. Liu, D. Su, H. L. Xin,
2006683 (9 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH
Small 2021, 17, 2006683

www.advancedsciencenews.com
www.small-journal.com
H.-T. Fang, B. Eren, S. Zhang, C. B. Murray, M. B. Salmeron, Nat.
Catal. 2019, 2, 78; d) F. Tao, M. E. Grass, Y. Zhang, D. R. Butcher,
J. R. Renzas, Z. Liu, J. Y. Chung, B. S. Mun, M. Salmeron,
G. A. Somorjai, Science 2008, 322, 932.
[22] a) D.-S. Wang, Q.-A. Chen, S.-M. Lu, Y.-G. Zhou, Chem. Rev. 2012,
112, 2557; b) A. Karakulina, A. Gopakumar, I. Akçok, B. L. Roulier,
T. LaGrange, S. A. Katsyuba, S. Das, P. J. Dyson, Angew. Chem.,
Int. Ed. 2016, 55, 292; c) R. Adam, J. R. Cabrero-Antonino,
A. Spannenberg, K. Junge, R. Jackstell, M. Beller, Angew. Chem., Int.
Ed. 2017, 56, 3216.
[21] C. Fan, Y.-A. Zhu, X.-G. Zhou, Z.-P. Liu, Catal. Today 2011, 160,
234.
Small 2021, 17, 2006683
2006683 (10 of 10)
© 2020 The Authors. Small published by Wiley-VCH GmbH