Bimetallic nanosized solids with acid and redox properties for catalytic activation of C-C and C-H bonds

Article abstract of DOI:10.1039/C6SC03335K

A new approach is presented to form self-supported bimetallic nanosized solids with acid and redox catalytic properties. They are water-, air- and H₂-stable, and are able to activate demanding C-C and C-H reactions. A detailed mechanistic study on the formation of the Ag-Fe bimetallic system shows that a rapid redox-coupled sequence between Ag⁺, O₂ (air) and Fe²⁺ occurs, giving monodisperse Ag nanoparticles supported by O-bridged diatomic Fe³⁺ triflimides. The system can be expanded to Ag nanoparticles embedded within a matrix of Cu²⁺, Bi³⁺ and Yb³⁺ triflimide.

Full text of DOI:10.1039/C6SC03335K

View Article Online
View Journal
Chemical
Science
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. Corma, A.
Leyva-Perez, J. R. Cabrero-Antonino, M. Tejeda, M. Quesada and A. VIDAL MOYA, Chem. Sci., 2016, DOI:
10.1039/C6SC03335K.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/chemicalscience

Please do not adjust margins
Chemical Science
Page 1 of 8
View Article Online
DOI: 10.1039/C6SC03335K
Chemical Science
EDGE ARTICLE
Bimetallic nanosized solids with acid and redox properties for
catalytic activation of C–C and C–H bonds
Jose R. Cabrero–Antonino,a María Tejeda–Serrano,a Manuel Quesada,a Jose A. Vidal–Moya,a
Antonio Leyva–Pérez,a* and Avelino Corma.a*
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
A new approach is presented to form self–supported bimetallic nanosized solids with acid and redox catalytic properties.
They are water–, air– and H2–stable, and are able to activate demanding C–C and C–H reactions. A detailed mechanistic
study on the formation of the Ag–Fe bimetallic system shows that a rapid redox–coupled sequence between Ag+, O2 (air)
and Fe2+ occurs, giving monodisperse Ag nanoparticles supported by O–bridged diatomic Fe3+ triflimides. The system can
www.rsc.org/
be expanded to Ag nanoparticles embedded within
a
matrix of Cu2+
,
Bi3+ and Yb3+ triflimide.
1. Introduction.
mechanism follows a simple redox–coupled sequence to
furnish self–supported solids that behave as efficient acid,
redox and acid/redox heterogeneous catalysts for different C–
H and C–C activation reactions.
Metal triflimides Mn+Tfn [M: metal cation, Tf: N(SO2CF3)2] are
strong Lewis acids due to the high delocalization of the
negative charge in the triflimide anion, which confers high
mobility to associated cations and protons. Contrary to other
-
soft acids with highly delocalized anions (i.e. 2. Results and Discussion.
tetrafluoroborates, hexaflurometalates), triflimide salts are
2.1.
Synthesis, characterization and structure of the solid
relatively stable and safe–to–handle1 and, for instance, LiNTf2
is commercially employed as an electrolyte in batteries. These
properties make triflimide salts powerful Lewis acid catalysts
to activate C–C bonds and, if the redox properties of the metal
cation are tuned with ligands, also very active catalysts to
oxidize reluctant C–H bonds.2 However, both acid/redox
functions are typically self–excluding and their concomitant
use in a single reaction is hampered. Besides that, triflimide
salts are expensive and difficult to recover in solution, which
translates their use basically to a laboratory scale. Thus, it is
clear that the synthesis of a recoverable solid triflimide
catalyst, insoluble in common solvents, and with its acid/redox
sites operative at the same time, would solve those problems.
However, to our knowledge, any solid triflimide has not been
reported yet.3,4 Here we present a new preparation concept
that allows the formation of a family of triflimide solids with
very strong acidity and readily available redox metal sites. The
Fe2O(NTf2)5@AgNPs.
Figure 1 shows the preparation of the solid material. The
procedure consists in mixing Fe3+ triflimide [Fe(NTf2)3] with Ag
triflimide (AgNTf2, 0.1–1.0 equivalent) in 1,4–dioxane at room
temperature, and then adding one equivalent of thiophenol
(PhSH) at once. A soft yellow solid precipitates in up to 90%
yield in 2 gram scale for Ag:Fe molar ratio= 0.5
.
material consists of metal (Fe3+, Cu2+, Yb3+ or Bi3+)
ꢀ–oxide
triflimides on Ag nanoparticles (Ag NPs) concomitantly formed
under ambient conditions, after adding thiophenol to a
solution of the corresponding metal triflimide and Ag+. The
Figure 1. Synthesis of Fe2O(NTf2)5@AgNPs. Reaction conditions: Fe
(NTf2)3 (6.16 mmol), AgNTf2 (0.1-1.0 equivalent), thiophenol (6.16
mmol), 1,4-dioxane (80 mL), room temperature, <1 min-1 h, 28-90%
yield. See details in ESI. A photograph of the yellow solid and a
model (not scaled) are also provided.
a.Instituto de Tecnología Química. Universitat Politècnica de València–Consejo
Superior de Investigaciones Científicas. Avda. de los Naranjos s/n, 46022,
Valencia, Spain.
*
+34963877800; Fax: +349638 77809.
Electronic Supplementary Information (ESI) available: General procedures,
additional Figures and Tables, compound characterization and NMR copies.. See
DOI: 10.1039/x0xx00000x
Corresponding author: acorma@itq.upv.es, anleyva@itq.upv.es. Phone:
Analysis by inductively coupled plasma–atomic emission
spectroscopy (ICP–AES) gives 5.8 wt% of Fe and 6.2 wt% of Ag for
the solid prepared with an Ag:Fe molar ratio= 0.5, which accounts
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 2 of 8
View Article Online
DOI: 10.1039/C6SC03335K
ARTICLE
Journal Name
for the amounts of Fe and Ag atoms initially added. Elemental the individual absorption spectra of independently–prepared
analysis (CHNS) of the solid gives a sulfur–to–nitrogen ratio of 2:1, Fe(NTf2)3 and AgNPs (see SI).
that corresponds exactly to triflimide anions, and that in principle
Figure 3 (top–right) shows the Fourier transform–infrared (FT–
discards the presence of thiophenol in the structure. Analysis of the
IR) spectrum of the material with the presence of two different
organic molecules present in the solid after dissolution in HNTf2,
triflimide peaks (see Table S1 in SI for a complete list of bands and
extraction with diethyl ether and quantification by gas
discussion), in accordance with the TG analysis. The triflimide
chromatography–mass spectrometry (GC–MS) with n–dodecane as
anions are coordinated to Fe cations either through the oxygen or
nitrogen atoms of the triflimide groups.5 The IR spectrum also
an external standard, confirms the lack of thiols and shows the
presence of ~15 wt% of 1,4–dioxane. Thermogravimmetric (TG)
unveils that a ꢀ–hydroxo or ꢀ–oxo bridge between two Fe atoms is
present.6 These results indicate that the solid is formed by oxo–
analysis (Figure S1 in SI) confirms the amount of 1,4–dioxane
present (16 wt%, loss at ~100 ºC) and also shows that the material
bridged Fe2O(NTf2)5 species.
contains 4 wt% of water and two types of triflimide anions, that
desorb at ~200 ºC (10 wt%) and 300 ⁰C (47 wt%). With all these
RD–UVvis
FT–IR
data in hand, we can calculate the initial formula
Ag0.5Fe(NTf2)2.5·(dioxane)2–3·H2O1–2 for the solid material.
Figure
2 shows transmission electron microscopy (TEM)
photographs of the material (Ag loading 5.1 wt%, 0.5 equivalents of
starting Ag) and well–dispersed, crystalline Ag NPs can be seen.
Electron–dispersive X–ray (EDX) analysis indicates that the analyzed
area of the NP is mainly Ag (98.5%), and mapping of Ag and Fe
atoms in the whole micrograph shows that, while all Ag atoms are
concentrated in the NP, Fe atoms are dispersed throughout the
micrograph. In average, the material presents the expected
molecular Ag:Fe mol ratio (0.5) and the photographs suggest that
the Ag NPs are somehow embedded within a matrix of Fe.
NTf2
Fe–O(H)–Fe
NTf2
15N–NMR
13C–NMR
NTf2 terminal
(x4)
NTf2
Dioxane
*
NTf2 bridged
*
*
*
*
*
NTf2
I
I
I
I
I
I
I
200
250
0
-250
-500
100
0
ESI-MS
Fe2(NTf2)5O·(Et2O)
(F.W.= 1601 a.u.)
Figure 2. Left:
A transmission electron microscopy (TEM)
photograph of the Fe2O(NTf2)5@AgNPs solid (Ag:Fe mol ratio=0.5)
where well–dispersed Ag nanoparticles (~12 nm average size) can
be appreciated, with the corresponding histogram for at least 5
different photos and 150 particles. The inset shows a scanning
transmission electron microscopy–dark field (STEM–DF) photograph
of a well–faceted Ag NP. Right: Scanning transmission electron
microscopy–dark field (STEM–DF) micrograph of an Ag NP with the
corresponding electron–dispersive X–ray (EDX) analysis of the
squared area and the mapping for Ag (top) and Fe (bottom).
Fe2(NT
(F.W
Figure 3. Top: Fourier transform–infrared (FT–IR) and diffuse
reflectance ultraviolet–visible (RD–UV vis) spectra of the
Fe2O(NTf2)5@AgNPs solid. For the sake of comparison, the UV–Vis
spectra of Fe(NTf2)3 and 10 nm Ag nanoparticles in solution are also
shown. Middle: 13C and 15N nuclear magnetic resonance–magic
angle spinning (NMR–MAS) spectra of the 15N isotopically–labelled
yellow solid (*= rotational peaks). Bottom: Electrospray ionization
mass spectrum (ESI–MS) of the solid extracted with diethyl ether.
The classical analytical test for Fe2+ with ICl does not give any trace
of Fe2+ in the solid after dissolution with HNTf2. In contrast, the
analytical test with SnCl2 gave Fe3+ as the only iron species present
in the solid. The diffuse reflectance ultraviolet–visible (RD–UVvis)
spectrum of the solid shown in Figure 3 (top–left) confirms the
absence of d–d transition bands for low spin Fe(II) complexes, that
should appear at around 600 nm. These results indicate that Fe3+ is
the main iron species present in the solid, with less than 0.1–0.01 %
of Fe2+. Notice that the RD–UVvis spectrum fits well with the sum of
Figure 3 (middle) shows the 13C nuclear magnetic resonance–magic
angle spinning (NMR–MAS) spectrum of the solid, with the
expected peaks for 1,4–dioxane (~69 ppm) and triflimide (~120
ppm). Since no differentiation by 13C NMR can be expected for
different triflimides, we prepared a sample of isotopically–labelled
15N–solid, i.e. Fe2O(15N–Tf2)5@AgNPs, by using Fe(15NTf2)3 and
Ag15NTf2 as starting metal salts during the synthetic procedure
2 | Chem. Sc., 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 3 of 8
View Article Online
DOI: 10.1039/C6SC03335K
Journal Name
ARTICLE
depicted in Figure 1 (see Figure S2 in SI for details). The reasonable dispersion of Ag NPs in the solid must be related to the
corresponding 15N MAS–NMR spectrum in Figure 3 shows two well– very fast (<1 min) formation of the material, which prevents further
resolved 15N NMR peaks, one main peak that corresponds to agglomeration, giving finally Ag NPs without the aid of any ligand or
triflimide anions bound to one Fe3+ atom (–144 ppm), and a second additional support.8
peak upshifted –63 ppm that integrates for ¼ of the former. This
2.2.
Catalytic results of Fe2O(NTf2)5@AgNPs.
result is in good agreement with the existence of two different
triflimide anions as indicated by FT–IR and TG measurements.
Moreover, the 4:1 ratio fits with the ꢀ–hydroxo or ꢀ–oxo bridged
Fe2O(NTf2)5 species.
2.2.1. Acid–catalysed C–C activation. Taking into account the
potential strong acidity of the solid, its catalytic activity was
tested for reactions that require an acidity value H0<-12, i.e.
reactions that can only be performed with concentrated H2SO4
or stronger acids. The catalytic solid material was that directly
obtained from the synthetic procedure described above (i.e
precipitated, washed in hexane and dried under vacuum), and no
grounding or sieving was needed to achieve reproducible results.
Figure 5 shows the results obtained for the different acid–
catalysed reactions.
To obtain a more direct evidence for the presence of the
dimeric species, an exhaustive extraction of the material was
performed in different solvents, and we found that treatment of
the solid in hot diethyl ether for 3 days gives a solution that
contains one single compound in amounts large enough to be
analysed by electrospray ionization mass spectrometry (ESI–MS).
The mass obtained for this compound is 1660.9 Daltons, as shown
in Figure 3, which can be assigned to Fe2(NTf2)5O·(Et2O) (F.W.= 1601
D). Matrix–assisted laser–desorption/ionization coupled to a time–
of–flight mass spectrometer (MALDI–TOF–MS) confirms that no
heavier species up to 40000 m/z are present in this solution. These
results indicate that the solid contains significant amounts of
Fe2O(NTf2)5 species.
Electronic paramagnetic resonance (EPR) measurements of the
solid showed a lack of EPR signals for Fe in the solid (Figure S3 in SI)
which can only be explained by either
a very strong
antiferromagnetic between two Fe3+ atoms, or because the Fe
centers are low spin Fe2+.7 The latter can be rejected since the
material does not present any Fe2+ according to the analytical tests
and RD–UVvis spectroscopy (vide supra). Thus, the fact that the
material is diamagnetic at room temperature (EPR silent and no
alteration in the NMR shifts) can only be explained by a strong
antiferromagnetic coupling between two oxo–bridged Fe3+ atoms.
Indeed, a semi–quantitative empirical correlation gives a J=
1.753x1012•e–12:663•R, where R is half of the shortest exchange
pathway, equivalent to the Fe–O distance, and agrees well with a
O–bridging two Fe3+ atoms. The existence of a ꢀ–hydroxo or ꢀ–oxo
bridge between two Fe3+ atoms explains the 5 triflimides in 4:1 ratio
observed in the TG, FT–IR and 15N MAS–NMR techniques: 4
triflimides bound to a single Fe atom (two per Fe atom) and the fifth
one bridging the two Fe3+ atoms, which close the hexagonal
coordination sphere of each Fe3+ in the dimer Fe2(NTf2)5O(H), as
shown in Figure 4. Notice that, electronically, the Fe3+ dimer with 5
triflimides is compensated with the ꢀ–hydroxo bridge, however, a
ꢀ–oxo bridge could also exist if the H+ remains associated to 1,4–
dioxane or water molecules.
Figure 5. Acid–catalysed reactions with Fe2(NTf2)5O@AgNPs.
Isolated yields. From top to bottom: A) head–to–tail dimerization of
styrenes, B) Markovnikov hydration of alkynes, C) addition of
methyl acetoacetate to styrene, and D) hydrodeoxygenation of
cyclohexanol. a GC yield.
The head–to–tail dimerization of styrenes (Figure 5, A) is a
useful reaction to obtain branched styrenes in a 100% atom-
economical manner,9 and it has been catalyzed previously
with, among others, Pd triflate complexes,10 a combination of
metal complexes such as Pd/In triflate/phosphines11 or
Co/Zn,12 and with Fe(NTf2)3,13 all of them in homogenous
solution. Thus, a simple solid catalyst for this reaction remains
Figure 4. Structure of the Fe3+–triflimide (Tf) dimer
in the solid. Coordination of the triflimides can
occur through oxygen or nitrogen atoms.
a
challenge. Indeed, only the linear (not branched)
dimerization of styrenes has been reported with
a
heterogeneous catalysts, i.e. Ru supported on CeO2, and using
formaldehyde or ethanol as promoters to give a 63% yield of
1,4–diaryl–1–butenes.14 Fe2(NTf2)5O@AgNPs catalyzes the
regioselective head–to–tail dimerization of styrenes 1a-c to 2a-
With all these data in hand, we can say that the structure of the
yellow solid obtained is formed by well–dispersed Ag nanoparticles
embedded in a matrix of O,triflimide–dibridged Fe3+ dimers, namely
thereafter as Fe2(NTf2)5O@AgNPs and depicted in Figure 1. The
c
in yields up to 98%. The hot filtration test for styrene 1a
showed that the catalytic activity completely stops when the
This journal is © The Royal Society of Chemistry 20xx
Chem. Sc., 2016, 00, 1-3| 3
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 4 of 8
View Article Online
DOI: 10.1039/C6SC03335K
ARTICLE
Journal Name
solid catalyst is filtered off, which confirms the heterogeneous Fe2O(NTf2)5@AgNPs as a catalyst for the oxidation of alkanes
nature of the catalysis.
under similar conditions to those reported, but without the
need of adding AcOH to activate the iron catalyst.34,35
The Markovnikov hydration of alkynes (Figure 5, B) was
early catalysed by Hg salts at the industrial level15 and, due to
toxicity issues, alternative catalysts based on transition metals
are currently been explored, mainly gold.16-18 It is difficult to
find efficient catalysts that operate under mild conditions
beyond noble metals,7 and systems based on Brönsted acids
either homogeneous or heterogeneous require high wt%
loadings, harsh reaction conditions (>150 ºC) or very particular
modifications of the catalytic system (additives, singular
reaction media such as microemulsions, surface modifications
in solids).19-21 Fe2O(NTf2)5@AgNPs catalyzes regioselectively the
hydration of alkynes 3a-d to ketones 4a-d without any additive
in up to 97% yield.
The addition of methyl acetoacetate to styrenes (Figure 5,
C) was originally reported with Fe complexes in stoichiometric Figure 6. Redox–catalyzed selective CH2 and Baeyer–Villiger
amounts22 and further developed with noble metal catalysts,23 oxidations with Fe2O(NTf2)5@AgNPs (10 and 20 mol% of Fe for
until Fe could be employed catalytically.24 It is not easy to find alkanes and ketones, respectively). X: Conversion. S: Selectivity. GC
a heterogeneous catalyst for this reaction. Fe2O(NTf2)5@AgNPs results.
is able to catalyze the reaction between styrene 1c and methyl
acetoacetate 5 to give the product 6 in 90% yield.
The results in Figure 6 show that Fe2O(NTf2)5@AgNPs converts
The hydrodeoxygenation reaction (Figure 5, D) is of
interest in the biorefining industry to obtain alkanes from
oxygen–
rich biomass derived chemicals.25 For that, a catalytic
linear and cyclic alkanes to a mixture of alcohols and ketones 9a-d,
in reasonable conversions after
a few minutes at room
temperature, with good selectivity towards sterically accessible
methylene groups. The solid catalyst is suitable not only for
methylene oxidation but also for the Baeyer–Villiger oxidation of
cyclic ketones to the corresponding cyclic esters 10a-c,36 under the
same reaction conditions.
combination of a soluble very acid metal salt, such as triflates,
and a solid hydrogenation catalyst, such as Pt/C, is employed.26
Thus, it would be desirable to have both functions on a solid,
either as a single solid or as a composite. A catalytic composite
of Pt on TiO2/C has been reported27 and operates at
2.2.3. Acid/redox–catalysed C–C and C–H activation. The
performance of redox catalysis in acid media gives access to
otherwise elusive reactions. Since Fe2O(NTf2)5@AgNPs is
operative in acid and redox reactions, separately, it was tested the
catalytic activity of Fe2O(NTf2)5@AgNPs for reactions that require
both functions present (see Table S2 for comparison with state–of–
the–art catalysts).The results are shown in Figure 7.
temperatures
Fe2O(NTf2)5@AgNPs and Pt/C catalyzes the hydrodeoxygenation
of cyclohexanol to cyclohexane in 98% yield at 150 ºC.
~300
ºC.
Here,
a
composite
of
7
8
In summary, Fe2O(NTf2)5@AgNPs catalyzes the synthetically
useful, acid strength–demanding reactions shown above in very
good yields, with similar catalytic activity to the state–of–the–
art soluble catalysts (Table S2). For the sake of comparison, the
dimerization of styrene 1a and the hydration of
phenylacetylene 3a were also carried out with representative
The Markovnikov hydrothiolation of styrenes is catalyzed
by a Fe3+/Fe2+ manifold mechanism with the participation of
protons in the redox catalytic cycle.37 The results obtained
(Figure 7, A) shown that Fe2O(NTf2)5@AgNPs is as active as the
state–of–the–art soluble catalysts,38 with good yields for the
representative products 12a-d. Besides, the solid catalyst could be
reused up to 9 times after simple filtration, without a significant
loss of its inherent catalytic activity (Figure S5).
strong solid acid catalysts such as nafionTM 28 H–USY zeolite,29 and
,
sulfated zirconia,30 under the indicated conditions (Figure S4). The
results showed that only nafionTM improves the catalytic activity of
Fe2(NTf2)5O@AgNPs, while H–USY and sulfated zirconia are much
poorer acid catalysts, which places Fe2O(NTf2)5@AgNPs among the
most acid solid catalysts reported.
The selective demethylation of N,N–dimethylanilines is a
2.2.2. Redox–catalysed C–H activation. Selective oxidation of C–
H bonds with environmentally benign reagents such as O2 or
H2O2 under mild reaction conditions is one of the main
challenges in organic synthesis.2 Nature makes use of enzymes
with non–heme O–bridged bisiron centers,31 and successful
bio–mimetic lines of research have been reported with O–
bridged bisiron complex catalysts having low–coordinating
ligands for C–H activation and C–C cleavage.32–35 Given the
resemblance between these adducts32–33 and the Fe3+ dimer
present in Fe2O(NTf2)5@AgNPs (see Figure 4 above), we tested
biomimetic reaction catalyzed by iron complexes in the
presence of strong oxidants, such as PhIO.39 Nature makes use
of O2 as the terminal oxidant, thus the synthetic use of O2 as
environmental-friendly oxidant for this reaction would be a
significant advance. Recently, it has been reported that this
reaction can moderately occur under aerobic conditions (~40%
combined yield with trimers and formamides) if a metal triflate
salt, i.e. Zn2+, Ba2+ or Y3+, is used in combination with the Fe3+
complex to form a triflate bridge between the two metal
4 | Chem. Sc., 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 5 of 8
View Article Online
DOI: 10.1039/C6SC03335K
Journal Name
ARTICLE
cations.40 It was envisioned that the solid triflimide composed accessible to the reactant molecules, in–situ low–temperature
of Fe dimers could mimic this system and catalyze the FT–IR experiments with carbon monoxide (CO), as a probe
demethylation of
N,N–dimethylanilines under similar aerobic molecule, were carried out (Figure S7 in SI). The results show
conditions (Figure 7, B). Indeed, the reaction worked well with that there is not interaction between CO and Ag0 or Agδ
Fe2O(NTf2)5@AgNPs, with similar yields and better selectivity to atoms,45 which unequivocally indicates that the Ag atoms in
the products than with the soluble metal complex since, the solid are not accessible to external chemicals. The lack of
presumably, the higher acidity of the triflimide solid overrides interaction between CO and Fe3+ was expected due to the
+
undesired biphenyl formation by the non
electron trasnfer.40
–
acidic aryl/Fe low–coordinating nature of the triflimide anions and reflects
the high acidity of the Fe3+ atoms in the solid. Considering the
amount and size of Ag NPs in the material for Ag:Fe ratios ≤0.5
It was also tested the possibility of engaging the hydration
of the aryl alkyl alkyne 15 and the methylene oxidation
reaction in cascade,41 with Fe2O(NTf2)5@AgNPs as a single
catalyst for both processes (Figure 7, C). The result shows that
the solid triflimide indeed preserves its redox catalytic activity
and the number of Fe atoms, a simple calculation gives that
the number of Fe dimers is at least 104 higher than the number
of Ag nanoparticles. Thus, there are enough Fe2O(NTf2)5
molecules to embed all the NPs. In fact, only 5% of Fe dimers
are enough to cover the whole surface of the Ag NPs. It
appears then that the Fe3+ matrix acts as a chemical shield for
the Ag NPs, conferring them protection against air, water or
any other external atmosphere. For instance, the material
(Ag:Fe mol ratio=0.5) is stable after heating at 200 ºC for 1 day
under a H2 atmosphere of 10 bars and no agglomeration of Ag
NPs was observed after this treatment.
after acting as an acid catalyst, and the 1,5
formed in one pot. Notice that the regioselective formation of
the first carbonyl group by acid catalyzed hydration directs the
–diketone 16 is thus
–
–
later oxidation of the methylene group.
A)
SH
R
11a-d
R = NO2, 12a 85%
R = OMe, 12b 88%
R = Br, 12c 77%
R = H, 12d 82%a
R
S
Fe2O(NTf2)5@AgNPs
(10 mol% Fe)
Dry 1,4-dioxane
80 ºC, 24 h
Cl
B)
Cl
12a-d
1a
NH
R = H, 14a 45%b
R = Me, 14b 33%c
R = Br, 14c 27%c
R = CN 14d 15%c
Fe2O(NTf2)5@AgNPs
(5 mol% Fe)
N
O2 (balloon), CH3CN
40 ºC, 24 h
R
R
13a-d
14a-d
10 nm
Accesible Ag
C)
atoms
Fe2O(NTf2)5@AgNPs
(40 mol% Fe)
O
O(H)
Then H2O2 (5 eq.)
3
CH3CN (0.25 M),
15 min, rt
H2O (3 eq.),
dry 1,4-dioxane,
110 ºC, 6 h
16, 28%d
15
Figure 8. Scanning transmission electron microscopy–dark field
(STEM–DF, left) and high–resolution transmission electron
microscopy (HR–TEM, right, black bar accounts for 10 nm)
photographs of Fe2O(NTf2)5@AgNPs synthesized with a Ag:Fe3+ mol
ratio= 1. The inset magnifies a particular area (inverted).
O
3
Figure 7. Acid/redox reactions catalysed by Fe2(NTf2)5O@AgNPs.
Isolated yields. From top to bottom: hydrothiolation of styrenes,
However, it should be in principle possible to make accessible the
Ag NPs in the solid by just increasing the Ag to Fe3+ ratio, in such a
way that the Fe3+ triflimide dimers will not cover completely the Ag
NPs. Figure 8 shows that, indeed, the solid material has clearly
visible ~10 nm Ag NPs on the surface when the synthesis is
performed with an Ag:Fe mol ratio=1. The resultant material with
accessible Ag NPs is active for the classical Ag–catalyzed reaction of
aniline to give azobenzene.44 The solid catalyst now gives a 20%
conversion with a turnover number (TOF) relative to the exposed
Ag atoms of 100 ( (see Figure S6 in SI). Comparatively, the Ag:Fe mol
ratio=0.5 solid is completely inactive and commercially available
Ag/C gives only 7% conversion. These results confirm that the
accessibility of the reactants to the Ag NPs in the solid material can
be regulated by varying the relative amount of metals during the
synthesis.
demethylation of N,N-dimethylanilines, and one–pot hydration
b
a
of alkyne 15–CH2 oxidation. After 8 uses of Fe2(NTf2)5O@AgNPs.
c
Combined yield with the corresponding trimers. Combined yield
with the corresponding formamides. d GC yield.
2.2.4. Making the Ag NPs accessible to reactants. When the solid
with a Ag:Fe ratio=0.5 was tested for reactions typically
catalyzed by Ag NPs of ~10 nm, including the epoxidation of
styrene,42 the aerobic dehydrogenation of alcohols43 and the
synthesis of azocompounds from anilines (see Figure S6 in
SI),44 no conversion was found for any of these reactions. Blank
experiments with independently synthesized Ag NPs of ~10 nm
showed the expected reactivity for the Ag catalyst.42–44
Therefore, the lack of catalytic activity of the Ag nanoparticles
in the Fe2O(NTf2)5@AgNPs solid suggests that the contact
between Ag NPs and the reactants may be hampered in some
way. To further assess if the Ag NPS in the solid were
This journal is © The Royal Society of Chemistry 20xx
Chem. Sc., 2016, 00, 1-3| 5
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 6 of 8
View Article Online
DOI: 10.1039/C6SC03335K
ARTICLE
Journal Name
In summary, the results shown in section 2.2 provide a picture bridge according to the slight shift of the corresponding IR band
of the rather unique acid/redox catalytic behavior of the respect to the non–labelled yellow solid (see Figure 3).6
Fe2O(NTf2)5@AgNPs solid.
Additionally, we found that H218O is formed in solution, as
determined by CG–MS. These results support that Fe2+ oxidizes to
Fe3+ not only with Ag+ but also with the O2 present in the
atmosphere, and that O2 is then reduced to H2O. This reaction
complements the oxidation with Ag+ and closes the mass balance.
The bridged oxygen atom would then come from reduced
molecular oxygen, either directly or after reduction to H2O. To test
this last possibility, H218O was added to an anhydrous solution of
triflimides prior to the formation of the solid under air. As occurred
previously when working with a 18O2 atmosphere, the formation of
the isotopically–labelled Fe218O(NTf2)5@AgNPs was observed, with
the concomitant consumption of H218O observed by GC–MS. These
results support that the oxygen atom for the O–bridged Fe3+ dimer
comes from water molecules present in the medium. If so, the
bridge should also be produced under an inert atmosphere
provided that water is in the medium. Indeed, a 65% yield of solid
was obtained under N2 when 0.5 equivalents of Ag were used. It
2.3.
Mechanism of formation of Fe2O(NTf2)5@AgNPs.
A possible mechanism for the formation of the Fe2O(NTf2)5@AgNPs
solid is depicted in Figure 9. This mechanism is supported by
reactivity tests and isotopic experiments, and it consists in the one–
electron reduction of Fe(NTf2)3 to Fe(NTf2)2 by PhSH, forming a
coordinatively unsaturated Fe2+ species that re–oxidizes at
expenses of reducing Ag+ to Ag NPs or, alternatively, with air.
Figure 9. Proposed mechanism for the formation of must be noticed, however, that the formation of the bridge is less
Fe2O(NTf2)5@AgNPs.
efficient in the absence of O2 since a 90% yield is obtained when the
process is run aerobically, which is explained by the better
oxidation of Fe2+ to Fe3+ when molecular oxygen is present.
The one–electron reduction of Fe(NTf2)3 to Fe(NTf2) is supported by
the quantitative transformation of the one–electron reductant
PhSH to diphenyl disulfide Ph2S2, observed by GC–MS and liquid
NMR measurements of the solution, when PhSH is added to a
mixture of Fe(NTf2)3 and AgNTf2, to form the yellow solid. In
contrast, if PhSH is added to a solution of Fe(NTf2)3 without AgNTf2
present in the medium, the solid does not form despite Ph2S2 is
quantitatively produced. Furthermore, if PhSH is added to a
solution of AgNTf2 without Fe(NTf2)3 present, no reaction occurs.
These results indicate that the one–electron reduction occurs to
form Fe2+, in agreement with the well–known single electron
All the steps described above occur in less than 1 minute at
room temperature, and the formation of
a coordinatively
unsaturated Fe(NTf2)2 species seems to play a key role in the
formation of the solid, accommodating the bridging NTf2 and O
species. In agreement with this, when a sample of independently
prepared Fe2+(NTf2)2 was used as starting material to form the
solid,13 with the whole hexacoordination sphere saturated (2
bridging triflimides and 4 bisoxo–coordinated terminal triflimides, 2
per Fe2+ atom), no solid was formed. This result evidences the need
of in–situ reducing the Fe3+ triflimide.
transfer (SET) process between PhSH and Fe3+ salts at room 2.4.
Extension of the work to other metallic systems: Synthesis
temperature.37 Other potential reductants for Fe3+ such as of Cu, Bi and Yb–triflimide@AgNPs.
isopropanol and Ce3+ failed to form the solid.
According to the mechanism in Figure 9, it might be expected that
The yield of solid decreases significantly if an excess of PhSH respect other metal cations could participate in the proposed redox
to metal triflimide is employed during the synthesis (90%, 65%, 28% sequence, as long as the corresponding metal triflimide can suffer
for 1, 3 and 10 equivalents of PhSH, respectively), due to the the SET reduction with PhSH and the generated reduced metal
formation of stable Ag+–thiolate complexes. When Au, Pd and Pt cation is re–oxidized back with O2 and/or Ag+. If so, a family of
salts were tested instead of Ag no solid formation occurred. These ligand–free, self–supported bimetallic solids could be available
results, in principle, discard the possibility that thiophenol acts as a following the preparation method described here. After examining
“seed” for nanoparticle formation, and the SET transfer reaction the redox potentials of tabulated metal cations and testing those
seems to be predominant in the mechanism.
suitable, we found that the triflimides of Cu2+, Yb3+ and Bi3+ were
able to form the corresponding solids in moderate to good yields
(84% for copper, 75% for bismuth and 63% for ytterbium) under the
same reaction conditions than for Fe2O(NTf2)5@AgNPs. Other
cations such as Ru3+ or Co3+ failed to produce the corresponding
solids.
The second step in the mechanism of formation of the
bimetallic solid is a redox reaction between the in–situ formed Fe2+
and Ag+, to give Fe3+ and Ag NPs, as expected from their respective
redox potentials. Accordingly, the pH of the solution is very acidic
(<1) by the released triflimidic acid, and the protons may stay either
coordinated to the oxygen atoms of the O–bridges or to the Figure 10 shows the characterization of the solids by FT–IR and
dioxane molecules.
TEM. The results suggest that, in principle, these solids have a
similar structure than Fe2O(NTf2)5@AgNPs, i.e. they are formed by
well–dispersed Ag NPs embedded within a metal triflimide matrix.
In the case of Cu, the Ag NPs flourish to the surface (see left
microphotograph in Figure 10).
The third and last step is the formation of the ꢀ–oxo or ꢀ–
hydroxo bridge between two Fe3+ atoms. To shed light on this step,
we used an atmosphere of 18O2, and the corresponding FT–IR
spectrum of the yellow solid showed the formation of a Fe–18O–Fe
6 | Chem. Sc., 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 7 of 8
View Article Online
DOI: 10.1039/C6SC03335K
Journal Name
ARTICLE
bimetallic solids here formed can act as an acid catalyst, a
redox catalyst or a bifunctional acid/redox catalyst.
ꢀ–oxo
ꢀ–oxo
ꢀ–oxo
Acknowledgements
,2
Bi
Yb
,0
,8
,6
,4
,2
,0
,2
,4
,6
Cu
Financial support by “Severo Ochoa” program, RETOS program
NTf2
(CTQ2014-55178-R) and Ramón y Cajal Program (A. L.–P.) by
NTf2
NTf2
MINECO (Spain), and also by “Convocatoria 2014 de Ayudas
Fundación BBVA a Investigadores y Creadores Culturales” are
acknowledged. The Electron Microscopy Service of the UPV is
also acknowledged.
500
1000
1500
2000
2500
3000
3500
500
1000
1500
2000
ν (cm-1
2500
3000
3500
500
1000
1500
2000
2500
3000
3500
ν (cm-1
)
)
ν
(cm-1)
Notes and references
1
S. Antoniotti, V. Dalla, E. Duñach, Angew. Chem. Int. Ed.,
2010, 49, 7860.
M. C. White, Science, 2012, 335, 807.
2
3
4
5
F.-X. Felpin, E. Fouquet, ChemSusChem, 2008, 1, 718.
Figure 10. Top: FT–IR spectra of the solids made by the same
procedure than Fe2O(NTf2)5@AgNPs but with Cu(NTf2)2, Bi(NTf2)3
and Yb(NTf2)3 as starting materials, from left to right, respectively.
Bottom: Transmission electron microscopy (TEM) photographs with
the corresponding histograms for, at least, 5 different photos.
U. Diaz, D. Brunel, A. Corma, Chem. Soc. Rev., 2013, 42, 4083.
I. Rey, P. Johansson, J. Lindgren, J. C. Lassègues, J. Grondin, L.
Servant, J. Phys. Chem. A, 1998, 102, 3249.
6
7
D. M. J. Kurtz, Chem. Rev., 1990, 90, 585.
J. R. Cabrero-Antonino, A. Leyva-Pérez, A. Corma, Chem. Eur.
J., 2012, 18, 11107.
8
9
A. Desireddy, B. E. Conn, J. Guo, B. Yoon, R. N. Barnett, B. M.
Monahan, K. Kirschbaum, W. P. Griffith, R. L. Whetten, U.
Landman, T. P. Bigioni, Nature, 2013, 501, 399.
J. Skupiçska, Chem. Rev., 1991, 91, 613.
The Cu and Bi material were tested as catalysts in the vinylation of
1,3–diphenylpropargyl alcohol46-48 and in the the hydrothiolation
of styrenes,37 respectively. Figure 11 shows that products 18
and 12d were obtained in good yields, similar to those previously
reported.
10 J. H. Choi, J. K. Kwon, T. V. RajanBabu, H. J. Lim, Adv. Synth.
Catal., 2013, 355, 3633.
11 R. B. Bedford, M. E. Betham, M. Blake, A. S. Garcés, S. L.
Millar, S. Prashar, Tetrahedron, 2005, 61, 9799.
12 C.-C. Wang, P.-S. Lin, C.-H. Cheng, Tetrahedron Lett., 2004,
45, 6203.
13 J. R. Cabrero-Antonino, A. Leyva-Pérez, A. Corma, Adv. Synth.
Catal., 2010, 352, 1571.
14 S. Shimura, H. Miura, S. Tsukada, K. Wada, S. Hosokawa, M.
Inoue, ChemCatChem 2012,
15 L. Hintermann, A. Labonne, Synth., 2007,
4, 2062.
8, 1121
16 R. E. Ebule, D. Malhotra, G. B. Hammond, B. Xu, Bo, Adv.
Synth. Catal., 2016, 358, 1478.
17 S. Liang, J. Jasinski, G. B. Hammond, B. Xu, Org. Lett., 2015,
17, 162.
18 A. Leyva, A. Corma, J. Org. Chem., 2009, 74, 2067.
Figure 11. Top: Vinylation of 1,3–diphenylpropargyl alcohol 17 with
styrene 1c catalyzed by the Cu–triflimide@AgNPs solid. Bottom:
Hydrothiolation of 4-chlorostyrene 1a catalyzed by the Bi–
triflimide@AgNPs solid. GC yields.
19 Z. Nairoukh, D. Avnir, J. Blum, ChemSusChem, 2013,
20 N. Mameda, S. Peraka, M. R. Marri, S. Kodumuri, D.
Chevella, N. Gutta, N. Nama, Appl. Catal. A, 2015, 505, 213.
21 F. Alonso, I. P. Beletskaya, M. Yus, Chem. Rev., 2004, 104
3079
6, 430.
,
22 P. Lennon, A. M. Rosan, M. Rosenblum, J. Am. Chem. Soc.,
1977, 99, 8426.
23 H. Qian, R. A. Widenhoefer, J. Am. Chem. Soc., 2003, 125
2056.
,
Conclusions
24 J. Kischel, D. Michalik, A. Zapf, M. Beller, Chem. Asian J.,
2007, , 909.
2
25 G. W. Huber, S. Iborra, A. Corma, Chem. Rev., 2006, 106
The direct synthesis of self–supported triflimide metal
matrices on Ag nanoparticles has been accomplished by a
conceptually–new procedure that involves a redox–coupled
sequence of two metal salts under air.49,50 This
straightforward, sustainable and high–scale method of
preparation of a catalytic triflimide solid opens a new way of
synthesizing strong solid acids with redox sites. The nanosized
,
4044.
26 Z. Li, R. S. Assary, A. C. Atesin, L. A. Curtiss, T. J. Marks, J. Am.
Chem. Soc., 2014, 136, 104.
27 A. Corma, O. de la Torre, M. Rentz, N. Villandier, Angew.
Chem. Int. Ed., 2011, 50, 2375.
28 G. Gelbard, Ind. Eng. Chem. Res., 2005, 44, 8468.
29 J. R. Cabrero-Antonino, A. Leyva-Pérez, A. Corma, Angew.
Chem. Int. Ed., 2015, 54, 5658.
This journal is © The Royal Society of Chemistry 20xx
Chem. Sc., 2016, 00, 1-3| 7
Please do not adjust margins

Please do not adjust margins
Chemical Science
Page 8 of 8
View Article Online
DOI: 10.1039/C6SC03335K
ARTICLE
Journal Name
30 K. Arata, Green Chem., 2009, 11, 1719.
31 C. E. Tinberg, S. J. Lippard, Acc. Chem. Res., 2011, 44, 280.
32 R. A. Leising, J. Kim, M. A. Pérez, L. Que Jr., J. Am. Chem. Soc.,
1993, 115, 9524.
33 N. Raffard-Pons Y Moll, F. Banse, K. Miki, M. Nierlich, J.-J.
Girerd, Eur. J. Inorg. Chem., 2002, 8, 1941.
34 M. S. Chen, M. C. White, Science, 2010, 327, 566.
35 P. E. Gormisky, M. C. White, J. Am. Chem. Soc., 2013, 135
,
14052.
36 A. Corma, L. T. Nemeth, M. Renz, S, Valencia, Nature, 2001,
412, 423.
37 J. R. Cabrero-Antonino, A. Leyva-Pérez, A. Corma. Adv. Synth.
Catal., 2012, 354, 678.
38 R. Castarlenas, A. Di Giuseppe, J. J. Pérez-Torrente, L. A. Oro.
Angew. Chem. Int. Ed., 2013, 52, 211.
39 E. Baciocchi, O. Lanzalunga, A. Lapi, L. Manduchi J. Am.
Chem. Soc., 1998, 120, 5783.
40 J. Zhang, Y. Wang, N. Luo, Z. Chen, K. Wu, G. Yin. Dalton
Trans., 2015, 44, 9847.
41 A. Leyva-Pérez, P. García-García, A. Corma. Angew. Chem.
Int. Ed., 2014, 53, 8687.
42 R. Xu, D. Wang, J. Zhang, Y. Li, Chem. Asian J., 2006, 1, 888.
43 T. Mitsudome, Y. Mikami, H. Funai, T. Mizugaki, K. Jitsukawa,
K. Kaneda, Angew. Chem. Int. Ed., 2008, 47, 138.
44 S. Cai, H. Rong, X. Yu, X. Liu, D. Wang, W. He, Y. Li, ACS Catal.,
2013, 3, 478.
45 Z. Qu, S. Zhou, W. Wu, C. Li, X. Bao, Catal. Lett., 2005, 101
,
21.
46 G.-B. Huang, X. Wang, Y.-M. Pan, H.-S. Wang, G.-Y. Yao, Y.
Zhang, J. Org. Chem., 2013, 78, 2742.
47 J. Kischel, K. Mertins, D. Michalik, A. Zapf, M. Beller, Adv.
Synth. Cat., 2007, 349, 865.
48 P. Moriel and A. B. García, Green Chem., 2014, 16, 4306.
49 Z. Wang, G. Chen, K. Ding, Chem. Rev., 2009, 109, 322.
50 H. Chen, D. Wang, Y. Yu, K. A. Newton, D. A.; Muller, H. D.
Abruña, F. J. DiSalvo, J. Am. Chem. Soc., 2012, 134, 18453.
8 | Chem. Sc., 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins