Synergistic catalysis of the Br?nsted acid and highly dispersed Cu on the mesoporous Beta zeolite in the intermolecular aminoazidation of styrene

Article abstract of DOI:10.1016/j.apcata.2020.117907

Acidic mesoporous Beta zeolite supported Cu (Cu/Beta-M) catalyst shows superior catalytic performance in the difunctionalization of styrenes with N-fluorobenzenesulfonimide (F-N(SO₂Ph)₂) and azidotrimethylsilane, compared with basic

Full text of DOI:10.1016/j.apcata.2020.117907

Journal Pre-proof
Synergistic catalysis of the Bro¨nsted acid and highly dispersed Cu on the
mesoporous Beta zeolite in the intermolecular aminoazidation of styrene
Wenqian Fu (Validation) (Investigation) (Writing - original draft)
(Funding acquisition), Chengyou Yin (Investigation), Yu Feng
(Formal analysis), Lei Zhang (Formal analysis) (Funding
acquisition), Feng Cheng, Zhongxue Fang (Data curation), Chaojie
Zhu (Visualization), Tiandi Tang (Conceptualization) (Methodology)
(Writing - review and editing) (Funding acquisition)
PII:
S0926-860X(20)30500-7
DOI:
https://doi.org/10.1016/j.apcata.2020.117907
APCATA 117907
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
28 July 2020
26 October 2020
29 October 2020
Please cite this article as: Fu W, Yin C, Feng Y, Zhang L, Cheng F, Fang Z, Zhu C, Tang T,
Synergistic catalysis of the Bro¨nsted acid and highly dispersed Cu on the mesoporous Beta
zeolite in the intermolecular aminoazidation of styrene, Applied Catalysis A, General (2020),
doi: https://doi.org/10.1016/j.apcata.2020.117907

This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Synergistic catalysis of the Brönsted acid and highly dispersed Cu on the
mesoporous Beta zeolite in the intermolecular aminoazidation of styrene
Wenqian Fu, Chengyou Yin, Yu Feng, Lei Zhang, Feng Cheng, Zhongxue Fang,
Chaojie Zhu, Tiandi Tang*
Advanced Catalysis and Green Manufacturing Collaborative Innovation Center,
Jiangsu Key Laboratory of Advanced Catalytic Materials and Technology, School of
Petrochemical Engineering, Changzhou University, Changzhou, Jiangsu 213164, P. R.
China.
* To whom correspondence should be addressed.
E-mail: tangtiandi@cczu.edu.cn
Graphical abstract
Highlights
 Cu/Beta-M exhibits high activity in the intermolecular aminoazidation of
styrene.
1

 Brönsted acid and Cu+ (Cu2+) species have a synergetic catalysis.
 Brönsted acidic sites on zeolite chemisorb the styrene and activate the
C=C bond.
 Cu+
(Cu2+)
favor
the
formation
of
electrophile
from
N-fluorobenzenesulfonimide.
ABSTRACT
Acidic mesoporous Beta zeolite supported Cu (Cu/Beta-M) catalyst shows superior
catalytic performance in the difunctionalization of styrenes with
N-fluorobenzenesulfonimide (F-N(SO2Ph)2) and azidotrimethylsilane, compared with
basic mesoporous ETS-10-M and acid-free Silicalite-1-M support Cu catalysts. The
Brönsted acid, copper species and mesoporosity in the Cu/Beta-M catalyst jointly
contribute to this reaction. The Cu/Beta-M catalyst can induce the chemisorption of
the styrenes through interaction of C=C bond with the Brönsted acidic sites,
promoting the activation of the C=C bond. Meanwhile, the Cu species in Cu/Beta-M
can catalyze F-N(SO2Ph)2 to transform into the -N(SO2Ph)2 electrophile, which
attacks the activated C=C bond of styrenes, leading to the catalyst with high activity
and product selectivity. In addition, the styrenes with electron-donating groups have
high electron density of the π-conjugated system, which can strengthen the interaction
of C=C bond with the Brönsted acidic sites, leading to the substrates with high
reactivity.
2

Keywords: Mesoporous Beta zeolite, Cu catalyst, Brönsted acid, Chemisorption,
Styrene difunctionalization
1. Introduction
Alkene difunctionalization reactions are simple and smart synthetic strategies for
simultaneous construction of C-N, C-O or C-X (X=halogens) bonds in organic
molecules [1-3], which are widely used in preparation of biologically active
compounds for the pharmaceuticals and agrochemicals industries [1-3]. Currently,
these kinds of transformations are commonly studied using homogeneous
transition-metal-salts as catalysts [1-3]. To improve the reaction activity, organic
ligands are required to modify the electronic properties or steric-hindrance effects of
the metal atoms [2-4]. In addition, some reaction systems require the addition of
different additives, such as organic oxidants and organic or inorganic alkali salts [5-7].
The use of homogeneous metal-salts, ligands, and various additives lead to heavy
metal residuals and harmful waste products, which does not meet the requirement of
sustainable green synthetic chemistry nowadays. As an example, the aminoazidation
of styrenes with N-fluorobenzenesulfonimide (F-N(SO2Ph)2) and azidotrimethylsilane
(TMSN3) can be catalyzed by CuCl salt, but this reaction must be proceeded in the
presence of 1,10-phenanthroline ligand [8]. Another example is the azidoarylation of
alkenes with TMSN3, in which organic iodine and K2S2O8 oxidants or NaOAc alkali
additives are needed [9,10]. Therefore, the complicatedness of the reaction system
limits the practical application. From the view of sustainable and practical point, the
essential solution is to develop a high-efficiently heterogeneous catalyst with
multi-functionality to realize alkene difunctionalization reaction in green and
economic routes.
3

As we know, the adsorption of reactive substrate on active sites is a critical process
in heterogeneous catalysis, which can promote the chemisorption and activation of the
substrate and improve its activity. On the other hand, for the supported metal catalyst,
the size of the metal species greatly influences its electronic properties and geometric
construction, which can affect its catalytic performance. It has been reported that
when the size of metal species on support is smaller than 1 nm, the electronic
properties of metal species were further changed due to the metal-support interaction,
enhancing the catalyst activity [11-13].
Therefore, it motivated us to endeavor to hunt a support material with unique
properties for the preparation of functional metal catalysts. In this case, the prepared
catalyst not only has unique surface properties that are beneficial for substrate
chemisorption, but also has small active metal whose electronic properties can be
modified by metal-support interaction to improve the catalytic performance. Zeolites
are one class of porous crystal materials with large specific surface area, which are
widely used as adsorbents, catalysts or metal supports in the petrochemical industry
and fine chemistry [14-16]. The unique zeolitic framework can confine the metal
species, which results in the highly dispersed metal clusters with particular electronic
properties showing high activity [17-22]. Recently, we found that the styrene can be
chemosorbed on mesoporous acidic zeolite through interaction of C=C bond with
Brönsted acidic sites [23], which could benefit the C=C bond activation and promote
its functionalization reaction. Herein, we used an acidic, mesoporous Beta zeolite as a
support for assembling Cu clusters catalyst (Cu/Beta-M) to achieve the styrenes
difunctionalization reaction with high efficiency. As a contrast, to investigate the
effect of acid-base property of zeolite support on the catalytic activity, the basic
mesoporous titanosilicate ETS-10 zeolite (ETS-10-M) that consists of SiO4 tetrahedral
4

and TiO6 octahedral linked by corner-sharing oxygen atoms was also synthesized. For
this material, each TiO6 unit contributes two minus charges to framework, leading to
the framework oxygen atoms in TiO6 unit bearing negative charges that are regarded
as the basic sites [24-26]. The ETS-10-M and acid-free mesoporous Silicalite-1
(Silicalite-1-M) assembled Cu catalysts (Cu/ETS-10-M and Cu/Silicalite-1-M) were
also prepared and applied in the styrenes functionalization reaction. The Cu/Beta-M
catalyst shows higher activity and target product selectivity than Cu/ETS-10-M,
Cu/Silicalite-1-M, Cu/Beta, Cu/Silicalite-1 and homogeneous CuCl catalysts. This
feature is due to the fact that the Brönsted acid and highly dispersed Cu species play a
synergistic catalysis on this difunctionalization reaction. The Brönsted acid can
chemisorb and activate the C=C bond in styrenes, and the small Cu species can
catalyze the F-N(SO2Ph)2 transformation into N-electrophile of -N(SO2Ph)2, which
subsequently attacks the activated C=C bond, enhancing the reaction activity and
product selectivity. In addition, the abundant mesopores in the Cu/Beta-M could not
only favor the diffusion of the reactants and products, but also enhance the
accessibility of the reactants to the active sites, further improving its catalytic activity.
2. Experimental Section
2.1 Material synthesis.
Mesoporous Beta zeolite (Beta-M) was hydrothermally synthesized from an
aluminosilicate
gel
with
a
molar
composition
of
1.9Na2O/1.0Al2O3/31.5SiO2/5.9TEAOH/1.57RCC/319.0H2O, where RCC was a
random cationic copolymer that contained quaternary ammonium groups and that was
synthesized according to our previous work [27]. In a typical run, 0.528 g NaOH and
1.128 g NaAlO2 were dissolved into 28.4 mL deionized water with stirring for 30 min.
5

Beta seeds (1.32 g) were added to this solution with stirring for 30 min, followed by
addition of 16.7 mL tetraethylammonium hydroxide (TEAOH, 35 wt.%) with stirring
for 1 h. Subsequently, 13.2 g silica gel was added and a homogeneous gel was formed.
After stirring at room temperature for 4 h, 7 mL RCC was added to this gel. Finally,
the aforementioned gel was transferred into a Teflonlined stainless steel autoclave and
placed in an oven for 7 days at 140 °C. The resulting product was filtered, washed,
and dried at 100 °C overnight and calcined in air at 550 °C for 4 h. The H-form
Beta-M sample (HBeta-M) was prepared through two consecutive ion-exchanging in
aqueous ammonium nitrate solution (1 M, 80 °C, 4 h, 10 mL per gram of zeolite) and
followed by drying and calcination (450 °C, 4 h). The acidic mesoporous ZSM-5
(ZSM-5-M) was synthesized according to the previously reported literature [28].
Dealuminated Beta-M sample was prepared as follows: 1 g Beta-M zeolite powder
was added into the nitric acid solution (1 M, 20 mL) under stirring, and this mixture
was stirred at 80 °C for 8 h. After that, this mixture was filtered, washed, dried and
calcined at 550 °C for 2 h. The obtained solid was labelled as D-Beta-M.
Mesopore-free Beta zeolite was synthesized according to the reported literature [29].
The Si/Al ratio of the as-prepared Beta-M and D-Beta-M is 18 and 117, respectively,
which is determined by inductively coupled plasma (ICP) analysis performed on a
Perkin-Elmer 3300DV emission spectrometer.
Mesoporous Silicalite-1 zeolite (Silicalite-1-M) was hydrothermally synthesized
from a silicate gel according to our previous work [30]. In a typical run, 5.0 mL of
tetraethyl orthosilicate was added into the mixture of 16.0 mL tetrapropylammonium
hydroxide (25 wt.%) and 29.0 mL deionized water with stirring for 180 min. Then,
5.0 mL N,N-dimethyl-N-octadecyl-N-(3-triethoxysilylpropyl) ammonium bromide
([(C2H5O)3SiC3H6]N(CH3)2C18H37]Br, TBAOT) was added to this mixture with
6

stirring for 1 h. Finally, the mixture was transferred into a teflonlined stainless steel
autoclave and placed in an oven for 3 days at 180 °C. The resulting product was
filtered, washed, and dried at 100 °C overnight and calcined in air at 550 °C for 4 h.
The mesopore-free Silicalite-1 zeolite was synthesized by the same procedure except
for the addition of the TBAOT.
Mesoporous ETS-10 zeolite (ETS-10-M) was hydrothermally synthesized from a
titanosilicate
gel
with
a
molar
composition
of
TiO2:5.7SiO2:4.7Na2O:1.3K2O:163H2O:0.7TBAOT. The TBAOT was acted as a
mesoscale template. The typical procedure of the synthesis process referred to our
previously published work [31].
The Cu catalyst was prepared by simple wet impregnation method. Typically, the
support was impregnated by copper nitrate aqueous solution, and was placed under air
atmosphere for 12 h, and then dried at 100 °C for 12 h followed by calcination in air
at 450 °C for 4 h. The Cu loading in the sample was 2 wt.%. The obtained catalyst
samples were denoted as Cu/Beta-M, Cu/ETS-10-M, Cu/Silicalite-1-M, Cu/Beta,
Cu/Silicalite-1, Cu/HBeta-M, Cu/ZSM-5-M, Cu/D-Beta-M and the Cu content in
these catalysts was 1.9, 2.1, 2.0, 1.9, 2.0, 2.1, 1.9 and 2.1 wt.%, respectively, which
was analysed by inductively coupled plasma (ICP) analysis performed on a
Perkin-Elmer 3300DV emission spectrometer.
2.2 Characterization
X-ray powder diffraction (XRD) pattern was recorded on a D/MAX 2500/PC powder
diffractometer (Rigaku) using a Cu Ka radiation source. Nitrogen physisorption was
carried out at liquid nitrogen temperature (-196 °C) on a Micromeritics ASAP 2020M
apparatus. The sample was degassed at 200 °C for 8 h before the measurement.
7

Specific surface area was calculated from the adsorption data using the
Brunauer-Emmett-Teller (BET) equation. Mesopore size distribution was obtained
using the desorption data according to the Barrett–Joyner–Halenda model.
Scanning electron microscopy (SEM) was performed a field-emission
scanning electron microscope (SUPRA55). Transmission electron microscopy
(TEM) image was obtained on a JEM-2100 microscope with a limited line resolution
capacity of 1.4 Å at 200 kV. X-ray photoelectron spectroscopy (XPS) experiment was
performed on an ESCALAB MK II system.
The acidity of the sample was measured using temperature-programmed
desorption of ammonia (NH3-TPD) on a Micromeritics ASAP 2920 instrument.
The analysis method was the same to the previous published work [29]. The
desorbed NH3 was collected in dilute hydrochloric acid and then titrated with a
dilute sodium hydroxide solution to determine the total acidic site density of the
materials. The acidic nature (Brönsted/Lewis) of the zeolite was investigated
through pyridine adsorption infrared spectroscopy (Py-IR) on a Bruker
TENSOR 27 spectrophotometer equipped with a reactor cell. The experiment
procedure was followed: the sample (20 mg) was pressed into self-supporting
wafers and degassed under vacuum (1 × 10-2 Pa) at 200 °C for 2 h, and
subsequently exposed to pyridine vapour after being cooled to 30 °C. The
Py-IR spectrum was then recorded at 30 °C after the sample was vacuumed at
30 °C for 30 min. The basic property of the ETS-10-M sample was
characterized by pyrrole adsorption infrared spectroscope. The pyrrole molecule
can be chemisorbed on the framework oxygen atom bearing negative charges
that are regarded as Lewis basic sites [26,32], leading to a decrease in the IR
NH-stretching vibration from 3430 cm-1 for the pure pyrrole down to around
8

3200-3400 cm-1 for pyrrole interacted with the framework oxygens. The shift of
the IR NH-stretching vibration band is used as an evaluation of the zeolite basic
strength. The IR spectrum of the pyrrole adsorbed on ETS-10-M was also
measured on the same apparatus.
13
The C solid-state NMR spectra of chemisorbed zeolite catalyst samples as well
as corresponding pure substrate were obtained on a Bruker AVANCE III 400WB
spectrometer operated at 9.4 T with a frequency of 100.62 MHz. The chemisorbed
zeolite catalyst sample was pre-outgassed overnight at 100 °C and 10-1 Pa and then
packed into a 4 mm zirconia rotor. The contact time was 1.5 ms, and the recycle delay
was 0.5 s for the chemisorbed zeolite catalyst sample, and 60 s for pure substrate,
1
respectively, which were adjusted to three times H T1 values. A ramp-contact and
13
spinal 64 decoupling pulse program was used. The numbers of scans of C CP-MAS
were 156,000 for the chemisorbed zeolite catalyst sample, and 16 for pure substrate,
with rotor spinning rates of 14 kHz. The spectra were calibrated using the methine
carbon atoms of adamantane as an external standard (δ=29.47 ppm).
The chemisorption of the p-methylstyrene reactant on the Cu/Beta-M catalyst was
investigated by infrared spectrum (IR) on a Nicolet iS50 instrument. Before
measurement, the adsorbed sample was evacuated to 10-3 Pa at 50 °C for 24 h, and
then the temperature was cooled to room temperature. The spectrum was obtained in
the transmittance mode and was shown after subtraction of a KBr background
spectrum. For comparison, the IR spectra of corresponding pure substrates were also
obtained through dropping the pure substrates on KBr piece at room temperature.
2.3 Activity test
All reagents unless otherwise noted were obtained from commercial sources (purity >
99%) and used without further purification. The intermolecular aminoazidation of
9

p-methylstyrene with N-fluorobenzenesulfonimide (NFSI) and azidotrimethylsilane
(TMSN3) was carried out in a tubular reactor with N2 atmosphere. Typically, 5 mg
catalyst, 0.3 mmol p-methylstyrene (1a), 0.42 mmol NFSI (2a), 0.6 mmol TMSN3 (3a)
and 2.0 mL DCE as solvent were placed into a Schlenk tube (10 mL). The reaction
was proceeded at 70 °C for desired time.
2.4 Adsorption experiment
The adsorption experiment was carried out in a Schlenk tube (25 mL) with a magnetic
stirrer bar. For example: 1 mmol substrate was dissolved in 50 mL DCE. After stirring
homogeneously, 7.8 g this solution and 0.15 g zeolite catalyst were added into
Schlenk tube. The mixture was heated at 80 °C for 3 h under stirring with N2
atmosphere. After cooling to room temperature, the catalyst was separated by
centrifugation and filtration to obtain the liquid phase. The liquid phase was analyzed
by an Agilent 7890B gas chromatograph equipped with a flame ionization detector.
The solid sample was washed with DCE as many times as possible to remove the
physically adsorbed substrate, and then the resulting solid sample was dried at 50 °C
for 24 h and underwent degassing treatment before IR and solid state NMR
characterizations.
2.5 DFT Calculation
All calculations were performed with the Gaussian 09 package [33]. Density
functional with B3LYP functional and the 6-31G(d) was used to perform geometry
optimization of all the minima. Default convergence criteria were used. The
10

vibrational frequency calculations were conducted at the same level of theory as
geometry optimization to confirm whether each optimized structure is a minimum.
The silicon atoms and related O atoms in the model were fixed in order to maintain
the shape of the cluster during geometry optimizations [34,35], and the rest of the
atoms of the cluster were allowed to relax in the calculations. To obtain the accurate
interaction energies, the single point energies are obtained from the
ωB97X-D/6-311+G(2df,2p). The interaction energy (ΔEinter) is calculated by the
equation ΔEinter=ΔEzeolite-mol-ΔEmol-ΔEzeolite, where ΔEzeolite-mol is the total energy of the
zeoilite system after the adsorption of molecules, ΔEzeolite is the energy of zeolite and
ΔEmol is the energy of molecule.
3. Results and Discussion
3.1 General properties of the Catalysts
As shown in Fig. 1a, the XRD patterns of Beta-M, Silicalite-1-M and ETS-10-M
show the well-resolved peaks in the range of 5-50° associated with Beta, MFI and
ETS-10 structure, respectively. After introduction of metal Cu into support,
diffraction peaks related to the crystalline phases of the Cu species, such as CuO
(PDF: 45-0937) and Cu2O (PDF: 34-1354) [36,37], are not detected in the XRD
patterns of the Cu/Beta-M, Cu/Silicalite-1-M and Cu/ETS-10-M (Fig. 1b), implying
that the Cu species with small particle size dispersed in the zeolite or the Cu species
located in ion-exchanged sites. A similar phenomenon was also observed in the XRD
patterns of ZSM-5-M and Cu/ZSM-5-M (Fig. S1). In contrast, new peaks of CuO
(38.7°) and CuAlO2 (39.6°) phases were observed in the XRD pattern of Cu/Beta
sample, and the CuO (38.7°) phase was also observed for the Cu/Silicalite-1 XRD
11

pattern (Fig. S2). The nitrogen sorption isotherms of the Beta-M, Silicalite-1-M and
ETS-10-M exhibit hysteresis loops at a relative pressure of 0.45-0.95, which are
assigned to the presence of mesostructure in the zeolite crystals (Fig. 2a) [38]. The
mesopore size of the Beta-M, Silicalite-1-M and ETS-10-M is mainly centered at 6.5,
11.3 and 5.7 nm (Fig. 2a, insert), respectively. After loading of Cu species, the
catalysts have the equivalent mesopore sizes as to the corresponding supports (insert,
Fig. 2b), indicating that the mesopore size was impregnable to introduction of the
metal Cu into the zeolite, and the Cu species with small size could be formed. Sample
textural parameters are presented in Table S1. The Cu catalysts have high external
surface areas (100-237 m2/g) and mesoporous volumes (0.12-0.32 cm3/g), which
would favor the diffusion of the reactant molecule.
To characterize the acid-base property of the supports, the NH3-TPD experiments
for Beta-M, Silicalite-1-M and ETS-10-M samples, and IR spectroscopy of pyrrole
adsorption experiment for ETS-10-M was performed. The Beta-M zeolite has two
desorption signals with curves around 225 and 373 °C, and Silicalite-1-M has very
weak TCD signal (Fig. 3). This result indicates that Beta-M has abundant weak and
strong acidity sites, and Silicalite-1-M is almost no acidity. While the ETS-10-M
sample displays a significant TCD signal centered at 209 °C, indicating the presence
of weak acidic sites. The acidity site density of the zeolite was also calculated based
on the NH3-TPD curve, and the total acidity site density of Beta-M and ETS-10-M is
497 and 310 µmol·g-1, respectively. To further inverstigate the acid nature (Lewis acid
or Brønsted acid) of the Beta-M and ETS-10-M, the Py-IR was performed (Fig. S3).
Both the IR spectra show two absorption bands at 1444 cm-1 and 1548 cm-1, which
belong to the vibrations of pyridine ring adsorbed on the Lewis and Brönsted acid
sites on Beta-M and ETS-10-M [23,39]. Nevertheless, the intensity of absorption band
12

for the Brönsted acid on ETS-10-M is very weak. The few Brønsted acid sites on
ETS-10-M could be attributed to the unsaturated coordination of the framework Ti,
leading to the presence of few titanol groups (TiOH) [40,41].
The IR spectroscopy of pyrrole adsorbed on ETS-10-M shows the characteristic
peaks of N-H stretching at 3220 cm-1 and the bands at 2922 and 2851 cm-1, revealing
the presence of basic sites on the ETS-10-M sample (Fig. S4) [26]. It is well known
that aluminosilicates zeolites are composed by SiO4 and AlO4 tetrahedra linked by
one oxygen atom. The replacement of a Si atom (Si4+) by an Al atom with a lower
valency (A13+) in a zeolite framework creates a negative charge which is neutralized
by a proton or a metal cation [26]. The framework oxygen atom bearing negative
charge generates the basic character of the zeolites [26,31]. Increasing the Al content
in zeolite can strengthen the basic characters [26]. While for Al-free silicate
Silicalite-1 zeolite that consists of SiO4 tetrahedra has a neutral framework, and thus,
the Silicalite-1 zeolite has almost no basic sites. In addition, compared with
aluminosilicate zeolite, the electronegativity of Ti is slightly lower than Al atom [25].
Furthermore, TiO6 octahedra in ETS-10 gives two minus charges, while AlO4
tetrahedra in aluminosilicate zeolite gives one minus charge. The electron density of
oxygen atom around Ti in ETS-10 is higher than that around Al in aluminosilicate
zeolite, leading to the basic strength stronger than aluminosilicate zeolite [25]. Thus,
in our case, the synthesized Silicalite-1-M should have no basicity, and the Beta-M
zeolite, with Si/Al ratio of 18, should have fewer basic sites than ETS-10-M zeolite.
The temperature-programmed desorption of CO2 (CO2-TPD) curves demonstrated
that the basic density of ETS-10-M is higher than that of Beta-M zeolite (Fig. S5).
The SEM image in Fig. 4a reveals that the Beta-M sample has an irregular spherical
morphology and the particle size is around 400-700 nm. The Silicalite-1-M sample
13

exhibits round cake morphology (size of 300-500 nm) with rough surface (Fig. 4b). In
contrast, ETS-10-M sample presents large particle (Fig. 4c). The quadrate shape of
ETS-10-M with particle size of 5 μm is seemingly composed by nanosheets with a
thickness of 500 nm (Fig. 4c). To survey the porous structure, and metal particle size
on the catalysts, the TEM characterization of the thin-sectioned catalyst was
performed. TEM image shows that Cu/Beta-M has abundant mesopores (bright spots),
and very small metal nanoclusters (black dot, less than 1 nm) are inconspicuously
scattered in the mesopore channels and on the outer surface (Fig. 5a). Interestingly,
when the zone in Fig. 5a was exposed by a strong electron beam and the zeolitic
crystal fringes were destroyed, some particles (2-5 nm) with a lattice fringe of 0.22
nm corresponding to the (111) crystal plane of Cu2O are clearly observed (Fig. 5b)
[42]. This result could be attributed to the aggregation of the small subnanometric
metal Cu oxide clusters within the micropores or on the mesopores. The
temperature-programmed reduction of H2 (H2-TPR) analysis also demonstrated that
besides Cu+ and Cu2+ ions, some Cu species in the form of CuO are present in the
Cu/Beta-M catalyst (Fig. S6, discussed in Supporting Information). In addition, to
further verify the presence of Cu species, Cu/Beta-M was studied by X-ray energy
dispersive spectroscopy (EDS, Fig. S7a). The Fig. 5c shows the EDS analysis result,
which confirms the presence of Cu in the crystalline microporous structure, and the
Cu content is 1.97 wt.% (Fig. 5c), which is close to the Cu loading for the catalyst
preparation. For the TEM image of Cu/Silicalite-1-M catalyst, the elliptic bright spots
are the mesopores, and many metal nanoclusters are observed that are sparsely
dispersed on the mesoporous surface or in the micropore channels (Fig. 6a). These
nanoclusters were gathered to form nanoparticle when the zeolite framework structure
was destroyed by the strong electron beam (2-3 nm, Fig. S8). Partial nanoparticles
14

existed in the form of Cu2O in the Cu/Silicalite-1-M (Fig. 6b). While for
Cu/ETS-10-M sample, obvious zeolite lattice fringes are observed and metal clusters
are barely visible (Fig. 6c). But some metal nanoparticles appeared when the crystal
lattice fringes were destroyed (Fig. 6d), indicating that very small metal nanoclusters
existed in the micropores and located on mesopores channels of Cu/ETS-10-M sample.
All these results indicate that very small Cu oxides nanoclusters existed in the
micropores or located on the mesopores of the zeolite catalysts
To investigate the electron state of the Cu species in the different catalysts, the XPS
analysis was performed. For Cu/Beta-M catalyst, the XPS spectrum of the Cu2p3/2
shows two binding energies approximate at 933.2 and 935.6 eV (Fig. 7), which is
characteristic of the Cu+ and Cu2+ species [43], respectively. While for the
Cu/Silicalite-1-M and Cu/ETS-10-M, the binding energy of the Cu+ appears at 933.5
and 933.6 eV, indicating that Cu+ species on the Cu/Silicalite-1-M and Cu/ETS-10-M
have low electron cloud density. From the XPS spectra, it can be concluded that the
Cu+ content in catalysts is higher than Cu2+ content, which could be caused by the
presence of strong metal-support interaction on the catalyst.
3.2 Catalyst Performance
Table 1 gives the catalytic performance of different catalysts in the difunctionalization
of p-methylstyrene with NFSI and TMSN3. Notably, the acidic Cu/Beta-M catalyst
gives high p-methylstyrene conversion (90%) and target product selectivity (91%),
compared to acid-free Cu/Silicalite-1-M (61 and 82%), basic Cu/ETS-10-M (6 and
83%), mesopore-free Cu/Beta (40 and 76%) and Cu/Silicalite-1 (25 and 84%) at the
reaction time of 35 min (entries 1-5). When the reaction time reached 45 min,
complete conversion (100%) and high product selectivity (90%) are achieved with the
15

Cu/Beta-M catalyst (entry 6). As a comparison, we also tested the activity of the CuCl
salt, which has been reported in the literature [8]. Unfortunately, the CuCl catalyst
shows no activity in the absence of the 1,10-phenanthroline ligand (entry 7), which is
in line with the literature [8]. When the Cu content in the reaction mixture was the
same as that of the 5 mg Cu/Beta-M, the reaction still does not occur even the addition
of phenanthroline ligand into the reaction mixture (entry 8). When the Cu content in
the reaction mixture was four times higher than that in the 5 mg Cu/Beta-M catalyst
and the 1,10-phenanthroline ligand was added, the p-methylstyrene conversion on
CuCl catalyst was only 54% (entry 9), which is still much lower than that when using
the Cu/Beta-M catalyst (90%, entry 1). In the control experiment using Beta-M
(without Cu) as a catalyst, no noticeable conversion of p-methylstyrene was observed
(entry 10), confirming that Beta-M alone was inactive in this transformation and Cu
was indispensable. The turnover frequency (TOF) was also calculated based on the Cu
content. The TOF (0.086 s-1) of Cu/Beta-M is eight times higher than that of the CuCl
catalyst (0.01 s-1, entries 1 and 9). These results indicate that the Cu/Beta-M catalyst
shows superior catalytic performance; and in the homogeneous CuCl catalytic system,
the 1,10-phenanthroline ligand and high CuCl content are necessary.
The scope of the styrenes aminoazidation over Cu/Beta-M was also investigated,
and the results are shown in Table 2. Prolongation of the reaction time to 60 min, all
substrates presented high conversions and good product selectivities. Styrene and
styrenes with electron-donating groups such as tert-butyl, methyl, 2,4,6-trimethyl
were almost completely transformed in conversions of 94-100% (entries 4a-4e). In
contrast, relatively low conversions of the styrenes with electron-withdrawing groups
on Cu/Beta-M catalyst were obtained (83-91%, entries 4f-4l). Meanwhile, the
substrates of 4-Vinylbiphenyl and 1-vinylnaphthalene with large molecular
16

dimensions were also suitable for this transformation and gave good conversions (4m,
4n). Notably, when the allylbenzene was used as a substrate, the aminoazidation of
the allylbenzene did not occur (4o).
3.3 Reason for the high activity of the Cu/Beta-M catalyst
3.3.1 The acidic sites promoted C=C bond activation on the Cu/Beta-M catalyst
As shown in Table 1, Cu/Beta-M catalyst shows higher activity and product
selectivity than Cu/Silicalite-1-M in the difunctionalization of styrenes with NFSI and
TMSN3, and the Cu/ETS-10-M catalyst is almost no activity. As discussions in
characterization above, the Cu loadings in the Cu/Beta-M, Cu/Silicalite-1-M and
Cu/ETS-10-M catalysts are the same to each other; Cu species with size of less than 1
nm are highly dispersed on the three catalysts; and both of Cu+ and Cu2+ species are
present on the three catalysts. The essential difference among these catalysts is the
acidic-basic properties of the supports. Compared with acid-free Cu/Silicalite-1-M
and basic Cu/ETS-10-M catalysts, the Cu/Beta-M catalyst has abundant acidic sites,
and the acid density of Beta-M is 497 µmol·g-1. The acidic or basic sites on catalysts
should influence the chemisorption and activation of reactant on the catalysts. This is
supported by the experimental results of p-methylstyrene adsorption on different
catalysts, as shown in Table 3.
The adsorption quantity of p-methylstyrene on the acidic Beta-M (64 mg/gcat.) is
much higher than that on the acid-free Silicalite-1-M (27 mg/gcat.) and basic
ETS-10-M (0.6 mg/gcat.) zeolite. After introducing Cu into the supports, the
adsorption quantity of the p-methylstyrene on catalyst is smaller than corresponding
support, this could be due to the fact that a fraction of Cu species could cover the
adsorption sites. Nevertheless, the acidic Cu/Beta-M catalyst still has higher
17

p-methylstyrene adsorption capacity (52 mg/gcat.) than Cu/Silicalite-1-M (7 mg/gcat.)
and Cu/ETS-10-M (0.6 mg/gcat.) catalysts. Notably, the 1a adsorption capacity of the
catalyst is proportional to the reaction activity of the catalyst. These results confirm
that the acidic sites in catalyst can benefit the adsorption of the p-methylstyrene, and
the basic sites in catalyst disfavor the adsorption of the p-methylstyrene.
To illustrate the adsorption properties of p-methylstyrene on the acidic Cu/Beta-M
catalyst and the intrinsic interaction of the p-methylstyrene molecule with acidic site,
the adsorbed sample (p-methylstyrene adsorbed on the Cu/Beta-M catalyst,
designated as Cu/Beta-M-pm) was investigated by 13C NMR and IR spectroscopy.
13
The C NMR spectrum of pure p-methylstyrene (pm) shows the peaks at 136.7 and
112.8 ppm characteristic of the carbons in the C=C bond (Fig. 8a). However, these
two peaks are not found in the low-field region and two new peaks appear in the high
field region, at 50.2 and 39.0 ppm for the Cu/Beta-M-pm sample (Fig. 8a), which is
characteristic of its counterpart with C(sp3)-C(sp3) single bonds. These results confirm
that the p-methylstyrene was chemisorbed on the acidic hydroxyl sites (Brönsted acid)
of the Cu/Beta-M catalyst, and the C=C bond in p-methylstyrene should be activated
by acidic hydroxyl groups. The IR spectra of pure p-methylstyrene and
Cu/Beta-M-pm also verified this inference. From Fig. 8b, the absorption bands at
2920 and 2863 cm-1 are assigned to C-H stretching vibrations of the methyl group in
p-methylstyrene [44]. When p-methylstyrene was adsorbed on Cu/Beta-M catalyst
(Cu/Beta-M-pm), the IR spectrum of Cu/Beta-M-pm gives two new absorption bands
at 2957 and 2871 cm-1. By comparison with the IR spectrum of pure p-methyl
ethylbenzene (pme), the absorption bands at 2963 and 2872 cm-1, corresponding to
C-H stretching vibrations in -CH2, were similar to the new bands at 2957 and 2871
cm-1 for Cu/Beta-M-pm. These results further suggest that the C=C bond can be
18

activated by Brönsted acid after p-methylstyrene was adsorbed on the Cu/Beta-M
catalyst.
The chemisorption behavior of the p-methylstyrene on the Brönsted acid and the
interaction energy (ΔEinter) of the C=C bond in p-methylstyrene with hydroxyl group
on Cu/Beta-M and Cu/Silicalite-1-M was calculated based on the density functional
theory (DFT, Table 4). The optimized structure of the p-methylstyrene molecule
adsorbed on the Cu/Beta-M and Cu/Silicalite-1-M is shown in Table S2. The ΔEinter of
the C=C bond in p-methylstyrene with the silicon hydroxyl group of Silicalite-1-M is
-46.5 kJ/mol. It implied that the silicon hydroxyl group could also interact with the
C=C bond in p-methylstyrene, promoting the reaction. The additional experiment
shows that Cu/Silicalite-1-M gives 87% conversion and 85% selectivity when the
reaction was extended to 5 h (Table S3, entry 1). Nevertheless, the ΔEinter of the C=C
bond in p-methylstyrene with the acidic hydroxyl group of Beta-M (-72.2 kJ/mol) is
much higher than that of the C=C bond with the silicon hydroxyl group of
Silicalite-1-M (-46.5 kJ/mol, Table 4). These results indicate that p-methylstyrene is
easily chemisorbed on the Brönsted acid of the Cu/Beta-M catalyst and the C=C bond
can be activated, as compared with that on silicon hydroxyl group in
Cu/Silicalite-1-M. The activated C=C bond in styrenes easily reacts with NFSI and
TMSN3, leading to the enhancement of the reaction activity.
On the other hand, the C=C bond electron density of the styrenes should be
influenced by their substituent group properties, which results in the different
chemisorption behavior of the styrenes and then affects the reaction activity of the
substrates. This is supported by the additional adsorption experiment and ΔEinter of the
styrenes with different substituent groups on the Cu/Beta-M (Table 4). It is clearly
that the adsorption quantity of the styrenes with electron-donating groups
19

(p-methylstyrene and p-tert-butyl-styrene) is much higher than styrene and styrenes
with electron-withdrawing groups (p-chloro-styrene and p-fluoro-styrene, Table 4).
Correspondingly, the ΔEinter of the p-methylstyrene and p-tert-butyl-styrene is much
higher than styrene, p-chloro-styrene and p-fluoro-styrene, and the dC-H distance
between p-methylstyrene (p-tert-butyl-styrene) and acidic hydroxyl group is shorter
than that between p-chloro-styrene (p-fluoro-styrene) and acidic hydroxyl group on
the Cu/Beta-M (Table 5). Expectedly, the substrate reaction activity on the Cu/Beta-M
follows the order of p-tert-butyl-styrene ˃ p-methylstyrene ˃ styrene ˃
p-chloro-styrene ˃ p-fluoro-styrene (Table 4). These results implied that the electron
density of the C=C bond in styrenes strongly influences the chemisorption behavior
and reaction activity of the substrates. The electron-donating groups on the benzene
ring of styrene can enhance the electron density of the C=C bond, resulting in the
relatively strong interaction between C=C bond and the acidic hydroxyl group of the
catalyst and further enhancement of the reaction activity. On the contrary, the
electron-withdrawing groups on the benzene ring of styrene can decrease the electron
density of the C=C bond, leading to the relatively weak interaction between C=C
bond and acidic hydroxyl group and a decrease in the reaction activity of the
substrate.
The effect of the electron density of the C=C bond in styrenes on chemisorption
behavior can further be proofed by the additional adsorption experiment of the
methylbenzene and allylbenzene molecules. The methylbenzene molecule was not
absorbed on the Cu/Beta-M catalyst (Table 4), implying that the delocalized π
electron cloud in the benzene ring is difficulty to be interacted with the acidic sites.
When the allylbenzene that the C=C bond is isolated by methylene and does not
conjugate with the π-system in benzene ring was used as an adsorbate, the adsorption
20

quantity of the allylbenzene on Cu/Beta-M catalyst is also almost none (Table 4).
Coincidentally, the C=C bond in allylbenzene molecule cannot be activated and the
allylbenzene was not transformed in this reaction (4o, Table 2). From these results, we
can conclude that the conjugation between the benzene ring and C=C bond in styrenes
is very important for the substrate absorption and activation. In this case, the electron
density between the benzene ring and C=C bond in styrenes was delocalized, and the
p-orbital in C=C bond would get electron from the π-system of benzene ring. Thus,
styrenes can be chemisorbed on the Cu/Beta-M catalyst through the C=C bond
interaction with Brönsted acid (Si-OH-Al).
3.3.2 The small Cu species facilitated the transformation of NFSI into
N-electrophile on the Cu/Beta-M catalyst
Besides the Brönsted acid, the highly dispersed Cu species in the Cu/Beta-M also play
a critical role in the enhancement of activity and selectivity for the difunctionalization
reaction. Form Table 1, in the CuCl-salt catalytic system, only in the presence of
1,10-phenanthroline ligand, the reaction can proceed, but CuCl catalyst has a low
activity. This phenomenon indicated that the 1,10-phenanthroline ligand was
indispensable. In the case of Cu/Beta-M-catalytic system, Cu/Beta-M catalyst shows
high activity without 1,10-phenanthroline ligand. The role of small Cu species in the
Cu/Beta-M-catalytic systems can be interpreted as followed.
For the aminoazidation of styrenes, the transformation of the NFSI into
electrophilic -N(SO2Ph)2 on metal Cu catalyst via a redox process is necessary [8,45].
Typically, when the low valency Cu species reduce NFSI, it could be transformed to
high valency Cu species. During this process, the F-N bond in F-N(SO2Ph)2 molecule
is cleaved, yielding two covalent bonds that bonded to the metal Cu and form a
F-Cu-N(SO2Ph)2 complex. This process is oxidation addition. Afterthat, the
21

-N(SO2Ph)2 in the F-Cu-N(SO2Ph)2 easily leaves and transforms into a electrophilic
reagent, resulting in the formation of the Cu species with a low valency, which is
reductive elimination. Therefore, under the reaction conditions, the Cu species should
have a variable electron structure, whose electronic valence can be reversibly varied
to activate the NFSI. This reversible transformation in the homogeneous CuCl-salt
catalytic system without ligand is difficulty. This is due to the fact that the electronic
structure of the mononuclear metal Cu is not easy to be changed in the absence of
ligand. In the presence of 1,10-phenanthroline ligand, the CuCl can be coordinated
with 1,10-phenanthroline that has enriched charge to form various Cu-complexes with
a variable electron structure under reaction conditions. In this case, F-N(SO2Ph)2 can
be activated to form to N-electrophile. The literature [8] also illustrated that the CuCl
coordinated with 1,10-phenanthroline can form different valence Cu-complexes.
Avelino Corma has reported that the electronic structures of the metal-complex
strongly depend on the ligands [46].
In the case of Cu/Beta-M catalyst, the Beta-M support with anion framework
characteristics has large amounts of acidic hydroxyl groups, and the framework
oxygen atoms in acidic hydroxyl groups (Si-O(H)-Al) present a negative
characteristic. The highly dispersed Cu species with positive charge are more inclined
to interplay with framework oxygen atoms, leading to a strong metal-support
interaction and electron transfer from each other. Thus, the negative charge
framework of Beta-M can act as a rigid electron ligand for the Cu species, which
could favor the reversible transformation of Cu species with variable electronic
valence under reaction conditions to realize the activation of the NFSI. The zeolitic
framework is regarded as an electron reservoir, which may either support or impair
electronic to metal clusters during the reaction, leading to improvement of the
22

catalytic activity [47]. The adsorption experiment shows that the NFSI (2a) adsorption
capacity of Cu/Beta-M is higher than that of Beta-M (Table 3), indicating that the
introduced Cu species in the catalyst benefit the adsorption of 2a molecule. The
adsorbed 2a molecule could interact with Cu species, which favors the 2a
transformation into N-electrophile under reaction conditions.
From the discussions above, it can be concluded that the Brönsted acid and the Cu
species, such as the surface metal atoms of the CuO clusters and Cu+ (Cu2+) ions on
Cu/Beta-M have a synergistic effect in the enhancement of reaction activity in this
aminoazidation of styrenes. Brönsted acidic sites chemisorb the styrenes and activate
the C=C bond. Meanwhile, the active Cu species can benefit the NFSI transformation
into N-electrophile of -N(SO2Ph)2 through a redox process. The formed -N(SO2Ph)2
electrophile easily attack the activated C=C bond, leading to the reaction to proceed
smoothly. Both Brönsted acid and active Cu sites are indispensable for this reaction
with high efficiency.
3.3.3 The facilitation role of the porous structure in the reaction
Besides the Brönsted acid and active Cu species, the porous structure in catalyst also
plays an important role in this reaction. Compared with mesopore-free Cu/Beta and
Cu/Silicalite-1, corresponding Cu/Beta-M and Cu/Silicalite-1-M have high catalytic
activity. This could be caused by the difference in the porous structure. Because the
molecular dimension (1.2 nm) of the target product 4a is much larger than micropore
size of Beta (7.6×6.4 Å) and Silicalite-1 (5.1×5.5Å) [48], thus, the reaction only
occurs in the mesoporous channels and on the outer surface of the catalysts. The
abundant mesopores and high mesopore surface area in the Cu/Beta-M and
Cu/Silicalite-1-M could not only favor the diffusion of the reactants and products, but
also enhance the accessibility of the reactants to the active sites, leading to them
23

presenting higher activity than microprous zeolite supported Cu catalysts. In addition,
the Cu/ETS-10-M has lower mesoporous surface area (89 m2/g) and mesopore volume
(0.12 cm3/g) than Cu/Beta-M (168 m2/g and 0.22 cm3/g ) and Cu/Silicalite-1-M (257
m2/g and 0.22 cm3/g, Table S1), which could also lead to Cu/ETS-10-M with low
catalytic activity.
As for Beta-M zeolite, it has interconnected mesopore and micropore channels,
and the micropore mouth opens on the mesopore surface (Fig. S9). In this manner, the
micropore openings can provide abundant Brönsted acid sites for mesopore surface.
When the bulky p-methlystrene and NFSI molecules diffused into the mesopores
channels, the p-methlystrene could be accessible to the Brönsted acid sites on the
micropore openings and the C=C bind in p-methlystrene can be activated by Brönsted
acidity. Meanwhile, the NFSI is facile to be activated by Cu species located in the
mesopores. The Brönsted acid sites located on the micropore mouth opening to the
mesopore surface could facilitate the catalytic conversion of the bulky reactant
molecules.
3.4 Reaction mechanism
According to the above analyses and literature reports [8,45], the difunctionalization
mechanism of the C=C bond in p-methylstyrene with NFSI and TMSN3 on
Cu/Beta-M is illustrated in Fig. 9. First, the C=C bond in p-methylstyrene (1a) can be
activated through the chemisorption of the C=C bond on Brönsted acidic sites on the
Cu/Beta-M catalyst to form the intermediate B. At the same time, oxidative addition
of NFSI occurs on Cu active sites in Cu/Beta-M catalyst and intermediate C is formed.
Subsequently, C undergoes a reductive elimination reaction to generate an
N-electrophile of -N(SO2Ph)2 [45], which attacks the activated C=C bond to form D.
24

-
In the presence of N3 , which comes from TMSN3, D can be transformed to the final
product 4a.
3.5 Catalyst recycle performance
To inverstigate recycle performance of the catalyst, the recycling experiment of
Cu/Beta-M catalyst was carried out. Fig. S10 shows the p-methylstyrene conversion
and product selectivity on the reused Cu/Beta-M catalyst. Cu/Beta-M exhibited a
good catalytic stability. After the fifth cycle, the p-methylstyrene conversion reached
85%, and the target product selectivity was still over 87%. Furthermore, the Cu
content in the spent Cu/Beta-M slighly was reduced to 1.8 wt.%. These results
indicate that the Cu/Beta-M has a good catalytic stability. In addition, the acidic
Cu/HBeta-M, Cu/HBeta-MI (prepared by ion-exchanged method), and Cu/ZSM-5-M
were successfully applied in this reaction, giving high p-methylstyrene conversion and
product selectivity (entries 2-4, Table S3, the detailed discussion placed in the
Supporting Information). While, the dealuminated Beta-M supported Cu catalyst
(Cu/D-Beta-M) shows lower activity than Cu/Beta-M (entry 5, Table S3). This is due
to the fact that acidity amount of D-Beta-M zeolite was much lower than that of
Beta-M sample (Fig. S11). This result suggests that the Brönsted acidic sites do play a
facilitation role in the enhancement of the catalytic activity.
4. Conclusion
In summary, based on the unique adsorption properties of acidic zeolites for styrenes
and the interaction of metal clusters with zeolite, we synthesized a series of
mesoporous zeolites with different acid-base properties for the preparation of Cu
catalysts and applied them for styrenes difunctionalization reaction. The acidity of
25

Cu/Beta-M can induce styrene molecule to chemisorb on catalyst through interaction
of the C=C bond with Brönsted acidic sites, leading to C=C bond activation and
enhancing its functionalization activity. In addition, electron-donating groups on the
benzene ring of styrene can enhance the electron density on the C=C bond, resulting
in the stronger interaction between the C=C bond and Brönsted acidic site to further
improve the activity of the C=C bond. Meanwhile, the highly dispersed
subnanometric Cu oxides clusters interacted strongly with the zeolite framework,
which would benefit the mutual transformation of Cu species with different electronic
valence under reaction conditions, leading to the F-N(SO2Ph)2 easy transformation
to electrophilic -N(SO2Ph)2 reagent that reacts with the activated p-methylstyrene to
form the desired product. In addition, the mesopores in Cu/Beta-M enhance the
accessibility of the reactants to the Brönsted acid and active Cu species, further
improvement of catalyst activity.
Credit author statement
Tiandi Tang: Conceptualization, methodology, Writing - Review & Editing, Funding
acquisition.
Wenqian Fu: Validation, Investigation, Writing- Original draft preparation, Funding
acquisition
Chengyou Yin: Investigation
Yu Feng: Formal analysis
Lei Zhang: Formal analysis, Funding acquisition
Feng Cheng: Soft
Zhongxue Fang: Data Curation
Chaojie Zhu: Visualization
Declaration of interests
26

The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work was supported by the National Natural Science Foundation of China
(21776022, 21972011 and U1662139), the Natural Science Foundation of Jiangsu
Province of China (BK20150258), and PetroChina Innovation Foundation
(2018D-5007-0401), supported by the Project of Jiangsu Key Laboratory of Advanced
Catalytic Materials and Technology, Changzhou University
Appendix A. Supplementary material
Supplementary material related to this article can be found, in the online version, at
doi:https://doi.org/10.1016/j.apcata.2020.XXXXXX.
27

References
[1] F. Cardona, A. Goti, Nat. Chem. 1 (2009) 269-275.
[2] R.I. McDonald, G. Liu, S. Stahl, Chem. Rev. 111 (2011) 2981-3019.
[3] J.K. Cheng, T.P. Loh, J. Am. Chem. Soc. 137 (2014) 42-45.
[4] G.-S. Liu, Y.-Q. Zhang, Y.-A. Yuan, H. Xu, J. Am. Chem. Soc. 135 (2013)
3343-3346.
[5] B. Zhao, H. Du, S. Cui, Y. Shi, J. Am. Chem. Soc. 132 (2010) 3523-3532.
[6] T. Wu, J. Cheng, P. Chen, G. Liu, Chem. Commun. 49 (2013) 8707-8709.
[7] W. Kong, P. Feige, C. Nevado, Angew. Chem. Int. Ed. 52 (2013) 2469-2473.
[8] B. Zhang, A. Studer, Org. Lett. 16 (2014) 1790-1793.
[9] K. Matcha, R. Narayan, A.P. Antonchick, Angew. Chem. Int. Ed. 52 (2013)
7985-7989.
[10] X. Tang, C.S. Thomoson, W.R. Dolbier Jr., Org. Lett. 16 (2014) 4594-4597.
[11] A. Corma, H. Garcꢀa, F.X. Llabrés i Xamena, Chem. Rev. 110 (2010) 4606-4655.
[12] M. Yang, S. Li, Y. Wang, J.A. Herron, Y. Xu, L.F. Allard, S. Lee, J. Huang, M.
Mavrikakis, M. Flytzani-Stephanopoulos, Science 346 (2014) 1498-1501.
[13] W. Fu, Y. Feng, Z. Fang, Q. Chen, T. Tang, Q. Yu, T. Tang, Chem. Commun. 52
(2016) 3115-3118.
[14] P. Hua, K. Nakamura, H. Matsubara, K. Iyoki, Y. Yanaba, K. Okumura, T.
Okubo, N. Katada, T. Wakihara, Appl. Catal. A 601 (2020) 117661-117668.
[15] X. Wang, J. Xu, G. Qi, C. Wang, W. Wang, P. Gao, Q. Wang, X. Liu, N. Feng, F.
Deng, J. Catal. 345 (2017) 228-235.
[16] G. Feng, P. Cheng, W. Yan, M. Boronat, X. Li, J.-H. Su, J. Wang, Y. Li, A.
Corma, R. Xu, J. Yu, Science 351 (2016) 1188-1191.
28