RhNPs supported onN-functionalized mesoporous silica: effect on catalyst stabilization and catalytic activity

Article abstract of DOI:10.1039/d0dt04213g

Amine and nicotinamide groups grafted on ordered mesoporous silica (OMS) were investigated as stabilizers for RhNPs used as catalysts in the hydrogenation of several substrates, including carbonyl and aryl groups. Supported RhNPs on functionalized OMS were prepared by controlled decomposition of an organometallic precursor of rhodium under dihydrogen pressure. The resulting materials were characterized thoroughly by spectroscopic and physical techniques (FTIR, TGA, BET, SEM, TEM, EDX, XPS) to confirm the formation of spherical rhodium nanoparticles with a narrow size distribution supported on the silica surface. The use of nicotinamide functionalized OMS as a support afforded small RhNPs (2.3 ± 0.3 nm), and their size and shape were maintained after the catalyzed acetophenone hydrogenation. In contrast, amine-functionalized OMS formed RhNP aggregates after the catalytic reaction. The supported RhNPs could selectively reduce alkenyl, carbonyl, aryl and heteroaryl groups and were active in the reductive amination of phenol and morpholine, using a low concentration of the precious metal (0.07-0.18 mol%).

Full text of DOI:10.1039/d0dt04213g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
RhNPs supported on N-functionalized
mesoporous silica: effect on catalyst stabilization
and catalytic activity†
Cite this: Dalton Trans., 2021, 50,
3289
a
a
Israel T. Pulido-Díaz,
Alejandro Serrano-Maldonado,
a
a
Carlos César López-Suárez, Pedro A. Méndez-Ocampo,
Benjamín Portales-Martínez, Aída Gutiérrez-Alejandre,
and Itzel Guerrero-Ríos
b
c
a
Karla P. Salas-Martin
a
*
Amine and nicotinamide groups grafted on ordered mesoporous silica (OMS) were investigated as stabil-
izers for RhNPs used as catalysts in the hydrogenation of several substrates, including carbonyl and aryl
groups. Supported RhNPs on functionalized OMS were prepared by controlled decomposition of an
organometallic precursor of rhodium under dihydrogen pressure. The resulting materials were character-
ized thoroughly by spectroscopic and physical techniques (FTIR, TGA, BET, SEM, TEM, EDX, XPS) to
confirm the formation of spherical rhodium nanoparticles with a narrow size distribution supported on
the silica surface. The use of nicotinamide functionalized OMS as a support afforded small RhNPs (2.3 ±
0
.3 nm), and their size and shape were maintained after the catalyzed acetophenone hydrogenation. In
Received 10th December 2020,
Accepted 27th January 2021
contrast, amine-functionalized OMS formed RhNP aggregates after the catalytic reaction. The supported
RhNPs could selectively reduce alkenyl, carbonyl, aryl and heteroaryl groups and were active in the reduc-
DOI: 10.1039/d0dt04213g
tive amination of phenol and morpholine, using
a low concentration of the precious metal
rsc.li/dalton
(0.07–0.18 mol%).
nation of two or more of these stabilizers impact the catalytic
activity and selectivity observed in colloidal rhodium
Introduction
4
,5
Among metal catalyzed reactions, rhodium represents one of nanoparticles.
the most versatile and active metal catalysts in a plethora of favors the formation of anilines from nitroarene,
transformations. In the field of metal nanoparticles (MNPs), aryl alcohols from acetophenone and the complete hydro-
Varying the rhodium nanoparticle setup
8
–10
secondary
1
10
7
the activity attained by rhodium is related to its great π-acidic genation of aryl derivatives, to name a few examples.
character, applied in the activation of dihydrogen and in
Despite the great performance of RhNPs in colloidal
2
hydrogenation reactions. The use of stabilizers of MNPs to systems, their actual industrial application requires ease of
avoid the formation of bulk metals can increase the selectivity separation and recyclability of the RhNPs, enabled when the
3
–7
2
towards the hydrogenation of different functional groups.
nanoparticles are supported on solids. RhNPs have been sup-
Ligands, polymers, non-conventional media and a combi- ported over amorphous to highly ordered materials (graphene,
CNTs, mesoporous inorganic oxides). Other types of MNP
approaches have used the architecture of ordered mesoporous
a
Depto. de Química Inorgánica y Nuclear, Facultad de Química, Universidad
Nacional Autónoma de México, Av. Universidad 3000, 04510 CDMX, Mexico.
E-mail: itzelgr@unam.mx
materials (OMS) in order to form ultrafine metal
1
1,12
nanoparticles.
The most common methodologies used to
support MNPs comprise impregnation (wet or dry), deposition/
precipitation or laser deposition, whilst chemical methods
make use of the infiltration of metal salts in OMS followed by
a reaction with borohydrides, urea or hydrazine to reduce the
b
CONACYT, Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada,
Laboratorio Nacional de Conversión y Almacenamiento de Energía, Instituto
Politécnico Nacional, Calzada Legaría 694, Col. Irrigación, Ciudad de México,
1
c
1500, Mexico
Depto de Ingeniería Química, Facultad de Química, Universidad Nacional Autónoma metal salts. The advantages of supporting MNPs include stabi-
de México, Av. Universidad 3000, 04510 CDMX, Mexico
Electronic supplementary information (ESI) available: Characterization spectra
13–15
lity towards sintering,
as well as an active role in the cata-
†
lytic transformation, among other possibilities that to date are
the focus of several investigations both in industry and acade-
mia. Overall, the catalytic performance of the supported MNPs
of NAM (NMR, 2D), silicas (TGA, IR, BET), and RhNPs (IR, TGA, BET, SEM, TEM,
XPS, EDX) and the tables of selected catalytic reactions. See DOI: 10.1039/
d0dt04213g
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3289–3298 | 3289

View Article Online
Paper
Dalton Transactions
strongly depends on the nature of the interaction of MNPs tions by supported RhNPs is influenced by the coordinating
1
6–18
with the support and solvent.
within the pores of ordered mesoporous supports (OMS) these groups by pyrolysis is needed to afford active and selec-
showing remarkable activity. tive catalysts, an approach without thermal treatment is pro-
Thus, RhNPs can be found groups grafted on silica supports and that partial removal of
Regarding the catalytic applications, rhodium has been posed. In this regard, amine or pyridine functional groups
immobilized on silica type MCM (Mobil Composition of grafted on the surface of silica were evaluated towards the
Matter No. 41) by an ionic exchange reaction of Rh(III) with the stabilization of RhNPs after the catalytic hydrogenation of
1
9
template cation, and more recently an organometallic chem- different functional groups as well as multistep transform-
istry approach was used to support RhNPs with phosphine ations, with the focus on the selectivity and the recyclability of
2
0
ligands grafted on silica. One important challenge to address the catalytic systems.
is the formation of stable metal nanoparticles that do not
change throughout the catalytic reactions to extend the shelf
life. The robustness of the supported RhNPs has been assessed
by studying the recyclability of the catalyst through several
catalytic runs, or by observing morphological changes on the
Experimental
General
catalyst after the reaction.
All manipulations were carried out under a dry, oxygen-free
Among the catalytic applications of supported RhNPs, it is nitrogen atmosphere using standard Schlenk techniques.
possible to obtain important products such as amines through Organic solvents were distilled from an appropriate desiccant
a sustainable catalysis approach using one-pot methodologies under a nitrogen atmosphere and stored over activated 3 Å
(
Scheme 1). In this regard, the production of amines from molecular sieves. Unless otherwise stated, all reagents were
2
1,22
nitroarenes and benzaldehydes (I in Scheme 1),
or from acquired from commercial suppliers and used as received. The
2
3
amines and phenols (II in Scheme 1), has been explored. precursor methoxy(1,5-cyclooctadiene)rhodium(I) dimer [Rh
Moreover, the selective hydrogenation of phenol to cyclohexa- (μ-OMe)(1,5-COD)]2 was prepared following previously
Infrared spectra were recorded on a
a
2
5
none by RhNPs is favored by specific sites of rhodium nano- reported method.
particles. It is crucial to control the reactivity of the RhNPs’ PerkinElmer FTIR/FIR Spectrum 400 spectrometer. H NMR,
1
1
13
1
13
sites that could favor the complete hydrogenation of aryl { H} COSY (Correlation Spectroscopy), C{1H}, 2D H– C
groups. The latter was demonstrated by the use of substituted HSQC (Heteronuclear Single-Quantum Correlation spec-
1
13
triethoxysilanes grafted onto silica that bond to RhNPs selec- troscopy) and 2D H– C HMBC (Heteronuclear Multiple Bond
1
6
tively, deactivating them towards aryl hydrogenation. In this Correlation spectroscopy) spectra were recorded on a Varian
1
regard, thiolate and amine groups grafted onto RhNP–silica VNMRS spectrometer at 9.4 T (400.13 MHz for H) and chemi-
catalysts could modulate the selectivity towards the partial cal shifts (δ, ppm) were calibrated relative to the residual
hydrogenation of phenol to cyclohexanone in 8–10 hours solvent peak. Thermogravimetric analysis (TGA) data were
−
1
(
TOF 14 h calculated as mmol of cyclohexanone × (mmol of recorded on a PerkinElmer TGA 4000 analyzer from room
−
1
−1
Rh × h) ).
temperature to 900 °C with a speed of 1 °C min . For surface
Given the efforts towards the development of active and area analysis a Micromeritics TriStar 3000 surface area and
stable catalysts based on RhNPs supported on ordered meso- pore size analyzer was used to obtain nitrogen physisorption
porous materials, in the present work a methodology for the isotherms at 77 K on the synthesized silicas; the data were fit
preparation of RhNPs based on our previous experience in the using a Brunauer–Emmett–Teller (BET) model to determine
9
,10,24
formation of colloidal RhNPs is proposed.
The catalytic the apparent surface areas of the materials. The average pore
systems were prepared through the controlled reduction of a diameter and cumulative pore volumes were calculated using
well-defined and readily prepared complex [Rh(μ-OMe)(1,5- the Barrett–Joyner–Halenda (BJH) model for mesopores and
COD)] with dihydrogen in the presence of N-functionalized the Non-Local Density Functional Theory model for micro-
2
silicas. Considering that the selectivity in hydrogenation reac- pores (Tarazona NLDFT, Cylindrical Pores, Esf = 30.0 K).
Samples were degassed under vacuum at 130 °C for 18 h prior
to the analysis. Scanning electron microscopy (SEM) images of
the synthesized materials were obtained on a JEOL JSM-5900-
LV microscope. The particle diameter was determined by
counting at least 100 individual particles with the software
2
6
package Digimizer 4.6.1. High-angle annular dark-field scan-
ning transmission electron microscopy (HAADF-STEM) images
and transmission electron microscopy (TEM) bright field
images of the synthesized materials before and after recycling
were obtained on a JEOL ARM200F microscope operating at an
accelerating voltage of 200 kV. Nanoparticle measurements
were performed with the software package DigitalMicrograph
2
7
Scheme 1 One-pot methodologies to produce amines.
3.30.2017.0.
X-Ray photoelectron spectroscopy (XPS)
3290 | Dalton Trans., 2021, 50, 3289–3298
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
measurements were performed using a Thermo Scientific 23.8 mmol) was then added using a syringe pump with a speed
K-Alpha spectrometer, with a MAGCIS ion source for depth pro- of 3 mL h− under vigorous stirring. After the addition, the solu-
filing analysis and surface cleaning. The assignment of the tion was heated at 80 °C for 2 hours and stirred for another
chemical components of core level Rh 3d was made by compari- 12 h at room temperature. The resulting suspension was filtered
1
28–30
son to the referenced values reported in the literature.
The via a cannula and the resulting solid was washed with cold
synthesis of supported rhodium nanoparticles was carried out methanol (2 × 50 mL). The white solid obtained was dried
in a 100 mL Fisher–Porter vessel. Microwave Plasma Atomic under vacuum for 12 h to afford a fine white powder. In a
Emission Spectroscopy (MP-AES) was used to determine the Rh Schlenk flask, the synthesized silica (1 g) was suspended in dry
content on the samples before and after catalytic experiments toluene (70 mL) under nitrogen and treated with the corres-
and was performed on an Agilent MP-AES 420 equipment, after ponding silicate (NAM 1.6 g, 5.1 mmol or APTES 1 g,
digesting the sample with nitric acid and microwaves (CEM 4.5 mmol). The mixture was heated at reflux temperature and
MDS 2000). Catalytic reactions were performed in a Parr Multi stirred for 24 h. The reaction mixture was cooled to room temp-
Reactor 500 system. Catalytic conversions were determined on a erature. Subsequently ammonium nitrate NH
4 3
NO (60 mg,
Varian 3800 gas chromatograph with a capillary column 0.75 mmol) and EtOH (60 mL) were added to remove the
33
DB-WAX (30 m × 0.32 mm × 0.25 mm) coupled to an FID detec- residual surfactant, and this mixture was heated to 60 °C and
tor, using decane as an internal standard.
stirred for 24 h. The reaction mixture was cooled to room temp-
erature and the solid was separated by centrifugation (300 000
rpm for 10 min), washed with ethanol five times and isolated by
Synthesis of N-(3-(triethoxysilyl)propyl)nicotinamide (NAM)
This compound was synthesized according to a previous centrifugation. The resulting white solid (SiO
2
–APTES or SiO
report with some modifications. In a Schlenk vessel, nic- NAM) was dried under reduced pressure for 12 h.
otinic acid (1.99 g, 16.10 mmol) was cooled in an ice bath General synthesis of supported RhNPs on N-functionalized
under a nitrogen atmosphere for 10 min. Subsequently, silicas. In a Fisher–Porter reactor, [Rh(μ-OMe)(1,5-COD)]
thionyl chloride (5 mL, 68 mmol) was added dropwise at 0 °C. (23.5 mg, 0.049 mmol) and 90 mg of functionalized silica
The resulting mixture was stirred for 24 h at 25 °C. The (SiO –APTES, SiO –NAM or non-functionalized SiO ) were sus-
2
–
3
1
2
2
2
2
respective acid chloride was obtained as a white solid after the pended in dry THF (50 mL) under a nitrogen atmosphere. This
removal of volatiles under reduced pressure and was used mixture was sonicated for one min to obtain a homogeneous
without any further purification. A suspension of the corres- pale-yellow suspension. The reactor was then charged with 3
ponding acid chloride in dry THF (60 mL) under nitrogen at bar of H
5 °C was treated with (3-aminopropyl)triethoxysilane APTES After the reaction, the reactor was depressurized, and the
3.8 mL, 16.23 mmol) and triethylamine (7 mL, 50.2 mmol). resulting black homogeneous suspension was transferred to a
2
at room temperature and allowed to stir for 18 h.
2
(
After the addition, the mixture was heated at reflux tempera- Schlenk flask. All volatiles were removed under reduced
ture and stirred for 24 h. The reaction mixture was cooled to pressure to afford a black solid. The supported RhNPs (black
room temperature, and the suspension was filtered through a solid) obtained were dried under vacuum at 80 °C for 24 h and
fritted funnel under nitrogen, obtaining an orange solution. stored under a nitrogen atmosphere.
After the removal of all volatiles at reduced pressure, a highly
viscous orange oil was obtained. Yield (4.27 g, 81%). H NMR supported RhNPs. To a 25 mL stainless steel reactor flask
General procedure for catalytic hydrogenation reactions with
1
3
,4
(
400 MHz, CDCl
4.7, 1.3 Hz, Py), 8.12 (1H, dq,
3
) δ(ppm): 8.98 (1H, s, Py), 8.64 (1H, dt,
J
H,H equipped with a glass vessel and a stirring magnet were added
3
,4
=
7
JH,H = 7.9, 1.6 Hz, Py), supported catalyst (5 mg), heptane (2 mL), and substrate(s)
.37–7.32 (1H, m, Py), 7.14 (1H, br, (–CO)–NH–), 3.77 (6H, qd, (2 mmol). The reaction mixture was then pressurized with hydro-
3
,4
3
JH,H = 7.0, 0.9 Hz, –SiO–CH
2
–CH
3
), 3.43 (2H, q, JH,H = 6.3 gen (5 or 40 bar) and stirred at 100 °C for the specified time.
3
Hz, –CH –NH(CO)–), 1.73 (2H, p, J
CH
(
= 7.1 Hz, Si–CH –CH – After the reaction was completed, the reactor vessel was cooled
), 0.66 in an ice bath and depressurized. The crude reaction residue
2 2
JH,H = 7.9 Hz, –Si–CH –CH –). C NMR (101 MHz, was filtered through a column prior to being analyzed by GC.
2
H,H
2
2
3,4
2
–), 1.16 (9H, td,
2H, t,
J
H,H = 7.0, 0.9 Hz, –SiO–CH
2
–CH
3
3
13
CDCl ) δ(ppm): 165.60 (CvO), 151.74 (C ), 148.03 (C ),
General procedure for recycling of supported RhNPs. After
), 42.30 the reaction was completed, cooled and depressurized, all vola-
). These data were tiles were removed under reduced pressure and collected in a
consistent with the previous report and were confirmed with nitrogen trap for GC analysis. More solvent and substrate were
D NMR experiments (ESI S1–S5†). added to the reactor vessel containing the catalyst and the pro-
General synthesis of N-functionalized mesoporous silica. cedure for the catalytic reaction was repeated.
Silica MCM-41 type was synthesized according to a previous
3
Py
Py
1
35.05 (CPy), 130.56 (CPy), 123.30 (CPy), 58.40 (CH
2
(
CH ), 22.71 (CH ), 18.18 (CH ), 7.77 (CH
2
2
3
2
3
1
2
3
2
report with some modifications. In a 1 L plastic container,
cetyltrimethyl ammonium bromide CTAB (1.0 g, 2.74 mmol)
was dissolved in 500 mL of deionized water, stirred for 15 min
and sonicated for another 30 min. Subsequently, 2 M NaOH
Results and discussion
Catalyst characterization
solution (3.5 mL, 7 mmol) was added to CTAB solution and Two different catalysts were prepared to evaluate the presence
stirred for 5 min. Tetraethyl orthosilicate TEOS (5 mL, and nature of grafted capping agents in the stabilization of
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3289–3298 | 3291

View Article Online
Paper
Dalton Transactions
RhNPs. Porous silicas were functionalized with (3-aminopro- interaction of rhodium with pyridine and amine of the nicoti-
pyl)triethoxysilane (APTES) or N-(3-(triethoxysilyl)propyl)nicoti- namide group was indicated by the presence of bands at 1429
namide (NAM). The functionalization took place after the syn- and 1463 cm− (Fig. 1b). The carbonyl interaction with
1
thesis of porous silicas using TEOS and the templating agent rhodium was discarded with the appearance of an upfield
M). After signal at 1648 cm− as previously observed for several metal
1
CTAB in basic aqueous solution (NaOH
2
3
6,37
N-functionalization, the removal of the surfactant was per- complexes with NAM as a ligand.
formed using NH NO . This type of functionalization locates The interaction between RhNPs and the support was
4
3
the functional groups on the external surface and at the pore further corroborated by X-ray photoelectron spectroscopy
mouth, as suggested in previous reports for coumarin functio- (XPS). The Rh 3d5/2 region from the high resolution XPS
3
4,35
nalized silicas.
The formation of supported RhNPs (Scheme 2) was Rh(0), Rh(I) and Rh(III) (Fig. 2).
achieved by the reduction of the metal precursor [Rh(μ-OMe) Rh is due to the anchoring of Rh nanoparticles onto the silica
1,5-COD)] with dihydrogen in the presence of the corres- surface, which takes place via oxygen bonding, as previously
ponding porous support SiO (RhNPs-A), SiO –APTES (RhNPs- reported. Considering the structural magic numbers for a
spectra of RhNPs-C shows a chemical environment made up of
2
8–30
The presence of oxidized
(
2
3
8
2
2
2
B) or SiO –NAM (RhNPs-C). The initial yellow suspension face-centered cubic closed-packed Rh nanoparticle, around
turned black within minutes after pressurization with dihydro- 63% of atoms are on the surface for a nanoparticle without
gen. The resulting catalysts RhNPs-(A–C) were thoroughly imperfection of 1.9 nm, and 52% of atoms for a particle of
3
9
characterized by different techniques (IR, TGA, SEM, TEM, 2.4 nm. For this reason, the presence of Rh(I) and Rh(III) is
BET and XPS) to assess the effect of the different amine-stabil- mostly on the nanoparticle surface. Comparable high-resolu-
izers grafted on mesoporous silicas.
tion spectra were observed for RhNPs-B (ESI S7 and S8†).
The presence of functional groups was examined by FT-IR
Thermal analyses of the silicas without rhodium showed
where the signals of the grafted aminopropyl or nicotinamide the loss of organic material between 150 and 450 °C (ESI S10
propyl groups were observed in addition to the expected and S11†). The amount of the organic fraction in SiO –APTES
2
signals for siloxane, silanol groups and physisorbed water. In was determined to be 21 wt% and 27.4 wt% for SiO
the case of RhNPs-B, a slight change in the frequencies of the Two important fragmentations in SiO –NAM were observed
corresponding rhodium-free silica support corroborated the probably due to NH–(CO) bond cleavage (ESI S11†). For all the
coordination of rhodium (Fig. 1a). In the case of SiO –NAM, a catalysts RhNPs-(A–C), the presence of Rh enhanced the
2
–NAM.
2
2
−
1
strong band at around 1640 cm is observed; it could be weight loss during thermogravimetric analyses (Fig. 3 and ESI
attributed to the carbonyl group of the nicotinamide (NAM) S12–S14†) that could arise from the weakening of organic
functional group, as well as IR bands were assigned to the bonds upon an interaction with rhodium. The rhodium
alkyl, pyridine and amine groups (Fig. 1b). For RhNPs-C, an content on each support was determined by microwave plasma
Scheme 2 Synthesis of RhNPs supported on N-functionalized silicas.
3292 | Dalton Trans., 2021, 50, 3289–3298
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Fig. 3 Thermogravimetric data of RhNPs-(A–C). (For parent functiona-
lized silicas see ESI S10 and S11.†)
pared to the amount in pristine silica in RhNPs-A could be
explained considering the presence of coordinating groups
grafted onto silica. In the absence of functional groups such as
propylamine, RhNPs aggregate and the amount of rhodium
incorporated in pristine silica could be higher than the
amount of rhodium incorporated when propylamine is
present. Propylamine stabilizes smaller RhNPs and this could
account for the low concentration of rhodium.
The synthesized materials were also characterized by elec-
tronic microscopy techniques. The size of silicas without
rhodium was characterized by scanning electron microscopy
(
SEM), showing particles with a mean diameter of 206 ± 20 nm
4
0
for pristine silica, 393 ± 74 nm for SiO
2
–APTES and 767 ±
1
37 nm for SiO –NAM (ESI S15–S18†). The pore diameter of
2
Fig. 1 IR spectra of functionalized silicas and the respective supported
the silicas was measured by high-angle annular dark-field
scanning transmission electron microscopy (HAADF-STEM),
showing a size of around 1.4 nm for pristine silica, 3.5 nm for
RhNPs. (A) SiO
Legend: SiO
ν (–CvO), ν (–CvN)py). (For RhNPs-A and SiO
2
–APTES and RNPs-B, (B) SiO
lattice, physisorbed water, ν (N–H), ν (–CH
see ESI S6.†)
2
–NAM and RhNPs-C.
(
2
2
–),
2
2 2
SiO –APTES and 3.3 nm for SiO –NAM (ESI S19, S20 and
S22†). Upon the formation of RhNPs supported on the silicas,
transmission electron microscopy (TEM) analyses showed the
presence of nanoparticles exclusively at the silica external
surface (Fig. 4 and 5) and in the cases of RhNPs-B and RhNPs-
C, energy dispersive X-ray (EDX) mapping showed the presence
of Rh homogeneously dispersed at the silica surface, as well as
N. The morphology of RhNPs was influenced by the nature of
the functional group at the silica surface. In the absence of
functional groups like in RhNPs-A, mainly metal agglomerates
were observed by TEM on the silica surface (ESI S24†). RhNPs-
B showed the presence of small aggregates of Rh over the
surface (Fig. 4A), while RhNPs-C presented a better dispersion
(
Fig. 5A). The size of the silicas was not modified after support-
ing Rh. The mean diameter of the supported RhNPs was deter-
mined by HAADF-STEM, obtaining 3.2 ± 0.5 nm for RhNPs-B
Fig. 2 High resolution-XPS Rh 3d of RhNPs-C.
(
Fig. 4B) and 2.3 ± 0.3 nm for RhNPs-C (Fig. 5B). The for-
mation of homogeneously dispersed spherical metal nano-
atomic emission spectroscopy (MP-AES) and it was 4.3 wt% for particles of smaller size in RhNPs-C compared to RhNPs-B,
RhNPs-A, 2.9 wt% for RhNPs-B and 7.6 wt% for RhNPs-C. The could be a consequence of the stronger coordination capabili-
small amount of Rh in SiO –APTES found in RhNPs-B com- ties of nicotinamide groups. HR-TEM and FFT evidenced the
2
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3289–3298 | 3293

View Article Online
Paper
Dalton Transactions
Fig. 4 TEM images of RhNPs-B (A); HAADF-STEM image; (B) TEM
image of RhNPs; (C) size distribution of RhNPs (3.2 ± 0.5 nm); (D) HR
HAADF-STEM image with FFT of a single nanoparticle, an interplanar
distance of 2.2 Å; (E) EDX mapping of a section of the catalyst, (F) Si, (G)
O, (H) Rh and (I) N.
Fig. 6 Adsorption–desorption isotherms of RhNPs-B (red circles) and
RhNPs-C (green triangles). Filled figures correspond to adsorption and
hollow figures correspond to desorption.
2
apparent surface area in RhNPs-B compared to SiO –APTES is
associated with the blockage of the pore mouth (aperture) of
the original functionalized silica by the metal nanoparticles.
This loss of porosity was further corroborated by the pore
width calculated using the NLDFT data treatment.
Catalytic activity in the hydrogenation of different substrates
The initial assessment in the hydrogenation of different sub-
strates was performed using the RhNPs-B catalyst, and the
results are summarized in Table 1. High conversion and
selectivity were achieved in the hydrogenation of nitrobenzene
(
1), terminal and internal alkenes (2–4), 4-phenyl-3-buten-2-
one (5), benzylidene aniline (6) and acetophenone (7) (Table 1,
entries 1–7). In comparison with other catalysts with heteroge-
4
2
nized Rh on magnetic silica or hollow nano-mesosilica
4
3
spheres with N-doped carbon used as catalysts at 1 bar H2
pressure, the activity towards the conversion of 1 was higher
for RhNPs-B at 5 bar H pressure. The activity and selectivity of
2
RhNPs-B in the hydrogenation of 5 (entry 5 in Table 1) were
Fig. 5 TEM images of RhNPs-C (A); HAADF-STEM image; (B) TEM
image of RhNPs; (C) size distribution of RhNPs (2.3 ± 0.3 nm); (D) HR
HAADF-STEM image, an interplanar distance of 2.2 Å; (E) EDX mapping
of a particle of the catalyst, (F) Si, (G) O, (H) Rh and (I) N.
superior to those of RhNPs supported on phosphine-functio-
2
0
nalized silica. RhNPs-B showed very low conversion of benz-
aldehyde (8) and toluene (9) under similar reaction conditions
presence of a crystalline structure of Rh with an interplanar (Table 1, entries 8 and 9), whereas alkyl derivatives such as
distance of 2.2 A, typical of the {111} planes of a face-centered nitro-decane, 1-octene, nonanal and cyclohexanone were not
cubic closed-packing in both RhNPs-B and RhNPs-C (Fig. 4D hydrogenated (see ESI Table S6†). The catalytic results of
and 5D, respectively).
RhNPs-B in hydrogenation reactions suggested that for the
Concerning the textural properties of the supported RhNPs hydrogenation to take place, π-conjugated and constrained
in functionalized OMS, the catalysts presented nitrogen geometry substrates are required. The latter has been observed
4
1
9,10,44
adsorption–desorption curves of Type II (Fig. 6), and no hys- for ligand-capped colloidal metal nanoparticles
where
teresis, characteristic of nonporous materials, with more π-conjugated substrates favor the formation of metal di-σ
gradual curvature (indistinguishable knee) indicating mono- bonded species and enhance the hydrogenation of one unsatu-
layer accumulation and the onset of multilayer adsorption rated group. This could explain why 3 is hydrogenated to ethyl-
resulting in a small apparent surface area (SBET) for RhNPs-B benzene (3a) by RhNPs-B and 1-octene is not transformed.
2
−1
2
−1
(
16 m
parent silica material SiO
mesoporous materials with a surface area (SBET) of 415 m g
and a pore diameter of 2 nm (ESI S37†). A decrease in the runs (Fig. 7). The reused RhNPs-B catalyst was further analyzed
g
) and for RhNPs-C (9 m
g
). Interestingly, the
The reuse of the RhNPs-B catalyst in the hydrogenation of
2
–APTES presented a typical curve of cyclohexene was evaluated after product separation by simple
2
−1
distillation, and the catalyst was active for up to 5 catalytic
3294 | Dalton Trans., 2021, 50, 3289–3298
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Table 1 Catalytic hydrogenation by RhNPs-B at low H
2
pressure in
toluene
a
Conv.
TOF
a
−1 b
Entry
1
Substrate
Product
(sel.) %
(h
)
c
>99 (92)
649
2
3
4
>99 (100)
706
649
600
Fig. 7 (A) Recycling hydrogenation of cyclohexene in toluene with
RhNPs-B, and (B) TEM images of catalysts after the recycling reactions.
c
>99 (92)
(
(
Reaction conditions:
2 mmol cyclohexene, 5 mg of RhNPs-B
0.07 mol%), H (5 bar), 100 °C, 2 h.)
2
85 (100)
95 (100)
5
6
670
473
acetophenone (7) in toluene (Table 1, entry 10). High activity
with high selectivity for ketone hydrogenation to form 1-phe-
nylethanol (7a, 93%) was observed. In the case of RhNPs-A, a
catalyst with poor stability for RhNPs due to the lack of amine-
functional groups (as shown in the previous section) showed
poor conversion of acetophenone with a high selectivity to 7a
67 (100)
(
Table 1, entry 11).
To apply the supported catalysts RhNPs-B and RhNPs-C in
d
7
8
98 (81)
560
the hydrogenation of more challenging substrates, a higher
dihydrogen pressure and heptane as an innocuous solvent
were used. In general, the activity observed for both catalysts
was similar. The hydrogenation of acetophenone (7) at 40 bar
(Table 2, entry 1) afforded the fully reduced product cyclohexyl
ethanol (7b) with moderate selectivity with acetyl cyclohexane
as the remaining product. This suggested that none of the cat-
alysts can reduce the initially formed acetyl cyclohexane into
cyclohexyl ethanol.
6 (n. d.)
n.d.
e
9
1 (n. d.)
81 (93)
n.d.
219
f
1
0
1
g
The stability of RhNPs-B and RhNPS-C was assessed after
1
10 (80)
48
2
the hydrogenation of acetophenone in heptane at 40 bar of H .
HAADF-STEM analyses and EDX mapping of the samples
revealed the formation of Rh aggregates at the surface in the
Reaction conditions: 2 mmol substrate, 5 mg of RhNPs-B (0.07 mol%), case of RhNPs-B (Fig. 8A), whereas RhNPs-C preserved the
a
2
mL of toluene, H
2
(5 bar), 100 °C, 2 h. Determined by GC using
homogeneous distribution of RhNPs over the functionalized
silica surface (Fig. 8D) with a similar nanoparticle size distri-
b
decane as an internal standard. TOF = mmol of the principal product
Completely hydrogenated byproduct.
Cyclohexyl ethanol (8%) and acetyl cyclohexane (11%) were detected bution (2.3 ± 0.3 nm, ESI S42†). Moreover, rhodium loss from
by GC. Performed in EtOH. RhNPs-C (0.18 mol%) used as the cata-
lyst. RhNPs-A (0.10 mol%) used as the catalyst.
−1
c
×
(mmol of Rh
×
h)
.
d
e
f
RhNPs-C during the hydrogenation of acetophenone was dis-
carded by the MP-AES analysis of the organic phase collected
after the catalytic reaction took place (0.19 ppm of rhodium
detected, corresponding to 0.2 × 10− % of supported rhodium
g
3
by HAADF-STEM, showing the formation of rhodium aggre- leaching.).
gates separated from the support as well as at the support, as a
Both catalysts, RhNPs-B and RhNPs-C, were active in the
result of metal sintering. The coordination ability of alkyl hydrogenation of benzaldehyde (8) and quinoline (10) at high
amines used as stabilizers of metal nanoparticles changes
2
H pressure with high selectivity to benzylic alcohol (8a) and
under hydrogen pressure, leading to nanoparticle aggregates 1,2,3,4-tetrahidroquinoline (10a), respectively (Table 2, entries
4
5
as previously shown for PdNPs.
The observed catalytic 2 and 3). When toluene (9) was used as the substrate, it was
activity in the hydrogenation of cyclohexene by reused RhNPs- moderately converted to methylcyclohexane (9a) (Table 2, entry
B could be feasible due to the presence of small rhodium 4). Benzonitrile (11) was converted entirely to 5 different pro-
nanoparticles.
Under similar reaction conditions, the RhNPs-C catalyst both catalysts showed similar behavior (Table 2, entry 5).
with nicotinamide groups grafted on silica as the stabilizer of We further studied the hydrogenation of phenol (12) by the
rhodium nanoparticles was evaluated in the hydrogenation of supported RhNPs at high H pressure (40 bar) and observed
ducts, with aryl benzyl imine (11a) as the major product and
2
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3289–3298 | 3295

View Article Online
Paper
Dalton Transactions
Table 2 Hydrogenation reactions by RhNPs-B and RhNPs-C in heptane
RhNPs-B
RhNPs-C
a
a
TOF (h−1)
b
a
a
TOF (h−1)
b
Entry
1
Substrate
Main product
H
2
pressure (bar)
Conv. (sel.) %
Conv. (sel.) %
c
c
40
99 (78)
545
99 (76)
203
d
d
2
3
40
40
65 (89)
408
664
96 (91)
235
251
99 (95)
94 (>99)
4
40
40
61 (>99)
>99 (48)
430
335
45 (>99)
97 (56)
120
147
e
5
f
6
40
5
>99 (75 for 12b)
524
473
>99 (54 for 12a)
144
44
f
7
87 (77)
25 (66)
Reaction conditions: 2 mmol substrate, 5 mg of catalyst (RhNPs-B (0.07 mol%), RhNPs-C (0.18 mol%)), 2 mL of heptane, 100 °C, 2 h.
a
b
−1 c
Determined by GC using decane as an internal standard. TOF = mmol of the principal product × (mmol of Rh × h)
. Acetyl cyclohexane as
d
e
byproduct. Benzoic acid as byproduct. Products were determined by GC-MS, and benzylidene imine and dibenzyl amine were detected as
byproducts. Reaction run for 4 h yielding cyclohexanone and cyclohexanol.
f
RhNPs-B the activity was superior (Table 2, entry 7). The latter
has been observed for colloidal RhNPs stabilized with
N-heterocyclic carbene ligands, and it was suggested that the
selectivity for cyclohexanone in the hydrogenation of phenol at
2
low H pressure is a result of the difference in the rates
between the isomerization and hydrogenation of the enol
4
6
intermediate. Considering the amount of supported Rh in
both catalytic systems, a similar TOF was calculated at 40 bar
−
1
−1
H pressure (TOF 89 h for RhNPs-B and 76 h for RhNPs-C,
2
−
1
calculated as mmol of cyclohexanone × (mmol of Rh × h) ),
and at 5 bar H pressure the difference was remarkable
239 h− for RhNPs-B and 23 h−1 for RhNPs-C). The latter are
among the best catalytic results for the hydrogenation of
2
1
(
1
6
phenol to cyclohexanone by supported RhNPs.
Overall, the different catalytic activities of RhNPs-B and
RhNPs-C depend on the coordination ability of amine or nico-
tinamide groups at the silica surface towards rhodium. The
Fig. 8 HAADF-STEM image and EDX mapping of RhNPs-B after aceto-
phenone hydrogenation at 40 bar of pressure: (A) HAADF image; (B) EDX
Rh mapping; (C) EDX
N mapping, HAADF-STEM image and EDX
mapping of RhNPs-C after acetophenone hydrogenation; (D) HAADF strong coordination of nicotinamide to the rhodium surface in
image; (E) EDX Rh mapping; (F) EDX N mapping.
RhNPs-C affords a more robust catalytic system, however with
a low conversion rate in the hydrogenation of phenol.
Given the activity and selectivity shown by the supported
RhNPs in the hydrogenation of phenol, the catalytic systems
the complete conversion of 12. The use of RhNPs-B afforded a were applied in the reductive amination of phenol with mor-
mixture of cyclohexanone (12a) and cyclohexanol (12b) in a pholine as a one-pot methodology to produce tertiary amines
3
: 1 ratio, respectively, whereas RhNPs-C favored a 46 : 54 ratio (Scheme 3). In this multistep reaction, phenol is catalytically
of products (Table 2, entry 6). At low H pressure (5 bar), both reduced to cyclohexanone, which reacts with morpholine to
catalysts showed similar selectivity to 12a and in the case of form an enamine intermediate; the latter is catalytically
2
3296 | Dalton Trans., 2021, 50, 3289–3298
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
reduced to afford the desired tertiary amine. The reductive
amination of phenol and morpholine catalyzed by both sup-
Conflicts of interest
2
ported catalysts RhNPs-B and RhNPs-C at low H pressure The authors have no competing interests to declare.
showed high conversion to the tertiary amine. In the reaction
catalyzed by RhNPs-C, a positive effect of the added morpho-
line on the conversion of phenol was observed.
Acknowledgements
This work was financially supported by Consejo Nacional de
Ciencia y Tecnología through the grant number 283094. The
authors acknowledge the access to the experimental facility of
Laboratorio de Conversión y Almacenamiento de Energía del
IPN for XPS studies and the Unidad de Investigación en
Catálisis of the School of Chemistry at UNAM for BET sorption
studies. Special thanks to Dr Juan Olguín for helpful
discussions.
Conclusions
The controlled reduction of [Rh(μ-OMe)(1,5-COD)] using dihy-
2
drogen in the presence of functionalized porous silica afforded
the formation of supported RhNPs. The supported catalyst was
thoroughly characterized through diverse analytical techniques
(FT-IR, TGA, XPS, SEM and TEM) to confirm the formation of
spherical RhNPs with a narrow size and homogeneous distri-
bution on the surface of ordered functionalized silicas.
RhNPs supported on amine-functionalized silica, RhNPs-B,
were active in the hydrogenation of several substrates and
showed remarkable selectivity to alkene, nitro, ketone and
imine groups at low hydrogen pressure. The catalytic activity
and selectivity of the supported RhNPs reached after only
References
1
2
3
4
5
6
7
8
9
P. A. Evans, Modern Rhodium-Catalyzed Organic Reactions,
John Wiley & Sons, 2005.
Y. Yuan, N. Yan and P. J. Dyson, ACS Catal., 2012, 2, 1057–
2
0
hours of the hydrogenation reaction using less than
.2 mol%, are among the most active supported catalysts
1
069.
N. J. S. Costa and L. M. Rossi, Nanoscale, 2012, 4, 5826–
834.
based on RhNPs. RhNPs-B suffered from morphological
changes after the catalytic reduction of cyclohexene as well as
after the reduction of acetophenone. The presence of nicotina-
mide groups grafted on ordered mesoporous silica stabilized
spherical and homogenously distributed RhNPs. The dis-
persion and morphology of RhNPs supported on the nicotina-
mide functionalized ordered silica, RhNPs-C, were preserved
after the catalytic reduction of acetophenone at high hydrogen
pressure. In addition, the selective reduction of acetophenone
could be tuned by varying the hydrogen pressure from 5 to 40
bar, achieving 1-phenylethanol at low pressure and cyclohexyl-
ethanol at high pressure.
The strong interaction between nicotinamide and rhodium
did not affect the reduction of several substrates. The catalyst
could selectively hydrogenate benzaldehyde, quinoline and
toluene, whereas benzonitrile and phenol showed moderate
selectivity to benzyl imine and cyclohexanone, respectively.
Finally, the catalysts RhNPs-B and RhNPs-C showed a
remarkable selectivity in the reductive amination of phenol
and morpholine to afford tertiary amines under a low hydro-
gen pressure (5 bar).
5
L. M. Rossi, J. L. Fiorio, M. A. S. Garcia and C. P. Ferraz,
Dalton Trans., 2018, 47, 5889–5915.
S. Campisi, M. Schiavoni, C. Chan-Thaw and A. Villa,
Catalysts, 2016, 6, 185.
M. A. Ortuño and N. López, Catal. Sci. Technol., 2019, 9,
5
173–5185.
F. Viñes, J. R. B. Gomes and F. Illas, Chem. Soc. Rev., 2014,
3, 4922–4939.
J. Won Park and Y. Keun Chung, ACS Catal., 2015, 5, 4846–
859.
4
4
A. Serrano-Maldonado, S. S. Rozenel, J. L. Jimenez-
Santiago, I. Guerrero-Ríos and E. Martin, Catal. Sci.
Technol., 2018, 8, 4373–4382.
10 A. Serrano-Maldonado, E. Martin and I. Guerrero-Ríos,
Eur. J. Inorg. Chem., 2019, 2863–2870.
1
1
1 Q. L. Zhu and Q. Xu, Chemistry, 2016, 1, 220–245.
2 R. J. White, R. Luque, V. L. Budarin, J. H. Clark and
D. J. Macquarrie, Chem. Soc. Rev., 2009, 38, 481–494.
3 G. Prieto, M. Shakeri, K. P. De Jong and P. E. De Jongh, ACS
Nano, 2014, 8, 2522–2531.
1
1
1
1
1
1
4 I. E. Ertas, M. Gulcan, A. Bulut, M. Yurderi and
M. Zahmakiran, J. Mol. Catal. A: Chem., 2015, 410, 209–220.
5 L. Zhang, M. Zhou, A. Wang and T. Zhang, Chem. Rev.,
2020, 120, 683–733.
6 H. Zhang, A. Han, K. Okumura, L. Zhong, S. Li, S. Jaenicke
and G. K. Chuah, J. Catal., 2018, 364, 354–365.
7 A. Sánchez, M. Fang, A. Ahmed and R. A. Sánchez-Delgado,
Appl. Catal., A, 2014, 477, 117–124.
8 I. Geukens and D. E. De Vos, Langmuir, 2013, 29, 3170–
3178.
Scheme 3 Reductive amination of phenol and morpholine. Reaction
conditions: 2 mmol phenol, 2 mmol morpholine, 5 mg of catalyst, H (5
2
bar), 2 mL of heptane, 100 °C, 6 h. Conversion and selectivity were
determined by GC using decane as an internal standard.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 3289–3298 | 3297

View Article Online
Paper
Dalton Transactions
1
9 M. Boutros, G. Moarbess, T. Onfroy and F. Launay, 34 N. K. Mal, M. Fujiwara and Y. Tanaka, Nature, 2003, 421,
C. R. Chim., 2018, 21, 514–521. 350–353.
0 J. Llop Castelbou, K. C. Szeto, W. Barakat, N. Merle, 35 N. K. Mal, M. Fujiwara, Y. Tanaka, T. Taguchi and
2
C. Godard, M. Taoufik and C. Claver, Chem. Commun.,
017, 53, 3261–3264.
1 M. A. Hussein, A. H. Dinh, V. T. Huynh and T. V. Nguyen,
M. Matsukata, Chem. Mater., 2003, 15, 3385–3394.
36 S. Bayari, A. Ataç and Ş. Yurdakul, J. Mol. Struct., 2003, 655,
163–170.
2
2
2
2
2
2
Chem. Commun., 2020, 56, 8691–8694.
2 L. Huang, Z. Wang, L. Geng, R. Chen, W. Xing, Y. Wang
and J. Huang, RSC Adv., 2015, 5, 56936–56941.
3 P. Tomkins, C. Valgaeren, K. Adriaensen, T. Cuypers and
D. E. D. Vos, ChemCatChem, 2018, 10, 3689–3693.
4 A. Serrano-Maldonado, A. Reina, B. Portales-Martínez and 39 C. P. Poole and J. J. Frank Owens, Intruduction to
I. Guerrero-Ríos, Eur. J. Inorg. Chem., 2020, 2506–2511. Nanotechnology, John Wiley & Sons, 2003.
5 R. Uson, L. A. Oro, J. A. Cabeza, H. E. Bryndza and 40 K. P. Ramírez-Cuellar, K. P. Salas-Martin, P. A. Méndez-
37 A. Ataç, Ş. Yurdakul and S. Berber, Spectrochim. Acta, Part
A, 2011, 81, 684–689.
38 V. Mévellec, A. Nowicki, A. Roucoux, C. Dujardin,
P. Granger, E. Payen and K. Philippot, New J. Chem., 2006,
30, 1214–1219.
M. P. Stepro, Inorg. Synth., 1985, 23, 126–130.
Ocampo, L. Martínez-dlCruz, A. García-Márquez and
I. Guerrero-Ríos, Catal. Today, 2020, 358, 45–50.
41 M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier,
F. Rodriguez-Reinoso, J. Rouquerol and K. S. W. Sing, Pure
Appl. Chem., 2015, 87, 1051–1069.
2
2
2
6 Digimizer 4.6.1, MedCalc Software, Belgium, 2005–2012.
7 DigitalMicrograph 3.30.2017.0, Gatan Inc., 1996–2018.
8 A. M. Dennis, R. A. Howard, K. M. Kadish, J. L. Bear, J. Brace
and N. Winograd, Inorg. Chim. Acta, 1980, 44, 139–141.
2
3
3
3
9 M. Carvalho, L. F. Wieserman and D. M. Hercules, Appl. 42 H. Tian, J. Zhou, Y. Li, Y. Wang, L. Liu, Y. Ai, Z. N. Hu, J. Li,
Spectrosc., 1982, 36, 290–296.
0 Y. Abe, K. Kato, M. Kawamura and K. Sasaki, Surf. Sci.
Spectra, 2001, 8, 117–125.
R. Guo, Z. Liu, H. Bin Sun and Q. Liang, ChemCatChem,
2019, 11, 5543–5552.
43 S. Wang, J. Dai, Z. Shi, Z. Xiong, Z. Zhang, S. Qiu and
R. Wang, ChemPlusChem, 2020, 85, 247–253.
1 Q. M. Wang and B. Yan, Cryst. Growth Des., 2005, 5, 497–
5
03.
44 J. S. Zhu and Y. S. Shon, Nanoscale, 2015, 7, 17786–17790.
2 N. Singh, A. Karambelkar, L. Gu, K. Lin, J. S. Miller, 45 M. Moreno, F. J. Ibañez, J. B. Jasinski and F. P. Zamborini,
C. S. Chen, M. J. Sailor and S. N. Bhatia, J. Am. Chem. Soc.,
011, 133, 19582–19585.
3 A. Pourjavadi and Z. M. Tehrani, Mater. Sci. Eng., C, 2016,
1, 782–790.
J. Am. Chem. Soc., 2011, 133, 4389–4397.
46 F. Martinez-Espinar, P. Blondeau, P. Nolis, B. Chaudret,
C. Claver, S. Castillón and C. Godard, J. Catal., 2017, 354,
113–127.
2
3
6
3298 | Dalton Trans., 2021, 50, 3289–3298
This journal is © The Royal Society of Chemistry 2021