Pd@tetrahedral hollow magnetic nanoparticles coated with N-doped porous carbon as an efficient catalyst for hydrogenation of nitroarenes

Article abstract of DOI:10.1002/aoc.5229

Hollow magnetic nanoparticles (MNPs) with tetrahedral morphology were synthesized and then covered by a shell prepared by coating with melamine–formaldehyde followed by the introduction of glucose-derived carbon. Subsequently, Pd nanoparticles were immobi

Full text of DOI:10.1002/aoc.5229

Received: 2 July 2019
DOI: 10.1002/aoc.5229
Revised: 8 August 2019
Accepted: 16 August 2019
F U L L P A P E R
Pd@tetrahedral hollow magnetic nanoparticles coated with
N‐doped porous carbon as an efficient catalyst for
hydrogenation of nitroarenes
Samahe Sadjadi1
| Majid M. Heravi2
1 Gas Conversion Department, Faculty of
Petrochemicals, Iran Polymer and
Petrochemicals Institute, 15km Tehran‐
Karaj Highway, Pajuhesh Science and
Technology Park,Pajuhesh Boulevard PO
Box 14975‐112, Tehran, Iran
2 Department of Chemistry, School of
Science, Alzahra University, PO Box
1993891176, Vanak, Tehran, Iran
Hollow magnetic nanoparticles (MNPs) with tetrahedral morphology were syn-
thesized and then covered by a shell prepared by coating with melamine–
formaldehyde followed by the introduction of glucose‐derived carbon. Subse-
quently, Pd nanoparticles were immobilized and the core–shell nanocomposite
was carbonized. The obtained magnetic catalyst was successfully applied for
the hydrogenation of nitroarenes in aqueous media. To investigate the effects
of the morphology of MNPs, the nature of carbon shell, and the order of incor-
poration of Pd nanoparticles, several control catalysts, including the MNPs
with different morphologies (disc‐like and cylinder); MNPs coated with differ-
ent shells (sole glucose‐derived carbon or melamine–formaldehyde carbon
shell); and a nanocomposite, in which Pd was immobilized after carbonization,
were prepared and examined as catalyst for the model reaction. To justify the
observed different catalytic activities of the catalysts, their Pd loadings,
leaching, and specific surface areas were compared. The results confirmed that
tetrahedral MNPs coated with porous N‐rich carbon shell exhibited the best
catalytic activity. The high catalytic activity of this catalyst was attributed to
its high surface area and the interaction of N‐rich shell with Pd nanoparticles
that led to the higher Pd loading and suppressed Pd leaching.
Correspondence
Samahe Sadjadi, Gas Conversion
Department, Faculty of Petrochemicals,
Iran Polymer and Petrochemicals
Institute,15km Tehran‐Karaj Highway,
Pajuhesh Science and Technology Park,
Pajuhesh Boulevard PO Box 14975‐112,
Tehran, Iran.
Email: samahesadjadi@yahoo.com;
s.sadjadi@ippi.ac.ir
Majid M. Heravi, Department of
Chemistry, School of Science, Alzahra
University, PO Box 1993891176, Vanak,
Tehran, Iran.
Email: mmh1331@yahoo.com; mheravi.
alzahra@ac.ir
KEYWORDS
catalyst, hydrogenation, magnetic nanoparticles, N‐doped carbon
Funding information
Iran National Science Foundation, Grant/
Award Number: 97009384; the Iran Poly-
mer and Petrochemical Institute and
Alzahra University
1 | INTRODUCTION
immobilization or adsorption of other functional
groups.[2,3] Notably, the properties of MNPs are depen-
In recent years, magnetic nanoparticles (MNPs) and their
composites have attracted considerable attention[1] in
nanotechnology, bioimaging, drug delivery, and heteroge-
neous catalysis due to their magnetic properties, large
surface‐to‐volume ratio, and very active surface for the
dent on the size and the morphology of the nanoparticles.
Among MNPs, maghemite nanoparticles (γ‐Fe2O3), espe-
cially hollow structures,[4,5] are well‐known paramagnetic
materials, which have been widely used in catalysis
because of their unique properties such as easy
Appl Organometal Chem. 2019;e5229.
wileyonlinelibrary.com/journal/aoc
© 2019 John Wiley & Sons, Ltd.
1 of 14
https://doi.org/10.1002/aoc.5229

2 of 14
SADJADI AND HERAVI
preparation processes, facile separation, and environmen-
tal stability that provide facile recovery and
recycling.[4,6,7]
catalytic activity, herein, we aim to scrutinize the effects
of the morphology of MNPs, the nature of a carbon shell
on the periphery of MNPs, and the order of the immobi-
lization of zero‐valence Pd nanoparticles on the carbon‐
coated MNPs on the catalytic activity of the resulting cat-
alysts for the hydrogenation of nitroarenes.
Unfortunately, MNPs have a strong tendency to
agglomerate because of their large surface area‐to‐volume
ratio and also magnetic dipole–dipole interactions.[8]
These drawbacks are the main challenges for the synthesis
of stable magnetic nanocomposites with good dispersion.
By contrast, in the case of loading of other nanoparticles
on MNPs, the involvement of iron in the redox processes
decreases the catalytic activity.[9] It is noticed that the
coating of outer surface of the MNPs with inorganic or
organic inert materials such as polymer, silica, or carbon
can furnish a solution to this problem.[10]
2 | EXPERIMENTAL
2.1 | Materials and instruments
All the chemicals and solvents including FeCl3·6H2O, D‐
(+)‐glucose, urea, melamine, formaldehyde, NaBH4, Pd
(OAc)2, EtOH, deionized water, and toluene were pur-
chased from Sigma‐Aldrich, Germany‐Taufkirchen and
used as received. The hydrogenation reaction was per-
formed using nitroarenes. All nitroarenes were of analyt-
ical grade, provided from Sigma‐Aldrich, Germany‐
Taufkirchen, and used without further purification.
The formation of Pd@TH‐Fe2O3@MFC as well as con-
trol samples was confirmed by applying Brunauer–
Emmett–Teller (BET), vibrating‐sample magnetometry
(VSM), X‐ray diffraction (XRD), thermogravimetric anal-
ysis (TGA), field emission scanning electron microscopic
(FESEM)/energy‐dispersive X‐ray spectroscopy (EDS),
Fourier‐transform infrared (FTIR), transmission electron
microscopy (TEM), Raman spectroscopy, and inductively
coupled plasma (ICP) atomic emission spectroscopy. XRD
patterns were obtained using a Siemens, D5000 diffrac-
tometer with Cu‐Kα radiation at 60 KeV and 15 Ma.
FESEM images of the catalyst and control samples were
recorded on a MIRA 3 TESCAN‐XMU instrument using
Au‐coated samples and an acceleration voltage of 20 Kv.
TEM images of the catalyst were obtained using Philips
CM30300Kv field emission transmission electron micro-
scope. TGAs were carried out using a METTLER
TOLEDO instrument under nitrogen atmosphere over
the range of 40 to 800 °C with a heating rate of 10 °
C min−1. FTIR spectra were collected on a PERKIN‐
ELMER‐Spectrum 65 FTIR spectrometer using KBr pel-
lets. BET analyses were performed using a BELsorp Mini
II instrument. The degassing process of samples was car-
ried out at 423 K for 3 hr. The magnetic properties of the
samples were estimated using VSM (Lakeshore7407) at
room temperature. Teksan N1‐541 spectrometer was used
for recording Raman spectrum at λ = 532 nm. The ultra-
sonic apparatus employed for the synthesis of the catalyst
was Bandelin HD 3200 with output power between 100
and 200 W and tip TT13. Using an ICP analyzer of type
Varian Vista‐pro, measurements of the Pd and Fe loading
and leaching of the catalyst were performed.
Lately, given their widespread applications in numer-
ous fields such as catalysis,[11] fuel cell,[12] gas storage,
and separation,[13] porous carbon materials have received
growing attention.[14,15] Porous carbons based on polymer
are an attractive generation of porous materials created
from light elements (H, B, C, N, and O) that are known to
form strong covalent bonds.[16,17] General methods for
the preparation of these carbons are based on the use of
nitrogen precursors such as aniline,[18] pyrrole,[19] and
acrylonitrile[20] that lead to the synthesis of N‐doped
porous carbons. Among available molecular monomers,
melamine,[21,22] which is a cheap and abundant chemical
with 66% N by mass, is recognized as an ideal candidate
for the fabrication of N‐doped porous carbons. Melamines
have many advantages over conventional N‐rich precur-
sors due to their rigid molecular backbone, high nitrogen
content, low density, and high thermal and/or chemical
stability.[23] More importantly, the pore size of the
melamine‐derived carbons can be tuned by wise choice of
monomer, controlling polymerization condition, etc.[24]
Anilines and their derivatives are key chemicals for
the production of dyes, agricultural chemicals, pharma-
ceuticals, and biologically active materials.[25,26] By con-
trast, catalytic hydrogenation of nitroarene compounds
is one of the most powerful tools for the excellent synthe-
sis of primary amines. Importantly, about 85% of aniline
is produced via the environmentally benign catalytic
hydrogenation of nitrobenzene using gaseous hydro-
gen.[27] This chemical transformation has been promoted
using a range of homogeneous and heterogeneous cata-
lysts.[28–33] However, many of these catalytic systems
have some limitations such as tedious and time‐
consuming workup processes, necessity of extra additives,
and incomplete noble metal recovery from such complex
systems. Hence, development of a facile, fast, and conve-
nient method for the synthesis of aniline derivatives is of
great importance.
Considering the importance of the morphology of
MNPs and their coating with inert materials for the

SADJADI AND HERAVI
3 of 14
reaction mixture was allowed to cool down to room tem-
perature, and then magnetically separated. The TH‐Fe2-
O3@MF was washed with a mixture of deionized
water/EtOH repeatedly and dried at 100 °C for 10 hr to
obtain pure product. The yield of this process was 75%.
2.2 | Synthesis of the catalyst
2.2.1 | Synthesis of nanomagnetic hollow
tetrahedral Fe2O3 (TH‐Fe2O3): Compound 1
Typically, to a solution of FeCl3·6H2O (10 mmol) in
deionized water (65 mL), a solution of urea (16 mmol)
in deionized water (60 mL) was added. The resulting mix-
ture was then stirred vigorously. After 10 min, the reac-
2.2.3 | Preparation of TH‐Fe2O3@MF‐Glu
core–shell: Compound 3
tion mixture was transferred into
a Teflon‐lined
An aqueous solution of glucose (40 mL, 0.3 M) was pre-
pared and then added to an aqueous suspension of TH‐
Fe2O3@MF (10 g L−1). The mixture was thoroughly dis-
persed by sonication with power of 200 W for half an
hour and then transferred into a Teflon‐lined stainless
steel autoclave and kept at 200 °C for 24 hr. At the end
of the hydrothermal process, the reactor was cooled down
to room temperature and the MNPs were magnetically
stainless‐steel autoclave (150 mL capacity) and main-
tained at 200 °C for 8 hr. At the end of the hydrothermal
process, the reactor was cooled down to room tempera-
ture and the MNPs were separated using an external
magnet, washed with a mixture of deionized water–EtOH
repeatedly, and dried at 100 °C for 12 hr in an oven.
2.2.2 | Synthesis of the TH‐Fe2O3@MF
core–shell: Compound 2
separated, washed with
a
mixture of deionized
water/EtOH, and dried at 60 °C for 12 hr.
Initially, melamine (100 mg) and formaldehyde (40 mg)
were mixed in 10 mL deionized water. Next, the pH of
the mixture was adjusted to 8.0 by adding Na2CO3 solution.
Under continuous magnetic stirring at 70 °C, pre‐
polymerization of melamine and formaldehyde began
and continued for 1 hr. Then, TH‐Fe2O3 (0.5 g) was dis-
persed in 10 mL deionized water. Subsequently, 100 mg
formaldehyde was added and the pH of the mixture was
adjusted to 9.0 by adding Na2CO3 solution. The mixture
was then well‐dispersed under ultrasonic irradiation for
1 hr. After sonication, the pre‐polymer solution was mixed
with the TH‐Fe2O3 mixture. In the following steps, the pH
of the mixture was adjusted to 5.0 by adding HCl aqueous
solution. Then, the prepared mixture was transferred into
a glass tube and heated to 120 °C and kept at this tempera-
ture for half an hour. Upon completion of the reaction, the
2.2.4 | Immobilization of Pd NPs on the
TH‐Fe2O3@MF‐Glu core–shell: Compound 4
The Pd@TH‐Fe2O3@MF‐Glu was synthesized by conven-
tional impregnation‐reduction method. Typically, TH‐Fe2-
O3@MF‐Glu (3 g) was stirred in toluene (40 mL) in a flask
for half an hour. Then, a solution of Pd salt, Pd(OAc)2, in
toluene (40 mL, 0.01 M) was added under stirring condi-
tion for 6 hr. In the following steps, an aqueous solution
of NaBH4 in MeOH (15 mL, 0.2 N) that served as the reduc-
ing agent was added into the solution under N2 atmo-
sphere and then the resulting mixture was stirred
vigorously for 2 hr. After magnetic separation, the solid
product was washed with a mixture of deionized
water/EtOH several times and then dried at 60 °C for 12 hr.
FIGURE 1 Schematic routes for the synthesis of Pd@TH‐Fe2O3@MFC. r.t., room temperature

4 of 14
SADJADI AND HERAVI
2.2.5 | Synthesis of Pd@TH‐Fe2O3@MFC
core–shell: Compound 5
2.2.6 | Synthesis of control catalysts
To further examine the effect of different morphologies of
MNPs and the carbon sources, several control catalysts,
compounds 17, 9, 7, 11, 13, and 16, were prepared and
their catalytic activities were compared with that of com-
pound 5. Compound 8 was synthesized by using a similar
procedure employed for the synthesis of compound 1,
except that a FeCl3·6H2O:urea ratio of 10:12 (mmol/
mmol) was applied instead of the ratio of 10:16.
To prepare palladated, magnetic, N‐rich porous carbon
core–shell system 5, Pd@TH‐Fe2O3@MF‐Glu was car-
bonized at 450 °C with a heating rate of 3 °C min−1 for
7 hr in a tubular furnace under argon flow followed by
the heat treatment at 750 °C for 1 hr. In Figure 1, the
schematic procedure for the synthesis of catalyst 5 is
depicted.
FIGURE 2 Schematic routes for the synthesis of the control catalysts. r.t., room temperature

SADJADI AND HERAVI
5 of 14
Compound 6 was prepared using the same procedure used
for compound 1, except that glucose was also applied as
one of the precursors. In this sample, the ratio of
FeCl3·6H2O:urea:glucose was 10:8:1.2 (mmol/mmol/
mmol). Compounds 7 and 9 were produced by palladating
compounds 6 and 8, respectively. Compound 17 was syn-
thesized by palladating compound 1. The procedure for
the immobilization of Pd nanoparticles was similar to the
method used for the synthesis of catalyst 5, except that dif-
ferent morphologies of bare h‐Fe2O3 were used instead of
TH‐Fe2O3@MF‐Glu. Compound 11 was also synthesized
by immobilization of Pd nanoparticles on the surface of
compound 10, which was achieved through carbonization
of compound 3. To prepare compound 13, compound 2 was
successfully palladated and then carbonized. To synthesize
compound 16, the outer shell of compound 1 was directly
covered by glucose, without using the melamine–
formaldehyde polymeric shell, to furnish compound 14.
Next, the product (TH‐Fe2O3‐Glu) was palladated to form
compound 15 and then carbonized through the similar car-
bonization procedure used for the synthesis of compound
5. In Figure 2, the schematic procedures for the synthesis
of the control catalysts are depicted.
MNPs on the catalytic activity. Hence, three different
morphologies of hollow MNPs, namely, disc‐like, cylin-
der, and tetrahedral (compounds 1, 8, and 6 in
Figure 2), were simply prepared by altering the
FeCl3·6H2O:urea:glucose ratio and characterized by using
FESEM analysis (Figure 3). As shown in Figure 3, the
three targeted morphologies were successfully prepared.
The EDS analyses also confirmed the success of the pro-
cedures for the synthesis of MNPs. Verifying the three
morphologies, Pd nanoparticles were immobilized on the
three as‐prepared MNPs using a similar protocol (see the
“Experimental” section). To elucidate whether the mor-
phology of MNPs can affect the catalytic activity, the cata-
lytic activities of three palladated MNPs (compounds 7, 9,
and 17, Figure 2) for the hydrogenation of nitrobenzene as
a model substrate were studied (Table 1). Prior to the
hydrogenation test, the reaction condition was optimized
(Table S1) by measuring the yield of the reaction using dif-
ferent solvents, temperatures, and catalyst loading. The
highest yield of the product was achieved at room temper-
ature in water using a 1.2 mol% catalyst. The results indi-
cated that compound 17, the palladated tetrahedral
MNPs, and compound 9, disc‐like MNPs, showed the
highest and the lowest catalytic activities, respectively.
This result established that the morphology of the MNPs
can effectively influence the catalytic activity. The higher
catalytic activity of MNPs with tetrahedral morphology
can be attributed to the presence of more angles in this
morphology that can promote more Pd anchoring. To ver-
ify this assumption, the Pd loadings of compounds 7, 9,
and 17 were estimated via ICP.
The results in Table 1 show that the contents of Pd in
those samples decreased in an order similar to that of the
catalytic activity (compound 17 > compound 7 > com-
pound 9), indicating the role of the MNPs morphology in
the Pd loading. To further investigate the origin of the
higher catalytic activity of the tetrahedral MNPs, the spe-
cific surface area of three compounds 7, 9, and 17 were
measured via BET. The results indicated higher specific
surface area of compound 17 (Table 1). Considering all of
these results, the superior catalytic activity of compound
17 was attributed to the higher Pd loading and specific sur-
face area. It is worth noting that Pd states immobilized on
different support can also affect the catalytic activity.
To find the most active morphology, the effect of the
nature of the carbon shell on the MNPs was investigated
by examining three different shells. First, compounds 13
and 16 in Figure 2 were prepared by coating tetrahedral
MNPs with melamine–formaldehyde polymer and
glucose‐derived carbon followed by palladating and car-
bonization. As tabulated, compound 13, in which the car-
bon shell is nitrogen rich, showed higher catalytic activity
than compound 16, in which the carbon shell is nitrogen
2.2.7 | Hydrogenation of nitrobenzene
In a two‐necked flask containing a solution of nitroarene
(1 mmol in 5 mL of deionized water), Pd@TH‐
Fe2O3@MFC was added as a catalyst (1.2 mol%). Subse-
quently H2 gas as a reducing agent was purged (1 atm.).
The glass reactor was stirred vigorously (400 rpm) and
the progress of the reduction reaction was then moni-
tored via thin‐layer chromatography using an n‐hexane‐
to‐ethyl acetate ratio of 9:1. Upon completion of the reac-
tion, the reactor was cooled and Pd@TH‐Fe2O3@MFC
catalyst was separated using an external magnet. The
recovered Pd@TH‐Fe2O3@MFC was washed several
times with a mixture of distilled water and EtOH, dried
in oven, and reused for the next run. The resulting filtrate
was collected and diluted with EtOH to obtain the final
product. All the obtained products were identified and
their formation was confirmed by FTIR spectroscopy.
3 | RESULTS AND DISCUSSION
3.1 | Investigation of the roles of MNPs
morphology, nature of carbon shell carbon
shell, and order of Pd immobilization in the
catalytic activity: Selecting the most
efficient catalyst
As mentioned, one of the aims of this research is investi-
gation of the effect of the morphology of bare hollow

6 of 14
SADJADI AND HERAVI
FIGURE 3 Field emission scanning electron microscopic and energy‐dispersive X‐ray spectroscopy analyses of h‐Fe2O3 with (a,b) disc‐like
morphology, (c,d) cylinder morphology, and (e,f) tetrahedral morphology

SADJADI AND HERAVI
7 of 14
TABLE 1 Comparison of the catalytic activity of the present
catalyst with the other prepared catalysts for the hydrogenation
reactiona
carbonization of the shell. The results (Table 1) showed
that the catalytic activity, specific surface area, and Pd
loading of this sample were higher than those of com-
pounds 7, 9, 17, and 13, but lower than those of com-
pound 5. This result indicated the importance of the
order of Pd incorporation.
Loading of Pd
nanoparticles SBET
Entry Catalyst
Yieldb (%) Time (min) (mmol/g)
(m2 g−1
)
1
2
3
4
5
6
7
Compound 7
Compound 9
45
30
75
75
75
75
75
75
75
0.021
0.013
0.028
0.06
75
64
3.2 | Characterization of the catalyst of
the choice, compound 5
Compound 17 65
Compound 5 100
Compound 11 85
Compound 13 80
Compound 16 68
78
451
155
124
92
0.047
0.041
0.033
In the previous section the Pd loading and specific surface
area of all the control catalysts were measured. In the
next step, compound 5 that showed the best catalytic
activity was further characterized. First, the morphology
of compound 5 was disclosed by recording the FESEM
image (Figure 4). As shown in Figure 4, although coating
with carbon shell disordered the morphology of the
MNPs, to some extent, the tetrahedral morphology is still
observable. The observation of C and N atoms in the EDS
analysis of the catalyst confirmed the incorporation of N‐
doped carbon shell. Moreover, the presence of Pd in this
analysis confirmed the successful immobilization of Pd
nanoparticles.
The TEM images of the catalyst are illustrated in
Figure 5, in which the large MNPs can be seen covered
with carbon shell. On the carbon shell, the Pd nanoparti-
cles with average size of 20 nm can be observed.
To further characterize the nature of the carbon shell
of the catalyst, Raman spectroscopy was applied
(Figure 6). The observation of two characteristic D and
G bands at 1359 and 1594 cm−1 is a proof for the graphitic
nature of the carbon shell.[34–36]
The porous nature of the catalyst was confirmed by
BET analysis. As depicted in Figure 7, the N2
adsorption–desorption isotherm of the catalyst is of type
IV, implying its porous nature. In addition, the specific
surface area of the catalyst was estimated to be 416
m2 g−1. According to the Barrett–Joyner–Halenda (BJH)
plot, the pore‐size distribution of Pd@TH‐Fe2O3@MFC
is narrow and the mesoporous size distribution centered
at 2.4 nm. The total pore volume of the catalyst was esti-
mated to be 35.8 m3 g−1. To understand the effect of cov-
ering MNPs with carbon shell, the specific surface area of
compounds 1, 3, and 4 was also calculated and compared
with that of the catalyst, compound 5. The results showed
that the bare MNPs had a low specific surface area
(51.4 m2 g−1). Upon covering the surface of MNPs with
melamine–formaldehyde and glucose‐derived carbon
sheet, this value increased to 362 m2 g−1. The specific sur-
face area of compound 4 (319 m2 g−1) was lower than that
of Pd‐free compound 3. This observation confirms that
the Pd nanoparticles are deposited on the surface of
aReaction condition: nitrobenzene (1 mmol), catalyst (1.2 mol%) in water
(5 mL) at room temperature (under 1 atmosphere H2 gas).
bIsolated yield.
free. This observation can be justified based on the capa-
bility of N‐rich carbon shell to effectively anchor Pd
nanoparticles through electrostatic interactions.
Again, ICP analysis was exploited to verify this assump-
tion. As shown in Table 1, the Pd loading in compound 13
was higher than that in compound 16, confirming the role
of the nature of the carbon shell and its nitrogen function-
alities in the catalytic activity of the resulting catalyst.
Notably, the BET results confirmed that the specific sur-
face area of compound 13 was higher than that of com-
pound 16. This can also contribute to the catalytic
activity through improving the dispersion of Pd nanoparti-
cles on the support. To further investigate the effect of car-
bon shell, another catalyst, compound 3, with
combinatory carbon shell was prepared through initial
coating of MNPs with melamine–formaldehyde followed
by introduction of a secondary shell (Glu) through hydro-
thermal treatment of glucose. Compound 3 was then
palladated and carbonized to furnish compound 5. Inter-
estingly, use of compound 5 as a catalyst for the hydroge-
nation of nitrobenzene indicated the superior catalytic
activity of this sample. The ICP analysis of this compound
implied the highest loading of Pd in this nanocomposite,
indicating that the dual shell was more effective for
anchoring of Pd nanoparticles. Moreover, the BET results
showed the high specific surface area of this sample that
was significantly higher than that of compounds 13 and
16. Taking these results into account, the high catalytic
activity of compound 5 can be attributed to its high spe-
cific surface area and Pd loading.
Finally, to elucidate the role of the order of Pd loading
in the catalytic activity, another control catalyst, com-
pound 11, was prepared through a similar protocol that
was used for the preparation of compound 5, except that
Pd loading was carried out at the final stage and after

8 of 14
SADJADI AND HERAVI
FIGURE 4 The (a,b) field emission scanning electron microscopic and (c) energy‐dispersive X‐ray spectroscopy (EDS) analyses of
compound 5
compound 3. Finally, upon carbonization, the specific
surface area showed an increase and reached to 416
m2 g−1. This can be due to the formation of porous car-
bon with high specific surface area.
To verify the formation of MNPs and Pd nanoparticles,
the XRD pattern of the catalyst was also obtained
(Figure 8). As shown, the XRD pattern of the catalyst con-
tains several characteristic bands. First, a broad band at
2θ = 20–27° is observable that can be attributed to the
graphite‐like carbon shell on the structure of the cata-
lyst.[37] The second group of the characteristic bands is
related to the Pd nanoparticles (observed at 2θ = 40°,
49°, 68°, 82.1°, and 86°). The third group of the character-
istic bands can be attributed to the MNPs (observed at
2θ = 30.3°, 35.6°, 43.2°, 54.0°, 57.3°, 63.0°, and 74.6°;
JCPDS card no. 39–1346).[6,38]
To further characterize the catalyst, FTIR spectroscopy
was performed. In the FTIR spectrum of the catalyst

SADJADI AND HERAVI
9 of 14
FIGURE 5 Transmission electron microscopy images of Pd@TH‐Fe2O3@MFC. NP, nanoparticle
FIGURE 6 Raman spectrum of
compound 5 (Pd@TH‐Fe2O3@MFC)
FIGURE 7 Nitrogen adsorption–desorption (ADS–DES) isotherm and Barrett–Joyner–Halenda (BJH) plot of Pd@TH‐Fe2O3@MFC

10 of 14
SADJADI AND HERAVI
FIGURE 8 X‐ray diffraction pattern of
Pd@TH‐Fe2O3@MFC
FIGURE 9 Fourier‐transform infrared
spectrum of compound 5 (Pd@TH‐
Fe2O3@MFC)
(Figure 9) the observed bands at 450–560 cm−1 can prove
the Fe–O stretching vibration.[38] The presence of
hydroxyl group can be confirmed by observing a broad
band at 3421 cm−1. The characteristic bands at 1658 and
1629 cm−1 can be attributed to the –C=N and –C=C
functionalities, respectively, in the carbon shell on the
periphery of MNPs.
Next, the magnetic property of compound 5 was
studied by room‐temperature VSM analysis. The mea-
sured maximum saturation magnetization (Ms) value
of Pd@TH‐Fe2O3@MFC was estimated to be
25.78 emu g−1 (Figure 10), which was lower than that
of bare MNPs (45.28 emu g−1). The relatively low Ms
value can be assigned to the presence of carbon shell
on the periphery of MNPs.[39] It is worth noting that
although the non‐magnetic carbon shell decreased the
magnetic feature of the catalyst, it was still magnetic
enough to be separated magnetically by using an
external magnet. To further investigate the effect of
the nature of the carbon shell on the magnetic proper-
ties of the final catalyst, VSM analysis of compounds
16 (the sample without the melamine–formaldehyde
shell) and 13 (the sample without glucose‐derived car-
bon) was performed. The results showed that similar
to Pd@TH‐Fe2O3@MFC, the Ms values of compounds
16 and 13 were lower than that of bare MNPs. However,
the magnetic properties of these samples (37.8 and
36.0 emu g−1, respectively) that possessed only one shell
were higher than that of Pd@TH‐Fe2O3@MFC that
contained both melamine–formaldehyde and glucose‐
derived carbon shells. Moreover, the comparison of this
value for both samples confirmed that the nature of car-
bon shell had an insignificant effect on the magnetic
property of the catalyst and the Ms values of com-
pounds 16 and 13 were very similar (37.8 and
36.0 emu g−1, respectively).

SADJADI AND HERAVI
11 of 14
FIGURE 10 Vibrating‐sample
magnetometry analysis of compound 5
are summarized. As tabulated, a diverse range of Pd‐
based catalysts have been utilized for promoting this reac-
tion. In some protocols, the chemical reducing agents
were applied. Despite the efficiency of the chemical
reducing agent, use of hydrogen gas is more appealing
due to the environmental concerns. By contrast, in some
of the H2‐assisted procedures, high pressures of hydrogen
gas were used that are not only costly, but also unsafe.
Regarding the reaction solvent and temperature, use of
water as green and available solvent and performing the
reaction at room temperature are more economical and
environmentally benign. Comparing the reaction times,
it can be seen that Pd@TH‐Fe2O3@MFC could promote
the reaction in reasonable reaction time. Considering all
of the results, it can be concluded that Pd@TH‐
Fe2O3@MFC can be considered as an efficient catalyst
for the hydrogenation of nitroarenes.
3.3 | Generality of the developed protocol
Notably, Pd@TH‐Fe2O3@MFC not only could promote
the hydrogenation of nitrobenzene, but also could effec-
tively catalyze the hydrogenation of steric 1‐
nitronaphthalene (Table 2, entry 2) and NO2‐substituted
aromatic ketones (Table 2, entry 3) selectively. Moreover,
the hydrogenation of electron‐withdrawing and electron‐
donating nitrobenzenes (Table 2, entries 4 and 5) resulted
in the corresponding anilines with moderate yields in
longer reaction times.
To show the merit of Pd@TH‐Fe2O3@MFC, its cata-
lytic activity for the hydrogenation of nitrobenzene was
compared with some of the Pd‐based catalysts reported
in the literature. In Table 3, the efficiency of the previ-
ously reported catalysts and the reaction details such as
reaction time, temperature, solvent, and reducing agent
TABLE 2 Pd@TH‐Fe2O3@MFC catalyzed hydrogenation of nitro
arenesa
3.4 | Catalyst recyclability
Yieldb
Recyclability of Pd@TH‐Fe2O3@MFC for the hydrogena-
tion of nitrobenzene was examined by repeating the
recovery–reusing cycle for 10 consecutive reaction runs
under similar hydrogenation conditions (Figure 11). As
illustrated in Figure 11, Pd@TH‐Fe2O3@MFC exhibited
excellent catalytic activity and recycling up to four reac-
tion runs and only an insignificant decrease of aniline
yield was observed upon each recycling. Upon further
recycling, the decrease of the yield of the reaction
increased and upon recycling for 10 reaction runs, 32%
loss of the catalytic activity was observed. From a struc-
tural point of view, the FTIR spectroscopy of the recycled
catalyst confirmed that recycling did not induce signifi-
cant change in the structure of the catalyst and both fresh
and recycled Pd@TH‐Fe2O3@MFC for 10 reaction runs
showed similar FTIR spectra (Figure 12).
Entry Reagent
Product
Time (h:min) (%)
1
1:15
100
2
3
3:00
90
2:00
94
4
5
2:25
3:13
70
65
aReaction condition: Nitro arene (1 mmol), Pd@TH‐Fe2O3@MFC (1.2 mol%)
in water (5 mL) at room temperature (under 1 atmosphere H2 gas).
bIsolated yield.
Finally, the Pd leaching of the catalyst after 10 reaction
runs was measured and compared with that of control

12 of 14
SADJADI AND HERAVI
TABLE 3 Comparison of the catalytic activities of Pd@TH‐Fe2O3@MFC with some of the reported catalysts in the literature for the
hydrogenation of nitrobenzene
Reducing
agent
Time (h:min) Temperature
(°C)
Yield
(%)
Entry Catalyst
Solvent
Ref.
1
Pd@TH‐Fe2O3@MFC (1.2 mol%)
Pd@Hal/di‐ureaa (1.5 wt%)
H2/1 atm
H2/1 bar
H2/1 atm
1:15
1:00
3:00
1:00
r.t.
50
r.t.
40
r.t.
r.t.
50
r.t
H2O
H2O
THF
EtOH
100
100
100
99
This work
[40]
2
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[45]
[45]
3
PdNP(0.5%)/Al2O3 (0.3 g)
4
Pd/PPh3@FDU‐12b (8.33 × 10−4 mmol Pd) H2/10 bar
5
Pd‐(CH3)2NHBH3 (6 mol%)
APSNPc (1 mol%)
(CH3)2NHBH3 0:10
H2O/MeOH 99
6
H2/40 atm
NaBH4
2:00
1:30
1.15
2:00
1:30
1:30
1:30
EtOH
100
91
90
99
98
98
85
7
Pd/graphene
H2O/EtOH
EtOAc
8
Pd0‐AmP‐MCFd (0.5 mol%)
PdNP@PPh2‐PEGPIILPe (0.05 mol%)
Pd NPs/RGO (6 mg)
H2/1 bar
NaBH4
9
r.t.
50
50
50
H2O
10
11
12
NaBH4
H2O/EtOH
H2O/EtOH
H2O/EtOH
PdCu/graphene (2 mol% Pd)
PdCu/C (2 mol% Pd)
NaBH4
NaBH4
aPd immobilized on functionalized Hal.
bPd nanoparticles (1.1 nm) with triphenylphosphine (PPh3) cross‐linked in the nanopore of FDU‐12.
cActivated palladium sucrose nanoparticles
dPd nanoparticles supported on aminopropyl‐functionalized siliceous mesocellular foam.
ePalladium nanoparticles stabilized by lightly cross‐linked phosphine‐decorated polymer immobilized ionic liquids and their PEGylated counterparts (PEGPIIL).
FIGURE 11 Yield of aniline after each run of recycling the Pd@TH‐Fe2O3@MFC catalyst and the Pd leaching
catalysts (Table 4). As tabulated, the bare samples (com-
pounds 7, 9, and 17) that do not have any shell showed
higher Pd leaching compared with that of the encapsu-
lated ones. By comparing the Pd leaching of the sample
with glucose‐derived carbon shell with that of the sample
with melamine–formaldehyde shell (compounds 13 and
16), it was confirmed that the N‐rich shell suppresses
the Pd leaching more effectively. This can be attributed
to the interactions of the heteroatoms in carbon shell
with Pd nanoparticles. The measurement of Pd leaching
of compound 11 with that of 5 showed that the order of
Pd loading did not affect the Pd leaching significantly
and both of the catalysts showed similar leaching (0.005
mmol g−1).

SADJADI AND HERAVI
13 of 14
FIGURE 12 Fourier‐transform infrared
spectra of A: fresh and B: recycled
Pd@TH‐Fe2O3@MFC after 10 runs
Fe2O3@MFC, with excellent catalytic activity and recy-
clability. To study the effect of the nature of shell on the
catalytic activity, some control samples with N‐rich car-
bon derived from melamine–formaldehyde and N‐free
glucose‐derived carbon were also prepared and their cat-
alytic activities were compared with that of Pd@TH‐
Fe2O3@MFC. The results confirmed the higher catalytic
activity of the latter. This was attributed to the higher spe-
cific surface area and Pd loading of this sample. Examin-
ing the effect of the order of Pd immobilization showed
that incorporation of Pd prior to carbonization was more
effective than after thermal treatment.
TABLE 4 Comparison of the Pd leaching of the catalyst and
control samples after 10 reaction runsa
Leaching of Pd nanoparticles
Entry
Catalyst
(mmol/g)
1
2
3
4
5
6
7
Compound 7
Compound 9
Compound 17
Compound 5
Compound 11
Compound 13
Compound 16
0.012
0.0092
0.008
0.005
0.005
0.0062
0.0069
aReaction condition: nitrobenzene (1 mmol), catalyst (1.2 mol%) in water
(5 mL) at room temperature (under 1 atmosphere. H2 gas).
ACKNOWLEDGMENTS
The authors appreciate the partial supports from the
Iran Polymer and Petrochemical Institute and Alzahra
University. S. Sadjadi and Heravi thank Iran National
Science Foundation (INSF) for the Individual given grant
(No. 97009384). M. Malmir's contribution is also
appreciated.
4 | CONCLUSION
To investigate the effects of structural parameters on the
catalytic activity of magnetic core–shell catalyst, three
MNPs with three different morphologies, disc‐like, cylin-
der, and tetrahedral, were prepared and used as Pd
nanoparticles support. The results showed that the mor-
phology can affect the specific surface area and Pd
loading and leaching of the final catalyst, with the tetra-
hedral morphology identified as the best choice, as it
had the highest catalytic activity for the hydrogenation
of nitrobenzene. To improve the catalytic activity of
the catalyst and prevent the catalyst from aggregation of
MNPs, the catalyst was covered by a dual shell by
introduction of melamine–formaldehyde shell, followed
by coating with glucose‐derived carbon. The Pd immobi-
lization and carbonization led to a catalyst, Pd@TH‐
CONFLICT OF INTEREST
There are no conflicts to declare. The Iranian authors
declare that none of them are employed by a government
agency that has a primary function other than research
and/or education. Moreover, none of them are official
representative or on behalf of the government.
ORCID
Samahe Sadjadi https://orcid.org/0000-0002-6884-4328
Majid M. Heravi https://orcid.org/0000-0003-2978-1157

14 of 14
SADJADI AND HERAVI
[30] Y. Chen, L. Li, L. Zhang, J. Han, Colloid Polym. Sci. 2018, 296,
REFERENCES
567.
[1] G. Han, X. Li, J. Li, X. Wang, Y. S. Zhang, R. Sun, ACS Omega
2017, 2, 4938.
[31] F. Figueras, B. Coq, J. Mol. Catal. A: Chem. 2001, 173, 223.
[32] M. Liu, J. Zhang, J. Liu, W. W. Yu, J. Catal. 2011, 278, 1.
[33] M. Liu, W. Yu, H. Liu, J. Mol. Catal. A: Chem. 1999, 138, 295.
[34] J. Hou, C. Cao, F. Idrees, X. Ma, ACS Nano 2015, 9, 2556.
[2] X. Zhang, Y. Niu, Y. Yang, Y. Li, J. Zhao, New J. Chem. 2014,
38, 4351.
[3] T. Zare, N. Sattarahmady, Clin. Cancer Res. 2015, 7, 29.
[35] F. Yang, Z. Zhang, K. Du, X. Zhao, W. Chen, Y. Lai, J. Li, Car-
bon 2015, 91, 88.
[4] S. Sadjadi, M. M. Heravi, M. Malmir, J. Taiwan Inst. Chem. Eng.
2018, 86, 240.
[36] A. Esmaeili, M. H. Entezari, J. Colloid Interface Sci. 2014, 432,
[5] C. Jin, Y. Wang, H. Tang, K. Zhu, X. Liu, J. Wang, J. sol‐Gel Sci.
Technol. 2016, 77, 279.
19.
[37] B. Manoj, A. G. Kunjomana, IOP Conf. Ser.: Mater. Sci. Eng.
2015, 73, 1.
[6] S. Sadjadi, M. Malmir, M. M. Heravi, RSC Adv. 2017, 7, 36807.
[7] W. Zhou, L. Guo, Chem. Soc. Rev. 2015, 44, 6697.
[38] A. Elhampour, M. Malmir, E. Kowsari, F. Boorboor Ajdari, F.
[8] W. Wu, Q. He, C. Jiang, Nanoscale Res. Lett. 2008, 3, 397.
Nemati, RSC Adv. 2016, 6, 96623.
[9] S. Kazemi Movahed, N. Farajinia Lehi, M. Dabiri, J. Catal.
2018, 364, 69.
[39] S. Sadjadi, M. M. Heravi, M. Malmir, Appl. Organomet. Chem.
2018, 32, e4029.
[10] Y. Yang, X. Jiang, J. Chao, C. Song, B. Liu, D. Zhu, Y. Sun, B.
Yang, Q. Zhang, Y. Chen, L. Wang, Sci. China Mater. 2017,
60, 1129.
[40] S. Dehghani, S. Sadjadi, N. Bahri‐Laleh, M. Nekoomanesh‐
Haghighi, A. Poater, Appl. Organomet. Chem. 2019, 33, e4891.
[41] S. Agrahari, S. Lande, V. Balachandran, G. Kalpana, R. Jasra,
[11] J. K. Shon, X. Jin, Y. S. Choi, J. G. Won, Y. K. Hwang, D. J.
J. Nanosci. Curr. Res. 2017, 2, 2572.
You, C. Li, J. M. Kim, Carbon Lett. 2016, 20, 66.
[42] M. Guo, H. Li, Y. Ren, X. Ren, Q. Yang, C. Li, ACS Catal. 2018,
8, 6476.
[12] Z. Z. Jiang, Z. B. Wang, H. Rivera, W. L. Qu, D. M. Gu, Fuell
Cells 2014, 14, 660.
[43] N. M. Patil, M. A. Bhosale, B. M. Bhanage, RSC Adv. 2015, 5,
[13] T. A. Centeno, A. B. Fuertes, J. Membr. Sci. 1999, 160, 201.
86529.
[14] G. Han, X. Wang, J. Hamel, H. Zhu, R. Sun, Sci. Rep. 2016, 6,
[44] D. Samsonu, M. Brahmayya, B. Govindh, Y. Murthy, S. Afr. J.
Chem. Eng. 2018, 25, 110.
38164.
[15] Z. Shen, Y. Luo, Q. Wang, X. Wang, R. Sun, ACS Appl. Mater.
Interfaces 2014, 6, 16147.
[45] Y.‐S. Feng, J.‐J. Ma, Y.‐M. Kang, H.‐J. Xu, Tetrahedron 2014,
70, 6100.
[16] M. Mastalerz, Angew. Chem. Int. Ed. 2008, 47, 445.
[46] O. Verho, K. P. Gustafson, A. Nagendiran, C. W. Tai, J. E.
[17] F. J. Uribe‐Romo, J. R. Hunt, H. Furukawa, C. Klöck, M.
O'Keeffe, O. M. Yaghi, J. Am. Chem. Soc. 2009, 131, 4570.
Bäckvall, ChemCatChem 2014, 6, 3153.
[47] S. Doherty, J. Knight, T. Backhouse, A. Bradford, F. Saunders,
R. Bourne, T. Chamberlain, R. Stones, A. Clayton, K. Lovelock,
Cat. Sci. Technol. 2018, 8, 1454.
[18] G. A. Zaharias, H. H. Shi, S. F. Bent, Thin Solid Films 2006,
501, 341.
[19] J. Du, F. Cheng, S. Wang, T. Zhang, J. Chen, Sci. Rep. 2014,
4, 4386.
[48] M. Nasrollahzadeh, S. M. Sajadi, A. Rostami‐Vartooni, M.
Alizadeh, M. Bagherzadeh, J. Colloid Interface Sci. 2016,
466, 360.
[20] Y. Zhao, K. Watanabe, K. Hashimoto, J. Am. Chem. Soc. 2012,
134, 19528.
[21] I. Ghiviriga, D. C. Oniciu, Chem. Commun. 2002, 22, 2718.
SUPPORTING INFORMATION
[22] J. S. Lee, G. S. Park, S. T. Kim, M. Liu, J. Cho, Angew. Chem.
Int. Ed. 2013, 52, 1026.
Additional supporting information may be found online
in the Supporting Information section at the end of the
article.
[23] X. Lin, X. Li, F. Li, Y. Fang, M. Tian, X. An, Y. Fu, J. Jin, J. Ma,
J. Mater. Chem. A 2016, 4, 6505.
[24] N. B. McKeown, P. M. Budd, Chem. Soc. Rev. 2006, 35, 675.
[25] J. H. Wang, Z. L. Yuan, R. F. Nie, Z. Y. Hou, X. M. Zheng, Ind.
Eng. Chem. Res. 2010, 49, 4664.
How to cite this article: Sadjadi S, Heravi MM.
Pd@tetrahedral hollow magnetic nanoparticles
coated with N‐doped porous carbon as an efficient
catalyst for hydrogenation of nitroarenes. Appl
Organometal Chem. 2019;e5229. https://doi.org/
10.1002/aoc.5229
[26] P. Lara, K. Philippot, Cat. Sci. Technol. 2014, 4, 2445.
[27] Z. Y. Sun, Y. F. Zhao, Y. Xie, R. T. Tao, H. Y. Zhang, C. L.
Huang, Z. M. Liu, Green Chem. 2010, 12, 1007.
[28] B. N. Saïb, P. Grange, P. Verhasselt, F. Addoun, V. Dubois,
Appl. Catal. A 2005, 286, 74.
[29] N. Arora, A. Mehta, A. Mishra, S. S. Basu, App. Clay Sci. 2018,
151, 1.