W. Xie et al.
Applied Catalysis A, General 623 (2021) 118249
ꢀ 1
(namely Co-Co3O4/NGr@C) can provide the best activity for the hy-
drogenation of nitroarenes with H2 gas, revealing that the Co-Nx sites are
ultrahigh turnover frequency (TOF) up to 310 mol4-NP mol minꢀ 1
.
Co
vital to the catalytic activity. Shortly after that, such
a
2. Experiment section
Co-Co3O4/NGr@C catalyst was used to catalyze the CTH of nitroarenes
by using formic acid as the hydrogen donor, in which excellent selec-
tivity was proved by the unprecedented tolerance toward functional
groups, such as olefins, ketone, aldehyde, nitrile, halides, ester, and
amide [29]. Since then, much effort has been devoted to the preparation
and utilization of the Co-N/C catalysts, in which the methods used to
prepare these catalysts are basically the same, involving the pyrolysis of
either carbon-supported/unsupported Co(II)/N-ligand complexes [30]
or Co-salt/carbon-precursor mixture followed by nitrogenation [31].
Among these Co(II)/N-ligand complexes, the zeolite-type Co-imidazo-
late framework, namely ZIF-67, is one of the most commonly used
precursor [32–36]. Li and co-workers reported, for the first time, the use
of ZIF-67-derived Co-N/C catalyst in the CTH of nitroarenes [37].
Although the nitroarenes were hydrogenated to the corresponding
aminoarenes with high chemoselectivity, the activity of Co-N/C catalyst
remained to be improved due to the low BET specific surface area (SBET
≈ 130 m2 gꢀ 1), which would be adverse to the exposure of active sites as
well as the mass transfer. Later on, the catalytic activity of
ZIF-67-derived Co-N/C catalysts has been greatly enhanced through the
use of hard templates (i.e., SiO2 [34], SBA-15 [38], graphene oxide [39],
carbon nanotubes [40], etc.) and bimetallic ZIF precursors [41], or
through the pre-treatment of ZIF-67 with heteroatom-enriched organic
molecules (i.e., urea, thioacetamide) [42,43] and polymers [44,45]. In
these ways, the Co-N/C catalysts can be engineered to have high SBET
and/or to possess highly dispersed active sites (i.e., Co-Nx species), in
which the high SBET together with high mesoporosity could lead not only
to efficient mass transport of reactants and/or products, but also to high
exposure of catalytic sites [46,47]. Meanwhile, the highly dispersed
active sites that are easily accessible to reactants are also a reason for the
improved activity. Apart from 2-methylimidazole that involved in
2.1. Synthesis of POP-gm and POP-gm/Co(II) complex
In a typical synthesis (Scheme 1), melamine (19.1 g, 151 mmol),
formic acid (10 mL), and DMF (700 mL) were added to a flask and stirred
at 80 ◦C under a N2 atmosphere. Then glyoxal (40 wt% aqueous solution,
26.0 mL, 229 mmol) was injected via a syringe. The reaction mixture
was heated to reflux for 96 h, after cooling to room temperature, the
resultant POP-gm was filtered, washed thoroughly with DMF and
ethanol, and then dried in vacuum at 60 ◦C for 12 h (yield, 85 %).
Subsequently, the as-synthesized POP-gm (10.0 g) was dispersed in 300
mL of dry DMF under a N2 atmosphere, and then Co(NO3)2⋅6H2O (10.0
g, 34.0 mmol) was added. The mixture was stirred at 80 ◦C for 4 h. After
cooling to room temperature, the resulting POP-gm/Co(II) complex was
filtered, washed with ethanol, and dried in vacuum.
2.2. Synthesis of Co-N/C-x catalysts
The as-synthesized POP-gm/Co(II) complex was separately pyro-
lyzed in a tube furnace at 850 ◦C, 950 ◦C, and 1050 ◦C for 2 h in a N2
atmosphere at a heating rate of 4 ◦C minꢀ 1. The resultant carbon
products were washed with 1.0 M HCl aqueous solution at 60 ◦C for 2 h
to remove CoOx nanoparticles. After vacuum drying, the obtained Co/N-
codoped porous carbons were named as Co-N/C-x (x = 850, 950, and
1050, denoting the pyrolysis temperatures). For comparison, Co/N-
codoped carbon without washing with 1.0 M HCl was also prepared at
950 ◦C and was named as Co-CoOx-N/C-950. In addition, N-doped car-
bon (namely N/C-950) was prepared by directly pyrolyzing POP-gm at
950 ◦C for 2 h in a N2 atmosphere.
ZIF-67, also other ligands, such as α-diimine [30], benzimidazole [47],
2.3. CTH of 4-nitrophenol (4-NP) to 4-aminophenol (4-AP)
phthalocyanine [48], and histidine [49], were used to synthesize such
Co-N/C catalysts.
In a typical experiment, the Co-N/C-950 catalyst (2.0 mg) was
dispersed in 30 mL of 4-NP aqueous solution (18 mM, 0.54 mmol) by
sonicating for 20 s followed by stirring at 25 ◦C. And then NaBH4 (0.30 g,
8.0 mmol) was added into the mixture under vigorous stirring, the re-
action progress was monitored by UV–vis spectroscopy.
Although considerable progress has been achieved in the preparation
and application of Co-N/C catalysts, the catalytic performance and the
precursor utilization efficiency are still needed to improve. Herein, we
present a poly(Schiff base) precursor for the scalable synthesis of
ultrahighly efficient Co-N/C catalyst with high mesoporosity for the
selective CTH of nitroarenes using NaBH4 as the hydrogen donor. The
poly(Schiff base) precursor, namely POP-gm, was firstly synthesized by
a facile, one-pot, Schiff base polycondensation of glyoxal and melamine
in DMF in the presence of formic acid catalyst. After coordination to Co
(II) ion, the obtained POP-gm/Co(II) complex was directly pyrolyzed at
950 ◦C, yielding Co-N/C-950 catalyst with a Co loading of 7.04 wt%.
Benefiting from the porous structure and the strong Co(II)-binding af-
finity of POP-gm precursor, the as-fabricated Co-N/C-950 catalyst has a
high SBET value (865 m2 gꢀ 1) with external-porosity up to 98.8 %, a high
N content with high proportions of pyrrolic-N, pyridinic-N and Co-Nx,
together with highly dispersed Co(0) nanoparticles with average size
below 20 nm. Considering the strong coordination interaction between
Co atoms and pyridinic-N/pyrrolic-N, the high percentage of such N-
configurations is helpful to stabilize Co species. In addition, the high
external-porosity of Co-N/C-950 catalyst is also important to enhance its
activity and stability, since smaller mesoporosity is crucial for the full
exposure of active sites along with effective mass transfer, while bigger
mesoporosity and macroporosity can act as the reservoirs for reactants,
therefore favoring the CTH reaction. As a result, the Co-N/C-950 catalyst
showed ultrahigh activity, selectivity, and stability for the CTH of
nitroarenes with NaBH4 as the hydrogen donor. More importantly, the
Co(0) particles embedded in the Co-N/C-950 catalyst can be further
transformed to Co4N phase by a facile nitridation reaction. By compar-
ison, the as-nitridated Co-N/C-950 catalyst (namely Co4N-N/C-950) has
an even higher activity for the CTH of 4-nitrophenol (4-NP), with an
3. Results and discussion
3.1. Synthesis and characterization of Co-N/C-x catalysts
In this study, the POP-gm precursor was synthesized by formic acid-
catalyzed Schiff base condensation of glyoxal and melamine in DMF
(Scheme 1). The formation of imine-linked, porous polymeric network
was verified by FT-IR spectroscopy and porosity analysis. As illustrated
in Fig. S1, a strong imine (C N) stretch at 1628 cmꢀ 1 was observed for
–
–
–
–
POP-gm [50], clearly confirming the generation of C N bonds in the
polymeric network. The N2 adsorption/desorption analysis indicated
that the SBET and total pore volume (Vt) values were around 281 m2 gꢀ 1
and 0.612 cm3 gꢀ 1, respectively, for the POP-gm (Fig. S2). Given the
strong coordination ability of Schiff base toward transition metal ions,
the coordination of POP-gm to Co(NO3)2 was performed at 80 ◦C in dry
DMF, yielding POP-gm/Co(II) complex (Scheme 1). The formation of
POP-gm/Co(II) complex was also characterized by FT-IR spectroscopy
–
and porosity analysis. As expected, the imine (C N) stretch at 1628
–
cmꢀ 1 for free POP-gm was shifted to 1596 cmꢀ 1 after the coordination to
Co(II) ion [50] (Fig. S1), whereas the SBET and Vt values were reduced to
91 m2 gꢀ 1 and 0.286 cm3 gꢀ 1, respectively, for the POP-gm/Co(II)
complex (Fig. S2). And meanwhile, the mesopore-size of the complex
became smaller by comparison with the free POP-gm, confirming the
pore-filling of Co(II) in POP-gm network. The Co loading in the
POP-gm/Co(II) complex was determined by TGA analysis. Compared
2