Scale-up biopolymer-chelated fabrication of cobalt nanoparticles encapsulated in N-enriched graphene shells for biofuel upgrade with formic acid

Article abstract of DOI:10.1039/c9gc01720h

Exploring both high-performance catalytic materials from non-edible lignocellulosic biomass and selective hydrodeoxygenation of bioderived molecules will enable value-added utilization of renewable feedstocks to replace rapidly diminishing fossil resources. Herein, we developed a scale-up and sustainable method to fabricate gram-quantities of highly dispersed cobalt nanocatalysts sheathed in multilayered N-doped graphene (Co@NG) by using a biomacromolecule carboxymethyl cellulose (CMC) as a raw material. The ionic gelation of CMC, urea and Co²⁺ ions leads to uniform dispersion and chelation of different species, consequently resulting in the formation of highly distributed Co nanoparticles (NPs) (10.91 nm) with N-enriched graphene shells in the solid-state thermolysis process. The usage of urea as a non-corrosive activation agent can introduce a porous belt-like nanostructure and abundant doped nitrogen. Among all the prepared catalysts in this work, the optimized Co@NG-6 with the largest specific surface area (627 m² g^-1), the most and strongest basic sites, and the highest proportion of pyridinic-N (37.6%) and mesopores exhibited excellent catalytic activity (99% yield of 2-methoxy-p-cresol) for base-free transfer hydrodeoxygenation (THD) of vanillin using bioderived formic acid (FA) as a H source at 160 °C for 6 h. The poisoning tests and electron paramagnetic resonance (EPR) spectra verified that the strong interaction between N atoms and encapsulated Co NPs provided synergistic effects, which were essential for the outstanding catalytic performance of Co@NG-6. The deuterium kinetic isotope effect study clearly demonstrated that the formation of Co-H^-via β-hydride elimination and protonation was the rate-determining step, and protic N-H⁺ and hydridic Co-H^- were considered to be active intermediate species in the THD reaction. Furthermore, Co@NG-6 was highly stable for recycling owing to the graphene shells preventing Co NPs from corrosion and aggregation.

Full text of DOI:10.1039/c9gc01720h

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. Zhou, F. Dai, C.
Dang, M. Wang, F. Lu, D. Liu and H. Qi, Green Chem., 2019, DOI: 10.1039/C9GC01720H.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
Scale-up Biopolymer-Chelated Fabrication of Cobalt Nanoparticles Encapsulated
in N-enriched Graphene Shells for Biofuel Upgrade with Formic Acid
Shenghui Zhou, a Fanglin Dai, a,c Chao Dang, a Ming Wang, a Detao Liu, a Fachuang Lu,
a and Haisong Qi*, a,b
a
State Key Laboratory of Pulp and Paper Engineering, South China University of
Technology, Guangzhou 510640, China.
b
Guangdong Engineering Research Center for Green Fine Chemicals, Guangzhou
510640, China.
c
School of Chemistry and Chemical Engineering, Guangxi University, Nanning
530004, China.
Abstract: Both exploring high-performance catalytic materials from non-edible
lignocellulosic biomass and selective hydrodeoxygenation of bioderived molecules will
enable the value-added utilization of renewable feedstocks to replace rapidly
diminishing fossil resources. Herein we developed a scale-up and sustainable method
to fabricate gram-quantities of highly dispersed cobalt nanocatalysts sheathed in
multilayered N-doped graphene (Co@NG) by using biomacromolecule carboxymethyl
cellulose (CMC) as raw material. The ionic gelation of CMC, urea and Co2+ ions lead
to the uniformly dispersion and chelation of different species, consequently resulting in
the formation of high distributed Co nanoparticles (NPs) (10.91 nm) with N-enriched
graphene shells in solid-state thermolysis process. The usage of urea as non-corrosive
activation agents can introduce a porous belt-like nanostructure and abundant doped
nitrogen. Among all the prepared catalysts in this work, the optimized Co@NG-6 with
the largest specific surface area (627 m2 g−1), the most and strongest basic sites, the
highest proportion of pyridinic-N (37.6%) and mesopore exhibited excellent catalytic
activity (99% yield of 2-methoxy-p-cresol) for base-free transfer hydrodeoxygenation
(THD) of lignin molecule vanillin using bioderived formic acid (FA) as H source at 160
°C for 6 h. The poisoning tests and electron paramagnetic resonance spectra (EPR)
verified that the strong interaction between N atoms and encapsulated Co NPs provided
synergistic effects, which were essential for the outstanding catalytic performance of
Co@NG-6. The deuterium kinetic isotope effect study clearly demonstrated the
formation of Co-H- via β-hydride elimination and protonation was rate-determining
step, and the protic N-H+ and hydridic Co-H− were considered to be active intermediate
species in THD reaction. Furthermore, Co@NG-6 was highly stable for recycling
owing to graphene shells preventing Co NPs from corrosion and aggregation.

Green Chemistry
Page 2 of 30
View Article Online
DOI: 10.1039/C9GC01720H
1. Introduction
As global energy systems and traditional chemical industry gradually shift from
depleted fossil fuels to renewable resources, an urgency exists to explore new
technologies to drive the transformation of biomass to bio-fuels and industry-related
green products.1-4 The pyrolysis of lignocellulosic biomass is a prospective approach
for producing valuable liquid bio-oil.5-8 As an effective strategy, hydrodeoxygenation
is of vital importance for improving energy density of bio-oil by selectively removing
oxygen-containing groups.9-12
The traditional hydrodeoxygenation technology requires explosive fossil resources-
derived hydrogen gas that is dangerous and uncontrollable, often leading to
overhydrogenation and C–C cleavage.13-16 In contrast, as a versatile renewable reagent
with considerable H content (4.4 wt%), formic acid (FA) can be regarded as a secure,
cheap and convenient liquid H source for THD.17-18 Notably, FA could be accessible
from biorefinery processing, or by hydrogenation of CO2.19-21 However, a large number
of basic compounds (e.g., sodium/kalium salts, amines) were introduced as additives in
most transition metal-catalyzed FA transfer hydrogenation processes. This can lead to
the high expense in the separation and purification of target product.22-23 For instance,
M. Beller and collaborator reported catalytic transfer hydrogenation of nitroaromatics
with FA on cubane-type Mo3S4 cluster, but the base triethylamine was required.24
Moreover, acidic FA can result in severe corrosion and leaching of catalytic active
sites.25 Therefore, it is significant to design an acid-resistant catalyst that is low-cost,
sustainable and efficient for base-free THD of biomass molecules with FA.
In recent years, transition metal nanoparticles (e.g., Fe, Co, Ni and Ru) sheathed in
graphene shells or carbon nanotubes have emerged as a novel catalyst (called as
“chainmail catalyst”) and quickly attracted wide attentions due to their high activity
and stability under harsh conditions.26-28 Bao and co-workers reported that the
interaction between metallic core and graphene shells can change the work function of
shell, and therefore endow it with superior chemical activities.29 Later on, several
materials have been reported for high-performance electrocatalysts for water splitting
or evolution reactions.30-32 However, most of preparation processes are uneconomical,
such as the undesirable employment of expensive and non-renewable precursors, or the
complicated preparation process (e.g. chemical vapor deposition) with high energy
consumption or special instruments.33-34 These may limit its practical application.

Page 3 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
Another important issue is the realization of large-scale production of catalysts to meet
the needs of industrial applications, which remains a great challenge.33-34
Recently, several researchers have found that the edible monosaccharide or
disaccharide (e.g. glucose, sucrose, fructose, D-xylose, D-glucosamine hydrochloride)
can be used as building blocks for preparing analogous chainmail catalyst, indicating
that fabrication of catalysts from natural resources has great advantages for practical
applications owing to the low-cost, non-toxicity, renewability.35-40 Nevertheless, the
use of monosaccharide as raw materials is a controversial issue since it is principally
used as food or intermediate for drug synthesis.41 In contrary, the use of non-edible
lignocellulosic biomass is more meaningful for the preparation of catalytic materials
owing to the huge worldwide availability of plant-derived lignocelluloses including
agricultural wastes and forest residues. The utilization and transformation of non-edible
biomass components, such as lignin, cellulose and its derivatives, chitin, etc., into
platform molecules and catalytic materials play a key role in sustainable development.42
However, compared with soluble glucose, most lignocellulosic biomass is insoluble in
conventional solvents owing to the H bond interactions, making it difficult for the
dispersion of active species and manufacture of catalysts.43 Therefore, the application
of non-edible lignocellulosic biomass as raw material for preparing chainmail catalyst
is seldom reported. Cellulose, which is available from wood and agricultural residues,
is abundant, low-cost, renewable and environmentally benign.44 In our previous work,
we tried cellulose as raw material and NaOH–urea aqueous solution as solvent to
prepare a chainmail catalyst.45 Although failed, we found this catalyst can effectively
promote transfer hydrogenation of furfural. However, the agglomeration of large MnOx
particles (100–200 nm) was unfavourable for the exposure of catalytic active sites.
Besides, the massive use of corrosive strong bases NaOH as activating agent was not
practical.
Carboxymethylcellulose (CMC) is
a
type of cellulose derived from
carboxymethylation of the hydroxyl groups in cellulose molecules. It can dissolve in
hot water and chelate with a variety of metal ions (e.g., Ca2+, Mg2+, Zr4+ and Fe3+) due
to the existence of rich carboxyl groups.46 Here, we developed a novel approach to
synthesize gram-quantities of highly dispersed Co nanocatalysts sheathed by
multilayered N-enriched graphene shells by pyrolysis of CMC. In this method, water-
soluble CMC was chelated and gelled with urea and Co2+ ions, enabling the uniformly
distribution and coordination of each component on the CMC chains. Consequently,

Green Chemistry
Page 4 of 30
View Article Online
DOI: 10.1039/C9GC01720H
the solid-state thermolysis process of this precursors rendered the formation of high
dispersed Co NPs (10.91 nm) with N-enriched graphene shells (Co@NG). The urea
acted as effective and non-corrosive activation agents, which introduced a porous
structure with high specific surface area. More importantly, the doped-N derived from
urea served as abundant basic sites to promote the reaction. Therefore, the prepared
composite was a highly active and acid-tolerant catalyst for base-free THD of various
biomass-derived organic compounds. The structural properties of as-prepared catalysts
are systematically characterized by SEM, TEM, HAADF-STEM, XRD, BET, XPS,
CO2-TPD and Raman spectroscopy. To further investigate the THD mechanism by
Co@NG-6, poisoning tests, EPR tests, deuterium kinetic isotope effect experiments
were performed as well. In addition, heterogeneity tests of catalysts, leaching and
recyclability experiments were also carried out.
2. Experimental Section
2.1 Materials and reagents
Sodium carboxymethyl cellulose (CMC) (M.W. 90000, 50-100 mPa.s),
hydroxyethyl cellulose (250~450 mPa.s), hydroxypropyl cellulose (M.W. 100,000),
methyl cellulose (40000 mPa.s), α-cellulose (25μm), Co(CH3COO)2·4H2O (>99.0%),
urea (>99.5%) were bought from Shanghai Macklin Biochemical Co., Ltd, China.
Vanillin (>99.5%), furfural (>99.5%), vanillic alcohol (>99.0%), cinnamaldehyde
(>99.5%), phenylpropyl aldehyde (>99%), benzaldehyde (>99%), phenylacetaldehyde
(>99%), quinoline (>99.0%), nitrobenzene (>99.0%) were obtained from Aladdin
Industrial Corporation. 95 wt.% DCOOD in 5 wt.% D2O (98 atom % D), 95 wt.%
DCOOH in 5 wt.% D2O and 95 wt.% HCOOD in 5 wt.% D2O were bought from the
Sigma-Aldrich Industrial Corporation.
2.2 Synthesis of different samples
Preparation of catalysts. The N-enriched graphene shells encapsulated cobalt
nanocatalysts (Co@NG) were fabricated according to the following method. Typically,
0.93 g CMC were dissolved in 60 mL distilled hot water (100 °C) under mechanical
agitation to get a sticky and transparent CMC solution. Afterwards, 0.15g
Co(CH3COO)2·4H2O and 6.5 g urea were completely dissolved in ultrapure water. This
purple solution was added to the previous CMC solution under mechanical agitation for
4 min. This CMC gel was then heated for ionic gelation and further dehydration at 70 °C
in an oven for 20 h. Subsequently, the obtained dry block was smashed, placed in a

Page 5 of 30
Green Chemistry
covered porcelain boat, followed by pyrolysis in a horizontal tube furnace at 800 °C foVriew Article Online
DOI: 10.1039/C9GC01720H
2 h at a heating rate of 10 °C·min-1 under N2 flow. The carbonized samples were
magnetic stirred in 60 mL hydrochloric acid solution (7 wt.%) for 12 h to remove
residual chemicals and soluble cobalt. Then, the sample was separated by filtration and
totally washed with ultrapure water until the filtrate was neutral. After drying in vacuum
at 60 °C for 10 h, the sample was marked as Co@NG-6. For other Co@NG-U, U stands
for the amount of added urea when 1g CMC was used as carbon source. Replacing
Co(CH3COO)2·4H2O
with
other
metal
salts
(Mn(CH3COO)2·4H2O,
Ni(CH3COO)2·4H2O or Fe(NO3)3·9H2O) synthesized Mn@NG, Ni@NG or Fe@NG,
respectively. Replacing CMC with hydroxypropyl cellulose (HPC), microcrystalline
cellulose (MCC), methyl cellulose (MC) or hydroxyethyl cellulose (HEC) synthesized
Co@NG-HPC, Co@NG-MCC, Co@NG-MC or Co@NG-HEC. The preparation
procedure of nitrogen-doped carbon (NC-6) was same as that of Co@NG-6, but no
Co(CH3COO)2·4H2O was added in CMC solution.
A scale-up fabrication of Co@NG-6. The process route was same as that of Co@NG,
except for increasing the added amount of CMC, Co(CH3COO)2·4H2O, urea and
distilled water to 5 g, 0.8g, 35g and 300 mL, respectively. The obtained dry block as
precursors were carbonized in tube furnace in twice. It is worth noting that the formed
and sublimated carbon nitride during preparation may block the gas outlet. Hence, the
tube diameter of the outlet is 1 cm. Finally, 1.74 g Co@NG-6 catalysts can be obtained
(Figure S1). Due to the facile preparing process and the use of low-cost and abundant
biomass as raw material, this synthesis method can be expected to realize the large-
scale production for industrial application.
2.3 Catalyst characterization
The texture structures of as-prepared catalysts were evaluated by low temperature N2
adsorption–desorption at -196 ℃ using a Micromeritics ASAP 2020. Before
measurements, the catalysts were outgassed at 120 °C for 7 h. The BET specific surface
areas were calculated using the Brunauer−Emmett−Teller method. The average pore
volume and pore size were obtained by multipoint Barrett−Joyner−Halenda (BJH)
method. X-ray diffraction (XRD) patterns were carried out using a Rigaku
diffractometer (D/MAX/IIIA, 3 kW) with Cu Kα radiation (λ = 0.1543 nm, 40 kV, 30
mA). The cobalt content of the catalysts was measured with Spectro Arcos FHX22
inductively coupled plasma-optical emission spectrometer (ICP-OES). The structures

Green Chemistry
Page 6 of 30
View Article Online
and morphologies of samples were characterized by a high-resolution scanningDeOlIe: 1c0t.1r0o3n9/C9GC01720H
microscopy (SEM, MERLIN of ZEISS) and a high-resolution transmission electron
microscope (JEM-2100F) with EDX analysis (Bruker Xflash 5030T) operated at 200
kV. Raman spectra were obtained on a confocal laser LabRAM Aramis Raman
Spectrometer (HORIBA Jobin Yvon) operating with 532 nm excitation and the wave
number range used in the measurement was from 400 to 2000 cm−1. X-ray
photoelectron spectroscopy (XPS) were carried out on a Kratos Axis Ultra DLD system
with a base pressure of 10−9 Torr. The surface basicity of catalysts was mearsured via
temperature-programmed desorption of CO2 (CO2-TPD) via a Micromeritics
AutoChem II 2920 instrument. The catalyst was pretreated under a flow of helium gas
(30 mL · min−1) at 140 °C for 2 h. Subsequently, the catalyst was cooled to 50 °C under
helium gas. After adsorption of carbon dioxide, the catalyst was flushed off under a He
atmosphere at 50 °C. Finally, the TPD data were obtained from 50 °C-750 °C under
helium gas (heating rate of 20 °C · min−1). Electron paramagnetic resonance (EPR)
spectra were performed on a Bruker EMX spectrometer. The EPR experiments were
conducted with a center field of 3507.815 G and a frequency of 9.83 GHz using an
Elexsys probe head with 20 mg of catalyst in a 4 mm tube. The water contact angle of
samples was analyzed using a drop shape analysis system (Krüss, DSA100).
2.4 Procedures for THD of vanillin
The THD process of vanillin was performed in a glass tube reactor (25 mL) under
oil-heating conditions. Typically, vanillin (0.5 mmol), water (10 mL), and catalyst (50
mg) were added into the glass tube reactor. This reactor was then purged with nitrogen
gas and sealed. Afterward, the reactor was placed in a preheated oil bath pot. After the
THD reaction for a desired time under magnetically stirring, the reactor was rapidly
transferred into cold water. The liquid aqueous mixtures were collected and extracted
by ethyl acetate for analysis. We used GC-MS (Agilent 7890B-5977A) equipped with
HP-5MS capillary column (30.0 m × 250 mm × 0.25 mm) to identify the products in
the reaction mixture. We used GC (Shimadzu Nexis GC-2030) equipped with flame
ionization detector and HP-5 capillary column (30.0 m × 250 mm × 0.25 mm) to
quantitatively analyze the reactants and products on the basis of standard samples.
2.5 Catalyst recycling experiments

Page 7 of 30
Green Chemistry
View Article Online
After THD reaction, the catalysts were isolated from reaction products by aDmOI:a1g0.n10e3t9,/C9GC01720H
washed several times with distilled water, ethyl acetate, and ethanol and dried in a
vacuum drying oven at 50 °C. Furthermore, the recycled catalysts were kept for the
next cycle directly under the identical reaction conditions without further treatment.
2.6 Deuterium kinetic isotope effect experiments
For deuterium kinetic isotope effect experiments of THD of vanillin, we used
different deuterated formic acid (HCOOH, HCOOD, DCOOH and DCOOD) as H
source to demonstrate the transfer hydrodeoxygenation mechanism and the rate-
limiting step. After the THD reaction, the reaction products were extracted by ethyl
acetate and analyzed by GC (Shimadzu Nexis GC-2030) and GC-MS (Agilent 7890B-
5977A).
3. Results and Discussion
3.1 Characterization of catalysts
Scheme 1. The fabrication process of Co@NG.
The schematic illustration for the in-situ preparation of Co@NG catalysts is
depicted in Scheme 1. The utilization of CMC is particularly meaningful not only for
its renewable feature but also the intrinsic abundant carboxyl and hydroxyl groups for
further gelation by chelating with Co2+ ions and urea. After the gelling processes, CMC
chains can be aggregated into interconnected nanofibrils with uniform coordination of
Co2+ and urea. During the thermolysis process of dry precursors, CMC transformed into
carbons and urea decomposed and released NH3 and CO2. The resulting gases can
activate the carbon skeleton and trigger nitrogen doping. Co2+ ions can be reduced by
carbon into high dispersed Co NPs, which can serve as catalysts promoting the

Green Chemistry
Page 8 of 30
View Article Online
DOI: 10.1039/C9GC01720H
formation of N-enriched graphene shells. Gram-scale quantity of Co@NG can be
obtained by one run, indicating the feasibility for realizing bulk production of this
catalyst (Figure S1).
Figure 1. SEM images of unpyrolyzed Co@NG-6 precursor (a) (the insert is digital
photo). SEM image of Co@NG-6 (b). Typical TEM image (c, d), high-resolution TEM
image (e), particle-size distribution histogram (f), HAADF-STEM image (g), and
elemental mapping (g) of Co@NG-6.
Afterwards, the physical–chemical properties of Co@NG-U (U stands for the amount
of added urea) were characterized by several techniques to clarify the relationship
between the structure of as-prepared catalysts and their catalytic reactivity. The
morphologies and metal dispersions of the samples were analyzed by scanning electron
microscopy (SEM) and transmission electron microscopy (TEM). As can be seen from
Figure 1a, the flower-like unpyrolyzed Co@NG-6 precursor confirmed tightly bind of
three components (CMC, urea and Co2+ ions) after ionic gelation and chelation process.

Page 9 of 30
Green Chemistry
View Article Online
Without urea, or CMC, the precursor is a flat sheet or scattered rod shape (FigDuOrIe: 10S.120a39,/C9GC01720H
b). Moreover, NC-6 precursor also showed the shape of interlinked polymers,
indicating the chelation of CMC and urea (Figure S2c). After the pyrolysis process,
Co@NG-0 without urea exhibited an overall block structure with very few
interconnected macropores (Figure S3a). In contrary, Co@NG-6 and NC-6 all showed
typical three-dimensional hierarchical porous nanostructure with abundant
continuously cross-linked mesopores and macropores (Figure 1b, S3b). Other samples
(Mn@NG-6, Fe@NG-6, Ni@NG-6) all showed similar morphologies with a large
number of channels (Figure S3c-e). This result indicated that the gelation and chelation
of CMC, urea and Co2+ ions and presence of urea in preparing catalysts can promote
the formation of uniform porous structure by releasing NH3 and CO2, thus activating
the carbon skeleton.
TEM images in Figure 1c, f showed that the metallic Co NPs with a mean diameter
of 10.91 nm were well-distributed in carbon skeleton. Further high-resolution (HRTEM)
analysis indicated that Co NPs were well encapsulated inside 3-10 layers of graphene
shells with obvious interlayer distance of 0.341 nm. All Co NPs had a diameter of ~ 7
nm and displayed a clear lattice fringe with interplanar distance of 0.205 nm, which
could be assigned to (111) plane of β-Co (Figure 1d, e). Obviously, in this encapsulated
catalyst, the electron of Co core could penetrate through the self-catalysis-formed
graphene surface to boost the catalytic process on external graphene shell. Meanwhile,
the graphene layers acted as shells can prevent reaction substrates from contacting the
Co core and therefore can protect the Co core from corrosion and aggregation during
the following harsh acidic THD reaction conditions.47 The HRTEM-EDS (energy-
dispersive X-ray spectrometry) in Figure S4 confirmed sample was consisted of four
components, that is Co, N, C, and O. The EDS elemental mapping from representative
high-angle annular dark-field scanning TEM (HAADF-STEM) validated uniform
distribution of Co, N and C in Co@NG-6 (Figure 1g, h).

Green Chemistry
Page 10 of 30
Figure 2. N2 adsorption-desorption isotherms (a), corresponding pore size distributionVsiew Article Online
DOI: 10.1039/C9GC01720H
(b) and specific surface area (c) of Co@NG-U.
The porous properties of Co@NG-U were examined by nitrogen
adsorption/desorption analysis. The isotherm curves can be assigned to typical type IV
patterns with a characteristic hysteresis loop at high relative pressures, demonstrating
the existence of abundant mesopores in all Co@ NG-U samples (Figure 2a). The BJH
pore size distribution calculated from the adsorption branch of the isotherms (Figure
2b) showed that all the samples exhibited intensive pore size distribution centered at
∼2–100 nm, confirming the hierarchical porous structure. Notably, as the elevated
amount urea from 0 g to 6.5 g, the specific surface area increased remarkably from
177.16 to 626.66 m2 g−1 and the pore size changed from macropore to mesopore (Figure
2c). However, the continuous increase of urea led to the obvious decline of specific
surface area and pore volume. This is because CMC transformed into amorphous
carbons and urea decomposed and released NH3 and CO2 in the pyrolysis process. The
resulting gases can etch the carbon skeleton, thus promoting the formation of porous
structure and high specific surface area, simultaneously triggering nitrogen doping.
Further increasing the amount of urea to a high level, however, the structure of carbon
skeleton was broken and collapsed, resulting in the decrease of specific surface area.
Figure 3. XRD patterns (a), Raman spectra (b), ID/IG and I2D/IG ratios (c) of Co@NG-
U.
The XRD patterns in Figure 3a showed an obvious diffraction peak at 26.1°, which
was assigned to (002) reflection of crystallized carbon. Besides, the characteristic
diffraction peaks at 44.3° and 51.4° could be ascribed to (111) and (220) reflections of
detectable cubic cobalt, respectively. Raman spectroscopy was performed to get more
insights into the graphitic degree and defects of Co@NG. Three pronounced peaks at
1350, 1580 and 2750 cm−1 in Figure 3b could be ascribed to D, G and 2D bands of
carbon materials, respectively. From the comparison of ID/IG ratios and I2D/IG, it can be

Page 11 of 30
Green Chemistry
View Article Online
seen that the ID/IG ratios value increased and I2D/IG decreased when the amountDoOfI:a1d0.d10e3d9/C9GC01720H
urea increased from 0 g to 10 g, indicating the increase of defective structure in carbon
materials (Figure 3c).
Figure 4. XPS spectrum of survey scan of Co@NG-U samples (a); High-resolution Co
2p (b), N 1s spectrum (c), C 1s spectrum (d) of Co@NG-6; Configurations of N dopants
(e) and CO2-TPD (f) of Co@NG-U.
X-ray photoelectron spectroscopy (XPS) investigations indicated that Co@NG-U
consisted of Co, N, C and O (Figure 4a). As manifested in Figure 4b and S5, the Co
2p spectrum of Co@NG could corroborated the coexistence of three cobalt species,
metallic Co (778.6 eV), Co(III)–N/O (780.5 eV) and Co(II)–N/O (782.5 eV). The
existence of Co–O/Co–N confirmed the strong interaction between Co species and N-
doped graphene shell via Co–O–C or Co–N–C bonds, which resulting in lower electron
density in cobalt site. The percentage of Co0 increased from 17.3% to 35.0% with an
increased amount of urea from 0 g to 6.5 g but then decreased to 27.6% at 10 g. This
indicated that the generated NH3 from urea can reduce Co ions into Co0. The Co0
species are the active sites in transfer hydrogenation reaction.18, 25 However, the excess
NH3 can also promote the formation of Co(III)–N species. Furthermore, the N 1s
spectrum of Co@NG suggested that there existed four N species, pyridinic N (398.6
eV), pyrrolic N (400.6 eV), graphitic N (401.2 eV) and pyridine oxide N (403.9 eV)
(Figure 4c, S6). Co@NG-6 possessed highest content of pyridinic N sites and

Green Chemistry
Page 12 of 30
Co@NG-8 had the maximum nitrogen content (Figure 4e, Table S2). TDhOeI: 1C0.1013Vs9i/eCw9AGrtCic0le1O72nl0inHe
spectrum of Co@NG could be deconvoluted into three peaks located at 284.4, 285.6,
and 288.6 eV, which could be indexed to C-C, C-N and C-O bonds, respectively
(Figure 4d). The presence of C-N bond verified the successful introduction of nitrogen
in carbon skeleton. In addition, we further carried out CO2-TPD to mearure the surface
basicity of catalysts. As can be seen in Figure 4f, Co@NG-6 exhibited the highest
strength and density of basic site in all samples based on the temperature and peak area
of carbon dioxide-desorption peaks. All these Co@NG-U samples had a similar Co
content of around 1 wt% on the basis of inductively coupled plasma optical emission
spectrometry (ICP-OES) results in Table S1. The nitrogen content of Co@NG-U
catalysts was determined to be between 0.2 and 7.04 at% from Co@NG-0 to Co@NC-8,
respectively, based on XPS results (Table S1). Further elevating the amount of urea in
preparing the catalyst did not lead to a higher nitrogen content of Co@NG, but resulted
in a lower yield of Co@NG catalysts due to the over etching of carbon skeleton by
generated gases from urea.
Figure 5. Photographs of a water droplet with a contact angle of Co@NG-0 (a),
Co@NG-3 (b), Co@NG-6 (c), Co@NG-8 (d).
The wettability of prepared Co@NG catalysts was studied using water contact angle
measurement method and the results were present in Figure 5. Co@NG-0 and
Co@NG-3 without N-doping or low N-doping displayed hydrophobicity with a high
contact angle of about 147.1° and 134.2°, respectively. In contrary, Co@NG-6 and
Co@NG-8 all displayed superhydrophilicity with low contact angle of 51.9°and 20.8°.

Page 13 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
In addition, a digital photo in Figure S7 showed that Co@NG-6 can be well-dispersed
in water due to the doped N increased the hydrophilic property, which might enhance
the exposure of catalyst to vanillin, thereby significantly increasing the catalytic
performance.
3.2 Evaluation of the catalytic performance of the various catalysts
Table 1. Catalytic results for different catalysts.a
MPC
formation rate TOF
FA eff.
(%) d
Entry
Catalysts
(mmolꢀg−1ꢀh−1) (h-1) c
Conversion MPC selectivity MPC yield
b
(%)
(%)
(%)
1
2
Blank
Co Powder
Co(OAc)2
C Powder
0
0
0
0.0
59.5
-
0
3.2
93
3.0
0.28
0.18
-
0.5
0.4
0.0
1.7
0.1
4.3
4.6
10.5
18.1
10.2
7.3
0.6
2.5
0.3
9.0
8.9
6.8
2.1
3
2.1
92
1.9
38.6
4
0
0
0.0
0.0
5
NC-6
9.3
100
99
9.3
186.0
7.9
-
6
Co@NG-0
Co@NG-3
Co@NG-5
Co@NG-6
Co@NG-6
Co@NG-8
Co@NG-10
Mn@NG-6
Fe@NG-6
Ni@NG-6
Co@NG-HEC
Co@NG-HPC
Co@NG-MC
0.4
0.4
0.03
2.03
2.18
4.97
3.57
4.81
3.43
0.30
1.18
0.13
4.26
4.19
3.19
1.00
7
23.3
25.1
57.1
98.5
55.3
39.4
3.4
100
100
100
100
100
100
99
23.3
25.1
57.1
98.5
55.3
39.4
3.4
466.0
502.0
1142.0
820.8
1106.0
788.0
67.3
8
9
10 e
11
12
13
14
15
16
17
18
19
13.6
1.5
100
100
100
100
100
99
13.6
1.5
272.0
30.0
49.0
48.2
36.7
11.5
49.0
48.2
36.7
11.4
980.0
964.0
734.0
227.7
Co@NG-MCC
a
Reaction conditions: 0.5mmol vanillin in 10 mL water, 250 mg FA, catalyst (2.3
b
mol% metal), 160 °C, 5 h. MPC formation rate is defined as (mol of formed
MPC)/(catalyst amountꢀ×ꢀtime). c TOF is defined as mol (converted vanillin)/[mol (total

Green Chemistry
Page 14 of 30
View Article Online
DOI: 10.1039/C9GC01720H
metal added)ꢀ×ꢀh (time)]. d The efficiencies of FA (FA eff.) were calculated taking into
e
account that 1 mol of FA is needed to produced 0.5 mol of MPC. 6 h, catalyst (4.6
mol%).
Subsequently, vanillin, a representative model compound of liquid bio-oil from
biomass lignin was used as a substrate to investigate the THD performance of various
catalysts with FA as H source. The blank experiment revealed that the THD reaction
did not take place without catalyst (Table 1, Entry 1). Then, different cobalt-containing
compounds (Co Powder and Co(OAc)2) were tested. All these Co catalysts failed to
catalyze this THD reaction (Table 1, Entries 2-3). Compared to the carbon powder
catalyst without nitrogen doping, NC-6 acquired a slight conversion of vanillin (9.3%),
suggesting that nitrogen species were favorable for promoting the transfer reduction of
vanillin (Table 1, Entries 4-5). The importance of nitrogen species in improving the
activity will be discussed in the part of Mechanism study. Gratifyingly, the as-
fabricated Co@NC-6 gave the highest conversion (57.1%) with 100% selectivity to 2-
methoxy-p-cresol (MPC) with the highest formation rate (1142ꢀmmolꢀg−1ꢀh−1) and TOF
(4.97ꢀh−1) at 160ꢀ°C for 5ꢀh among all tested samples (Table 1, Entry 9). Moreover,
when the metal center Co of Co@NG-6 was replaced with other metals such as Mn, Fe
and Ni, the conversion rate for THD of vanillin with FA decreased distinctly (Table 1,
Entries 13–15). It not only confirmed that Co NPs were catalytic active center in THD,
but also revealed that the nature of metallic catalytic center was critical for this reaction.
After optimizing the reaction conditions, the THD of vanillin can obtain a near
quantitative yield (98.5%) with the highest FA efficiency (18.1%) (Table 1, Entry 10).
Moreover, the conversion of vanillin was remarkably increased from 0.4% to 57.1% as
the amount of urea added in preparation process being increased from 0 to 6.5 g.
However, further increasing the initial loading amounts of urea gradually decreased the
vanillin conversion (Table 1, Entries 6–12). There were two main reasons accounting
for the best performance of Co@NG-6 combined with the previous characterization:
(1) The calculated BET surface area of Co@NG-6 was the highest (626.66 m2 g−1)
along with the maximum pore volume and highest mesoporous proportion, which
would be favorable for exposing more catalytic active sites and improving the
adsorption and transformations of reaction substrate. (2) Co@NG-6 had the most and
strongest basic sites (CO2-TPD from Figure 4f) along with the highest proportion of
pyridinic N (XPS from Figure 4e). The doped-nitrogen (especially electron-rich
pyridinic N) can promote the electrons transfer from N-doped carbon/graphene to

Page 15 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
transition metal, reduce the energy barrier for reactant adsorption, thus boosting the
catalytic reaction.48-49 (3) It is noteworthy that Co@NG-8 also exhibited excellent
catalytic performance although a little worse than Co@NG-6. The possible reasons are
that Co@NG-6 also possess high content of N and defective structure along with high
superhydrophilicity. This result revealed that urea loading played a crucial role on
modulating the physical microstructure and chemical properties of final pyrolytic
product, thus remarkably influencing the THD reaction.
3.3 Effect of different cellulose derivatives on catalysts
In order to reveal the importance of dissolution, chelation and gelation process of
CMC, other cellulose derivatives including insoluble microcrystalline cellulose (MCC),
soluble hydroxypropyl cellulose (HPC), hydroxyethyl cellulose (HEC) and methyl
cellulose (MC) were respectively used as carbon source (Table 1, Entries 16-19). The
as-prepared Co@NG-MCC gave a slightly improved vanillin conversion, indicating the
complete dissolution of biomass macromolecule was important for the highly
dispersion of each components. Co@NC-HEC and Co@NC-HPC with hydroxyethyl
and hydroxypropyl groups in HEC and HPC exhibited comparable vanillin conversion
(48~49%), formation rate (960~980ꢀmmolꢀg−1ꢀh−1) and TOF (4.1~4.3ꢀh−1), while
Co@NC-MC only gave moderate activity (36.7% conversion) compared to Co@NG-
6. These results indicated that the dissolution, dispersion, chelation (carboxy, or
hydroxyl with urea and Co2+) and gelation process of biomass precursors played an
important role in homogeneous coordination and dispersion of active sites.
3.4 Effect of solvents and reaction conditions
Owing to that the solvents play an important role in determining reaction rate and
product selectivity, we tried various solvents in THD of vanillin. The corresponding
conversion, selectivity and polarity of solvents are depicted in Figure 6a. We found
that strongly polar water and N, N-dimethylformamide (DMF) as solvent provides
better conversion than low polar solvents, such as tetrahydrofuran (THF), 1,4-dioxane,
and ethyl acetate. This law was consistent with the reported study that high polar
solvents benefit from the hydrogenation of less polar substrates.35 On the other hand, a
digital photo in Figure S7 showed that Co@NG-6 can be well-dispersed in water due
to the doped N increased the hydrophilic property (confirmed by Contact angle
measurement in Figure 5). The good dispersion capability in solvent water might

Green Chemistry
Page 16 of 30
View Article Online
enhance the exposure of catalyst to vanillin, thereby significantly increaDsOinI:g10.1t0h3e9/C9GC01720H
catalytic performance. Besides, water is an ideal green solvent that is environmentally
friendly and low-cost when compared with DMF.
Figure 6. Optimization of THD reaction parameters. Reaction conditions: (a) 0.5 mmol
vanillin in 10 mL solvent, 250 mg FA, Co@NG-6 catalyst (2.3 mol% metal), 160 °C,
5 h; (b) 0.5 mmol vanillin in 10 mL water, 250 mg FA, Co@NG-6 catalyst (2.3 mol%
metal), 5h; (c) 0.5mmol vanillin in 10 mL water, Co@NG-6 catalyst (2.3 mol% metal),
160 °C, 5 h; (d) TOF values for conversion of vanillin with amount of FA. (e) 0.5mmol
vanillin in 10 mL water, Co@NG-6, 250 mg FA, 160 °C, 5 h. (f) GC spectrum of the

Page 17 of 30
Green Chemistry
View Article Online
conversion of vanillin with varied reaction time. 0.5mmol vanillin in 10 mL wDaOteI:r1,02.105309/C9GC01720H
mg FA, 160 °C, Co@NG-6 catalyst (4.6 mol% metal).
The reaction temperature significantly affected THD of vanillin (Figure 6b). The
conversion over Co@NG-6 increased sharply from 0 to 57.1% when elevating the
reaction temperature from 100 to 160 °C. Due to the use of glass reactor, the reaction
temperature was not raised further. As the FA amount increased from 0 to 250 mg
(Figure 6c), the vanillin conversion increases remarkably. Further increasing the FA
amount to 350 mg, however, resulted in an obvious decrease of conversion to 37%,
indicating the high concentration of FA can aggravate the etching, leaching and
inactivation of Co NPs. The plot of reaction rate (TOF value) versus the concentration
of FA is depicted in Figure 6d. The THD reaction rate followed a linear relationship
versus the concentration of FA over the range 25−250 mg. The first-order dependence
of the H source suggested a reaction mechanism in which the proton and/or hydride
transfer is involved in, or precede, the rate-determining step. Further insights about the
transfer hydrodeoxygenation mechanism over Co@NG-6 will be discussed in the part
of Reaction mechanism studies.
Catalyst loading and reaction time both has a large impact on the catalytic activity
(Figure 6e and S8), and an increasing catalyst usage or reaction time led to the
complete conversion of vanillin. Moreover, it is worth mentioning that no vanillyl
alcohol was formed and MPC selectivity remains >99.0% in all reaction (Figure 6f).
This result suggested a different reduction mechanism in comparison to other report
using H2 as H source, in which vanillyl alcohol is firstly generated and then
hydrogenated to MPC.50 The possible reasons will be discussed in the part of Reaction
mechanism. Compared with the recent other reported noble metal-based catalysts
using high pressure H2 as H source (Table S4), our developed catalytic system with
liquid H donor FA can also produce nearly quantitative MPC yield (99%) under mild
conditions. Compared with the recent reported Co-based catalyst or the THD reaction
using alcohols as H donors, the Co@NG-6 can also provide high conversion at lower
reaction temperatures with less amount of catalyst (4.6 mol% Co).
3.6 Extension of the catalyst system.
The catalytic transfer hydrogenation of a series of unsaturated compounds was
investigated further. For vanillyl alcohol, the reaction rate is faster than that of vanillin
and only 2.5 h was needed to complete the transformation (Table 2, Entries 2).

Green Chemistry
Page 18 of 30
Furthermore, Co@NG-6 was also broadly applicable for transfer reductionDoOfI: 1o0t.1h0e3V9ri/eCw9AGrtCic0le1O72nl0inHe
lignocellulose-derived oxygenates. For instance, the conversions of furfural,
benzaldehyde were 100% and selectivities to the target alcohols were more than 90%
(Table 2, Entries 3-4). Additionally, the developed Co@NG-6 also exhibited also
superior catalytic performance for cinnamaldehyde resulting in more than 98%
conversion with 96% selectivity to cinnamyl alcohol (Table 2, Entries 5). These
reduced alcohols are significant reagents for preparing pharmaceuticals, flavors and
cosmetics in fine chemical industry.51-52 Besides the THD of bioderived platform
molecules, other unsaturated chemical compounds from petrochemical industry were
also tested as substrates to evaluate the catalytic reactivity of Co@NG-6 (Table 2,
Entries 6-9). For quinoline, the catalyst also exhibited outstanding catalytic activity
and the yield of 1,2,3,4-tetrahydroquinoline are more than 96%. Because of the high
value of aniline,53 catalytic transfer hydrogenation of nitrobenzene was carried out and
90.8% yield of aniline was achieved within 6 h. Therefore, we concluded that Co@NG-
6 was a highly versatile catalyst for catalytic THD of both bioderived chemicals and
organic compounds from petrochemical industry with FA as H donor.
Table 2. Results of Co@NG-6-catalyzed transfer hydrogenation of various unsaturated
chemical compounds. a
Entry
1
Substrates
Products
t (h) Conv. (%) Selec. (%)
HO
CHO
HO
CH3
6
98.5
100
O
O
HO
HO
CH2OH
CH3
2
3
2.5
6
99.2
100
100
O
O
O
O
90.3
OH
O
O
OH
4
5
6
100
98
97.2
96
OH
OH
O
10
O
6
7
6
6
98.3
95.4
99
99
OH
O
N
H
N
8
9
6
6
100
95.6
99
NO2
NH2
90.8

Page 19 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
a Reaction conditions: 0.5 mmol substrates in 10 ml water, 250 mg FA, Co@NG-6
catalyst (4.6 mol% Co), 160 °C.
3.7 Heterogeneity and recyclability of Co@NG-6
To prove Co@NG-6 was heterogeneous, we separated the catalyst from reaction
solution after 2 h at 160 °C. Afterwards, the THD reaction was stirred for another 4 h
without Co@NG-6. The vanillin conversion still remains constant (Figure 7a),
suggesting the heterogeneous character of Co@NG-6. Furthermore, the recycling
experiment of Co@NG-6 for THD of vanillin was studied in aqueous medium at 160 °C
for 6 h. After reaction, the Co@NG-6 was easily separated by a strong magnet (Figure
7b). As shown in Figure 7c, conversion of vanillin was only slightly decreased from
98.5% to 91% in seven reaction runs. The selectivity of MPC remained unchanged in
all cycles. After catalytic reactions, diffraction peaks of the Co phase and characteristic
peaks of carbon materials were almost identical to the fresh one, as confirmed by XRD
and Raman spectra (Figure 7d, e). Co 2p, N 1s, C 1s XPS results of recycled Co@NG-6
(Figure 7f, S9) were nearly identical to the fresh Co@NG-6 catalyst, which indicated
that the chemical state of Co, N, C species didn’t change remarkably. However, the Co
content reduced slightly after seven cycles (Table S3), suggesting the leaching of little
Co sites during seven consecutive recycling process contributed to slightly decrease of
catalytic reactivity.

Green Chemistry
Page 20 of 30
View Article Online
DOI: 10.1039/C9GC01720H
Figure 7. Heterogeneity (a) of Co@NG-6; Magnetization curve (b) of Co@NG-6 and
the inset shows the magnetic separation of Co@NG-6 by using a magnet; Reusability
(c) of Co@NG-6; XRD patterns (d), Raman spectra (e), Co 2p spectrum (f) of Co@NG-
6 before and after reaction.
3.8 Mechanism study
There are few reports on the mechanism study of THD of biomass molecules with
FA. In general, basic compounds were needed as to assist the transfer hydrogenation
with FA.54 However, we didn’t add any base in THD of vanillin. In our previous
experiments (Table 1, entry 4-6 and 10), we demonstrated that the doped-N in
catalysts were of vital importance for high catalytic performance. We speculated that
the N atoms in Co@NG-6 should serve as a Lewis base. To prove this, 250 mg H3PO4
as base poisoning reagent were added into the reaction mixture. The conversion of
vanillin decreased significantly (Table 3, Entry 1). The strong acidic H3PO4 could
react with basic N atoms to form NH+, thus poisoned the Co@NG-6. In contrast, the
addition of triethylamine as additional basic sites remarkably improved the catalytic
performance of Co@NG-6 (Table 3, Entry 2). We speculated that the cooperative role
between the basic nitrogen-doped carbon/graphene and cobalt NPs via electron transfer
at metal-N-graphene interface could beneficially modulates the conversion rate of THD.
In fact, it was also reported in recent years that metal@C−N catalysts can serve as a
Mott−Schottky type nanocatalyst with electron-deficient metal and electron-rich basic
nitrogen-doped carbon/graphene.55-56 The doping of N atoms into carbon skeleton can
adjust the position of valence band or conducting band of carbon phase. The flat band
potential of N-doped carbon was higher than that of metal. As a result, N-enriched
carbon can accept electrons from metal until their Fermi level were balanced. The
electron at interface of metal and N-doped carbon can be redistributed, consequently
resulting in enriched positive charges on metal. The interface Mott−Schottky effect in
metal-nitrogen-carbon heterojunction can promote electron transfer to an electron-rich
surface and, thus, leading to an enhanced Lewis basicity of the carbon supports
(Scheme 2a).57-58
In our experiment, the presence of Lewis basic sites was well verified by the CO2-
TPD analysis (Figure 4f). Furthermore, electron paramagnetic resonance spectra (EPR)
spectra (Figure 8) demonstrated that there was a strong interaction between the
nitogen-doped carbon/graphene and metallic Co, which was favorable for high catalytic
performance by enhancing the electron transfer between metallic catalytic center and

Page 21 of 30
Green Chemistry
basic N-doped support. These results further confirmed the cooperative role of LewiVsiew Article Online
DOI: 10.1039/C9GC01720H
basic N and encapsulated Co NPs at Mott−Schottky interface in the THD reaction.
Combined with the above analysis and catalyst characterization, we proposed a feasible
reaction mechanism (Scheme 2b).
Figure 8. EPR spectra of NC-6 and Co@NG-6.
Firstly, electronegative N atoms served as Lewis basic sites in Co@NG-6 captured
the H+ from FA to form NH+. Simultaneously, electron negative formate intermediates
can coordinate with electron-deficient Co NPs via Co0 empty d orbitals. As a result,
Co-formate intermediates were generated. In recent years, some research workers also
considered that metal-formate were intermediates in the dehydrogenation reaction of
FA into H2, when the catalysts were doped with nitrogen or basic compounds was used.
59-61 Afterwards, Co-hydride species (Co-H−) were generated after releasing of CO2.
The formation of NH+ and Co-H− resembled the heterolytic cleavage of molecular
hydrogen, in which doped-N or added basic compounds could promote the heterolytic
cleavage. For instance, Zhang and co-workers found the doped N in Pd/N-C acted as
base could enhance the heterolytic cleavage of molecular hydrogen to form protic N-
H+ and hydridic Pd-H- species.62 Therefore, the formation of NH+ and Co-H− was
reasonable based on previous discuss and designed experimental results and these
intermediates were catalytic active sites for THD of C=O bonds in vanillin. Recently,
protic H+ and hydridic H− were also generally proposed as active species for
hydrogenation of polar C=O or C=N bonds with H2.18, 63-64 Besides, the deuterium
kinetic isotope effect (KIE) experiments (Table 3, Entry 3-6) revealed that a large KIE
of 2.79 and 3.41 was observed using DCOOH and DCOOD for THD. But the KIE was

Green Chemistry
Page 22 of 30
View Article Online
only 1.08 when HCOOD was used as H doonor. This result corroboratedDtOhI:a1t0.1t0h3e9/C9GC01720H
generation of Co-H− via β-hydride elimination and protonation was the rate-limiting
step for THD. 18, 60
Table 3 Effect of additives or D-substituted FA on THD of vanillin with over the
Co@NG-6 catalyst.a
Reaction rate
Entry
Formic acid
Conv. (%)
KIE d
(μmol min−1 g catalyst−1)
1b
2c
3
HCOOH
HCOOH
HCOOH
HCOOD
DCOOH
DCOOD
18.05
87.05
51.04
47.24
18.33
14.99
5.01
24.18
14.18
13.12
5.09
-
-
-
4
1.08
2.79
3.41
5
6
4.16
a
Reaction conditions: 0.5 mmol vanillin in 10 mL water, 250 mg FA, Co@NG-6
catalyst (4.6 mol% metal), 160 °C, 3 h.
b Basic sites of Co@NG-6 were poisoned by adding 300 mg phosphoric acid.
c 120 mg triethylamine was added as additional basic sites.
d KIE = rate (entry 3) / rate (entry n); n = 4–6.
Moreover, vanillyl alcohol is firstly generated and then hydrogenated to MPC when
H2 was used as H source. However, vanillyl alcohol cannot be detected in all reactions
and MPC was the sole products (Figure 6). This result suggested a different reduction
mechanism in comparison to other report using H2 as H source.46 Indeed, H2 can form
in our catalytic system via decomposition of FA because formic acid is excessive. It is
reported that the generation of H2 from active H requires more energy to overcome the
overpotential of the chemisorbed H at the metal surface.65-66 In contrary, the transfer of
active H to reduce vanillin over Co@NG-6 is much faster and easier than to form H2
gas. Therefore, the formation and activation of H2 gas was not essential for the THD of
vanillin although we can detect a small amount of hydrogen from tail gas in the reactor.
As a result, the reduction of vanillin proceeded by consuming four active hydrogen
bonded on the surface of the Co NPs in one step. Notably, the steric hindrance caused
by the encapsulation of ultrafine Co NPs in N-enriched graphene shells and strong
interactions of substrate and intermediate product (vanillyl alcohol) could result in the
blocking of reaction pathway toward vanillyl alcohol by steric constraints (Scheme
2a).67 Besides, we found that the conversion of vanillyl alcohol was much faster than
that of vanillin (Table 2, Entry 2). Although vanillyl alcohol can be formed in the

Page 23 of 30
Green Chemistry
system, it can be hydrogenated into MPC immediately. Thus, vanillyl alcohol aVsiew Article Online
DOI: 10.1039/C9GC01720H
byproducts could not be detected in all reaction. On the whole, all the catalyst
characterization, experimental results, poisoning tests and KIE experiments confirmed
that the proposed mechanism was reasonable.
Scheme 2 Possible reaction pathways (a) and mechanism (b) for the Co@NG-6-
catalyzed THD of vanillin with FA.
4. Conclusions
In conclusion, we have explored a scale-up biomass-annealing approach to
fabricate Co NPs encapsulated in N-doped graphene shells (Co@NG) as highly

Green Chemistry
Page 24 of 30
View Article Online
DOI: 10.1039/C9GC01720H
efficient catalysts toward base-free THD of lignin model compound vanillin with
biomass-derived FA. The ionic gelation and chelation of CMC, urea and Co2+
contribute to uniformly dispersed Co NPs anchored on high surface area carbon
skeleton, enabling the full exposure of catalytic active sites. The protected graphene-
sheathed Co NPs improved the stability. Mechanism studies demonstrated that the
synergistic effect between Co NPs and incorporated N via can promote the generation
of Co-formate intermediates, giving impetus to form active protic H+ and hydridic H−
for hydrogenating C=O bonds. Given that all the reaction substrate, hydrogen source
and catalysts in this reaction system are based on renewable biomass, we believe that
our findings can pave a sustainable and promising avenue for both developing stable
chainmail catalysts and efficiently upgrading of biofuel.
Acknowledgements
The authors thank the financial support for this work by the National Natural Science
Foundation of China (21774036), Guangdong Province Science Foundation
(2017B090903003, 2017GC010429).
Reference
[1] J. S. Luterbacher, J. M. Rand, D. M. Alonso, J. Han, J. T. Youngquist, C. T.
Maravelias, B. F. Pfleger, J. A. Dumesic, Nonenzymatic sugar production from biomass
using biomass-derived γ-valerolactone, Science, 2014, 343, 277-280.
[2] A. J. Ragauskas, G. T. Beckham, M. J. Biddy, R. Chandra, F. Chen, M. F. Davis,
B. H. Davison, R. A. Dixon, P. Gilna, M. Keller, Lignin valorization: improving lignin
processing in the biorefinery, Science, 2014, 344, 1246843.
[3] R. Nie, H. Yang, H. Zhang, X. Yu, X. Lu, D. Zhou, Q. Xia, Mild-temperature
hydrodeoxygenation of vanillin over porous nitrogen-doped carbon black supported
nickel nanoparticles, Green Chem., 2019, 21, 314-320
[4]S. Song, J. Zhang, G. Gözaydın, N. Yan, Production of Terephthalic Acid from Corn
Stover Lignin, Angew. Chem., Int. Ed., 2019, 58, 4988-4991.
[5] S. Y. Lee, H. U. Kim, T. U. Chae, J. S. Cho, J. W. Kim, J. H. Shin, D. I. Kim, Y. S.
Ko, W. D. Jang, Y. S. Jang, A comprehensive metabolic map for production of bio-
based chemicals, Nat. Catal., 2019, 2, 18-33.
[6] R. Fan, C. Chen, M. Han, W. Gong, H. Zhang, Y. Zhang, H. Zhao, G. Wang, Highly
Dispersed Copper Nanoparticles Supported on Activated Carbon as an Efficient
Catalyst for Selective Reduction of Vanillin, Small, 2018, 14, 1801953.
[7] G.H. Wang, J. Hilgert, F. H. Richter, F. Wang, H.J. Bongard, B. Spliethoff, C.

Page 25 of 30
Green Chemistry
View Article Online
Weidenthaler, F. Schüth, Nat. Mater. Platinum–cobalt bimetallic nanoparDtiOcI:le10s.10i3n9/C9GC01720H
hollow carbon nanospheres for hydrogenolysis of 5-hydroxymethylfurfural, Nat.
Mater., 2014, 13, 293-300.
[8] Y. T. Cheng, J. Jae, J. Shi, W. Fan, G. W. Huber, Production of renewable aromatic
compounds by catalytic fast pyrolysis of lignocellulosic biomass with bifunctional
Ga/ZSM-5 catalysts, Angew. Chem., Int. Ed., 2012, 51, 1387-1390.
[9] L. Chen, C. Fink, Z. Fei, P. J. Dyson, G. Laurenczy, An efficient Pt nanoparticle–
ionic liquid system for the hydrodeoxygenation of bio-derived phenols under mild
conditions, Green Chem., 2017, 19, 5435-5441.
[10] G. H. Wang, Z. Cao, D. Gu, N. Pfänder, A. C. Swertz, B. Spliethoff, H. J. Bongard,
C. Weidenthaler, W. Schmidt, R. Rinaldi, Nitrogen-Doped Ordered Mesoporous
Carbon Supported Bimetallic PtCo Nanoparticles for Upgrading of Biophenolics,
Angew. Chem., Int. Ed., 2016, 55, 8850-8855.
[11] K. A. Goulas, A. V. Mironenko, G. R. Jenness, T. Mazal, D. G. Vlachos,
Fundamentals of C–O bond activation on metal oxide catalysts, Nat. Catal., 2019, 2,
269-276.
[12] J. Zhang, L. D. Ellis, B. Wang, M. J. Dzara, C. Sievers, S. Pylypenko, E. Nikolla,
J. W. Medlin, Control of interfacial acid–metal catalysis with organic monolayers, Nat.
Catal., 2018, 1, 148-155.
[13] D. Wang, D. Astruc, The golden age of transfer hydrogenation, Chem. Rev., 2015,
115, 6621-6686.
[14] S. Zhou, F. Dai, Y. Chen, C. Dang, H. Qi, Sustainable hydrothermal self-assembly
of hafnium–lignosulfonate nanohybrids for highly efficient reductive upgrading of 5-
hydroxymethylfurfural, Green Chem., 2019, 21, 1421-1431.
[15] G. H. Wang, X. Deng, D. Gu, K. Chen, H. Tüysüz, B. Spliethoff, H. J. Bongard,
C. Weidenthaler, W. Schmidt, F. Schüth, Co3O4 Nanoparticles Supported on
Mesoporous Carbon for Selective Transfer Hydrogenation of α,β-Unsaturated
Aldehydes, Angew. Chem., Int. Ed., 2016, 55, 11101-11105.
[16] S. Zhou, F. Dai, Z. Xiang, T. Song, D. Liu, F. Lu, H. Qi, Zirconium–lignosulfonate
polyphenolic polymer for highly efficient hydrogen transfer of biomass-derived
oxygenates under mild conditions, Appl. Catal. B: Environ., 2019, 248, 31-43.
[17] D. Mellmann, P. Sponholz, H. Junge, M. Beller, Formic acid as a hydrogen storage
material–development of homogeneous catalysts for selective hydrogen release, Chem.
Soc. Rev., 2016, 45, 3954-3988.

Green Chemistry
Page 26 of 30
View Article Online
[18] P. Zhou, Z. Zhang, L. Jiang, C. Yu, K. Lv, J. Sun, S. Wang, A versatiDleOI:c1o0b.10a3l9t/C9GC01720H
catalyst for the reductive amination of carbonyl compounds with nitro compounds by
transfer hydrogenation, Appl. Catal. B: Environ., 2017, 210, 522-532.
[19] H. H. Khoo, W. L. Ee, V. Isoni, Bio-chemicals from lignocellulose feedstock:
sustainability, LCA and the green conundrum, Green Chem. 2016, 18, 1912-1922.
[20] L. Xiang, S. Li, Y. Liu, C. Yong, Formic acid: A versatile renewable reagent for
green and sustainable chemical synthesis, Chin. J. Catal. 2015, 36, 1461-1475.
[21] K. Mori, T. Sano, H. Kobayashi, H. Yamashita, Surface Engineering of a
Supported PdAg Catalyst for Hydrogenation of CO2 to Formic Acid: Elucidating the
Active Pd Atoms in Alloy Nanoparticles, J. Am. Chem. Soc., 2018, 140, 8902-8909.
[22] L. Liu, J. Li, Y. Ai, Y. Liu, J. Xiong, H. Wang, Y. Qiao, W. Liu, S. Tan, S. Feng,
A ppm level Rh-based composite as an ecofriendly catalyst for transfer hydrogenation
of nitriles: triple guarantee of selectivity for primary amines, Green Chem. 2019, 21,
1390-1395.
[23] G. Wienhöfer, I. Sorribes, A. Boddien, F. Westerhaus, K. Junge, H. Junge, R.
Llusar, M. Beller, General and selective iron-catalyzed transfer hydrogenation of
nitroarenes without base, J. Am. Chem. Soc., 2011,133, 12875-12879.
[24] I. Sorribes, G. Wienhöfer, C. Vicent, K. Junge, R. Llusar, M. Beller,
Chemoselective Transfer Hydrogenation to Nitroarenes Mediated by Cubane-Type
Mo3S4 Cluster Catalysts, Angew. Chem., Int. Ed., 2012, 51, 7794-7798.
[25] G. Li, H. Yang, H. Zhang, Z. Qi, M. Chen, W. Hu, L. Tian, R. Nie, W. Huang,
Encapsulation of Nonprecious Metal into Ordered Mesoporous N-Doped Carbon for
Efficient Quinoline Transfer Hydrogenation with Formic Acid, ACS Catal., 2018, 8,
8396-8405.
[26] J. Deng, P. Ren, D. Deng, X. Bao, Enhanced electron penetration through an
ultrathin graphene layer for highly efficient catalysis of the hydrogen evolution reaction,
Angew. Chem. Int. Ed., 2015, 54, 2100-2104.
[27] J. Shi, M. Zhao, Y. Wang, J. Fu, X. Lu, Z. Hou, Upgrading of aromatic compounds
in bio-oil over ultrathin graphene encapsulated Ru nanoparticles, J. Mater. Chem. A,
2016, 4, 5842-5848.
[28] A. Kasprzak, M. Bystrzejewski, M. Koszytkowska-Stawinska, M. Poplawska,
Grinding-induced functionalization of carbon-encapsulated iron nanoparticles, Green
Chem., 2017, 19, 3510-3514.
[29] X. Cui, P. Ren, D. Deng, J. Deng, X. Bao, Single layer graphene encapsulating

Page 27 of 30
Green Chemistry
View Article Online
non-precious metals as high-performance electrocatalysts for water oxidationD,OEI:n10e.r10g3y9/C9GC01720H
Environ. Sci., 2016, 9, 123-129.
[30] Y. Xu, W. Tu, B. Zhang, S. Yin, Y. Huang, M. Kraft, R. Xu, Nickel Nanoparticles
Encapsulated in Few-Layer Nitrogen-Doped Graphene Derived from Metal-Organic
Frameworks as Efficient Bifunctional Electrocatalysts for Overall Water Splitting, Adv.
Mater., 2017, 29, 1605957.
[31] D. Das, A. Das, M. Reghunath, K. Kar Nanda, Phosphine-free avenue to Co2P
nanoparticle encapsulated N,P co-doped CNTs: a novel non-enzymatic glucose sensor
and an efficient electrocatalyst for oxygen evolution reaction, Green Chem., 2017, 19,
1327-1335
[32] F. Bu, P. Xiao, J. Chen, M. F. Aly Aboud, I. Shakir, Y. Xu, Rational design of
three-dimensional graphene encapsulated core–shell FeS@carbon nanocomposite as a
flexible high-performance anode for sodium-ion batteries, J. Mater. Chem. A, 2018, 6,
6414-6421.
[33] J. Deng, D. Deng, X. Bao, Robust catalysis on 2D materials encapsulating metals:
concept, application, and perspective, Adv. Mater., 2017, 29, 1606967.
[34] Y. Li, T. Gao, Y. Yao, Z. Liu, Y. Kuang, C. Chen, J. Song, S. Xu, E. M. Hitz, B.
Liu, R.J. Jacob, M.R. Zachariah, G. Wang, L. Hu, In Situ “Chainmail Catalyst”
Assembly in Low-Tortuosity, Hierarchical Carbon Frameworks for Efficient and Stable
Hydrogen Generation, Adv. Energy Mater., 2018, 8, 1801289.
[35] Z. Wei,. J. Wang, S. Mao, D. Su, H. Jin, Y. Wang, F. Xu, H. Li, H. Y. Wang, In
Situ-Generated Co0-Co3O4/N-Doped Carbon Nanotubes Hybrids as Efficient and
Chemoselective Catalysts for Hydrogenation of Nitroarenes, ACS Catal., 2015, 5,
4783-4789.
[36] H. Yang, R. Nie, W. Xia, X. Yu, D. Jin, X. Lu, D. Zhou, Q. Xia, Co embedded
within biomass-derived mesoporous N-doped carbon as an acid-resistant and
chemoselective catalyst for transfer hydrodeoxygenation of biomass with formic acid,
Green Chem., 2017, 19, 5714-5722.
[37] L. Liu, S. Ci, L. Bi, J. Jia, Z. Wen, Three-dimensional nanoarchitectures of Co
nanoparticles inlayed on N-doped macroporous carbon as bifunctional electrocatalysts
for glucose fuel cells, J. Mater. Chem. A, 2017, 5, 14763-14774.
[38] H. Xu, S. Ci, Y. Ding, G. Wang, Z, Wen, Recent advances in precious metal-free
bifunctional catalysts for electrochemical conversion systems, J. Mater. Chem. A, 2019,
7, 8006-8029.
[39] S. Li, G. Zhang, X. Tu, J. Li, Polycrystalline CoP/CoP2 structures for efficient full
water splitting, ChemElectroChem, 2018, 5, 701-707.

Green Chemistry
Page 28 of 30
[40] P. Ren, S. Ci, Y. Ding, Z. Wen, Molten-salt-mediated synthesis of poDroOIu: 1s0.1F0e3V9-i/eCw9AGrtCic0le1O72nl0inHe
containing N-doped carbon as efficient cathode catalysts for microbial fuel cells, Appl.
Surf. Sci., 2019, 481, 1206-1212.
[41] A. Szabolcs, M. Molnár, G. Dibó, T.M. László, Microwave-assisted conversion of
carbohydrates to levulinic acid: An essential step in biomass conversion, Green Chem.,
2017, 15, 439-445.
[42] P. Bhaumik, P. L. Dhepe, Conversion of biomass into sugars. Biomass sugars for
non-fuel applications. Royal Society Chemistry, 2015, Chapter 1, 1-53.
[43] H. Qi, B. Schulz, T. Vad, J. Liu, E. Mäder, G. Seide, T. Gries, Novel carbon
nanotube/cellulose composite fibers as multifunctional materials, ACS Appl. Mater.
Interfaces., 2015, 7, 22404–22412.
[44] Y. Chen, P. Pötschke, J. Pionteck, B. Voit, H. Qi, Smart cellulose/graphene
composites fabricated by in situ chemical reduction of graphene oxide for multiple
sensing applications, J. Mater. Chem. A, 2018, 6, 7777-7785.
[45] S. Zhou, G. Chen, F. Xiao, W. Ming, H. Qi, In situ MnO x/N-doped carbon aerogels
from cellulose as monolithic and highly efficient catalysts for the upgrading of
bioderived aldehydes, Green Chem. 2018, 20, 3593.
[46] C. B. Godiya, X. Cheng, D. Li, Z. Chen, X. Lu, Carboxymethyl
cellulose/polyacrylamide composite hydrogel for cascaded treatment/reuse of heavy
metal ions in wastewater, J. Hazard. Mater. 2019, 364, 28-38.
[47] W. Yang, X. Liu, X. Yue, J. Jia, S. Guo, Bamboo-like Carbon Nanotube/Fe3C
Nanoparticle Hybrids and Their Highly Efficient Catalysis for Oxygen Reduction, J.
Am. Chem. Soc. 2015, 137, 1436-1439.
[48] Q. Bi, J. Lin, Y. Liu, H. He, F. Huang, Y. Cao, Dehydrogenation of Formic Acid
at Room Temperature: Boosting Palladium Nanoparticle Efficiency by Coupling with
Pyridinic-Nitrogen-Doped Carbon, Angew. Chem., Int. Ed., 2016, 55, 11849-11853.
[49] H. Su, P. Gao, M. Y. Wang, G. Y. Zhai, J. J. Zhang, T. J. Zhao, J. Su, M. Antonietti,
X. H. Li, J. S. Chen, Grouping Effect of Single Nickel−N4 Sites in Nitrogen-Doped
Carbon Boosts Hydrogen Transfer Coupling of Alcohols and Amines, Angew. Chem.,
Int. Ed., 2018, 57, 15194-15198.
[50] S. Tian, Z. Wang, W. Gong, W. Chen, Q. Feng, Q. Xu, C. Chen, C. Chen, Q. Peng,
L. Gu, H. Zhao, P. Hu, D. Wang, Y. Li, Temperature-Controlled Selectivity of
Hydrogenation and Hydrodeoxygenation in the Conversion of Biomass Molecule by
the Ru1/mpg-C3N4 Catalyst, J. Am. Chem. Soc., 2018, 140, 11161-11164.
[51] W. Feng, H. Dong, L. Niu, X. Wen, L. Huo, G. Bai, A novel Fe3O4@nSiO2@

Page 29 of 30
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01720H
NiPd–PVP@mSiO2 multi-shell core–shell nanocomposite for cinnamic acid
hydrogenation in water, J. Mater. Chem. A, 2015, 3, 19807-19814.
[52] N. S. Date, A. M. Hengne, K. W. Huang, R. C. Chikate, C. V. Rode, Single pot
selective hydrogenation of furfural to 2-methylfuran over carbon supported iridium
catalysts, Green Chem., 2018,20, 2027-2037.
[53] L. Lin, S. Yao, R. Gao, X. Liang, Q. Yu, Y. Deng, J. Liu, M. Peng, Z. Jiang, S. Li,
Y. Li, X. Wen, W, Zhou, D. Ma, A highly CO-tolerant atomically dispersed Pt catalyst
for chemoselective hydrogenation, Nat. Nanotechnol., 2019, 1, 354-361.
[54] R. V. Jagadeesh, K. Natte, H. Junge, M. Beller, Nitrogen-doped graphene-
activated iron-oxide-based nanocatalysts for selective transfer hydrogenation of
nitroarenes, ACS Catal., 2015, 5, 1526-1529.
[55] Y. Liu, H. Wang, T. Zhao, B. Zhang, H. Su, Z. Xue, X. Li, J. Chen, Schottky
Barrier Induced Coupled Interface of Electron-Rich N-Doped Carbon and Electron-
Deficient Cu: In-Built Lewis Acid–Base Pairs for Highly Efficient CO2 Fixation, J. Am.
Chem. Soc., 2018, 141, 38-41.
[56] H. Zhou, H. Xu, Y. Liu, Aerobic oxidation of 5‑hydroxymethylfurfural to 2, 5-
furandicarboxylic acid over Co/Mn-lignin coordination complexes-derived catalysts,
Appl. Catal. B: Environ. 2019, 244, 965-973.
[57] H. Su, K. Zhang, B. Zhang, H. Wang, Q. Yu, X. Li, M. Antonietti, J. Chen,
Activating cobalt nanoparticles via the Mott–Schottky effect in nitrogen-rich carbon
shells for base-free aerobic oxidation of alcohols to esters, J. Am. Chem. Soc., 2017,
139, 811-818.
[58] Y. Cai, X. Li, Y. Zhang, X. Wei, K. Wang, J. Chen, Highly Efficient
Dehydrogenation of Formic Acid over a Palladium‐Nanoparticle‐Based Mott–Schottky
Photocatalyst, Angew. Chem., Int. Ed., 2013, 52, 11822-11825.
[59] B. Albert, M. Dörthe, G. R. Felix, J. Ralf, J. Henrik, P. J. Dyson, L. Gábor, L. Ralf,
B. Matthias, Efficient dehydrogenation of formic acid using an iron catalyst, Science,
2011, 333, 1733-1736.
[60] Q. Bi, X. Du, Y. Liu, Y. Cao, H. He, K. Fan, Efficient subnanometric gold-
catalyzed hydrogen generation via formic acid decomposition under ambient conditions,
J. Am. Chem. Soc., 2012, 134, 8926-8933.
[61] Q. Liu, X. Yang, Y. Huang, S. Xu, X. Su, X. Pan, J. Xu, A. Wang, C. Liang, X.
Wang, A Schiff base modified gold catalyst for green and efficient H2 production from
formic acid. Energy Environ. Sci, 2015, 8, 3204-3207.