Selective electrochemical hydrogenation of furfural to 2-methylfuran over a single atom Cu catalyst under mild pH conditions

Article abstract of DOI:10.1039/d0gc03999c

Furfural is regarded as one of the most promising bio-based feedstocks in the bio-refinery industry. Selective hydrogenation of the carbonyl bond in furfural plays a vital role in its conversion to downstream products. Electrochemical hydrogenation (ECH) method provides a green and sustainable way for this reaction. Yet, it still suffers from harsh pH conditions and low selectivity for highly reduced products, such as 2-methylfuran. In this study, high faradaic efficiencies of over 90% for furfuryl alcohol and 60% for 2-methylfuran were obtained in a near-neutral environment (pH = 5) at ?0.75 V and ?0.90 Vvs. the reversible hydrogen electrode, respectively. The key to this success is the integration of single atom copper active sites and the oxophilic phosphorus dopants in a single catalyst. Single atom Cu sites are found to be the active centers for this reaction and decreasing the size of Cu sites to a single atom enhances the efficiencies of the ECH reactions by suppressing the competing hydrogen evolution reaction. Phosphorus doping facilitates furfural hydrogenation to 2-methylfuranviaa sequential two-step reduction process. This study opens up possibilities for the selective electrochemical hydrogenation of furfural to 2-methylfuran under mild conditions.

Full text of DOI:10.1039/d0gc03999c

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Selective electrochemical hydrogenation of
furfural to 2-methylfuran over a single atom Cu
catalyst under mild pH conditions†
Cite this: Green Chem., 2021, 23,
028
3
a
b
a
a
c
a
Peng Zhou, Yu Chen, Peng Luan, Xiaolong Zhang, Ziliang Yuan, Si-Xuan Guo,
d
d
a
a
Qinfen Gu, * Bernt Johannessen,
David R. Turner
Mamun Mollah, Alan L. Chaffee,
a
a
and Jie Zhang
*
Furfural is regarded as one of the most promising bio-based feedstocks in the bio-refinery industry.
Selective hydrogenation of the carbonyl bond in furfural plays a vital role in its conversion to downstream
products. Electrochemical hydrogenation (ECH) method provides a green and sustainable way for this
reaction. Yet, it still suffers from harsh pH conditions and low selectivity for highly reduced products, such
as 2-methylfuran. In this study, high faradaic efficiencies of over 90% for furfuryl alcohol and 60% for
2-methylfuran were obtained in a near-neutral environment (pH = 5) at −0.75 V and −0.90 V vs. the
reversible hydrogen electrode, respectively. The key to this success is the integration of single atom
copper active sites and the oxophilic phosphorus dopants in a single catalyst. Single atom Cu sites are
found to be the active centers for this reaction and decreasing the size of Cu sites to a single atom
enhances the efficiencies of the ECH reactions by suppressing the competing hydrogen evolution reac-
tion. Phosphorus doping facilitates furfural hydrogenation to 2-methylfuran via a sequential two-step
reduction process. This study opens up possibilities for the selective electrochemical hydrogenation of
furfural to 2-methylfuran under mild conditions.
Received 25th November 2020,
Accepted 17th March 2021
DOI: 10.1039/d0gc03999c
rsc.li/greenchem
for adhesive production, while MF is an excellent additive for
gasoline due to its high octane number. Conventional
Introduction
Biomass is a promising sustainable carbon source that can methods for selective transformation of furfural include hydro-
3
,4
replace traditional fossil fuels as raw materials for the pro- deoxygenation (HDO) and catalytic hydrogen transfer (CHT)
1
2a,5
duction of downstream value-added chemicals and fuels.
reactions.
However, the large quantity of explosive hydrogen
Highly efficient and selective transformation, especially hydro- gas, which is mainly derived from non-renewable coal and
genation, of the platform molecules derived from biomass is natural gas, needed in the HDO reactions and the high temp-
2
important in the bio-refinery industry. According to the U.S. erature required for the CHT reactions constitute a large frac-
Department of Energy report in 2014, furfural, obtained from tion of the overall operational and energy costs. Moreover, the
acid-catalyzed dehydration of xylose, has been identified as a CHT reactions are not atom economic since the use of formic
3
key platform moiety in the bio-refinery industry. Value-added acid and alcohol as a hydrogen donor results in chemical
2
a,5b
reduction products from furfural, such as furfuryl alcohol (FA) wastes.
Therefore, the traditional methods in the bio-refin-
and 2-methylfuran (MF), are of great importance in the chemi- ery industry may lead to the consumption of fossil resources
cal and energy industries. For instance, FA is a raw material and the production of chemical waste, which could contribute
6
to climate change and soil acidification.
Electrochemical hydrogenation (ECH) provides a sustain-
able alternative to traditional methods, especially when the
electricity used is generated from renewable energy sources,
such as wind and solar. It can be operated under mild con-
a
School of Chemistry, Monash University, Wellington Road, Clayton 3800, Victoria,
Australia. E-mail: jie.zhang@monash.edu
b
Monash Center for Electron Microscopy, Monash University, Wellington Road,
Clayton 3800, Victoria, Australia
c
School of Chemistry, South-Central University for Nationalities, Wuhan, 430074,
ditions, avoiding the use of both flammable and explosive H
2
P.R. China
d
gas and high temperature. In the ECH, the H atom comes
from chemisorbed hydrogen (Hads) generated in situ from non-
toxic water by renewable electricity, thus providing a sustain-
able way to store energy in the chemical form of easily trans-
Australian Synchrotron, ANSTO, 800 Blackburn Rd, Clayton, VIC 3168, Australia.
E-mail: qinfeng@ansto.gov.au
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0gc03999c
3028 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
petition from the HER and low carbon balance due to the for-
7
b,8a
mation of by-products.
Therefore, the conversion of fur-
fural to MF in a mildly acidic condition might also provide a
high carbon balance. Low current density and high overpoten-
tial also diminish the product selectivity as the acid-catalyzed
7
b,14
polymerization of furfural or FA may occur.
Hence, there is
clearly still considerable room for improvement with more
advanced Cu catalysts for this reaction.
Recently, isolated single atom catalysts (SACs) have shown
significant advantages in electrocatalytic processes such as the
HER, oxygen reduction, and CO2 reduction reaction.
Scheme 1 Schematic diagram of an ECH cell and the main products of
the ECH of furfural (standard reduction potential (E° vs. SHE) values are
obtained from literature ).
6
15
16
17
Homogeneously distributed and isolated active sites of the
SACs closely resemble the active sites in homogeneous mole-
portable liquid fuels such as MF and 2-methyl tetrahydrofuran cular catalysts, bridging the gap between heterogeneous and
1
8
(
MTHF) (Scheme 1). However, the ECH method still suffers homogeneous catalysts. In comparison with the traditional
from low product selectivity due to the competing hydrogen nanoparticle based heterogeneous catalysts where the utiliz-
evolution reaction (HER) and the polymerization side reaction. ation of metal atoms is poor, all metal sites in SACs are
H
ads from water can either dimerize to form H
2
or react with involved in the reactions. Therefore, catalyst loading can be
furfural to form FA via a 2 Hads process or MF via a 4 Hads minimized. This is particularly important for the noble metal-
1
8
process. Additionally, after receiving a Hads, furfural can based catalysts. The structurally uniform active sites could
7
dimerize via C–C coupling and further polymerize. Thus, the enhance the selectivity of the reaction, which is vital in the
faradaic efficiency (F.E.) and product distribution of the ECH ECH of furfural. However, the applications of SACs in the ECH
reactions are closely related to the HER overpotential on the of biomass have not been widely explored. Several strategies
7
,8
8e
19
catalysts used. Among the catalysts tested (such as Cu, Pb,
have been developed to synthesize SACs. High temperature
8
e
6,9
21
10
8d
Fe, Pt, Pd, Ti, graphite ), Cu is the most promising pyrolysis of metal containing organic polymers provides a
one for electroreduction of furfural due to its high abundance rather convenient route to synthesize carbon supported
7
a,11
20
in nature and moderate HER overpotential.
Competition SACs. However, tedious procedures and expensive precursors
also exists between the formation of FA and MF from furfural for the synthesis of organic polymers still limit large scale syn-
7
a
as FA cannot be further reduced due to its weak adsorption thesis of SACs via this method.
and consequently the low reactivity on Cu surface found in
Here we report a simple and easily scaled-up fabrication
heterogeneous catalysis. It was found that the concentration of method for the formation of high surface area phosphorus-
+
H in the bulk solution plays an important role in product dis- doped carbon supported Cu catalysts for the ECH of furfural
tribution by influencing the Hads coverage on the catalyst sur- with controllable size of Cu sites ranging from nanoparticles to
7
a
face. A high pH environment is usually needed to generate single atoms. This strategy involves pyrolysis of the bio-waste
8
e
2+
FA with high F.E. A lower Hads coverage in a higher pH chitosan with coordinated Cu ions. Phosphoric acid is chosen
environment benefits the formation of FA, which requires less to dissolve chitosan, which serves as an effective activation
H
ads compared to MF. For instance, Liu et al. employed a Cu agent and provides phosphorus dopants for the formation of a
foil for the electroreduction of furfural in a buffer solution at porous phosphorus-doped carbon nanostructure upon pyrol-
21
pH 10. At a potential of −1.4 V vs. the reversible hydrogen elec- ysis. The Cu SACs supported on phosphorus-doped carbon
trode (RHE), selectivity and F.E. for FA of 87% and 71% were (PC) exhibit superior performance in electrochemical hydrogen-
8
e
achieved respectively. More recently, Zhang et al. reported ation of furfural and show an unexpected potential dependent
nearly 100% carbon balance and F.E. for FA in a 0.1 M KOH product distribution in a near-neutral acetate buffer (pH 5)
electrolyte solution (pH = 13) using copper phosphide as the environment. Formation of MF is favored at more negative
8
c
catalyst. However, to produce MF, the reaction has to be potentials than that for FA. Control experiments showed that
+
carried out in strongly acidic media to provide more H and single atom Cu sites and phosphorus dopants are two essential
hence higher Hads coverage, which also inevitably results in components in facilitating the production of MF from furfural
7
b,8a,12
more polymerization products.
For example, the pro- via a two-step reduction process. Use of mild pH conditions
duction of a 4-electron reduced product, 2-methylfurfuryl, has instead of strong acid/base allows less corrosion resistive
only been reported in low pH environments via a one-step materials to be used to construct the reactors. Furthermore,
1
1–13
reaction from furfural.
By optimizing the size of Cu par- when the furfural feedstock is derived from solid acid catalysed
22
ticles and electrolysis parameters, Biddinger and co-workers transformation, strong mineral acids such as H SO , which
2
4
achieved a F.E. of 73% for MF on nanocrystalline Cu (400 nm) may introduce safety and health hazards, can be completely
in a 0.5 M H SO electrolyte solution (pH = 0.3) at an applied avoided in the biomass upgrading process. This work may
2
4
potential of −0.8 V vs. RHE, but with a low carbon balance of inspire new discoveries in the refining of oxygen-containing
8
a
6
7%. Thus, the major challenge in producing MF from fur- biomass molecules via a greener route under more benign reac-
fural lies in the low faradaic efficiency due to the severe com- tion conditions with less energy consumption.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3028–3038 | 3029

View Article Online
Paper
Green Chemistry
Results and discussion
Fabrication and characterization of Cu catalysts
The process of fabrication of Cu catalysts is shown in
Scheme 2. By controlling the Cu loading in the catalyst, we
have succeeded in synthesizing Cu catalysts with different Cu
site sizes ranging from a single atom to a nanoparticle.
Chitosan was chosen as the carbon source due to its wide
availability from the shells of crabs and its rich oxygen/nitro-
gen content. Oxygen/nitrogen in chitosan and phosphorus in
phosphoric acid assist the chelating of Cu and prevent the
agglomeration during pyrolysis, thus aiding the formation of
SACs. Moreover, phosphoric acid also provides the low pH con-
dition needed for the homogeneous dispersion of chitosan in
water. The amount of copper ions was carefully controlled to
avoid agglomeration which results in copper sites with larger
sizes. Three catalysts with different Cu loadings were prepared.
For convenience of presentation, these catalysts with increas-
1 2 3
ing Cu loading are named Cu /PC, Cu /PC and Cu /PC. The
exact loadings of copper in each catalyst were determined to
be 0.45, 2.2 and 5.7 wt% using inductively coupled plasma-
1 2 3
mass spectrometry (ICP-MS) for Cu /PC, Cu /PC, and Cu /PC,
respectively.
The morphologies of the as-synthesized catalysts and the
size and distribution of Cu sites were first investigated using
transmission electron microscopy (TEM) and aberration-cor-
rected high-angle annular dark-field scanning transmission
electron microscopy (HAADF-STEM). TEM image of Cu
showed no agglomerated Cu nanoparticles (Fig. 1a). Atomic
resolution HAADF-STEM image of Cu /PC (Fig. 1b) shows
numerous uniformly distributed bright dots with the size (ca.
.2 nm) comparable to that of a Cu atom, suggesting the for-
1
/PC
1
Fig. 1 (a) TEM and (b) atomic-resolution HAADF-STEM images for the
Cu
1
/PC catalyst. (c) TEM image and the corresponding EDX mapping for
0
Cu
1
/PC. TEM images for (d) Cu /PC and (e) Cu /PC. (f) PXRD patterns
2
3
mation of homogeneously distributed Cu single atoms. for the Cu /PC catalysts.
x
Furthermore, the presence of uniformly distributed C, N, O, P
and Cu elements in the Cu /PC sample was confirmed by
1
elemental mapping using energy-dispersive X-ray spectroscopy
(
EDX; Fig. 1c). TEM image of Cu
Cu nanoparticles with an average size of 7.2 nm (Fig. 1d), Cu
which is consistent with the calculated value (10.3 nm) from for the (002) and (101) planes of graphitic carbon (Joint
PXRD based on the Scherrer equation while Cu
/PC shows ran- Committee on Powder Diffraction Standards (JCPDS) no: 01-
domly distributed Cu nanoparticles with larger sizes (20 nm, 089-8487). The broadening of these two peaks could be due
2
/PC shows the formation of
Powder X-ray diffraction (PXRD) patterns of the as-prepared
/PC catalyst (Fig. 1f) only showed two characteristic peaks
1
3
2
3
Fig. 1e).
to the amorphous graphite structure caused by heteroatom
doping. No Bragg peaks corresponding to metallic Cu were
observed, which is consistent with the HAADF-STEM results.
2
4
In contrast, PXRD patterns of both Cu
2 3
/PC and Cu /PC show
an additional Bragg peak at 43.5°, which can be assigned to
2
5
the Cu (111) plane (JCPDS no: 00-004-0836).
The chemical states of Cu were first investigated by high-
resolution X-ray photoelectron spectroscopy (XPS) for the Cu
2
p peaks of Cu /PC (Fig. S1a†). In the Cu 2p core level XPS
1 3/2
region, the binding energy of 935.7 eV was found. The detec-
tion of the satellite peak at 963.3 eV confirms the presence to
2
+
Cu . The high-resolution XPS spectra of C 1s (Fig. S1b†) can
be deconvoluted into the C–C (284.9 eV), C–P (285.5 eV), C–N
2
6
(
286.5 eV), and C–O (287.2 eV) bonding. The high-resolution
x
Scheme 2 Schematic illustration of the fabrication process for the Cu /
PC catalysts.
XPS spectra of O 1s (Fig. S1c†) can be deconvoluted into 3
3030 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
peaks, which are assigned to CvO (531.5 eV), P–O (533 eV), shows a predominant peak at around 1.5 Å which is much
2
6,27
and C–O–C (534.5 eV).
Fig. S1d† shows the high-resolution smaller than that of ca. 2.2 Å of the Cu foil sample, and no
XPS spectra for N 1s with contributions from pyridinic N peak for Cu–Cu bond could be observed, indicating that Cu
2
6
(
398.2 eV), pyrrolic N (399.8 eV) and graphitic N (401.0 eV).
most likely coordinates with the atoms that are lighter than
2
8
In principle, the N–P bonding could also contribute to the itself, such as O or N in PC. EXAFS wavelet transform (WT)
peak at 401.0 eV. The elemental composition of Cu /PC was analysis, which can provide not only radial distance resolution
1
2
5
obtained from XPS and elemental analysis (Table S1†). The but also wavenumber (k)-space resolution, was employed to
results show that carbon and oxygen form the major part of further ascertain the atomic Cu dispersion. As shown in
this material followed by phosphorus and the amount of nitro- Fig. 2c and d, one intensity maximum at 4.0 Å− was obtained
1
gen is far less. X-ray absorption spectroscopy (XAS) was further from the WT analysis of Cu
used to obtain the valence state and structural information of different from the value found on Cu foil (6.5 Å− ). Thus, the
the Cu species in Cu /PC at the atomic scale. Fig. 2a shows the Cu species in Cu /PC are confirmed to be atomically distribu-
X-ray absorption near edge structure (XANES) spectra of Cu ted and coordinated with O/N. High-resolution synchrotron
PC at the Cu K-edge with Cu foil, copper(II) oxide (CuO), powder X-ray diffraction pattern (Fig. 2e) also shows no visible
copper(I) phosphide, and copper(I) oxide (Cu O) as the refer- Bragg peaks for CuO or Cu phases, indicating the wt% of these
ence samples. The absorption edge of Cu /PC is located at phases are below 0.1 wt%. Pair distribution function (PDF)
997.2 eV, which is similar to that of CuO, and principal com- G(r) analysis of the Cu /PC sample is shown in Fig. 2f. The
1
/PC, which is significantly
1
1
1
1
/
2
1
8
1
1
ponent analysis of XANES data of Cu /PC indicated 73% simi- peak positions of the PDF pattern correspond to the distance
larity to the CuO reference sample, implying the average of atomic pairs in the material. The predominant peaks in
valence state of the Cu atom is close to the +2 oxidation PDF pattern of Cu /PC centered at 1.4 Å and 2.4 Å correspond
1
28
state, which is consistent with the XPS results. The coordi- to the two shortest distances between carbon atoms sitting on
nation environment of Cu was confirmed by the extended the vertices of hexagonal rings with (sp )–(sp ) bonding, indi-
2
2
29
X-ray absorption fine structure (EXAFS). The Fourier-trans- cating the high content of C in the Cu /PC sample. The much
1
2
formed (FT) k -weighted EXAFS spectrum of Cu
1
/PC (Fig. 2b) weaker peaks observed at 1.9 Å and 3.0 Å could be assigned to
Cu–N/Cu–O bonds, while the shortest distances of the Cu–Cu
bond (2.58 Å, 3.66 Å, 4.48 Å) are absent. We further carried out
1
least-squares EXAFS fitting analysis of Cu /PC (Fig. 2g,
Table S2†), and the fitting result shows that Cu coordinates
with 4 planar O atoms with a distance of 1.95 Å and 2 vertical
O atoms with a larger distance of 2.95 Å.
Nitrogen adsorption–desorption measurements were con-
ducted at 77 K using a Micromeritics 3 Flex instrument to
evaluate the BET surface areas and pore size distributions of
the as-prepared Cu
x 2
/PC catalysts. N adsorption isotherms of
these samples were type I isotherms as defined by IUPAC
3
0
(
Fig. S2†). The BET surface areas for the Cu
1 2
/PC, Cu /PC and
2
−1
Cu /PC samples are determined to be 751, 997 and 844 m g
3
respectively. All Cu /PC samples contain mesopores with an
x
average pore size of 2.6–2.8 nm as determined by the Barrett–
Joyner–Halenda method (Table S3†). These materials also
2
−1
exhibit a significant microporous area of about 200 m
g
based on the t-plot analysis (Fig. S2c–e and Table S3†). By con-
trast, the BET surface area of Cu/C (prepared in the same way
as Cu /PC except acetic acid was used instead of phosphoric
1
2
−1
acid) is as small as 245 m g , suggesting the important role
of the phosphorus dopants in the formation of the porous
structure of PC.
Electrochemical hydrogenation of furfural
Fig. 2 (a) Cu K-edge XANES spectra of Cu
Cu metal. (b) Fourier transformation of the k -weighted Cu K-edge
EXAFS spectra obtained on Cu
1 2 3
/PC, Cu O, CuO, Cu P, and
2
Considering the poor carbon balance reported in harsh pH
7
b,8a
1
/PC, CuO and Cu metal. (c and d) conditions
and the high F.E. of FA and carbon balance
2
8c
Wavelet transformations for the k -weighted Cu K-edge EXAFS signals
of Cu /PC (c) and Cu foil (d). (e) Synchrotron PXRD pattern of Cu /PC (λ
0.58842 Å). (f) PDF pattern for Cu /PC. (g) The corresponding Cu
/PC at R-space inset of (g) is sche-
obtained in pH 8, the experiment was initially conducted in
mildly acidic pH conditions in order to provide a balance for
the formation of MF with a higher selectivity. As the HER is a
major competing reaction for the ECH reactions, the HER
activity of the as-prepared catalysts was first examined in N2-
1
1
=
1
K-edge EXAFS fitting curves for Cu
1
1
matic model for Cu /PC (the blue ball and the red balls represent Cu
and O atoms respectively).
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3028–3038 | 3031

View Article Online
Paper
Green Chemistry
process, which has not been reported previously, could either
be assigned to the reduction of FA to MF (2-Hads process) or
furfural to MF (4-Hads process). To identify the second
reduction process, electrochemical reduction of FA was also
undertaken on the Cu /PC catalyst under the same conditions.
1
A reduction process was observed at around −1.1 V vs. RHE
(
Fig. 3b), suggesting that the second reduction process associ-
ated with furfural may be attributed to the reduction of FA.
These results indicate that the Cu /PC catalysts may catalyze
x
the ECH of furfural to MF at pH 5 via two sequential 2-Hads
transfer steps, which is different from the previously reported
one-step 4-Hads transfer process mainly observed in strongly
7b
acidic media. Direct transformation of furfural to MF via a
-Hads transfer step may require a higher Hads surface coverage
4
Fig. 3 Cyclic voltammograms recorded in
2
a N -saturated acetate
compared to the 2-Hads transfer process, thus may not be
buffer solution (pH 5) using (a) Cu
Cu /PC modified GC electrode in the presence and absence of 20 mM
of furfural or FA, (c) Cu /PC and (d) Cu /PC modified GC electrodes in
the presence and absence of 20 mM of furfural. Scan rate = 50 mV s
x
/PC modified GC electrodes, (b) a
7a
favourable in a near neutral environment. Multi-cycle cyclic
1
voltammograms were also recorded in a N -saturated acetate
2
2
3
−1
.
buffer solution (pH 5) containing 20 mM furfural using Cu
PC, Cu /PC and Cu /PC modified GC electrodes with a scan
rate = 50 mV s− (Fig. S4†). A slight decrease in current density
1
/
2
3
1
satrurated acetate buffer (pH = 5) (Fig. 3a) and phosphate was observed upon cycling of the potential due to the mass
buffer (pH = 8) (Fig. S3a†) solutions using cyclic voltammetry transport effect under the stationary conditions while other key
(CV). CVs were recorded in the potential region where metallic voltammetric characteristics remained unchanged. Cyclic vol-
copper is thermodynamically stable. In the acetate buffer solu- tammetric experiments were also undertaken in a phosphate
tion (Fig. 3a), the onset potential (arbitrarily defined as the buffer solution of pH 8 in the presence and absence of furfural
potential where the current density reaches 0.5 mA cm− ) for or FA (Fig. S3b–d†). Only one reduction process was observed
2
the HER process on Cu
more negative than that of Cu
PC, (−0.48 V vs. RHE) indicating that the HER process is process for FA was observed (Fig. S3b–d†) suggests that FA
heavily suppressed at Cu /PC. A similar trend was also found cannot be reduced to MF via a 2-Hads transfer process in a
in the phosphate buffer solution (Fig. S3a†). The heavily sup- basic environment as expected due to the low Hads coverage.
pressed HER at Cu /PC may be due to the change of the HER
Controlled potential bulk electrolysis of furfural was per-
1
/PC (−0.77 V vs. RHE) is significantly which may be assigned to the reduction of furfural via a 2-H
ads
8c
2
/PC (−0.59 V vs. RHE) and Cu
3
/
transfer according to the literature. The fact that no reduction
1
1
mechanism. Since there is no adjacent active site on single formed at −0.75 V vs. RHE (IR corrected potentials are reported
+
atom catalysts, the HER follows the Volmer–Heyrovrsky (H + in all cases) in an acetate buffer solution (pH = 5) containing
−
+
−
e
→ H*; H + e + H* → H ; * represents the adsorption 40 mM furfural and the results are shown in Fig. 4a. In all
2
+
−
species) mechanism instead of the Volmer–Tafel (H + e
H*; 2H* → H ) mechanism.
2
→
cases, an excellent carbon balance of nearly 100% obtained
implies the absence of the dimerization reaction and no MTHF
3
1
The ECH of furfural in an acetate buffer solution (pH = 5) was detected using a gas chromatograph. This is consistent
was investigated by CV (Fig. 3b). The current density decreased with the reported results from other groups as the dimerization
significantly upon addition of furfural, implying that the inter- and polymerization reactions mainly occur in highly acidic
7b,8a
action between the proton and catalytic sites is inhibited by media.
The highest F.E. for furfural reduction products was
furfural. This is consistent with the literature findings and obtained with Cu /PC and FA was the only reduction product of
1
could be attributed to the stronger interaction of the catalyst furfural with a F.E. of over 90%. Although the total current
−
2
−2
with the aromatic furan ring of furfural via the oxygen atom in density increased from 4.3 mA cm for Cu /PC to 5.1 mA cm
1
3
2
−2
the carbonyl group. As shown in Fig. 3b–d, voltammograms for Cu /PC and 6.7 mA cm for Cu /PC, the partial current
2
3
−2
obtained with these three catalysts in the presence of 20 mM density for FA decreased from 4.0 mA cm for Cu /PC to
1
furfural showed cathodic peaks at approximately −0.8 V vs. 3.1 mA cm− for Cu /PC and Cu /PC, suggesting a diminished
2
2
3
−2
RHE with current densities of around 3 mA cm , which are F.E. for FA upon enhanced HER activity, consistent with the vol-
assigned to the reduction of furfural, and are consistent with tammetric results (Fig. 3a). This trend indicated that the ECH of
7b,8c
the results reported in the literature.
Furthermore, the furfural was more favourable on the single atom Cu (Cu /PC)
1
reduction potentials for furfural are slightly more negative catalyst and as the size of the Cu sites increased, the HER
than the onset potential for the HER, indicating that the became more favourable. No MF was detected under these con-
electrocatalytic formation of Hads is a prerequisite step for the ditions on Cu /PC even when more charges were consumed.
1
ECH reaction. Another reduction process was observed at a The F.E. for FA decreased slightly upon consumption of more
more negative potential of about −1.1 V vs. RHE on all three charges (Fig. 4b). This is attributed to the increasing contri-
Cu /PC catalysts, but was better defined on Cu /PC. This bution of the HER when the concentration of furfural decreases,
x
1
3032 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
potential of −0.76 V vs. RHE where the reduction of FA does
not occur. As the potential became more negative than −0.84
V, FA was reduced to form MF with a low F.E. of approximately
1
5%. The highest F.E. of 26% for MF was reached at −0.91 V
vs. RHE. However, more negative applied potential resulted in
a dominant HER with a F.E. for H of nearly 100%. Comparing
2
with the electrolysis with furfural as the substrate, F.E. of MF
is much lower with FA as the substrate. This may due to the
weaker adsorption of FA on the surface of the catalyst, which
may benefit the competing HER.
The effect of substrate and product concentrations on the F.
E. of the product was further investigated at −0.90 V vs. RHE in
acetate buffer (pH = 5) and the results are summarised in
Table S4.† Increasing the concentration of furfural to 50 mM
leads to a lower F.E. for H
Table S4,† entry 1 vs. 2). This could be explained by lower Hads
availability on surface due to higher concentration of furfural,
2
and higher F.E. for FA and MF
Fig. 4 Bulk electrolysis data. (a–c) The F.E. and carbon balance data for
the bulk electrolysis of 40 mM furfural in an acetate buffer solution
obtained with (a) different catalysts after 10 C of charge were consumed
(
7a
at an applied potential of −0.75 V vs. RHE; (b) Cu
/PC after different which could block the sites for H adsorption. Moreover, the
amounts of charge were consumed at an applied potential of −0.75 V competition between FA and MF formation moves towards FA
1
1
vs. RHE; (c) Cu /PC at different applied potentials after 10 C of charge
slightly, giving a higher concentration ratio of FA/MF. Addition
of 10 mM of FA to 40 mM furfural shows a similar but smaller
influence, indicating a weaker adsorption of FA compared to
furfural (entry 1 vs. 3) while addition of 10 mM MF shows little
effect on the F.E.s (entry 1 vs. 4), implying that the interaction
between MF and the catalyst surface could be even weaker.
were consumed. (d) The F.E. and carbon balance data for the bulk elec-
trolysis of 40 mM furfuryl alcohol in an acetate buffer solution obtained
1
with Cu /PC over different applied potentials after 10 C of charge were
consumed.
since the catalyst is stable during electrolysis as discussed later Efforts for producing MF from furfural on Cu /PC and Cu /PC
2
3
in this article (vide infra). Control experiments were conducted turned out to be a failure as H was the major product when the
2
2
to investigate whether H from HER would reduce furfural or applied potential became more negative (Fig. S6†). As further
FA. First, the electrolysis of the buffer solution was conducted reduction of FA occurs at a more negative potential, it will be
in an airtight H-cell at −1.1 V in the absence of furfural or FA to more subject to the influence of HER. Thus, single atom size
produce H
electrodes. After passing 10 C of charge, furfural or FA was petition. Increasing the size of Cu can lead to severe HER and
injected into the cathodic compartment to verify whether H
hence fail to form MF. As the pH value of the electrolyte solu-
2 1
using the Cu /PC catalyst modified carbon paper Cu is important to this reaction for suppressing the HER com-
2
would reduce furfural or FA. The solution was stirred for 12 h. tion was increased to 8, no MF was detected in the product of
No FA or MF was detected confirming that furfural or FA can furfural reduction even after 30 C of charge was consumed at a
not be reduced by H under our experimental conditions. In our potential of −0.75 V vs. RHE (Fig. S7a†). The only product
2
study, the electrolysis time is far shorter than 12 h. Therefore, obtained from furfural reduction was FA with a F.E. of over 90%
the contribution of the homogeneous chemical reduction pro- after consuming 10 C of charge. The F.E. for FA decreased
cesses is negligible. Electrolysis was also performed on the Cu
1
/
slightly as the consumed charges increased due to the decrease
PC catalyst at more negative potentials where the second in the substrate concentration and the HER became more sig-
reduction process occurs and the results are shown in Fig. 4c, nificant. Applying more negative potential also showed no effect
and the current density and the partial current density are on the product distribution except diminishing the F.E. for FA
shown in Fig. S5a.† The formation rate of FA and MF was also (Fig. S7b†). Attempts to reduce FA in a basic environment also
calculated and normalized to the mass of Cu (Fig. S5b†) for the failed. When electrolysis was conducted in a phosphate buffer
readers who may not be familiar with electrochemistry. As solution containing 40 mM FA, H was the only product with a
2
expected, MF was produced at a potential of −0.82 V and F.E. of nearly 100%. These results confirm that the reduction
reached the highest F.E. of over 60% at −0.90 V (vs. RHE). As process of furfural observed in the voltammetric measurements
the applied potential became more negative than −0.96 V (vs. in a basic environment (pH = 8) is due to the 2-Hads transfer
RHE), the HER became dominant with a F.E. of 55% for H
and the F.E. for MF decreased to approximately 38% although it and neither FA nor furfural could be reduced to MF in this
became the only product of furfural reduction.
basic environment. This could be attributed to the much lower
2
,
process instead of a 4-Hads transfer one associated with furfural,
Controlled potential electrolysis was also performed with Hads coverage under these conditions. A detailed comparison of
4
0 mM FA as the substrate in an acetate buffer solution at our results with the previously reported ones can be found in
different potentials and the results are shown in Fig. 4d. Table S5.† Significantly, Cu /PC is the first reported catalyst for
1
Consistent with the results obtained from the electrolysis of the selective production of MF from furfural in a near neutral
furfural, the charges were entirely consumed by the HER at a environment.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3028–3038 | 3033

View Article Online
Paper
Green Chemistry
1
Reusability of the Cu /PC catalyst
The recyclability and durability of the catalyst was evaluated by
performing 6 consecutive electrolysis experiments using a Cu
1
/
PC modified carbon paper electrode. In each experiment, the
electrolysis was conducted at an applied potential of −0.90 V
in a freshly prepared acetate buffer (pH = 5) solution contain-
ing 40 mM furfural, and 10 C of charge was consumed. After
1
each experiment, the Cu /PC modified carbon paper electrode
was washed carefully with water and dried before use for the
next experiment. The results of each experiment are shown in
Fig. 5a. The current density and the F.E. for MF remained
essentially unchanged in the six separate electrolysis experi-
ments. The Cu K-edge EXAFS spectrum taken after electrolysis
is similar to the one obtained before electrolysis (Fig. 5b), indi-
cating the high reusability of our Cu /PC catalyst.
1
Investigation of mechanisms
To identify the catalytically active sites of Cu /PC, a series of
1
control experiments were undertaken. Voltammetric experi-
ments were first performed using P-doped carbon material Fig. 6 Comparison of voltammograms obtained in an acetate buffer
(
PC) modified GC electrodes in an acetate buffer in the pres- (pH = 5) for (a) PC and (b) Cu/C in the presence and absence of 20 mM
furfural; (c) Cu/C in the presence and absence 20 mM FA.
ence and absence of 20 mM furfural (Fig. 6a). In the presence
of furfural, no obvious additional reduction process was
observed although the HER current density decreased due to
the adsorption of furfural. Controlled potential bulk electroly-
of electrons transferred during these processes based on
Faraday’s law (Q = nNF, where Q is the amount of charge con-
sumed during the process which can be obtained from the
2
sis with PC-modified carbon paper electrodes showed H was
the only product, suggesting that copper is the active site. To
obtain the information about the valence state of copper
under the catalytic turnover conditions, voltammetric experi-
3
3
integration of the background subtracted voltammogram,
n
represents the number of electrons involved per copper
species, N is the amount of copper involved in moles and F is
the Faraday constant). A value of 2.1 for n was obtained, which
ments were performed using the Cu /PC modified electrode in
1
an acetate buffer (pH 5, Fig. S8a†). A broad reduction process
was observed at 0.3 V vs. RHE, although it is not well-defined
due to the slow electron transfer rate and the very low loading
of Cu on the catalyst. To make it better defined, the same
2
+/0
is close to the theoretical value of 2 for the Cu
process, indi-
cating that Cu is reduced to a valence state of 0 under the cata-
lytic turnover conditions. Since the reduction of furfural
occurs at a far more negative potential of at least −0.8 V vs.
RHE, Cu is the active site of the catalyst. And after the reaction,
experiment was conducted in a 0.01 M H
2 4
SO aqueous (pH =
1
.7) solution where charge neutralization can be facilitated via
2
+
the Cu site maybe re-oxidized to Cu by air according to the
proton transfer. Two closely spaced reduction processes were
observed at around 0.2 V vs. RHE (Fig. S8b†), which were not
found with the PC modified electrode, and are attributed to
the reduction of copper. The total amount of charge associated
with these processes can be calculated to estimate the number
Cu k-edge EXAFS spectrum of Cu /PC catalyst taken after elec-
1
trolysis (Fig. 5b).
To illustrate the role of P in this reaction, we have prepared
a reference sample without a P dopant (Cu/C) using the same
procedure for Cu
amount as phosphoric acid was used to dissolve chitosan
PXRD shown in Fig. S9†). The reduction of FA to MF was not
1
/PC except that acetic acid of the same molar
(
found in the acetate buffer solution (pH = 5) containing fur-
fural or FA under voltammetric conditions when a Cu/C modi-
fied GC electrode was used (Fig. 6), suggesting the important
roles of P in the formation of MF. A bulk electrolysis experi-
ment was performed using the Cu/C modified carbon paper
electrode in the acetate buffer solution containing 40 mM fur-
fural at an applied potential of −0.8 V vs. RHE, and FA was the
only furfural reduction product with a F.E. of over 80%. These
results demonstrated that P dopants also play an important
role in the ECH of furfural to form MF.
1 1
Fig. 5 Reusability of the Cu /PC catalyst: (a) Recyclability test of Cu /
PC. Each electrolysis experiment was conducted with an applied poten-
tial of −0.90 V in an acetate buffer (pH 5) solution containing 40 mM
furfural. Products were analysed after 10 C of charge were consumed.
The doping effect of phosphorus on carbon was examined
by Raman spectroscopy. The results are shown in Fig. S10.†
(
b) Comparison of the Cu K-edge EXAFS spectra of the Cu
1
/PC catalyst
before (red line) and after (blue circle) electrolysis.
3034 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
The observation of the characteristic D and G bands at 1390
and 1550 cm− , respectively, confirmed the presence of a gra-
1
x D G
phene structure in the Cu /PC and Cu/C samples. Similar I /I
ratios of approximately 1.9 obtained for both samples indi-
cated the abundance of defects in the carbon supports, and
confirmed that phosphorus doping did not appreciably distort
the carbon structure. This defect in the carbon matrix may be
derived from oxygen and nitrogen contained in the chitosan Scheme 3 Reaction pathways accompanying the formation of MF
molecules, as the PXRD pattern of the C sample, which was using traditional copper catalyst (Pathway A) and the Cu /PC catalyst
Pathway B).
1
(
synthesized from the pyrolysis of chitosan dissolved in acetic
acid, also shows a broad Bragg peak (Fig. S9†). This similarity
in carbon structure of these two catalysts despite their signifi-
cantly different activities for MF formation suggests that
carbon does not directly contribute to the catalytic conversion
of FA to MF. The valence state of phosphorus and the nature
of chemical bonding involved were then examined by XPS. The
high-resolution spectra of P 2p (Fig. 7a) can be deconvoluted
into three peaks at 132.9 eV, 133.8 eV, and 135.5 eV which can
for MF formation. By contrast, in this study, the oxophilic P is
3
8,41
used to enhance the adsorption of FA,
which consequently
facilitates the conversion of FA to MF. Thus, MF could also be
formed in a near neutral environment via the FA intermediate.
+
Moreover, limited H concentration in a near-neutral environ-
7
b,12
2
6,34
ment also inhibits the polymerization side reaction.
be assigned to P–C, P–N, P–O respectively.
The dominant
According to the above results, we propose a new pathway for
MF formation by a two-step Hads transfer pathway via the FA
intermediate (Pathway B). The product distribution can be
easily controlled by the applied potential instead of pH. At less
negative potential, FA is formed as a major product, and as the
potential becomes more negative MF becomes the dominant
product. Our results suggest that MF could be produced in a
38 mild near-neutral environment by enhancing the adsorption
of FA and suppressing the HER reaction.
presence of the P–O bond was also confirmed by Fourier trans-
formed infrared (FT-IR) spectroscopy as shown in Fig. 7b.
Compared to the FT-IR spectrum of Cu/C, additional peaks are
−
1
−1
present in the regions around 1050 cm and 1150 cm in
Cu /PC which can be assigned to P–O and PvO stretching,
respectively. PO
has been shown to enhance the selectivity for Pd catalysed
1
3
5
36
x
can act as a Lewis acid in catalysis, which
3
7
ECH reactions and promote the deoxygenation of alcohol.
Furthermore, phosphorus in its high oxidation state is oxophi-
lic, which is highly desirable since it can enhance the adsorp-
tion of alcohol and promote its deoxygenation on the Cu
3
9
38 Conclusions
surface, hence facilitate the conversion of FA to MF.
Similar effects could also be expected in the current catalytic
system.
In summary, we demonstrated a novel sustainable strategy for
the selective production of 2-methylfuran or furfuryl alcohol
by the electrochemical hydrogenation of furfural in a near-
neutral environment using a P doped carbon supported single
atom Cu catalyst. The product selectivity of this reaction can
be conveniently controlled by varying the applied potential.
Furfuryl alcohol and 2-methylfuran were obtained with high F.
E. at the potentials of −0.8 V vs. RHE and −1.0 V vs. RHE,
respectively. Control experiments showed that the high activity
and selectivity can be attributed to the presence of single
copper atoms, which suppress the HER, and the doping effect
of phosphorus, which enhances the adsorption of FA and con-
sequently facilitates the conversion of FA to MF. Based on the
experimental evidence, a two-step Hads transfer route to MF via
a FA intermediate is proposed. Our study points to a new direc-
tion for the electrochemical hydrogenation of furfural in a
near-neutral environment.
The reaction pathways of the ECH of furfural which involve
Hads transfer have already been revealed by other researchers
when Cu nanoparticles and bulk Cu material were used as the
Both MF and FA were produced directly from fur-
fural due to the much weaker adsorption of FA on Cu surface
7
b,8e
catalysts.
4
0
(
pathway A, Scheme 3). Therefore, product selectivity can
also be tuned by varying the Hads coverage. Competition
between the formation of FA and MF in consuming Hads
requires high Hads coverage and hence harsh pH conditions
Experimental
Material and apparatus
3 2
Copper nitrate (Cu(NO ) , 99%), chitosan (from crab shell),
furfural, furfuryl alcohol, 2-methylfuran, and phosphoric acid
were purchased from Sigma-Aldrich. Deionized water from a
Fig. 7 (a) High resolution P 2p XPS spectra, and (b) FT-IR spectra
1
obtained on Cu /PC and Cu/C.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3028–3038 | 3035

View Article Online
Paper
Green Chemistry
MilliQ-MilliRho purification system was used to prepare all Preparation of working electrode
aqueous solutions.
Powder X-ray diffraction (PXRD) data were collected with a
Bruker D8 ADVANCE powder diffractometer (Cu K radiation,
x
To prepare the coating dispersion, 20 mg of the Cu /PC catalyst
and 0.1 mL of Nafion (5%) were dispersed uniformly in a 2 mL
mixed solution of water and ethanol (4 : 1 v/v) for 30 min
under ultrasonication. To fabricate the working electrode for
bulk electrolysis, the coating dispersion was added dropwise
onto a carbon paper (∼1 cm × 1 cm) and placed under an infra-
red lamp to dry. The loading of the catalysts on the carbon
α
λ = 0.15406 nm). Transmission electron microscopic (TEM)
images were taken on a FEI Tecnai G2 T20 TWIN TEM
instrument. EDX mapping was taken on an FEI Tecnai G2
F20 S-TWIN TEM. HAADF-STEM image was recorded on a
double-corrected FEI Titan3 FEG-TEM instrument with an
electron acceleration energy of 300 keV. X-ray photoelectron
spectroscopy (XPS) measurements were conducted on a
Thermo VG scientific NEXA spectrometer with a monochro-
−2
paper was calculated to be approximately 1 mg cm . The
modified glassy carbon (GC) electrodes used for voltammetric
experiments were fabricated in the same way with a catalyst
−2
loading of ∼0.14 mg cm
.
α
matized Al K source (1486.6 eV) at a constant analyzer pass
energy of 25 eV. The accuracy of the binding energy was esti-
mated to be within 0.2 eV. All binding energies were cor-
rected referencing to the C 1s (284.6 eV) peak of the con-
tamination carbon as an internal standard. N2 adsorption–
Electrochemical measurements
All electrochemical experiments were performed at 20 ± 2 °C
using a CHI 760e potentiostat (CHI instrument, Austin, Texas,
USA). Voltammetric experiments were carried out using a three-
electrode setup with a modified glassy carbon electrode (3 mm
diameter, CHI) as the working electrode, a Pt wire as the counter
electrode and an Ag/AgCl (1 M KCl) as the reference electrode. A
scan rate of 50 mV s− was used in all measurements. Prior to
each experiment, the electrode was pre-conditioned at the initial
potential for 10 s. Bulk electrolysis was carried out in a two com-
partment H-cell separated by a glass frit with a piece of modified
carbon paper as the working, a graphite rod as the counter and
Ag/AgCl (1 M KCl) as the reference electrodes. The potential vs.
Ag/AgCl was converted to the RHE scale using the Nernst
equation of the form ERHE = EAg/AgCl + 0.059 pH + 0.235 V. A 0.1
M buffer solution (12 mL) of a certain pH was used in both com-
partments. Before the addition of the substrate into the cathodic
compartment for bulk electrolysis, the electrode was pre-con-
ditioned at the applied potential for 5 min to reduce any higher
oxidation state copper to metallic copper. All solutions were
2
desorption isotherms were surveyed through N adsorption
measurement by using a Micromeritics 3 Flex instrument
under liquid nitrogen conditions. The samples were
degassed at 100 °C for 12 h before testing. The surface area
was determined by the Brunauer–Emmett–Teller (BET)
method. Pore size distribution was determined by the BJH
method from the adsorption branch. Inductively coupled
plasma-mass spectrometric (ICP-MS) measurements were
undertaken on a FLEXAR NEXION 300X ICP-MS spectro-
meter. Fourier-transform infrared spectroscopy (FT-IR) data
were collected on an Agilent Cary 630 FT-IR spectrometer.
Elemental analysis was conducted at The Campbell
Microanalytical Laboratory, University of Otago in New
Zealand. X-ray absorption spectroscopy (XAS) and synchro-
tron PXRD were performed at XAS and PD beamline at the
Australian Synchrotron, respectively. Athena and Artemis soft-
1
4
2
ware ver. 0.9.25 implemented in the Demeter package were
used for the analysis of the obtained XAS spectra. Least-
squares curve-parameter fitting was carried out using the
2
purged with N for at least 10 min before the experiments.
4
2
Analysis of products
ARTEMIS module of the IFEFFIT software packages
to
obtain the quantitative structural parameters around the Liquid products were analyzed with an Agilent 7820A gas chro-
4
3
central atoms. DiffPy-CMI was used for the analysis of syn- matogram equipped with a flame ionization detector (FID) and
chrotron PXRD total scattering data.
2
a DB-Wax column. N was used as the carrier gas. The tempera-
tures of the inlet and FID were set at 200 °C and 230 °C respect-
ively and the temperature program of the oven was set as
Synthesis of Cu /PC, PC and Cu/C
x
In a typical synthesis procedure, 2.5 mg Cu(NO
3
)
2
was dis- follows: maintained at 40 °C for 5 min; increased to 200 °C with
solved in 30 mL of 1.5 M phosphoric acid aqueous solution a ramping rate of 15 °C min− . The products were extracted
and then 0.5 g chitosan was added slowly with vigorous stir- with dichloromethane for gas chromatographic analysis to
ring. After stirring for another 12 h, the obtained solution was avoid the possible damage of column by the electrolyte and
1
dried at 50 °C to give a yellow gel. Cu /PC catalyst was obtained water. In a typical procedure, 500 μL of the electrolyte solution
1
by pyrolyzing the yellow gel under a N
2
atmosphere at 800 °C was collected from the cathode compartment after electrolysis,
−
1
for 2 h at a ramping rate of 3 °C min from room tempera- saturated with NaCl (to enhance the extraction efficiency) and
ture. Cu /PC and Cu /PC were prepared following the same extracted with 500 μL of dichloromethane for three times, and
2
3
procedure as Cu
1
3 2
/PC except the amount of Cu(NO ) was the organic phase was collected. Toluene was then added into
changed to 10 mg and 20 mg respectively. Cu/C was syn- the organic phase as an internal standard and subjected to gas
thesized following the same procedure as Cu /PC except that chromatography analysis. Peaks of products were identified by
1
acetic acid was used instead of phosphoric acid. PC was syn- comparing with the authentic samples. Gas products were ana-
thesized following the same procedure as Cu
salt was added.
1
/PC except no Cu lyzed with an Agilent 7820A gas chromatogram equipped with a
HP-plot molesieve (5 Å) column and a thermal conductivity
3036 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
detector (TCD). Typically, 0.2 mL of gas was taken from the
headspace of the cathode compartment and subjected to gas
chromatography analysis directly. The carbon balance and F.E.
are calculated as the following:
2016, 116, 2275–2306; (c) G. W. Huber, J. N. Chheda,
C. J. Barrett and J. A. Dumesic, Science, 2005, 308, 1446–
1450.
2 (a) L. Vaccaro, F. Valentini, V. Kozell, C. Petrucci,
A. Marrocchi, Y. Gu and D. Gelman, Energy Environ. Sci.,
nFA þ nMF þ nFF
Carbon balance ¼
ꢀ 100%
2
019, 12, 2646–2664; (b) S. Zhou, F. Dai, Y. Chen, C. Dang,
C. Zhang, D. Liu and H. Qi, Green Chem., 2019, 21, 1421–
431.
n′FF
2
nFA ꢀ F
1
F:E:FA ¼
F:E:MF ¼
ꢀ 100%
ꢀ 100%
ꢀ 100%
Q
3
4
N. S. Date, A. M. Hengne, K. W. Huang, R. C. Chikate and
C. V. Rode, Green Chem., 2018, 20, 2027–2037.
(a) S. Kim, E. E. Kwon, Y. T. Kim, S. Jung, H. J. Kim,
G. W. Huber and J. Lee, Green Chem., 2019, 21, 3715–3743;
(b) Y. Yang, L. Chen, Y. Chen, W. Liu, H. Feng, B. Wang,
X. Zhang and M. Wei, Green Chem., 2019, 21, 5253–5262;
4nMF ꢀ F
Q
2nH ꢀ F
2
F:E:H ¼
2
Q
(
c) Y. Deng, R. Gao, L. Lin, T. Liu, X. D. Wen, S. Wang and
where nMF, n , n and nFF are the mole amounts of MF, FA, H2,
FA H2
D. Ma, J. Am. Chem. Soc., 2018, 140, 14481–14489.
(a) L. Jiang, H. Guo, C. Li, P. Zhou and Z. Zhang, Chem.
Sci., 2019, 10, 4458–4468; (b) J. Du, J. Zhang, Y. Sun, W. Jia,
Z. Si, H. Gao, X. Tang, X. Zeng, T. Lei, S. Liu and L. Lin,
J. Catal., 2018, 368, 69–78.
and furfural after the reaction and n′FF is the mole amount of fur-
fural at the beginning of the reaction. F is the Faraday constant.
5
6
Author contributions
S. K. Green, J. Lee, H. J. Kim, G. A. Tompsett, W. B. Kim
and G. W. Huber, Green Chem., 2013, 15, 1869–1879.
P. Z. designed, synthesized and characterized the catalyst, per-
formed the electrochemistry experiments and wrote the manu-
script with input from other authors. Y. C. conducted the
HAADF-STEM experiments. P. L. carried out the XRD and XPS
measurements. X. L. Z. conducted the TEM and EDX experi-
ments. Z.L. Y. conducted the ICP measurements. S.
X. G. helped analyzing the electrochemistry data and revising
the manuscript. Q.F. G. processed and analyzed the XAS and
synchrotron XRD data. B. J. conducted the XAS experiment
and analyzed the XAS data. M. M. and A.L. C. perform the BET
measurements. D.R. T. helped revising the manuscript. J. Z.
conceived the project and supervised the research work.
7 (a) X. H. Chadderdon, D. J. Chadderdon, J. E. Matthiesen,
Y. Qiu, J. M. Carraher, J. P. Tessonnier and W. Li, J. Am.
Chem. Soc., 2017, 139, 14120–14128; (b) S. Jung and
E. J. Biddinger, ACS Sustainable Chem. Eng., 2016, 4, 6500–
6508.
8 (a) S. Jung, A. N. Karaiskakis and E. J. Biddinger, Catal.
Today, 2019, 323, 26–34; (b) Y. Sun and K. Li, Chem. – Eur.
J., 2018, 24, 18258–18270; (c) X. Zhang, M. Han, G. Liu,
G. Wang, Y. Zhang, H. Zhang and H. Zhao, Appl. Catal., B,
2019, 244, 899–908; (d) P. Nilges and U. Schröder, Energy
Environ. Sci., 2013, 6, 2595; (e) L. Liu, H. Liu, W. Huang,
Y. He, W. Zhang, C. Wang and H. Lin, J. Electroanal. Chem.,
2
017, 804, 248–253.
9 B. Zhao, M. Chen, Q. Guo and Y. Fu, Electrochim. Acta,
014, 135, 139–146.
Conflicts of interest
2
There are no conflicts to declare.
1
0 (a) F. Wang, M. Xu, L. Wei, Y. Wei, Y. Hu, W. Fang and
C. G. Zhu, Electrochim. Acta, 2015, 153, 170–174; (b) D. Chu,
Y. Hou, J. He, M. Xu, Y. Wang, S. Wang, J. Wang and
L. Zha, J. Nanopart. Res., 2009, 11, 1805–1809.
Acknowledgements
1
1 J. Carneiro and E. Nikolla, Annu. Rev. Chem. Biomol. Eng.,
The authors thank the Monash–Warwick Alliance for funding
support through the Accelerator Grant and the assistance from
Yvonne Hora at the Monash X-ray Platform. The authors also
acknowledge use of the facilities within the Monash Centre for
Electron Microscopy, the Monash X-ray Platform, the XAS and
PD beamline, Australian Synchrotron, ANSTO and the
Campbell Microanalytical Laboratory, University of Otago, New
Zealand. This research used equipment funded by an
Australian Research Council grant (LE0454166).
2019, 10, 85–104.
12 M. J. Gilkey, P. Panagiotopoulou, A. V. Mironenko,
G. R. Jenness, D. G. Vlachos and B. Xu, ACS Catal., 2015, 5,
3988–3994.
13 N. Singh, Y. Song, O. Y. Gutieŕrez, D. M. Camaioni,
C. T. Campbell and J. A. Lercher, ACS Catal., 2016, 6, 7466–
7470.
1
4 (a) M. Choura, N. M. Belgacem and A. Gandini,
Macromolecules, 1996, 29, 3839–3850; (b) T. Kim,
R. S. Assary, H. Kim, C. L. Marshall, D. J. Gosztola,
L. A. Curtiss and P. C. Stair, Catal. Today, 2013, 205, 60–66.
5 (a) L. Cao, Q. Luo, W. Liu, Y. Lin, X. Liu, Y. Cao, W. Zhang,
Y. Wu, J. Yang, T. Yao and S. Wei, Nat. Catal., 2018, 2, 134;
(b) N. Cheng, S. Stambula, D. Wang, M. N. Banis, J. Liu,
Notes and references
1
1
(a) Z. Zhang and G. W. Huber, Chem. Soc. Rev., 2018, 47,
351–1390; (b) B. M. Upton and A. M. Kasko, Chem. Rev.,
1
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3028–3038 | 3037

View Article Online
Paper
Green Chemistry
A. Riese, B. Xiao, R. Li, T.-K. Sham, L.-M. Liu, G. A. Botton
and X. Sun, Nat. Commun., 2016, 7, 13638; (c) H. Fei,
and T. F. Jaramillo, Proc. Natl. Acad. Sci. U. S. A., 2017, 114,
5918–5923.
J. Dong, M. J. Arellano-Jiménez, G. Ye, N. D. Kim, 26 J. Wu, X. Zheng, C. Jin, J. Tian and R. Yang, Carbon, 2015,
E. L. Samuel, Z. Peng, Z. Zhu, F. Qin and J. Bao, Nat. 92, 327–338.
Commun., 2015, 6, 1–8; (d) L. Fan, P. F. Liu, X. Yan, L. Gu, 27 J. Zhou, X. Shan, J. Ma, Y. Gu, Z. Qian, J. Chen and
Z. Z. Yang, H. G. Yang, S. Qiu and X. Yao, Nat. Commun.,
016, 7, 1–7.
6 (a) Z. Miao, X. Wang, M.-C. Tsai, Q. Jin, J. Liang, F. Ma,
H. Feng, RSC Adv., 2014, 4, 5465–5468.
2
28 B. H. Lee, S. Park, M. Kim, A. K. Sinha, S. C. Lee, E. Jung,
W. J. Chang, K. S. Lee, J. H. Kim, S. P. Cho, H. Kim,
K. T. Nam and T. Hyeon, Nat. Mater., 2019, 18, 620–626.
1
T. Wang, S. Zheng, B.-J. Hwang, Y. Huang, S. Guo and
Q. Li, Adv. Energy Mater., 2018, 8, 1801226; (b) R. Jiang, 29 V. Petkov, Y. Ren, S. Kabekkodu and D. Murphy, Phys.
L. Li, T. Sheng, G. Hu, Y. Chen and L. Wang, J. Am. Chem. Chem. Chem. Phys., 2013, 15, 8544–8554.
Soc., 2018, 140, 11594–11598; (c) Y. Wang, L. Tao, Z. Xiao, 30 K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou,
R. Chen, Z. Jiang and S. Wang, Adv. Funct. Mater., 2018, 28,
705356; (d) C. H. Choi, H.-K. Lim, M. W. Chung, G. Chon,
R. R. Pierotti, J. Rouquerol and T. Siemieniewska, Pure
Appl. Chem., 1985, 54, 2021–2218.
1
N. R. Sahraie, A. Altin, M.-T. Sougrati, L. Stievano, H. S. Oh 31 (a) L. Zhang, W. Liu, Y. Dou, Z. Du and M. Shao, J. Phys.
and E. S. Park, Energy Environ. Sci., 2018, 11, 3176–3182;
e) S. H. Lee, J. Kim, D. Y. Chung, J. M. Yoo, H. S. Lee,
M. J. Kim, B. S. Mun, S. G. Kwon, Y.-E. Sung and T. Hyeon,
Chem. C, 2016, 120, 29047–29053; (b) M. D. Hossain, Z. Liu,
M. Zhuang, X. Yan, G.-L. Xu, C. A. Gadre, A. Tyagi,
I. H. Abidi, C.-J. Sun, H. Wong, A. Guda, Y. Hao, X. Pan,
K. Amine and Z. Luo, Adv. Energy Mater., 2019, 9, 1803689.
(
J. Am. Chem. Soc., 2019, 141, 2035–2045.
1
7 (a) Y. Pan, R. Lin, Y. Chen, S. Liu, W. Zhu, X. Cao, W. Chen, 32 C. J. Bondue and M. T. M. Koper, J. Am. Chem. Soc., 2019,
K. Wu, W. C. Cheong, Y. Wang, L. Zheng, J. Luo, Y. Lin, 141, 12071–12078.
Y. Liu, C. Liu, J. Li, Q. Lu, X. Chen, D. Wang, Q. Peng, 33 A. M. Bond, Broadening Electrochemical Horizons, Oxford
C. Chen and Y. Li, J. Am. Chem. Soc., 2018, 140, 4218–4221; Press, New York, 2002.
b) J. Gu, C.-S. Hsu, L. Bai, H. M. Chen and X. Hu, Science, 34 Y. Li, D. Liu, J. Gan, X. Duan, K. Zang, M. Rønning,
(
2
019, 364, 1091–1094; (c) H. B. Yang, S.-F. Hung, S. Liu,
L. Song, J. Luo and D. Chen, Adv. Mater. Interfaces, 2018, 6,
K. Yuan, S. Miao, L. Zhang, X. Huang, H.-Y. Wang, W. Cai
1801623.
and R. Chen, Nat. Energy, 2018, 3, 140–147; (d) Q. He, 35 G. Bekiaris, C. Peltre, L. S. Jensen and S. Bruun,
D. Liu, J. H. Lee, Y. Liu, Z. Xie, S. Hwang, S. Kattel, L. Song Spectrochim. Acta, Part A, 2016, 168, 29–36.
and J. G. Chen, Angew. Chem., Int. Ed., 2020, 59, 3033–3037; 36 S. Wang, D. Xu, J. Zhao, W. Zheng, C. Hu, X. Wen, Y. Yang
e) Y. Cheng, S. Zhao, B. Johannessen, J. P. Veder, and Y. Li, Catal. Sci. Techol., 2019, 9, 5712–5724.
M. Saunders, M. R. Rowles, M. Cheng, C. Liu, 37 K. Koh, U. Sanyal, M. S. Lee, G. Cheng, M. Song,
(
M. F. Chisholm, R. De Marco, H. M. Cheng, S. Z. Yang and
S. P. Jiang, Adv. Mater., 2018, 30, e1706287.
8 X. Cui, W. Li, P. Ryabchuk, K. Junge and M. Beller, Nat.
Catal., 2018, 1, 385–397.
V. A. Glezakou, Y. Liu, D. Li, R. Rousseau, O. Y. Gutierrez,
A. Karkamkar, M. Derewinski and J. Lercher, Angew. Chem.,
2019, 59, 1501–1505.
1
1
2
2
2
2
2
2
38 S. Chen, R. Wojcieszak, F. Dumeignil, E. Marceau and
S. Royer, Chem. Rev., 2018, 118, 11023–11117.
39 P.-J. Hsu, J.-W. Jiang and Y.-C. Lin, ACS Sustainable Chem.
Eng., 2017, 6, 660–667.
40 (a) S. Sitthisa, T. Sooknoi, Y. Ma, P. B. Balbuena and
D. E. Resasco, J. Catal., 2011, 277, 1–13; (b) Y. Shi, Y. Zhu,
Y. Yang, Y.-W. Li and H. Jiao, ACS Catal., 2015, 5, 4020–
4032.
9 S. Ji, C. Chen, Q. Peng, D. Wang, Y. Chen and Y. Li, Joule,
2
018, 2, 1242–1264.
0 J. Wang, Z. Li, Y. Wu and Y. Li, Adv. Mater., 2018, 30,
801649.
1 C. H. Choi, S. H. Park and S. I. Woo, J. Mater. Chem., 2012,
2, 12107.
1
2
2 L. Zhang, L. Tian, R. Sun, C. Liu, Q. Kou and H. Zuo,
Bioresour. Technol., 2019, 276, 60–64.
3 H. Jin, H. Zhang, H. Zhong and J. Zhang, Energy Environ.
Sci., 2011, 4, 3389–3394.
41 M. E. Witzke, A. Almithn, C. L. Coonrod,
M. D. Triezenberg, D. D. Hibbitts and D. W. Flaherty, J. Am.
Chem. Soc., 2019, 141, 16671–16684.
4 C. Liu, G. Shi, G. Wang, P. Mishra, S. Jia, X. Jiang, P. Zhang, 42 B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12,
Y. Dong and Z. Wang, RSC Adv., 2019, 9, 6898–6906. 537–541.
5 C. Hahn, T. Hatsukade, Y. G. Kim, A. Vailionis, 43 P. Juhas, J. N. Louwen, L. van Eijck, E. T. C. Vogt and
J. H. Baricuatro, D. C. Higgins, S. A. Nitopi, M. P. Soriaga
S. J. L. Billinge, J. Appl. Crystallogr., 2018, 51, 1492–1497.
3038 | Green Chem., 2021, 23, 3028–3038
This journal is © The Royal Society of Chemistry 2021