Efficient and Selective Electrochemical and Photoelectrochemical Reduction of 5-Hydroxymethylfurfural to 2,5-Bis(hydroxymethyl)furan using Water as the Hydrogen Source

Article abstract of DOI:10.1021/acscatal.5b02586

Reductive biomass conversion has been conventionally conducted using H₂ gas under high-temperature and -pressure conditions. In this study, efficient electrochemical reduction of 5-hydroxymethylfurfural (HMF), a key intermediate for biomass conversion, to 2,5-bis(hydroxymethyl)furan (BHMF), an important monomer for industrial processes, was demonstrated using Ag catalytic electrodes. This process uses water as the hydrogen source under ambient conditions and eliminates the need to generate and consume H₂ for hydrogenation, providing a practical and efficient route for BHMF production. By systematic investigation of HMF reduction on the Ag electrode surface, BHMF production was achieved with the Faradaic efficiency and selectivity nearing 100%, and plausible reduction mechanisms were also elucidated. Furthermore, construction of a photoelectrochemical cell (PEC) composed of an n-type BiVO₄ semiconductor anode, which uses photogenerated holes for water oxidation, and a catalytic Ag cathode, which uses photoexcited electrons from BiVO₄ for the reduction of HMF to BHMF, was demonstrated to utilize solar energy to significantly decrease the external voltage necessary for HMF reduction. This shows the possibility of coupling electrochemical HMF reduction and solar energy conversion, which can provide more efficient and environmentally benign routes for reductive biomass conversion.

Full text of DOI:10.1021/acscatal.5b02586

Research Article
pubs.acs.org/acscatalysis
Efficient and Selective Electrochemical and Photoelectrochemical
Reduction of 5‑Hydroxymethylfurfural to 2,5-
Bis(hydroxymethyl)furan using Water as the Hydrogen Source
John J. Roylance, Tae Woo Kim, and Kyoung-Shin Choi*
Department of Chemistry, University of WisconsinMadison, Madison, Wisconsin 53706, United States
*
S Supporting Information
ABSTRACT: Reductive biomass conversion has been con-
ventionally conducted using H gas under high-temperature
2
and -pressure conditions. In this study, efficient electro-
chemical reduction of 5-hydroxymethylfurfural (HMF), a key
intermediate for biomass conversion, to 2,5-bis-
(
hydroxymethyl)furan (BHMF), an important monomer for
industrial processes, was demonstrated using Ag catalytic
electrodes. This process uses water as the hydrogen source
under ambient conditions and eliminates the need to generate
and consume H for hydrogenation, providing a practical and
2
efficient route for BHMF production. By systematic
investigation of HMF reduction on the Ag electrode surface, BHMF production was achieved with the Faradaic efficiency
and selectivity nearing 100%, and plausible reduction mechanisms were also elucidated. Furthermore, construction of a
photoelectrochemical cell (PEC) composed of an n-type BiVO semiconductor anode, which uses photogenerated holes for
4
water oxidation, and a catalytic Ag cathode, which uses photoexcited electrons from BiVO for the reduction of HMF to BHMF,
4
was demonstrated to utilize solar energy to significantly decrease the external voltage necessary for HMF reduction. This shows
the possibility of coupling electrochemical HMF reduction and solar energy conversion, which can provide more efficient and
environmentally benign routes for reductive biomass conversion.
KEYWORDS: biomass conversion, 5-hydroxymethylfurfural, 2,5-bis(hydroxymethyl)furan, electrochemical hydrogenation,
aldehyde reduction, solar biomass conversion, photoelectrochemical cell
1
. INTRODUCTION
Biomass conversion can provide a viable pathway to minimize
our dependence on petroleum products for generating both
fuels and organic chemicals used in various industrial
1
−6
Figure 1. Conversion of HMF to BHMF (a) by conventional
hydrogenation using H as the hydrogen source and (b) by
processes.
In particular, the conversion of cellulosic matter,
7
,8
2
the most abundant organic material on earth, is of great
interest because it can provide numerous chemicals and
+
electrochemical hydrogenation using H from H O as the hydrogen
2
source.
1
−6
materials.
5-Hydroxymethylfurfural (HMF) is one of the
most important intermediates of biomass conversion and is
obtained by the dehydration of hexoses, most predominantly
be produced from other primary sources with an energy input,
9
−12
fructose and glucose.
HMF can undergo oxidation,
enabling reductive biomass conversion without consuming H
would be highly beneficial.
2
20,21
reduction, and etherification to produce a variety of important
2
,3,11−13
chemicals.
Of the many HMF reduction products, 2,5-
Recently, the possibility of electrochemically reducing HMF
+
bis(hydroxymethyl)furan (BHMF), which is produced by
reduction of the formyl group of HMF, is an important starting
molecule for various industrially important polymerization or
etherification processes (e.g., biodiesel, polyester, and polyur-
to BHMF (Figure 1b) and other reduced species utilizing H or
22−24
H
O as the hydrogen source was demonstrated.
2
This
process is very appealing because it does not consume H
.
2
However, in order to make electrochemical biomass conversion
a practical and efficient alternative pathway for reductive
biomass conversion, two major issues need to be addressed.
2
,13−15
ethane foams).
To date, the conversion of HMF to
BHMF has been mainly achieved by using H as the hydrogen
2
source in a pressurized cell (28−350 bar) operating at high
temperatures (403−423 K) using various heterogeneous
Received: November 17, 2015
Revised: January 27, 2016
16−19
catalysts, usually containing precious metals (Figure 1a).
Considering that H itself is a valuable energy source that must
2
©
XXXX American Chemical Society
1840
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
First, since electrochemical HMF reduction in aqueous media
placed in the AgNO solution. Following deposition, the film
3
always competes with water reduction to H , the Faradaic
was rinsed gently with water and dried in an air stream.
Morphologically flat Ag electrodes (hereafter referred to as
Ag ) were prepared by sputter coating. To prepare Ag
2
efficiency (FE) for HMF reduction is expected to be low.
Therefore, the development of catalytic electrodes and
reduction conditions that can significantly increase FE as well
as selectivity is critical. Previous studies did not report FE for
the production of BHMF, and the highest selectivity reported
sp
sp
electrodes, Cu foil cleaned using aforementioned procedures
was placed in a sputter coater (Anatech USA, DC/RF Dual
Source Sputtering System) where 100 nm of Ag was sputter-
coated onto the Cu substrate. The films were then made into
electrodes in the same manner as for the Cu-foil electrodes.
The Pt counter electrodes were prepared by sputter-coating a
100 nm platinum layer over a 20 nm titanium layer onto
cleaned glass slides. The dimension of the working area of the
Pt counter electrode was 2.0 cm × 2.0 cm. The nanoporous
BiVO4 electrode coated with oxygen evolution catalysts
(FeOOH and NiOOH) used in the construction of a PEC
22−24
for BHMF production was 77%.
Second, since the
electrochemical reduction of HMF requires electrical energy
input, a strategy to minimize that input while achieving high FE
and selectivity is essential. Any conditions and strategies that
can address these issues will be advantageous not only for HMF
reduction but also for the reduction of related biomass
intermediates such as furfurals.
In this study, we report electrochemical reduction conditions
using catalytic Ag cathodes that can convert HMF to BHMF
26
was prepared using the procedures reported in a recent paper.
using H O as the hydrogen source with FE and selectivity
2.3. Characterization. Aggd and Agsp electrodes were
characterized by powder X-ray diffraction (XRD) (Bruker D8
Advanced PXRD, 298 K, Ni-filtered Cu Kα radiation, λ =
1.5418 Å). Aside from the underlying copper substrate, only
crystalline Ag peaks were observed for both films (Figure S1a in
the Supporting Information). No other crystalline impurity
2
nearing 100% at ambient pressure and temperature. We also
demonstrate the construction of a photoelectrochemical cell
(PEC) by simply replacing a dark anode with an n-type BiVO4
semiconductor electrode (photoanode) that can directly utilize
solar energy for HMF reduction, significantly decreasing the
external energy input necessary for reduction. Previously, for
cleaner and more sustainable biomass conversion, it was
peaks (e.g., Ag O, AgO) were detected.
2
When XRD patterns of the electrodes were obtained again
after passing 20 C for HMF reduction at −1.3 V vs Ag/AgCl,
no change was observed (Figure S1b in the Supporting
Information). The surface morphologies of the Aggd and Agsp
electrodes were examined by scanning electron microscopy
(SEM) using a LEO 1530 microscope at an accelerating voltage
of 5 kV before and after electrochemical HMF reduction at
envisioned that the H gas required for reductive biomass
2
conversion would be provided from H produced by renewable
2
1
,25
sources such as solar water splitting.
The PEC presented in
this study demonstrates the possibility of directly utilizing solar
energy and water for HMF reduction without involving the
formation and use of H2.
−
1.3 V vs Ag/AgCl. Again, no change was observed (Figure S2
in the Supporting Information).
2
. EXPERIMENTAL SECTION
.1. Materials. Copper foil was purchased from Nimrod
Hall Copper Foil Co. Boric acid (≥98.5%), sodium hydroxide
≥97%), and 5-hydroxymethylfurfural (≥99%) were purchased
2
2.4. Electrochemical HMF Reduction. To investigate the
catalytic abilities of Cu-foil, Aggd, and Agsp electrodes to reduce
HMF, linear sweep voltammetry (LSV) was performed using
solutions with and without HMF. The general experimental
setup consisted of a three-electrode system (Cu or Ag working
electrode, Pt counter electrode, and a Ag/AgCl (4 M KCl)
reference electrode) in an undivided cell controlled by a Bio-
Logic VMP2 potentiostat. A 0.5 M borate buffer solution (pH
9.2) with and without 0.02 M HMF was used as the electrolyte,
(
from Sigma-Aldrich. Silver nitrate (99.9+%) was purchased
from Alfa Aesar. To assist in method development, standard
calibration, and product identification, 2,5-bis(hydroxymethyl)-
furan (97%) was purchased from Polysciences, Inc. All
chemicals were used without further purification.
2.2. Preparation of Electrodes. Cu electrodes used in this
study were prepared by cutting Cu foil into pieces with a
dimension of 1.5 cm × 2.5 cm. Cu tape was attached to the top
of the Cu-foil electrode to enable connection to the
potentiostat lead. The back and top 0.5 cm of the Cu foil
were then covered with Teflon tape to yield a 3.0 cm working
area. High-surface-area Ag electrodes (hereafter referred to as
which was purged with N to remove dissolved oxygen prior to
2
obtaining LSVs. The potential was swept to the negative
direction from the open-circuit potential to −1.5 V vs Ag/AgCl
(4 M KCl) at 5 mV/s without stirring. The results were plotted
against both Ag/AgCl and the reversible hydrogen electrode
(RHE). The latter allows easy interpretation of the data against
2
Ag ) were prepared by galvanic displacement, which involved
the thermodynamic water reduction potential to H under the
2
gd
immersing clean Cu-foil electrodes (1.5 cm × 2.0 cm working
pH conditions used in this study. All current densities reported
in this study were calculated on the basis of the geometric area
of the working electrode.
area) into a 50.0 mM solution of AgNO for 30 s. The
3
concentration of the Ag+ solution and the duration of
deposition reported here are optimum conditions to produce+
uniform and adherent Ag electrodes. Increasing the Ag
concentration, which increased the deposition rate of Ag,
resulted in poor adhesion of Ag on the Cu surface. Decreasing
Constant-potential electrolysis experiments were conducted
to quantify products, selectivities, and FE. A three-electrode
setup was employed using an H-shaped cell divided by a glass
frit. The Cu or Ag working electrode and the reference
electrode were immersed in the cathodic compartment
containing 14 mL of a 0.5 M borate buffer solution (pH 9.2)
with 0.02 M HMF. The counter electrode was immersed in the
anodic compartment containing 14 mL of a 0.5 M borate buffer
solution (pH 9.2) with 1 M sodium sulfite. Because sulfite
oxidation occurs at a potential less positive than that of water
oxidation, using sulfite oxidation as the counter electrode
reaction prevented the drift of the anode potential to the
+
the Ag concentration resulted in nonuniform coverage of Ag
on the Cu surface. In a similar manner, increasing or decreasing
the deposition time affected the adhesion or uniformity of the
Ag deposition. To achieve good coverage and adhesion of Ag
on the Cu, the Cu-foil electrodes were cleaned just prior to
deposition by rinsing with 2-propanol and water followed by
immersing in 1 M HCl for 1 min to remove surface oxides.
They were then rinsed with water, dried, and immediately
1
841
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
potential limit of the potentiostat, which occasionally occurred
when water oxidation was used as the counter electrode
reaction because the divided cell used in this study was not
optimized and caused a significant IR drop during electrolysis.
For a three-electrode setup, the counter electrode reaction does
not affect the current generated at the working electrode.
Therefore, using sulfite oxidation as the counter electrode
reaction allowed us to prevent experimental issues unrelated to
HMF reduction without affecting the results obtained at the
working electrode.
Just prior to use in electrolysis, the working electrode was
reduced in a separate borate buffer solution by performing a
potential sweep from open circuit potential to −1.5 V vs Ag/
AgCl at 5 mV/s to ensure reduction of any surface oxide
species so as to minimize Faradaic loss during HMF reduction.
The designated potential was applied to the working electrode
by a Bio-Logic VMP2 potentiostat for a duration of 20 C, after
which time the solution was collected and analyzed. Mass
transport of the HMF solution during the electrolysis was
facilitated by a magnetic stirrer at approximately 1600 rpm.
needed to be evaluated accurately. The cell reactions are
summarized in the Results and Discussion.
When a two-electrode cell is used, while the potential
difference between the cathode and the anode is fixed at 1.0 V,
the absolute potentials applied to the working electrode and the
counter electrode are not fixed and can be influenced by various
conditions. As discussed below, the results of a three-electrode
potential-dependence study showed that −1.3 V vs Ag/AgCl
(equivalent to −0.56 V vs RHE) is an optimum potential for
the Ag electrode to reduce HMF to BHMF with the highest
gd
selectivity and FE. Therefore, while 1.0 V was applied between
the working electrode and the counter electrode, we optimized
other conditions to ensure that the potential applied to the Aggd
electrode was near −1.3 V vs Ag/AgCl by monitoring the
potential between the Aggd and a Ag/AgCl (4 M KCl) reference
electrode, which was not connected to a potentiostat, using a
multimeter. This included making the size of the BiVO (2
4
2
2
cm ) 4 times larger than the size of the Aggd electrode (0.5 cm )
and stirring only in the BiVO anode compartment. The latter
4
condition facilitated the current generation at the anode, which
in turn resulted in the application of a more negative potential
to the Aggd cathode to generate a matching current. As a result,
a desired potential, ∼−1.3 V vs Ag/AgCl, was achieved at the
Aggd electrode during the constant-potential electrolysis at 1.0
2
.5. Product Analysis. Products were detected and
quantified using a Bruker Avance III 400 MHz nuclear
magnetic resonance (NMR) spectrometer. Calibration curves
were generated for both BHMF and HMF by obtaining NMR
spectra for a series of solutions of known concentration for each
species and plotting the area obtained for signature HMF and
BHMF peaks for each concentration. The area for HMF and
BHMF peaks of unknown concentrations from the product
solution were then plotted against the calibration curves to
determine their concentrations. The HMF peaks in the product
solution were also compared to those of the initial solution to
determine the amount of HMF consumed. Since the samples
were aqueous solutions, the NMR spectra were collected and
V between the Aggd cathode and the BiVO photoanode, which
4
resulted in the production of BHMF with a FE of 94% and a
selectivity of 95%.
3
. RESULTS AND DISCUSSION
When HMF is reduced in an aqueous solution, water reduction
to H is the major competing reaction. Therefore, identifying
2
good HMF reduction catalysts among poor H2 evolution
catalysts can be a good strategy for minimizing Faradaic
efficiency (FE) loss due to water reduction. This is one of the
major reasons lead, mercury, and cadmium electrodes have
often been used for electroorganic synthesis in aqueous
1
processed using a water suppression method to remove the H
signal from water. As a result, signals for peaks near water (4.7
ppm) were significantly decreased. Water suppression did not
affect quantifications of HMF and BHMF. See Figure S3 in the
Supporting Information for NMR spectra of an HMF solution
20,21,27−29
media.
In this study silver and copper electrodes,
which are also known to be poor H evolution catalysts while
2
30
after passing 20 C at −1.3 V vs Ag/AgCl to an Ag electrode,
not being toxic, were investigated for HMF reduction. High-
surface-area Ag cathodes, which were discovered to be
particularly catalytic for BHMF formation in this study, were
prepared by galvanically displacing a piece of commercially
gd
as well as of a fresh HMF solution and a BHMF solution.
2
.6. Photoelectrochemical HMF Reduction. LSVs for
photoelectrochemical HMF reduction were performed in an
undivided cell without stirring using the same solution and
procedures described above. The only difference was that the Pt
available Cu foil with Ag in a 0.05 M AgNO solution for 30 s
3
(referred to as Aggd in this study). The displacement reaction
+
anode was replaced by a BiVO photoanode, and the reference
was kinetically fast and quickly depleted Ag ions at the Cu
4
electrode was removed to form a two-electrode system.
surface. As a result, the growth of Ag was limited by diffusion of
+
Illumination of the BiVO photoanode was achieved through
Ag ions, and Ag electrodes with a dendritic fractal morphology
4
31,32
the FTO substrate (back-side illumination) with an Oriel LCS-
were obtained (Figure 2a).
This morphology is favorable
1
00 solar simulator with the intensity of the incident light
for increasing surface area while maintaining good electrical
2
32
calibrated to be AM1.5G, 100 mW/cm . The result was
compared with those for LSVs obtained using a two-electrode
continuity between Ag crystals.
The catalytic ability of the Ag electrode for HMF reduction
gd
system composed of a Ag working electrode and a Pt counter
as the half-cell reaction was first investigated by linear sweep
voltammetry (LSV) with and without 0.02 M HMF. The LSVs
were measured in an undivided three-electrode cell containing a
0.5 M borate buffer solution (pH 9.2). The LSVs of the Cu foil
used as the substrate for preparing Ag electrodes and a
morphologically flat Ag electrode prepared by sputter-coating
gd
electrode. For this comparison, the sizes of all working and
2
counter electrodes were kept as 1 cm .
Constant-potential electrolysis was performed by applying
1
.0 V between the Ag cathode and the BiVO photoanode (a
gd 4
two-electrode cell) and passing 5 C under illumination (AM
2
1
.5G, 100 mW/cm ) using a divided cell. A 0.5 M borate buffer
Ag onto Cu foil (Figure 2b), referred to as Ag , are also shown
sp
solution (pH 9.2) was used as the electrolyte, and the solution
in the cathodic compartment contained 0.02 M HMF. This
time sulfite was not added to the anodic compartment because
the potential necessary for full cell operation with the presence
of only HMF and water as electrochemically active species
for comparison.
The cathodic currents observed in the solution without HMF
are due to the reduction of water to H (Figure 3a). When 0.02
M HMF was added to the solution, all electrodes showed a shift
of the cathodic current onset potential to the positive direction
2
1
842
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
for the case of HMF reduction (Figure 3b), the Ag electrode
gd
shows a distinctively more positive onset potential in
comparison with the Agsp electrode, indicating that the Aggd
electrode is particularly more catalytic for HMF reduction. The
Aggd electrode, which is composed of dendritic nanocrystals,
likely offers effective binding sites for HMF molecules that can
facilitate interfacial charge transfer processes for HMF
2
reduction. The potentials required to generate 1 mA/cm of
current density when HMF was present were −0.33, −0.44, and
−
0.48 V vs RHE, respectively, for Ag , Ag , and Cu-foil
gd sp
electrodes. We note that the reduction of HMF to BHMF is
not reversible on these electrodes, judging from the cyclic
voltammograms (CVs) obtained using the same conditions
Figure 2. SEM images showing the surface morphologies of (a) Aggd
and (b) Agsp electrodes.
(
Figure S4 in the Supporting Information).
The reduction products and conversion efficiencies for HMF
reduction were analyzed by performing HMF reduction in a
divided cell at a constant potential of −1.3 V vs Ag/AgCl
(
equivalent to −0.56 V vs RHE) by passing 20 C in 14 mL of
0
.5 M borate buffer solution containing 0.02 M HMF. After
1
reduction, the electrolyte was analyzed by H NMR to quantify
the amounts of BHMF and remaining HMF (Figure S3 in the
Supporting Information). On the basis of these results, FE and
selectivity for BHMF formation, which are summarized in
Table 1, were calculated using the equations
mol of BHMF formed
total charge passed/(F × 2)
FE (%) =
× 100%
(1)
selectivity of BHMF (%)
mol of BHMF formed
mol of HMF consumed
=
× 100%
(2)
The FE and selectivity for BHMF formation by the Aggd
electrode were, remarkably, both 99%. This means that
formation of H2 and other HMF reduction products is
negligible. The FE and selectivity for BHMF formation were
7
3
2% and 84%, respectively, for the Ag electrode and 19% and
9%, respectively, for the Cu-foil electrode. The decrease in
sp
selectivity was due to the formation of C−C bonds between
HMF molecules, forming dimeric and polymeric species on the
basis of the NMR data. No other reduction products were
Figure 3. LSVs of Cu (black), Agsp (blue), and Aggd (red) electrodes
in 0.5 M borate buffer (pH 9.2) (a) without and (b) with 0.02 M
HMF (scan rate 5 mV/s).
33
detected. The FEs of the Ag and Cu-foil electrodes, which are
sp
(
Figure 3b). This indicates that HMF reduction is easier than
much lower than their selectivities for BHMF formation,
indicate that part of the cathodic current was used for H2
production.
water reduction on the Ag and Cu electrodes. For the case of
both the Aggd and Ag electrodes, a distinctive cathodic wave
sp
for HMF reduction could be observed. In particular, the Aggd
electrode showed the most significant changes in onset
potential shift and current density for HMF reduction. Figure
After the reduction process, the composition, crystallinity,
and morphology of the Aggd electrode were reexamined. No
sign of instability or deformation of the Ag surface or
composition was observed (Figures S1 and S2 in the
Supporting Information). When the same Aggd electrode was
used for multiple HMF reduction experiments, similar current
density−time (j−t) plots were obtained (Figure S5 in the
3
a shows that the onset potentials of the Aggd and Agsp
electrodes for water reduction are comparable to each other,
although the current of the Aggd electrode after the onset
increases more rapidly due to its higher surface area. However,
a
Table 1. Results Obtained from Electrochemical Reduction of HMF to BHMF by Ag , Ag , and Cu Electrodes
gd
sp
av current density
HMF consumed
(mmol)
BHMF formed
(mmol)
rate for BHMF production
selectivity
(%)
2
2
electrode
(mA/cm )
(mmol/(h cm ))
FE (%)
Aggd
Agsp
Cu
6.90
2.53
1.32
0.104
0.089
0.053
0.102
0.75
0.127
0.034
0.0047
99(2)
72(1)
19(1)
99(3)
84(7)
39(9)
0.020
a
Average values obtained from three measurements are shown. For FE and selectivity, standard deviations are shown in parentheses. Reaction
conditions: E = −1.3 V vs Ag/AgCl (=−0.56 V vs RHE) in a 0.5 M borate buffer (pH 9.2) containing 0.02 M HMF for 20 C passed.
1
843
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
a
Table 2. Results Obtained from Electrochemical Reduction of HMF to BHMF by a Aggd Electrode
potential (V vs
Ag/AgCl)
av current density
HMF consumed
(mmol)
BHMF formed
(mmol)
rate for BHMF production
selectivity
(%)
2
2
(mA/cm )
(mmol/(h/cm ))
FE (%)
−
−
−
−
−
−
−
−
−
−
1.0
1.05
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
0.460
1.14
2.47
3.21
6.90
7.06
7.97
10.2
12.0
18.0
0.059
0.097
0.101
0.101
0.104
0.102
0.100
0.096
0.088
0.059
0.028
0.065
0.082
0.091
0.102
0.101
0.099
0.094
0.086
0.058
0.0023
0.013
0.036
0.052
0.127
0.128
0.14
27(3)
63(5)
79(2)
87(4)
99(2)
97(2)
96(2)
90(3)
87(7)
56(7)
48(5)
68(8)
81(2)
89(7)
99(3)
99(3)
98(3)
97(7)
98(4)
98(2)
0.17
0.19
0.19
a
Average values obtained from three measurements are shown. For FE and selectivity, standard deviations are shown in parentheses. Reaction
conditions: −1.0 V ≥ E ≥ −1.8 V vs Ag/AgCl in a 0.5 M borate buffer (pH 9.2) containing 0.02 M HMF for 20 C passed.
Figure 4. Plausible HMF hydrogenation mechanisms. Blue arrows represent hydrogenation pathways that do not involve the formation of adsorbed
H atom (H ). Red arrows indicate additional pathways available when the formation of H is enabled.
ads
ads
Supporting Information), also confirming the stability of the
Aggd electrode.
The second potential region is −1.1 V ≥ E > −1.6 V vs Ag/
AgCl, where both the FE and selectivity for BHMF formation
increase considerably and become near 100%. This suggests
that the reactions between HMF molecules considerably
diminish in this potential region. LSVs obtained without
We also performed electrochemical HMF reduction using
the Aggd electrode at various applied potentials, which can
provide mechanistic insights for potential-dependent HMF
reduction. The results are summarized in Table 2. On the basis
of the FEs and selectivities of BHMF, the reduction potential
could be divided into three regions. The first region is −1.0 V ≥
E > −1.1 V vs Ag/AgCl. In this region, the amount of BHMF
formed was much less than the amount of HMF consumed,
meaning that there is a competing reaction consuming HMF,
which lowers both the FE and selectivity for BHMF formation.
The major competing reaction was the reaction between HMF
molecules, which is the same competing reaction observed for
the Cu and Agsp electrodes at −1.3 V vs Ag/AgCl discussed
above. The LSVs obtained without HMF (Figure 3a) show that
H evolution on the Ag electrode initiates at −1.1 V vs Ag/
HMF (Figure 3a) indicate that H evolution and, therefore, the
2
formation of Hads are possible on the Aggd electrode surface in
this region. This means new mechanisms that involve the
formation of Hads for BHMF production are possible in this
potential region, which are indicated with red arrows in Figure
4
. As the coverage of Hads increases on the electrode surface
and the rate of BHMF formation increases, the probability of
HMF molecules adsorbing in close enough proximity on the
electrode surface decreases, and the HMF-HMF reactions can
be effectively suppressed.
The third potential region is E ≤ −1.6 V vs Ag/AgCl, where
H evolution competes with BHMF formation, lowering the FE
2
2
gd
for BHMF formation. However, since there is no competing
reaction consuming HMF in this region, the amount of HMF
consumed is lowered accordingly. As a result, the selectivity for
BHMF remains high in this region. In addition, since the
current density in this high overpotential region is significantly
increased, the conversion rates for BHMF are higher than those
in the low-overpotential regions (Table 2).
We also monitored the change in FE and selectivity during
electrochemical reduction using 0.02 and 0.10 M HMF
solutions buffered with borate (pH 9.2) to examine the effect
of HMF concentration on the FE and selectivity (Tables S1 and
S2 in the Supporting Information). For the case of 0.02 M
solution, the FE and selectivity remained above 89% and 99%,
respectively, during the complete conversion of HMF to
AgCl. Therefore, in this potential region, the reduction
mechanism of HMF most likely does not involve the formation
of surface-adsorbed hydrogen atoms (H ), which is the first
ads
34
step for H evolution. Then, HMF reduction proceeds either
through e transfer to an HMF molecule to form an anionic
intermediate that further reacts with H or through
simultaneous e transfer and H transfer to HMF. Plausible
mechanisms describing these pathways are shown in Figure 4
2
−
+
−
+
with blue arrows. In this case, the surface of the Ag electrode
gd
is expected to be covered only with HMF or its reduction
intermediates. As a result, reactions that form C−C bonds
between HMF molecules and/or their reduction intermediates
can be possible competing reactions in this potential region.
1
844
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
BHMF. For the case of 0.10 M HMF solution, both the FE and
selectivity were lower than those obtained for the 0.02 M HMF
solution because the higher concentration of HMF resulted in
enhanced adsorption of HMF on the Ag surface, which
facilitated the undesired dimerization and polymerization
reactions of HMF molecules.
After evaluating the Aggd electrode as an optimum catalytic
cathode for HMF reduction, a PEC was constructed to
demonstrate how simply replacing a metal anode in an
electrochemical cell with a semiconductor anode (photoanode)
can significantly decrease the external energy input necessary
for HMF reduction (Figure 5a) by utilizing solar energy. For
than water splitting (ΔG° = 237.14 kJ/mol). This means that
direct electrochemical reduction of HMF using water as the
hydrogen source is more energy efficient than the two-step
procedure composed of water reduction to H , followed by the
2
use of H for hydrogenation.
2
+
−
cathode reaction: HMF + 2H + 2e → BHMF
(3)
+
−
anode reaction: H O → 1/2O + 2H + 2e
(4)
(5)
2
2
overall: HMF + H O → BHMF + 1/2O
2
2
Since the photoexcited electrons in the CB of BiVO already
4
26
possess a significantly negative potential (∼0.0 vs RHE), the
additional potential necessary to reduce HMF to BHMF can be
decreased considerably (Figure 5a). Also, since the valence
band (VB) edge of BiVO is located at ∼2.4 V vs RHE, the
4
photogenerated holes in the VB of BiVO have sufficient
4
overpotential for water oxidation before applying any external
bias, which decreases the external bias necessary for the cell
operation. This is quite different from the case of an
electrochemical cell, where sufficient overpotentials for both
the cathode reaction (HMF reduction) and for the anode
reaction (water oxidation) need to be provided by the external
bias (Figure 5b).
The advantage of using solar energy in decreasing external
bias to reduce HMF is well demonstrated in Figure 5c,
comparing an LSV obtained from a two-electrode cell
composed of a Ag cathode and a Pt anode with that obtained
gd
from a PEC composed of a Aggd cathode and a BiVO /OEC
4
photoanode under AM 1.5G illumination. For this experiment,
the use of a two-electrode system was critical because the goal
was to evaluate the electrical energy input (i.e., applied bias
between the working and the counter electrode) necessary for
HMF reduction. LSVs obtained using a three-electrode cell
show the change in current density when the potential between
the working electrode and the reference electrode is varied, as
shown in Figure 3. Experiments using a three-electrode cell are
important for analyzing the half-cell reaction occurring at the
working electrode at a precisely controlled potential against the
reference electrode. However, these measurements do not
provide any information on the potential applied between the
working electrode and the counter electrode and, therefore,
cannot be used for evaluating the decrease in external energy
input owing to the utilization of solar energy.
Figure 5. Schematic diagrams comparing the external bias necessary to
achieve HMF reduction for (a) a photoelectrochemical cell and (b) an
electrochemical cell. (c) LSVs obtained from a photoelectrochemical
cell (red) and an electrochemical cell (black) for HMF reduction using
a two-electrode cell (scan rate 5 mV/s, illumination for photo-
2
electrochemical cell AM 1.5G, 100 mW/cm ). A 0.5 M borate buffer
solution (pH 9.2) containing 0.02 M HMF was used as the electrolyte.
the purpose of demonstrating this concept, we chose to use n-
type BiVO as a photoanode, which has been recently identified
The electrolyte was a 0.5 M borate buffer solution (pH 9.2)
containing 0.02 M HMF. In order to achieve a current density
4
as one of the most efficient and practical oxide-based
26,35
2
photoanodes for use in solar water splitting.
As discussed
of 1 mA/cm in an electrochemical cell, application of 2.76 V
below, it has favorable conduction band and valence band edge
positions suitable to utilize photogenerated charge carriers for
HMF reduction and water oxidation while being stable at pH
between the Aggd cathode and the Pt anode was necessary.
However, when a PEC was used, 1 mA/cm was achieved at a
potential of 0.92 V between the Aggd cathode and the BiVO4/
OEC anode, saving about 2 V. Results and discussion of a two-
2
9.2 used for HMF reduction in this study.
Under illumination, electron−hole pairs were generated and
electrode cell where a BiVO /OEC electrode was used as the
4
separated in BiVO . The holes were used at the BiVO surface
anode both for an electrochemical cell in the dark and for a
photoelectrochemical cell (e.g., LSV comparisons with and
without HMF as well as with and without illumination) can be
found in Figure S6 in the Supporting Information.
4
4
to oxidize water to O , while the photoexcited electrons in the
2
conduction band (CB) of BiVO were transferred to the Ag
4
cathode to reduce HMF to BHMF (Figure 5a). Because the
bare surface of BiVO is poorly catalytic for water oxidation, the
The reduction products, FE, and selectivity for photo-
electrochemical reduction of HMF were analyzed after
performing photoelectrochemical reduction at 1.0 V between
the BiVO /OEC photoanode and the Ag cathode (a two-
4
surface of the BiVO electrode was coated with layers of oxygen
4
evolution catalysts (OECs) (i.e., FeOOH and NiOOH) to
26
facilitate water oxidation to O2. The cell reactions are
4
gd
summarized in eqs 3−5). The standard free energy change
electrode cell). It should be noted that this potential is not
sufficient for electrochemical reduction of HMF using a two-
electrode cell according to the LSVs shown in Figure 5c and
(
ΔG°) of the overall reaction (eq 5) at 298 K is calculated to be
36
2
11.20 kJ/mol, which is thermodynamically more favorable
1
845
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
Figure S6 in the Supporting Information. The photo-
electrochemical reduction results show that BHMF was
produced with a FE of 94% and a selectivity of 95%, confirming
that the results obtained from electrochemical HMF reduction
can be reproduced by a PEC while decreasing the necessary
electrical energy input. When better photoelectrodes (photo-
anode and/or photocathode) are developed with a smaller
band gap and a more negative CB position and the cell design is
improved to minimize the IR drop within the cell, photo-
electrochemical HMF reduction can be achieved with no
external energy input.
In this study we demonstrated the use of a photoanode to
construct a PEC. However, it will also be possible to construct a
PEC by replacing the Ag cathode with a p-type semiconductor
electrode (photocathode) when efficient and inexpensive p-
type semiconductors that possess a proper CB position and
stability around pH 9.2 are developed. In this case, the surface
of photocathodes can be decorated with Ag catalysts to achieve
BHMF production with high efficiency and selectivity.
Number DE-SC0008707. We thank Professor Ronald T. Raines
for drawing our attention to the possibility of employing
electrochemistry for biomass conversion.
REFERENCES
■
(
4
1) Huber, G. W.; Iborra, S.; Corma, A. Chem. Rev. 2006, 106, 4044−
098.
(
2) van Putten, R.-J.; van der Waal, J. C.; de Jong, E.; Rasrendra, C.
B.; Heeres, H. J.; de Vries, J. G. Chem. Rev. 2013, 113, 1499−1597.
3) Rosatella, A. A.; Simeonov, S. P.; Frade, R. F. M.; Alfonso, C. A.
(
M. Green Chem. 2011, 13, 754−793.
(
(
4) Bozell, J. J.; Petersen, G. R. Green Chem. 2010, 12, 539−554.
5) Bond, J. Q.; Upadhye, A. A.; Olcay, H.; Tompsett, G. A.; Jae, J.;
Xing, R.; Alonso, D. M.; Wang, D.; Zhang, T.; Kumar, R.; Foster, A.;
Sen, S. M.; Maravelias, C. T.; Malina, R.; Barrett, S. R. H.; Lobo, R.;
Wyman, C. E.; Dumesic, J. A.; Huber, G. W. Energy Environ. Sci. 2014,
7, 1500−1523.
(6) Huber, G. W.; Dumesic, J. A. Catal. Today 2006, 111, 119−132.
(
(
7) O’Sullivan, A. C. Cellulose 1997, 4, 173−207.
8) Klemm, D.; Heublein, B.; Fink, H.-P.; Bohn, A. Angew. Chem., Int.
Ed. 2005, 44, 3358−3393.
4
. CONCLUSIONS
(
(
(
9) Lewkowski, J. Arkivoc 2001, i, 17−54.
In summary, we identified electrochemical reduction conditions
that can achieve selective HMF reduction to BHMF with the
FE nearing 100% using Ag catalytic electrodes. The Aggd
electrode showed a wide potential region where hydrogenation
of HMF can occur without inducing H2 evolution. By
systematic and thorough analysis of potential-dependent FE
and selectivity for BHMF production, optimum reduction
conditions were identified and plausible hydrogenation
mechanisms on the Ag surface were elucidated. We also
demonstrated the possibility of developing a PEC for HMF
reduction by simply replacing a Pt anode with a BiVO4
photoanode as a practical and inexpensive means to decrease
the external voltage necessary for HMF reduction, which
provided one example of coupling solar energy conversion and
biomass conversion.
10) Moye, C. J. Pure Appl. Chem. 1964, 14, 161−170.
11) Binder, J. B.; Raines, R. T. J. Am. Chem. Soc. 2009, 131, 1979−
1985.
(12) Gallo, J. M. R.; Alonso, D. M.; Mellmer, M. A.; Dumesic, J. A.
Green Chem. 2013, 15, 85−90.
(13) Balakrishnan, M.; Sacia, E. R.; Bell, A. T. Green Chem. 2012, 14,
1626−1634.
(14) Cottier, L.; Descotes, G.; Soro, Y. Synth. Commun. 2003, 33,
4
(
6
(
1
(
2
(
285−4295.
15) Goswami, S.; Dey, S.; Jana, S. Tetrahedron 2008, 64, 6358−
363.
16) Schiavo, V.; Descotes, G.; Mentech, J. Bull. Soc. Chim. Fr. 1991,
28, 704−711.
17) Faury, A.; Gaset, A.; Gorrichon, J. P. Informations chimie 1981,
14, 203−209.
18) Alamillo, R.; Tucker, M.; Chia, M.; Pagan-Torres, Y.; Dumesic,
The electrochemical and photoelectrochemical processes
reported in this study offer practical and efficient routes for
reductive biomass conversion, as they utilize water as the
hydrogen source for hydrogenation without needing preformed
J. Green Chem. 2012, 14, 1413−1419.
(19) Ohyama, J.; Esaki, A.; Yamamoto, Y.; Arai, S.; Satsuma, A. RSC
Adv. 2013, 3, 1033−1036.
(20) Albert, W. C.; Lowy, A. Trans. Electrochem. Soc. 1939, 75, 367−
375.
H , eliminating multiple issues regarding the production,
2
(
21) Ellis, K. G.; ul-Islam, N.; Soper, D. W.; Utley, J. H. P.; Chum, H.
L.; Ratcliff, M. J. Electrochem. Soc. 1987, 134, 3058−3062.
22) Kwon, Y.; de Jong, E.; Raoufmoghaddam, S.; Koper, M. T. M.
storage, and use of H . Similar approaches may be developed
2
for other reductive biomass conversion processes.
(
ChemSusChem 2013, 6, 1659−1667.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b02586.
■
(
23) Nilges, P.; Schro
̈
der, U. Energy Environ. Sci. 2013, 6, 2925−2931.
*
S
(24) Kwon, Y.; Birdja, Y. Y.; Raoufmoghaddam, S.; Koper, M. T. M.
ChemSusChem 2015, 8, 1745−1751.
(25) Lewis, N. S. Electrochem. Soc. Interface 2013, 22, 43−49.
(26) Kim, T. W.; Choi, K.-S. Science 2014, 343, 990−994.
(27) Swann, S., Jr. Ind. Eng. Chem. 1937, 29, 1339−1341.
(28) Couper, A. M.; Pletcher, D.; Walsh, F. C. Chem. Rev. 1990, 90,
XRD patterns, SEM images, CVs, NMR spectra, LSVs,
and electrochemical test results (PDF)
8
37−865.
AUTHOR INFORMATION
■
(29) Goodridge, F.; Lister, K.; Plimley, R. E.; Scott, K. J. Appl.
Corresponding Author
Electrochem. 1980, 10, 55−60.
*E-mail for K.-S.C.: kschoi@chem.wisc.edu.
(30) Trasatti, S. J. Electroanal. Chem. Interfacial Electrochem. 1972, 39,
1
(
63−184.
Notes
31) Meakin, P. In Fractals, Scaling and Growth Far from Equilibrium;
The authors declare no competing financial interest.
Chirikov, B., Cvitanovic, P., Moss, F., Swinney, H., Eds.; Cambridge
University Press: Cambridge, U.K., 1998; pp 326−400.
ACKNOWLEDGMENTS
■
(
(
32) Lopez, C. M.; Choi, K.-S. Langmuir 2006, 22, 10625−10629.
33) Huang, Y.; Hu, S.; Zuo, S.; Xu, Z.; Han, C.; Shen, J. J. Mater.
This material is based upon work supported by University of
WisconsinMadison and the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, Division of
Chemical Sciences, Geosciences, and Biosciences, under Award
Chem. 2009, 19, 7759−7764.
(34) Conway, B. E.; Salomon, M. Electrochim. Acta 1964, 9, 1599−
1615.
1
846
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847

ACS Catalysis
Research Article
(
35) Park, Y.; McDonald, K. J.; Choi, K.-S. Chem. Soc. Rev. 2013, 42,
321−2337.
36) Assary, R. S.; Redfern, P. C.; Hammond, J. R.; Greeley, J.;
Curtiss, L. A. Chem. Phys. Lett. 2010, 497, 123−128.
2
(
1
847
DOI: 10.1021/acscatal.5b02586
ACS Catal. 2016, 6, 1840−1847