Journal of the American Chemical Society
Article
spectroscopic studies of the DHPS-mediated HER reaction
unveil a distinct hydrogen formation mechanism involving site-
specific dehydrogenation reactions of DHPS-2H on Pt,
accounting for the fast hydrogen generation as compared to
the conventional electrocatalytic HER reaction. With these
encouraging results, we anticipate that the redox-mediated
approach would provide an intriguing alternative for hydrogen
production and should adequately engineer studies especially
those on the reactor design to be conducted in future.
frequency of 9.84 GHz (X-band) and a 2 mW microwave power with
a 0.2 G modulation amplitude at 100 kHz. The three EPR samples
were collected from the fully charged negolyte (DHPS-2H) of the
flow cell, and two of them were subsequently added with different
amounts of commercial Pt/C catalyst (1 and 10 mg) prior to the test.
The solutions were sealed in high-purity quartz capillary tubes with
vacuum grease at two ends of the tubes. Continuous small bubbles
were generated in the quartz capillary tubes of samples added with Pt/
C catalyst during the EPR test.
Computational Details. The spin-polarized DFT computations
were performed using the VASPsol, a software package that
incorporates solvation into the Vienna Ab-initio Simulation Package
EXPERIMENTAL SECTION
■
(
VASP) within a self-consistent continuum model. Core electrons
Electrochemical Measurements of Electrocatalysts. Cyclic
voltammetry (CV) and linear scan voltammetry (LSV) were
conducted on an Autolab electrochemical workstation (Metrohm,
PSTA30). A three-electrode cell system was used to test the CV of
were treated using a projector augmented wave (PAW) method with
generalized gradient approximation (GGA) designed by Perdew,
Burke, and Ernzerhof (PBE) with a plane wave kinetic energy cutoff
of 350 eV. The atomic positions were fully relaxed with an energy
4
−/3−
DHPS and [Fe(CN)6]
the counter and reference electrodes, respectively. The electrolyte was
M NaOH aqueous solution bubbled with N for 1 h prior to use.
. Graphite rod and Hg/HgO were used as
−4
convergence of 10 eV per cell between two consecutive self-
−2
−1
consistent steps and a force convergence of 10 eV Å to obtain the
equilibrium configuration. To guarantee that the supercell size of
surface model was sufficiently large enough to avoid the effect of
DHPS between their mirror images, the calculations were performed
with a 3 × 3 Pt (111) surface with four layers. A k-mesh grid of 2 × 2
4
2
For the CV test of DHPS, the cells were sealed with N protection.
2
The potentials were presented versus the reversible hydrogen
electrode on the basis of the following equation:
×
1 was set to sample the Brillouin zone during the structural
E (V vs RHE) = EHg/HgO + 0.098V + 0.0591V × pH
optimization. The vacuum layer thickness was adopted to be more
than 20 Å in order to avoid the artificial interaction between the layers
and their periodic images. The zero-point vibration energy (ZPE)
corrections were made over the frequencies at Γ point, on the basis of
the harmonic approximation on the most energetically favorable
structure. The energy of DHPS at different intermediate states was
calculated by the Gaussian 09 program suite, with B3LYP hybrid
exchange-correlation functional. A solvation model based on density
The LSV curves in Figure 2c were carried out by using a H-cell
system. Both the working and reference electrodes were placed on
one side and sealed with N protection, while the counter electrode
2
(
graphite rod) was placed on the other side without sealing. The
membrane was a Sustainion X37-50 membrane. The structure of the
double-layer electrode (DLE) is shown in the inset of Figure 2c,d.
Considering that the redox mediator should diffuse fast from one side
to the other, thin carbon cloth was used as the working electrode here.
(SMD) was used with water as a solvent.
2
A piece of carbon cloth (CC, 1.0 × 1.0 cm ), glass fiber separator
2
2
(
GF, 1.2 × 1.2 cm ), and catalyst@Ni foam (1.0 × 1.0 cm ) were
ASSOCIATED CONTENT
sı Supporting Information
■
clamped together. All the LSV measurements in Figure 2c,d were
carried out at a scan rate of 0.1 mV/s.
*
Electrolytic Flow Cell Test and Hydrogen Evolution
Reaction. The cell was assembled by sandwiching two pieces of
carbon felt as the cathode and anode. The active area of the electrode
2
Discussions of materials, syntheses, and characterization
methods, cyclic voltammetry tests of redox mediators,
hydrogen and oxygen production during decoupled
water splitting, comparison of two different anion-
exchange membranes, operando UV−vis spectroscopic
measurement, comparison of decoupled and direct water
splitting, DFT calculations of DHPS-mediated HER
reaction, EPR measurement of DHPS-2H upon
dehydrogenation, NMR measurement of DHPS-2H
upon dehydrogenation, and comparison of different
was 5 cm . Each half-cell had a graphite plate as the current collector
connected to the external electrical circuit. An anion-exchange
membrane (Sustainion X37-50 membrane) was used as the separator.
The posolyte consisted of 50 mL of 0.6 M K Fe(CN) in 4 M NaOH,
3
6
while the negolyte consisted of 12 mL of 0.6 M DHPS in 4 M NaOH.
The electrolytes were circulated through the cell stack and tanks using
peristaltic pumps. The Pt−Ni(OH) catalyst was loaded in the
2
cathodic tank, while the NiFe(OH) catalyst was in the anodic tank.
2
2
2
Pt−Ni(OH) catalyst (10 cm ) and 15 cm NiFe(OH) catalyst were
used in the test (Figure 4a,c). The voltage profiles of the flow cell
2
2
were recorded in galvanostatic mode with an Arbin battery tester. The
1
13
negolyte was purged with N before charging. Gases that are produced
decoupled water splitting methods, figures of H and C
NMR spectra, peak current vs. the square root of scan
rate, CV curves, SEM and TEM images, EDS mapping
and EDX spectrum, XPS spectra, LSV curves, voltage
profiles, GITT curve, EIS analysis, configuration of the
spectroelectrochemical setup for operando UV−vis
spectroscopic measurements, absorbance changes, com-
parison of overall energy efficiency, H2 yield, and
collection rate, possible reaction pathways, models of
the adsorption configuration, reactant, intermediate, and
product structures, initial state and final state structures,
and EPR experimental and simulated spectra, and tables
of values used for the free energy, entropy, zero-point
energy corrections, and Gibbs reaction free energy and
comparison of the spatially decoupled water electrolyzer
reported here with other decoupled water splitting
systems reported in the literature (PDF)
2
in the tank were collected by water displacement. A measuring
cylinder filled with water was placed upside-down in a water bath. The
gas produced in the tank was fed into the water-filled measuring
cylinder through a silicone tube. The gas production was then
determined by the volume of displaced water.
Gas Chromatography Headspace Measurement. GC head-
space analysis was performed using a Shimadzu GC-2010 Plus system
by direct autoinjection of gas from the headspace of the negolyte
holding into the GC through a silicone tube. For the decoupled water
splitting system, the flow battery and electrolyte were prepared as
described above. For the conventional water splitting system, the cell
was assembled by sandwiching NiFe(OH)2 and Pt−Ni(OH) as
2
anode and cathode, respectively. Fifty milliliters of 4 M NaOH and 12
mL of 4 M NaOH were used as the posolyte and negolyte,
respectively. Each negolyte was purged with N before charging. The
2
2
electrolysis process was operated at a constant current of 40 mA/cm .
EPR Test. The EPR spectra were obtained with an EMX-plus 10/
2 spectrometer from Bruker at room temperature operating at a
1
G
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX