Engineering Bipolar Interfaces for Water Electrolysis Using Earth-Abundant Anodes

Developing efficient and low-cost water electrolyzers for clean hydrogen production to reduce the carbon footprint of traditional hard-to-decarbonize sectors is a grand challenge toward tackling climate change. Bipolar-based water electrolysis combines the benefits of kinetically more favorable half-reactions and relatively inexpensive cell components compared to incumbent technologies, yet it has been shown to have limited performance. Here, we develop and test a bipolar-interface water electrolyzer (BPIWE) by combining an alkaline anode porous transport electrode with an acidic catalyst-coated membrane. The role of TiO2 as a water dissociation (WD) catalyst is investigated at three representative loadings, which indicates the importance of balancing ionic conductivity and WD activity derived from the electric field for optimal TiO2 loading. The optimized BPIWE exhibits negligible performance degradation up to 500 h at 400 mA cm–2 fed with pure water using earth-abundant anode materials. Our experimental findings provide insights into designing bipolar-based electrochemical devices.


Materials characterization
The surface morphology of the PTEs and MEAs were characterized using scanning electron microscopy (FEI Quanta FEG 250).Freshly prepared samples were placed into the specimen chamber under high vacuum conditions (< 2 × 10 -5 Torr).The energy level of the beam was set to 10 kV.To obtain cross sectional images, samples were fractured under liquid nitrogen, and were carefully mounted onto the substrate.The X-ray diffraction (XRD) of the cobalt oxide and titanium dioxide was analyzed using a Rigaku Smartlab X-ray diffractometer equipped with a HyPix-3000 high energy resolution 2D multidimensional semiconductor detector.The powder samples were placed on the sample holder.The XRD measurements were performed by setting the same Rigaku SmartLab diffractometer to Bragg-Brentano mode at room temperature.The surface compositions of cobalt oxide and titanium dioxide were characterized using X-ray photoelectron spectroscopy (XPS) Kratos Axis Ultra DLD system at takeoff angles of 0° and 60° relative to the surface normal.
The measurements were conducted under room temperature and ultrahigh vacuum of 7.5 × 10 -9 Torr.A monochromatic Al Kα source (hν = 1486.6eV) was used to excite the core level electrons of the material.Spectral analysis was performed using CasaXPS analysis software.Transmission electron microscopy (TEM) imaging was performed on a JEOL-2100 electron microscope (JOEL Ltd.).Images were collected with an accelerating voltage of 200 kV and a beam current of ~16 pA.

Anode porous-transport-electrode fabrication
For anode inks, 100 mg of the cobalt oxide catalyst was added to 10 mL of a 1:9 v/v water/nPA mixture and sonicated for 10 min using a horn sonicator (CEX500, Cole-Parmer) at a power of 38%.Afterwards, 600 mg of 5 wt% PiperION A dispersion (I:Cat = 0.20) was added to the ink, and bath sonicated in an ice bath for at least 30 min.The porous transport electrodes were fabricated using a handheld airbrush (Iwata).The bare PTLs were cut to a precise size with an area of 30.25 cm 2 and massed before spraying.The inks were loaded into a hand-held airbrush connected to house nitrogen and deposited on the substrate over many passes until the desired loading was reached.Figure S5 shows SEM images of the PTE catalyst layer.

Fabrication of cathode catalyst layer and water dissociation catalyst layer
For the fabrication of cathode catalyst layer and water dissociation catalyst layer, ultrasonic spray coating method was used.The cathode catalyst ink was deposited onto the Nafion 212 membrane to fabricate a half-CCM, and water dissociation catalyst ink was deposited onto the other side of the Nafion 212 membrane to fabricate the water dissociation catalyst layer.The ink for the cathode side was prepared by mixing platinum supported by carbon (TEC10V50E 46.8% Pt, Tanaka) with DI water and n-propanol at a volumetric ratio of 1:1 and adding PFSA ionomer solution (Nafion 5 wt%, Ion Power D521) until ionomer to catalyst ratio of 0.45 was achieved.The ink mixture was bath sonicated for 30 min at 10°C.Immediately after cathode ink was prepared, the Sono-Tek spray coater at 120 kHz sonication rate was used to deposit cathode catalyst onto the membrane over the vacuum plate at 80°C.The platinum loading on the cathode side was measured to be 0.1 mg Pt /cm 2 ) using X-ray fluorescence.
Water dissociation catalyst inks were prepared by mixing commercially titanium dioxide catalysts with Milli-Q DI water, ethanol, and n-propanol at a volumetric ratio of 1:1:2.The catalyst ink was sonicated using a horn sonicator (CEX500, Cole-Parmer) at a power of 38% for 30 min under an ice bath.The Sono-Tek ultrasonic spray coater was used for deposition of the water dissociation catalyst ink with sonicating nozzle set to 120 kHz.The mass of titanium dioxide powders was reduced by a factor of 10 to fabricate the low-loaded water dissociation catalyst layer with other ink composition unchanged.

Electrolyzer-cell assembly and testing
A single electrolyzer-cell hardware (Fuel Cell Technology, FCT) with a platinum-coated titanium single parallel-channel flow field on the anode, and a single serpentine-channel graphite flow field on the cathode was used for all the experiments.On the cathode side, carbon paper without microporous layer (Toray 120) with 5% PTFE was used as the gas diffusion layer.20% compression of the cathode GDL was achieved by controlling the ethylene tetrafluoroethylene gaskets while anode gasket thickness was matched to the PTE thickness.Electrolyzer cells were torqued uniformly up to 4.5 Nm.The active areas were designed to be 5 cm 2 for all tests.
A potentiostat (VSP 300, Biologic) equipped with a 20 A booster was used for electrochemical analyses.In-house modified FCT test station was used for conducting electrolyzer testing.Milli-Q DI water was fed into the anode at 80˚C while cathode inlet was capped and produced hydrogen was vented through cathode outlet.The anode water was recirculated for the duration of the experiment.A rod heater was used to maintain the electrolyzer cell at 80°C.Polarization curve was measured at various currents with 20 s holds each current until cell voltage reached 2.5 V followed by measurement of galvanostatic EIS at each current step measured from polarization curves between 1 MHz and 100 mHz.The amplitude of the alternating current was optimized for each step to ensure a sufficient signal to noise ratio while maintaining a linear system response.

Figure S1
. Reproducibility of the bipolar interface water electrolyzer (BPIWE) performance from three independent experiments.Experiential conditions for Run 1 and Run 2: Anode: 1.3±0.1 mg Co3O4 cm -2 , Cathode: 0.1 mg Pt cm -2 (Pt/C).Proton exchange membrane: Nafion 212 and water dissociation catalyst: TiO 2 (6 µg cm -2 ).Run 3 was conducted with an anode at 2.0±0.1 mg Co3O4 cm -2 with other conditions the same as Run 1 and Run 2. All tests were done with DI water fed only to anode side at 80 °C using 5 cm 2 single cell device.

Figure S5 .
Figure S5.Scanning electron microscopy (SEM) images of the cobalt oxide porous transport electrode (PTE) catalyst layer.

Table S1 .
Summary of BPM/BPI water electrolyzer performance comparison.