Promoting in-situ stability of hydroxide exchange membranes by thermally conductive network for durable water electrolysis

Abstract

Hydroxide exchange membrane (HEM) water electrolysis is promising for green hydrogen production due to its low cost and excellent performance. However, HEM often has insufficient stability in strong alkaline solutions, particularly under in-situ electrolysis operation conditions, hindering its commercialization. In this study, we discover that the in-situ stability of HEM is primarily impaired by the locally accumulated heat in HEM due to its low thermal conductivity. Accordingly, we propose highly thermally conductive HEMs with an efficient three-dimensional (3D) thermal diffusion network to promote the in-situ stability of HEM for water electrolysis. Based on the 3D heat conductive network, the thermal conductivity of polymeric HEM is boosted by 32 times and thereby reduce the HEM temperature by up to 4.9 °C in a water electrolyzer at the current density of 1 A cm^-2. Thus, the thermally conductive HEM exhibits negligible degradation after 20,000 start/stop cycles and reduces the degradation rate by 6 times compared to the pure polymeric HEM in a water electrolyzer. This study manifests the significance of thermal conductivity of HEM on the durability of water electrolysis, which provides guidelines on the rational design of highly durable HEMs in practical operation conditions for water electrolysis, fuel cells, and beyond.

Fig. 1. In-situ stability of HEM affected by localized temperature and thermal conductivity.

a Schematic diagram of the setup for the temperature measurement of HEMWE, where the thermocouples are inserted along the GDL layer near the cathode side (T_a), cathode catalyst layer near the HEM (T_b), and GDL layer near the anode side (T_c), respectively. b Temperature (T_a, T_b, T_c) evolution of HEMs operated in a water electrolyzer assembled with commercial HEM of FAAM−40 at various current densities. c Voltage evolution of the water electrolyzer assembled with FAAM−40 operated at different cell temperatures and the corresponding voltage degradation rates. d Schematic diagram of components in HEMWE for the numerical simulation and simulated 2D temperature contour plot of HEMs with k = 0.4 and k = 6.2 W m⁻¹ K⁻¹. e Simulated 1D temperature profile within the HEMWE assembled with HEMs with different thermal conductivities at a current density of 2 A cm⁻². Unit of k is W m⁻¹ K⁻¹. HEM is indicated by a red shadow. f Simulated temperature contours of HEMs operated in HEMWE at various current densities assembled with HEMs with different thermal conductivities. The bipolar plate, gas diffusion layer, cathode catalyst layer, and anode catalyst layer are respectively abbreviated as BP, GDL, C-CL, and A-CL in panels (a, d, and e).

Fig. 2. Highly thermally conductive composite HEM.

a Schematic diagram of HEMWE in which the BN composite HEM with a 3D thermally conductive network is applied. Composite HEM exhibits fast heat diffusion, thereby potentially reducing the localized temperature and promoting the in-situ stability of HEM in a water electrolyzer. b Design of BN composited HEM, where phonon and hydroxide transport along the 3D network. c, d Side-view SEM image of the QCS/3.8%BN and QCS/10.6%BN membranes, respectively. The structure of QCS/BN membranes is indicated in insets. e Thermal conductivity of QCS/BN membranes with different BN volume fractions. Inset: Photo of pure QCS and QCS/10.6%BN. Data in (e) are presented as mean values  ±  standard deviation based on at least three trials. f Infrared photos of membranes with a square region cooling from 60 °C to 10 °C.

Fig. 3. Ionic transport properties of QCS/BN composite HEM.

a Ion conductivity in water (room temperature) and tensile stress of QCS/BN membranes with different BN volume fractions. b IEC of QCS/BN membranes. c Arrhenius plots of QCS and QCS/10.6%BN membranes. d Dielectric relaxation spectra and (e) relaxation times of different QCS/BN membranes. f Relationship between relaxation time and temperature. g Schematic diagram of water and hydroxide transport of QCS with open space and QCS/BN composites with different space confinement effects. Data in (a, b) are presented as mean values  ±  standard deviation based on at least three trials.

Fig. 4. In-situ stability of HEMs for water electrolysis.

a Temperature profiles of the cathode catalyst layer near the HEM (T_b). b T_b of the HEMWE with QCS/BN composite HEMs with different thermal conductivities measured at different current densities. c Voltage evolution of HEMs with different thermal conductivities at 1 A cm⁻². d Relationship between in-situ voltage degradation rates and temperature vs. thermal conductivity of HEMs. e Possible degradation pathways and (f) the accordingly energy barriers calculated by DFT. g ¹H NMR spectrum and (h) the accordingly QA content retention of the HEMs after the in-situ stability test. Peaks of GLU and QA are indicated by the purple shadows in panel (g). The peak at 3.8–4.2 ppm ascribed to the glucosamine (GLU) unit of chitosan is considered a reference peak. The HEM after the stability test was treated by dissolving it in acetic acid solution and centrifugal filtration to remove the catalyst and boron nitride for the ¹H NMR test. i Long-term in-situ stability of FAA-3, QCS/10.6%BN, and QCS for water electrolysis. Unit of k is W m⁻¹ K⁻¹.

Fig. 5. Performance of water electrolysis using QAPPT/BN composite HEM.

a Ionic conductivity and thermal conductivity of QAPPT and QAPPT/8.9%BN HEMs. The inset shows the chemical structure of QAPPT. b Polarization curves measured in 15% KOH at 90 °C and 110 °C. c Transient in-situ stability test based on transiently oscillated current densities between 1 A cm^–2 and 0.1 A cm^–2, holding for 2 s and standing for 1 s for each current density every cycle. d Resistance changes of the electrolyzer before and after the transient stability test. e In-situ stability test in 15% KOH at a constant current density of 1 A cm⁻². The inset shows the polarization curves of the pristine HEM and the HEM after the in-situ stability test at 90 °C. f Comparison of voltage degradation rate and temperature between the proposed QAPPT/BN composite HEM and reported HEMs^,,,–. Testing parameters, including current density and KOH concentration are indicated by bubble size and color contrast, respectively. Lifetime is placed beside the bubbles. The samples in this work are indicated by the blue shadow. g Comparison of ex-situ stability and in-situ stability of the QAPPT/BN composite HEM and reported HEMs in HEMWE^{,,,,–,,,,,,–}. The reported HEMs are indicated by the blue shadow.