Intermediate Temperature Steam Electrolysis with Phosphate-Based Electrolytes

Abstract

Water electrolysis for hydrogen production has been predicted to get a prominent role in the energy system of the future. Current low temperature technologies rely on expensive noble metal catalysts and high temperature systems requires special construction materials to withstand the high temperatures. Electrolysis in the intermediate temperature (IT) region (200-400 °C) is of interest as it would allow for the use of non-noble metal catalysts, due to the improved kinetics, and a wide range of construction materials as a result of the more benign temperature. At these temperatures water is supplied as steam. This work centred on the design and development of a novel steam electrolysis concept based on phosphate electrolytes capable of operating in the IT range. Central for the work was the selection and evaluation of the materials and components for the test setup and cells as well as the technological issues and challenges faced. A setup suitable for intermediate temperature electrolysis has been constructed in order to accommodate testing in the IT region. This included the evaluation of multiple generations of components such as end plates and flow plates. Chemical vapour deposition of tantalum was used to protect stainless steel components from the highly oxidative environment of the oxygen side of the electrolyser. While such protection should not be necessary on the hydrogen side, it was found that the best results were obtained using tantalum coated stainless steel flow plates not only on the oxygen side but at the hydrogen side as well. Additional key steps and components for electrolysis testing are detailed in this thesis. This includes gas diffusion layers (GDL), sealing, cell assembly techniques, test operation, electrolytes and electrocatalysts. Gas diffusion layers of carbon with a PTFE bound micro-porous layer was used for the cathode side and tantalum coated stainless steel felt was used for the anode side due to the need of corrosion protection. For the cathode side a platinum electrocatalyst was used as benchmark (Pt-black ≈ 8 mg/cm2) and iridium oxide was used for the anode (≈ 3 mg/cm2). Symmetrical cell testing for hydrogen pumping at 200 _C revealed the cathode gas diffusion layers to be unstable over time. After 60 hours, the electrode resistance was more than tripled. The most prominent reason for this was thought to be a softening of the PTFE in the cathode micro-porous layer. CsH2PO4 and Sn0.9In0.1P2O7 were used as proof-of-concept electrolytes, with emphasis on the latter electrolyte. Evaluation of electrolysis cells with these electrolytes was done with a range of tools constantly under development. These tools included regression analysis of I-V curves, reference electrode measurements and electrochemical impedance spectroscopy (EIS). While reference electrode measurements were found hard to optimise, EIS, and especially complex non-linear least-square (CNLS) fitting, was found very useful. CNLS allowed for the estimation of electrolyte resistance and polarisation resistances giving a detailed view of the novel system. Electrolysis with CsH2PO4 as electrolyte revealed a need for steam on both cathode and anode in order to prevent dehydration of the electrolyte. Additional stabilisation in the form of SiC fibres was found to increase longevity considerably. Highest achieved current density was 60 mA/cm2 at 2.0 V and 250 °C. Measurements using Sn0.9In0.1P2O7 as electrolyte, Pt black as cathode electrocatalyst and IrO2 as anode electrocatalyst gave current densities as high as 313 mA/cm2 at 1.9 V and 200 °C. The stability of the electrolyte was found to be high at 200 °C and a water partial pressure of 0.05 atm. For stabilisation of the electrolyte at 250 °C a higher water partial pressure is needed. Variation of temperature from 200-250 °C showed both signs of activation of electrode processes and electrode degradation. Efforts were done to optimise the synthesis of Sn0.9In0.1P2O7 in order to establish a reproducible synthesis procedure. The synthesis used in this work required two heat treatment steps. Fourier transform infrared spectroscopy (FT-IR) shows an O-H band in the IR spectrum from 1500 cm-1 to 3800 cm-1 strongly dependent on the first heat treatment step of the synthesis. It was found that initial heating of the synthesis precursors to 270 _C gave a high quality sample in a reproducible fashion. Investigations of two additional novel phosphates was attempted. These were phosphoric acid treated Nb5P7O30 and a mixture of Bi2P4O13, BiPO4 and 2 wt.% Polybenzimidazole (PBI). Both were found to be lacking in stability. As a central attraction of IT electrolysis is the possible use of non-noble electrocatalysts such materials were tested using Sn0.9In0.1P2O7 as an electrolyte. Tungsten carbide was tested as an alternative cathode electrocatalyst. Two samples were prepared by carburisation and tested. One prepared from WO3 and one prepared fromW2N. Current densities at 1.9 V were measured as high at 129 mA/cm2 and 73 mA/cm2 for the two samples respectively. While WC gave lower performance than Pt black its stability was found to be higher. Furthermore, testing WC and Pt black at increasing temperatures revealed an increase in performance for a WC cathode but a decrease for a Pt black electrode. This bodes well for the further use of WC as cathode electrocatalyst. Ni foam tested as a cathode electrocatalyst showed fair stability but lacked activity. Ni was tested as anode electrocatalyst as well. Here it showed low activity and poor stability. LaNiO3 was tested as cathode electrocatalyst where it was found to be lacking in activity. It was however stable over polarisation indicating the possible use of similar materials in this role.