Investigation of a reversible PEM fuel cell with integrated metal-hydride hydrogen storage

The overall aim of this project was to investigate experimentally for the first time the technical feasibility of integrating a metal hydride (MH) hydrogen material into a proton exchange membrane (PEM) unitised regenerative fuel cell (URFC).

To investigate experimentally the technical feasibility of the proton flow battery − that is, a PEM URFC with integrated MH electrode − four electrodes with different structures but made of the same MH material, were designed and fabricated:

1. The ‘MH-NiMesh electrode’: a mixture of the MH material and PTFE binder cold pressed onto a nickel substrate
2. The ‘MH Powder 1 electrode’: the MH powder cold pressed (with no binder) onto the Nafion membrane
3. The ‘MH Powder 2 electrode’: the MH powder hot pressed (with no binder) onto the Nafion membrane
4. The ‘MH-Nafion composite electrode’: a mixture of MH powder and Nafion solution formed into a solid composite, and then hot pressed onto the Nafion membrane.

Each of these electrodes was integrated into a modified URFC and was tested in both operational modes: Electrolysis (E-mode) and fuel-cell (FC-mode). The test results were used to determine if hydrogen was absorbed in E-mode of operation into the MH electrode, and if the electrical discharge capacity of the cells in FC-mode could be related to the hydrogen stored in the electrode.

The MH-Nafion electrode had the best performance in E-mode with an estimated hydrogen storage capacity of 0.62 wt% H/M. The maximum hydrogen storage capacity of the MH material used was around 1.3 wt% H/M, which shows the first trial of this new composite electrode did achieve a reasonable hydrogen storage capacity at ambient pressure and temperature. The hydrogen storage capacities for the MH Powder 2 and MH Powder 1 electrodes in E-mode were 0.15% and 0.07% respectively, so that some storage capacity was demonstrated, albeit limited in magnitude. The MH-NiMesh electrode did not show any signs of hydrogen storage in E-mode.

However, only the MH Powder 1 electrode was able to use all of the hydrogen stored in the MH electrode in FC-mode. This cell had the best discharge capacity, with an equivalent hydrogen storage capacity of 0.065 wt%, which was almost the same as its storage capacity of 0.07 wt% in E-mode. Hence this electrode showed a high level of reversibility. The discharge capacities for the MH Powder 2 and MH-Nafion electrodes were 0.01 wt% and 0.007 wt% respectively, which were very low compared to that for the MH Powder 1 electrode, and also much lower than their own storage capacity in E-mode (0.15 and 0.62 wt% respectively). These results show that although these electrodes have better hydrogen storage performance in E-mode, they were nowhere near fully reversible in their present forms.

These results provide initial confirmatory evidence that the concept of a PEM URFC with an integrated MH storage electrode is technically feasible, though additional research is still required to improve the performance of the experimental cells investigated in this project in terms of storage capacity and reversibility.