No abstract

Gross fuel starvation at the anode of a proton exchange membrane fuel cell (PEMFC) can result in cell reversal. High positive anode potentials are reached to sustain the current by water oxidation and carbon corrosion at the Pt based anode electrocatalysts. The membrane electrode assembly (MEA) is irreversibly damaged and the flow field plates and stack material in electrical contact with the MEA can be corroded. One of the best approaches to minimise the impact of fuel starvation is to build cell reversal tolerance into the MEA. A protocol was developed to investigate the impact of cell reversal in single cell and in stack hardware. By monitoring the anode outlet gas for O 2 and CO 2 the relative amounts of water oxidation and carbon corrosion were established for MEAs with varying degrees of cell reversal tolerance. Moving to more corrosion resistant carbon supports, adding Teflon to the anode catalyst layer to retain water, and deploying a water oxidation electrocatalyst in the anode significantly improved the cell reversal tolerance of the MEA.

Carbon supported Pt and PtCo alloy cathodes have been evaluated under the DoE proscribed accelerated MEA durability protocol (0.7 - 0.9 V iR free, square wave cycle, H2/air). The Pt MEA showed reasonable stability up to 2500 cycles, after which significant performance loss was observed. Post mortem analysis indicated that this was due to Pt sintering and dissolution. The PtCo MEA showed similar stability up to 2500 cycles showing a mass activity of twice that of the Pt MEA. However, after this, activity loss was observed to a value similar to pure Pt, without significant loss in electrochemical area. The MEA was tested to 83,000 cycles with only a gradual further activity loss. Post-mortem analysis showed only modest Pt sintering and loss of Pt into the membrane. This early loss of PtCo activity has been correlated with surface loss of Co through activity testing of acid pre-leached catalysts.

In this work it is shown that PtCo/C catalysts give a 2× mass activity enhancement in oxygen reduction activity, compared to Pt/C. Through the use of Pt and PtCo catalysts of varying surface area, clear evidence of a particle size effect on oxygen reduction activity was found. Investigation of the effects of cathode potential cycling in both wet electrolyte cells (1M H 2 SO 4 ) and as small MEAs revealed high metal surface area stability and stable oxygen performance over 10,000 cycles for the PtCo/C alloys compared to Pt/C. TEM analysis of MEA cross-sections after cycling show clear evidence of Pt migration towards and into the PEMFC membrane for Pt/C cathodes, whilst PtCo/C materials show minimal deterioration. The larger crystallite sizes present in PtCo particles therefore confer higher stability to cycling and resistance to dissolution, whilst use of an alloy produces high mass activity, despite a lower metal surface area compared to Pt/C.

The mechanism of activity loss in Pt and PtCo alloy systems has been investigated during 0.6 - 1.0 V and 0.6 - 1.2 V vs RHE cycling regimes using an electrochemical cell that allows electrochemical metal area (ECA) loss measurement and electrolyte sampling. Dissolution of platinum and base metal into the electrolyte was quantified using ICP-MS and TEM used to look for changes in catalyst morphology with cycles. ECA loss and base metal removal were found to be potential dependent. PtCo alloys were found to be more stable to ECA loss by Ostwald-ripening and dissolution than Pt. In terms of base metal loss the PtCo alloy was unstable and cycling was found to remove similar amounts of Co by dissolution as chemical acid leaching. Loss of Co from the surface layers of the catalyst was found to remove the enhanced stability and activity benefits of this catalyst making it more platinum like.