Recoverable Degradation Losses in PEM Fuel Cells

Abstract

A major problem with performance for transportation applications is associated with degradation. The performance degradation losses can be separated into two categories: irreversible and recoverable losses. Significantly more work has examined the irreversible degradation losses compared with the recoverable losses because the recoverable losses often can be simply recovered by performing polarization curves[1]. Delineating the component contributions of recoverable performance losses and mitigation of these losses appears to be a growing concern for fuel cell developers. While these degradation mechanisms are recoverable, recovery processes are not necessarily trivial to accomplish in operando in a vehicle. MEA performance degradation is sensitive to operating conditions and both irreversible and reversible materials changes. Some known recoverable losses for PEMFCs include the loss of cathode activity due to surface oxide (hydroxide) formation at high potentials,[2] catalyst poisoning by membrane degradation products [3] and recoverable transport losses due to water transport [1,4]. Platinum oxidation at the cathode results in decreased ORR (Oxygen Reduction Reaction) and with Pt often considered to have two different Tafel slopes; one for a metallic Pt surface, and one for an oxidized Pt surface. In other cases, indications are that membrane degradation products, such as (bi)sulfate are readily adsorbed onto the catalyst and are a large source of recoverable degradation. These require removal from the catalyst surface and then from the electrode layer for the performance recovery.[5] To examine the relative contributions to recoverable degradation, we have conducted in situ and ex situ experiments to separate the effects of Pt oxidation, (bi)sulphate adsorption, water management plus other operational effects including spatial degradation mapping, potential cycling and different levels of RH. One method to identify the effect of membrane degradation products on the performance decay is to compare the effect of non-chemically-stabilized membranes versus the effect of chemically-stabilized membranes. This is shown in Figure 1, where the OCV is shown during two-24-hr periods (with recovery in-between the 24-hr periods) for MEAs with (a) a non-chemically stabilized membrane and (b) a chemically stabilized membrane. Past results have shown significantly more degradation products of fluoride and sulphate anions without the chemical stabilization. Both MEAs show decreasing OCVs, however the decay is significantly more for MEA using the non-chemically stabilized membrane. Segmented cell measurements, made in a 10x10 co-flow cell, do not indicate a substantial spatial difference in the rate of decay during these types of tests. Various methods of recovering the losses were examined to separate the Pt oxidation from the other effects, e.g. decreasing the potential to 0.6 V to reduce the Pt oxide films compared with much lower recovery voltages to desorb adsorbed anions. Recovery protocols in fuel cell mode (H2/air) are also compared without current generation (H2/N2) at potentials of 0.6 V, 0.4 V, 0.3V and 0.2 V including with and without liquid water injection. Liquid water injection was found to be detrimental to the recovery of the fuel cell performance and a voltage of 0.3 V required to good recovery with little difference below that potential. Acknowledgments This research is supported by DOE Fuel Cell Technologies Office, through the Fuel Cell Performance and Durability (FC-PAD) Consortium; Fuel Cells program manager: Dimitrios Papageorgopoulos.