Durability of Low-PGM Electrodes Under Automotive Drive Cycles

Abstract

We are evaluating the potential of reaching 8000-h durability of automotive fuel cell systems with a de-alloyed Pt3Co catalyst supported on a high surface area carbon (HSAC) for >440 A/gPt mass activity at a low Pt loading (<0.1 mg-Pt/cm2) for meeting the 30 $/kWe cost target. The supporting experimental data was obtained on a membrane electrode assembly (MEA) operated under differential cell conditions. The MEA has d-PtCo/C in cathode, Pt supported on Vulcan carbon in anode, thin 12-mm reinforced PFSA membrane, and 200-mm GDL with microporous layers. The Pt loadings are 0.1 mg/cm2 in the cathode catalyst and 0.025 mg/cm2 in the anode catalyst.The durability data were obtained by running catalyst accelerated stress tests (AST) for 5k, 10k and 30k cycles. The standard AST protocol consists of trapezoidal cycles with 3-s holds at 0.60/0.95 V lower/upper potential limits (LPL/UPL) and exposure at 80oC and 100% relative humidity (RH). For model development, additional single-variable tests were run at different UPL (0.95 V, 0.90 V, 0.85 V and 0.80 V) and RH (50% and 100%) at 80ᵒC. The ORR kinetics and O2 transport resistances in degraded electrodes were characterized by measuring mass activity and H2/N2 EIS impedance (EIS) at beginning (BOT) and end of tests (EOT) and, in some tests, polarizations curves at different pressures (1-2.5 atm), temperatures (60-95oC), RH (30-100%), and O2 mole fraction (5-21%). In parallel with the differential cell campaign, two-factor interactions were investigated by conducting catalyst ASTs on 50-cm2 cells. In all, 16 tests were conducted for different combinations of UPL (0.95 V and 0.85 V), RH (100% and 40%), temperature (95°C and 55°C), and hold time at UPL (1 s and 5 s) for 30k and 60k cycles. Also, one standard condition (0.9 V UPL, 70% RH, 75ᵒC and 3 s hold time at UPL) was repeated to track reproducibility in each test series.We constructed a durability model that calculates the aqueous dissolution of Pt and Co from a d-PtCo/C catalyst for which online ICP-MS measurements are available as a function of potential cycle parameters. We correlated the calculated Pt dissolution rate to the measured ECSA loss in catalyst AST tests on differential cells at different UPL, temperature and RH. The correlation was validated against the ECSA losses measured in two-variable interaction tests on the 50-cm2 cell.We used the performance and durability models to understand ECSA and voltage losses on urban (UDDS) and highway (HWFET) duty cycles to measure fuel economy of light duty vehicles. These simulations show that the requirement of <10% acceptable degradation in rated power over lifetime can be met by restricting the ECSA loss to 55.3%. A sensitivity analysis shows that potential and RH are the two most important parameters that control electrode degradation. A potential-RH-temperature acceptability map was constructed to determine the envelope of operating conditions for achieving target ECSA loss rate. We used this envelope to guide us in determining the parameters and controls to reach 8,000-h durability on drive cycles in a mid-size vehicle powered by an 80-kWe FCS with a 1.6 kWh battery. We observed that the minimum FCS power (4 kWe) is an important parameter since it determines the peak potential reached during a drive cycle transient. If the power demand falls below 4 kWe, a drive-cycle FC stack shutdown is initiated during which the anode and cathode channels are isolated and electrochemically depleted of H2 and O2. The subsequent stack restart depends on power sharing between FCS and the hybrid battery. We found that the CEM turndown (ratio of air flow rate at rated power and at idle) and cell temperature are important parameters in controlling RH and maximum cell voltage. Figure 1(left) indicates 1,800-h lifetime at the baseline operating conditions, CEM turndown of 20 and 66oC average stack temperature, that lead to 111%/116% average outlet RH in UDDS/HWFET, and cell voltage varying between 760-866 mV. Figure 1(middle) shows that the electrode lifetime can be extended to 5,000 h by reducing the CEM turndown to 12.5 such that the average outlet RH is 90%/102% during UDDS/HWFET, and the cell voltage varies between 760-850 mV. These operating conditions may be benign toward membrane stability. Reaching 8,000-h electrode lifetime requires the CEM turndown to be further lowered to 10 and the average cell temperature to be raised to 70oC, see Fig. 1(right), so that the average outlet RH is 52%/78% during UDDS/HWFET, and the cell voltage varies between 760-825 mV. These dry operating conditions may prove challenging to membrane stability and require further investigation. Figure 1