Modeling Platinum Oxide Growth of PEMFC Cathode Catalysts

Introduction

Catalyst performance and durability are important factors that impact the cost, which is critical for successful deployment of PEMFCs in automotive applications. Oxygen reduction reaction (ORR) in a fuel cell occurs on a partially oxidized platinum (Pt) surface and the durability of the Pt surface also depends on the extent of Pt oxidation. Therefore, Pt oxide is intricately related to both durability and performance of cathode catalysts. A well-established in-situ method for characterizing the oxide type and amount do not yet exist. In this work, we applied a coulometric method for measuring the growth of cathode Pt oxide in an N₂/H₂O atmosphere as a function of potential and temperature. An oxide growth/reduction model was developed and characterized against the measurements made on Pt/Vulcan and PtCo/HSC catalysts.

Experimental

A detailed description of the measurement of Pt oxide coverage in PEMFCs is provided in the publication by Y. Liu et al.^[1]. The time dependence of Pt oxide formation was measured by changing the potential-hold time from 5 sec to 120 min at 0.75, 0.8, 0.85… 1 V_RHE, respectively. The temperature effect on the Pt oxide coverage was tested at 30 to 90°C at fully saturated conditions and at the corresponding total pressure of 101.3 kPa_abs plus water vapor pressure at each temperature.  Before each measurement, the cell was pre-conditioned for 1hr at the open circuit voltage for each set of operating conditions (10% H₂ (in N₂) on the anode and N₂ on cathode with a flow rate of 1000 sccm on both sides).  The CVs (Gamry^® G750) were collected by scanning negatively from the hold potential at a scan rate of 20mV/s down to 0.05V_RHE followed by three sweeps with potential ranging from 0.05 to 0.6V_RHE. The cumulative charge from HAD curves was used to calculate total oxide coverage on the basis of 2e^-for each Pt site.

Results and Discussion

Figure 1 below shows the growth in oxide coverage on log time basis for PtCo/HSC and Pt/V catalysts at various hold potentials. The data shows that there are two time constants associated with two types of oxide growth; the first type was predominant below 300s (at all hold potentials) followed by a slower oxide growth. The mechanisms for oxide growth mechanisms was extensively discussed in the literature and recent studies from CV, EQCM, AES and XAFS established a mechanism ^[2,3,4] wherein water activation or physisorption of H₂O on the Pt surface was the first step to occur even at low potentials. With  potentials >0.75V vs RHE, oxidation of physisorbed water molecules lead to adsorbed oxide species (O-Pt or OH-Pt) followed by a interfacial place between O(H)_ads and Pt atoms to form a quasi- 3D layer. Based on this mechanism, oxide growth kinetics was modeled with two reversible reactions; Pt to PtO (Pt oxidation) and OPt to OPtO (Place-exchanged Pt oxidation). It is reported in the literature that Pt site crystallographic orientation also has significant impact on the type of oxide formed and also the growth kinetics ^{[5, 6]}.  Trustworthy methods to quantify fraction of crystal sites and a more quantitative understanding of this mechanism is needed to model such effects, therefore we did not include that level of detail in the present model.

The kinetic parameters derived from the data reveal that Pt oxidation is faster than the place-exchanged Pt oxidation by several orders of magnitude. Model-based analysis also shows that Pt oxidation is two times faster and place-exchanged oxidation is 100 times faster on PtCo/HSC catalysts when compared on Pt/V. Effect of temperature on kinetics was found to be similar for both Pt/V and PtCo/HSC catalysts. The model can be used to quantify the growth and reduction of oxide coverages during voltage cycling. This will enable understanding of the hysteresis in ORR kinetics as a function of oxide growth/reduction as well as identification of the catalyst degradation factors in a voltage cycle.

References

1. Y. Liu, M.F. Mathias, J. Zhang, Electrochem.Solid-State Lett., 13, B1 (2010).
2. G. Jerkiewicz, G. Vatankhah, J. Lessard, M.P. Soriaga, Y.S. Park, Electrochimica Acta 49 (2004).
3. M. Alsabet, M. Grden, G. Jerkiewicz, J. Electroanalytical Chemistry 589 (2006).
4. B.E. Conway, Progress in Surface Science, Vol 49, No 4 (1995).
5. A. Kongkanand, J. M. Ziegelbauer, J. Phys. Chem. C 116 (2012).
6. E. L. Redmond, B. P. Setzler, F. M. Alamgri, T. F. Fuller, Phys. Chem. Phys. 16 (2014).