(18e) Kinetic Monte Carlo Simulation of Surface Segregation in Platinum-Based Alloys Under O2 Adsorption and Reaction

One of the most important challenges in low-temperature fuel cell technology is improving the catalytic efficiency, especially at the electrode-catalyst where the oxygen reduction reaction takes place. Platinum is the most popular catalyst for this reaction; however, it is expensive and scarce. Pt-based alloys are being studied to substitute platinum while maintaining the efficiency and life-time of the catalyst. Since the reaction takes place in acid medium, both activity and durability of the catalyst are affected continuously while being exposed to such harsh environment. This work focuses on characterizing the evolution of the surface composition of platinum-based alloys under reaction conditions. The characterization is performed by applying first principles computational methods that allow us to follow changes on the surface composition at the atomistic level, and at the same time to identify how those changes impact the catalytic activity.

During the reaction, oxygen is adsorbed on the catalyst surface. We have recently shown that its presence has a strong influence on the surface segregation phenomenon and on the stability of the catalytic surface. Dramatic changes in the surface segregation have been observed where the system reverts from antisegregation to segregation and vice versa. The mechanisms of oxygen adsorption on catalytic alloy surfaces are complex and the presence of adsorbed oxygen affects the surface segregation phenomenon as well as the stability of the catalyst with respect to metal dissolution in the electrolyte. Experiments and ab initio simulations provide information about the possible microscopic states of oxygen adsorption and dissociation. However, the connection between such microscopic states and the time evolution of the rates of adsorption, desorption and change of the catalytic surface remain unexplained.

Molecular dynamics simulations can provide the dynamic evolution of a real system if the potential gives a correct description of the atomic forces present in the modeled material. However, a serious limitation is that accurate integration of the equations of motion requires short enough time steps (~10^-15 s) to resolve the atomic vibrations. Thus, the total simulated time is usually limited to less than one microsecond whereas the steps involved in the complete surface reaction, including surface reconstruction and side reactions such as metal dissolution, may take place on much longer time scales. In order to overcome this problem, we are implementing a coarse-grained 3-D Kinetic Monte Carlo method which utilizes the thermodynamic and kinetic information obtained from density functional theory (DFT). Hence, we are able to study the dynamic evolution of the surface composition of catalytic alloy surfaces in long time scales, and under electrochemical conditions. This approach provides useful insight of the behavior of the alloy surface under reaction conditions, and may give firm guidelines for the design of improved catalysts.