Electrochimica Acta 101 (2013) 326–333 Contents lists available at SciVerse ScienceDirect Electrochimica Acta j our nal homep age : www.elsevier.com/locate/electacta Dealloying of platinum-based alloy catalysts: Kinetic Monte Carlo simulations Rafael Callejas-Tovar a , C. Alex Diaz a , Julibeth M. Martinez de la Hoz a,b , Perla B. Balbuena a,b,∗ a Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX 77843, United States b Materials Science and Engineering Program, Texas A&M University, College Station, TX 77843, United States a r t i c l e i n f o Article history: Received 30 September 2012 Received in revised form 8 January 2013 Accepted 12 January 2013 Available online 20 January 2013 Keywords: Kinetic Monte Carlo Dealloying Nanoparticles Oxygen reduction reaction Porous structures a b s t r a c t Kinetic Monte Carlo simulations are performed to study the dealloying of Pt-based nanoparticles typically used as oxygen reduction reaction catalysts. The Kirkendall effect is represented to emulate the synthe- sis of hollow nanoparticles by removing the Ni core in a Ni-core/Pt-shell nanoparticle. It is found that initial shell vacancies are required to completely dissolve the non-noble core. The evolution of porosity is followed by dealloying Ni 0.75 Pt 0.25 and Co 0.63 Pt 0.37 nanoparticles. Two critical potentials define regions where the bimetallic particles may exist as core–shell, porous, and hollow structures, accompanied with clear variations in the dissolution rates. The phenomena are characterized by the dynamic evolution of the surface coordination numbers, and that of the surface area per platinum mass. © 2013 Elsevier Ltd. All rights reserved. 1. Introduction Dealloying is the selective removal of elements from an alloy. According to recent experimental reports, dealloying that takes place during synthesis of oxygen reduction reaction (ORR) catalysts may yield remarkable activity enhancement by creating character- istic porous or hollow structures [1–6]. However, the dealloying process that occurs during fuel cell operation is enhanced by the concomitant presence of oxygen that promotes metal dissolution and causes the degradation of the catalyst. Therefore, it is crucial to understand the dealloying mechanism in order to use it for tailoring the synthesis of improved catalysts and to prevent the degradation of such catalysts under operating conditions. Several theoretical studies have demonstrated the usefulness of coarse-grained approaches such as kinetic Monte Carlo (KMC) sim- ulations for the molecular understanding of processes taking place in relatively long time scales, comparable to experiments [7–12]. Here we study the driving forces and the effect of dealloying on the structure of alloy nanocatalysts during their synthesis using KMC simulations. The KMC method requires the correct represen- tation of all the relevant processes in the system in the form of rate expressions. The algorithm selects the processes taking place at each time step based on the magnitude of their rates within a random approach. We present the results of a 3D simulation ∗ Corresponding author at: Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX 77843, United States. Tel.: +1 979 845 3375; fax: +1 979 845 6446. E-mail address: balbuena@tamu.edu (P.B. Balbuena). code capable of studying the degradation and dealloying processes in ORR nanocatalysts including the environmental effects of the electrolyte. Two cases of study are reported here: the synthesis of hollow nanoparticles resulting from removal of the alloy core through dealloying using the Kirkendall effect, and the synthesis of porous nanoparticles by selective dealloying of the less-noble com- ponent. Both cases are relevant for improving the understanding of the possible structures of state-of-the-art ORR catalysts. 2. Computational details The initial configuration of the system in the KMC model consists of a nanoparticle with a FCC structure surrounded by electrolyte particles as depicted in Fig. 1a. For the Kirkendall effect simula- tions (Fig. 1b) either pure or bi-metallic core/shell systems with and without vacancies in the shell were considered. For the evolution of porosity simulations, Ni 0.75 Pt 0.25 and Co 0.63 Pt 0.37 nanoparticles with an initial radius between 25 and 40 ˚ A were studied. All the species occupy a site in the 3D simulation lattice and they may participate in two reactions: diffusion and dissolution. The rate k p,i for the reaction p involving site i is calculated with Eq. (1), where A p,i is the prefactor in sites/s, E B,i is the total bonding energy, given by Eq. (2), which depends on the number of nearest neighbor sites of type n j and the pair bonding energy parameter E b,ij (Eq. (3)); ˛ p,i is the transfer coefficient parameter, E is the applied potential, E 0p,i is the reference potential, T is the temperature, and k B is the Boltzmann constant. k p,i = A p,i exp  - E B,i + ˛ p,i (E - E 0p,i ) k B T  (1) 0013-4686/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.electacta.2013.01.053