Ab Initio Atomistic Thermodynamics of Water Reacting with Uranium Dioxide Surfaces P. Maldonado,* ,† L. Z. Evins, ‡ and P. M. Oppeneer † † Department of Physics and Astronomy, Uppsala University, Box 516, SE-75120 Uppsala, Sweden ‡ Swedish Nuclear Fuel and Waste Management Co., Blekholmstorget 30, SE-10124 Stockholm, Sweden ABSTRACT: Using first-principles simulations, we study the temperature- and pressure-dependent adsorption reaction of water on the flat (111) and (211) and (221) stepped surfaces of uranium dioxide. Our calculations are based on the density functional theory (DFT) corrected for on-site Coulomb interactions (DFT+U) for describing the chemical interaction of water with UO 2 , in combination with ab initio molecular dynamics simulations to capture the temperature dependence of the reaction. We compute the pressure-temperature phase diagrams and establish the thermodynamic boundaries which govern the feasibility of water adsorption at these surfaces. Effects of water coverage on the surface adsorption reaction have been taken into account. We find that the dissociative adsorption reaction of water on stepped surfaces can be analyzed as two separated reactions, the dissociative water adsorption on the step edge and the water adsorption on the terrace. The most stable water adsorption upon modification of the water partial pressure and temperature is adsorption on the (211) step edge, followed by adsorption on the (221) step edge and being the least favorable for the (111) surface. We conclude that these UO 2 surfaces will always react with water at room temperature and atmospheric pressure, leading to water dissociation and a modification of the step morphology. ■ INTRODUCTION The treatment of spent nuclear fuel is currently the issue of an intense debate. Different strategies have been proposed to solve this issue. Nuclear transmutation 1 and nuclear reprocessing 2 are promising strategies that however do not represent a complete solution to the problem, since they also produce nuclear waste. In addition, these routes are only compatible with a commitment to nuclear power for the foreseeable future. If nuclear power is to be phased out and replaced with other energy sources, direct disposal of the spent nuclear fuel in deep geological repositories 3 emerges as the most promising solution for effective long-term isolation of the spent fuel. However, its implementation requires a fundamental understanding of fuel corrosion processes in order to present a safety case based on a scientifically sound estimation of the environmental impact of the planned repository. Uranium dioxide UO 2 is the most commonly used fuel in the nuclear reactors operating today, leading to a spent nuclear fuel consisting of mainly UO 2 with only a small fraction of highly radiotoxic long-lived actinides and fission products. Uranium dioxide has a very low solubility in water if conditions are reducing; 4 however, it has been shown that oxidizing conditions enhance the dissolution. 5 In a reducing repository environment, which is expected, for example, in the Swedish KBS-3 system, 6 oxidative corrosion could still be affecting the spent fuel due to the radiolysis of water, which produces oxidants that can attack the UO 2 surface. However, in the KBS-3 system, oxidative dissolution is expected to be suppressed by the so-called hydrogen effect, which has seen to be effective while there is anoxic corrosion of metallic iron. 7 Thus, nonoxidative dissolution of UO 2 must be considered as an important process by which the radionuclides contained within the spent fuel matrix will be released to the water. The first step in fully understanding this process is to investigate the energetics of water reactions with UO 2 surfaces. Uranium dioxide has been the object of extensive studies during the last decades, both experimentally 4,7-13 and computationally. 14-23 The complexity of the material, involving Received: February 18, 2014 Revised: April 1, 2014 Published: April 1, 2014 Article pubs.acs.org/JPCC © 2014 American Chemical Society 8491 dx.doi.org/10.1021/jp501715m | J. Phys. Chem. C 2014, 118, 8491-8500