Effect of impurity and Alloying Elements on Zirconium (Zr) Grain Boundary Strength and Iodine Adsorption, Dissociation, and Diffusion from First-Principles Computations

The combination of mechanical properties, corrosion resistance and low neutron absorption cross section makes zirconium-based alloys very important structural materials in the nuclear industry. To maintain functionality of zirconium alloys in nuclear environments, it is important to understand effects of alloying elements and elements resulting from fission, transmutation, and coolant interactions on material properties. Of particular concern is weakening caused by accumulation of impurities at grain boundaries. Resistance of zirconium-based alloys to radiation embrittlement is highly valued in reactor design. Even after significant irradiation, Zircaloy fracture surfaces exhibit ductile fracture features. However, during power transients in commercial nuclear power plants, brittle fracture can occur and is known as pellet-clad interaction (PCI). The responsible mechanism is believed to be fission product induced stress corrosion cracking, with iodine suspected as the primary detrimental specie. However, mechanistic details for how iodine promotes brittle failure are not well understood, with two main mechanisms proposed. One mechanism involves iodine being adsorbed on a crack face lowering Zr-Zr bond energies, leading to separation along crystallographic planes (transgranular cleavage crack path) or grain boundaries (intergranular crack path). Another mechanism postulates that crack advance occurs by formation of Zrl4(g) which transports Zr away from crack tips. The importance of iodine concentration and Zrl4 partial pressure on SCC (stress corrosion cracking) susceptibility of Zircaloy has been shown in several studies. While it is difficult to obtain experimental information on effects of impurity atoms on grain boundary (GB) strength, first-principles calculations can provide insight. This work first studied the effect of twenty impurity and alloy elements on the strength of a Zr(0001)/ Zr(0001) Σ7 twist grain boundary. Modeling based on ab initio density functional theory (DFT) was used for all calculations as implemented in the Vienna ab initio Simulation Package (VASP). For each impurity investigated the preferred occupancy site (substitutional, interstitial, or surface adatom), segregation energy (from bulk to surface or grain boundary), and change in Zr grain boundary cohesion energy were calculated. The influence of twenty different impurity elements on the Zr grain boundary work of separation was ranked in order of most weakening to most strengthening: Cs, I, He, Te, Sb, Li, Sn, Cd, O, H, Si, C, N, U, B, Ni, Hf, Nb, Cr, and Fe. These results demonstrate that strength of a Zr Σ7 (0001) twist grain boundary is quite insensitive to most impurity elements. In part, this is due to similar behavior of impurities at grain boundaries and at free Zr surfaces. Elements that prefer a substitutional site in bulk Zr also prefer substitutional sites at surfaces and in Zr Σ7 (0001) grain boundaries, Exceptions are Li, I, and s, that prefer bulk substitutional sites but prefer an adatom position at free surfaces (He which just leaves free surfaces is another exception). These elements are also among those with the largest reduction in calculated grain boundary cohesion. No element in this study showed a large beneficial effect on grain boundary strength. Comparison of modeling and experimental reports reveals consistency for iodine embrittlement results. Both computations and experiments show iodine to weaken grain boundaries, however, results for Cs and He are inconsistent between modeling and literature. Experimentally He does not promote brittle failure, and is readily understood as there is no driving force for helium to interact with zirconium. For Cs, neither brittle intergranular nor transgranular cleavage was reported even though calculations show Cs to be the most embrittling for a Zr Σ7 (0001) grain boundary. This may be attributed to details of failure mechanisms that have not yet been explored, such as how detrimental elements interact at a crack tip to lower strength of a boundary and enhance kinetic transport phenomena. To further investigate kinetics of detrimental effects of iodine on Zr cohesion, adsorption, dissociation, and diffusion of iodine on a Zr surface were studied using a first-principles approach. Different static and dynamic simulations were run for adsorption and dissociation of molecular iodine on a Zr(0001) surface. Molecular iodine (I2) impinging on a Zr (0001) surface shows no barrier to surface approach and no precursor state, resulting in spontaneous dissociation of 12 molecules on this surface. The equilibrium surface coverage of iodine atoms on zirconium was calculated as a function of temperature and partial pressure of molecular iodine in the gas phase. Computed adsorption isotherms show significant adsorption site occupancy even at extremely low I2 partial pressures. Diffusion of iodine atoms on a Zr (0001) surface was calculated by considering jumps between stable FCC sites. Tracing an energy profile along a diffusion path shows minima exist at FCC sites and local minima occur at HCP sites. A diffusion coefficient for iodine can be expressed using an Arrhenius relationship as D = Doe-Q/RT with Q = 6.77 kJ/mol and 2.0 x 10-4 cm2/sec. Diffusion of iodine atoms on a Zr surface was calculated to be quite fast with little activation energy required. These results indicate that iodine, which is shown computationally and experimentally to be detrimental to Zr, has a high tendency to react with a zirconium surface and a high mobility once dissociated on a surface. These behaviors support an iodine-enabled Zr cracking mechanism. (authors)