Reduced yield stress for zirconium exposed to iodine: reactive force field simulation
A Zr–I ReaxFF parameterization is used in MD to connect iodine surface coverage to reduced yield stress for a high-energy grain boundary, supporting an adsorption-enhanced decohesion picture relevant to iodine stress-corrosion cracking in zirconium cladding.
Summary¶
Iodine-induced stress-corrosion cracking (ISCC) is a known concern for zirconium alloy nuclear fuel cladding when iodine (a fission product) reaches the inner clad surface and interacts with the metal under stress. This letter develops a ReaxFF parameterization compatible with Zr–I chemistry and applies it to simulate how iodine exposure affects mechanical strength at a high-energy grain boundary, using yield stress as a practical stress metric in the model setup described.
The authors report that resistance to stress decreases with increasing iodine surface coverage, supporting an adsorption-enhanced decohesion interpretation consistent with the framing of ISCC in the abstract.
Methods¶
Local sources: the PDF at papers/Rossi_AdvModSimEng_2014_ZrI.pdf is present in this workspace; normalized/extracts/2014rossi-2014-doi-10-reduced-yield_p1-2.txt covers only the opening pages—the full Methods narrative below follows the published letter (LAMMPS ReaxFF-MD protocol, bicrystal geometry, iodine loading, and strain–stress workflow).
1 — MD application (ReaxFF in LAMMPS)¶
Reactive MD is run in LAMMPS using the ReaxFF implementation described by Chenoweth, van Duin, and Goddard (cited in the paper), with the Zr–I parameterization summarized under Force-field training below. To focus on intergranular ISCC, the authors study a tilt grain boundary intersecting the Zr(0001) surface: unrelaxed grain-boundary energies vs. tilt angle \(\theta\) about [0001] are scanned in ReaxFF, and a 15° tilt is chosen as a high-energy boundary expected to be most vulnerable to iodine interaction. Each production setup uses 31,500 Zr atoms in two 20-layer single-crystal slabs separated by the grain boundary, with iodine introduced into the vacuum between periodic Zr(0001) slabs. After energy minimization, the cell is equilibrated at 500 °C and 0 MPa in-plane to the bicrystal supercell; MD uses a 10 fs timestep. Temperature: 773 K (500 °C) for equilibration and the 1 ns tensile deformation stage. Thermostat: N/A — explicit algorithm (Berendsen, Nosé–Hoover, etc.) not restated in this wiki summary—see pdf_path Methods. Atomic iodine is placed in the vapor region (radiolytic-release idealization); the effective iodine partial pressure is inferred from the equilibrium cell volume and iodine count (0–12 MPa bracketing pellet-gap fission-gas conditions). Uniaxial strain is applied perpendicular to the grain boundary at \(10^8\,\mathrm{s}^{-1}\) (10% strain over 1 ns), at constant 500 °C, with stress computed and normalized to the slab cross section every 50 steps to build stress–strain curves vs. iodine exposure. Yield is identified near \(\sim\)4% strain (Figure 4 in the article); Additional file 1 tabulates raw stress–strain data. Replica / enhanced sampling / electric field: N/A — not used. Barostat: in-plane 0 MPa equilibration is followed by NVT-style constant-volume deformation reporting (see article for how lateral stress is handled during uniaxial strain); NPT during the mechanical test — N/A — confirm wording in pdf_path.
2 — Force-field training¶
The Zr–I extension builds on prior molecular and solid-state DFT models of Zr–I aggregation and related states (article refs. [22–24]); parent parameters are the broader ReaxFF framework with element coverage extended for Zr–I chemistry. QM reference data feed the training/bond-order refinements (see letter). Training set / optimization: the communication points to those QM references rather than reprinting every training geometry here—full training tables: pdf_path + SI. External validation: stress–strain responses are compared against the iodine-free baseline and iodine-loaded cases in the model; experimental PCI details are discussed qualitatively in framing.
Findings¶
The letter reports that the maximum yield stress of the 15° high-energy grain boundary falls as iodine exposure increases. Relative yield stress drops sharply at low iodine pressure (up to \(\sim\)20% of the iodine-free yield by \(\sim\)0.5 MPa), then more gradually, reaching up to \(\sim\)80% reduction at \(\sim\)11 MPa when pressure is used as the control variable. At low pressure, iodine is largely chemisorbed (with I\(_2\) dissociation treated as barrierless on Zr in the discussion), so the authors replot yield stress vs. chemisorbed iodine coverage (iodine atoms per surface Zr) and find an approximately linear weakening—supporting adsorption-enhanced decohesion tied to zirconium iodide surface film formation rather than bulk grain-boundary iodine loading. Trajectory inspection near yield shows crack initiation where the grain boundary meets the free surface, with ingress of iodine at the yield point; at higher pressure a surface phase transformation (intermingled Zr and I) adds driving force for opening. The authors summarize that chemisorption-mediated iodide films reduce the stress needed to initiate failure at the GB–surface junction, consistent with an ISCC framing.
Limitations¶
- A single boundary geometry and simplified environmental chemistry do not exhaust real PCI conditions (other halides/irradiation effects).
- Classical reactive FF cannot capture full electronic-structure details of bond rupture under all conditions.
Relevance to group¶
Demonstrates ReaxFF applied to nuclear materials degradation pathways with van Duin coauthorship.