Atomistic insights on the influence of pre-oxide shell layer and size on the compressive mechanical properties of nickel nanowires
Summary¶
Metal nanowires gain oxidation shells in ambient processing, so their mechanical response reflects coupled metal plasticity and amorphous oxide chemistry rather than ideal single-crystal scaling laws. This Journal of Applied Physics study uses ReaxFF molecular dynamics in LAMMPS to compare [001]-oriented nickel nanowires at three pristine diameters (about 5.0, 6.5, and 8.0 nm) against the same sizes after spontaneous oxidation in O\(_2\) at 300 K, which builds an approximately 1 nm amorphous Ni\(_x\)O\(_y\) shell (final diameters near 6.0, 7.6, and 8.9 nm). Uniaxial compression along [001] at 300 K applies engineering strain rates near 0.01% ps\(^{-1}\) (order 10\(^8\) s\(^{-1}\) in the article’s framing) to 14% strain while tracking virial stress. The central comparison is how native oxide lowers yield stress and changes size scaling relative to pristine wires, with emphasis on dislocation nucleation near the disordered oxide–metal interface.
Methods¶
Initial nickel cylinders are relaxed with conjugate-gradient minimization, NPT equilibration at 300 K without applied stress, then compressed under NVT at 300 K with Nosé–Hoover thermostats. The loading cell is periodic along the compression axis with free lateral surfaces so the wire can expand or contract laterally. Oxidation runs expose pristine wires to O\(_2\) under NVT until a self-limiting amorphous shell forms. Integration uses velocity Verlet with a 0.5 fs timestep; stresses derive from the virial expression. Interactions use a Ni/O/H ReaxFF parameterization with variable charges so metallic bonding, ionic oxide motifs, and covalent oxygen chemistry share one reactive framework.
MD engine and system sizes. Molecular dynamics uses LAMMPS with the cited Ni/O/H ReaxFF on nanowire cylinders of order 10³ atoms (pristine diameters ~5.0, 6.5, 8.0 nm; oxidized shells ~1 nm thick). Boundaries: periodic along [001] compression with free lateral surfaces. Timestep: 0.5 fs as stated above. Ensemble: NPT at 300 K for initial relaxation, then NVT compression at 300 K with Nose–Hoover thermostats. Barostat: NPT only during the zero-stress relaxation leg; compression uses NVT without additional hydrostatic pressure servo. External electric field: N/A. Enhanced sampling: N/A.
Findings¶
For pristine nanowires, yield stress and strain rise with diameter in the simulated diameter window, interpreted through surface-stress–mediated dislocation nucleation in defect-free cylinders—a trend that can differ from polycrystalline experiments where sources and statistics dominate. Oxide-coated wires yield at lower stress than their pristine counterparts and show reduced sensitivity to diameter, because the rough oxide–metal interface seeds plasticity earlier and homogenizes size effects. Stress–strain curves progress through elastic and nonlinear elastic regimes into plastic flow with zigzag character; the oxide shifts both the onset of plasticity and the post-yield hardening/softening pattern relative to bare nickel. Operators validating these trends should compare the reported stress definitions and engineering-strain conventions directly with papers/Aral_Ni_nanowires_JAP_2019.pdf, because high strain rates and nanometer dimensions amplify sensitivity to thermostat choice and stress averaging windows.
Limitations¶
High strain rates and idealized cylindrical geometry differ from experiments; oxidation protocol is simulation-specific but consistent with prior tensile study from the same group.
Relevance to group¶
van Duin-parameter ReaxFF Ni/O/H for chemomechanics of oxidized metal NWs.
Citations and evidence anchors¶
papers/Aral_Ni_nanowires_JAP_2019.pdf