Do nickel and iron catalyst nanoparticles affect the mechanical strength of carbon nanotubes?
ReaxFF molecular dynamics in LAMMPS is used to tensile-test defect-free single-wall carbon nanotubes in contact with Ni and Fe nanoparticles in metallic and oxidized forms. Simulations resolve stress–strain behavior, bond-order evolution, and failure pathways to explain how catalyst particles that are essential for CNT growth can weaken the tubes.
Summary¶
Reactive MD with ReaxFF explores how Ni and Fe catalyst nanoparticles (pure and oxidized) alter the tensile strength and failure mechanism of defect-free single-wall CNTs. Simulations use LAMMPS in the NVT ensemble with a Nosé–Hoover thermostat at 300 K, an initial timestep of 1 fs with restarts at selected strains using 0.5 fs to reduce premature bond cleavage from numerical noise, stress from stress/atom, and a fixed nanotube wall thickness of 0.34 nm, with multiple tubes per nanoparticle case for statistics.
The framing problem is that CVD growth often leaves Ni or Fe catalyst particles embedded in or attached to otherwise pristine tubes, so understanding whether metallic versus oxidized catalyst residues weaken C–C bonds—and by which chemistry—guides interpretation of measured nanotube strengths and failure statistics.
Methods¶
1 — MD application (tensile failure with catalyst contact). ReaxFF trajectories are integrated in LAMMPS in the NVT ensemble with a Nosé–Hoover thermostat at 300 K; stresses/strains come from stress/atom. Oxidized Fe/Ni nanoparticles are prepared with hybrid GC-MC + MD oxygen insertion under a µVT oxygen reservoir, including NPT segments (10 ps, 300 K, 0.1 MPa) to equilibrate oxides before contact with (10,0) single-wall CNTs (defect-free, ~0.34 nm wall thickness as fixed in the protocol). Tensile tests use an initial 1 fs timestep, restarting at strains 0.08–0.12 with 0.5 fs to reduce catastrophic bond cleavage from integration noise; ≥6 replicas per case are averaged. System sizes follow the Extreme Mechanics Letters construction tables (full atom counts in PDF). PBC: three-dimensional PBC for the supercells described in the article. Barostat during tensile pulls: N/A — tensile production uses NVT without hydrostatic barostat control. Electric fields / enhanced sampling: N/A — not used beyond the stated GC-MC oxidation pretreatment.
Findings¶
Outcomes: pure metal (Ni, Fe) nanoparticles lower CNT failure stress/strains more strongly than oxidized counterparts at comparable oxygen content; among oxides, Fe-oxide contacts are more damaging than Ni-oxide for the same oxygen loading. Mechanisms: Ni failures concentrate on C–C bonds directly interacting with Ni, whereas Fe cases include weakening of C–C bonds not directly bonded to Fe.
Comparisons: trends are interpreted relative to oxidation state and element (Ni vs Fe) for industrially relevant CVD catalyst residues.
Sensitivity: oxygen content and metal identity shift both failure strain and bond-order evolution along the stress–strain curve.
Limitations: the ingested file is an Elsevier proof; compare final VOR pagination. Rates at 300 K tensile MD are not direct CVD growth maps.
Corpus honesty: protocol numbers above are from papers/Ostadhossein_Ni_CNT_ExtremeLetters_2018.pdf; confirm any updated values against the published article.
Limitations¶
The ingested PDF is an Elsevier proof; final pagination and minor edits may differ from the version of record.
Relevance to group¶
Adri C. T. van Duin is a co-author; study highlights ReaxFF for mechanical failure of carbon nanostructures in contact with transition-metal catalysts.