Skip to content

Atom vacancies on a carbon nanotube: to what extent can we simulate their effects?

Curators

Body text summarizes the JCTC article identified in YAML; numeric barriers and structures should be checked against the publisher PDF.

Summary

This Journal of Chemical Theory and Computation article benchmarks classical interatomic models against spin-polarized density functional theory for intrinsic vacancies in a (10,0) zigzag single-wall carbon nanotube. The study focuses on single and double vacancies and reports relaxed structures, vacancy formation energies, and energy barriers for elementary processes such as reconstruction, migration, and coalescence. On the DFT side, calculations use CPMD with PBE and BLYP exchange–correlation functionals and include selected hybrid PBE0 repeats; on the classical side, the authors compare widely used carbon potentials including AIREBO, LCBOP (LCBOPI as primary, LCBOPII partly in Supporting Information), ReaxFF15, REBO (more fully in Supporting Information), and a Tersoff parameterization oriented toward amorphous carbon. Adri C. T. van Duin is a co-author, placing ReaxFF15 explicitly in a comparative validation setting rather than as a standalone claim of accuracy.

Methods

Reference system. (10,0) zigzag single-wall carbon nanotube models with single and double vacancies; quantities reported include relaxed structures, vacancy formation energies, and barriers for reconstruction, migration, and vacancy coalescence.

Static QM (DFT). Spin-polarized plane-wave DFT in CPMD with PBE and BLYP functionals, Γ-only sampling, and selected PBE0 repeats as described; lowest spin multiplicity (singlets) is enforced in the benchmark set noted in the article.

Classical potentials (energy landscapes). The same defect motifs are evaluated with AIREBO, LCBOPI (LCBOP family; LCBOPII partly in SI), ReaxFF15, REBO (SI), and Tersoff (discussed largely in SI) using the same tetragonal supercells (~359–360 atoms for most cases, 598 atoms for the slowly converging 5r8r5r-Z double vacancy). Formation energies use the carbon chemical potential definition printed in the article.

Auxiliary classical MD (motif sampling). The authors additionally report high-temperature AIREBO molecular dynamics on a large ~100,000-atom, ~1 µm tube model followed by simulated annealing to propose vacancy geometries that are subsequently relaxed in DFT; this exploratory MD is not the same object as the small-cell barrier tables.

Production reactive MD (ReaxFF). N/A — ReaxFF15 enters as static energies and barriers on the (10,0) benchmark cells, not as long finite-temperature ReaxFF MD production trajectories. The separate high-temperature AIREBO MD on a large tube followed by annealing is exploratory geometry sampling later relaxed in DFT; it is not the quantitative ReaxFF15 benchmark path. Timestep, thermostat law, NPT barostat, replica exchange, and electric fields: N/A — not applicable to the small-cell minimization/barrier workflows (and not separately tabulated for the scouting MD in the sections indexed for this note). PBC: 3D periodic tetragonal supercells for the DFT / classical small-cell work as stated in the article.

Findings

The central empirical result is lack of uniform agreement: vacancy-related structures, formation energies, and barriers differ markedly among DFT choices and among classical potentials, and no single classical model matches PBE benchmarks across all processes examined. ReaxFF15, AIREBO, LCBOP, Tersoff, and related schemes each show distinct strengths and failures depending on the vacancy process, underscoring that reactive and bond-order potentials require careful, property-specific validation for defective nanotubes. The authors frame the compilation as guidance for constructing or refining DFT-informed classical schemes for carbon nanostructures rather than as endorsement of a single potential for all vacancy dynamics.

The introduction additionally notes that vacancies can influence binding of adsorbates, oxidation, and coalescence phenomena in nanotubes, and that growth or fracture simulations that move atoms between gas-like and condensed environments especially depend on reliable classical energetics for defect rearrangements, motivating systematic benchmarking against spin-polarized DFT rather than isolated literature values. The authors also highlight exchange-correlation sensitivity within DFT itself by comparing gradient-corrected functionals with partial exact-exchange hybrid checks for selected quantities.

Limitations

Γ-point supercells and specific vacancy topologies; classical models inherit known transferability limits for charged/spin states and long-range relaxation.

Relevance to group

Adri C. T. van Duin co-authorship; directly contextualizes ReaxFF against DFT for defective CNT energetics.

Citations and evidence anchors

DOI 10.1021/acs.jctc.5b00292.

Reader notes (navigation)