Atomic oxygen chemisorption on carbon nanotubes revisited with theory and experiment
Evidence and attribution¶
Evidence
Prose below summarizes the peer-reviewed article (DOI 10.1021/jp310332y).
Summary¶
DFT (PBE, with additional BLYP/PBE0 comparisons) studies low-coverage atomic oxygen chemisorption on (10,0) and (8,4) single-walled carbon nanotubes, addressing ether vs epoxide site preferences and electronic structure near STM/STS probes. ReaxFF (C/O reactive FF) is used to compare classical predictions of structure and relative stability against the DFT benchmarks for these chemisorption motifs.
Methods¶
DFT calculations used the Kohn–Sham scheme with PBE exchange–correlation (with additional BLYP and hybrid PBE0 data reported in places), norm-conserving pseudopotentials, and a plane-wave cutoff of 100 Ry, as implemented in CPMD. Supercells for (10,0) and (8,4) SWNTs (~1 nm diameter) were converged with respect to size (models from 40 to 360 C atoms are discussed; much of the analysis uses 360 C in a periodic orthorhombic cell with stated lattice parameters in the article).
ReaxFF (C/O reactive parameterization cited in the paper) was used in the authors’ ReaxFF code for geometry optimization of chemisorbed O adducts, enabling classical comparison to DFT structures and relative energetics. NEB was used where barriers between ether and epoxide-related configurations were computed.
STM/STS measurements on purified HiPco SWNTs on Au(111) at 5 K used in situ atomic oxygen from a catalytic cracker; dI/dV maps were recorded to probe gap-state changes near chemisorption sites.
1 — MD application (thermally sampled RMD). Aside from CPMD/DFT and NEB pathway work, the authors’ ReaxFF usage summarized above is geometry optimization of O-chemisorbed adducts, not a reported long NVT/NPT production run with quoted timestep/thermostat in this wiki’s summary. N/A — treat fs timestep, ps/ns duration, Berendsen/Nosé–Hoover controls, barostat, and hydrostatic pressure targets as not stated here; read papers/ReaxFF_others/Kroes_JPC_2013.pdf if production molecular dynamics appears beyond the optimization narrative.
Findings¶
For an isolated O atom on the sidewall, PBE finds the open ether configuration more stable than closed epoxide by about 0.4 eV on (10,0), with supercell-size convergence requiring >200 C atoms (smaller cells artificially stabilize epoxide-like motifs). Electronic differences are strong: epoxide introduces characteristic gap states detected in STS, whereas ether-related spectra differ, matching the experimental dI/dV trends discussed in the paper.
Prior literature disagreements on low-coverage energetics are attributed mainly to undersized simulation cells that do not represent dilute oxygen on nanotubes. ReaxFF reproduces DFT structural trends and relative stability ordering for these O chemisorption motifs in the authors’ tests (with the usual limitations of a classical model for electronic spectra).
Sensitivity / outlook: Supercell size and functional choice (PBE vs hybrid) shift relative stability of ether vs epoxide motifs and gap features compared to STS.
Corpus honesty: extraction_quality: good here refers to metadata alignment; quantitative claims should still be checked against pdf_path pages and figures.
Limitations¶
DFT functional dependence (PBE vs hybrid) affects gaps and relative energies; classical FF cannot fully reproduce quantum electronic spectra.