Ferromagnetism at the edges of the stacked graphitic fragments: an ab initio study

Hosik Lee, Noejung Park, Young‐Woo Son, Seungwu Han, Jaejun YuChemical 2004 · DOI

TL;DR

This paper investigates the origin of ferromagnetism in pure carbon systems, specifically in graphitic fragments. Using first-principles calculations, they demonstrate that ferromagnetism is stable at the edges of isolated graphitic fragments, but this magnetism disappears upon stacking these fragments due to interlayer interactions. Crucially, they identify exposed "dangling bonds" at the edges as the essential ingredient for sustained ferromagnetism in stacked graphitic structures.

What Problem Does This Paper Address?

The existence of intrinsic ferromagnetism in carbon-based materials has been a long-standing question, fueled by experimental observations in materials like doped fullerenes and highly oriented pyrolytic graphite. While some experiments suggest stable ferromagnetic states, others caution that these signals might arise from extrinsic magnetic impurities. Theoretical studies have pointed to localized electronic states at the zigzag edges of graphitic fragments as a potential source of magnetism. However, it remained unclear whether these edge-localized spins could contribute to the magnetism of larger, stacked graphitic structures, and what role, if any, hydrogen passivation plays in this phenomenon. This paper aims to clarify the role of edge structure and stacking on the magnetic properties of graphitic fragments.

Methods

The researchers employed first-principles density-functional theory (DFT) calculations using a plane-wave basis set and ultrasoft pseudopotentials. They utilized the Perdew-Wang parameterization for the exchange-correlation potential and a kinetic energy cutoff of 30 Ry. To model graphitic fragments, they used one-dimensional strips with zigzag edges, employing a supercell method to allow for antiferromagnetic ordering along the strip. For stacked fragments, they fixed the interlayer distance to that of bulk graphite (3.335 Å) and included vacuum regions to prevent spurious interactions. They investigated various spin configurations, including ferromagnetic (FM-F, FM-A) and antiferromagnetic (AF-E), as well as spin-unpolarized (PM) states. The calculations were performed using the PWSCF package.

Datasets

Not discussed in the paper. The study relies on theoretical calculations rather than experimental datasets.

Key Insights and Ideas

  • Dangling Bonds are Key: The most significant finding is that exposed dangling bonds at the edges of graphitic fragments are indispensable for sustained ferromagnetism in stacked structures.
  • Hydrogen Passivation Kills Edge Magnetism: While isolated graphitic strips with hydrogen-passivated zigzag edges exhibit stable ferromagnetism, this magnetism vanishes upon stacking these fragments. This highlights the sensitivity of edge magnetism to the local atomic environment.
  • Stacking Reorients Spins: Interlayer interactions in stacked graphitic fragments with dangling bonds favor specific spin orientations. AB stacking promotes ferromagnetic ordering between layers, while AA stacking favors antiferromagnetic ordering.
  • Edge States vs. Dangling Bonds: The paper distinguishes between the magnetism arising from edge-localized π-orbitals (which are susceptible to stacking effects) and the magnetism from dangling bonds (which are more robust to stacking).
  • Implications for Experimental Observations: The findings suggest that experimental observations of ferromagnetism in pure carbon systems might be explained by the presence of exposed dangling bonds, potentially in structures with hollow spaces or defects, rather than solely relying on hydrogen-passivated edges.

Key Results

  • Isolated graphitic strips with hydrogen-terminated zigzag edges show stable ferromagnetism with an exchange splitting of approximately 0.2 eV.
  • Graphitic strips with dangling bonds at the edges exhibit multiple self-consistent spin solutions, with ferromagnetic ordering (FM-F and FM-A) being energetically favored over antiferromagnetic (AF-E) and spin-unpolarized (PM) states.
  • The magnetic moment of the FM-F and FM-A configurations in isolated strips is substantial, exceeding 1 µB per edge atom, due to contributions from both dangling bond states and edge-localized π states.
  • Upon stacking, the ferromagnetism in hydrogen-passivated graphitic fragments disappears immediately, even with just two layers.
  • In contrast, stacked graphitic fragments with dangling bonds retain their magnetic moments. AB stacking is found to be more stable than AA stacking, with AB-FM being the most stable configuration.
  • The exchange splitting of dangling-bond states (Δσ) remains significant in stacked structures, while the exchange splitting of π-states (Δπ) is more affected by stacking.

Limitations and Future Work

The authors acknowledge that their study focuses on idealized graphitic strips with zigzag edges. They suggest that more complex edge geometries and the presence of defects, as might be found in proton-irradiated graphite, could lead to different magnetic behaviors. Future work could explore these more realistic scenarios and investigate the role of different types of hydrogen passivation in more detail. The paper also hints that the observed ferromagnetism in proton-irradiated graphite might be explained by hollow spaces within the graphite, retaining ferromagnetically interacting dangling bonds, which warrants further investigation.

Summary updated

BibTeX
@inproceedings{lee2004ferromagnetism,
  author = {Hosik Lee and Noejung Park and Young‐Woo Son and Seungwu Han and Jaejun Yu},
  title = {{Ferromagnetism at the edges of the stacked graphitic fragments: an ab initio study}},
  booktitle = {Chemical},
  year = {2004},
  doi = {10.1016/j.cplett.2004.09.069},
}
Figure