Atomic and electronic structure of ideal and reconstructed α-Sn (111) surface

Zhong-Yi Lu, G. Chiarotti, Sandro Scandolo, Erio TosattiPhysical 1996 · DOI · arXiv

Abstract

We have carried out an ab initio study of α-Sn (111), with the aim of predicting and understanding its structure, reconstructions, and electronic states. We consider a variety of structural possibilities and optimize them by moving atoms according to Hellmann-Feynman forces. Our results indicate that the unreconstructed surface is highly unstable, while a variety of reconstructions compete for the true ground state. Extrapolated trends from diamond to Si to Ge are well borne out, with a 2×1 π-bonded chain reconstruction prevailing in the absence of adatoms, and a c(4×2) or (2×2) basic adatom-restatom unit reconstruction otherwise. Accompanying surface bucklings are in both cases larger than in Si and Ge, with consequently large ionic charge transfers predicted. Search for a β-Sn-like metallic state of the surface turned out to be inconclusive.

Introduction

In the study of semiconductor physics, the (111) surfaces of Group-IV elements—diamond, silicon, and germanium—serve as the primary models for understanding surface reconstruction. As the heaviest member of this group, gray tin (α-Sn) represents the logical conclusion of these structural trends. However, α-Sn also sits near a phase transition to the metallic β-Sn form, raising fundamental questions: does its surface follow the covalent reconstruction patterns of its lighter cousins, or does its proximity to a metallic phase lead to entirely different behaviors?

This research provides a foundational first-principles investigation into the α-Sn (111) surface, predicting "gigantic" physical distortions and a surprising shift toward ionic character on an otherwise covalent surface.

Computational Methodology

The study employed Density Functional Theory (DFT) within the Local Density Approximation (LDA). To model the electron-ion interactions, the researchers used norm-conserving pseudopotentials in the Kleinman-Bylander form. The surface was represented using a repeating slab model consisting of 12 atomic layers separated by a vacuum region to prevent interaction between slab images.

Structural optimization was achieved by calculating Hellmann-Feynman forces to iteratively relax the atoms from their ideal bulk-terminated positions until the total energy of the system was minimized. By comparing the total energies of various reconstruction candidates—including the π-bonded chain and various adatom configurations—the authors identified the most energetically favorable states and mapped their electronic band structures.

A primary contribution of this work is the identification of a clear trend in vertical "buckling" across Group-IV elements. In the 2×1 π-bonded chain reconstruction, the vertical displacement between surface atoms increases significantly with atomic weight:

  • Diamond (C): 0 Å
  • Silicon (Si): 0.4 Å
  • Germanium (Ge): 0.8 Å
  • Gray Tin (α-Sn): 1.15 Å

This 1.15 Å buckling in α-Sn is described as "gigantic." Such extreme displacement is a direct consequence of the larger atomic size and the specific electronic requirements to minimize dangling bond energy in the heavier element.

Electronic States and Ionic Character

In lighter Group-IV semiconductors, surface bonding is predominantly covalent. However, the extreme buckling predicted for α-Sn (111) induces a significant charge transfer. Electrons migrate from the "lowered" atoms to the "raised" atoms. This creates a surface with a high degree of ionic character, which is highly unusual for a Group-IV element.

The electronic band structure calculations revealed that these reconstructions effectively open a gap in the surface states. By reconfiguring the dangling bonds into filled and empty states separated by an energy gap, the system lowers its total energy, explaining why the ideal (1x1) surface—with its unsaturated bonds—is highly unstable.

Preferred Reconstructions and Stability

The study evaluated several competing models for the ground state of the surface:

  1. π-bonded Chain (2×1): This structure is predicted to prevail in the absence of adatoms, following the trend of Si and Ge but with the aforementioned enhanced buckling.
  2. Adatom-Restatom Unit: The authors found that a c(4×2) or (2×2) reconstruction involving adatoms in the $T_4$ site is the most energetically stable configuration for an annealed surface. This provides a surface energy gain of approximately 0.33 eV per (1x1) cell relative to the ideal surface.

The Question of Metallization

Given that α-Sn is energetically close to the metallic β-Sn phase in bulk, the authors investigated whether the surface might spontaneously transform into a metallic overlayer. Their results were inconclusive. While a metallic state was searched for, the significant lattice mismatch between the diamond-structure α-Sn and the body-centered tetragonal β-Sn makes a metallic surface layer energetically unfavorable at low temperatures compared to the standard reconstructions.

Conclusion and Future Outlook

The ab initio study of α-Sn (111) confirms that tin continues and amplifies the structural trends observed in silicon and germanium. The prediction of "gigantic" buckling and significant ionic charge transfer marked a significant departure from the purely covalent models of lighter semiconductors.

Because no experimental data for the α-Sn (111) surface existed at the time of the study, the authors concluded with a call for experimental verification. They suggested that techniques such as photoemission or vibrational spectroscopy would be essential to confirm these dramatic structural predictions and the predicted energy gaps in the surface electronic states.

Summary updated

Abstract

We have carried out an ab initio study of \ensuremath{\alpha}-Sn (111), with the aim of predicting and understanding its structure, reconstructions, and electronic states. We consider a variety of structural possibilities, and optimize them by moving atoms according to Hellmann-Feynman forces. Our results indicate that the unreconstructed surface is highly unstable, while a variety of reconstructions compete for the true ground state. Extrapolated trends from diamond to Si to Ge are well borne out, with a 2\ifmmode\times\else\texttimes\fi{}1\ensuremath{\pi}-bonded chain reconstruction prevailing in the absence of adatoms, and a c(4\ifmmode\times\else\texttimes\fi{}2) or (2\ifmmode\times\else\texttimes\fi{}2) basic adatom-restatom unit reconstruction otherwise. Accompanying surface bucklings are in both cases larger than in Si and Ge, with consequently large ionic charge transfers predicted. Search for a \ensuremath{\beta}-Sn-like metallic state of the surface turned out to be inconclusive. \textcopyright{} 1996 The American Physical Society.
BibTeX
@inproceedings{lu1996atomic,
  author = {Zhong-Yi Lu and G. Chiarotti and Sandro Scandolo and Erio Tosatti},
  title = {{Atomic and electronic structure of ideal and reconstructed α-Sn (111) surface}},
  booktitle = {Physical},
  year = {1996},
  doi = {10.1103/physrevb.54.11769},
}
Figure