No abstract

We have identified and characterized a charge-density wave transition ͑T c ϳ 260 ± C͒ in the lowcoverage a phase of the Sn͞Ge(111) interface both experimentally and theoretically. Charge ordering is accompanied by a structural distortion from ͑ p 3 3p 3 ͒R30 ± to ͑3 3 3͒ symmetry. Density-functional theory calculations are unable to correctly reproduce the observed ground state and, more importantly, indicate that Fermi surface nesting does not play a role in this transition. Both signal the importance of many-body effects in this system. Experiment and theory indicate that the Sn͞Ge(111) overlayer is fundamentally different from the Pb͞Ge (111) overlayer previously reported. [S0031-9007(97)04249-X] PACS numbers: 73.20.At, 61.16.Ch, 68.35.Md, 71.45.LrA charge-ordered state, or charge-density wave (CDW), incorporates a symmetry-lowering periodic redistribution of valence charge driven by a reduction in the system's total electronic energy, resulting in a small periodic lattice distortion [1]. This phenomenon is most likely to occur in reduced dimensions [2,3], for example, in the layered perovskites [4]. The loss of coordination at a crystal surface might also be expected to invite CDW formation, but few genuine instances of surface charge-density waves have been reported to date. One clear example was recently discovered at the Pb͞Ge(111)-a vacuum interface [5]. This low-density overlayer transforms from the room-temperature (RT) ͑ p 3 3p 3 ͒R30 ± structure to the low-temperature (LT) ͑3 3 3͒ ground state. Charge ordering (ϳ0.5 Å corrugation) and accompanying lattice distortion occur gradually and reversibly with T c ϳ 220 ± C. Density function theory calculations indicate the chargeordered ͑3 3 3͒ structure is the ground state of this system [5]. The opening of a E g ϳ 65 meV band gap below T c (later confirmed by photoemission measurements [6]) must be the consequence of correlation effects [5]. We therefore conjectured that although enhanced electron-phonon coupling (enabled by Fermi surface nesting) drives this transition, many-body interactions stabilize the ground state.Our motivation to study the a phase of Sn͞Ge(111) was threefold. First, seeking verification that Pb͞Ge(111) charge ordering is not an isolated quirk of nature resulting from Pb's exotic properties, we considered the other isostructural adsorbate/substrate combinations that exist. Al, Ga, In, Sn, and Pb all form the same low-density overlayer atop both Si(111) and Ge(111) [7,8]. However, overlayers composed of the trivalent species (Al, Ga, In) will probably not undergo a CDW distortion, as they are semiconducting [7] with an even number of electrons per unit cell at RT.Second, it is imperative to test several theoretically derived concepts, including Fermi surface nesting, and the idea that the CDW ground state is a Mott-Hubbard insulator [5]. Intuitively, the Sn overlayer bandwidth w should be less than that of the Pb overlayer, because the metallic r Sn is ϳ10% smaller than r Pb . For the same reason one expects the on-site Coulomb repuls...

First-principles electronic-structure methods are used to study a structural model for Ag/Si(111)3×1 recently proposed on the basis of transmission electron diffraction data. The fully relaxed geometry for this model is far more energetically favorable than any previously proposed, partly due to the unusual formation of a Si double bond in the surface layer. The calculated electronic properties of this model are in complete agreement with data from angle-resolved photoemission and scanning tunneling microscopy.The surfaces of silicon reconstruct in strikingly diverse ways. This diversity provides a rich proving ground for simple, physically intuitive ideas about the stability of semiconductor surfaces-ideas which are invaluable for understanding more complex dynamical phenomena such as growth, etching, and reactivity. Two such simple concepts-elimination of surface dangling bonds and relief of surface stress-explain the frequent appearance of elementary "building blocks" in silicon reconstructions. For example, dimers appear on Si (001) One reconstruction that remains controversial is the metal-induced M /Si(111)3×1 (where M =Li, Na, K, Ag, Mg), which is widely believed to have a single common structure. Starting from known building blocks, we and others have recently proposed two models for this reconstruction, both with low dangling-bond density and low surface stress [2][3][4][5][6]. First-principles total-energy calculations showed these models to be stable relative to previously proposed ones [6]. The more stable of the two, the extended Pandey chain model, was also consistent with scanning tunneling microscopy (STM) images, but the calculated surface-state band structure of both models was in serious disagreement with angle-resolved photoemission (ARPES) data [7].In this Letter, we examine theoretically a new model for the M :3×1 surface proposed very recently by CollazoDavila, Grozea, and Marks (CGM) on the basis of direct phasing of transmission diffraction data [8], and independently by Lottermoser et al. from surface x-ray diffraction and total-energy calculations [9]. First, starting from coordinates obtained by CGM, we further relax the atomic positions so as to minimize the calculated total energy. The resulting model is by far the most stable of any proposed to date. Second, we show how a nearly perfect "surface symmetry" of this reconstruction neatly resolves a long-standing puzzle regarding ARPES data for Li:3×1 (and may also explain the apparent insulating nature of Mg:3×1, which has an odd number of electrons per unit cell). Third, we show that this model derives its remarkable stability from the formation of a true Si double bond-a "building block" not seen on any other surface of silicon. Finally, we show that this model completely accounts for the appearance of the existing STM images (including the differences between M =Li and Ag).The model proposed by CGM was based on data from Ag:3×1 [8]. Previous experimental results from lowenergy electron diffraction [10], STM [11], and core level spe...

"Noncompensated n-p codoping" is established as an enabling concept for enhancing the visible-light photoactivity of TiO2 by narrowing its band gap. The concept embodies two crucial ingredients: the electrostatic attraction within the n-p dopant pair enhances both the thermodynamic and kinetic solubilities, and the noncompensated nature ensures the creation of tunable intermediate bands that effectively narrow the band gap. The concept is demonstrated using first-principles calculations, and is validated by direct measurements of band gap narrowing using scanning tunneling spectroscopy, dramatically redshifted optical absorbance, and enhanced photoactivity manifested by efficient electron-hole separation in the visible-light region. This concept is broadly applicable to the synthesis of other advanced functional materials that demand optimal dopant control.

Superconductivity is inevitably suppressed in reduced dimensionality1-9 . Questions of how thin superconducting wires or films can be before they lose their superconducting properties have important technological ramifications and go to the heart of understanding coherence and robustness of the superconducting state in quantum-confined geometries 1-9 . Here, we exploit quantum confinement of itinerant electrons in a soft metal to stabilize superconductors with lateral dimensions of the order of a few millimeters and vertical dimensions of only a few atomic layers 10 . These extremely thin superconductors show no indication of defect-or fluctuationdriven suppression of superconductivity and sustain supercurrents of up to 10% of the depairing current density. The extreme hardness of the critical state is attributed to quantum trapping of vortices. This study paints a conceptually appealing, elegant picture of a model nanoscale superconductor with calculable critical state properties.

Ferromagnetic Mn5Ge3 thin films were grown on Ge(111) with solid-phase epitaxy. The epitaxial relationship between the alloy film and substrate is Mn5Ge3(001)//Ge(111) with [100]Mn5Ge3//[11̄0]Ge. The alloy films exhibit metallic conductivity and strong ferromagnetism up to the Curie temperature, TC=296 K. These epitaxial alloy films are promising candidates for germanium-based spintronics.

We present a variable temperature scanning tunneling microscopy and spectroscopy study of the Si(553)-Au atomic chain reconstruction. This quasi-one-dimensional system undergoes at least two charge density wave (CDW) transitions, which can be attributed to electronic instabilities in the fractionally filled 1D bands of the high-symmetry phase. Upon cooling, Si(553)-Au first undergoes a single-band Peierls distortion, resulting in period doubling along the chains. This Peierls state is ultimately overcome by a competing 3 CDW, which is accompanied by a 2 periodicity in between the chains. These locked-in periodicities indicate small charge transfer between the nearly 1=2-filled and 1=4-filled bands. The presence and the mobility of atomic-scale dislocations in the 3 CDW state indicates the possibility of manipulating phase solitons carrying a (spin, charge) of 1=2; e=3 or 0; 2e=3 . DOI: 10.1103/PhysRevLett.96.076801 PACS numbers: 73.20.At, 68.37.Ef, 71.10.Pm, 73.20.Mf According to the Mermin-Wagner theorem [1], thermodynamic fluctuations preclude the formation of a longrange ordered broken symmetry state in one dimension, except at T 0 K [2]. For all practical purposes, however, thermodynamic phase transitions may still be possible in finite size 1D systems. Furthermore, fluctuations are inevitably suppressed if the 1D chains are weakly coupled, or if the chains are coupled to a substrate [2,3]. Prototypical 1D metallic systems such as the transition metal trichalcogenides, organic charge transfer salts, blue bronzes, and probably all atomic Au-chain reconstructions on vicinal Si substrates exhibit symmetry breaking phase transitions at finite temperatures [4,5]. For a band filling of 1=n, the phase transition opens up a gap in the single particle excitation spectrum at wave vector k F =na, and the corresponding broken symmetry state adopts the new periodicity of =k F na, where a is the lattice parameter of the high-symmetry phase [4].Fractional band fillings other than half filling provide an interesting subset of 1D systems which often exhibit exotic physical phenomena. Depending on the relative magnitude of bandwidth and electron-electron interaction, charge density wave (CDW) states often compete with spin density waves, Mott insulating states, or a Luttinger liquid state. Atomic-scale STM observations of surface phase transitions provide important insights into the complexity of symmetry breaking phenomena in reduced dimensionality [6]. For instance, the recently reported 4 1-to-8 2 phase transition in quasi-1D indium chains on Si(111) [6] involves a gap opening in a complex triple band Peierls system, resulting in a doubling of the periodicity along the atom chains. Another recently discovered system with three fractionally filled bands is the Si(553)-Au surface. Angle-resolved photoemission spectroscopy (ARPES) [7] revealed three metallic bands, but despite theoretical efforts to understand the electronic structure [7,8], the atomic structure and real space location of the surface state orbitals remain...

Many quasi-one-dimensional ͑1D͒ materials are experimental approximations to the textbook models of Peierls instabilities and collective excitations in 1D electronic systems. The recently observed self-assembly of atom wires on solid surfaces has provided fascinating new insights into the nature of their structural and electronic instabilities, from both real-space and momentum-space perspectives. In this Colloquium, three of the most studied atom wire arrays are highlighted, all featuring multiple surface-state bands. One of these is made of indium atoms on a flat silicon ͑111͒ surface, while the two others consist of gold atoms on surfaces that are vicinal to Si͑111͒. The experimental and theoretical results are discussed with a focus on the detailed mechanisms of the observed phase transitions and on the role of microscopic defects.