Density waves in atomic necklaces

Abstract

O (1D) systems have always captured the imagination of physicists. In the ultimate 1D limit, a crystal is nothing but a string of atoms, akin to a necklace of tiny pearls. Most physics graduates undoubtedly associate 1D systems with the simple models used in introductory level quantum mechanics or statistical mechanics courses. Indeed, 1D model systems represent pedagogical and often most useful illustrations of the basic physics principles that govern more realistic condensed matter systems. These principles tell us, for instance, that an ideal atomic chain with a single metallic band should be unstable with respect to a Peierls distortion. A Peierls distortion is a condensed state of the 1D electron gas whose formation is triggered by the strong coupling between electrons and quantized lattice vibrations (phonons). The upshot is that such a wire would not conduct electricity at low temperature. Alternatively, atom wires could display exotic many-body physics. It has been predicted that, due to strong electron-electron interactions in 1D, the concept of single electrons should even break down and that it is only meaningful to discuss their collective behavior (i.e., excitations). Such a state of matter is referred to as the Luttinger-Tomonaga liquid. But even though the theory of ideal 1D systems has advanced much, experimental realization of truly 1D condensed matter systems remains a daunting task; it is impossible to suspend a long string of atoms in free space. Nonetheless, due to the progress in nanoscience, it is now possible to synthesize and tailor novel nanophase materials that mimic the ultimate 1D limit quite closely. As a result, crucial questions regarding the physics of atom wires can be experimentally addressed. For instance: would a string of atoms conduct if it is laid down on a substrate? How can the often detrimental effects of the inevitable structural imperfections or those of thermodynamic fluctuations be alleviated? Or perhaps the most fundamental and practical issue: would atom wires actually be thermodynamically stable? Experimental approaches toward creating 1D systems are often classified as ‘top down’ or ‘bottom up’ fabrication. The top-down procedure refers to artificial (lithographic) creation of 1D nanostructures. A natural way to produce atom wires is a bottom-up method that employs the most common 1D line-defects on surfaces, namely atomic step-edges, as a template for self-assembly. Moreover, artificially created nanostructures are generally unstable, whereas self-assembled (nano-) structures instead can be thermodynamically stable, just because the driving force in their formation is not a kinetically enforced low-dimensional ordering, but purely a drive toward the lowest energy state [1,2]. We use the symmetry breaking properties of surfaces and the reconstructions of their topmost atom layers to realize self-assembled macroscopic arrays of atom wires. To this end, we have used vicinal (stepped) Si surfaces. Besides the thermodynamic stability inherent to self-assembly based fabrication, an extra advantage of this approach is that the spatial access to these atom wire arrays allows for a detailed investigation of electronic instabilities in these wires, even at a local scale. The atom wires can reach macroscopic lengths on vicinal surfaces; their length is only limited by occasional step edges crossing the wires. On the other hand, producing atom wires via this bottom-up method inevitably introduces (intrinsic) defects. This should not necessarily be viewed as a drawback, since it provides a much-needed opportunity to study the dramatic influence of imperfections in atom wires on their (in-)stabilities. Because they are located at surfaces, the wires can be studied conveniently using well-developed techniques such as Scanning Tunneling Microscopy and Spectroscopy (STM and STS), and Angle Resolved Photoemission Spectroscopy (ARPES).