Project Intro.

This project aims at exploring novel avenues to uncover the intrinsic properties of metal atomic chains, involving mainly one experimental research group at the University of Geneva, with collaborations with other experimental and theoretical research groups in England and Japan. It combines expertise in low-temperature scanning probe imaging and spectroscopy and in growth by self-assembly of atomic-scale structures on silicon substrates.

General problem

One dimensional structures, of which atomic chains of metal atoms are perhaps the ultimate embodiment, host a range of fascinating and exotic physical phenomena such as Luttinger liquid behaviour, spin-charge separation, magnetic and charge ordering, quantum phase transitions and possibly novel properties not present in higher dimensional assemblies of the same atoms, such as superconductivity. In particular, reduction in the coordination of metal atoms on a surface leads to enhanced magnetic properties such as the magnetic anisotropy energy. It may even induce magnetism in nonmagnetic elements. Understanding these phenomena is of fundamental interest, and their control holds huge potential for a number of applications in spintronics, magnetic storage, and quantum computing, to name only a few.

Accessing the intrinsic properties of single atom chains in an experiment is very challenging. They cannot usually be made free standing, and there will always be some degree of coupling between the electronic states of the chain atoms and those of a substrate. The magnetic properties of surface metal atoms depend strongly upon the degree of interaction with the substrate which is especially large in metal-on-metal systems. We plan to grow metal atomic chains on a weakly-interacting semiconducting substrate to produce systems closer to that of an ideal isolated atomic chain.

Details of plan

We first grow Bi nanolines on the Si(001) surface, to form a 1D template. The structure of the Bi nanoline is well-understood, and there are no electronic states in the bulk band gap, which might interfere with the metal atomic chains. After growth of the Bi nanolines, the background silicon dimers are passivated with atomic hydrogen. According to the recipe used, we can either maintain the Bi dimers intact, or else strip them off, leaving the underlying Haiku core of the Bi nanoline exposed as a long, straight trench. Modelling suggests that deposition of metals such as Au, Ag, Cu, and Mn onto these templates will generate 1D chains of adsorbed metal atoms, with metallic character. An alternative route to 1D structures is to adsorb metal at elevated temperatures, so that the metal atoms can move subsurface. Preferential interstitial adsorption sites exist in the 5 and 7-membered rings of Si in the Haiku core, so that chains of transition metals, such as Fe or Mn may result.

Having fabricated the atomic chains, we will characterise them with STM and STS, and use Spin-resolved STM to study the electronic and magnetic properties.