News

Parity matters in a wire

Return to C. Untiedt





Parity matters in a wire (23-7-2003)

(Observation of a length-dependent oscillation of conductance in atomic wires. R.H.M. Smit, C. Untiedt, G. Rubio-Bollinger, R.C. Segers, J.M. van Ruitenbeek. Phys. Rev. Lett  91,076805 (2003) and cond-mat/0303039)

The common experience tells us that the electrical conductance (the inverse of the resistance) of a macroscopic wire is dependent of its length. This can be easily understood when we think that the electrical current is carried by electrons and the longer these have to move, more chances they will have to find an obstacle that delays their movement. If we reduce the size of a wire below the characteristic length that an electron moves without finding such an obstacle, its resistance will no longer depend on the length of the wire in what is called the ballistic regime of the electrical conduction. However a recent collaboration of researchers of the AMC group of the University of Leiden (The Netherlands) and the Laboratorio de bajas temperaturas of the Universidad Autónoma de Madrid (Spain) have shown that this is not the case for an atomic chain.
The thinnest wire possible is the chain of atoms. These can be formed at the final stages of breaking a macroscopic wire and their conductance can be measured wile increasing their length atom by atom. In the case of chains made out of gold the conductance is close to the quantum of conductance and small variations in its value are observed wile pulling the chain. From these variations, the new research has shown dependence between the conductance of the wire and the number of atoms which form the chain. This time the dependence is of a new kind, in such a way that it depends on whether the number of atoms in the chain is even or if it is odd.
The even-odd behavior is a consequence of connecting the chain to the electrodes, the extremes of the chain will act as a barrier to the electrons in the chain and this will make the chain to work as an interferometer of electrons. This interference effect will be the responsible of the alternating conductance with the number of atoms in the chain. The conductance of a chain containing an even number of atoms will be generally smaller than the one containing an even number which will give the maximum conductance possible.
This behavior was predicted for metallic wires made of materials with a simple electronic structure were the conduction was carried by electrons of “s” type (electrons residing in a spherical electron cloud) but this research shows that this holds for all the metals which forms chains (Au, Pt and Ir). On top of that, for the last two which are not “s”-type, an unpredicted continuous decrease of conductance is also observed while increasing the chain length.
These new phenomena show how different things are when going to small.    
 
(More information at www.physics.leidenuniv.nl/sections/cm/amc/  and http://www.ua.es/personal/untiedt/ )