Theory and application for chain formation in break junctions

We introduce a generic model for chain formation in break junctions by formulating separate criteria for the stability and producibility of suspended monoatomic chains based on total energy arguments. Using first-principles calculations [1], we apply our model to break junctions of $4d$ and $5d$ transition-metals (TMs), as well as Ag and Au, including the effects of spin-polarization and spin-orbit coupling. Thereby, we can explain the physical origin of the experimentally observed trend of increasing probability for the creation of long suspended chains in break junctions for $5d$-TMs at the end of the series [2] and suppressed chain elongation for $4d$ elements. Moreover, we clearify why the probability of chain elongation is greatly enhanced by the presence of oxygen in experiments with Au and Ag. Our model also allows us to make predictions on the ballistic transport properties of suspended chains. [1] Y.Mokrousov {\sl et al.}, Phys.\ Rev.\ B\ {\bf 72}, 045402 (2005). [2] R.H.M.Smit {\sl et al.}, Phys.\ Rev.Lett.{\bf 87}, 266102 (2001).