The effective Kondo model for a trimer on a metallic surface

TL;DR

This paper derives an effective Kondo model for a three-atom trimer on a metallic surface, explaining how the Kondo temperature varies with atomic distances and aligning with experimental observations.

Contribution

It introduces a generalized Schrieffer-Wolf transformation to derive a Kondo model for a trimer, connecting microscopic parameters to Kondo behavior.

Findings

Kondo temperature is negligible for certain atomic arrangements.

Maximum Kondo temperature occurs at specific interatomic distances.

Results match experimental data for Cr trimers on Au(111).

Abstract

I consider a Hubbard-Anderson model which describes localized orbitals in three different atoms hybridized both among themselves and with a continuum of extended states (ES). Using a generalized Schrieffer-Wolf transformation, I derive an effective Kondo model for the interaction between the doublet ground state of the isolated trimer and the ES. For an isoceles trimer with distances a, l, l between the atoms, the Kondo temperature is negligigle for l < a and has a maximum for finite l > a when a is small. The results agree with experiments for a Cr trimer on Au(111). [T. Jamneala et al., Phys. Rev. Lett. 87, 256804 (2001)].