Kondo effect and valence fluctuations ...

Transition-metal (TM) and rare-earth (RE) condensed-matter compounds are known to present remarkable properties, such as Kondo, intermediate-valence, or heavy-fermion behavior. These phenomena are intrinsically related to the localized character of the d or f electrons and to their interactions with the delocalized conduction-band states. They reflect the competition between electron delocalization and local charge fluctuations. A typical manifestation of the unconventional properties of such strongly correlated systems is the presence of small energy scales in the excitation spectrum that lead to striking low-temperature properties.

The size reduction effects in clusters and nanoparticles can drastically modify and eventually suppress these phenomena due to the discreteness of the energy spectrum and to the changes in the number of electronic states near the Fermi level, or even as a consequence of a change in the underlying lattice structure.  For example, one would like to understand the changes occurring in the Kondo screening as the Kondo cloud of the solid in confined into a small nanoparticle.  

Electron-correlation effects and temperature dependent properties of doped metal clusters have been investigated in the framework of the Anderson model by using variational approaches and exact diagonalization methods. A global optimization of the cluster structure, taking into account all possible topologies and impurity positions, reveals a remarkable interplay between the cluster structure and the magnetic behavior. The dependence of the singlet-triplet spin gap ?E as a function of size, band-filling, cluster structure, and impurity position can be interpreted in terms of the environment-specific sd hybridizations and conduction- electron spectrum. Exact calculated finite-temperature properties (e.g., specific heat, effective impurity moment, and magnetic susceptibility) and many-body spectral densities allow us to identify the characteristics of a finite-size equivalent of the Kondo effect and to quantify its dependence on size, structure, and impurity position, for example.

This project is developed in collaboration with Dr. J. L. Ricardo-Chávez from Instituto Potosino de Investigación Científica y Técnica (San Luis Potosí, Mexico) and is supported by the DAAD-CONACyT (Mexico) international exchange program PROALMEX.