The Kondo effect in ultracold atoms

点击翻译按钮获取中文摘要

The simplicity of impurity models, combined with the often fundamental importance of the impurity on the emergent properties of the system, makes them a fruitful starting point for studies of open questions at the forefront of physics. We propose to implement the Kondo impurity model with ultracold atomic mixtures as experimental benchmark for open questions such as dynamical properties of highly correlated systems as well as the emergence of non-Fermi liquids. The 'single-channel' Kondo model describes the antiferromagnetic coupling of a localized impurity to a sea of spin-1/2 Fermions. The impurity can be treated perturbatively at high temperatures, but for low temperatures the impurity binds a Fermion into a singlet. In the ground state, the impurity is therefore entirely screened from the other Fermions by the Kondo screening cloud. The transition in the region of the Kondo temperature cannot be described by perturbative approaches and is a standard benchmark problem for non-perturbative methods. The situation changes dramatically in the 'multi-channel' Kondo model, where the impurity is coupled to more than one channel of spin-1/2 Fermions. In this case, the ground state is the superposition of spin singlets, each formed by the impurity and a Fermion from one of the channels. This degeneracy leads to strong quantum correlations between the channels, competing for the singlet formation with the impurity. These quantum correlations can even manifest themselves macroscopically in the formation of a non-Fermi liquid state. The excitations are losing their fermionic character in such a state. The emergence of these new quasi-particles and the prediction of their properties remains an open problem of fundamental physics. The goal of this proposal is to establish atomic mixtures as a new platform for investigating the Kondo model. The experimental parameters of these extremely clean and tunable systems enjoy a high versatility and they are now established as a benchmarking platform for a variety of model systems. Despite this attractiveness, no experimental results have been reported so far on the Kondo model. The multi-channel Fermi sea will be realized by a large cloud of Fermions with large spin. They will be trapped in a weak optical trap such that the spin is free to flip. The spin impurity will be realized by a few bosonic atoms that are localized in the ground states of a deep optical trap, such that its motion is frozen out. The exchange interaction will arise due to spin-changing collisions between the Bosons and the Fermions. We expect the interaction strength to be sufficiently high to obtain a Kondo temperature in the order of the Fermi temperature, which is experimentally accessible. The Kondo effect will be observed through local spin correlations between the Bosons and Fermions. We will detect non-Fermi liquid behavior through the non-fermionic propagation of correlations after a quench.