Kondo interaction

CeAl2 represents a special case. It may order antiferromagnetically but its behavior appears to be markedly affected by the Kondo interaction of the Ce3+ ion. 

The distinction between Kondo metals, etc. and HF systems is fuzzy at best. As pointed out, the Kondo interaction is, among others, a basic ingredient of HF behavior. In a Kondo-lattice material one observes the effects of the Kondo interaction, for example on the magnetic properties, but very heavy quasiparticles are not formed and in consequence, the Sommerfeld constant is only slightly enhanced, That at least is the basis for a distinction we shall adopt. The hybridization between 4f and conduction electrons can lead to a hybridization gap in the density of states at the Fermi surface. The exact mechanism of gap formation is still under debate and also may vary from compound to compound. If a gap is present, one leaves the realm of Kondo metals and has, depending on the form of the gap (e.g., whether it is open in all crystallographic directions) and on its width, either a Kondo semimetal, semiconductor or insulator. The latter are certainly the most challenging class of Kondo compounds to understand. 

CEF levels of Ce (see sect. 5.1.1). Dalmas de Reotier et al. (1996) discuss in general the connection between CEF and Kondo interactions. They point out that the difference in T between the neutron and pSR measurements in CePt2Sn2 could be a straightforward result of CEF interactions. The absence of information on the full CEF level scheme does not allow a quantitative discussion of this aspect (for some more details, see the discussion of the HF compound YbAuCiLi). 

Fig. 157. Comparison of the temperature dependenee of experimental values of (solid symbob) and Fq (open symbob) for YbAuCu, with model caleulations. Left Predictions based on the Korringa interaction (k-f exchange) only with a coupling strength of yy/i(E p) = 0.37 (solid and broken lines). Cubic CEF splitting is taken into account. Right Predictions of the NCA formalism of Kondo interaction with a Rondo scale of 20 K and cubic symmetry for the Yb site (solid and broken lines). The dash-dotted and the dotted lines assume a trigonal distortion of the Yb site due to the presence of a stopped p with a term of -1.75 and -2.35, respectively. The muon-4f coupling constant was set to 208 MHz in all cases corresponding to the calculated value for the 4d interstitial stopping site. After Bonville et al. (1996).

Whether the NCA model is a proper description of Kondo interaction in the present case could be questioned since Kondo temperature and CEF splittings are roughly equal. In addition, magnetic correlations have been completely n lected. It is true that the discussion is of a temperature range far above 7n, but on the other hand, it is known (see, for example sect. 5.3) that pSR senses such correlations far into the paramagnetic regime. 

Detailed magnetoresistance measurements were performed for YbNiSn (Adroja et al. 1998, D Onofrio et al. 1991), emphasizing the influence of the Kondo interactions. 

The groimd-state properties of Kondo-lattice systems, for example, are expected to be very sensitive to external pressure. The latter may modify the strength of the exchange interaction J between the localized 4f and the conduction electrons and thereby the competition between the intersite (RKKY) and intrasite (Kondo) interactions. The balance between these effects has been theoretically described by Doniach (1977) for a magnetic phase diagram depending on the exchange parameter J. 

Discussing results on k of heavy-fermion systems one must distinguish two temperature regions a low-temperature one, T< Tr, and a high-temperature one, T> Tr. The Kondo temperature, Tr, corresponds to the delocahzation of f-electrons due to Kondo interaction with free electrons ... 

---