Curie temperature of metallic ferromagnets Gd and Ni as a function of number of layers compared and contrasted

Experimentally it is known for metallic Gd that, if the ferromagnetism is studied as a function of the number of layers, say nL, the outermost layer has a Curie temperature TC some 20 K higher than its bulk value of around 290 K. This is in marked contrast to the measured form of TC(nL) in Ni, showing that TC(nL) approaches its bulk value from below, with substantial reduction of the bulk TC occurring for nL less than 10 atomic layers. Besides this marked difference of behaviour, we also anticipate that above the Curie temperature the two metals will behave in such a way that Gd will have disordered local moments while Ni will not. Following this, we use analogies with known results on Friedel oscillations to assess the changes in the range of RKKY interactions in Gd due to mesoscopic size effects and dimensionality. Possibly, surface reconstruction may need consideration in quantitative work. Here, we propose the RKKY exchange as the essential mechanism in discussing the behaviour of TC in Gd, and we compare this proposal with an earlier theoretical study of Freeman and Wu (1991) and later works. In contrast, in Ni, we appeal to the concept of SF (SF) temperature TSF introduced into the phenomenology of bulk transition metal ferromagnets by G.G. Lonzarich and L. Taillefer. [J. Phys. C: Condens. Matter 18 (1985), p. 4339 and propose its sensitivity to the number of atomic layers nL in Ni. Reference to experiments on Ni in relation to Fe and Co then suggest that variation of nL could readily reduce TC(nL) down to 300 K and perhaps even less for a monolayer of Ni. Specific heat measurements on mesoscopic ferromagnetic Ni may be revealing in the present context.