Is room temperature ferromagnetism possible in K-doped SnO2?

Ab initio studies have theoretically predicted room temperature ferromagnetism in crystalline SnO2, ZrO2 and TiO2 doped with non magnetic element from the 1A column as K and Na. Our purpose is to address experimentally the possibility of magnetism in both Sn1-xKxO2 and Sn1-xCaxO2 compounds. The samples have been prepared using equilibrium methods of standard solid state route. Our study has shown that both Sn1-xCaxO2 and Sn1-xKxO2 structure is thermodynamically unstable and leads to a phase separation, as shown by X-ray diffraction and detailed micro-structural analyses with high resolution transmission electron microscopy (TEM). In particular, the crystalline SnO2 grains are surrounded by K-based amorphous phase. In contrast to Ca: SnO2 samples we have obtained a magnetic phase in K: SnO2 ones, but no long range ferromagnetic order. The K: SnO2 samples exhibit a moments of the order of 0.2 {mu}B/K /ion, in contrast to ab-initio calculations which predict 3{mu}B, where K atoms are on the Sn crystallographic site. The apparent contradictions between our experiments and first principle studies are discussed.