Recently a large positive magneto-resistance (MR) effect was discovered in semiconduct ฮฑ-silver selenide samples with a composition between "Ag 2,01 Se" and "Ag 2,33 Se" [1]. MR effects up to 120% have been observed at room temperature and magnetic fields of 5,5 T. The magnetic field dependence showed no evidence of saturation up to 5,5 T and was linear down to 10 mT. As known from the phase diagram, samples of ฮฑ-Ag 2ฯฉฮด Se with a silver excess higher than ฮด max 10 ฯช4 are heterogeneous. Therefore we prepared homogeneous ฮฑ-Ag 2ฯฉฮด Se samples with precisely defined silver excess within the phase field and heterogeneous samples with high silver excess up to "Ag 2,3 Se". Homogeneous samples of ฮฑ-Ag 2ฯฉฮด Se show a positive and small ordinary magneto-resistance (OMR) with quadratic field dependence. The dependence of the MR effect on the non-stoichiometry can be well understood on the basis of the defect structure of the material. For heterogeneous samples we find both a positive MR effect with linear field dependence and a negative MR effect, which saturates at higher fields. A change from one effect to the other is observed on thermal treatment and is most probably due to a change in the microstructure and related defect concentrations. In order to separate the influence of microstructure of the heterogeneous specimens from other parameters we started to prepare thin films of silver selenide and of silver selenide/silver multilayers by suitable methods. Results from these experiments are presented. Further measurements on other binary silver and copper chalcogenides are performed in order to explore the behavior of materials which are chemically closely related to Ag 2 Se. Typically, we observe OMR, but again find deviations from this simple behavior for metal-rich phases. Finally, we discuss a possible influence of an inhomogeneous/ heterogeneous distribution of excess metal on the MR properties on the basis of existing theories for the MR effect. This study is funded by the DFG within the project Ja 648/5-1/2.
[