We suggest a mechanism causing large positive magnetoresistance (MR) in dilute magnetic semiconductors when hopping via nonmagnetic donor impurities dominates the conductivity. The effect is due to the increase in the characteristic width σ of the donor energy distribution with increasing magnetic field B, caused by exchange interactions between magnetic Mn atoms and the electrons localized on nonmagnetic Cl donor impurities. Using general scaling arguments based solely on the dependencies of hopping rates on temperature and on the energies of hopping sites we show that this mechanism accounts quantitatively for our experimental data on MR in n-type Zn 1−x Mn x Se:Cl. We suggest a method for extracting the dependence of σ on magnetic field from the MR data. The mechanism explains the experimentally observed universal dependence of the MR effect on the ratio B/T at different temperatures T under the premise that transport is due to the nearest-neighbor or Mott hopping mechanism.

No abstract

Articles you may be interested inThe effects of group-I elements co-doping with Mn in ZnO dilute magnetic semiconductor Electron carrier concentration dependent magnetization and transport properties in ZnO:Co diluted magnetic semiconductor thin films

Composition and carrier-concentration dependence of the electronic structure of In y Ga 1 − y As 1 − x N x films with nitrogen mole fraction of less than 0.012 J. Appl. Phys. 98, 093714 (2005); 10.1063/1.2127126Effect of lithium ion irradiation on the transport and optical properties of Bridgman grown n-type InSb single crystals

A series of samples consisting of alternating stripes of ZnO grown by molecular-beam epitaxy (MBE) and radio-frequency (rf) sputtered Ga-doped ZnO stripes was laterally microstructured with a self-aligned pattern transfer method. We measured as a function of temperature the Seebeck coefficient S and the electrical resistivity r in-plane of the samples with the transport direction perpendicular to the stripe direction. Throughout the series the bar width and hence the number of interfaces was kept constant, but the interface profile was varied yielding different interface lengths and geometries. The dependence of S, r and the power factor S 2 /r on the interface length at room temperature were simulated using an empirical network model and it was demonstrated that the macroscopic transport coefficients are very sensitive to the interface region and that even this rather simple modelling yields useful information about the interface region.