We formulate a theory of doped magnetic semiconductors such as Ga(1-x)Mn(x)As which have attracted recent attention for their possible use in spintronic applications. We solve the theory in the dynamical mean field approximation to find the magnetic transition temperature T(c) as a function of magnetic coupling strength J, carrier density n, and Mn density x. We find that T(c) is determined by a subtle interplay between carrier density and magnetic coupling.

We have investigated the structure and electronic properties of ferrimagnetic double perovskites, A 2 FeReO 6 (A= Ca, Sr, Ba). The A=Ba phase is cubic (Fm3m) and metallic, while the A=Ca phase is monoclinic (P2 1 /n) and nonmetallic. 57 Fe Mossbauer spectroscopy shows that iron is present mainly in the high-spin (S=5/2) Fe 3+ state in the Ca compound, while it occurs in an intermediate state between high-spin Fe 2+ and Fe 3+ in the Ba compound. It is argued that a direct Re t 2g -Re t 2g interaction is the main cause for the metallic character of the Ba compound; the high covalency of Ca-O bonds and the monoclinic distortion (which lifts the degeneracy of t 2g states) seem to disrupt the Re-Re interaction in the case of the Ca compound, making it non-metallic for the same electron count.

We present a thorough and quantitative comparison of double-exchange models to experimental data on the colossal magnetoresistance manganese perovskites. Our results settle a controversy by showing that physics beyond double-exchange is important even in La0.7Sr0.3MnO3, which has been regarded as a conventional double-exchange system. We show that the crucial quantity for comparisons of different calculations to each other and to data is the conduction band kinetic energy K, which is insensitive to the details of the band structure and can be experimentally determined from optical conductivity measurements. The seemingly complicated dependence of Tc on the Hund's coupling J and carrier concentration n is shown to reflect the variation of K with J, n and temperature. We present results for the optical conductivity which allow interpretation of experiments and show that a feature previously interpreted in terms of the Hund's coupling was misidentified. We also correct minor errors in the phase diagram presented in previous work.

We formulate a theory of double perovskite coumpounds such as Sr2FeReO6 and Sr2FeMoO6 which have attracted recent attention for their possible uses as spin valves and sources of spin polarized electrons. We solve the theory in the dynamical mean field approximation to find the magnetic transition temperature Tc. We find that Tc is determined by a subtle interplay between carrier density and the Fe-Mo/Re site energy difference, and that the non-Fe same-sublattice hopping acts to reduce Tc. Our results suggest that presently existing materials do not optimize Tc.Identification of a ferromagnet with high spin polarization at room temperature and stable surface properties is an important goal in the field of magnetic materials. Such a system would allow, for example the fabrication of 'spin-valve' devices of greatly improved efficiency for magnetic field sensing [1], the development of new magnetic recording media [2], and perhaps the construction of improved sources of spin-polarized electrons for 'spintronic' applications [3]. One promising family of materials are the 'double perovskites' [4][5][6]. These are compounds of chemical formula ABB ′ O 6 , with A an alkaline earth such as Sr, Ca or Ba, and B, B ′ two different transition metal ions. Double perovskites in which B is F e and B ′ is M o or Re are of particular recent interest because they seem [5] to be metallic ferrimagnets with very high magnetic transition temperatures and highly spin-polarized conduction bands. However, neither the physics nor the materials science of these compounds is yet well understood. The magnetic transition temperature and whether the ground state is metallic or insulating vary as A is changed from Ba to Sr to Ca [7]. Mis-site (B − B ′ ) disorder has a pronounced effect [8].In this paper we take a step towards a theoretical understanding for these materials. We derive a many-body Hamiltonian, using band theory calculations [5] to fix important parameters. We calculate the magnetic transition temperature T c , and determine how different material parameters affect it. Our results should provide guidance in attempts to design double perovskite materials with improved properties, and an appropriate starting point for calculations of other properties.Double perovskites have a crystal structure which generalizes the familiar ABO 3 perovskite structure by hav-ing two B-site ions, which in the ideal structure alternate in a simple two sublattice pattern. The band theory has been determined [5]. The conduction bands are derived from transition metal B-site t 2g d-orbitals, in agreement with quantum chemical considerations [7]. There are six conduction bands per spin direction per unit cell; roughly, one triplet arises mainly from the d xy,xz,yz orbitals on the F e and the other from the same orbitals on Mo/Fe. The occupied bands are fully polarized at T = 0.Because the near-fermi-surface bands are derived from transition metal d-orbitals, we argue that a simple tightbinding parametrization of the band theory is adequate. We therefore mod...