Magnetism at the edge: New phenomena at oxide interfaces
- PDF / 828,955 Bytes
- 8 Pages / 585 x 783 pts Page_size
- 20 Downloads / 234 Views
Introduction With advances in thin-film growth, atomic-level control of interfaces between dissimilar materials has reached an unprecedented level of sophistication. Thanks to the adoption of MgO-barrier magnetic tunnel junctions in hard-disk readheads, wafer-scale mastery of oxide barriers a few atomic layers thick has been achieved. Coupled with the broad compatibility of the oxygen sublattices of different materials and the sensitivity of charge, spin, and orbital degrees of freedom on the atomic structure of the interface, this technical control has made oxide interfaces a playground for exploring phenomena that are absent in the bulk constituents.1–3 Many scientists are mesmerized by the unexpected phenomena that are emerging at the oxide interface. The emergence and modification of magnetism at an interface is the topic of this overview. We consider both thin films and multilayers of oxides of d-block elements, but also thin films and interfaces between oxides with no magnetic cations. Broken spatial symmetry, strain, and rotation of oxygen octahedra required to accommodate differing lattice constants, atomic, or electronic reconstruction to restore local charge neutrality, and the associated electric fields will influence the magnetic order, as well as the charge or orbital order appearing in an interfacial region that may be as little as one or two atomic layer thick.
These thin films and multilayers are usually grown on an oxide substrate by means of a physical vapor technique such as pulsed-laser deposition, sputtering, e-beam evaporation, or molecular beam epitaxy, where the deposition temperature and oxygen pressure are far from ambient. A cap layer (e.g., Au) may be needed to protect the structure in air, but even a simple oxide thin film has two dissimilar upper and lower interfaces. Materials scientists enjoy ready access to measurement techniques capable of detecting magnetic moments with a sensitivity better than 10–11 Am2.4 The moment of a 5 × 5 mm2 monolayer of ferromagnetically aligned atoms, each with a magnetic moment of one Bohr magneton (µB) on a square lattice with a cell size of 0.4 nm, is 1.5 × 10–9 Am2, so it is perfectly feasible to measure the magnetic moment of a fraction of a monolayer. This high level of sensitivity is a mixed blessing. Oxide substrates lack the purity of electronic-grade silicon, and most are intrinsically diamagnetic with a magnetic susceptibility, defined as the ratio of magnetization M to applied field H, of order –10–5 (see Table I). In an applied field of 1 MAm–1 (equivalent to 1.25 Tesla), a 5 × 5 mm2 substrate, which is 0.5 mm thick, has an induced moment on the order of –10–7 Am2. Often this is greater in magnitude than the magnetic signal we wish to measure. The Curie law susceptibility of paramagnetic impurities, such as Fe2+ in MgO, which varies
J.M.D. Coey, Trinity College Dublin; [email protected] Ariando, National University of Singapore; [email protected] W.E. Pickett, University of California, Davis; [email protected] DOI: 10.1557/mrs.2013.283
1040
M
Data Loading...
