SGER: New Magnetic Tunnel Junctions Using Semiconductor Sandwich Layers with Controllable Band Gaps

Iowa State University, Ames IA

Investigators

Abstract

D.C.Jiles and J.E.Snyder Progress beyond data storage densities of 100 Gbits/in2 requires a revolution in materials. Magnetic tunnel junctions provide a possible solution and are expected to form an essential component of future magnetic disk drive read heads, part of the $50 billion/year hard disk drive industry. Tunnel junctions with a R/R of 30% have been announced in the last year. However there are serious problems with the current generation of tunnel junctions, which are based exclusively on an alumina barrier layer. These layers are now down to 0.7 nm in thickness and still the resistances of the tunnel junctions are too high for the intended applications. Essentially the alumina tunnel junction has reached its ultimate performance limits and there is a concern if alumina tunnel junctions will ever be able to be used in read heads and MRAM because of the high resistance. Therefore it is time for new tunnel junction materials to be brought forward based on alternative barrier layers. Recently studies of alternative insulator materials by Fert et. al. showed that the barrier layer strongly affects the spin polarization in the magnetic layers. Freitas et al. have studied the effect of nitrogen additions to the alumina which changed the barrier height and resulted in an increase in tunnel magnetoresistance from 22% to 25% as the composition changed from Al2O3 to AlN. However none of this has addressed the central problem of the overall resistance of the devices. In our proposed SGER we intend to make the radical change of completely replacing the alumina with other semiconductors. Of course this is a high risk endeavor, but the enormous benefits to the magnetic data storage industry that will accrue if the project is success make this a worthwhile exploratory investigation.