Devices based on the GMR effect consist of a sandwich of a nonmagnetic metal layer a few atoms thick between two thin ferromagnetic layers, the bottom layer of which is grown on top of an antiferromagnetic substrate. An external magnetic field can switch the magnetization direction of the outer ferromagnetic layer, but the magnetization of the layer grown on the antiferromagnet is pinned in one position by exchange biasing. The flipping of the magnetization orientation of the two ferromagnetic layers from parallel to antiparallel changes the resistance state of the material from low to high. Although exchange bias is important to a multibillion-dollar industry, the basic physics is not well understood, largely because of the lack of techniques capable of studying in detail the magnetic moments near interfaces.

To attack this problem, the ALS/IBM/Arizona State "PEEM team" combined polarized synchrotron radiation from the ALS with the new PEEM2 microscope with a spatial resolution as good as 20 nm. For some years, experimenters have been able to image ferromagnetic domains in thin layers by means of x-ray magnetic circular dichroism (XMCD), in which the absorption of circularly polarized x rays depends on the relative orientation of the polarization and the magnetization of the domain, thereby providing magnetic contrast. Antiferromagnetic materials have typically posed a bigger problem because the alternating orientations of the spins at each atom result in no net magnetic moment. Recently, the team showed that use of x-ray linear magnetic dichroism (XMLD) could be employed to image antiferromagnetic domains in LaFeO₃. Here the contrast arises at an absorption edge that splits into two peaks (multiplets) with heights that depend in opposite ways on the angle between the x-ray polarization and the antiferromagnetic axis.

Putting it all together, the team studied a sample comprising a thin (1.2-nm) cobalt layer on a 40-nm thick layer of LaFeO₃. Owing to the thinness of the cobalt, electrons emitted from the underlying layer were able to escape, and the magnetic structure on both sides of the interface could be imaged. XMLD images made at the split L₃ edge of iron resulted in a pattern of dark and light areas, according to the orientation of the antiferromagnetic axis. XMCD images made at the L_3,2 edge of cobalt resulted in a pattern highly correlated with that from LaFeO₃. Horizontally oriented domains, which showed up as light areas in the LaFeO₃, appear as gray areas in the cobalt. Vertically oriented domains, which showed up as dark areas in LaFeO₃, appear either dark or light in the cobalt, depending upon spin orientation (up or down, respectively). These results and the local remanent hysteresis loops recorded in individual ferromagnetic domains imply that the alignment of the ferromagnetic spins is determined, domain by domain, by the spin orientation in the antiferromagnetic layer.

A Step Closer to Solving a Magnetic Puzzle

At the frontier of advanced magnetic devices are tiny structures comprising multiple layers, each with dimensions measured in nanometers, hence the name magnetic nanostructures. These devices are based on the giant magnetoresistance (GMR) effect in which an external magnetic field dramatically changes the electric current flowing through the device. In the newest computer disks, the high- and low-current states are used to represent the digital states (0 and 1) in the read head that skims over the disk surface sensing the magnetic fields associated with the stored bits. The magnetic storage industry is developing similar devices as random access memory (RAM) for computers. One of the unsolved puzzles associated with GMR-based devices is a phenomenon called exchange biasing, which poses a scientific challenge of fundamental interest whose solution might also lead to improved technology. Researchers working at the ALS have developed an improved x-ray microscope (x-ray photoemission electron microscope, or PEEM) and used it to directly observe exchange biasing in action, thereby opening the door to a definitive understanding of the mechanism of this effect.

Linear dichroism at Fe L edge images antiferromagnetic domains (left). Circular dichroism at Co L edge images ferromagnetic domains (right). Comparison of the images shows that the Co domains align with the antiferromagnetic domains (light and dark regions inside outlined areas).

Research conducted by F. Nolting, A. Scholl, S. Anders, and H.A. Padmore (ALS); J. Stöhr, J. Lüning, E.E. Fullerton, and M.F. Toney (IBM Almaden Research Center); J.W. Seo (University of Neuchâtel and IBM Zürich Research Laboratory); J. Fompeyrine, H. Siegwart, and J.-P. Loquet (IBM Zürich Research Laboratory); and M.R. Scheinfein (Arizona State University).

Research funding: Office of Basic Energy Sciences (BES), U.S. Department of Energy; Swiss National Science Foundation. Operation of the ALS is supported by BES.

Publications about this research: F. Nolting, A. Scholl, J. Stöhr, J.W. Seo, J. Fompeyrine, H. Siegwart, J.-P. Loquet, S. Anders, J. Lüning, E.E. Fullerton, M.F. Toney, M.R. Scheinfein, and H.A. Padmore, "Direct observation of the alignment of ferromagnetic spins by antiferromagnetic spins," Nature 405, 767 (2000). A. Scholl, J. Stöhr, J. Lüning, J.W. Seo, J. Fompeyrine, H. Siegwart, J.-P. Loquet, F. Nolting, S. Anders, E.E. Fullerton, M.R. Scheinfein, and H.A. Padmore, "Observation of antiferromagnetic domains in epitaxial thin films," Science 287, 1014 (2000).