Jump to content

Spin-polarized electron energy loss spectroscopy

From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by 192.108.69.177 (talk) at 14:11, 8 December 2009. The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Introduction

spin waves are quantized collective excitations in the magnetic solida. The physical properties of spin wave excitations depends strongly on theier wavelength. For longe wavelength, which are order of magnitude longer than the lattice constant, the resulting spin waves have very low frequencies. Both ferromagnetic resonance (FMR) and Brillouin light scattering (BLS) provide information on these spin waves in thin films and nanostructures. However, if the wavelength is comparable to the lattice constant, the spin waves are entirely determined by the microscopic exchange coupling.

Up to now, the spin polarized electron energy loss spectroscopy (SPEELS) is the only technique, which can be used to measure the dispersion of such short wavelength spin waves in ultrathin films and nanostructures.

The First Experiment

By studying ultrathin ferromagnetic films Kirschner's group [1] in Max-Planck institute of Microstructure Physics [2] showed for the first time that the signature of the short wavelength spin waves can be detected by spin polarized electron energy loss spectroscopy (SPEELS) [1] [2]. Later, with a better momentum resolution, the spin wave dispersion was fully measured in 8 ML fcc Co film on Cu(001) [3] and 8 ML hcp Co on W(110) [4], respectively. Those spin waves were obtained up to the surface Brillouin zone (SBZ) at the energy range about few hundreds of meV.

Basic Principle of the Experiment

Fig. 1. A schematic representation of the SPEELS experiment.


In a spin polarized electron energy loss spectroscopy (SPEELS) experiment, a spin-polarized monochromatic electron beam, created by a strained GaAs photocathode, is scattering from the sample. The scattered electrons are analyzed with respect to their energy to determine the energy and wave vector transferred to the sample. Furthermore, the intensity of the scattered electron beam, for the two possible orientations of the incoming electron spin (parallel and antiparallel) with respect to the sample magnetization, is recorded [5]. A schematic representation of the SPEELS experiment and the scattering process taking place in this experiment is given in Fig.1.

Fig. 2. Series of normalized SPEELS intensity spectra of 2 ML Fe on W(110).


An example of thier recent work is the study of 1 and 2 monolayer Fe films on W(110) measured at 120 K and 300 K, respectively [6] [7]. The corresponding (normalized) energy loss spectra of a 2 monolayer sample are shown in Fig. 1 for various wave vector transfers. The excitation of a spin wave appears as a well defined peak in the loss spectrum, and the full spin wave dispersion of the Fe films were obtained up to the surface Brillouin zone (SBZ) boundary along Fe[001] direction. It has been observed that a peak emerges as a shoulder from the elastic peak for small wave vector transfer spectra, and this peak shifts toward higher energy loss range with increasing of wave vector. (The peaks at 70 meV and 130 meV are vibrational loss features of O and H, respectively). This feature can be explained on the basis of formation of spin wave excitations in the I channel (incoming electron spin is parallel to the spin of a minority electron in the sample). The total angular momentum conservation during the scattering process prohibits the excitation of spin waves for incoming electrons having a majority spin character (I). However, since the spins of electrons of the sample are not aligned perfectly in the remanence state, a much weaker spin wave peak shows up in I channel. One can avoid the non-spin-dependent effects by taking the difference of the spectra ΔI = I - I, shown in (c). By plotting the peak position (energy) versus wave vector transfer one obtains the dispersion on the spin waves.

References

  1. ^ M. Plihal, D. L. Mills, and J. Kirschner, Phys. Rev. Lett. 82, 2579 (1999).
  2. ^ H. Ibach, D. Bruchmann, R. Vollmer, M. Etzkorn, P. S. Anil Kumar, and J. Kirschner, Rev. Sci. Instrum. 74 4089 (2003).
  3. ^ R. Vollmer, M. Etzkorn, P. S. Anil Kumar, H. Ibach, and J. Kirschner, Phys. Rev. Lett. 91, 147201 (2003).
  4. ^ M. Etzkorn, P. S. Anil Kumar, W.X. Tang, Y. Zhang, and J. Kirschner, Phys. Rev. B 72, 184420 (2005).
  5. ^ R. Vollmer, M. Etzkorn, P. S. Anil Kumar, H. Ibach, and J. Kirschner, Phys. Rev. Lett. 91, 147201 (2003).
  6. ^ W. X. Tang, Y. Zhang, I. Tudosa, J. Prokop, M. Etzkorn, and J. Kirschner, Phys. Rev. Lett. 99, 087202 (2007).
  7. ^ J. Prokop, W. X. Tang, Y. Zhang, I. Tudosa, T. R. F. Peixoto, Kh. Zakeri, and J. Kirschner, Phys. Rev. Lett. 102, 177206 (2009).