Evidence for Spin-Dependent Electron-Hole-Pair Excitations in Spin-Polarized Secondary-Electron Emission from Ni(110)

Abstract

and which thus rapidly increases as the incommensurate-paraeiectric transition T T is approached. In the vicinity of the defects the phase will be pinned, whereas in the region between the defects we shall have large phase fluctuations (i.e., the modulation wave will be rough). The radius of the "pinned" region (where the root mean square phase fluctuation amplitude is smaller than TT/2) slowly decreases with increasing temperature, whereas the "floating" part (where the root mean square phase fluctuation amplitude exceeds TT/2) increases with the mean distance between "pinning" defects. The fraction £ of the pinned nuclei giving rise to the broad NMR line thus decreases critically as the incommensurateparaeiectric transition temperature is approached;Thus NMR lines from the pinned and the floating regions should be observed below T l if the number of defects is not too large.The experimental data show that y ~ 2. The appearance of the sharp NMR lines is thus due to a decrease in the strength of the phase pinning potential of impurities with increasing temperature and the corresponding increase in the mean distance between pinning defects. This increase in distance results in an increase in the mean square vibration amplitude of the modulation wave, which thus becomes "rough" and float-Primary electrons with sufficient energy impinging on a solid generate low-energy secondary electrons via a cascade process. Within the cascade process, the number of hot electrons is multiplied due to scattering of the excited elec-ing over an increasing part of the crystal. The observed phenomenon is thus similar to a roughening transition. Similar effects should be observable close to T l in other incommensurate systems provided the pinning defect concentration is not too large. . trons primarily by conduction-band electrons. 1 In the absence of spin-dependent scattering processes it had been thought that the spin polarization of the secondaries directly reflected the net spin polarization of the conduction bands, P = («t The energy-resolved spin polarization P of secondary electrons from Ni(110) has been measured with high energy resolution. At the lowest kinetic energy p reaches a maximum of (17 ±2)% which is significantly higher than the mean conduction-band polarization (5.5%). Within about 3 eV towards higher kinetic energy the spin polarization decreases to about 8% and displays pronounced structures at higher energies. Spindependent electron-hole-pair excitations are suggested to be responsible for the observed spin polarization.

Monte Carlo study of the spin polarization enhancement in the low‐energy secondary electron emission from ferromagnetic materials

“…We take the value of 16 eV for the present model, which is determined so that the calculated peak of the secondary spin polarization matches the experimental one for Ni by Hopster et al 20 Actually, there is a probability of non-spin-flip scattering for spin-down electrons below 16 eV. Because we do not understand the cross-section of the Stoner excitation, the probability of the Stoner excitation is assumed to be unity in the case of point (3) above.…”

“…The incident energy and angle of primary electrons are 400 eV and 30°, respectively. The Monte Carlo results with two models are compared with the experimental results of Hopster et al 20 The simulation result with the previous model does not reproduce the enhancement of the spin polarization at low energies. The threshold energy of Stoner excitation of 16 eV in the present model is determined so that the calculated peak value matches that of this experimental result of 17%.…”

“…The humps at 4 and 16 eV are considered to be attributed to the maxima in the density of states. 20 However, a full understanding has not been obtained yet. Our simulation cannot reproduce these humps.…”

Analysis of the spin polarization of secondary electrons emitted from ferromagnetic materials

“…7,8 The other is that the spin polarization increases with primary energy and reaches a saturation value at high primary energies. 6,9 The former characteristic has been explained qualitatively by Kisker et al 3 and by Hopster et al 6 However, there are few quantitative explanations given so far on the energy distribution of spin-polarized secondary electrons and, especially, on the fine structures at higher energies. Also, the primary energy dependence of spin polarization has not been explained yet.…”

“…Many experiments have been carried out to study the properties of spin polarization of secondary electrons for several types of magnetic materials such as single crystals, 3 metallic glasses 4,5 and polycrystalline permalloys. 3,6 Two major characteristics for the spin polarization of secondary electrons were found and regarded as very important. One is that the spin polarization of secondary electrons is much larger than the average conduction band polarization at low energies and it decreases with increasing kinetic energy of electrons.…”