Presented here are experimental/theoretical studies of the geometric structure of a Cu(001) surface and the near-surface layer magnetization of a ferromagnetic Ni(111) surface, both using spin-polarized low energy electron diffraction (SPLEED). This probe shows great promise both as a high quality structural probe, and also in a variety of magnetic studies.
Significant advances have been made over the past several years in both experimental and calculational LEED techniques. The introduction r-factor analysis and streamlining of calculational codes now allows the evaluation of increasingly complex structures by LEED I-V analysis. In addition, experimental advances now allow the investigation of not only the intensities, but also the spin-dependence of scattering from many surfaces.
Spin dependence in low-energy electron diffraction is found for scattering from all surfaces, and is due to two effects--the spin-orbit interaction and the exchange interaction. The former is sensitive to the positioning of ions that make up the surface and provides complementary probe of surface geometric structure in addition to LEED intensity analysis. Here we introduce a simple r-factor analysis to evaluate the structure of Cu(001). This surface was chosen because it is well studied, and a highly precise structural determination of the surface based on LEED I-V analysis is available for comparison. Experimental SPLEED A-V profiles are compared with calculated profiles to obtain the best near-surface structure. In this study, the comparison between calculated and experimental spin-asymmetry profiles obtained from this surface is unusually good, and further, the best fit is for a structure that agrees well with the previous LEED study.
The present study is the first use of r-factor in a SPLEED structural determinations. Several things are readily noted from the present SPLEED r-factor analysis. First the precision of structural conclusions obtainable r-factors is much greater than possible using the visual inspection method found elsewhere. Further, because A-V's are self-normalizing, the r-factors are cleaner, requiring no arbitrary scaling factor. Also the spin asymmetries show more sensitivity to structural variations than do the comparable LEED intensities.
The exchange interaction is sensitive to the electron spin-population density in the surface, an thus to surface magnetism. Our preliminary study of the ferromagnetic Ni(111) surface shows very small exchange asymmetry effects. The results are consistent with some enhancement of the sample magnetization at the surface, but a precise layer magnetization determination is beyond the limits of present technique.
| Identifer | oai:union.ndltd.org:RICE/oai:scholarship.rice.edu:1911/16087 |
| Date | January 1987 |
| Creators | LIND, DAVID MELVIN |
| Source Sets | Rice University |
| Language | English |
| Detected Language | English |
| Type | Thesis, Text |
| Format | application/pdf |
Page generated in 0.0188 seconds