| Summary: | Clean silicon surfaces and the reactions of carbon and hydrogen with the surfaces are studied in UHV with STM and LEED. The STM is the first STM to achieve atomic resolution in UHV in Canada. The details of its construction as well as the methods for resistively heating silicon surfaces, transferring the sample and changing the STM tip In UHV are described. The clean Si(111) 7x7 surface consists of reconstructed (111) oriented terraces separated by atomic steps, as seen by STM. Stress is applied to the Si(1 11) 7x7 surface by bending the crystal about the [1 ī 0] direction. The stressed surface that is thermally cleaned In UHV, has a 7x7 LEED pattern and no changes are observed by STM in the 7x7 reconstruction or in the structure of the surface steps. The sample becomes plastically deformed after heating and each of the LEED spots is stretched in the same direction. We postulate that the surface after heating consists of (111) oriented facets which are separated by parallel lines that have outward normals that are tilted with respect to one another. From the LEED pattern we determine that the lines separating the facets are oriented along the [0 ī 1] direction. The tilt boundary between neighbouring facets is accounted for by a row of dislocations below the surface with axes in the [0 ī 1] direction and alternating Burgers vectors. Carbon is deposited onto silicon surfaces in UHV by three methods: hydrocarbon exposure of a heated sample, contamination by the residual carbon containing gases in the chamber, and carbon evaporation by resistively heating carbon fibres. The STM shows that for all three methods, the result of heating the sample with deposited carbon to 1000 °C is a surface covered by identically shaped objects which are interpreted as tip images formed by sharp clusters of f-SiC on the surface. No evidence for the SiC is seen in the LEED patterns of the surface. It is postulated that the SiC forms pyramids with surfaces of (110) orientation, with a triangular base on the Si(1 11) surface and a square base on the Si(100) surface. The crystallites are observed to initially grow at the surface steps. The Si(111) 7x7 surface is reacted with atomic hydrogen and at each stage of the reaction the surface is studied by STM and LEED. For low exposures, the hydrogen atoms bond to the adatoms of the 7x7 reconstruction, removing the surface states associated with the reacted dangling bonds and causing the reacted adatoms to appear to be missing in the STM images of the surface. A greater probability for hydrogen reactions with the middle adatoms than with the corner adatoms is observed for increasing surface temperatures during the exposure. By counting the reacted adatoms in the STM images of the surface, we determine that a hydrogen atom has a 0.11 eV lower binding energy when it is bonded to a middle adatom than when it is bonded to a corner adatom and we measure an activation barrier of 1.1 eV for the surface diffusion of a hydrogen atom from a corner to a middle adatom. With higher hydrogen exposures the surface Is etched producing SiH4. It is determined that the SiC crystallites can also be etched by atomic hydrogen. For the higher exposures, the surface appears rough as seen by STM but well defined changes in the LEED patterns of the surface as a function of hydrogen exposure are observed. For the highest hydrogen exposures, the surface has a 1x1 LEED pattern. The stages in the formation of the 7x7 reconstruction on the 1x1 surface are seen when the 1x1 hydrogen etched surface is heated. A √3x√3 R30° LEED pattern is observed for the shortest heatings and the STM images show a disordered arrangement of adatoms on the surface. Further heating results in a 7x7 surface consisting of reconstructed terraces covered by pits. Eventually, with more heating all of the pits are removed from the terraces and the surface has the 7x7 reconstruction.
|
|---|
