The high-temperature decomposition of thin ( ~ 100 A) Si02 layers on Si(001) under ultrahighvacuum annealing conditions has been studied by means of ion scattering and microscopy techniques. Si02 is removed from the Si surface by the formation and lateral growth of holes in the oxide, exposing regions of atomically clean Si, while the surrounding oxide retains its initial thickness. Surface diffusion of Si inside the holes supplies Si for reaction with Si02 at the periphery, so that a volatile product (presumably SiO) can be formed.

No abstract

No abstract

Initial stages of Si epitaxial growth on vicinal Si (001) substrates were investigated using scanning tunneling microscopy. For a growth temperature of about 750 K it was found that initial growth occurs almost exclusively at one of the two nonequivalent types of step edge. This leads to the formation of a single-domain surface with an array of evenly spaced straight steps with biatomic height. This structure can be preserved by quenching the sample to room temperature. PACS numbers: 61.50.Cj, 68.55.Bd, 82.40.DmThe interest in the use of silicon molecular-beam epitaxy (Si MBE) on Si(001) surfaces for the production of novel electronic devices has motivated many studies of the initial stages of growth and the influence of the preceding cleaning treatment. One topic of interest has been the preparation of single-domain surfaces, where the steps between adjacent terraces have a height of an even number of monolayers (ML). This is important in reducing the amount of antiphase boundaries in III-V films grown on Si substrates. 1 It has been reported in the literature that a prolonged annealing of samples with a misorientation less than 0.2° at about 1250 K may result in a single-domain surface. 2 " 5 On samples with a misorientation of more than about 4° towards [110], biatomic height steps were found after cleaning and annealing at about 1150 K. 6 " 8 During MBE on a sample with a misorientation of 0.5° towards [110], a single-domain reflection high-energy electron diffraction pattern was observed. 9,10 However, Sakamoto et al. 10 mentioned that this single-domain surface is not stable. When growth is terminated and the sample is kept at the growth temperature of about 750 K, this surface changes back into a two-domain surface within a few minutes. 10 We present scanning-tunneling-microscope (STM) images of epitaxially grown single-domain Si (001) surfaces that were quenched to room temperature immediately after termination of growth.The STM uses a mechanical approach 11 and is described in more detail elsewhere. 12,13 It is mounted in an ultrahigh-vacuum chamber with a base pressure of lxlO -8 Pa, together with a low-energy electron diffraction (LEED) system. Si can be evaporated in situ from resistively heated Si strips at a growth rate of about 1 ML/min, as calibrated with Rutherford backscattering spectrometry on carbon substrates, and at a pressure better than 2xl0" 7 Pa. The 5x20-mm 2 substrates were cut from commercially available wafers (Wacker, floating zone, n type, 1 a cm) and ultrasonically rinsed in ethanol before loading into the vacuum chamber. The misorientation, as determined by x-ray diffraction, was 0.52° in the [110] and 0.09° in the [TlO] directions, respectively. Inside the chamber the samples were heated resistively. Temperatures were measured with an infrared pyrometer with an accuracy of 50 K. After outgassing the sample and the holder for several hours, the sample was thermally cleaned at 1450 K for 2 min and cooled at an average rate of 25 K/min. Si films have been deposited on these clean s...