Adatom vibrations on Si(111) reconstructed surfaces

Abstract

The vibrational electron energy-loss spectrum of the Si(l 1 0(7x7) surface exhibits a well-defined loss at 570 cm -1 and a structure at 200-270 cm -1 . A similar spectrum is displayed by a Si surface with a pseudo (1 x l) structure. Ab initio calculations of the dynamical matrix show that these features are due to localized vibrations involving Si adatoms and the surface atoms lying underneath. The latter are fivefold-coordinated Si atoms exhibiting the electronic structure of an acceptor impurity localized at the surface.PACS numbers: 68.35.Ja, 68.35.BsRecent studies by high-resolution electron energy-loss (EEL) spectroscopy of clean Si(ll 0(7x7) surfaces have found a continuum of energy losses with increasing intensities towards smaller loss energies accompanied by a large broadening of the elastic peaks. 1 " 4 These features, which have been interpreted to be indicative of a metallic band of surface states contributing to the surface loss function, 1,4 have appreciably hindered the observation of intrinsic vibrations of the surface. In this Letter, we demonstrate that it is possible to suppress both the high electronic background and the broadening of the elastic reflected signal, by allowing small amounts of impurities on the surface. We have obtained Si(ll 0(7x7) surfaces with some carbon that exhibit a sharp loss at 570 cm -1 and a broad weaker feature near 220 cm -1 . We have also produced SiC111) surfaces with a pseudo (lxl) low-energy electron diffraction (LEED) pattern, stabilized by traces of Ni, that display nearly the same EEL spectra, indicating that the local arrangements of the surface atoms are similar to that of the (7x7) surface. By comparison to ab initio calculations of the vibration spectrum of a Sij5 cluster we show that the observed EEL spectrum is a clear signature of Si adatoms. The 570-cm _1 loss, which has an unusually high frequency for a localized Si vibration, involves the stretching of two Si-Si bonds and is thus inconsistent with tetrahedral coordination. The calculation reveals a new bond connecting the adatom with the surface Si atom lying underneath, which is therefore fivefold coordinated. This Si atom has the electronic structure of an acceptor impurity localized at the surface and we suggest that it may play a role in explaining the metallic character of the clean Si(l 1 0(7x7) surface.The experiments have been performed in ultrahigh vacuum (UHV), in a recipient equipped with an air-lock system which enabled us to change the samples quickly without breaking the vacuum. For a coarse determination of the sample temperature a NiCr-Ni thermocouple was fixed to the sample holder. Figure 1 (a) shows the EEL and Auger spectra of a SiC 11 0(7x7) surface containing small amounts of carbon. The samples, which have been cut from 400-800-n cm /2-type doped Si(l 11) wafers, have been introduced in the recipient through air and flashed to 800 °C in UHV in order to remove the oxide. The Auger spectrum shows that besides some carbon (/c272/^si92 = 0.014) no other contaminants were on the su...

Photoelectron diffraction of the Si(111)-(3×3)R30°-Ga surface: Local atomic structureand vibrational correlation

“…22 The mode is characteristic of the Si͑111͒ surface having T 4 adatoms. 23,24 As far as high-energy photoelectron is concerned, forwardscattering is dominant, and only the wave emitted toward the scatterer contributes to the intensity. Therefore, the parity of the photoelectron wave function has no influence upon the intensity distribution.…”

Femtosecond nonlinear spectroscopy at surfaces: Second-harmonic probing of hole burning at the Si(111)7x7 surface and fourier-transform sum-frequency vibrational spectroscopy

“…There are also optical phonon bands at about 240 cm −1 [119,120]. It is possible that optical phonons are contributing significantly to the deduced dephasing times.…”

“…The simplest modification is to suppose that the 1γ contribution to the second-order nonlinear response is due to a pair of inhomogeneously broadened oscillators. This is a reasonable possibility given that the Si(111)7×7 surface band structure is not known in sufficient detail to determine whether the surface bands are indeed single bands or separate, narrow, closely The Si(111)7×7 reconstruction has a surface optical phonon band at 570 cm −1 [119,120], so the extra term in the fit may be a manifestation of scattering of photoexcited electrons or holes by optical phonons to a third band or to a different point in k-space of the ground or excited band giving rise to the hole at zero detuning. Although the fits involving the second Lorentzian term are fairly good, we must be cautious about the interpretation of show that the evolution of the population difference ∆ρ = ρ aa − ρ bb is governed by…”

Electron Dynamics of Silicon Surface States: Second-Harmonic Hole Burning onSi(111)−(7×7)

“…The deep hole at a probe-pump detuning of 570 cm ÿ1 implies strong coupling between the excited adatom electronic state and the 570 cm ÿ1 surface optical phonon of Si111-7 7 [20], which can be identified primarily with the out-of-plane motion of the adatom. Strong electron-phonon coupling at the adatoms has been implicated in the desorption of the adatoms of the Si111-7 7 surface [21].…”