Surface structure of Si(100) with submonolayer coverages of C
Introduction
The incorporation of C atoms into Si structures has been of great interest as a way to modify the properties of Si-based materials. The tensile strain due to the 35% mismatch in bond length between C and Si makes it possible to make Si1−xCx/Si heterostructures where electrons can be localized, forming an electron gas in the Si1−xCx layer [1]. The beneficial effect of a low concentration of C-atoms to suppress B diffusion during thermal processing is also well established [2]. Since the solid solubility of C is only of the order of 10−5, it is necessary to use low temperature growth techniques or ion implantation to introduce the C-atoms into the Si-lattice. One dramatic effect of C-atoms in the near surface layer of Si(100) surfaces is the large influence on the nucleation of Ge-islands on the surface [3], [4]. Another interesting effect predicted theoretically and supported experimentally is that C-atoms deposited on a Si surface in many cases prefer to occupy sub-surface sites [5], [6], [7].
In many surface studies where C has been intentionally or unintentionally deposited on Si(100) surfaces, a c(4×4) reconstruction has been observed by low-energy electron diffraction (LEED), reflection high-energy electron diffraction (RHEED) or scanning tunneling microscopy (STM). The estimated C-coverage for this reconstruction has varied from 0.0 to 0.5 ML. Correspondingly, there have been different models suggested for the surface, containing from zero to four C-atoms per c(4×4) unit cell [6], [8]. In this paper we report a combined STM, LEED and Auger electron spectroscopy (AES) study to establish the amount of C that is needed to get a full coverage of the c(4×4) reconstruction. We also describe a model for the surface that can explain all main features of the STM-images.
Section snippets
Experimental details
The studies were performed using a combined STM/MBE system with three separate sections: the STM chamber, the MBE chamber and the preparation chamber, that is also used as a load lock for new samples and STM tips. The MBE chamber contains four MBE-sources, two of them filled with Si and SiC, respectively. To monitor the deposition rates from each source, there is a quartz crystal monitor. There is also a cylindrical mirror analyzer for AES analysis and a rear-view LEED optics. The background
Results and discussion
Fig. 1 shows a filled state image of a Si(100) surface taken within 2 h after deposition of 1 Å from the SiC source and annealing at 600 °C. The amount of C on the surface corresponds to 0.07±0.01 ML. The surface coverage of the c(4×4) phase on this surface is 98%, with the rest being mainly dimer rows of very small 2×1 reconstructed areas. The appearance of the c(4×4) surface in Fig. 1 is very typical for the as-prepared surfaces, i.e. there is a coherent underlying c(4×4) reconstruction that
Conclusions
The formation of the Si(100) c(4×4) reconstruction has been studied in experiments using co-deposition from Si and SiC sources and post-annealing. It has been determined that 0.07 ML of C is enough to get a full coverage of the c(4×4) reconstruction. This excludes the possibility that there are one or more C atoms per c(4×4) unit cell. The freshly prepared c(4×4) surface contains a blend of mainly two types of unit cells. The features observed in STM-images are consistent with an underlying
References (9)
- et al.
Surf. Sci.
(1998) - et al.
Phys. Rev. Lett.
(1996) - et al.
Appl. Phys. Lett.
(1997) - et al.
Appl. Phys. Lett.
(1997)
Cited by (12)
Ge dots formation using Si(100)-c(4×4) surface reconstruction
2016, Journal of Crystal GrowthCitation Excerpt :Heteroepitaxial growth of Ge on Si is known to follow the Stranski–Krastanov (SK) growth mode [7] which forms layer-plus-island structure, so the Ge QDs without a wetting layer are difficult to form. Several studies [8–11] have reported that carbon (C) mediated growth enabled to form isolated small and dense Ge dots. However, the effects of C binding states with Si (100) surface and Si–C reaction temperature were not clear, so we investigated their effects on Ge dot formation in our previous studies [12–14].
General trends of the carbon penetration in Si(0 0 1) surfaces: influences of relevant parameters
2006, Progress in Surface ScienceCitation Excerpt :In presence of a dimer vacancy, carbon incorporation in the fourth layer sites located directly under the dimer vacancy is favoured, as reported by H. Nörenberg et al. [19], and Kim et al. [37,38]. In the case of a defect-free Si(0 0 1) surface, mixed configurations involving carbon in both surface and subsurface layers are favoured from rather low coverages (θ = 0.25) and the carbon penetration in the Si(0 0 1) subsurface becomes easier with increasing carbon concentration in agreement with the experimental observations by Simon et al. [28], C1s core binding energy measurements of Ref. [34,35] as well as Ref. [23–29]. Finally, taking the surface defects presence on Si(0 0 1) into account in the theoretical studies helps to clarify the controversy on the carbon atom positions with respect to the surface (carbon either at the surface, or in subsurface layers, or in both surface and subsurface sites): the role of the carbon concentration and of the number and nature of the surface defects (dimer vacancies or ad-dimers) in the carbon location with respect to the surface layer is clearly emphasized.
General trends of the carbon penetration in Si(001) surfaces
2005, Solid State CommunicationsInfluence of C incorporation on the initial growth of epitaxial NiSi <inf>2</inf> on Si(1 0 0)
2004, Applied Surface Science