Investigation of the (√3 × √3)R30°-Cu2Si/Cu(1 1 1) surface alloy using DFT
Highlights
► Evident and surprising similarities between the Si–Cu interaction for each of these phases, despite differences in the surface-2nd layer registry and proven Si–Cu interactions between these two layers. ► Si states in these systems do not undergo significant hybridization, and the Si–Cu bonds lack directionality and form either covalent or metallic states. ► Significant distortions in the 2nd layer Cu 3d orbitals which are exactly compensated by bond length differences between the surface and 2nd layer, suggesting the masking of these effects in a simple work function analysis. ► Preliminary energetic analysis suggestive that boundary and mass transport effects dominate the formation of the stable mixed phase surface.
Introduction
Surface science studies of semiconductor–metal systems can serve as an effective building block in understanding commercial electronic material growth [1] and concurrently can further the understanding of industrial catalytic processes. The field is developed and extensive, with several decades of research contributing to our current understanding of these systems. The expressed interest in bi-metallic, and higher, alloyed surfaces arises from the fact that by controlling even the most basic growth parameters of these surfaces, such as stoichiometry and chemical composition, ‘tuning’ of the reactivity and electronic properties of the surface is seen experimentally. A fundamental, theoretical understanding of the reasons why particular components in the alloy produce different effects on the surface is a rigorous way of developing more effective and commercially more attractive materials and processes.
The (√3 × √3)R30°-Cu2Si/Cu(1 1 1) surface alloy has recently been investigated [2], [3] and has clearly shown that the alloy grows in 3 phases, with Si atoms occupying FCC, HCP and 2-fold bridge sites in ratio 25:25:50. The structure of the FCC, HCP and 2-fold bridge phases are shown in Fig. 1. These investigations are entirely unique within the field of metal–semiconductor interface formation, notably with their identification of multiple phase formation. Investigations of the electronic structures that form during semiconductor deposition on metal surfaces are very few in number. The Cu(1 1 0) + c(2 × 2)–Si system has been investigated using photoemission spectroscopy [4]. The study showed that the alloy electronic states favour an in-plane orientation enabled by hybridization of the Si 3p and Cu 3d states. Later computational studies of the Si/Cu(1 1 0) system [5] confirmed the assignment of the c(2 × 2) structure and identified the energetic preference of the Si ion cores to incorporate themselves in the surface layer rather than in on-surface sites, or bulk incorporation. Studies of Ag/Cu(1 1 0) [6] showed the formation of a substitutional alloy which undergoes a transition from 1D and 2D electronic structures at low (<0.4 ML) Ag coverages to 3D structures at high coverages. Some discussion showed that geometrical factors at much larger (>0.65 ML) coverages were accompanied by 1D and 2D electronic features. These features were assigned to hybridization between the Ag sp and Cu d states.
The effects of local binding geometry in bulk Si–Cu alloys were investigated by Magaud et al. [7] who identified the notable effects of Si p–Cu d hybridization in bulk Cu3Si. The relevance of this is the comparison of the form of Si bonding in bulk compared to, for example, common molecular forms where strong hybridization between the s and p states is seen. Other works on Al/Pd(1 0 0) [8] have identified that significant reconstruction of the surface layer coupled with incorporation of Al ion cores in the 2nd layer are required to form a stable structure. In balance, however, submission also exists which identify weak bonding effects between the alloy components – for example, Cu/W(0 0 1) [9].
The current therefore predominantly addresses the interactions between Si and Cu in each phase of the (√3 × √3)R30°-Cu2Si/Cu(1 1 1) surface alloy system. The significance of the Si–Cu interaction between ion cores in both the surface layer, and between ion cores in the surface and 2nd layers is elucidated. Further, some discussion of the Cu–Cu interaction in these phases is presented.
Section snippets
Computational details
The results presented in this paper were performed using the SIESTA method [10]. The pseudopotentials were generated according to the improved Troullier–Martins method [11] with non-linear core corrections [12]. Exchange-correlation was estimated using the GGA method [13]. A double-ζ split-valence basis set including polarization orbitals was used [14] and a single Kleinman–Bylander projector represented each angular momentum channel [15]. The simulations were performed using 3s and 3p orbitals
Results and discussion
The current work predominantly address the electronic structures of the (√3 × √3)R30°-Cu2Si/Cu(1 1 1) system rather than structural issues which have been previously published [2], [3]. One of the most powerful probes of total electronic structure is the work function change Δϕ during alloy formation. This change will contain all the electronic information about the surface change, albeit in a non-specific way.
The changes in work function Δϕ during formation of each of the FCC, HCP and 2-fold
Conclusions
The current work has investigated the FCC, HCP and 2-fold bridge phases of the (√3 × √3)R30°-Cu2Si/Cu(1 1 1) surface alloy. Surprising similarities between the Si–Cu interaction for each of these systems has been noticed. Analysis has revealed that the Si–Cu bonds lack directionality and can form either covalent or metallic states. Investigations into the role of the Cu 3d states in these systems have shown that the Cu3dxy and orbitals are oriented to lie the surface plane whilst the Cu3d
References (17)
App. Surf. Sci.
(2009)- et al.
Surf. Sci.
(2001) - et al.
Surf. Sci.
(2000) - et al.
Surf. Sci.
(2004) - et al.
Surf. Sci.
(2005) - et al.
Physica B
(1996) - et al.
Surf. Sci.
(2008) - et al.
Surf. Sci.
(1989)
Cited by (8)
Magnetism in Cu <inf>2</inf> Si zigzag nanoribbon
2019, Computational Materials ScienceCitation Excerpt :The results are consistent with previous theoretical prediction and experimental value [37,42], i.e. 4.123 Å for lattice constant and 2.38 Å for Cu-Cu and Cu-Si bonds. The intrinsic SOC is weak in monolayer Cu2Si and so does the interaction between monolayer Cu2Si and the substrate [37,38]. Therefore, we will ignore both of them throughout the paper.
Controlled FCC/on-top binding of H/Pt(111) using surface stress
2016, Applied Surface ScienceInvestigation of the H-Cu and Cu-Cu bonds in hydrogenated Cu
2013, Journal of Physics and Chemistry of SolidsComparative investigation of the c(2 × 2)-Si/Cu(0 1 1) and (√3 × √3)R30°-Cu <inf>2</inf> Si/Cu(1 1 1) surface alloys using DFT
2012, Applied Surface ScienceCitation Excerpt :The Si 3s state has been shown to form covalent bonds with the Cu 4s and 4p states, and the Si 3p states have been shown to form a mixture of covalent and metallic bonds with the Cu 3d, 4s and 4p states. This is very similar to the Si-Cu interactions seen in the three phases of the (√3 × √3)R30°-Cu2Si/Cu(1 1 1) system [8] indicating that the Si ion cores bond promote hexagonal (high coordination) bonding even when in a rectangular (low coordination) environment and is heavily suggestive that the Si ion cores drive the structure of Cu–Si alloys. Distortions in the surface and second layer Cu 3d states have been thoroughly investigated and the loss of resolution of these states during alloying of the Cu(0 1 1) surface has been assigned to the loss of suitable Cu nearest neighbors in the surface layer rather than the inclusion of Si ion cores.
The catalytic properties of sub-surface H layers on Cu(1 1 1)
2011, Chemical Physics Letters