R. I. G. Uhrberg

Initial formation of the interface between a polar insulator and a nonpolar semiconductor: CaF2 on Si(111)

The initial stages of interface formation between CaF2 and Si(111) have been studied with both core‐level and angle‐resolved valence band photoemission spectroscopy. Both the Si 2p and Ca 3p core levels indicate that a Si–Ca bond is present at the CaF2–Si(111) interface. Annealing of thin CaF2 films at 700–800 °C results in preferential evaporation of F, and the surface undergoes a number of reconstructions until a stable Si(111):Ca 3×1 reconstruction is obtained on which there is calcium, but no fluorine. This surface exhibits a surface state with an upward dispersion of 0.4 eV towards the K point of the 1×1 surface Brillouin zone. Our results show that the CaF2 molecule can be dissociated on the Si(111) surface at typical epitaxial growth temperatures, with the Ca being more strongly adsorbed to the Si than is the F.

Formation of the interface between GaAs and Si: Implications for GaAs‐on‐Si heteroepitaxy

Results of photoemission core‐level spectroscopy measurements for coverages of around one monolayer of As, Ga, and GaAs on Si substrates are presented. The interfaces were formed on on‐axis Si(100) and Si(111) substrates using molecular beam epitaxy techniques. The bonding between As and the substrate surface leaves the As atoms fully coordinated and thus extremely unreactive. This causes the GaAs films to form islands at average coverages of less than one monolayer. The surface between the islands is found to be terminated by a single atomic layer of As. Use of a Ga predeposition technique shows evidence of decreasing the area between islands. Results for As interaction with stepped Si(100) surfaces and the implications for avoidance of antiphase domain boundaries are discussed.

Electronic and atomic structure of arsenic terminated Si(100)

A comparison between angle resolved photoemission data and ab initio pseudopotential calculations for a structural model of the As covered Si(100) surface is presented. After a coverage of one monolayer the Si(100):As surface shows a 2×1 low energy electron diffraction pattern. The arsenic acts as an effective passivating layer and the adsorption of further arsenic is highly reduced. We propose a symmetric As–As dimer model for the Si(100):As 2×1 surface derived from energy minimization calculations. The major feature of the electronic structure of this model is the occurrence of occupied π and π* surface state bands derived from bonding and antibonding combinations of the dangling hybrids on the As atoms. The calculated dispersions for the π and π* bands are found to be in very good agreement with the experimental surface state dispersions. The results discussed provide an example of how angle resolved photoemission in combination with ab initio pseudopotential calculations can lead to a conclusive deter...

Surface band dispersion of Ge(111)c(2×8) and Ge(111):As 1×1

The surface band dispersions of Ge(111)c(2×8) and Ge(111):As 1×1 were determined experimentally using angle resolved photoemission. The clean c(2×8) reconstructed Ge(111) surface exhibited a surface dispersion with a 2×2 periodicity, suggesting that the surface has a 2×2 local geometry. Addition of As to the surface converted the surface symmetry to 1×1 and produced a strongly dispersing surface state with a bandwidth of 1.7 eV. Energy minimization calculations based on the first principles pseudopotential method and the local density approximation were used to determine the structure of Ge(111):As 1×1. The dispersion of the lone pair band was calculated and compared to the experimental results. The agreement found between experiment and theory allows us to conclude that arsenic interaction with the Ge(111) surface results in the replacement of the outer Ge layer with a single As layer.