Influence of sulfur interlayers on the Mg/GaAs(100) interface formation

Abstract

The modification of clean GaAs͑100͒ surfaces by in situ deposition of molecular sulfur was investigated by soft x-ray photoemission spectroscopy. Upon S treatment of the clean GaAs͑100͒ sample at 435-455°C in ultrahigh vacuum the formation of a three monolayer thick gallium sulfide-like compound is observed, which exhibits a ͑2ϫ1͒ low-energy electron diffraction pattern. Due to the S modification on n-GaAs a reduction of the band bending by 0.35 eV is achieved, while the band bending on p-GaAs is increased by 0.17 eV. The subsequent Mg evaporation leads to the formation of a metal/semiconductor contact with a reacted magnesium sulfide-like compound at the interface. After 1 nm Mg deposition the Schottky barrier height of the S-modified Mg/n-GaAs(100) contact amounts to 0.44 eV, which is 0.18 eV lower than without S modification, while the Mg/p-GaAs͑100͒ Schottky contact exhibits an increase in the Schottky barrier height by 0.30 eV in comparison to the value of the unmodified Schottky contact ͑0.55 eV͒.

Photoemission Studies of Surface, Interface and Bulk Properties of GaAs (001) Treated by Mg

“…This is consistent with an earlier reported model [13]. Therefore, Ga3 is assigned to metallic Ga in agreement with earlier investigations, in which a component in the Ga 3d core level spectrum shifted by either 0.9 eV [14] or by 1.12 eV [16] relative to the bulk component was attributed to Ga metallic clusters. Together with data from a similar experiment on Mg/GaN [17], the above experimental data suggests that site exchange between Mg and Ga releases free Ga during the Mg deposition, and that Ga clusters have formed at the interface.…”

“…But the shift of this component relative to the bulk As1 decreases from 0.33 eV at 0.2 nm to 0.19 eV at 0.6 nm of nominal Mg coverage. Compared with previous results [12][13][14][15], this As4 is tentatively interpreted in terms of the formation of a Mg 3 As 2 -like compound. More detailed discussion will be given in the discussion section.…”

“…No contribution from surface Ga2 was observed following 0.2 nm Mg deposition, indicating that emission from Ga atoms on the surface was either fully attenuated by the Mg overlayer, or that these atoms have formed clusters with other released free Ga. This situation is similar to the case of sulfur treated Mg/GaAs [14].…”

“…Following the above discussion, the Mg core level might be expected to consist of two components: a reacted component and a bulk component. Generally for metallic layers, there is no evidence for a surface component [14]. However, following 0.2 nm Mg deposition, the Mg core level was able to be fitted using one component (Mg1) of the 0.58 eV Voigt width as shown in Fig.…”

“…The reaction of a Mg thin film with a Gastabilized 4 Â 2 GaAs (001) surface was observed to produce elongated wire-like structures with widths of the order of 50 nm and lengths up to 1000 nm. Interfaces of Mg/GaAs [13] and influence of sulfur interlayers on the Mg/GaAs (100) interface [14] have been studied by synchrotron radiation photoemission, and an obvious core level shift has been observed. Investigations of the ZnSe/GaAs interface by treatment of the GaAs (2 Â 4) surface with Mg was also performed.…”