Cleaning and passivation of semiconductor surfaces are of great importance in crystal growth and device fabrication. Several cleaning procedures are available for InP wafers, for example, chemical cleaning, 1,2 thermal cleaning, 3-5 and atomic hydrogen irradiation. 6,7 Among them, the chemical cleaning is the simplest and easiest to control.Unlike Si which forms a very excellent interface with its oxide, SiO 2 , the native oxide of InP is detrimental due to the formation of surface traps and interface states at the InP/InP native oxide interface. Therefore, the removal of the native oxide from the InP surface is critical to various device fabrications using this material. 8 Evolution of the surface topography of semiconductors is also of practical importance, since they usually represent the starting material of most device processing. It is known 9 that the physical and physicochemical properties of a semiconductor are closely related to its surface roughness. For example, the presence of topological imperfections could create surface or interface scattering, and these affect the carrier transport properties. It is also clear that the downsizing of semiconductor devices requires a high-resolution imaging of the semiconductor surface.In this paper, we report the chemical treatment effects of InP(100) surfaces in aqueous HCl solutions studied using X-ray photoelectron spectroscopy (XPS), spectroscopic ellipsometry (SE), (ex situ) atomic force microscopy (AFM), and contact-angle measurement techniques. XPS is used to give an overview of the core-level XPS lines observed for the HCl-cleaned InP(100) surfaces. SE is a highly surface-sensitive technique which enables detection of not only submonolayer coverage of a surface by adsorbed species 10,11 but also the surface roughness of its size comparable or smaller than the wavelength of light. 12 We perform an effective medium approximation (EMA) 13 and a linear regression analysis (LRA) 14 to evaluate the surface microroughness of the HCl-cleaned InP surfaces. AFM is also used to independently assess surface morphology changes due to the HCl etching.
ExperimentalThe InP(100) wafers used were Sn-doped, n-type with carrier concentrations of 1-3 ϫ 10 18 cm Ϫ3 . They were mirror-finished and supplied by Nippon Kogyo Co., Ltd.The as-received InP(100) wafers were first degreased with organic solvents in an ultrasonic bath and then rinsed with deionized (DI) water. No further cleaning of the sample surface was performed. Then the sample surfaces to be studied were covered with a ϳ4 nm thick native oxide film. This oxide thickness was determined by SE.The concentrations of the aqueous HCl solutions studied were x ϭ 0.009-36 wt %. A large quantity of the solution was prepared to prevent the etching solution composition from varying during the experiment. Etching experiments were performed at 20ЊC in room light without stirring. After chemical treatment, the samples were rinsed in DI water.The XPS measurements were performed with an ULVAC-PHI model 5600 spectrometer equipped wit...