AVS 49th International Symposium
    Electronic Materials and Devices Tuesday Sessions
       Session EL+SC-TuM

Paper EL+SC-TuM1
Epitaxial Growth and Characterization of CdS Layers Grown on InP (001) using Molecular Beam Epitaxy from Atomic Sulfur Beam and Thermally Evaporated Cd

Tuesday, November 5, 2002, 8:20 am, Room C-107

Session: Heterojunctions
Presenter: J.W. Choi, University of North Carolina at Charlotte
Authors: J.W. Choi, University of North Carolina at Charlotte
M.-A. Hasan, University of North Carolina at Charlotte
A. Bhupathiraju, University of North Carolina at Charlotte
Correspondent: Click to Email
Recent progress in epi-ready InP(E@sub g@=1.35 eV) wafer growth has prompted new applications of heterojunction devices and quantum structures based on closely lattice matched epitaxial films on InP. Cadmium sulfide (CdS), a direct bandgap material (E@sub g@=2.5 eV), has a small lattice mismatch to InP (0.63%) that can facilitate optoelectronic integration on InP. In this work, growth of CdS on InP (001) substrates is investigated using modified molecular beam epitaxy (MBE). Two growth methods are investigated. In the first method, epitaxial growth is conducted by atomic sulfur (S) beam from an RF atomic source with H@sub 2@S as the source gas while Cd is supplied from a cracker cell. The RF source is operated at 13.56 MHz and produces an intense monoatomic sulfur beam (@lambda@= 930.5 nm emission peak measured by plasma spectroscopy). In the second method, deposition is performed using CdS sublimation from a solid CdS using a conventional Knudsen cell. Rinsing by methanol is used to clean the epi-ready InP (001) substrates. Final substrate cleaning is achieved by thermal desorption (530-550°C) in vacuum. RHEED shows InP (001) 2x1 surface after thermal cleaning. In both methods, in-situ RHEED measurement indicates epitaxial growth of zincblende CdS; Auger electron spectroscopy (AES) shows stoichiometric CdS within the AES resolution limit. AES depth profiles also indicate an abrupt InP/CdS interface for all temperatures investigated. The growth rate of CdS is primarily governed by Cd flux as indicated by the activation energy measured from the flux-dependence of the growth rate. Atomic force microscopy (AFM) shows measurable dependence of surface roughness on growth method. For growth using an RF atomic source, average surface roughnesses of 1-2 nm is measured for all growth temperatures, layer thicknesses, and flux ratios with no evidence of island formation; higher roughness values (10-20nm) were typically measured for growth from solid CdS.