XRD and TEM Characterization of Compound Semiconductor Solid Solutions: Sn(S,Se) and (Pb,Cd)S

X-ray powder diffraction (XRD) is an essential component of research into use of organometallic precursors as a route to low temperature (<500 “C) synthesis of bulk compound semiconductors. This paper illustrates the use of XRD to study solid solution behavior in the systems SnS-SnSe and PbS-CdS. XRD characterization demonstrated that phase pure intermediate members of the Sn(S,Se) solid solution prepared by this low temperature route showed the typical dependence of unit cell parameters on composition. PbS-CdS exhibits essentially no solid solution behavior at low temperatures according to established phase diagrams. Characterization of organometallic precursor thermolysis products with intermediate compositions initially suggested that a metastable solid solution with perhaps 30% of CdS in PbS could be prepared at 150°C. More detailed XRD and high resolution transmission electron microscopy (HRTEM) characterization indicated that the CdS was in fact present as a separate phase with such small crystallites (2-5 nm) that peak broadening made it essentially undetectable in concentrations below 30%. The actual amount of solid solution was 2% or less of CdS in PbS. Introduction Organometallic precursors have been found to provide a facile route to low temperature synthesis of bulk compound semiconductor metal chalcogenide solid solutions’. The primary advantage of producing solid solutions by organometallic precursor decomposition is intimate mixing of the inorganic constituents at the molecular level. This mixing occurs in the melt phase, during heating of mixtures of the organometallic precursor end-member molecules, prior to decomposition of these precursor molecules and subsequent crystallization of the inorganic end products, The organometallic compounds decompose at relatively low temperatures typically less than 300°C. Synthesis of solid solutions by conventional means, i.e. from the elements or binary end member compounds, would require temperatures of 500-lOOO”C, long heating times, intermediate grindings, and possibly explosive conditions. Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 XRD characterization, including the use of calculated X-ray powder diffraction patterns2, is an essential component of these studies. Here we describe results from studies of two binary chalcogenide systems, one in which XRD characterization was routine, and one in which the characterization was more challenging. The first binary system was that of the chalcogen anion-substituted solid solution, Sn(S,Se), in which complete solid solution between the isostructural (GeS structure) SnS and SnSe components is known3. Phase pure intermediate members of the Sn(S,Se) solid solution prepared at 450°C showed the typical dependence of unit cell parameters on composition. The equilibrium phase diagram for PbS-CdS4 indicates essentially no solid solution between the cubic (Fm3m) rocksalt structure PbS and the hexagonal (P63mc) wiirtzite structure CdS end members at temperatures below 500 “C. However, some solid solution of CdS in PbS is known to occur under equilibrium conditions at higher temperatures (maximum of ca. 40 mol% at the eutectic temperature of 1 052°C)4. In addition, complete solid solution with metastable retention of the rocksalt structure at ambient conditions has been reported for high pressure syntheses’. The organometallic precursor route was explored as a possible low temperature (<5OO”C), ambient pressure synthesis of metastable (Pb,Cd)S. Examination of routine powder diffraction patterns of products of syntheses of intermediate compositions suggested at first that a metastable solid solution with perhaps 30% of CdS in PbS could be prepared at 150°C. However, the rocksalt unit cell parameter did not show the substantial decrease that substitution of the smaller Cd for Pb would produce. We describe here the additional characterization by XRD and TEM that led to the conclusion that up to about 30%, the CdS was actually present in thermolysis products as a crystalline phase, but with such small crystallite sizes (2-5 nm) that peak broadening made it essentially impossible to detect in the X-ray diffraction patterns.