Abstract: Thin films for electro−optical devices

Publisher Website
Google Scholar
Abstract
The recent progress in the development of high−quality, binary, and ternary compound semiconductor films promises significant advancements for the electro−optical device field. In particular, the technology for producing epitaxial films of certain variable energy gap materials should result in a new generation of economical sensors, sensor arrays, and emitters for all useful portions of the infrared spectrum. Beyond economy, thin film photoelectric materials have other inherent advantages over their bulk or thick film counterparts which should contribute to the increased sophistication of electro−optical devices. Multicolor sensors, charge−coupled elements and elements for sensor integrated optics systems are but a few examples of areas in which thin film technology should make rather complex devices practical. Progress of this nature has been particularly noteworthy with the IV−VI compounds which include the interesting narrow and/or variable energy gap materials PbTe, PbSnTe, PbSnSe, PbSeTe, and PbSe. Many investigators have now succeeded in preparing excellent single crystal films of these compounds utilizing such deposition techniques as sputtering,1,2 evaporation,3−6 vapor phase,7 and liquid phase epitaxy.8 More recent results of these efforts are reviewed in this paper. This review shows that it has not only been possible to deposit structurally high−quality epitaxial films of these materials, but that the electrical properties of these films are excellent and approach those of good, single crystal bulk materials. Where the nature of the work permits, the review includes also a survey of the electro−optical device applications of the recently developed thin film materials. The preparation of high−quality electro−optical devices is obviously the ultimate goal of most of the work being pursued. Examples of the topics covered are work on high−quality Schottky barrier devices using thin film PbTe, PbSnTe, and PbSe9−11 and on ion−implanted (protons and Sb+ ions) p−n junction devices using thin film PbTe.12,13 In addition, some very noteworthy work on the preparation of p−n homo− and hetero−junctions of PbSnTe and PbTe will be considered. The second portion of the talk will consist of a more detailed exploration of some recent work by the author with emphasis on thin film Pb1−xSnxTe, presently the most viable IV−VI compound material. Specifically, some new approaches for the formation of thin film diodes will be discussed which evolved in the author’s laboratory. As already mentioned, thin film Pb1−xSnxTe which is equivalent in quality to well−annealed bulk single−crystal material of identical composition can be deposited by various techniques. In most cases, however, the films require an annealing treatment after deposition similar to that used for bulk materials to adjust the stoichiometry. Only then can film carrier concentrations and mobilities suitable for electro−optical devices be obtained (e.g., n?1016 − 1017 cm−3, μ ? 104 cm2/V sec). Annealing is one preparation step which is not only uneconomical but adds severe difficulties in the fabrication of diode and multicolor devices. Another difficulty inherent in most PbSnTe thin film device preparation approaches is the requirement for simultaneously controlling the composition, which defines the spectral response of the film and the structure and carrier concentration which define the sensitivity. All are controlled by interdependent deposition parameters such as substrate temperatures, and relative as well as absolute deposition rates of the compound elements. These are some examples of the reasons which led to a search for a new and simpler deposition technique for the preparation of thin film sensors. One such technique is bias sputtering from a prereacted PbSnTe target. It has been demonstrated to be an easily controlled deposition technique which yields as−deposited, that is, unannealed thin films of device quality. Even more important for device application, bias sputtering permits the control of the carrier type during deposition, i.e., n− or p−type films can be deposited for any desired stoichiometric film composition. A second technique, also a sputter deposition process, entails the use of gaseous impurities (O2 or N2) as additives to the sputtering gas. This method produces p− or n−type films in a manner analogous to ion implantation. Coupled with the control of the carrier types, the additives produce an apparent trapping effect which greatly enhances the photo response characteristics of Pb1−xSnxTe films. Bias sputtering entails supported discharge or triode sputtering onto substrates to which a dc bias has been applied. For each desired Pb1−xSnxTe composition, a critical bias voltage V0 can be defined which will yield near stoichiometric film. Small deviations from V0 produce p− or n−type films of identical composition and typically with very low, as−deposited carrier concentrations. The type depends on whether the deviation is positive or negative relative to V0. For a given target composition, p− or n−type films can be produced by this method which have a wide range of specifiable film compositions. For example, if the target used has the composition Pb1−ySnyTe, the film composition can be controlled to have x−values from x=0 to x=y. The low−carrier concentration that can be achieved at all compositions without annealing by bias sputtering seems to result from two effects: (a) the bias optimizes the film stoichiometry at the critical value V0, and (b) the bias increasingly reduces the impurity induced carrier sources as larger negative biases are applied. Both phenomena can be observed over the small bias range of ±30 V dc. Preferential adsorption or repulsion...

This publication has 0 references indexed in Scilit: