Charge Injection Devices

|The name "charge injection" describes the operation of a generic class of electronic and optoelectronic multiterminal semiconductor devices, based on the real space transfer of hot carriers over a potential barrier into adjacent and separately contacted layers. Charge injection devices enable the implementation of compact optoelectronic gates endowed with a powerful logic functionality. Keywords|Charge Injection, Functional Device, Real Space Transfer. I. Operation of a Charge Injection Device The concept of real-space transfer (RST) [1{3] describes the process in which electrons in a narrow semiconductor layer, accelerated by an electric eld parallel to the layer, acquire high average energy (become \hot") and then spill over an energy barrier into the adjacent layers. Redistribution of carriers between parallel layers of di erent mobility, leading to a negative di erential resistance (NDR) and an unstable current-voltage characteristic in a two-terminal device was the original intended application of RST [4]. Transistor applications of the RST began with the proposal by Kastalsky and Luryi [5] of a three-terminal structure where hot-electron injection occurs between separately contacted conducting layers. This structure, called the charge injection transistor or CHINT is illustrated in Fig. 1. One of the two layers (\emitter") has source and drain contacts and plays the role of a hot-electron cathode. The other layer (\collector") is separated by a potential barrier. When carriers in the emitter layer are heated by the source-drain eld, most of them do not reach the drain but are injected over the barrier into the collector layer; a strong NDR develops in the drain circuit. The transistor action results from an e cient control of the electron temperature Te and hence the injection current IC by the input voltage VDS . Experimentally, CHINT was demonstrated for the rst time at Bell Laboratories in 1984 [6] in a GaAs/AlGaAs heterostructure. Since then many other heterojunctions have been used to implement di erent versions of charge injection devices [7{9]. Below, we shall brie y review the main variations. Section II describes the operation of CHINT in more detail, while section III discusses a different operation of the same device that is based on the NDR e ect in the channel. Logic properties of charge injection devices are described in section IV. The last section V describes light emitting charge injection devices. II. Charge Injection Transistor The operation of CHINT is based on the control of the collector current by the heating voltage VDS . The maSerge Luryi is with State University of New York at Stony Brook, Stony Brook, NY 11794-2350, U.S.A. Marco Mastrapasqua is with LucentTechnologies Bell Laboratories, 600 Mountain Ave, Murray Hill, NJ 07974, U.S.A. EMITTER