Metal—semiconductor interfaces

The materials (metals and conjugated polymers) that are used in LED applications were introduced in the previous chapter. The polymer is a semiconductor with a band gap of 2-3 eV. The most commonly used polymers in LEDs today are derivatives of poly(p-phenylene-vinylene) (PPV), poly(p-phenylene) (PPP), and polythiophene (PT). These polymers are soluble and therefore relatively easy to process. The most common LED device layout is a three layer component consisting of a metallic contact, typically indium tin oxide (ITO), on a glass substrate, a polymer film r- 1000 A thick), and an evaporated metal contact4. Electric contact to an external voltage supply is made with the two metallic layers on either side of the polymer. 

In the following discussion of the electronic properties of metal-semiconductor interfaces, the properties of the electron-injecting contact are taken as the example for contacts. Most of our studies of metals on polymers have involved low work function metals with the 

Suppose that the metal and the semiconductor are both electrically neutral and separated from each other. Since the metal at the electron-injecting contact is assumed to have a low work function, the Fermi energy of the metal lies above that of the semiconductor, close to its conduction band. If the metal and the semiconductor are connected electrically, electrons will flow from the metal to the polymer in order to establish equilibrium, which is obtained when the Fermi energies of the two materials are aligned. Because of this flow of charge, the materials are no longer neutral after electric 

The difficulty facing the reviewer is the almost complete incompatibility between direct kinetic measurements and electronic/compositional studies of metal deposits. Kinetic results have been used to construct models of the surface metal phase with varying degrees of geometric complexity and/or thin film growth mode, but in all cases the metal is assumed a priori to be on top of the semiconductor and to desorb from it in some classical way. Unfortunately, direct measurement of the interface chemistry 

In the following sections, we will first consider metal—silicon and metal/III—V compound systems, covering electronic and chemical interface effects. We will then present a general treatment of the adsorption-desorption models which have been proposed, including some discussion of the interpretation of thermal desorption spectra, and finally we will discuss to what extent the two sets of data can be reconciled. We will not be concerned with metal films in device fabrication, nor the formation of additional phases by heat treatment. We shall only deal with interactions and interface chemistry which is directly relevant to adsorption-desorption behaviour. 

---

Katzenellenbogen N and Grischkowsky D 1991 Efficient generation of 380 fs pulses of THz radiation by ultrafast laser pulse excitation of a biased metal-semiconductor interface Appl. Phys. Lett. 58 222-4... 

Metals for Schottl Contacts. Good Schottky contacts on semiconductor surfaces should not have any interaction with the semiconductor as is common in ohmic contacts. Schottky contacts have clean, abmpt metal—semiconductor interfaces that present rectifying contacts to electron or hole conduction. Schottky contacts are usuaHy not intentionaHy annealed, although in some circumstances the contacts need to be able to withstand high temperature processing and maintain good Schottky behavior. 

Applications of CL to the analysis of electron beam-sensitive materials and to depth-resolved analysis of metal-semiconductor interfaces by using low electron-beam energies (on the order of 1 keV) will be extended to other materials and structures. 

A number of solid compounds have been examined with this time-domain method since the first report of coherent phonons in GaAs [10]. Coherent phonons were created at the metal/semiconductor interface of a GaP photodiode [29] and stacked GaInP/GaAs/GalnP layers [30]. Cesium-deposited [31-33] and potassium-deposited [34] Pt surfaces were extensively studied. Manipulation of vibrational coherence was further demonstrated on Cs/Pt using pump pulse trains [35-37]. Magnetic properties were studied on Gd films [38, 39]. 

The term photovoltaic effect is further used to denote non-electrochemical photoprocesses in solid-state metal/semiconductor interfaces (Schottky barrier contacts) and semiconductor/semiconductor pin) junctions. Analogously, the term photogalvanic effect is used more generally to denote any photoexcitation of the d.c. current in a material (e.g. in solid ferroelectrics). Although confusion is not usual, electrochemical reactions initiated by light absorption in electrolyte solutions should be termed electrochemical photogalvanic effect , and reactions at photoexcited semiconductor electodes electrochemical photovoltaic effect . 

Structure, Dynamics and Growth Mechanisms of Metal-Metal and Metal-Semiconductor Interfaces by Means of SEXAFS... 

Structural information on the atomic arrangements at the early stage of formation of metal-metal, metal-semiconductor interfaces and semiconductor-semiconductor heterojunctions is needed along with the determination of the structure of the electron states in order to put on a complete experimental ground the discussion of the formation of solid-solid junctions. Amongst the structural tools that have been applied to the interface formation problem Surface-EXAFS is probably the best... 

A surface sensitive version of the EXAFS technique has been attempted ten years ago, and has proven to be successful in a large variety of surface chemisorption and interface formation problems. In the following we recall very briefly what makes SEXAFS different from EXAFS and what is the specific information that can be withdrawn from the SEXAFS data, and address the problems of metal-metal interface formation, and metal-semiconductor interface formation with detailed examples. 

For metal/semiconductor interfaces the limitations coming from 1) are not very severe since the cross sections for photoemission and the lifetime broadening of such deep state photopeaks along with the poor energy resolution of the photon source... 

Chandesris, D., Roubin, P., and Rossi, G. Structure, Dynamics and Growth Mechanisms of Metal-Metal and Metal-Semiconductor Interfaces by Means of SEXAFS. 147, 95-119 (1988). Charton, M., and Motoc, I. Introduction, 114, 1-6 (1983). 

It is the electrode potential

electrochemical experiments it represents a potential difference between two identical metallic contacts of an electrochemical circuit. Such a circuit, whose one element is a semiconductor electrode, is shown schematically in Fig. 2. Besides the semiconductor electrode, it includes a reference electrode whose potential is taken, conventionally, as zero in reckoning the electrode potential (for details, see the book by Glasstone, 1946). The potential q> includes potential drops across the interfaces, i.e., the Galvani potentials at contacts—metal-semiconductor interface, semiconductor-electrolyte interface, etc., and also, if current flows in the circuit, ohmic potential drops in metal, semiconductor, electrolyte, and so on. (These ohmic drops are negligibly small under experimental conditions considered below.)... 

Electrical charges are produced on the metal-semiconductor interface and on the outer surface of the semiconductor, by means of a corona discharge (ionization of the air by high voltage). 

Figure 4.2(d) shows that an energy barrier forms at the semiconductor/redox electrolyte interface, similar to the Schottky barrier at a metal/semiconductor interface. The most important quantity is the barrier height (q ) or the flat band potential U, which essentially determines the surface band positions of the semiconductor with respect to the energy levels of solution species. The q B is given for an n-type semiconductor by... 

Figure 4.11 explains how the loaded metal can act as catalyst for both the reductive and oxidative reactions. The essential point is that the barrier height at the metal/semiconductor interface is changeable by the principle of Fig. 4 10 or other mechanisms. Thus, when the band bending is weak, photoexcited electrons mostly enter the metal and the metal acts as a catalyst for a reductive reaction (Fig. 4.11(A)). On the other hand, when the band bending is strong, the holes in the valence band mostly enter the metal and the metal acts as a catalyst for an oxidative reaction (Fig. 4.11(B)). The prevalence of one over the other depends on the magnitude of the band bending, i.e., on the relative rates of reaction of the electrons and holes at the metal-free semiconductor surface.

It is, of course, well known that metal-semiconductor interfaces frequently have rectifier characteristics. It is significant, however, that this characteristic has been confirmed specifically for systems that have been used as inverse supported catalysts, including the system NiO on Ag described above as catalyst for CO-oxidation. In the experimental approach taken, nickel was evaporated onto a silver electrode and then oxidized in oxygen. A space charge-free counter-electrode was then evaporated onto the nickel oxide layer, and the resulting sandwich structure was annealed. The electrical characteristic of this structure is represented in Fig. 8. The abscissa (U) is the applied potential the ordi-... 

Lawrence H. Dubois received his B.S. degree in chemistry from the Massachusetts Institute of Technology in 1976 and a Ph.D. in physical chemistry from the University of California, Berkeley, in 1980. Dubois then joined AT T Bell Laboratories in Murray Hill, NJ, to pursue studies of the chemistry and physics of metal, semiconductor, and insulator surfaces chemisorption and catalysis by materials formed at the metal-semiconductor interface and novel methods of materials growth and preparation. 

The application we have in mind for the metal-polymer interfaces discussed in this book is primarily that where the polymer serves as the electroactive material (semiconductor) in an electronic device and the metal is the electric contact to the device. Metal-semiconductor interfaces, in general, have been the subject of intensive studies since the pioneering work of Schottky, Stromer and Waibel1, who were the first to explain the mechanisms behind the rectifying behaviour in this type of asymmetric electric contact. Today, there still occur developments in the understanding of the basic physics of the barrier formation at the interface, and a complete understanding of all the factors that determine the height of the (Schottky) barrier is still ahead of us2. 

The aim of this chapter is to present a simple but general band structure picture of the metal-semiconductor interface and compare that with the characteristics of the metal-conjugated polymer interface. The discussion is focused on the polymer light emitting diode (LED) for which the metal-polymer contacts play a central role in the performance of the device. The metal-polymer interface also applies to other polymer electronic devices that have been fabricated, e.g., the thin-film field-effect transistor3, but the role of the metal-polymer interface is much less cruical in this case and... 

---