A solid state electronic device composed mainly of amorphous semiconductor materials. Although the amorphous semiconductor molecule is disordered on the whole, it still has the microstructure of a single crystal, so it has many special properties. In 1975, the British W.G. Speer doped successfully in the amorphous silicon film prepared by glow discharge decomposition silane method, so that the resistivity of the amorphous silicon film changed by 10 orders of magnitude, and promoted the development and application of amorphous semiconductor devices. Compared with single crystal materials, amorphous semiconductor materials have simple preparation process, no special requirements for substrate structure, easy to grow in large areas, and can be made into a variety of devices after doping. Amorphous silicon solar cells have large absorption coefficient, high conversion efficiency and large area, and have been applied to calculators, electronic watches and other commodities. Amorphous silicon thin film FET arrays can be used as addressing switches for large area LCD flat panel displays. Devices that record and store photoelectric information using the structural transformation of some chalcosulfide amorphous semiconductor materials have been used in computers or control systems. The charge storage and photoconductivity properties of amorphous films can be used to make photoreceptors for static images and target surfaces for TV pick-up tubes for dynamic images.
Amorphous materials having semiconductor properties. Amorphous semiconductor is an important part of semiconductor. In the 1950s, B.T. Kolomietz and others began to study chalcogenide glass, but little attention was paid until 1968. It was only after the publication of Ovshinsky's patent on the production of switching devices from sulfur films that interest in amorphous semiconductors was aroused. In 1975, W.E. Speer et al. realized doping effect in amorphous silicon prepared by silane glow discharge decomposition, which made it possible to control conductance and manufacture PN formation, thus opening up broad prospects for the application of amorphous silicon materials. On the theoretical side, P.W. Anderson and Mott, N.F. developed the electronic theory of amorphous semiconductors, for which they were awarded the 1977 Nobel Prize in Physics. At present, the research of amorphous semiconductors is developing rapidly both in theory and in application.
At present, there are two main categories of amorphous semiconductors.
Chalcogenide glass. Amorphous semiconductor containing chalcogenide elements. For example, As-Se and As-S are usually prepared by melt cooling or vapor deposition.
Tetrahedral bonded amorphous semiconductor. Such as amorphous Si, Ge, GaAs, etc., the amorphous state of such materials can not be obtained by melt cooling method, only by thin film deposition method (such as evaporation, sputtering, glow discharge or chemical vapor deposition, etc.), as long as the substrate temperature is low enough, the deposited film is amorphous structure. The properties of tetrahedral bonded amorphous semiconductor materials are closely related to the preparation process and conditions. Figure 1 Optical absorption coefficients of amorphous silicon prepared by different methods The optical absorption spectra of amorphous silicon prepared by different processes are given. The preparation process of a and b is silane glow discharge decomposition, the substrate temperature is 500K and 300K respectively, the preparation process of c is sputtering, and the preparation process of d is evaporation. The conductive and photoconductive properties of amorphous silicon are also closely related to the preparation process. In fact, the amorphous silicon prepared by silane glow discharge method contains a large amount of H, sometimes called amorphous silane hydrogen alloy; Different process conditions, different hydrogen content, directly affect the properties of the material. In contrast, the properties of chalcogenide glass have little to do with the preparation method. FIG. 2 The spectra of the light absorption coefficients of AsSeTe, a typical example, are given for the AsSeTe samples prepared by melt cooling and sputtering, which have the same curves.
The electronic structure of amorphous semiconductors Amorphous has a basic band structure similar to that of crystalline semiconductors, but also has conduction, valence, and bandgap bands (see energy bands of solids). The basic band structure of materials depends mainly on the conditions near the atoms, which can be qualitatively explained by the chemical bond model. Taking tetrahedral bonding amorphous Ge and Si as an example, the four valence electrons in Ge and Si are hybridized by sp, and covalent bonds are formed between the valence electrons of neighboring atoms, and their bonding states correspond to valence bands. The antibonding state corresponds to the conduction band. Whether Ge or Si is crystalline or amorphous, the basic bonding mode is the same, but there is a certain degree of distortion in the bond Angle and bond length in the amorphous state, so their basic band structure is similar. However, the electronic states in amorphous semiconductors are also fundamentally different from the crystalline states. The structure of the crystalline semiconductor is periodic ordered, or has translation symmetry, the electron wave function is the Bloch function, the wave vector is the quantum number associated with translation symmetry, and the amorphous semiconductor does not have periodicity and is no longer a good quantum number. The motion of electrons in crystalline semiconductor is relatively free, and the mean free path of electron motion is much larger than the atomic distance. The distortion of structural defects in amorphous semiconductors greatly reduces the mean free path of electrons, and when the mean free path approaches the order of magnitude of the atomic distance, the concept of electron drift motion established in crystalline semiconductors becomes meaningless. Amorphous semiconductors do not change the edge state density as sharply as the crystal state, but drag different degrees of band tail (as shown in Figure 3 the relationship between state density and energy of amorphous semiconductors). The electronic states in the amorphous semiconductor band are divided into two categories: one is called the extended state and the other is the localized state. Each electron in the extended state is shared by the whole solid and can be found in the entire scale of the solid. It behaves in the outer field like an electron in a crystal; Each electron in the local state is basically confined to a certain region, and its state wave function can only be significantly non-zero in a small scale around a certain point, and they need the help of phonons to conduct electricity by leaps and bounds. In a band, the central part of the band is an extended state, and the tail of the band is divided into local states, and there is a boundary between them, as shown in Figure 4, the sum of the extended state, local state and mobility edge of the amorphous semiconductor. This boundary is called the mobility edge. In 1960, Mott first proposed the concept of mobility edge. If mobility is regarded as a function of the energy of the electronic state, Mott believes that there is a sudden change in mobility at the boundary. The electrons in the local state conduct electricity by leaps and bounds, and jump from one local state to another by exchanging energy with the lattice vibration, so the mobility of the electrons in the local state tends to zero as the temperature approaches 0K. The conduction of electrons in the extended state is similar to that of electrons in a crystal, and the mobility tends to a finite value as it approaches 0K. Mott further argued that the mobility edge corresponds to the case where the mean free path of the electron is close to the atomic distance, and defined the conductivity in this case as the minimum metallization conductivity. However, there are still debates around the mobility edge and the minimum metallization conductivity.
Compared with the crystalline state, amorphous semiconductors have a large number of defects. These defects introduce a series of local energy levels in the band gap, which have important effects on the electrical and optical properties of amorphous semiconductors. Tetrahedron-bonded amorphous semiconductors and chalcogenide glasses, the defects of these two types of amorphous semiconductors are significantly different.
The defects in amorphous silicon are mainly vacancies and microvoids. A silicon atom has four valence electrons in its outer layer, which normally form four covalent bonds with four nearby silicon atoms. The presence of vacancy and microvoids causes some silicon atoms to have insufficient four neighboring atoms, resulting in suspension bonds with an unbonded electron on the neutral suspension bond. There are two possible charged states of the suspended bond: the release of unbonded electrons to become the positive center, which is the donor state; Accepting the second electron to become the negative center is the acceptor state. Their corresponding energy levels are in the band gap and are called donor and acceptor energy levels respectively. Because the acceptor state indicates that there are two electrons occupied on the suspension bond, the coulomb repulsion between the two electrons makes the acceptor energy level higher than the donor energy level, which is called positive correlation energy. Therefore, in general, the suspension bond remains in a neutral state occupied by only one electron, and spin resonances of unpaired electrons on the suspension bond have been observed experimentally. In 1975, Speer et al. used the silane glow discharge method to first achieve the doping effect of amorphous silicon, because the amorphous silicon prepared by this method contains a large amount of hydrogen, and the combination of hydrogen and suspension bonds greatly reduces the number of defect states. These defects are also effective recombination centers. In order to increase the lifetime of non-equilibrium carriers, the density of defect states must also be reduced. Therefore, controlling the defects in amorphous silicon has become one of the key problems in material preparation.
The form of the defect in the chalcogenide glass is not a simple suspension bond, but a \"exchange pair\". At first, it was found that chalcogenide glass, unlike amorphous silicon, could not observe the spin resonance of electrons in the defect state. In view of this abnormal phenomenon on the surface, Moter et al. proposed the MDS model according to Anderson's assumption of negative correlation energy. When two electrons are occupied in the defect state, the lattice distortion will be caused. If the energy reduced by the distortion exceeds the coulomb repulsion energy between electrons, there is a negative correlation energy, which means that the acceptor energy level is below the donor energy level. With D, D and D respectively representing the states of no possession, one possession and two possession of electrons on the defect, negative correlation energy means:
2 D -→D D
It's exothermic. Therefore, the defects mainly exist in the form of D and D, and there is no unpaired electron, so there is no spin resonance of the electron. Many people have analyzed the structure of D, D and D defects. Taking amorphous selenium as an example, selenium has six valence electrons, which can form two covalent bonds, usually in a chain structure, and two unbonded p electrons called lone pair electrons. There is equivalent to a neutral suspension bond at the end of the chain, which is likely to be distorted, bonding with a nearby lone pair electron and releasing an electron (forming D), which is combined with another suspension bond to form a pair of lone electrons (forming D), as shown in Figure 5. Therefore, it is also called D and D for exchange pairs. Due to coulomb attraction, D and D are usually close together in pairs, forming a close exchange pair. As long as there is a small change in the bonding mode in the chalcogenide glass, a group of tight exchange pairs can be formed, as shown in the self-reinforcing effect of the exchange pair in Figure 6. It requires only a small amount of energy and has a self-reinforcing effect, so the concentration of such defects is usually very high. A series of experimental phenomena such as photoluminescence spectra and photoluminescence electron spin resonance of sulfur amorphous semiconductors can be explained by the exchange pair model.
There is great potential for the application of amorphous semiconductors in the field of technology, and amorphous sulfur has long been widely used in photocopying technology, by S.R. Ovsinski's As-Te-Ge-Si series of glass semiconductor has been commercially produced, and the optical memory made of telluride microcrystalline film by optical pulses is being developed. The most studied application of amorphous silicon is solar cells. Amorphous silicon than crystalline silicon preparation process is simple, easy to make a large area, amorphous silicon for sunlight absorption efficiency is high, the device only needs to be about 1 micron thick film material, therefore, is expected to make a cheap solar cell, has been the attention of energy experts. Recently, amorphous silicon field-effect transistors have been tested for use in liquid crystal displays and integrated circuits.
The substrate is a clean single crystal sheet with specific crystal faces and appropriate electrical, optical and mechanical properties for growing epitaxial layers.
< br >
role
< br >
1. Support. A few microns or even a few nanometers thick film must be attached to the substrate to be difficult to break and destroy.
< br >
2. Participate in conducting electricity. Many substrates are semiconductors, such as silicon, which form heterojunctions with functional materials and participate in the implementation of device functions.
< br >
3. Growth. Some films must be on a suitable substrate in order to grow the required material, involving problems such as lattice structure.