Semiconductors, amorphous solids

The two extremes of ordering in solids are perfect crystals with complete regularity and amorphous solids that have little symmetry. Most solid materials are crystalline but contain defects. Crystalline defects can profoundly alter the properties of a solid material, often in ways that have usefial applications. Doped semiconductors, described in Section 10-, are solids into which impurity defects are introduced deliberately in order to modify electrical conductivity. Gemstones are crystals containing impurities that give them their color. Sapphires and rubies are imperfect crystals of colorless AI2 O3, red. 

Hosono, H. 2006. Ionic amorphous oxide semiconductors Material design, carrier transport, and device application. J. Non-Cryst. Solids 352 851-858. 

The difference between amorphous and crystalline solids shows up more clearly in phonon excitations than in electronic excitations. Contributions from the entire phonon density of states appear in the first-order Raman and infrared spectra of amorphous solids. All modes in elemental amorphous semiconductors are active in the infrared. 

It is necessary to note that in addition to the problems associated with the theoretical analysis of electronic conduction in disordered systems, experimental techniques of established success in the study of crystalline semiconductors have proven to yield ambiguous results when applied to amorphous solids. 

Bulk crystalline or amorphous solid-state materials whose conductivity is intermediate between metals and insulators and whose resistance decreases with increasing temperature. The valance band of an undoped semiconductor is completely filled, whereas its conduction band is empty. The energy difference between the valence and conduction bands (the band-gap) defines a semiconductor (see Fig. 95). 

See Proceedings of Symposium on Semiconductor Effects in Amorphous Solids, Edited by W. Doremus, North-Holland, Amsterdam, 1970. 

The question is now what sort of order, if any, is conserved in a solid semiconductor when the long-range order is lost altogether, not only slightly disturbed. Such a solid is called amorphous. 

One is also forced to admit that the crystallites are 10—20 A in size. The crystallites must be oriented at random in order to achieve the macroscopic isotropy of the amorphous solid. High-angle boundaries should therefore be present. The thickness of these often postulated completely disordered regions which separate the crystallites from each other must be of the same order as the size of the crystallites themselves. It is difficult to admit that about half the volume of an amorphous semiconductor could be completely disordered without also admitting a high fraction of the covalent bonds being unsatisfied (Warren (1937)). 

Davis and Mott s hypothesis simplifies considerably the following discussions and we shall adopt it. However, its validity is subject to justification in any particular theoretical model of the electronic structure of an amorphous solid. Hindley (1970) has shown that it follows from his random phase model for the wavefunctions in amorphous semiconductors. 

We have seen that there are many possible and plausible sources of potential fluctuations in amorphous solids which may be responsible for the broadening of the edges. It is rather surprising that with such a large variety of possible sources Parts B of many amorphous semiconductors are closely similar. This observation suggests that there may be some internal mechanism which adjust the average value of internal fields into relatively narrow limits. 

The third section deals with the electronic properties of amorphous semiconductors. These are compared with and described in terms of transport equations which treat the amorphous solid as a rather ideal homogeneous, uniform material. 

Doremus, W., ed., (1970) Semiconductor Effects in Amorphous Solids , / Non-Cryst. Solids, 2,1-575. 

A large class of amorphous solids — glasses — have been studied extensively over many decades from the point of view of their thermodynamical and structural properties. The first two chapters deal with this aspect which is considered essential for any deeper understanding of the nature of the amorphous state. The inherent metastalpility of the amorphous state compared to the crystalline state is responsible for many discrepancies in the experimental results. It should be understood not only for improving the reproducibility of the measurements but also because the control of the structural parameters is the heart of such applications of amorphous semiconductors which cannot be produced by devices constructed of crystalline materials. Indeed, most kinds of memory devices are based on the ability of glasses to exist in different structural states. 

---