Anderson-localized states

We discuss in this section the effect of short-range interaction on the Anderson-localized states of a Fermi glass described in Chapter 1, Section 7, and in particular the question of whether the states are singly or doubly occupied. Ball (1971) was the first to discuss this problem. In this section we consider an electron gas that is far on the metal side of the Wigner transition (Chapter 8) the opposite situation is described in Chapter 6, where correlation gives rise to a metal-insulator transition. We also suppose that Anderson localization is weak (cca 1) otherwise it is probable that all states are singly occupied. 

In addition to Mott-Hubbard localization, there is another common source of electron localization, which arises when a lattice is under a random potential (e.g. a random distribution of alkali metal ions in alkali metal containing transition metal oxides). For a metal, a practical consequence of a random potential is to open a band gap at the Fermi level. Insulating states induced by random potentials are referred to as Anderson localized states (see Anderson Localization)) ... 

Anderson localization is the localization of electrons on low-dimensional materials, which is induced by the irregularity of the periodic potential field [43]. Figure 17 gives a schematic representation of Anderson localization of a particle in one-dimensional box. The same is true for an electron on a polymer skeleton. A localized state in a completely periodic... 

If the Anderson criterion is not satisfied then, as first pointed out by Mott (1966), since states are likely to become localized in the tail of a band, there exists a critical energy Ec (the mobility edge )f separating localized from non-localized states (Fig. 1.21). The simplest definition of Ec in terms of the behaviour of the conductivity a(E) is as follows ... 

Buyers et al (1971) discussed evidence for the existence of Anderson-localized spin states m antiferromagnets in which there is disorder. They investigated the substitutionally disordered materials K(Co,Mn)F3 and (Co,Mn)F2, finding two branches of propagating spin waves corresponding to the two constituents and... 

Throughout we make use of the pseudogap model outlined in Chapter 1, Section 16- A valence and conduction band overlap, forming a pseudogap (Fig. 10.1). States in the gap can be Anderson-localized. A transition of pure Anderson type to a metallic state (i.e. without interaction terms) can occur when electron states become delocalized at EF. If the bands are of Hubbard type, the transition can be discontinuous (a Mott transition). 

Therefore the lack of an observable bleach can only be explained by the cancellation of all contributions to the pump-probe signal, which is the case for a perfect harmonic state. It can be shown that the anharmonicity of a vibrational exciton is a direct measure of its degree of delocalization [5]. Thus, we conclude that the free exciton state is almost perfectly delocalized at 90 K. As temperature increases, a bleach signal starts to be observed, pointing to a non-complete cancellation of the different contributions of the total pump-probe signal. Apparently, thermally induced disorder (Anderson localization) starts to localize the free exciton. The anharmonicity of the self-trapped state (1650 cm 1), on the other hand, originates from nonlinear interaction between the amide I mode and the phonon system of the crystal. It... 

Attempts to take into account both localization and percolation or, in other words, to allow for quantum effects in percolation go back to Khmel-nitskii s pioneer paper [68]. The experimental attempts to study quantum effects in conductivity close to the percolation threshold have been undertaken in Refs. [69-71]. The physical sense of these results is stated in Ref. [71] and could be described as follows. The percolation cluster is non-uniform it includes both big conductive regions ( lakes ) and small regions (weak links or bottlenecks) which connect lakes to each other. On approaching the percolation threshold from the metallic side of the transition, these weak links become thinner and longer, and at x = xc the cluster breaks or tears into pieces just in such areas. As a result, exactly these conditions start to be sufficient for the electron localization. Thus, a percolation provokes an Anderson localization in bottlenecks of the percolation cluster. Sheng and collaborators [36,37,72] tried to take into account the influence of tunneling on conductivity for systems in the vicinity of the percolation transition. Similar attempts have been made in papers [38,56]. The obtained results prove that the possibility of tunneling shifts the percolation threshold toward smaller x values and affects material properties in its vicinity. 

It has been seen in the previous section that the ratio of the onsite electron-electron Coulomb repulsion and the one-electron bandwidth is a critical parameter. The Mott-Hubbard insulating state is observed when U > W, that is, with narrow-band systems like transition metal compounds. Disorder is another condition that localizes charge carriers. In crystalline solids, there are several possible types of disorder. One kind arises from the random placement of impurity atoms in lattice sites or interstitial sites. The term Anderson localization is applied to systems in which the charge carriers are localized by this type of disorder. Anderson localization is important in a wide range of materials, from phosphorus-doped silicon to the perovskite oxide strontium-doped lanthanum vanadate, Lai cSr t V03. 

The existence of localized states was predicted early on in the studies of amorphous semiconductors by the Anderson localization theory (Section 1.2.5) and their presence is well established ex-... 

The next step in the theory is to calculate the conductivity above and below the mobility edge. In the Anderson model, locali2ed states are defined by a decreasing probability that the electron diffuses a larger distance from its starting point. Mott and Davis (1979 Chapter 1) prove that the dc conductivity in the localized states is zero at T = 0 K. They use the Kubo-Greenwood formula for the conductivity,... 

A fundamental result of Anderson localization is that there is no conductivity in localized states at zero temperature. However, transitions between localized states can take place at elevated temperatures and result in a hopping conductivity. The transition probability, between two states separated by distance R and energy Wis (see Section 1.2.5),... 

Anderson s simple model to describe the electrons in a random potential shows that localization is a typical phenomenon whose nature can be understood only taking into account the degree of randomness of the system. Using a tight-binding Hamiltonian with constant hopping matrix elements V between adjacent sites and orbital energies uniformly distributed between — W/2 and W/2, Anderson studied the modifications of the electronic diffusion in the random crystal in terms of the stability of localized states with respect to the ratio W/V. 

We emphasize first of all that the disorder in the subsurface region, i.e. in the region of localization of the surface exciton, may result from its own internal disorder or may be caused by other, external reasons (e.g. by absorbed molecules). The microscopic surface states under consideration are strongly affected by both types of disorder. This circumstance should be borne in mind even in the cases when SSSE are treated in isotopically disordered crystalline solutions. In such states, which interact weakly with internal crystal monolayers, the effect of internal and external disorder can result in equally serious consequences. We will now make some qualitative remarks on the Anderson localization of surface excitons. As before let us assume a molecular crystal, ignore the exciton-phonon interaction, but take into account, for instance, the diagonal disorder (i.e. the random energy distribution of molecular excitations). 

---