Search

Herein, a theoretical model is proposed for the weakly temperature‐dependent electrical conductivity of compensated crystalline semiconductors with hydrogen‐like impurities near the insulator–metal concentration phase transition (Mott transition). The model uses a simple non‐stoichiometric cubic "impurity lattice" formed by the doping and compensating impurities in crystal matrix. A shift of the c‐band bottom (v‐band top) into the bandgap due to overlap of the excited states of neighboring impurities is considered. The distribution of electron (and hole) density of states in the band of ground (unexcited) states of impurities is assumed to be Gaussian. Tunneling transitions of electrons between nearest donors in the charge states (0) and (+1), and tunneling transitions of holes between acceptors in the charge states (0) and (−1) are considered. It is shown that, at low temperatures, transitions of electrons (holes) near the Fermi level in the impurity band lead to electrical conductivity that weakly depends on temperature (in the form of a characteristic plateau). The results of calculating electrical resistivity in the zero‐temperature limit for the plateau region agree with the known experimental data for moderately compensated n‐ and p‐type Ge, Dia, Si, ZnSe, GaAs, InSb, and InP crystals. [ABSTRACT FROM AUTHOR]

A model of tunneling (jumping) migration of charge carriers near their mobility edge in the upper band of neutral states of majority hydrogen‐like impurities is proposed to calculate the energy of thermal activation of electrical ε 2 $\left(\epsilon\right)_{2}$ ‐conductivity of weakly compensated semiconductors. The difference from the known Hubbard model consists in the scheme of interimpurity transitions of charge carriers and in the method of calculating the position of their tunnel mobility edge. The drift mobility edge of free charge carriers corresponds to the thermal ionization energy of majority impurities ε 1 > ε 2 $\left(\epsilon\right)_{1}> \left(\epsilon\right)_{2}$ , which is located near the c‐band bottom or the v‐band top in n‐ and p‐type semiconductors, respectively, and is due to the overlap of excited states of electrically neutral majority impurities. The position of the tunnel mobility edge for ε 2 $\left(\epsilon\right)_{2}$ ‐conductivity is determined by taking into account the Coulomb interaction of the majority impurities in the charge states ( − 1 ) $\left(\right. - 1 \left.\right)$ and ( + 1 ) $\left(\right. + 1 \left.\right)$ . It is assumed that doping and compensating impurities form a single simple nonstoichiometric cubic lattice in a crystal matrix. The calculations of the activation energy ε 2 $\left(\epsilon\right)_{2}$ on the insulator side of the insulator–metal concentration phase transition for weakly compensated p‐Si:B, n‐Si:P, and n‐Ge:Sb crystals quantitatively agree with known experimental data. [ABSTRACT FROM AUTHOR]