COMPUTATIONAL APPROACHES TO THE CHEMICAL-SENSITIVITY OF SEMICONDUCTING TIN DIOXIDE

Some computational approaches to the chemical sensitivity of semicondu cting tin dioxide are presented. Chemical sensitivity is often observe d using conductance measurement. Therefore, the potential energy barri ers in grain contacts between adjacent grains of a polycrystalline sem iconductor are the key parameters for transducing the chemical surface sensitivity into the conductance response. The rate equation model de scribes the electronic exchange between the adsorbed oxygen species an d the bulk conduction band of a semiconductor. It predicts the type of the major negative oxygen ion (O-2(-) or O-) at the surface as a func tion of temperature in agreement with experimental findings. The grain geometry has only a small effect on the potential energy barrier at t he surface of finite grains. Even a small neck contact between grains, in the case of mobile donors, decreases strongly the potential energy barrier between grains compared to that in the case of an open grain contact. Results from Monte Carlo simulations with random barrier netw orks reveal that the current-voltage characteristic of a polycrystalli ne semiconductor is non-linear at higher voltages and the non-linearit y of the network increases with increasing width of the barrier distri butions. Electronic-structure calculations with clusters give qualitat ive information on the role of oxygen vacancies in different atomic pl anes in SnO2 and its unrelaxed and unreconstructed (110) surface. (C) 1998 Elsevier Science S.A. All rights reserved.