The conductivity of tin-doped indium oxide (ITO), a transparent conductor, is critically dependent on the amount of tin doping and oxygen partial pressure during preparation and annealing. Frank and Köstlin (Appl. Phys. A, 27, 197-206 (1982)) have rationalized the observed carrier concentration dependence by postulating the formation of two types of neutral defect clusters at medium tin-doping levels: "reducible" and "nonreducible" defect clusters, so named to indicate their ability to create carriers under reduction. According to Frank and Köstlin, both clusters are composed of one oxygen interstitial and two tin atoms substituting for indium, positioned in nonnearest and nearest coordination, respectively. This work, seeking to distinguish reducible and nonreducible clusters by means of an atomistic model, finds only a weak correlation of oxygen interstitial binding energies with the relative positioning of tin dopants. Instead, the number of tin dopants in the vicinity of the interstitial has a much larger effect on how strongly it is bound, a simple consequence of Coulomb interactions. We postulate that oxygen interstitials become nonreducible when clustered with three or more Sn In. This occurs at higher doping levels as reducible clusters aggregate and share tin atoms. A simple probabilistic model, estimating the average number of clusters so aggregated, yields a qualitatively correct description of the carrier density in reduced ITO as a function of tin-doping level.
|Number of pages||5|
|Journal||Journal of the American Ceramic Society|
|Publication status||Published - Oct 2003|
ASJC Scopus subject areas
- Ceramics and Composites
- Materials Chemistry