A statistical Seebeck coefficient model based on percolation theory in two-dimensional disordered systems.

In the presence of structural disorders, carrier conduction via localized hopping sites emerges in two-dimensional systems and results in a unique thermopower characteristic with T1/3 dependence. The disorders induced potential differences of hopping sites leading to energy variations along current-carrying paths. A systematic thermoelectric study is presently required in comprehending the statistical effects. Therefore, we proposed a statistical model of the Seebeck coefficient on the basis of percolation theory and hopping mechanisms. With this model, the carrier density and temperature dependences can be practically predicted. Key parameters can be extracted by calibration to molybdenum disulfide and black phosphorus experiments, providing a deeper insight into device physics. Moreover, a Mott-like analytical model is developed to investigate the parametric dependence. The thermopower deviations from the noninteracting Mott picture at high and low temperatures are analyzed. Finally, the temperature dependence on the thermoelectric figure of merit is evaluated in a variable range hopping regime. Our model is essential for a reliable prediction of the disorder induced statistical effects on thermoelectric behaviors, which guides both device optimization and material engineering. [ABSTRACT FROM AUTHOR]