Modeling Fermi energy, free-carrier density, and resistivity in degenerate n-Ge

Luigi Abenante

DOI: https://doi.org/10.1063/5.0163730

IF: 1.697

2024-02-01

AIP Advances

Abstract:A new expression for Fermi energy vs doping is derived using the standard model for free carriers in n-type semiconductors. The new expression is composed of the Fermi energy in non-degenerate semiconductors, a doping function for bandgap narrowing (BGN), and an adjustable energy variation. In non-degenerate semiconductors, the new expression is equivalent to the standard Boltzmann expression. Calculated curves of Fermi energy are assigned in the Fermi–Dirac expression for the donor ionization ratio, and reported data of electron density and resistivity measured in heavily doped n-Ge layers are fitted. Five reported doping functions for BGN are used. One of the BGN functions allows modeling frustrated incomplete ionization. Another allows modeling bandgap widening.

materials science, multidisciplinary,physics, applied,nanoscience & nanotechnology

What problem does this paper attempt to address?

The paper primarily addresses the issue where the standard model fails to accurately predict the free carrier density (i.e., electron density) in highly doped n-type germanium (n-Ge). The authors found that at high doping levels, the traditional model significantly underestimates the electron density due to the neglect of heavy doping effects. To solve this problem, the authors proposed a new expression for the relationship between the Fermi level and the doping concentration.

Specifically, the new model includes the Fermi level in non-degenerate semiconductors, the Bandgap Narrowing (BGN) effect, and an adjustable energy variation term. In the non-degenerate case, this model degenerates into the traditional Boltzmann expression. By substituting this new Fermi level expression into the Fermi-Dirac statistics describing the donor ionization ratio, the authors were able to better fit the experimentally measured electron density and resistivity data.

The authors also compared several different BGN functions and found that one of the functions could simulate the "frustrated incomplete ionization" phenomenon, while another function suggested the possibility of bandgap widening in certain materials. By adjusting the parameters in the model, the authors successfully fitted experimental data from different literature sources, indicating that the newly proposed model is more accurate and effective in handling highly doped semiconductors.