## Abstract

In characterizing thermoelectric materials, electrical and thermal transport measurements are often used to estimate electronic band structure properties such as the effective mass and band gap. The Goldsmid-Sharp band gap, E_{g} = 2e|S|_{max}T_{max}, is a tool widely employed to estimate the band gap from temperature dependent Seebeck coefficient measurements. However, significant deviations of more than a factor of two are now known to occur. We find that this is when either the majority-to-minority weighted mobility ratio (A) becomes very different from 1.0 or as the band gap (E_{g}) becomes significantly smaller than 10 k_{B}T. For narrow gaps (E_{g} ≲ 6 k_{B}T), the Maxwell-Boltzmann statistics applied by Goldsmid-Sharp break down and Fermi-Dirac statistics are required. We generate a chart that can be used to quickly estimate the expected correction to the Goldsmid-Sharp band gap depending on A and S_{max}; however, additional errors can occur for S < 150 μV/K due to degenerate behavior.

Original language | English |
---|---|

Article number | 022112 |

Journal | Applied Physics Letters |

Volume | 106 |

Issue number | 2 |

DOIs | |

State | Published - 12 Jan 2015 |

## Fingerprint

Dive into the research topics of 'Band gap estimation from temperature dependent Seebeck measurement - Deviations from the 2e|S|_{max}T

_{max}relation'. Together they form a unique fingerprint.