Analysis of the ‘cold finger effect’ in measuring the Seebeck coefficient

Abstract

This work analysed one of the most important uncertainty contributions in measuring the Seebeck coefficient of thermoelectric materials: the ‘cold finger effect’ (CFE). A didactic approach based on the fundamental laws of thermoelectric circuits is used to explain the influence of this effect on the Seebeck coefficient measured. Finite element simulations and practical measurements of an ISOTAN® sample were performed under different conditions relating to the thermal coupling of the probes to the sample, and the results are presented. Relative errors of the Seebeck coefficient of up to a few percent can occur if measurements are conducted under non-optimum measuring conditions. It is shown how the thermal coupling of the probes to the sample can minimize the influence of the CFE on the measured Seebeck coefficient.