Segmented FGM for linearization of the characteristics of β-FeSi2 based thermal sensors

Highly sensitive thermal sensors can be developed when replacing metallic alloys in high temperature thermoelectric (TE) detectors by semiconducting materials. Functionally graded materials (FGM) can be used to tune the temperature characteristics of the signal responsivity, which should be independent from temperature. A technique of preparing segmented TE FGM of FeSi₂ doped with Al, FeSi₂ doped with Mn, or FeSi₂ double doped with Mn and Al is demonstrated. A correlated variation of the TE properties caused by change of the doping is found. The doping content of the material was varied mainly to control the temperature dependence of the Seebeck coefficient S. For highly responsive heat flux sensors, high S and low K are desired. Al+Mn double doping was employed to achieve proper S values. The linearization of sensor characteristics needs in the easiest approach a temperature independent S. This aim was met using a stack of segments with different temperature curves of the Seebeck coefficient. Constant integral thermopower between -50°C and +500°C with a comparably high value of about 270 μV/K was numerically predicted for the stack with a dispersion below 5%. Accordingly, a segmented thermocouple has been prepared after the modeling results and has been characterized