Space-resolved Temperature and Heat Sensing by Thermoelectrics

The research on thermoelectricity is mainly focused on the development of technologies for energy harvesting, cooling and waste heat recovery. However, the first and most consolidated application of thermoelectricity was and still is in temperature sensing. Currently, technologies are available to extend metrological applications to achieve space-resolved measurements of temperature (and of heat fluxes thereof) by designing arrays of thermocouples with a space resolution up to 200 μm either for temperature or power density mapping. Temperature mapping will serve the aim of adding thermal sensitivity to a robot developed by ISC/Oversonics to be prospectively deployed in Phase IV of the Artemis space program. To this aim, a temperature resolution of 0.5 K is required for a device capable of reliably operation between –150 and +200 °C in the presence of ionizing radiation. Instead, heat flux maps are requested to support beam focalization in the Legnaro National Institute of Nuclear Physics (INFN) facility. To this purpose, incoming energy densities from 1 nW to 0.1 mW need to be measured. Although different by aim and context, both applications call for the making of an integrated matrix of thermocouples [1, 2]. In this presentation we will describe their layout, reporting the results of their preliminary design. In short, we will show how radiation hardness dictates materials choice, while sensitivity rules array layout and application context guides manufacturing. Concerning aptic sensing, bidirectional curvature of finger surfaces discourages the use of flexible substrates, suggesting instead direct printing/deposition of metallic thermocouples as the most reliable manufacturing protocol. Instead, since heat flux measurements will be carried out on flat surfaces, the use of more conventional microfabrication methods will be considered. The demanding issue is in this case that of a critically low heat flux detection threshold, estimated to be in the order of 100 nJ/pixel (with an integration time of 90 s). Although largely within the capabilities of semiconducting thermoelectric devices, preliminary evidence shows how the device design must grant a highly efficient heat withdrawal at the cold side to sustain the small temperature difference (≈ 1 K) generated by the full thermalization of the particle beam.