Thermoelectric antimonides for high temperature applications – HIGHTHERM

The first task is to reproduce the results obtained in a previous project, of course using the techniques available today (this will most likely generate different - hopefully only slightly - results). Following this, further optimisation of selected compostions will be done using calculation (DFT), rapid synthesis, standard processing and transport property measurements, this vertuous cycle leading quickly to the selection of the best materials. Once down-selected, these compounds will be produced in larger quantity and accelerated ageing studies will be performed. Only then, could the fabrication of thermoelectric legs be achieved, in a single step processing, joining the active materials to the other layers needed (diffusion barrier, current collector, etc). This crucial step will enable us to proceed further in the making of a demonstrator that could in turn be tested in real life conditions.

N/A

N/A

N/A

This project is focused on thermoelectric materials with potential industrial applications at very high temperature, from 600°C to 1000°C and possibly higher, to harvest waste heat and convert it into usable energy. This particular temperature range targets the steel, non-ferrous, ceramics and glass industries that use a lot of energy, 50% of which being lost during the production process. this project will target the research and development of high temperature stable thermoelectric materials based on the cubic structure Th3P4. In this family of intermetallics, n-type La3Te4-x are already known as good thermoelectric materials with ZT above the unity above 1000°C. This project therefore proposes to develop a p-type counterpart of the same structure type, e.g rare-earth antimonides crystallizing in the anti-Th3P4 structure, making it easier to fabricate p-n thermoelectric couples. However, there is only scarce information about the p-type counterparts, even if a few reports have shown very promising thermoelectric properties and stability at high temperature, for instance a ZT of 0.75 was reported in La0.5Yb3.5Sb3 at 1000°C. These materials, their optimization (using modeling tools) and their implementation within a thermoelectric uni-couple and the subsequent demonstrator tests and qualifications are the focus of this proposal.
The materials will be made via mechanical alloying followed by annealing and spark plasma sintering. These particular techniques have already proven that they allow the control of whole process, thus assuring the reproducibility of the obtained thermoelectric properties.
With the development of powerful methods to compute the electronic band structure of solids and the increasing complexity of the formulations of advanced thermoelectric materials such as those targeted in this project, quantum chemical calculations based on density functional theory (DFT) will be used for the optimization of thermoelectric material properties. DFT programs embedding the most advanced approximation of the exchange-correlation functionals and taking into account relativistic effects will be employed to calculate the electronic structures required to use band engineering approach for the optimization of the thermoelectric properties of the studied materials.
The third step of the project will focus on the making of the TE legs, namely, the active materials contacted on both side by metallic electrodes. This will be achieved using the LINK facilities and equipment and will be fed by the data available on the n type material (La4Te3-x) developed by NASA-JPL. CRISMAT will participate in the making of the metallized legs, metallographic studies will be performed on the different bondings and transport properties will be monitored upon ageing of the assemblies.
Finally, the last challenge will be to actually build a thermoelectric converter. In essence, it consists of several unicouples connected electrically in series to form a module. The power delivered by such device obviously depends on the number of unicouples. In order to keep the project realistic and in order to be able to respond quickly to necessary design modification, small demonstrators will be privileged over large units.
Besides characterization of the TE modules, these tests would serve to anticipate the applicability of the modules in industrial conditions, and anticipate potential modifications to the original design. The efficiency and durability of the module, will be used to estimate how much energy can be recovered and the economic advantage for an industrial application. Data will serve to identify other potential application domains for the modules, according to industrial process characteristics.
The consortium ideally combines the expertise of well know research center, CRISMAT laboratory, IRSN Rennes, NIMS Tsukuba via the UMI LINK, and an end user: St Gobain via the CREE research center and also via its belonging to the LINK UMI