Metal oxides for thermoelectrics
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
The goal of this work is to investigate the synthesis, the stability, and the thermoelectric properties of some binary and ternary early transition metal oxides with adaptive structures. The biggest advantages of metal oxides are their extremely low costs paired with low or no toxicity, high abundance, and long term stability. On the other hand, their break-through as thermoelectric materials has been limited by their low figure of merit (zT = S2T/ρκ) caused by the low mobility of the charge carriers and a generally high thermal conductivity. Nevertheless, the research on the thermoelectric oxides is still a relatively new field, and therefore there is much space for improvement. In this scenario, metal oxide with adaptive structures could overcome most of the limitation of the metal oxides. The combination of intrinsic defects and large unit cells along with the possibility of manipulating the electronic transport properties independently from the thermal properties, make them promising candidates for high temperature thermoelectric applications.
The focus of this thesis lies on the study of metal oxides systems with adaptive structures. In detail, the thermoelectric optimization of the tungsten Magnéli phase WO2.90, the thermoelectric properties and the stability of the tetragonal tungsten bronze series Nb8-xW9+xO47, and the synthesis of reduced molybdenum oxides have been thoroughly investigated.
In the literature for the tungsten Magnéli phase WO2.90 promising thermoelectric properties are reported. It shows a relatively low thermal conductivity (≈ 3.5 Wm-1K-1) due to the presence of crystallographic shear (CS) planes, coupled with a good electrical conductivity. The main reason of the low figure of merit of this material is caused by the high charge carrier concentration. Therefore, the optimization strategy relies on the reduction of the charge carrier density. Since this compound is an n-type conductor, the decrease of the electron count was achieved by increasing the oxygen content up to a nominal composition WO2.91. This manipulation was beneficial to the thermoelectric properties of the compound, resulting in an improvement of the figure of merit of about 30% compared to the reference WO2.90 compound. The highest measured value of zT was 0.24 at 1200 K.The n-type tetragonal tungsten bronze (TTB) series Nb8-xW9+xO47 shows intrinsic crystal defects which are similar to those observed in the Magnéli phases. This nanostructure containing crystallographic shear (CS) planes leads to low thermal conductivity values. The change of the cation composition and the concomitant change of the charge carrier concentration, does not substantially affect the crystal structure, and therefore the thermal transport properties for all the members of the series. Therefore, the substitution of niobium with tungsten allows a decoupled variation of the thermoelectric properties. To study the potential of these compounds, samples with low substitution degree x = 0, 0.075, 0.1, 1 and 2 were synthesized and structurally characterized. The thermal stability and the oxygen deficiency were studied together with the thermoelectric transport properties. Both cation substitution and the oxygen deficiency lead to an enhancement of the transport properties.
The second part of the investigation focused on the highly substituted compounds of the series: Nb5W12O47 (x = 3) and Nb4W13O47 (x = 4). These two compositions have been synthesized as pure polycrystalline samples. Subsequently, their chemical and physical stability when subjected to thermal cycling was studied. A complete thermoelectric characterization of the substituted TTB was carried out. A second set of measurements was done on the devices available at the University of Oslo (FERMIO), to verify the reproducibility of the obtained results. The X-ray analysis of powder patterns at different stages of the investigation, and the agreement of the cycled measurements results confirmed the high stability of the samples.
Finally, another transition metal oxide from the Magnéli-type phases was investigated. Some components of the molybdenum oxide family MoO3-x were synthesized by solid state reactions. After optimization of the synthesis conditions, it was possible to obtain polycrystalline powders of 𝛾-Mo4O11, Mo17O47, and Mo18O52 whose purity was confirmed by PXRD measurements. Spark plasma sintering (SPS) was used to consolidate dense pellets, but all attempts in SPS led to the formation of side phases. Therefore, a thorough study of the sintering conditions has to be performed to obtain pure phase pellets that can be used for the thermoelectric characterization.