Search

Solid solutioning of Mg2(Si, Sn) has been a promising approach in reducing thermal conductivity and leads to improvement of thermoelectric performance. In addition to the Mg2(Si, Sn) solid solutions, we have noticed a layered structure with a gradient composition, which is formed by nonequilibrium solidification and peritectic reaction process and can provide further reduction of thermal conductivity of the Mg2(Si, Sn) solid solutions. All layers of the layered structure have the same face-centered cubic-based structure but varying Sn/Si concentration ratios in each layer. The interfaces between the layers are semi-coherent, reticulating with different numbers of misfit dislocations. Such an interfacial structure brings large numbers of phonon-scattering sources, resulting in further reduction of thermal conductivity in the Mg2(Si, Sn) solid solutions. Consequently, the undoped Mg2Si0.75Sn0.25 containing a higher density of the layered structure has relatively lower thermal conductivity, 1.9 W m-1 K-1 at 523 K, than Mg2Si0.25Sn0.75 with a much lower density of the layered structure, 2.3 W m-1 K-1 at 523 K.

Thermoelectric composite alloys consisting of the β-FeSi2 matrix and SiO2 particles dispersion were fabricated by a so-called combined reactions sintering process using reduction and oxidation reactions between eutectoid Si decomposed from α-Fe2Si5 and added Fe-oxide powder. Typical microstructure may include some of residual eutectoid Si particles, intermediate product Fe2SiO4 particles, and/or remaining reduced Fe particles depending on the composite alloy compositions and the process conditions. Partitioning of doping element, n-type Co or p-type Mn, during the process plays an important role to control the optimum carrier concentration of the composite alloys. Thermal conductivity can be reduced, as expected, by the dispersion of SiO2 particles. The solubility of doping elements, Co, Mn, Al, and Ru was evaluated in α-Fe2Si5 at 1373 K and in β-FeSi2 at 1073 K being based on the isotherm determination. It is suggested that suitable dopants for the present process are n-type Co and p-type Mn, since they have sufficiently large solubility around 10 at% in both α-Fe2Si5 and β-FeSi2 phases.

Retained austenite plays a significant role in the strength-ductility balance of medium Mn steels. In this contribution, a new processing route (Flash-Austenite Reversion Treatment, Flash-ART) was proposed to design retained austenite with a compositional core-shell structure in an Fe-0.20C-7.76Mn-1.99Al (wt%) steel. It was interestingly found that the austenite core, which should transform into martensite based on its composition and size, was stabilized by the Mn enriched shell. Flash-ART allows us to obtain more retained austenite compared with the conventional ART.

The microstructure and thermoelectric properties of an off-stoichiometric quaternary (Ti 0.2 ,Zr 0.8 )Ni 1.1 Sn half-Heusler (HH) alloy was investigated in three heating cycles. A high density of coherent nanoprecipitates with an average diameter of ∼13 nm and an interprecipitate spacing of ∼6 nm was observed in the alloy. Formation of the extremely fine nanoprecipitates, most likely to be the ‘full’ Heusler (FH) phase, was not only strongly related to the degree of excess Ni concentration ratio in the alloy, but also appeared to be affected by the Ti–Zr substitutions. We noticed the behaviour of both electrical resistivity ( ρ ) and Seebeck coefficient ( S ) of the alloy was closely associated with the microstructure evolution of the FH-nanoprecipitates, which depended on their phase instability at elevated temperature and the cyclic heating process. The ρ and S reduced after the 1st heating cycle and stabilised thereafter in the subsequent heating cycles. Despite of the presence of the metallic FH-nanoprecipitates, the stabilised S maintained similar magnitudes to S of the ZrNiSn (without FH-nanoprecipitates) and did not show degradation of S as previously seen in the ZrNi 1.1 Sn containing relatively much larger FH-nanoprecipitates. The high density of FH-nanoprecipitates and the presence of Ti–Zr point defects were responsible for the significant reduction of thermal conductivity ( κ ) of the alloy, about 30% and 20% less than κ of the ZrNiSn and ZrNi 1.1 Sn alloys, respectively. Moreover, further reduction of κ was noticed due to formation of the diffuse HH/FH interfaced FH-nanoprecipitates from the cyclic heating process. Consequently, the alloy has shown a maximum dimensionless figure of merit ( ZT ) up to 0.81 at 870 K.