Iron disilicide based thermoelectric materials Fe0.92Mn0.08Six (1.9≤x≤2.5) were prepared by rapid solidification (melt spinning) and hot uniaxial pressing at 1248K with 50MPa for 30min, followed by annealing at 1073K for 20h. X-ray diffraction and scanning electron microscopy showed excess silicon phase for samples with x≥2.1, and both the configurations and the amounts of secondary silicon particles varied with an increase in x. Hall measurements carried out at room temperature showed that the carrier concentrations for Fe0.92Mn0.08Six (1.9≤x≤2.5) were between 2.6×10 18 and 5.6×10 18cm-3. The Seebeck coefficient, electrical conductivity and thermal conductivity were measured from room temperature to 973K. It was found that a little excess silicon in the sample, x=2.1, enhanced the Seebeck coefficient weakly, but was effective for decreasing the thermal conductivity. A maximum figure of merit, ZT=0.17, was obtained for Fe0.92Mn0.08Si2.0 at 973K.

FeSi2-and Fe2Si5-based thermoelectric alloys have been fabricated by melt spinning and levitation melting. It was found that the levitation-melted FeSi2-based alloy must be annealed at 800℃ for 10h to complete transformation of the β phase, while an anneal for only 6h was needed for the melt-spun alloy. On the other hand, annealing the levitation-melted Fe2Si5-based alloy for 4h was enough to complete β-phase formation, whereas 14h was required for the melt-spun alloy. Annealing temperature dependence of the Seebeck coefficient showed that the maximum rate of transformation to β phase occurred at about 800℃ for all samples. Application of the Johnson-Mehl-Avrami equation revealed that grain-boundary nucleation was predominant in the levitation-melted FeSi2-based alloy (time exponent n=1.1), while the zero nucleation mechanism was operative in the melt-spun alloy (n=3.1). For the eutectoid reactions in the Fe2Si5-based alloys, several kinds of nucleation site were active. However, nuclei formed at grain edges were dominant in the melt-spun alloy since n=1.6.

The skutterudite-type CoSb3 alloy was prepared using levitation melting. The electrochemical properties of CoSb3 were investigated using Li ion model cells, Li/LiPF6(EC/DMC)/CoSb3. It was found that the first reversible capacity is about 420mA h g-1. Ex-situ XRD (X-ray diffraction) experiments have been used to characterize the electrochemical lithiation and de-lithiation mechanism of CoSb3 alloy. The results revealed that when CoSb3 is intercalated by Li ions, it decomposes into Co and Sb first, followed by the formation of the Li3Sb phase, which is dispersed in the Co matrix. When Li ions are removed from Li3Sb, Co and Sb atoms cannot reconstruct CoSb3, while the Li-Sb alloying and de-alloying processes are reversible.

Some antimony or tellurium based thermoelectric alloys have been prepared and studied as new candidate anode materials for lithium-ion batteries. It was found that some thermoelectric antimonides yield a volume capacity for reversible lithium storage more than twice that of state of the art carbon based materials. Ex-situ XRD analyses show that the semimetals such as antimony, bismuth and tellurium in the semiconducting thermoelectric alloys are the active elements for the lithium adsorption. However, inactive elements are also necessary for an alloy electrode to ensure the electrochemical capacity retention during cycling.

Antimony zinc alloy, Zn4Sb3, was studied as a potential material for negative electrodes of lithium-ion batteries. It was found that the reversible capacity of ball-milled Zn4Sb3 in the first cycle reached 507mA h g-1 and increased up to 580mA h g-1 when the alloy was ball-milled with about 11.8wt.% graphite additives. Ex-situ XRD analyses of the pure Zn4Sb3 electrodes during cycling showed that several lithium-containing compounds such as LiZnSb, Li3Sb and LiZn have been formed successively during the insertion of lithium into Zn4Sb3. It was found that the ball-milled composite, Zn4Sb3/C7, possesses high initial reversible capacity, small voltage hysteresis and good capacity retention, which make this material an interesting potential electrode for lithium-ion batteries.