No abstract

The efficiency of thermoelectric devices is a function of the non-dimensional figure of merit ZT [1] of the material, whereT is the absolute temperature and Z = ra 2 /j; a = thermoelectric power or the Seebeck coefficient, r = electrical conductivity, j = thermal conductivity; the numerator of Z is referred to as the power factor. State-of-the-art thermoelectric materials usable for cooling and power generation have a ZT ∼ 1, whereas a ZT ∼ 4 is necessary to surpass competing technologies.[2] Factorial enhancements in ZT may be obtained by introducing nanoscopic geometrical confinement in one, two, or three dimensions, for example, nanolayered quantum-well superlattices, nanowires, and quantum-dot superlattices, respectively. [3][4][5][6] Increased scattering of heat carriers at interfaces and nanostructure boundaries are thought to decrease j, while changes in charge carrier density near the Fermi level offers possibilities for increasing r and a. [3,6,[7][8][9][10][11] Decreases in j have been observed across Bi 2 Te 3 /Sb 2 Te 3 nanolayer superlattices, [3] giving rise to approximately twofold increase of ZT (highest reported ZT ∼ 2.3). Even larger ZT increases (approximately fivefold compared with the bulk value) due mainly to the reduction of j have been observed in PbSeTe/ PbTe quantum-dot superlattices.[6] However, the magnitude of ZT is only ca. 1.6 since PbSeTe/PbTe bulk has a lower ZT than Bi 2 Te 3 /Sb 2 Te 3 . Lowering the thermal conductivity of materials with an inherently high ZT in the bulk form (e.g., bismuth telluride), by introducing 3D confinement, that is, producing nanoparticles, is expected to yield higher ZT values. This approach could also allow the tuning of quantum effects by tailoring the nanoparticle size and shape, to facilitate additional mechanisms for increasing ZT. Indeed, power-factor increases have been predicted to arise from such effects in wellorganized Si/Ge nanoparticle arrays.[11] Hence, synthesis and assembly of nanoparticles of bismuth telluride is of interest to enable the development of higher efficiency devices for thermoelectric cooling and power generation. Bismuth telluride nanoparticles with diameters of ca. 100 nm have been synthesized by several routes. Examples include melting the constituent elements in sealed vessels above 600°C; [12] low-temperature (ca. 30°C) precipitation by reacting tris(dimethylamine) bismuthine and bis(trimethylsilyl) telluride in hexane, and subsequent annealing at 160°C; [13] co-precipitation of bismuth and tellurium oxides in water followed by hydrogen reduction; [14] or through reduction of organometallic complexes. [15] Smaller (e.g., 20-40 nm) nanostructures can be realized by solvothermal techniques [16,17] using precursors in solvents such as N,N-dimethylformamide or water at 100-180°C in reducing ambients. The high temperatures or long reaction times in these reactions, however, are not conducive for obtaining nanostructures smaller than 20 nm, which is the size regime where quantumconfinement effects are expected to be signif...

Branched core/shell bismuth telluride/bismuth sulfide nanorod heterostructures are prepared by using a biomimetic surfactant, L-glutathionic acid. Trigonal nanocrystals of bismuth telluride are encapsulated by nanoscopic shells of orthorhombic bismuth sulfide. Crystallographic twinning causes shell branching. Such heteronanostructures are attractive for thermoelectric power generation and cooling applications.

Chalcogenides such as bismuth telluride and lead telluride exhibit high thermoelectric figures of merit ZT 1-3 in the bulk because of a combination of high electrical conductivity σ and Seebeck coefficient S, and low thermal conductivity κ, making them attractive for refrigeration and generating electrical power from heat. 4-6 Nanostructuring these materials along one dimension (e.g., nanolayers) has been shown to increase ZT because of size and quantum effects on S, σ, and κ. 7,8 Greater degrees of confinement (e.g., PbTe nanowires with characteristic dimensions <5 nm) 9 are expected to provide higher ZT increases and open up entirely new types of possibilities for novel applications such as cooling nanoscale hotspots in nanodevice interconnections comprising nanowires or nanotubes and creating nanoscale power generators.Growing one-dimensional nanocrystals of chalcogenides with a cubic crystal structure (e.g., PbTe, rocksalt structure) is a challenge because of the necessity of a symmetrybreaking pathway for enabling anisotropic growth. 10 Solvothermal decomposition or polyol reduction yield spherical 11 or cube-shaped 12,13 nanoparticles. Incorporating surfactants provides particle size control, but is ineffective in inducing shape anisotropy because of the high crystallographic degeneracy in cubic crystals. Alternative methods based on oriented attachment of nanoparticles in solution, 14-16 catalystmediated vapor-liquid-solid reactions, 17,18 and inorganic templating have been devised. 19,20 However, nanowire polycrystallinity and template removal issues limit the utility of these routes for fundamental studies and for reaping the maximal benefit of the properties of single crystals for applications. Recently devised sonochemical and autoclavebased techniques, 21-23 e.g., <10 nm diameter nanorods 12 with aspect ratios j7, and formation of single-crystal nanowires through nanoparticle coalescence and ripening, however, are promising approaches. 31 Here, we report a completely new two-step organicinorganic templating method to obtain single-crystal PbTe nanorods with aspect ratios up to 27, at moderate temperatures, without using an autoclave. We exploit anisotropic nanotube growth through organic surfactant templating and transform the nanotubes into single-crystal PbTe nanorods via reaction with a lead salt obviating any subsequent template removal. Nanorod to nanotube conversion has been reported, 25 but the reverse process reported here is new and can be applied to producing anisotropic nanostructures of materials with a cubic structure.In the first step, Te nanotubes were synthesized by solvothermal reduction of TeO 2 in the presence of cetyl ether,