Several novel types of nanostructured materials are emerging as potentially suitable for the development of high ZT thermoelectrics. Amongst them, Nanoparticle Embedded in Alloy Thermoelectric (NEAT) materials, offer a good opportunity for theoretical computations to predict thermoelectric properties and compare with experimental results. In this talk I will first discuss the theoretical approaches to compute the thermal conductivity of NEAT materials, including ab-initio methods. I will present concrete results for SiGe [1], Mg₂(SiSn), and (BiSb)₂Te₃ [2] based composites, discussing various expected limits. I will also discuss the case of other related systems, such as 2D single layered hybrid BN [3], or SiGe alloy nanowires. Finally, the thermoelectric properties of sintered nano-grained composites will be discussed within a simplified model, and a high throughput computational study of the power factor for 3000 sintered compounds will be presented [4]. [1] A. Kundu, N. Mingo, D. A. Broido, and D. A. Stewart, The role of lighter and heavier embedded nanoparticles on the thermal conductivity of SiGe alloys, Phys. Rev. B, 84, 125426 (2011). [2] N. Ayape-Katcho and N. Mingo, Lattice thermal conductivity of (Bi_xSb_1â^'x)₂Te₃ alloys with embedded nanoparticles, in preparation (2011). [3] H. Sevinçli, W. Li, N. Mingo, G. Cuniberti, and S. Roche, Effects of domains in phonon conduction through hybrid boron nitride and graphene sheets, submitted (2011). [4] S. Wang, Z. Wang, W. Setyawan, N. Mingo, and S. Curtarolo, Assessing the thermoelectric properties of sintered compounds via high-throughput ab initio calculations, submitted (2011).

Sculpting nanoscale building blocks with novel properties and retaining the nanostructuring-induced properties in larger scale assemblies are key to obtaining bulk nanomaterials with properties otherwise not attainable in non-nanostructured bulk materials. We have demonstrated a new class of both p- and n-type bulk nanomaterials with room-temperature thermoelectric figures of merit ZT as high as 1.1 from assemblies of pnictogen chalcogenide nanoplates obtained by a scalable (10 g/min) bottom-up approach. A unique aspect of our work is that our binary nanobulk thermoelectrics exhibit ZT~1 by a combination of doping and nanostructuring, without any alloying additions, resulting in multi-fold ZT improvements over non-nanostructured non-doped counterparts. For example, antimony telluride is a low figure-of-merit (ZT<~0.3) thermoelectric because of a low Seebeck coefficient α arising from degenerate carrier concentrations ~> 10²⁰ cm^-3 generated by antisite defects. Here, we report 10-25% higher α in sub-atomic percent sulfur doped nanocrystalline antimony telluride having up to hundred-fold lower carrier concentrations due to the suppression of antimony antisite defects. High resolution X-ray photoelectron spectroscopy (XPS) using synchrotron radiation reveals lattice site-occupancy of the sulfur dopant atoms, and combined with Hall and Nernst coefficient measurements and a defect model for chalcogenides allows correlating the sulfur doping with observed electronic structure and defect chemistry changes. We show the presence of sulfur raises the lattice bond-polarity and increases the activation energy for formation of antisite defects, strongly suppressing their concentration. XPS of valence bands of the nanobulk antimony telluride reveal sulfur-induced density of states modification. The α increase, together with the nanostructuring-induced thermal conductivity decrease, in our binary antimony telluride yields a factor-of-three higher ZT~1 at 423 K, thus also shifting the peak ZT to higher temperatures, outperforming state-of-art alloys. We conclude by presenting high-resolution XPS for p-type antimony telluride based alloys and n-type bismuth telluride providing insights for raising the power factors of pnictogen chalcogenides over bulk ceilings. Tuning the electron-crystal phonon-glass behavior by adapting our bottom-up approach for designing nanoscale building blocks should enable nanobulk thermoelectrics with further increases in ZT for transforming thermoelectric refrigeration and power harvesting technologies.

Incorporating nano-sized inclusions in bulk thermoelectric materials is expected to lower the lattice thermal conductivity by enhancing phonon scattering, thus improving their thermoelectric figure of merit ZT. We herein report a wet chemical process to fabricate BixSb2-xTe3 (x=0.4 or 0.5) nanocomposite with metal nanoparticles by using metal acetate precursors with ethyl acetate solvent as a medium for homogeneous incorporation, which is possible to apply various kinds of metal nanoparticles. Nano-scale metal particles were homogeneously dispersed in the Bi-Sb-Te alloy matrix after reduction and spark plasma sintering. The lattice thermal conductivities were reduced by enhanced long-wavelength phonon scattering by metal nanoparticles regardless of kinds of metal, while the Seebeck coefficients were enhanced for a few selected metals of work function of 3.5 to 5.5 eV, including palladium, cobalt, manganese, and silicon, by possible carrier energy filtering effect. Consequently, the ZTs of aforementioned metals-embedded nanocomposites were enhanced over 1.4 at 320 K.

Nanocomposite Bi2Te3 based alloys are attractive for their potentially high thermoelectric figure-of-merit (ZT) around room temperature. The nano-scale structural features, grains or precipitates etc., embedded in the matrix provide more scattering of phonons and can thus reduce the lattice thermal conductivity. To further take advantage of such nanocomposite structures, we focus on the development of nanocrystalline Bi(Sb)Te(Se) powders by high energy cryogenic mechanical alloying followed by optimized hot-pressing process. This approach is shown to successfully produce Bi(Sb)Te(Se) alloy powders with grain size averaging about 9 nm for n-type BiTeSe and about 16 nm for p-type BiSbTe respectively. This cryogenic process offers much less milling time and prevents thermally activated contaminations or imperfections from being introduced during the milling process. The nanocrystalline powders are then compacted at optimized high pressures and temperatures to achieve full density compactions and preserve the grain sizes effectively. The resulting nano bulk materials have low thermal conductivity, good electrical conductivity, optimal Seebeck coefficients and consequently improved ZT. Thermoelectric properties and microstructure studies by X-ray diffraction and transmission electron microscopy will also be presented and discussed.

This talk will describe a completely new class of doped-nanothermelectrics obtained by the assembly of surfactant-sculpted nanocrystals synthesized by a scalable microwave-solvothermal approach, where the surfactant serves as a doping agent. Sintered assemblies of the nanostructures exhibit up to 25-250% higher figure of merit (ZT) than their non-nanostructured counterparts for a variety of thermoelectric materials, attractive for transformative solid-state refrigeration and waste-heat harvesting technologies. The surfactant-mediated microwave-stimulated bottom-up synthesis approach can scalably sculpt large quantities (>10g/minute) of nanocrystals with controllable shapes and sizes that can be assembled into bulk samples to obtain both high power factors α²Ïf as well as unprecedentedly low thermal conductivity k. I will demonstrate that our method can be used to obtain both n- and p-typed pnictogen chalcogenide nanostructures that can be sintered to obtain ZT>1. While nanostructuring diminishes the lattice thermal conductivity k_L to ultralow values of 0.2-0.5 W/mK, sub-atomic-percent sulfur doping and heterostructuring of the nanoscale building blocks results in large Seebeck coefficients between -240 < α < 298 μV/K and high Ïf between 0.2-2.5 Ã- 105 Ω^-1m^-1. Photoemission spectroscopy combined with carrier concentration and mobility measurements using the Hall effect show that the high power factors are due to alterations in the density of states near the Fermi level. These effects can be accentuated by creating nanocomposites by mixing nanoplates of different materials (e.g., S-doped Sb₂Te₃ and Bi₂Te₃), that lead to factorially higher ZTs than the pure material at specific volume fractions, pointing to non-linear effects on the power factor and nanostructuring. We will finally illustrate the extendability of our approach of nanostructuring and doping in Al- In- and Bi-doped ZnO, where we obtain >50% ZT enhancement through factorial decreases in k_L , while retaining bulk-like power factor.

High-quality Bi_0.5Sb_1.5Te₃ nanoparticles were fabricated by spark-erosion with an unprecedentedly high production rate of 135 grams/hour using a relatively small laboratory apparatus. Spark erosion utilizes electric discharges to vaporize a desired material inducing nucleation of nanosized particles. The compacted nanocomposite samples made from these nanoparticles exhibit a well defined, 20 â?" 50 nm size nano-grain microstructure and show an enhanced figure of merit, ZT, of 1.36 at 360 K. This is attributed to a significant reduction in lattice thermal conductivity as a result of increased phonon scattering at grain boundaries. The ZT is comparable to the best values obtained by other nanoparticle techniques such as mechanical pulverization, but spark-erosion is a simpler, higher-production rate, readily-scalable method. Such a technique is essential for providing inexpensive, oxidation-free nanoparticles required for fabricating high performance thermoelectric devices for power generation from waste heat, and for refrigeration.

FeSb2 is a strongly correlated semiconductor that has been shown to have an extraordinarily large thermopower in single crystal samples. Bentien et. al. report a thermopower of 45000 microV/K at 10K. The dimensionless figure of merit(ZT) for single crystal samples is calculated to be approximately 0.005 at 10 K and is constrained by its relatively high thermal conductivity. In order to reduce thermal conductivity values, we studied thermoelectric properties of a series of FeSb2 nanocomposites (NC). Nanocomposites tend to decrease thermal conductivity by introducing phonon mismatches between crystal grains. We find the NC samples minimum value of Seebeck coefficient is reduced to about -200 to -475 microV/K when compared to single crystal, but this is combined with a dramatic decrease in thermal conductivity through nanocomposite technology. Our results reflect an improvement of about 160% increase in ZT over that obtained by single crystal methods. Further improvements in ZT were obtained by changes in stoichiometry and will be discussed.

In the 1960s, silicon was investigated as a potential thermoelectric material due to its excellent electronic properties, however it was determined that the thermal conductivity was over an order of magnitude too high for practical applications. As result the ZT remained about 0.1 (at 1275 K) for 50 years. Bulk nanostructuring of silicon has been proven to be an effective method of increasing the thermoelectric figure merit, leading to increases of up to 250% over single crystal Si (at 1275 K). This large increase is due in part to over an order of magnitude reduction in the lattice thermal conductivity while maintaining relatively high carrier mobility, thus making nanostructured Si a contender for high temperature thermoelectric applications. While significant gains have been, further improvements in the thermal-to-electric conversion efficiency are necessary for large scale applications such as waste heat recovery. Modeling suggests that the figure of merit can be further enhanced via a nanocomposite approach through the use of small (3-5 nm) inclusions. The inclusions are predicted to not only further reduce the lattice thermal conductivity but also enhance the Seebeck coefficient via carrier injection. In this presentation, an overview of the nanostructured Si work that has been done at Jet Propulsion Laboratory will be presented and discussed along with new strategies to improve the thermoelectric figure of merit.

Due to either a remarkable coincidence or an interesting fundamental connection that has not yet been established, small band gap bulk semiconductors based on Bismuth and Antimony act as both very good low temperature thermoelectric and as hosts for exotic electronic surface states. In fact it was this possible connection that led to the discovery of what has been called the â?ohydrogen atomâ? of topological insulators, Bi2Se3. In this talk I will describe some of the members of this materials family, which range from simple Bi,Sb alloys to much more complex intergrowth phases, from a structural and basic electronics perspective. Control of the carrier concentrations in the bulk phases is critical for both their thermoelectric properties and their study as topological insulators. Due to the small defect formation energies in these compound semiconductors based on constituents with similar electronegativities, this has posed an interesting materials challenge, which we have made some progress in addressing. This work will also be described.

The purpose of this talk is to discuss thermoelectric device physics from a Landauer perspective. The Landauer approach is a well-established and widely used technique in mesoscopic physics and nanoelectronics. Less well known is the fact that it can also be used to treat bulk thermoelectric problems. In its rigorous form, the Non-equilibrium Greenâ?Ts Function (NEGF) approach, it provides a detailed description of dissipative quantum transport. At a simpler level, it is mathematically equivalent to traditional, Boltzmann equation approaches. We find, however, that it provides a different perspective, and the insights obtained from viewing an old problem in a new light may prove useful in the quest to increase thermoelectric performance. Moreover, it provides a connection to the broader field of nanoelectronic devices and to the wide range of already-developed tools that may prove useful in thermoelectric research and development. In this talk, we will discuss what makes a good thermoelectric from a Landauer perspective. The basic viewpoint and its connection to traditional approaches will first be reviewed. Next, we will examine the question from a Landauer perspective of what bandstructure is best for thermoelectric performances. We will then show how to apply the same techniques to the problem of lattice thermal conductivity. In addition to a new viewpoint on thermoelectrics, the Landauer approach also provides a way to connect first-principles bandstructures and phonon dispersions with phenomenological models that are suitable for device analysis and design. Use of the approach in this role will also be discussed. Using NEGF simulations, a rigorous form of the Landauer approach, we will show that careful optimization of coupled electro-thermal transport may open up new possibilities. Finally, a Landauer view of the thermoelectric figure of merit will be presented. The overall goal of the presentation is to summarize the Landauer approach and to illustrate its use in thermoelectric research and design.

We consider phononic heat transport between two heat baths having different temperatures and mediated by a quantum particle using the generalized quantum Langevin equation. We derive expressions for the steady state heat flux and thermal conductance that depend on a small number of measurable parameters, such as the relaxation times of the oscillator's momentum and coordinate, the coupling constant, and the Debye cutoff frequency, using the Drude-Ullersma model. These expressions are analyzed at different temperatures regimes and for different strengths of the coupling parameter. Despite the fact that a molecular chain with strong coupling between its quantum systems violates Fourier's law, the law can be restored if one considers a chain of alternating nanoparticles and quantum systems. Connection to other models describing the heat transfer and to experimental observations is discussed.

Bismuth-Antimony alloys have been shown to have high ZT values below room temperature, especially for single crystals. For polycrystalline samples, impurity doping and magnetic field have proven to be powerful tools in the search for understanding and improving thermoelectric performance. Nanopolycrystalline Bi_.88Sb_.12 doped with .05, .5 and 3% Ce were prepared by ball milling and dc hot pressing technique. Electrical resistivity, Seebeck coefficient, thermal conductivity, carrier concentration, mobility, and magnetization are measured in a temperature range of 5-350 K and in magnetic fields up to 9 Tesla. The effects of Ce doping on the thermoelectric properties of Bi_.88Sb_.12 will be discussed.

Resonant doping impurities are known[1] to increase the thermoelectric power S of semiconductors at a given carrier concentration n above the value that the host semiconductor doped to the same value of n would ordinarily have. The known resonant impurities that do this are Tl in PbTe, In in SnTe and Sn in Bi2Te3. Here, we report for the first time that potassium has the same property in Bi1-xSbx alloys used for cryogenic cooling. Experimental results will be reported as function of temperature, x, and K concentration. For specific concentrations, the thermoelectric figure of merit is much increased, and reaches usable values from 50 K to near room temperature. Theoretical band structure calculations support the result, and identify the physical origin of the resonance. Work supported by AFOSR MURI â?oCryogenic Peltier Coolingâ? Contract #FA9550-10-1-0533 [1] J. P. Heremans, B. Wiendlocha and A. M. Chamoire, Resonant levels in bulk thermoelectric semiconductors, Energy Environ. Sci., DOI:10.1039/C1EE02612G (2011)

The first systematic measurements of the transverse Nernst-Ettingshausen Coefficient N of the elemental rare-earth metals are presented from 80 to 420 K. For those rare-earths that have hexagonal symmetry, measurements are given with the heat flux along the [100] and the [001] axes, for the cubic ones [100]. It is somewhat surprising that such data are not available in the literature, as the Nernst-Ettingshausen coefficients of these elemental solids are expected to be large, because they have a complex band structure and Fermi surface, a small thermopower and a multicarrier systems involving electron (e) and hole (h) pockets. Indeed, for such systems, the Seebeck coefficient (S) can be expressed as¹ S=(S_eÏf_e+ S_hÏf_h)/(Ïf_e+Ïf_h) while N=[(N_eÏf_e+ N_hÏf_h)(Ïf_e+Ïf_h)+(S_h-S_e)(R_HhÏf_h- R_HeÏf_e)Ïf_eÏf_h]/( Ïf_e+Ïf_h)², where Ïf is the electrical conductivity and R_H the Hall coefficient. Here, each of the carriers have partial Ïf, R_H, S and N coefficients denoted by the corresponding subscript e or h. of either carriers : electrons (e) and holes (h). Since S_h>0 and S_e<0, the resulting thermopower S should be low but the S_h â?" S_e term in N becomes large. This might make rare-earth metals useful in transverse Nernst-Ettingshausen cooler, especially at cryogenic temperatures. Peltier coolers are the usual thermoelectric heat pumps, but to reach large temperature differences Peltier coolers need to be cascaded. Using this thermomagnetic effect where heat and current flow are perpendicular to each other, a single material with a particular design, can be used to create a very large temperature difference, cancelling the problem of contact resistances in cascaded Peltier cooler modules.² The dimensionless figure of merit of Nernst-Ettingshausen coolers is zT_N=B²N²Ïf(B)T/κ(B), with B the magnetic field, T the absolute temperature and κ the thermal conductivity. Work supported by DOE-BES Energy Frontier Research Center on Revolutionary Materials for Solid State Energy Conversion, 61-3212B 1. E.H. Putley, The Hall Effect and Semiconductor Physics (New York: Dover, 1968). T) 2. H. J. Goldsmid Thermoelectricity, Springer Verlag, Berlin, 2010

Thermionic emission has long been an attractive idea for thermal energy conversion for its potentially high-efficiency conversion with no moving parts and simple device geometry. However, previous generations of convertors have had relatively poor actual performance due space charge effects and low output voltages, which has severely restricted their implementation and research. Here we discuss why modern microfabrication techniques and a new idea that combines both photovoltaic and thermionic effects together may reverse the fortunes of this technology, and lead to highly efficient conversion devices. We show that both theoretical and experimental data on MEMs thermionic devices show higher maximum currents than previously possible. Solar thermal applications also have a considerable quantity of light present, and we describe a new mechanism, photon-enhanced thermionic emission (PETE), that can benefit from both photon illumination and thermal energy. This process enables thermionic emission at much lower temperatures (~400C) than conventional thermionic emission, and may have higher overall efficiency. We will discuss experimental demonstrations of this process and the critical barriers to making lab prototypes into realistic conversion systems.

We investigate thermal nanostructures that allow the Surface Phonon Polariton guiding. The glass tubes were designed to guide monochromatic thermal energy to improve heat conduction or serve as radiative source. Surface Phonon Polaritons (SPPs) are surface waves produced by the coupling of atomic vibrations with the electromagnetic field. Because they are monochromatic waves in the Infra-Red, SPP might be able to provide monochromatic sources just based on simple heating. At ambient temperatures, they also might produce abnormal heat conduction phenomena related to ballistic regimes of transport. Considering heat transfer in the propagative direction of SPPs, quite recent works have tried to show heat conduction improvement in nanofilms. However, the coupling is not favoured in this configuration because the film is dissipating the electromagnetic energy and the coupling distance is too small. We aim at designing glass microtubes in order to measure the SPP heat flux and its abnormal behaviour. The modelling of the tube consists in solving the Maxwell equations in the approximation of small wall thickness and small outer diameter compared to the polariton wavelengths. A simplified polynomial solution was obtained and solved numerically to provide the free path spectrum and the SPP heat flux. The experimental set-up for proving abnormal heat conduction consists in sticking a glass capillary sample on an AC heating Pt wire on one end and measuring the temperature on the other end by using a deposited thermocouple. The phase between heating and temperature signals directly provides thermal diffusivity. The first theoretical results reveal a very strong enhancement of the Free Path appearing in a narrow range of tube radius. Experimental works have proven that the signal to noise ratio was satisfying and the spectrum of the temperature response was obtained. The temperature profile by using IR microscopy should yield the proof of the SPP contribution in the heat transfer along the tube.

We theoretically investigate the use of energetically-sharp resonances of core-shell nanoparticles embedded in semiconductors to selectively scatter carriers and thereby enhance the thermoelectric power factor and figure of merit. Appropriate selection of materials for the core-shell band structure can lead to the formation of quasi-bound states inside the nanoparticles, which strongly scatter carriers near these energy levels, making sharp features in the energy-dependent electron relaxation time. Particularly at low temperatures, these features can significantly enhance the semiconductorâ?Ts Seebeck coefficient and the power factor. We find that the power factor of PbTe at 80 K is enhanced by more than 80 % when single-sized core-shell nanoparticles of 3 nm core diameter and 1.5 nm shell width are introduced with density 1 Ã- 10¹⁸ cm^-3. The effect of nanoparticle size distribution and operating temperature are also quantified.

Lead Telluride is one of the best binary thermoelectric materials at slightly elevated temperatures. Nanostructuring has been shown to increase its large thermoelectric figure of merit even further mainly due to a decrease in lattice thermal conductivity. However, with approaching the alloy limit in state-of-the-art nanostructured PbTe-based materials, new concepts need to be exploited in order to achieve further improvements in zT. Here, we present a fully colloidal approach to nanostructured PbTe-based materials together with suitable surface treatment and spark plasma sintering conditions to allow for conductive nanostructured macroscopic samples. We introduce core-shell nanoparticles as a structural unit cell with great potential to manipulate the Seebeck coefficient in combination with effective phonon scattering. As an example, the full characterization of PbTe-PbSe core-shell heteronanostructures are presented and the results discussed in the light of a possible energy filtering effect vs. grain boundary scattering. Strategies to improve electric coupling between these structures are discussed.

Recently, solid state refrigeration based on the electrocaloric effect, EC, has attracted a great deal of attention due to the easy generation of large electric fields, high efficiency and relatively low cost of working media. It has been showed theoretically that the electrocaloric effect originates from the entropy changes in the material upon the application of electric field [1] and thus can be related to the degree of structural disorder in the material. In this work the effect of introduced cation disorder on the A-site positions in titanate perovskites on the ferroelectric properties and ultimately the electrocaloric response was evaluated. The degree of cation disorder in the perovskite structure can be expressed by A-site cation radii variance, Ïf², as Ïf²= â^'y_iR_i²-A<², where R_A - average ionic radius of A-site, R_i and y_i â?" ionic radii and occupancy of A-site of element. It has been shown that the A-site cation variance significantly influences electronic and magnetic properties in perovskites and perovskite related compounds [2] A series of perovskites with the general formula Ba_1-x-ySr_xCa_yTiO₃ (0â?¤xâ?¤0.35, 0â?¤yâ?¤0.22) were prepared by the solid state reaction. The average ionic radius of A-site cations was maintained constant at 1.5505 Ã., whereas the A-site cation variance was varied from 0.0066 Ã.² in Ba_0.65Sr_0.35TiO₃ to 0.0125 Ã.² in Ba_0.78Ca_0.22TiO₃. All studied specimens crystallizise in the tetragonal perovskite structure. Curie temperature, Tc, increased lineally with the cation variance from 22 °C for Ba_0.65Sr_0.35TiO₃ to 130 °C for Ba_0.78Ca_0.22TiO₃. The studied specimens showed typical ferroelectric behaviour. The remnant polarisation measured at 30 °C increased with the cation variance. Electrocaloric response of the studied ceramics was measured both directly by the modified DSC technique and indirectly from the P-E loops. A very good agreement between the values of EC heat measured by both techniques was obtained. The magnitude of the EC response was measured as a function of temperature and applied electric field. The maximum EC effect was observed close to the Tc. The maximum of the EC effect increased almost linearly with the variance. The maximum EC response of 1.2 °C was observed at 122 °C and 20 kV/cm in Ba_0.78Ca_0.22TiO₃. The model to describe the EC behaviour of (Ba,Sr,Ca)TiO₃ perovskites was proposed. The studied alkaline earth titanates are very promising EC materials which are environmentally friendly alternatives to current lead containing EC compounds (e.g. PMN-PT [3]). [1] L. J. Dunne et al., Appl. Phys Lett., 2008, 93, 122906. [2] L.M. Rodriguez-Martinez et al., Phys. Rev. B, 1996, 54, R15622. [3] G. Sebald et al., J. Appl. Phys., 2006, 12, 124112.