Nano-engineering of high performance thermoelectrics

In the present result we have carried out zone refinement synthesis and thermoelectric property characterization on Ge based clathrates. Melt/solid distribution coefficients have been determined for a large number of dopant atoms in Ba(8)Ga(16)Ge(30). We have furthermore shown that very accurate control of the composition is the key to controlling the thermoelectric properties. The crystal structures have been investigated with state of the art synchrotron radiation techniques and analysis of single crystal diffraction data. In particular, we have developed a new experimental method to obtain site selective valence information on the nanoinclusions in the system. We have used a different approach than in the typical DAFS experiment. Instead of collecting few reflections at many different wavelengths, we have collected many reflections at relatively few wavelengths around the Sr k-edge. This allows us to obtain the anomalous scattering component directly as a parameter in the crystallographic refinements. This anomalous scattering contains the information on the guest atom valence state. One key result is the finding that guest inclusions in Ge based clathrates are ionic and thus conform with the Zintl view of these solids. This aspect has been debated in the literature. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Electrical transport properties of high purity single crystal Zn(4)Sb(3) materials fabricated for the first time at UAA were evaluated at NEDO, Cardiff. Samples of Zn(4)Sb(3)were zone refined under various conditions at UAA and sent to DLR and NEDO for investigation. Prior to slicing, Seebeck coefficient profiling along the crystal ingot was made in 2mm steps to investigate homogenity of the refined crystal. Seebeck and electrical conductivity were measured (at 300K), together with carrier concentration and Hall mobility. Results indicate an increase in Seebeck coefficient is due to reduction in carrier concentration in the second batch material. Electrical conductivity is also noted to increase in the second batch material. Data is comparable to results published in the literature for this state-of-the-art material confirming that synthesis of this material by UAA is successful. Consolidated material will be manufactured from grinding these crystals and transport properties characterisation used to assess any increase in ZT after doping is implemented. Zn(4)Sb(3) is a state-of-the-art thermoelectric material and further increases in ZT would enable high performance thermoelectric devices using this technology to be developed within Europe. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

First principles DFT methods were applied to model thermoelectric materials, TMs. Use of local basis sets (gaussians) required special calibration, but it also allowed for an atomistic description of the electronic structure and of its change upon TMs structural changes. The effects of chemical doping/substitution was highlighted by projecting the density of states onto the set of atoms which are most responsible of such effects. Application of Baders theory allowed for chemical understanding of doping effects. TMs electronic structures were linked to their electronic transport properties (ETPs) using the semi-classical Boltzmanns transport theory and the approximation of constant relaxation time. Two different approaches were adopted to model the ETPs of chemically-doped TMs. In the first, the frozen bands of the undoped materials are used but with a number of electrons increased by an amount equal to actual doping and Fermi energy levels accordingly recomputed. In the second approach, the band structures of the fully chemical-doped systems are used to account directly for the modifications of bands due to dopants. To find the optimum doping level within either of the two considered approaches, we introduced the electronic figure of merit, Z(e)T=( S(2)[sigma]/[kappa](e) )T, where [kappa](e) is the electronic contribution to the thermal conductivity. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

In the present result we have developed a new route for synthesis of very high purity, single phase Zn(4)Sb(3). This material fills the gap in the intermediate temperature range (100-500 degrees Celsius), where there are no other materials with an adequate figure of merit. Conventional synthesis procedures have problems with phase purity, and therefore with adequate optimisation of thermoelectric performance. The application in the intermediate temperature range offers opportunities for economic recovery of waste or low-grade heat in several applications to increase the overall operation economy. Waste heat recovery can contribute to an increase of the total operation economic effectiveness of some 3-5%. Within vehicles there are several application for the conversion of heat to electric current; e.g. from car exhaust gases and breaks. The recovered electrical energy can thus be recycled. With the increased demand on clean air of European Cities, electrical cars, or hybrid cars containing electrical motors are becoming a reality. Electrical recovery of heat emitted from the breaks can be most efficiently recovered through TE conversion. This technology is also suitable for other electrically driven crafts, such as trains, tramways, etc. General Electric in the USA has tested proto-type models for car breaks. Only in this sector, there are several billions euros worth of products per year. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Two crystals of clathrate Ba(8)Ga(16)Ge(30) have been grown by CTH, cut to discs and sent to DLR for investigation of their thermoelectric properties. Both crystals are of different conduction type, thus n- and p-type legs for a module could be produced from the same material, probably having their optimal performance at the same application temperature, estimated to be near 900°C. The n-type crystal discs have been investigated with respect to the temperature dependence of electrical and thermal conductivity as well as Seebeck coefficient. From these measurements a maximum ZT of 1.1 at 800°C was calculated. The received p-type disc was characterized with respect to its homogeneity and its room temperature Seebeck coefficient by the Seebeck microprobe. Subsequently the crystal material was ground and hot pressed to obtain pellets more suitable for thermoelectric evaluation. Because of unknown optimum pressing conditions, several attempts had to be done to find a proper set of temperature, pressure and duration, with no positive final result, so far. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Nano-sized powders of CoSb(3) Skutterudites with partial substitution of Co by Ni and Fe, respectively, and partial substitution of Sb by Te, delivered by KTH, have been consolidated by hot pressing at DLR to bulk-like samples suitable for functional characterization. The samples have been characterized concerning spatial homogeneity of the Seebeck coefficient. The temperature dependence of the thermoelectric properties has been measured by different methods at DLR. Structure and chemical composition has been analysed by XRD, XRF, and SEM. Optimisation: The material has been functionally optimised by partial substitution of Sb by Te. Tests on these materials have shown an improvement in the TE properties. Most important, these measurements have shown, that nano-technology is indeed capable to produce materials with drastically reduced thermal conductivity. In comparison to known bulk values, the obtained thermal conductivity was reduced by a factor of 4 to 5. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Information of the atomic arrangement in thermoelectric materials is key to understanding their properties and being able to improve these in a rational and scientific manner. Using synchrotron radiation we have studied the oxidation states and local structures in a series of germanium or silicon based clathrate structures, which has aided in developing the knowledge on the structure of these intriguing and promising compounds. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

A method has been developed for pulling large single crystals (> 40mm in length) of germanium containing clathrates using the Czochralski method. The method is highly suitable for industrialisation of the promising clathrate materials. If further developed the method should enable increased control of framework stoichiometry Ga/Ge ratio, which is key factor in affecting the thermoelectric properties of this material. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

NEDO and DLR have developed and improved methods, standard procedures and special apparatuses for characterization of thermoelectric materials and measuring thermoelectric properties. In this project the combined work of NEDO and DLR included standardisation and round robin measurements to prove the reliability of evaluated materials data. Having access to the competence of doing standardized high precision measurements in the thermoelectric field is of great importance for research and industry in Europe. Efforts will be taken to make this competence and facilities available on a European level. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

All rational design of new materials is based on precise information about the microscopic (atomic) structure of the material. This is because the properties of materials are determined by their atomic composition and 3D structure. Based on 3D structural information structure-property relations may be developed, which can lead to improved materials. In the present result we have determined the crystal structure and charge density distribution of Zn(4)Sb(3) based on very accurate synchrotron radiation single crystal and powder diffraction measurements. The results are significantly at variance with previously reported results in the literature, and they provide the foundation for rationalising the outstanding thermoelectric properties of the material. The unique inter-metallic framework with amorphous regions of interstitial Zn makes it imperative to search for similar systems, which potentially could be extremely good thermoelectric materials. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Due to the close co-operation of one of the employees at DLR (Dr. Dieter Platzek) with the industrial partners THERMOGEN AB and CIDETE, Dr. Platzek decided after finishing the project to found his own company and to carry on with the fruitful cooperation with the above mentioned partners. The new company PANCO, Physics Technology Development of Facilities and Consulting, deals with the development of applications in thermoelectrics and with the development and distribution of measurement facilities for measuring physical properties. Furthermore the great experience of Dr. Dieter Platzek in measurement technology will be used for consulting. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Samples of Zn(4)Sb(3) have been zone-refined under different conditions by UAA and investigated by DLR. Quenched samples from UAA have been ground and ball-milled to powder and pressed under various temperatures at 100 MPa. For characterisation of the thermoelectric properties the samples were analysed by different methods. The electrical conductivity sigma, thermal conductivity kappa, and the Seebeck coefficient S were measured in the range from room temperature to 400°C and the figure of merit Z was calculated. The results so far published elsewhere could be largely attained. The homogeneity of the samples was investigated by means of the Seebeck microprobe, which gives a hint on the phase purity. DSC measurements have been performed to investigate the decomposition of Zn(4)Sb(3) during heat treatment resulting in a possibly limited to 350°C application temperature. Considering that the investigated material was still state of the art material. Great potential is in this material when doping. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

All rational design of new materials is based on precise information about the microscopic (atomic) structure of the material. This is because the properties of materials are determined by their atomic composition and 3D structure. Based on 3D structural information structure-property relations may be developed, which can lead to improved materials. In the present result we have determined the crystal structures of a series of nanostructured skutterudite materials synthesised by partner KTH. The crystal structures were determined by Rietveld refinement of neutron and X-ray powder diffraction data. Of particular importance is the determination of the exact positions of dopant Ni atoms in the Ni(x)Co(1-x)Sb(3) skutterudites, which have been determined for the first time using neutron diffraction measurements. Such information cannot be obtained by the conventional X-ray methods commonly used in materials research. The information on Ni sitting is vital for rational design of improved skutterudite materials since theoretical calculations by partner CNR have shown that different types of doping can have very different effects on the thermoelectric properties. Based on very accurate synchrotron radiation X-ray powder diffraction measurements obtained at the worlds strongest X-ray source (SPring8 in Japan) we have furthermore established the nature of impurity phases in the samples. These also cannot be determined by conventional X-ray methods. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Information of the atomic arrangement in thermoelectric materials is key to understanding their properties and being able to improve these in a rational and scientific manner. Using synchrotron radiation we have studied the oxidation states and local structures in the high performing Zn(4)Sb(3) structure, which has aided in developing the knowledge on the structure of this intriguing and promising compound. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

A method for synthesising nanoparticulate bismuth telluride has been developed. Small particles of Bi(2)Te(3) are of interest due to their reported higher thermoelectric efficiency compared to bulk Bi(2)Te(3). More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

First principles DFT methods were applied to calculate suitable models of 14 structurally modified C(o)Sb(3). Changes induced on the geometry, electronic structure and electronic transport properties (ETPs) from frame-substitution (at 8c and/or 24g positions), voids filling (2a position) and a combination of these two structural modifications of C(o)Sb(3) have been detailed. Dopant atoms included Ni, Fe, Te, Sn, La, Ba. Effects on geometry, electronic structure and ETPs of different structural sittings for the dopants allowed to assess which sittings realize or are most favourable for a given dopant (or combination of dopants) or which is the maximum amount allowed for a given dopant at a given position. For instance Te preferentially locates at 24g positions rather than at 2a position and La can fill only about 1/8 of the cubic voids of C(o)Sb(3). Density of states and their atomic projections on the atoms responsible of structural changes along with calculated electronic transport properties have enlightened the different role played by La, Ba, and Ni as fillers. Further information was provided by the atomic charges obtained within the Quantum Theory of Atoms in Molecules approach and by the atomic contributions to the electronic bands around the Fermi energy. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

A novel chemical alloying method has been developed for the fabrication of nanocrystalline Bi-Te alloys. The method consists of a combination of solution chemical method and thermal processing under controlled heating conditions. The components have been co-precipitated from a solution and the precursor consists of a solid solution of the different intermediate compounds, and exhibits high reactivity. Calcination and hydrogen reduction of the precursor at moderate temperature, 350degrees Celsius, for 2 hrs resulted in the alloying of these elements to obtain the pure phase of the thermoelectric material. The method developed in this work is simpler than conventional melt processing and produced a 95-98% yield in laboratory-scale. High concentration of grain boundaries provided by nanostructuring is expected to lower the thermal conductivity of the material and further increase the thermoelectric performance. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Nano-sized powders of C(o)Sb(3) Skutterudites with partial substitution of Co by Ni and Fe, respectively, and partial substitution of Sb by Te, delivered by KTH, have been consolidated by hot pressing at DLR to bulk-like samples suitable for functional characterization. The samples have been characterized concerning spatial homogeneity of the Seebeck coefficient. The temperature dependence of the thermoelectric properties has been measured by different methods at DLR. Structure and chemical composition have been analysed by XRD, XRF, and SEM. Optimisation. The material has been functionally optimised by partial substition of Sb by Te. Tests on these materials have shown an improvement in the TE properties. Most important, these measurements have shown, that nano-technology is indeed capable to produce materials with drastically reduced thermal conductivity. In comparison to known bulk values, the obtained thermal conductivity was reduced by a factor of 4 to 5. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

By using first-principles DFT computations, we deconvoluted structural information on Zn(4)Sb(3) from the space-time data averaging inherent to X-ray structural determination. We found that the material is a 0.184:0.420:0.396 mixture of 12-Zn atom cells and two kinds of 13-atoms cells, one with one vacancy and two interstitial Zn atoms, the other with two vacancies and three interstitial Zn atoms, respect to highly symmetric Zn12Sb10. We demonstrate that interstitial Zn atoms do the trick in Zn(4)Sb(3). They supply electrons and enhance the thermopower, besides lowering thermal conductivity. Adding interstitials to the 12-Zn structure yields a change from p- to n-doping. Zn12.82Sb10, with interstitials in 82% of the cells, is a p-doped semiconductor whose thermopower is very sensitive to composition and/or doping. We demonstrate the extreme sensitivity of electronic transport properties to subtle changes in the material composition. We find a maximum Z(e)T value of 4.68 at T=670K, for a p-doping level of 0.01h/cell with respect to Zn(13)Sb(10). This corresponds to Zn(12.995)Sb(10) stoichiometry, if the doping level is modified by adding interstitial Zn atoms only. A maximum ZT value of 2.11,at T=670K, is estimated from Z(e)T, assuming the lattice contribution to the total thermal conductivity independent of doping level. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

A novel chemical alloying method has been developed for the fabrication of nanocrystalline skutterudites with the general formula M(x)Co(1-x)Sb(3-y)Te(y) (M= Ni or Fe). The method consists of a combination of solution chemical coprecipitation and thermochemical processing under controlled conditions. The coprecipitated precursor consists of a homogeneous mixture, or even a solid solution, of the different intermediate compounds of the desired elements, and exhibits high reactivity for further processing. After the precursor is calcined at a moderate temperature 350degrees Celsius for 1-2 hrs, followed by hydrogen reduction at 450 350oC for 2 hrs, the nanocrystalline skutterudite is produced at a phase purity >95%. The method developed in this work is much simpler than conventional melt technique. High concentration of grain boundaries provided by nanostructuring, has been shown to lower the thermal conductivity of the material and so enhance the thermoelectric performance greatly. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

First principles DFT methods, using local gaussian basis sets, were applied to calculate the electronic structure of type I inorganic clathrates ( Sr(8)Ga(16)Ge(30), Ba(8)Ga(16)Ge(30) ). A careful tuning of the basis set enabled us to consider crystal groups with low or no symmetry, corresponding to different sittings for the Ge and Ga atoms. Use of the Quantum Theory of Atoms in Molecules (QTAM) on the resulting wave functions allowed us to put the guest-host interactions on a physically sound basis. The charge transfer from guests to the semi-conducting framework was obtained by integrating the charge within the guest basins, using the QTAM recipe, which is firmly rooted in quantum mechanics. The result is freed from any arbitrariness, at variance with estimates from other approaches (X-ray adsorption, magnetic measurements, etc.). Contrary to previous experimental and theoretical estimates, the guest atoms are found to be mostly ionic. We fully reconciled theory with the Zintl phase view that structural chemists have of these systems. We show this discrepancy to be related to wrong definitions used for electron transfer. Preliminary investigations ruled out any important relationship between the observed increase in the thermoelectric figure of merit with increasing external pressure, and the host-guest charge transfer. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Transport properties, namely Seebeck coefficient, electrical conductivity, thermal diffusivity, specific heat, carrier concentration and Hall mobility have been measured with confidence on an extensive number of nanostructured CoSb(3) skutterudite materials of the form R(y)Co(1-x) M(x)Sb(3) where R are rare earth materials and M=Ni, Fe, C etc. The dimensionless figure-of-merit ZT of these materials has been obtained from these measurements using the relationship ZT = [alpha](2) [sigma] / [lambda] where alpha is the Seebeck coefficient, sigma the electrical conductivity and lambda the total thermal conductivity of the material. Thermal conductivity was obtained from thermal diffusivity (lambda) measurements, sample density (d) and specific heat measurements (Cp) according to the relationship [lambda]=[lambda] C(p) d. Significant reductions of 4 to 5 times in the total thermal conductivity when compared to unfilled C(o)Sb(3) materials prepared by conventional methods have been obtained. Despite relatively low electrical conductivity values, a ZT of above 0.3 at 650K has been measured to-date on these materials. Further improvements in electrical conductivity through optimised doping and consolidation parameters are expected to yield an increase in ZT values. Increased ZT values will enable high performance thermoelectric devices to be realised, extending the range of possible applications for this technology throughout Europe and worldwide. More information on the NANOTHERMEL project can be found at: http://www.nanothermel.org

Deliverables

Share this page

Download