Jeff Sharp

Thermoelectric materials: New approaches to an old problem

Thermoelectrics is an old field. In 1823, Thomas Seebeck discovered that a voltage drop appears across a sample that has a temperature gradient. This phenomenon provided the basis for thermocouples used for measuring temperature and for thermoelectric power generators. In 1838, Heinrich Lenz placed a drop of water on the junction of metal wires made of bismuth and antimony. Passing an electric current through the junction in one direction caused the water to freeze, and reversing the current caused the ice to quickly melt; thus thermoelectric refrigeration was demonstrated (figure 1).

Enhanced thermoelectric properties of n-type Mg2.16(Si0.4Sn0.6)1−ySby due to nano-sized Sn-rich precipitates and an optimized electron concentration

Due to the rich reserves of the raw materials, along with their low cost and nontoxic nature, Mg2Si1−xSnx-based compounds have generated intense attention from the international thermoelectric community for their application in thermoelectric power generation within the intermediate temperature range. In this work, we have adopted a two-step solid state reaction followed by a spark plasma sintering process to prepare a series of Sb-doped Mg2.16(Si0.4Sn0.6)1−ySby (0 ≤ y ≤ 0.055) solid solutions. We discuss the influence of Sb doping and the microstructure on the thermoelectric properties. Our results confirm that Sb acts as an effective n-type dopant and we estimate the maximum amount of Sb the Mg2Si0.4Sn0.6 structure can accommodate to be ∼2.3% by XRD, DSC and EPMA analyses. The electron transport properties and low-temperature electronic heat capacity measurements reveal that both the light conduction band and the heavy conduction band contribute to the transport in n-type Mg2Si0.4Sn0.6 solid solutions. The highest density-of-states effective mass and power factor were observed for Mg2.16(Si0.4Sn0.6)0.985Sb0.015 with an electron concentration of n ≈ 1.67 × 1020 cm−3, which is likely to be due to the Fermi level positioned within ∼2kBT of both the heavy and light conduction bands providing contributions from both bands. In addition, doping with Sb does not seem to affect the lattice thermal conductivity above room temperature. TEM analysis indicates the presence of Sn-rich precipitates with the size of several tens of nanometers dispersed in the Mg2Si0.4Sn0.6 matrix. Such a nanophase may enhance the boundary scattering of phonons and contribute to a low lattice thermal conductivity. Owing to the above characteristics of the band structure and the microstructure, the Mg2.16(Si0.4Sn0.6)0.985Sb0.015 solid solution with n = 1.67 × 1020 cm−3 possessed excellent thermoelectric properties and achieved a high ZT value of 1.3 at 740 K. Further reductions in the lattice thermal conductivity could be achieved via optimization of the nanophase inclusions, leading to a further enhancement of the figure of merit for Mg2Si0.4Sn0.6-based solid solutions.

Low-temperature solid state reaction synthesis and thermoelectric properties of high-performance and low-cost Sb-doped Mg2Si0.6Sn0.4

Mg2Si1−xSnx compounds are a type of low-price, environment-friendly medium temperature thermoelectric materials with very important prospects for practical application, and the exploration of high performance Mg2Si1−xSnx compounds is currently attracting worldwide interest. In this study, Sb-doped Mg2Si0.6Sn0.4 compounds were prepared through a two-step, low-temperature solid state reaction method combined with the spark plasma sintering technique for rapid densification. The influence of Sb doping amount on the thermoelectric properties of Mg2Si0.6−ySn0.4Sby (0 ≤ y ≤ 0.015) compounds was investigated. The solid solubility limit of Sb in Mg2Si0.6Sn0.4 compounds was estimated around y = 0.0125. As y increased, the electrical conductivity of Mg2Si0.6−ySn0.4Sby (0 ≤ y ≤ 0.0125) compounds increased considerably, while the absolute value of the Seebeck coefficient and the lattice thermal conductivity decreased. The sample with y = 0.0125 had the highest ZT, reaching 1.11 at 860 K, and the samples with 0.005 ≤ y ≤ 0.015 all attained ZTmax > 0.95. The adopted synthesis process also showed very good repeatability and regularity in obtaining thermoelectric properties, together with the capability of precise composition control of Mg2Si0.6−ySn0.4Sby, making it promising for the practical application of Mg2Si based thermoelectric materials.

Thermoelectric properties of Tl2SnTe5 and Tl2GeTe5

We report on the thermoelectric properties of two ternary tellurides with known crystal structures. Both compounds have a very low lattice thermal conductivity. Tl2SnTe5 appears to have a p-type figure of merit about the same as that of Bi2Te3, the best thermoelectric material among binary compounds. We have synthesized mainly polycrystalline samples, but a few small crystals have been grown and used for electrical measurements. Prospects for further improvement of these materials are discussed.We report on the thermoelectric properties of two ternary tellurides with known crystal structures. Both compounds have a very low lattice thermal conductivity. Tl{sub 2}SnTe{sub 5} appears to have a {ital p}-type figure of merit about the same as that of Bi{sub 2}Te{sub 3}, the best thermoelectric material among binary compounds. We have synthesized mainly polycrystalline samples, but a few small crystals have been grown and used for electrical measurements. Prospects for further improvement of these materials are discussed. {copyright} {ital 1999 American Institute of Physics.}

Overview of Solid-State Thermoelectric Refrigerators and Possible Applications to On-Chip Thermal Management

The concept of thermal management in microelectronic components is changing, and so is the potential for solid-state cooling to solve emerging problems. There is a qualitative change that differentiates the past from the future. We discuss past practices and future trends in the electronic cooling markets, setting the stage for an outline of designs and processes that provide new and enabling options. We briefly review the science that is empowering these changes, and conclude with some thoughts on the future direction of thermal management of microelectronics

Assembly and measurement of a hybrid nanowire-bulk thermoelectric device

Bi1−xSbx nanowires are predicted to show an increased thermoelectric efficiency compared to bulk materials. The authors have synthesized dense Bi0.3Sb0.7 nanowire arrays by electrodeposition into porous anodic alumina and fabricated Ni electrical contacts to the wires. The nanowire/alumina composite was assembled into a hybrid nanowire-bulk thermoelectric device, and electrical measurements were used to calculate the device ZT. The nanowire array produced a temperature difference of 7°C and the hybrid couple had a ZT of 0.12, which is on par with an equivalent bulk couple.

Statistical Analysis of a Round-Robin Measurement Survey of Two Candidate Materials for a Seebeck Coefficient Standard Reference Material.

In an effort to develop a Standard Reference Material (SRM™) for Seebeck coefficient, we have conducted a round-robin measurement survey of two candidate materials—undoped Bi2Te3 and Constantan (55 % Cu and 45 % Ni alloy). Measurements were performed in two rounds by twelve laboratories involved in active thermoelectric research using a number of different commercial and custom-built measurement systems and techniques. In this paper we report the detailed statistical analyses on the interlaboratory measurement results and the statistical methodology for analysis of irregularly sampled measurement curves in the interlaboratory study setting. Based on these results, we have selected Bi2Te3 as the prototype standard material. Once available, this SRM will be useful for future interlaboratory data comparison and instrument calibrations.

Atomic Displacement Parameters: A Useful Tool in the Search for New Thermoelectric Materials?

The atomic displacement parameters (ADPs) measure the mean-square displacement amplitude of an atom about its equilibrium position in a crystal. It is demonstrated that the ADPs can be used to identify crystalline solids with unusually low lattice thermal conductivties. A low lattice thermal conductivity is essential in the design of thermoelectric materials with improved efficiencies.The atomic displacement parameters (ADPs) have been measured using powder neutron diffraction as a function of temperature for several clathrate-like compounds (R x Co 4-y Fe y Sb 12 , where R= La, Ce, Yb or TI, x=0.22, 0.8, 1, y=0, 1;Tl 2 SnTe 5 and Tl 2 GeTe 5 ). The ADP data show that in each of the compounds one of the atoms is weakly bound and “rattles” within its atomic cage. This atomic “rattling” severely reduces the ability of these crystals to conduct heat and in some cases the lattice thermal conductivity approaches the theoretical minimum value. In many clathrate-like compounds, the ADP can also be used to estimate the Einstein frequency of the “rattler”, and to predict the existence of localized vibrational modes.

Thermal conductivity measurements of bulk thermoelectric materials

Thermal conductivity is an important material property of the bulk thermoelectrics. To improve ZT a reduced thermal conductivity is always desired. However, there is no standard material for thermoelectrics and the test results, even on the same material, often show significant scatter. The scatter in thermal conductivity made reported ZT values uncertain and sometime unrepeatable. One of the reasons for the uncertainty is due to the microstructure differences resulting from sintering, heat treatment and other processing parameters. We selected commonly used bulk thermoelectric materials and conducted thermal conductivity measurements using the laser flash diffusivity and differential scanning calorimeter (DSC) systems. Thermal conductivity was measured as a function of temperature from room temperature to 500 K and back to room temperature. The effect of thermal cycling on the bulk thermoelectric was studied. Combined with measurements on electrical resistivity and Seebeck coefficient, we show the use of a ZT map in selecting thermoelectrics. The commercial bulk material showed very good consistency and reliability compared to other bulk materials. Our goal is to develop a thermal transport properties database for the bulk thermoelectrics and make the information available to the research community and industry.

Measurement of thermoelectric nanowire array properties

Numerous groups are studying nanowires of Bi/Sb and [Bi/Sb]2[Te/Se]3 as possible high ZT materials. Relative to the corresponding bulk compositions, it is feasible that nanowires will yield both improved electrical properties and reduced thermal conductivity. The measurement of nanowire properties is difficult and various approaches should be considered. We are attempting to infer the thermoelectric transport properties of nanowire arrays by making and testing miniature couples. The couples contain a nanowire array leg and a bulk material leg, and we measure AC resistance, DC voltage, ∆T and Seebeck coefficient. Motivation A growing body of theoretical and experimental work suggests that greater thermoelectric performance might be found in nanostructured, low-dimensional materials. [1] Many of these studies have focused on superlattices, which offer the opportunity for precise experimental control and relatively straightforward modeling. From the viewpoint of