Contrasting SnTe-NaSbTe2 and SnTe-NaBiTe2 Thermoelectric Alloys: High Performance Facilitated by Increased Cation Vacancies and Lattice Softening.

Defect chemistry is critical to designing high performance thermoelectric materials. In SnTe, the naturally large density of cation vacancies results in excessive hole doping and frustrates the ability to control the thermoelectric properties. Yet, recent work also associates the vacancies with suppressed sound velocities and low lattice thermal conductivity, underscoring the need to understand the interplay between alloying, vacancies, and the transport properties of SnTe. Here, we report solid solutions of SnTe with NaSbTe2 and NaBiTe2 (NaSnmSbTem+2 and NaSnmBiTem+2, respectively) and focus on the impact of the ternary alloys on the cation vacancies and thermoelectric properties. We find introduction of NaSbTe2, but not NaBiTe2, into SnTe nearly doubles the natural concentration of Sn vacancies. Furthermore, DFT calculations suggest that both NaSbTe2 and NaBiTe2 facilitate valence band convergence and simultaneously narrow the band gap. These effects improve the power factors but also make the alloys more prone to detrimental bipolar diffusion. Indeed, the performance of NaSnmBiTem+2 is limited by strong bipolar transport and only exhibits modest maximum ZTs ≈ 0.85 at 900 K. In NaSnmSbTem+2 however, the doubled vacancy concentration raises the charge carrier density and suppresses bipolar diffusion, resulting in superior power factors than those of the Bi-containing analogues. Lastly, NaSbTe2 incorporation lowers the sound velocity of SnTe to give glasslike lattice thermal conductivities. Facilitated by the favorable impacts of band convergence, vacancy-augmented hole concentration, and lattice softening, NaSnmSbTem+2 reaches high ZT ≈ 1.2 at 800-900 K and a competitive average ZTavg of 0.7 over 300-873 K. The difference in ZT between two chemically similar compounds underscores the importance of intrinsic defects in engineering high-performance thermoelectrics.

Chemical Insights into PbSe- x%HgSe: High Power Factor and Improved Thermoelectric Performance by Alloying with Discordant Atoms.

Thermoelectric generators can convert heat directly into usable electric power but suffer from low efficiencies and high costs, which have hindered wide-scale applications. Accordingly, an important goal in the field of thermoelectricity is to develop new high performance materials that are composed of more earth-abundant elements. The best systems for midtemperature power generation rely on heavily doped PbTe, but the Te in these materials is scarce in the Earth's crust. PbSe is emerging as a less expensive alternative to PbTe, although it displays inferior performance due to a considerably smaller power factor S2σ, where S is the Seebeck coefficient and σ is electrical conductivity. Here, we present a new p-type PbSe system, Pb0.98Na0.02Se- x%HgSe, which yields a very high power factor of ∼20 µW·cm-1·K-2 at 963 K when x = 2, a 15% improvement over the best performing PbSe- x%MSe materials. The enhancement is attributed to a combination of high carrier mobility and the early onset of band convergence in the Hg-alloyed samples (∼550 K), which results in a significant increase in the Seebeck coefficient. Interestingly, we find that the Hg2+ cations sit at an off-centered position within the PbSe lattice, and we dub the displaced Hg atoms "discordant". DFT calculations indicate that this feature plays a role in lowering thermal conductivity, and we believe that this insight may inspire new design criteria for engineering high performance thermoelectric materials. The high power factor combined with a decrease in thermal conductivity gives a high figure of merit ZT of 1.7 at 970 K, the highest value reported for p-type PbSe to date.

Absence of Nanostructuring in NaPb mSbTe m+2: Solid Solutions with High Thermoelectric Performance in the Intermediate Temperature Regime.

Thermoelectric devices directly convert heat into electrical energy and are highly desired for emerging applications in waste heat recovery. Currently, PbTe based compounds are the leading thermoelectric materials in the intermediate temperature regime (∼800 K); however, integration into commercial devices has been limited. This is largely because the performance of PbTe, which is maximized ∼900 K, is too low over the temperatures of interest for most potential commercial applications (generally under 600 K). Improving the low temperature performance of PbTe based materials is therefore critical to achieve usage outside of existing niche applications. Here, we provide an in-depth study of the cubic NaPb mSbTe m+2 system of compounds ( m = 1-20) and report that it is an excellent class of low- to medium-temperature thermoelectrics when m = 10-20. We show that the as-cast polycrystalline ingots exhibit degenerate p-type conduction and high maximum ZTs of 1.2-1.4 at 650 K when m = 6-20. Because the ingots are found to be extremely brittle, we utilize spark plasma sintering (SPS) to prepare more mechanically robust samples, and surprisingly, find that SPS results in an undesired change in charge transport toward n-type behavior. We show this unanticipated transition from p-type behavior as ingots to n-type after SPS is due to dissolution of secondary phases that are present in the ingots into the primary matrix during the SPS process, resulting in a transformation from an inhomogeneous state to a solid solution without any observable evidence of nanoscale precipitation. This is in sharp contrast to the seemingly similar AgPbmSbTe m+2 (LAST) system, which is heavily nanostructured. The SPSed NaPb mSbTe m+2 is doped p-type by tuning the cation stoichiometry, i.e., Na1+ xPb m- xSb1- yTe m+2. The optimized compounds have very low lattice thermal conductivities of 1.1-0.55 W·m-1·K-1 over 300-650 K, which enhances the low-intermediate temperature performance and gives rise to maximum ZT values up to 1.6 at 673 K as well as an excellent ZTavg of 1.1 over 323-673 K for m = 10, 20, making Na1+ xPb m- xSb1- yTe m+2 among the highest performing PbTe-based thermoelectrics under 650 K.

Melt-Centrifuged (Bi,Sb)2 Te3 : Engineering Microstructure toward High Thermoelectric Efficiency.

Microstructure engineering is an effective strategy to reduce lattice thermal conductivity (κl ) and enhance the thermoelectric figure of merit (zT). Through a new process based on melt-centrifugation to squeeze out excess eutectic liquid, microstructure modulation is realized to manipulate the formation of dislocations and clean grain boundaries, resulting in a porous network with a platelet structure. In this way, phonon transport is strongly disrupted by a combination of porosity, pore surfaces/junctions, grain boundaries, and lattice dislocations. These collectively result in a ≈60% reduction of κl compared to zone melted ingot, while the charge carriers remain relatively mobile across the liquid-fused grains. This porous material displays a zT value of 1.2, which is higher than fully dense conventional zone melted ingots and hot pressed (Bi,Sb)2 Te3 alloys. A segmented leg of melt-centrifuged Bi0.5 Sb1.5 Te3 and Bi0.3 Sb1.7 Te3 could produce a high device ZT exceeding 1.0 over the whole temperature range of 323-523 K and an efficiency up to 9%. The present work demonstrates a method for synthesizing high-efficiency porous thermoelectric materials through an unconventional melt-centrifugation technique.