The efficiency of a thermoelectric material is characterized by the dimensionless figure of merit (ZT) … A number of high ZT metallic thermoelectric materials have been developed in the past three decades such as tellurium (Te)-based compounds and selenium (Se)-based compounds. However, these materials are usually too heavy, expensive, or toxic to find universal application, particularly when the environmental impact is considered. Furthermore, they are limited in their ability to harvest electricity at high temperatures, such as from solar and industrial waste heat, due to their decomposition and volatilization at elevated temperatures.

Over the past decade, there has been growing interest in alternatives to these alloys, and this interest has led to the development of oxide-based materials. … The perovskite-type strontium titanate (SrTiO₃) has attracted significant interest owing to its large carrier effective mass, good thermal stability at high temperature, and strong structural tolerances for substitutional doping. A popular route for enhancing the thermoelectric performance of SrTiO₃, which is poor in its intrinsic form, is to increase its electrical conductivity by doping. SrTiO₃ can be easily modified by doping with either trivalent elements (e.g., Lanthanum (La) and dysprosium (Dy)) at the Sr sites, and pentavalent elements (e.g., Niobium (Nb)) at the Ti sites. Additionally, efforts have been made to reduce the thermal conductivity while preserving the electrical conductivity.

… Although significant progress has been made in the enhancement of the thermoelectric performance of SrTiO₃, major challenges remain. The critical parameters of S, σ, and κ are interdependent, which complicates the efforts to improve the average ZT of the material. The thermoelectric properties of SrTiO₃ continue to be inferior to those of the traditional metal materials. This low ZT has restricted the commercial application of oxide-based materials. Thus, SrTiO₃ based thermoelectric materials with higher ZT are desirable. Furthermore, thermoelectric materials have a “thermal window” within which they are able to convert heat energy into electrical energy, effectively an operating range within which the materials display useful properties. Outside this temperature range, which is unique to any particular material, the material has little or no ability to generate electrical energy in response to heat. Consequently, the exploitation of these materials in different applications is limited not only by their ZT values, but also by the range of temperature within which they operate.

—Lin et al.

Few thermoelectric materials exhibit a broad thermal window. This complicates application in a device in dynamic thermal conditions, because performance for a single material is never optimal. Therefore, for conditions with large temperature variations, systems are used combining thermoelectric materials with different thermal windows (a “cascade” structure). This complicates fabrication and increases cost.

Synthesizing a a material with a broad thermal window or a technique that can produce this kind of material is thus vitally important to optimize device performance and simplify device integration, the researchers explained.

Based on earlier work on La-doped strontium titanate, the Manchester team selected La_0.67Sr_0.9TiO₃ as their base formulation. They prepared the graphene/LSTO nanocomposites (G/LSTO) by mixing together dispersions of exfoliated graphene and LSTO, then by filtering, drying and milling the resultant mixture.

The LSTO composites incorporated one percent or less of graphene; the resultant materials were reduced and possessed a multiphase structure with nanosized grains.

They found that the thermal conductivity of the nanocomposites decreased upon the addition of graphene, whereas the electrical conductivity and power factor both increased significantly.

These factors, together with a moderate Seebeck coefficient, meant that a high power factor of ∼2500 μWm⁻¹ K⁻² was reached at room temperature at a loading of 0.6 wt % graphene.

The highest thermoelectric figure of merit (ZT) was achieved when 0.6 wt % graphene was added (ZT = 0.42 at room temperature and 0.36 at 750 °C), with >280% enhancement compared to that of pure LSTO.

Most significantly, the addition of graphene broadened the operational thermal window, an effect not seen with other dopants/additives. The ZT was >0.25 at room temperature to 750 °C. A prototype device achieved an open-circuit potential of 600 mV with cold side at 18.5 °C and a temperature difference of 400 °C. This discovery highlights an alternative strategy to nanostructuring for developing high-performance, environmental friendly, low cost thermoelectric materials and extending their use to lower-temperature applications such as automotive and low-temperature generators.

—Lin et al.

The authors acknowledge funding from the University of Manchester Intellectual Property, Engineering and Physical Sciences Research Council and the European Union Seventh Framework Programme.

Resources

• Yue Lin, Colin Norman, Deepanshu Srivastava, Feridoon Azough, Li Wang, Mark Robbins, Kevin Simpson, Robert Freer, and Ian A. Kinloch (2015) “Thermoelectric Power Generation from Lanthanum Strontium Titanium Oxide at Room Temperature through the Addition of Graphene” ACS Applied Materials & Interfaces doi: 10.1021/acsami.5b03522