Thermoelectric Measurements and Standards

Our goal is to develop standard reference materials (SRMs), measurement methodologies, and comprehensive data sets (Seebeck coefficient, electrical conductivity, thermal conductivity) for thin film and bulk thermoelectric materials to enable the development of these materials for applications involving waste heat recovery and solid-state cooling. Our approach will facilitate comparison of thermoelectric data between leading laboratories, and accelerate the commercialization of these materials.

Our approach is to develop standard reference materials, measurement methods, and  combinatorial  materials  methodologies  to  accelerate  the  commercial introduction of thermoelectric materials to the market place. Especially for the case of thin film thermoelectric materials, there are currently no methods to accurately and reproducibly  (laboratory to laboratory) measure the material properties that determine thermoelectric conversion efficiency, i.e., Seebeck coefficient (S), resistivity (r), and thermal conductivity (k). High-throughput combinatorial methodologies will be employed to generate comprehensive data sets (S, r, k) for industrially relevant bulk and thin film thermoelectric materials. We will also collaborate with several industrial, university and government laboratories to generate the appropriate data sets and standard reference materials.

Impact and Customers:

• Thermoelectric SRMs and measurement methods will allow for inter-laboratory validation of data, thereby accelerating the selection and optimization of thermoelectric materials. This will lead to more rapid commercialization of thermoelectric materials for waste heat recovery and solid-state cooling applications.

• The widespread use of thermoelectric converters for vehicular waste heat recovery would lead to a 10% improvement in fuel efficiency, translating to a fuel savings of $150 per year for every automobile, as well as decreased CO2 emissions.

• Customers for thermoelectric materials and devices include the automotive and consumer products industries, the military, NASA, and the energy sector.

In addition to equipping a new laboratory for thermoelectric property measurement, we accomplished three other goals: the completion of a round-robin survey of two potential bulk standard reference materials (SRMs) for measurement of the low-temperature Seebeck coefficient, development of a scanning tool for Seebeck coefficient and resistivity measurements on combinatorial thin film libraries of thermoelectric materials, and crystallographic measurements of novel thermoelectric materials.

The Seebeck coefficient round robin data were analyzed using a parametric model to generate common fitted curves for data generated from various laboratories using different measurement techniques and samples from the same batch. Of the two candidate materials, Bi2Te3 and Constantan (a copper-nickel alloy), we found that the coefficient of variance for Bi2Te3 was smaller across the entire temperature range compared to that of Constantan.

In addition, Bi2Te3 has a larger Seebeck coefficient than Constantan. Therefore, we have chosen Bi2Te3 as our prototype SRM, and production of Bi2Te3 low temperature Seebeck coefficient SRMs is proceeding according to schedule. The completion of this SRM will enable accurate instrument calibration, and therefore meaningful interlaboratory comparison of data.

The power factor (S2/r) is seen to peak between the Sr-rich and La-rich regions. From other data (not shown) we can attribute this to the higher electrical conductivity in the Sr-rich region (better carrier mobility due to lattice deformation), and the larger Seebeck coefficient in the La-rich region (decrease of the hole concentration).

Power factor of a (Ca,Sr,La)3Co4O9 film

Finally, in collaboration with G. Nolas of the University of South Florida, we successfully determined two crystal structures of novel thermoelectric materials related to type-II clathrates. In Cs8Na16Ge136-xMx (M=Ag, Cu) alloys, M was found to dope preferably in the most distorted tetrahedral Ge site. As a result, the values of S, r, k can be modified by doping.