2D cobblestone streets – a pleasant walk for electrons and a bumpy ride for phonons. The spark plasma sintered (SPS) sample of Bi2Te3 exhibits interfacial charged defects that can filter low energy electrons, scatter phonons and holes more efficiently than electrons to improve the figure-of-merit.
2D cobblestone streets – a pleasant walk for electrons and a bumpy ride for phonons. The spark plasma sintered (SPS) sample of Bi2Te3 exhibits interfacial charged defects that can filter low energy electrons, scatter phonons and holes more efficiently than electrons to improve the figure-of-merit.

As world supplies of fossil fuels dwindle, the quest for low-cost renewable energy devices is proving more urgent than ever [1]. Of the many clean energy choices (e.g. solar and wind energy), thermoelectric (TE) devices, which convert waste heat into electricity, are emerging as an important and economically viable alternative to power devices. When exposed to the waste heat, TE materials exhibit a voltage across the two ends (Seebeck effect) akin to the voltage across the two ends of an AA battery. In such devices, the efficiency of converting heat into electricity is governed by a set of strongly and inherently coupled materials properties, namely electrical conductivity, Seebeck coefficient, and thermal conductivity. To achieve a high ‘figure-of-merit’ (a measure of performance for TE material) in converting heat to electricity, an ideal TE material should have very high electrical conductivity and Seebeck coefficient with very low thermal conductivity. In this scenario, two important bottlenecks limit our ability to harness the true potential of TE materials: the inherent coupling between materials properties that lead to a natural and inevitable decrease in electrical conductivity (or Seebeck coefficient) when thermal conductivity (or electrical resistivity) is decreased and vice versa. Accordingly, these properties must be decoupled and made independent of each other to reach a high a figure-of-merit and a functional TE device must consist of multiple legs made up of p- and n-type TE materials (possessing high positive or negative Seebeck coefficients, respectively) just like a diode comprises a p-n junction. During device manufacture, different legs must be chosen in such a way that they have a similar and (ideally) temperature independent compatibility factor (i.e. ratio of current density to heat flux) to avoid parasitic losses. The compatibility factor depends upon the material properties, which are temperature dependent (e.g. electrical conductivity). Therefore, the temperature dependence of material properties must be tailored to make the compatibility factor nearly temperature independent.

Defects in materials science and engineering are often perceived as performance limiters, but in the case of TE materials defect engineering using so-called spark plasma sintering (SPS) can help overcome the two main challenges described above. SPS is a high energy, low voltage, pulsed plasma discharge in a low-pressure atmosphere that generates highly localized ohmic heating (up to a few thousand degrees Kelvin) in a few minutes. The SPS method is advantageous over other techniques such as hot-pressing since it goes beyond simple material densification by chemically activating interfaces through electric discharges between neighboring particulates in addition to ohmic heating at the grain boundaries/interfaces. For example, the SPS method can modify interfaces in single elemental polycrystalline Bi to result in ‘double-decoupling’ (simultaneous decoupling of Seebeck coefficient, electrical, and thermal conductivity) [1]. A simple variation in the duration for which the SPS current is held either ‘on’ or ‘off’ dictates the nature and extent of surfaces in nanosized Bi. The SPS-modified surfaces lead to new electronic states that can simultaneously increase electrical conductivity, Seebeck coefficient, and decrease thermal conductivity to result in a six-fold improvement in the figure-of-merit of Bi.

While the case of Bi is intriguing from a fundamental perspective, its efficiency is too low for any practical applications. Unlike Bi, n-type Bi2Te3 is a layered material, which can be viewed as a deck of playing cards with each card only a few atoms thick, with a relatively high figure-of-merit. However, the material properties of Bi2Te3 (and hence its figure-of-merit) are highly temperature dependent. Traditional nanosizing methods simply downgrade all materials properties simultaneously and consequently fail to alter the temperature dependence of the compatibility factor and figure-of-merit in n-type Bi2Te3. Bit n-type Bi2Te3 may be peeled into atomically thin sheets (i.e. the deck of cards is separated into individual cards) and then re-assembled into a pile different from the original arrangment via the SPS process [2].  This re-assembling using SPS enables one to tailor the materials properties of n-type Bi2Te3 for high TE performance. In this approach, the so-called 'interfacial charged defects' generated by SPS can scatter/filter: (i) low energy electrons to increase electrical conductivity; (ii) phonons more effectively than electrons to decrease thermal conductivity; and (iii) holes more effectively than electrons to improve the Seebeck coefficient. 

These improved properties support better compatibility factors that, in turn, open new vistas and opportunities for more highly efficient TE devices. The intriguing and noteworthy element and the significant scientific contribution of these advances is that defects, which often are associated with low performance or efficiency, can be tailored using the SPS process to tune the properties of materials to ones’ advantage. Although SPS processing is advantageous, much still remains to be understood in terms of its more specific mechanism of chemical activation of materials surfaces. Further fundamental research into a detailed understanding of the SPS process is necessary to make rapid progress in not only TE materials but also in the synthesis and manufacturing of other materials.

Pooja Puneet, Ramakrishna Podila, Jian He, Terry Tritt, and Apparao M. Rao.
Department of Physics and Astronomy and Clemson Nanomaterials Center, Clemson University, Clemson, SC 29634 USA.


1. Pooja Puneet, Ramakrishna Podila, Song Zhu, Malcolm J. Skove, Terry M. Tritt, Jian He and A. M. Rao. Enhancement of Thermoelectric Performance of Ball-Milled Bismuth Due to Spark-Plasma-Sintering-Induced Interface Modi?cations. Advanced Materials 25 (2013) 1033.

2. Pooja Puneet, Ramakrishna Podila, Mehmet Karakaya, Song Zhu, Jian He, Terry M. Tritt, Mildred S. Dresselhaus, and A. M. Rao. Preferential Scattering by Interfacial Charged Defects for Enhanced Thermoelectric Performance in Few-layered n-type Bi2Te3. Nature Scientific Reports 3 (2013), DOI: 10.1038/srep03212239.