Modeling and theoretical efficiency of a silicon nanowire based thermoelectric junction with area enhancement

Abstract

Recent experimental work suggests that individual silicon nanowires with rough surfaces possess a thermoelectric figure of merit as high as 0.6 near room temperature. This paper addresses the possibility of using an array of such nanowires in a thermoelectric junction for generation. Employing a model of frequency dependent phonon boundary scattering, we estimate the effective thermal conductivity of the array and investigate heat flow through the junction. We show that charge transport is largely unaffected by the roughness scales considered. Enhancing the area for heat exchange at an individual 200 μm × 200 μm p-n junction yields significant temperature differences across the junction leading to power >0.6 mW and efficiency >1.5% for a junction with effective thermal conductivity <5 W/mK, when the source and sink are at 450 K and 300 K, respectively. We show that relatively short nanowires of ∼50 μm length are sufficient for obtaining peak power and reasonable efficiency. This substantially reduces the challenge of engineering low resistivity electrical contacts that critically affect power and efficiency. This paper provides insight into how fundamental transport in relation to bulk heat transfer and charge transport, affects the performance of thermoelectric junctions based on nanostructured materials.

Seebeck Coefficient of Nanowires Interconnected into Large Area Networks

“…15−18 Differently from the phonon behavior, the electron mobility in heavily doped, few tens of nanometer wide, nanowires is only weakly affected by surface scattering, as demonstrated by theoretical analysis. 19,20 The high doping level increases the impurity scattering, reducing the electron mean free path to few nanometers (2−5 nm, depending on doping). Therefore, electron scattering due to surfaces plays its effect only within few nanometers (that is, a distance comparable with the bulk mean free path) close to the surfaces.…”

“…The electron mobility in a nanowire is still an open problem for nanowires narrower than 10 nm, when quantum effects begins to become important even at room temperature. , In our case, W > 50 nm so that at room temperatures we can exclude important quantum effect; modification of electron mobility can also derive from the surface scattering. Several works measured the SiNW electron mobility by different techniques. − Differently from the phonon behavior, the electron mobility in heavily doped, few tens of nanometer wide, nanowires is only weakly affected by surface scattering, as demonstrated by theoretical analysis. , The high doping level increases the impurity scattering, reducing the electron mean free path to few nanometers (2–5 nm, depending on doping). Therefore, electron scattering due to surfaces plays its effect only within few nanometers (that is, a distance comparable with the bulk mean free path) close to the surfaces .…”

Peak thermoelectric power factor of holey silicon films

“…4 for the Seebeck coefficient. We use results from our previous work on scattering rates for electrons and holes 20,27 to calculate the electrical conductivity. From our calculations, the optimal doping is in the range 4-6 × 10 19 cm −3 across doping polarities.…”

Silicon-based nanostructures for integrated thermoelectric generators

“…In this case, MFP in bulk is in the order of 2-5 nm, implying that significant σ reduction takes place only when the diameter of the nanowire is reduced below this threshold. This issue has been theoretically discussed by Seong et al and Shi et al [90,91] but, so far, there is no experimental evidence of carrier scattering at nanowire boundaries affecting thermoelectric properties.…”