†Department of Mechanical Science and Engineering, ‡Department of Electrical and Computer Engineering, and §Micro and Nanotechnology Laboratory, University of Ilinois at Urbana-Champaign, Urbana Illinois 61801, United States.
Nano Lett. 2015 May 13;15(5):3159-65. doi: 10.1021/acs.nanolett.5b00267. Epub 2015 Apr 8.
Existing theory and data cannot quantify the contribution of phonon drag to the Seebeck coefficient (S) in semiconductors at room temperature. We show that this is possible through comparative measurements between nanowires and the bulk. Phonon boundary scattering completely quenches phonon drag in silicon nanowires enabling quantification of its contribution to S in bulk silicon in the range 25-500 K. The contribution is surprisingly large (∼34%) at 300 K even at doping of ∼3 × 10(19) cm(-3). Our results contradict the notion that phonon drag is negligible in degenerate semiconductors at temperatures relevant for thermoelectric energy conversion. A revised theory of electron-phonon momentum exchange that accounts for a phonon mean free path spectrum agrees well with the data.
现有理论和数据无法量化室温下半导体中声子拖拽对塞贝克系数(S)的贡献。我们通过纳米线和体材料之间的对比测量证明了这一点。声子边界散射完全抑制了硅纳米线中的声子拖拽,从而能够量化其对 25-500 K 范围内体硅 S 的贡献。即使在掺杂浓度约为 3×10(19)cm(-3)时,300 K 时的贡献也非常大(约 34%)。我们的结果与声子拖拽在与热电能量转换相关的温度下对于简并半导体可以忽略不计的观点相矛盾。一个考虑声子平均自由程谱的电子-声子动量交换修正理论与实验数据吻合得很好。