Generic Model of Bulk Silicon and Nanowires

Summary

So far we have studied semiclassical theories for electrons and phonons for bulk and low‐dimensional structures (quantum well, nanowires, etc.). Silicon (Si) is considered as a future potential thermoelectric material due to its abundance, nontoxicity, large‐processing infrastructure for low cost, and high yield, in addition to its good electronic transport properties. However, bulk Si has a high thermal conductivity of ~150 W/mK, leading to a poor thermoelectric performance (ZT ≈ 0.01). Si nanowires have Seebeck coefficient and electrical conductivity that are comparable to bulk Si, but those with diameters of about 30 nm exhibit 100‐fold reduction in thermal conductivity, yielding ZT = 0.6 at room temperature [1]. In spite of a remarkable achievement on thermal conductivity with nanowires, there remains a question as to why electronic transport properties with nanowires did not show better performance than those with bulk Si, as expected from the work of Hicks and Dresselhaus (1993) [2, 3]. According to the literature, it seems difficult to determine electron scattering mechanisms for nanowires. Assuming a constant scattering rate [4, 5] or using experimental values [6], whereas phonon scattering mechanisms are given much attention. Actually, Monte Carlo and Ab initio simulations of detailed electronic transport for Si nanowires including electron scattering were reported by Ramayya et al. in 2012 [7] and Neophytou and Kosina in 2011 [8], respectively, wherein ZT does not show much enhancement with decreasing nanowire diameter. However, these give detailed information on electronic transport in nanowires, which would be valuable data for the present work. In the previous chapters, it was shown that transport properties for both bulk and nanowires give rise to changes only in their electron density of states (DOS). Therefore, we first develop a nonparabolic two‐band Kane model for bulk Si in terms of the DOS by comparing the computed results with experimental data, and try to extend it from bulk to nanowire by just switching the electronic DOS. We wish to see the realistic electronic transport behaviors for nanowires. The validity for this attempt is open to future work.