ABSTRACT
In 1993, Hicks and Dresselhaus proposed that the sharply peaked electronic density of states (DOS) in two-dimensional (2D) quantum wells and one-dimensional (1D) quantum wires can be used to increase the power factor, P ≡ S2σ, and the thermoelectric œgure of merit, ZT ≡ S2σT/κ, where S, σ, κ, and T are the Seebeck coe²cient, electrical conductivity, thermal conductivity, and absolute temperature, respectively (Hicks and Dresselhaus 1993a,b). In a subsequent theoretical study, Broido and Reinecke suggested that the ZT enhancement in realistic quantum well thin-œlm superlattice (SL) systems can be limited by several factors, including electron tunneling through the barrier layers that modiœes the DOS and limits the power factor, increased charge carrier-phonon scattering rates and reduced charge carrier mobility, and parasitic thermal conduction in the barrier layers (Broido and Reinecke 1995). šey further suggested that freestanding structures such as suspended nanowires may be needed to overcome these limitations in order to obtain a large ZT. Several calculations by Lin and coworkers as well as Mingo found
21.1 Introduction .................................................................................... 21-1 21.2 še 3ω Method ................................................................................ 21-3
21.4 Supported Microdevices for Field E¤ect šermoelectric Measurements ....................................................21-14
21.5 Mesa Structures for šin Film Cross-Plane Measurements .. 21-15 21.6 Time Domain šermal Re£ectance Techniques ..................... 21-17 21.7 Micro-Raman Spectroscopy Measurements ............................ 21-18 21.8 Conclusion ..................................................................................... 21-21 Acknowledgments .................................................................................... 21-22 References .................................................................................................. 21-22
that ZT > 2 at 77 K and 1.5 at 300 K, respectively, is possible in Bi and InSb nanowires with diameter <10 nm (Lin et al. 2000; Mingo 2004, 2006). In addition, the calculation by Rabina and coworkers further suggested that ZT > 2 at 77 K could be achieved in Bi1−xSbx nanowires of diameter <20 nm (Rabina et al. 2001). Because the electron e¤ective mass (m*) is very small in Bi, InSb, and Bi1−xSbx and the de Broglie wavelength (λe) of electrons is proportional to m*−1/2, the electron wavelength becomes comparable to the 10-20 nanowire diameter in these three systems, giving rise to 1D DOS.
