2021年第68回応用物理学会春季学術講演会

講演情報

一般セッション(口頭講演)

13 半導体 » 13.6 ナノ構造・量子現象・ナノ量子デバイス

[19a-Z23-1~6] 13.6 ナノ構造・量子現象・ナノ量子デバイス

2021年3月19日(金) 09:30 〜 11:30 Z23 (Z23)

中岡 俊裕(上智大)

10:30 〜 10:45

[19a-Z23-5] Thermal activation behavior of current in semiconductor thermionic cooling heterostructures

Xiangyu Zhu1、Marc Bescond1,2、Gerald Bastard3、Naomi Nagai1、Kazuhiko Hirakawa1,2 (1.IIS/INQIE,Univ. of Tokyo、2.LIMMS-CNRS、3.Ecole Normale Superieure)

キーワード:semiconductor, heterostructure, thermionic cooling

Managing rapid increase in thermal power densities associated with device miniaturization is a major technological challenge. Development of new efficient cooling technologies is therefore urgently required for future progress in electronics. We have been working on semiconductor double barrier heterostructures to utilize the thermionic cooling effect.
We fabricate an asymmetric AlGaAs/GaAs double barrier heterostructure. In the present heterostructure, cold electrons are first injected into the quantum well (QW) by resonant tunneling through the thin barrier (emitter barrier). Subsequently, hot electrons are removed by thermionic emission over the second thick barrier (collector barrier). This sequential two-step conduction process is essential for the cooling effect. Therefore, quantitative understanding on the conduction process is necessary. In this work, we have developed an analytical theory to calculate the two-step current and compared it with experiment.
To clarify the transport mechanism, we have systematically measured I-V curves at various temperatures and also performed theoretical calculations on the sample structure. Figure 1(b) plots the temperature-dependence of the current density measured at 0.1V (green dots), which shows a thermal activation behavior. In the theory, we use two different description for the thermionic emission process.The one is the standard three-dimensional (3D) Richardson theory (blue). The other is a more rigorous theory which takes into account the thermionic emission process from the two-dimensional (2D) QW to the continuum assisted by LO phonon scattering (red). The major difference between the two theories is that the thermal activation energy for the LO phonon-assisted process (red) is smaller than that for the 3D theory (blue) by hwLO. The agreement between theory and experiment is very reasonable and, furthermore, the activation energy determined from experiment agrees with that for the LO assisted process.