整改
二极管
整流器(神经网络)
热的
材料科学
光电子学
电流(流体)
数码产品
热流
热能
工程物理
凝聚态物理
电气工程
物理
热力学
计算机科学
工程类
功率(物理)
随机神经网络
机器学习
循环神经网络
人工神经网络
作者
M.J. Pérez,Antonio Fornieri,Francesco Giazotto
标识
DOI:10.1038/nnano.2015.11
摘要
A thermal diode with two orders of magnitude higher on/off ratio than that previously achieved can be obtained by combining normal metals and superconductors. Thermal diodes1,2—devices that allow heat to flow preferentially in one direction—are one of the key tools for the implementation of solid-state thermal circuits. These would find application in many fields of nanoscience, including cooling, energy harvesting, thermal isolation, radiation detection3 and quantum information4, or in emerging fields such as phononics5,6,7 and coherent caloritronics8,9,10. However, both in terms of phononic11,12,13 and electronic heat conduction14 (the latter being the focus of this work), their experimental realization remains very challenging15. A highly efficient thermal diode should provide a difference of at least one order of magnitude between the heat current transmitted in the forward temperature (T) bias configuration (Jfw) and that generated with T-bias reversal (Jrev), leading to ℛ = Jfw/Jrev ≫ 1 or ≪ 1. So far, ℛ ≈ 1.07–1.4 has been reported in phononic devices16,17,18, and ℛ ≈ 1.1 has been obtained with a quantum-dot electronic thermal rectifier at cryogenic temperatures19. Here, we show that unprecedentedly high ratios of ℛ ≈ 140 can be achieved in a hybrid device combining normal metals tunnel-coupled to superconductors20,21,22. Our approach provides a high-performance realization of a thermal diode for electronic heat current that could be successfully implemented in true low-temperature solid-state thermal circuits.
科研通智能强力驱动
Strongly Powered by AbleSci AI