[Back]

M. Feiginov:
"THz resonant-tunneling diodes";
Talk: Russia-Japan-USA-Europe Symposium on Fundamental & Applied Problems of Terahertz Devices & Technologies (RJUSE TeraTech-2017), Troy, NY, USA (invited); 10-01-2017 - 10-05-2017.

Resonant-tunneling diodes (RTDs) have reemerged in the last years as a promising and viable technology for sub-THz and THz radiation sources. RTD oscillators working above 1 THz have been demonstrated in 2010-2011 for the first time. Presently, their operating frequencies are getting close to 2 THz. The contemporary RTD oscillators can deliver ~0.5 mW at sub-THz frequencies and ~10 μW at THz frequencies. It was demonstrated that RTD oscillators could be as small as a fraction of mm2, they consume low DC power and operate at room temperature. RTD oscillators have been used in high-speed wireless data-transmission experiments recently. In meanwhile, we understand better the physics of RTD limitations at very high frequencies, we could find the ways, how to overcome the limitations and suggested novel types of RTD oscillators.
We were concerned with investigation of inherent limitation of RTDs and of relaxation tunnel processes (described by a time constant τrel) inside them. We have predicted that τrel is not limited by the tunnel lifetime (τ) of electrons in the quantum well of RTDs. It turns out, that the relaxation processes could be much faster, as well as much slower than τ, depending on the operating regime of RTDs. We have also predicted that RTDs with unusually highly-doped collector, should exhibit negative differential conductance at frequencies far beyond τ and τrel limits, i.e., when τω>>1 and τrelω>>1. We have proved the effects experimentally up to ~600 GHz. Following further along this line of development, with RTDs with heavily doped collector and rather low current density we could demonstrate oscillators working at ~1.1 THz in the regime τrelω≈1. The analysis of the oscillators suggests that multi-THz frequencies should be achievable with RTD oscillators and the RTD oscillators have much room for further optimization.
Further on, we have brought the concept of an RTD with heavily doped collector to its logical limit: we have increased the collector doping to the extend, that the lowest quantum-well subband of the RTD stays immersed under collector Fermi level. In such RTDs, the electrons are injected into the quantum well not only from the emitter side (which is usual for RTDs), but also from the collector. The oscillators with such RTDs were demonstrated to be working up to ~1.5 THz.
We have also analyzed theoretically the travelling-wave microstrip RTD oscillators. Such oscillators could be seen conceptually similar to THz quantum-cascade lasers (QCLs) with the metal-metal waveguide and with a single cascade period (an RTD) as their active core. However, contrary to THz QCLs, microstrip RTDs should be working at room temperature. Assuming realistic parameters of RTD layers (we took them from our past experimental studies) we show that microstrip RTD oscillators could be working up to ~1.5 THz. We expect that RTD layers specifically optimized for microstrip oscillators should extend the operating frequency even further.
In conclusion, there is much room for further development of RTDs and RTD oscillators, that should enable operation of RTD oscillators at multi-THz frequencies and increase their output power. RTD oscillators have potential to evolve into enabling technology for real-world THz applications.