Multi-THz resonant-tunneling-diode oscillators

Resonant tunnelling diodes (RTDs) are the highest-frequency active electronic semiconductor device existing nowadays. The first RTD sources (RTD oscillators) operating above 1 THz have been demonstrated in 2010-2011. RTD oscillators are quite simple, very compact and they work at room temperature. The frequencies of RTD oscillators have reached almost 2 THz in the last year, their output power is also increasing; spectroscopic, imaging and wireless high-speed data-communication experiments have been already demonstrated. The recent development in the field indicates that RTD oscillators could be a unique enabling technology for real-world THz applications. In addition, the previous studies indicate that there is much more room for further improvement of THz RTD sources and our aim in the project is to demonstrate that experimentally. We know from our previous studies of RTDs and RTD oscillators, that they should be working at frequencies above 2 THz. Our aim in the project is to achieve 2-3 THz with the RTD oscillators. An RTD oscillator consists of a resonator or a resonant antenna integrated with a small-size RTD. We are going to investigate two types of oscillators, one is more conventional, it consists of a resonant slot antenna fabricated on a chip surface and the chip is then mounted on a Si lens. The other type is technologically more challenging: it is an oscillator fabricated on a thin dielectric membrane. The membrane oscillators are much more compact and potentially could be more efficient. To achieve operating frequencies in the range 2-3 THz, one needs to optimise many parameters of the antenna/resonator and of RTDs, adjust them in a proper way to each other, one needs to minimise parasitics, develop an efficiently emitting THz antenna, etc. These requirements make design and fabrication or RTD oscillators a very challenging task. An electromagnetic simulator will be used for the oscillator design, design of RTD layers will stay similar to the one we used in our previous works, although the RTD barriers will be made thinner. The dimensions of THz RTDs should be in the sub- micrometre range, therefore a combination of optical and e-beam lithography will be used in fabrication of the RTD oscillators. Back-side processing will be required for the membrane oscillators. If the project is successful, we will extend the operating frequencies of RTD oscillators up to 3 THz. The oscillators will be covering the major part of THz spectrum and that will be the highest- frequency active semiconductor devices, the frequency-boundary of electronics will be extended. Although the project aims an evolutionary development of RTD oscillators, it should lead to revolutionary consequences: the RTD oscillators should make deep-THz frequency range accessible for real-world applications. 1

THz frequencies are thousand times higher than the usual operating frequencies of electronic devices and systems, e.g., of our cell phones. On the other hand, the THz frequencies are about thousand times lower than the optical frequencies, e.g., of visible light. Except for highly-sophisticated scientific instruments, THz frequency range is not accessible in practice nowadays. No applications in this frequency range exist. To change that, we are investigating resonant-tunneling diodes (RTDs) and RTD oscillators. RTDs were shown to be the highest-frequency active electronic device in existence presently. In this particular project, our aim was to clarify, what are the ultimate frequency and output-power limitations of THz RTD oscillators. Our focus was on slot- and patch-antenna types of RTD oscillators. That are the most simple and common types of RTD oscillators with their specific pro and contra. We carried out a thorough theoretical investigation into fundamental limitations of the above types of RTD oscillators. We found out that 2 THz is nearly an ultimate limit for slot-antenna RTD oscillators. We also identified the optimal designs and configurations for such type of oscillators to achieve the highest frequency and output power. We developed a general model for non-linear RTD dynamics, which is required for an accurate analysis of RTD-oscillator output power and their operating frequencies. On the experimental side, highly-advanced fabrication technology was developed both for patch- and slot-antenna RTD oscillators. We presented a special ``island'' design for on-chip slot-antenna RTD oscillators; at frequencies of around 2 THz, this design leads to a drastic reduction of the slot-antenna losses. With this design, we achieved a record-high output power of around 2.2 W at the fundamental frequency of 1.74 THz. Furthermore, we demonstrated 10-30 W at the fundamental frequency of 1.09-0.6 THz with double-RTD patch-antenna oscillators. That is the first demonstration of patch-antenna RTD oscillators with the fundamental oscillation frequency above 1 THz. The above parameters of the patch-antenna RTD oscillators are close to the state-of-the-art level for all other types of RTD oscillators at around 1 THz. The improvement of the parameters is due to an advanced design with dramatically reduced parasitic inductance and due to the use of state-of-the-art high-current-density RTDs. We made also several steps beyond the initial plan of the project: we demonstrated operation of our RTD oscillators in the radar applications, started looking into array configurations of RTD oscillators, and discovered a broad-band parasitic radiation in continuous-wave THz photomixing systems (which we tried to use to drive RTD oscillators); state-of-the-art results were demonstrated in all these cases as well. The project results led to completion of two PhD theses and a large number of high-level publications.