The physics of plasmonic gain in low-dimensional electronic systems

A number of theoretical studies, including our own, have recently shown that the properties of plasmons can be influenced strongly by dc currents flowing in the materials that support them. This, in principle, leads to the possibility of plasmon gain by a transfer of power from the dc current, with the strength of the interaction between the plasmons and the current being proportional to the dc electron drift velocity, which can be very high in semiconductors but is only small in metals, due to much more frequent electron collisions. Unfortunately, there have been only a few direct experimental observations of the interaction between plasmons and dc currents. These have, though, confirmed the basic prediction - that the plasmon wavevector depends on the strength and direction of the dc electron drift velocity. The current state-of-the-art in the field can be summarised as follows: 1) Several theoretical works predict THz oscillations by interaction of plasmons with dc currents in low-dimensional systems (LDSs); 2) Low-power THz emission from LDSs has been obtained, but the role of plasmons in these experiments has not been fully understood; 3) Current-driven plasmon gain has never been directly measured.

Further progress now requires basic experimental studies of the interaction between plasmons and dc currents, supported by theoretical interpretation of the mechanisms for plasmon gain. Previous emission experiments were ill-suited for this purpose. Coherent THz emission should appear at a threshold when plasmon gain (due to the interaction with a dc current) exceeds loss, as in a laser. However, in the sub-threshold regime when the gain is weak, no emission can be observed and, therefore, the gain could not be quantified in previous experiments. Moreover, weak plasmon emission may be obscured by other mechanisms, notably by thermal radiation. Likewise, many widely used methodologies based on photothermal plasmon detection are unsuitable since they do not permit interactions with dc currents to be probed.

Our experimental technique allows us to address these issues, and to study the propagation of plasmons through LDSs, and recover their full THz transmission spectra. Suited ideally for studying plasmon gain, it gives an unique ability to characterise the same plasmon device in the passive (no dc current), sub-threshold, and above-threshold regimes. At the same time, our new theoretical models will allow us both to analyse experimental results and to design optimized structures. These parallel advances will be crucial to understand and demonstrate plasmon gain for the first time. Step-by-step improvements will ultimately lead to THz emission - powerful, cheap, and tuneable (100 GHz - 10 THz) sources are a long-standing goal for the international community, but despite progress in many areas, no compact, room-temperature source exists with CW mW power output between 1 and 4 THz. Our disruptive technology offers the potential to solve this problem since semiconductor plasmons have resonant frequencies that fall in the THz range when confined to LDSs. It will also contribute to realising the potential of plasmons in LDS for THz detectors and sensors.

The global market for terahertz (THz) systems and devices has been predicted to grow [1,2] at an annual rate of 32% between 2016 and 2023, exceeding $1bn by 2024. The main drivers are the development of new systems and the growing demand for the THz technology in laboratory research, medical imaging, high-speed communications, and security. All THz systems require a THz source, which then largely defines its capabilities and limitations. A novel practical source with low cost (particularly with respect to Ti:sapphire laser systems, the workhorse of THz spectroscopy), would have a profound impact both on academic research and industrial applications, opening up a number of markets. Our guiding strategy here will be to first demonstrate and optimise our new plasmonic THz sources, while fully protecting the underlying IP, before we then investigate licensing to potential instrumentation manufacturers. Teraview are the leading THz company internationally, and we have a good track record of collaboration with them and licensing IP to them, for example, thus providing a possible exploitation route in due course. Before this can be contemplated, however, we need robust demonstrations of the emission, to be delivered here.

NPL has a long-standing interest in the development of new calibration standards for high frequency / THz radiation, which historically they have been able to exploit in calibration measurements for companies in the microwave instrumentation sector, which is now pushing into the THz range. Part of this work now involves the development of on-chip calibration standards, and protocols for the accurate measurement of dielectric and other samples in the frequency range between 100 GHz and 1.1 THz. Owing to the close synergy of this work with our programme, NPL have undertaken 1) to verify our measurements of emitted power, with reference to their own traceable THz power standards, and 2) to measure and confirm our measurements of waveguide losses using the on-chip calibration standards available at NPL. This will help both to quantify and (ultimately) to minimise the on-chip losses in our systems.

The research will provide an excellent opportunity for the PDRAs employed on the project to develop their careers by receiving training in a range of techniques of importance to the UK high-technology economy. The host School at Imperial has committed a PhD student to their PI (Sydoruk) for a project aligned with the current programme, subject to award of this grant (see Letter of Support), and there will be further opportunities for PhD studentships in topics aligned with this programme at both Imperial and Leeds during the course of this project.

IP protection will be carried out in collaboration with the technology transfer arms of the University of Leeds and of Imperial College. The research groups will liaise with their respective support teams before work commences to draw up a contract for the division of all IP coming out of this project after award, as well to cover discussions with companies under appropriate non-disclosure agreements during the programme. Representatives from NPL will attend our consortium meetings, along with representatives selected from other companies during the course of our translation activities.

We expect that our initial experimental demonstrations (during year 3) of amplifying and oscillating structures will generate heightened industrial interest, so at this point onwards the programme we will take the technology to industry, under non-disclosure agreements, and investigate licensing arrangements for the IP that we will have protected in the form of patents as appropriate. References: [1] http://tinyurl.com/zusx456 [2] http://tinyurl.com/ze2gjf7