Nonlinear polariton phenomena in GaN-based slab waveguides at temperatures up to 300 K

Novel quasi-particles, so-called polaritons, can be formed in an optically active semiconductor material due to mixing of photons and excitons (an exciton is an analogue of hydrogen in condensed matter). While two photons colliding in free space do not interact, polaritons strongly repel due to the exciton component in their wavefunctions. The Sheffield group has demonstrated that polariton-polariton interactions are several orders of magnitude stronger than effective photon-photon interactions in any other ultrafast photonic materials, where light is weakly coupled to matter. The efficient polariton-polariton scattering can be utilised for development of novel ultrafast light sources, frequency mixers and converters as well as scalable and compact devices performing control of light by light on a very fast timescale and at very low signal intensities (potentially at a single photon level). Potentially, this may have a strong impact on development of novel photonic signal processing and quantum technology hardware. The polariton platform is also well suited to the study of fundamental many-body phenomena ranging from Bose-Einstein condensation, superfluidity and solitons to quantum correlated phases in important physical systems, such as photonic analogues of topological insulators or quantum Hall systems.

So-far phenomena due to polariton interactions have been explored in GaAs microresonators only at 4-50 K. Our proposal capitalises on the recent demonstration of ultraviolet polaritons in waveguides based on AlGaN/GaN material, where polaritons are robust at 300 K given the large exciton binding energy and highly efficient exciton-photon coupling; this provides an opportunity to explore the novel room temperature physics of interacting polaritons and to bring polariton applications to reality.

In order to explore the fundamental physics of interacting GaN polaritons we will address polariton solitons, i.e non-spreading wavepackets stabilised by the nonlinearity. These interactions depend on many factors, such as the exciton Bohr radius, binding energy and the exciton fraction in the polariton wavefunction. A weak light pulse propagating in a polariton waveguide broadens with time, as different frequency components propagate with different velocities. By contrast, by increasing the pulse intensity polariton interactions are expected to cancel the spreading leading to formation of a soliton. Given the giant polariton nonlinearities solitons are expected to form at ultra-low thresholds and on a very short length-scale of ~10 micrometers.

Another advantage of GaN-based waveguides over GaAs counterparts is that exciton-photon hybridisation occurs over a broad range of frequencies enabling polariton-polariton scattering from a spectrally narrow pulse to a broad quasi-continuum of states. As a result very short UV polariton soliton pulses with a duration down to 10's femtoseconds are anticipated. The outcome of this research would also enable realisation of novel ultraviolet broadband pulsed sources and frequency converters operating at very low thresholds, which are important for many spectroscopy applications in biophotonics and molecular photochemistry.

Finally, we will demonstrate a prototype of a compact ultrafast all-optical polariton switch by exploiting nonlinear interactions between the polariton pulses with different central frequencies propagating in a photonic circuit based on coupled waveguides. These interactions induce nonlinear phase shifts in the optical signals enabling routing and switching of the ultrafast optical pulses. Given the very fast response of polariton system such a switch is expected to operate at THz rates. We note that exciton-polaritons in GaN-based nanostructures are also very robust against screening by hot electron-hole carriers, enabling study of the amplification of propagating polaritons in the presence of optical gain. This is essential for scalability of polariton devices.

Our proposal has significant potential for impact over a range of timescales in both the academic community and industry. The latter is demonstrated by the letters of support from two companies IBM and NANOPLUS. The other industrial firms having a strong expertise in nanophotonic technology, such as Intel, Sharp, Toshiba or Oclaro with strong R&D activities in Europe will be also interested in the discoveries we aim to make.

In the timescale of the programme the outcomes are likely to be of particular significance to researchers in the field, as summarized in the section on Academic Beneficiaries. In the longer term of 10-20 years or so, the research is likely to gain strong momentum in applied directions. Particularly, the outcomes of the proposal may be used in novel all-optical or mixed electro-optical signal processing and computing. The idea of using nanophotonic circuits along with electrical components is being explored by some industrial giants (IBM, Intel) and may significantly increase the speed of modern supercomputers. All-optical active components that enable control of light by light offer great potential for the development of ultrafast components operating in the Tb/s regime. Moving towards compact digital-optical signal manipulation has potential for transformative impact on both communication components in networks and signal processors operating at a speed significantly exceeding currently available electronic analogues.

The main aim of this application is to explore the fundamental physics of giant polariton nonlinearities in GaN-based nano-photonic structures operating at 300 K. These structures have all the properties necessary for the development of energy effective, miniature and very fast optical integrated circuits and signal processing devices (solitonic routers and switches). Currently there is a strong research activity on the growth and fabrication of GaN-based devices on Si: fabrication of GaN high electron mobility transistors on Si has been demonstrated. Since the study of polaritons in GaN-based structures is still at its fundamental stage here we focus on AlGaN/GaN structures grown on GaN templates. If the goals of this proposal are achieved the next step could be the study of polariton physics in GaN photonic structures grown on Si. The fabrication of both energy efficient GaN active photonic circuits and Si-integrated electronic devices on the same chip is a likely prospect in the longer term.

The giant polariton-polariton interactions also enable polariton-polariton scattering over a broad range of wavelengths (30-60 nm) in UV range leading to generation of ultrafast pulses with a duration down to 10's of femtoseconds. Such a quasi-continuum emission is expected to be observed at excitation powers 3-4 orders of magnitude lower than in other ultrafast nonlinear optical systems. Potentially, low threshold polariton sources of broadband pulsed UV emission could be used for many spectroscopy applications in biophotonics and molecular photochemistry. Our project partner NANOPLUS will be the direct beneficiary of this research.

As well as the ultrafast circuit applications, polariton platform is very promising for development of quantum photonic devices. Potentially, giant nonlinearity in polariton structures with strong confinement may lead to squeezed polariton solitons with reduced fluctuations in the amplitude or phase of the soliton field or enable operation of active polariton devices at a single photon level (single photon phase shifters). Realisation of such scalable devices would be a breakthrough in quantum computation technology where information is encoded into squeezed light or single photon degrees of freedom.

Not only researchers and entrepreneurs will benefit from the results of our project, but we also believe that our studies may provide a further stimulus for young people at schools and universities to become strongly involved in photonics and polaritonics.