Photonics @ Interface: Heterogeneous Integrations for Generation, Detection, Conversion, and Modulation

Lead Research Organisation: University of Southampton
Department Name: Electronics and Computer Science

Abstract

We will develop novel photonic devices by bonding two different semiconductor substrates with different spacing between atoms. It was very difficult to ensure an excellent quality at the interface, because the atoms cannot connect perfectly, if the lattice spacing is different. We will overcome this problem by making nano-scale tiny trenches to allow atoms to expand for releasing the strain accumulated at the interface. The quality of the interface is very important to make sure to reduce defects in the atomic scale. We will use the interface for making a highly sensitive detector to identify even just single photon (quantum of light), which is impossible to realise with defects due to the noise from additional carriers created by defects. This detector will be useful for future quantum technologies to enable secure communications and powerful commutations. We will also develop a novel laser and high speed optical switches by using this interface.

Our project is summarised as follows:
1. Novel Manufacturing Technologies for Enabling Heterogeneous Integrations: We will develop new wafer-scale bonding process technologies to allow excellent interface qualities without defects. Our challenges to overcome the difference of lattice constants and thermal expansion constants for bonded materials. We will accumulate comprehensive knowledge for new bonding techniques.
2. Si/Ge Avalanche-Photo-Diodes and Si/Ge Lasers: The strain engineered interface will enable us to reduce dark currents of Si/Ge Avalanche-Photo-Diodes (APDs) to the level useful for detecting single photons at room temperature. Si/Ge APDs are also useful for LiDAR (Laser Imaging Detection and Ranging). The improved interface quality also enables to achieve lasing of Ge on a Si substrate towards monolithic integrations.
3. Si/LiNbO3 Hybrid Optical Modulator and Second-Harmonic-Generators: We will also bond LiNbO3 on a Si substrate, which allows us to utilise the electro-optic and nonlinear effects of LiNbO3, while keeping the advantages of nanoscale patterning of Si. The hybrid optical modulator with a slot waveguide will be operated at attojoule power consumption by removing the 50 Ohm-termination. The hybrid Second-Harmonic-Generators (SHGs) will convert various wavelengths to generate green and UV lights for much denser data-writing on DVDs.

We think our approach will establish a new way of making heterogeneous interface with improved quality. Wafer-scale bonding of a patterned substrate is certainly well-known, but the nano-scale patterning to form perfect bonding in atomic-scale has not yet been achieved, yet. We will accumulate comprehensive knowledge on the developed interface, in terms of various physical parameters such as strains, voids, adhesions, and defects for researchers in nanoelectronics and photonics.

Planned Impact

The market share of semiconductor industries in the EU including the UK is significantly dropped down to 10% from the peak value of 34% at 1980, while the global market is still growing with the rate 7% p.a. We believe that the share drop was not coming from the weakness of science and technologies in the UK, but due to the lack of continuous large investments for Moore's scaling law. For this to happen about £10bn would need to be invested to create state-of the art foundries. It is not easy to compete for commodity-product markets, where the huge investments are critical while the profit margin is small. Our focus is instead to engage for a high-value and small-volume market for UK industries as based on famous Blue Ocean strategies. In particular, our identified niche is to focus on photonics applications, where the UK industry and academia are strong and particularly active. Here we will have less competition (a blue ocean), rather than heavily competitive stringent market (bloody red ocean). The Blue Ocean strategies also encourage us to develop disruptively new technologies, which are promising for the future in terms of scalabilities and cost reductions to change the figure-of-merit, completely. More specifically, we envisage following scenarios towards the impact for each device applications:

Si/Ge Avalanche-Photo-Diodess:
Phase 1: Free usages of our devices for the UK collaborators for single-photon-detections.
Phase 2: Providing proto-types for the international collaborators through University owned spin-out company, ECS Partner.
Phase 3: Applications for LiDAR (Light Detection And Ranging) towards an autonomous driving car and for long-hall communications, working together with industries.

Ge Lasers:
Phase 1: Demonstrating a continuous-wave lasing of Ge at room temperature by optical pumping.
Phase 2: Making a Ge laser diode on a Si substrate.
Phase 3: Integrating a Ge laser diode on a Si Photonics chip for ubiquitous short-reach optical communications.

Si/LiNbO3 hybrid optical modulators:
Phase 1: Demonstrating a attojoule power operation by employing a novel slot-waveguide configuration to remove the 50-Ohm termination.
Phase 2: Transferring the technologies to the UK industry for long-hall optical communications.
Phase 3: Opening up novel applications, such as polarisation modulators with passive photonic integrated circuits for communication technologies towards next generations and quantum technologies.

Si/LiNbO3 hybrid SHGs:
Phase 1: Demonstrating compact designs within a mm-scale chip to convert wavelengths in the 1550nm rage and to generate green and UV lights.
Phase 2: Increasing the efficiencies and reducing the coupling and the propagation loss of the hybrid waveguide.
Phase 3: Working together with the UK industries to open up novel applications including telecommunications and data storages.

Within the duration of this project, we are aiming to achieve Phase 1 of above targets. For some applications, we will proceed beyond Phase 1. We will check our progress at the Mid-Term Review meeting after 2-years from the start of the project, and address the key challenges and find out solutions to overcome the problems. In order to mitigate the potential risks, we have prepared a lot of ideas and contingence plans for process developments. We will also seek advices from experienced senior researchers in the advisory committee both from industry and academia. We will work together with industrial and academic collaborators. We will do our best for transferring our technologies to the existing UK industries, but if no company wishes to take over, we will consider to make a start-up to disseminate our technologies. We believe that our devices are novel based on bright ideas and competitive for the high-value photonics market.

Publications

10 25 50