Integrated Optics

Research Objectives

Our current research in integrated optics focuses on materials, devices, and device integration. In our view, the most important current challenges in photonics revolve around these central themes. Our research objectives can be summarized as follows:

• Chip-scale integration of multiple optical functions - A traditional challenge associated with integrated optics is that diverse materials are required to satisfy the full range of envisioned functionality. This has made it difficult to realize multi-function photonic circuits in a ‘monolithic’ fabrication process. Our research will develop new materials, processes, and devices to enable the desired integration of numerous functions (light emission, routing, amplification, switching, etc.).


• High-density integrated optics - A second historical challenge is the large size and low density of integrated optic devices, due to bending radius limitations in traditional waveguides and relatively weak light-matter interactions in active devices. Our research will lead to high-density integrated optics, sometimes termed microphotonics. This will be achieved through development of new waveguide structures (optical wires, photonic crystals, etc.), new artificial materials (nanocluster glasses, etc.), and new device approaches (three-dimensional integration, resonant structures, etc.).


• Integration of photonics and electronics – As discussed below, we view silicon as the photonics platform of the future. Our research aims for convergence of photonics and silicon electronics. Currently, this vision guides our choice of materials in our integrated optics research. In the longer term, we aim to demonstrate opto-electronic integration on silicon CMOS chips. This will be achieved through development of new materials and techniques that satisfy the thermal, mechanical, and contaminant restrictions imposed by the standard silicon IC process.


• Integration of photonics and microfluidics - Optical detection is the most common non-destructive method used in microfluidic lab-on-a-chip systems. Optical detection of cells and other analytes based on laser-induced fluorescence is superior in terms of selectivity and sensitivity. The volumes being probed in these systems are extremely small (pL to nL). However, the optical systems used to perform the measurements are generally constructed from bulk optical components. The alignment of the components is critical in order to properly focus the light onto the small volume being examined and the number of detector configurations is limited. A potentially simpler solution is to integrate the optical devices directly with the microfluidic components and increase both the flexibility and performance of fluorescence detection.

Optical materials and fabrication processes

A thorough understanding of the optical, thermal, and mechanical properties is required before a material can be considered a suitable candidate for optical integration. TRLabs is investigating novel optical materials (passive and active polymers, chalcogenide glasses, and silicon rich oxides) for integration on silicon. As with current CMOS based electronic technologies in silicon, integrated optical devices will not contain a single material system but be made up of a number of compatible materials, each performing a specific function.

Processes must also be developed for fabricating the photonic devices. Making use of the Nanofabrication Facility and the University of Alberta, TRLabs has an active effort to study the deposition and patterning of the materials being used to make photonic devices. This research is fundamental to carrying forward our photonic device research program. Sub-themes include:

• thin-film deposition – thermal and electron beam evaporation, pulsed laser deposition, spin-casting
  
• patterning and feature definition – direct patterning of waveguide devices in chalcogenide glass, development of wet and dry etching processes
  
• study of material compatibilities – adhesion, chemical compatibility, thermal and mechanical compatibility of diverse glass and polymer films

Passive microphotonics

Passive optical structures such as waveguides, Bragg gratings, and optical resonators are the key building blocks for integrated optics. TRLabs is researching passive micron scale structures fabricated in chalcogenide glass and polymers for use in optical networking and as optical interconnects in integrated circuits. These passive structures include sub-micron optical wires (see figure at right) and rib waveguides, low-loss waveguide bends, mode converters, power splitters, dispersion compensation devices, and wavelength mux/demux structures. Ideally, chip-scale microphotonics should retain some of the best features of electrical interconnects on ICs. Based on this, we believe a couple of key sub-themes emerge:

• Wavelength-scale waveguides and devices – advance waveguide structures such as high index contrast optical wires, photonic crystal waveguides, and plasmon-based waveguides can enable optical interconnect densities that compete with the wiring dimensions on current ICs (few hundred nm). Further, they make compact 90 degree bends and junctions possible.
• Three-dimensional integrated optics – since electrical signals can be easily routed between layers on a chip (using vias), it is highly desirable to develop structures that enable optical signals to be transferred efficiently between horizontal layers. Resonant couplers and mirrors might be employed for this purpose.

Optical sources and amplifiers

Optical gain devices (lasers and amplifiers) are probably the key enablers for silicon-based microphotonics. A practical, monolithically integrated laser technology on silicon promises to have a revolutionary impact on computing and optical transport systems. To underscore this, a critical challenge cited in the International Technology Roadmap for Semiconductors is “a high efficiency, high switching rate laser source, monolithically integratable into Si CMOS, (at low cost)…”. Further, it is expected that the next phase of development in optical communications will be driven by metro, access, and fiber-to-the-home markets. These markets are characterized by a need for low-cost, compact, mass-produced optical components. While the erbium-doped fiber amplifier (EDFA) is an established technology for long-haul fiber networks, it typically does not meet the cost and size requirements envisioned for more localized networks. Low-cost amplifier arrays on silicon, pumped electrically or optically, are expected to be key enablers for such applications.

TRLabs has been investigating rare-earth doped chalcogenide glasses as a potential gain medium that can be deposited on a silicon platform. Initial work has focused on Er-doped alloys, which can amplify signals in the standard 1550 nm fiber band. We have recently started to investigate Neodymium-doped glasses with a gain band in the 1060 nm range, which should have superior compatibility with silicon-based photodetectors. Also recently, we have kicked off a collaborative project to investigate waveguide amplifiers and sources based on silicon rich oxide glasses. These nanocomposite glasses offer great potential for realization of silicon-based lasers and amplifiers. Dr. Meldrum has developed a thin-film synthesis technique that requires significantly lower processing temperatures (<500 C) compared to competing techniques in the literature. This material fits nicely into our overall strategy for integration on silicon.

Nonlinear integrated optics

We are studying both 2nd and 3rd order nonlinearities in glasses and polymers. The electro-optic effect, a 2nd order nonlinear effect present in poled electro-optic polymers and chalcogenide glasses, is being exploited to design integrated amplitude and phase modulators. Traditional materials such as lithium niobate or III-V semiconductors cannot be directly integrated onto silicon platforms. 3rd order nonlinearities are used to realize devices in which light signals control each other. Such ‘all-optical’ devices, if they become practical, promise to greatly increase the capacity and functionality of computing devices and optical communications networks. Chalcogenide glasses exhibit fast, large 3rd-order nonlinearities at fibre communication wavelengths, and are thus promising materials for all-optical switches and pulse-shaping devices.

Integrated acousto-optic devices are unique enablers of tunable wavelength-selective elements and optical switching for optical networking. It is expected that integrated acousto-optic devices will play a role in future all-optical networks. Certain chalcogenide glasses have attractive acousto-optic properties in the IR region, and integrated devices based on these glasses have been demonstrated.

Building on recent progress in chalcogenide integrated waveguides, TRLabs will investigate active devices based on surface acoustic waves in chalcogenide films. Relative to most of the recent work in the literature, typically employing lithium niobate, our approach will have the advantage of being compatible with a silicon platform. Chalcogenide glass is not piezo-electric, so it will be necessary to integrate a piezo-electric film adjacent to the chalcogenide film. Possible options include ZnO, AlN, or poled polymers. Initial work will employ ZnO films deposited by pulsed laser deposition.