- CONTENT LIST

1. Project Summary

2. Project objectives

3. Participant List

4. Contribution to programme/key action objectives

5. Innovation

6. Community added value

7. Contribution to community social objectives

8. Economic development and scientific and technological prospects

1.	Project Summary

Objectives

Photonic Crystals (PCs) in 2 dimensions have recently given rise to striking validation of the basic concepts. The purpose of the PCIC (Photonic Crystal Integrated Circuits) project is to apply these concepts to Photonic Integrated Circuits (PICs) on InP, intented for telecommunication applications, for which we foresee three major advantages: ultra high compacity, simple and low-cost fabrication routes and implementation of novel functionnalities. The aim is to develop integrated-optics oriented PC fabrication routes and combine carefully validated PC modelling tools to achieve several important PC-based building blocks, including low-loss waveguides, bends, couplers, combiners and filters. Demonstrators consisting of integrated systems of active (lasers) and passive elements for WDM applications will be developed to validate the impact of PCs in integrated optics.

Description of the work

The workplan is divided into three technical workpackages, in addition to the management and dissemination workpackages 4 (WP4) and 5 (WP5). WP1 is devoted to modelling tools, and characterisation techniques for parameter determination. Modelling will use PC-oriented methods in integrated optics configurations and will provide a CAD toolbox for PC-based building blocks and demonstrators. Physical characterisation will allow to assess fabrication issues via the optical properties of test structures fabricated in WP2, both in GaAs and InP material systems. WP2 addresses the development of all technologies required for PC test-structures and building block fabrication (straight and bend waveguides, mirrors, filters, combiners) and will rely on high resolution electron-beam lithography and high aspect ratio dry-etching in the GaAs and InP material systems. Passive PCs will be characterised by polarised transmission and reflection spectroscopy. Modal propagation losses of the building blocks will be assessed. Based on the results of WP1 and WP2, WP3 is devoted to the design, fabrication and test of more complex PC-based demonstrators. These consist of : (i) a Passive Combiner (at least 4=>1) ; (ii) a Monolithic Source integrating two PC-based tunable lasers and a combiner which should demonstrate the feasibility of a broadband compact WDM transmitter at 1.55µm, (iii) a High Power compact source made of an array of at least four 1.48µm lasers and the combiner, (iv) a high quality-factor Drop Filter. We believe that this integrated effort towards physical characterisation, validated fabrication techniques and modelling tools, building blocks and demonstrator integration is quite unique in the field of photonic crystals and should lead to clearcut assessment of the impact of PCs in PICs results, owing to the wide range of competences of the partners in both PCs and integrated optics fields.

Milestones and expected results

Main milestones are : (i) Modelling tools adapted for Integrated Optics PC-based functions, (ii) Assessment of a PC fabrication route on InP, compatible with guided wave applications, (iii) Experimental demonstration of building blocks (waveguides, bends, combiners) Expected results: (a) CAD toolbox with a repertory of PC-based advanced functions (b) Technology roadmap to PC-based Photonic Integrated Circuits (PCICs) (c) Integration of PC building blocks into compact circuits with major improvements over classical WDM ones

2. Project objectives

The development of ultra-high-capacity optical networks such as those based on wavelength-division multiplexing (WDM) will require complex optoelectronic integrated circuits (OEICs, also called photonic integrated circuits, PICs, when restricted to active and passive optical functions) to achieve the functions no longer attainable by usual electronic techniques. So far, only few implementations of PICs exist, due to a mix of technological issues (varying according to materials systems): fabrication routes can be complex and require several critical steps like multiple regrowths. This often results in a limited number of achievable functions in PICs which does not correspond to the system requirements. Furthermore existing PICs use large surfaces (sometimes in the several square cm range) which results in high prices.

 

 

The present project aims at demonstrating for the first time good optical waveguiding properties, devices and integrated systems made of PCs. This is a medium-risk project ( although none of the properties and functions have been demonstrated, recent advances by the partners point to the sound basis of PC concepts, see refs of partners 2, 3 and 7 respectively) which could yield very high rewards if successful. We expect major improvements in the performance and density in such systems: PCs allow wavelength-sized device structures, such as ultra-short bends without losses, as propagating waves can only be coupled to evanescent waves in the photonic propagation bandgap. Therefore, coupling to other photon modes does not introduce losses in this context, on the contrary of usual waveguides. In the same vein, fabrication fluctuations are less severe than in usual integrated optics as they also only introduce coupling to evanescent, lossless waves. This is true up to relatively large hole-to-hole fabrication tolerances (~10 % size fluctuation) for PCs in the GaAs/GaAlAs system, a tolerance well within reach for InP-based systems. In addition, we expect to reach such effects with simpler fabrication schemes, hence leading to lower costs, as well as new functionalities.

The information age is bringing us in the "tera-era" : demands are terabits per second for information transport, tera-operations per second for processing power, terabyte information storage. It might appear that the transmission capacity is there to satisfy demand on the short, medium and long term, but it is not the case in an cost-effective manner, which would ensure universal fast access to networks, witness slowdowns of the web : Whereas modern optical fibers have seen their capacity increase dramatically in the recent years , thus decreasing the cost component associated to bit transmission over long distances, the same has not been true for the signal generation, routing and processing parts of networks (Fig.1 below). This is due to the fact that these functions are still mainly performed electronically, without taking advantage of the huge throughput of optics. Today, no electronic circuit is able to handle the data output of an optical fiber and this will be even more the case tomorrow. Therefore, the huge optical link capacity cannot be used at its full potential down to the customer.

The techniques used in PCIC, based on innovative concepts, promise to provide new routes to achieve most, if not all, of the building blocks required for complex optical processing functions in optical communications, and a number of new ones. This will be achieved using an eventually single, non-critical lithographic-and-etch step, which could allow the simultaneous fabrication of all the required functional elements. In addition, the optical circuits should be very compact, thanks to the strong waveguiding properties of PCs, thus allowing for very high functional density of optical circuits.

While at applications of PCs in integrated optics are at a very preliminary stage, the concepts are sound and the team assembled to undertake the project gathers some of the best european talent in a coherent and coordinated way. The size of the PCIC consortium is large because the task to be undertaken is wide. Given the potential impact of PCs in a critical technological field, no risk should be taken that might lead to a late start or partial coverage by the EU. Therefore, the consortium assembles highly qualified partners from different european countries, including a telecommunication systems industry, a service provider and research institutes, each of them with its specific skills and know-hows in the Photonic Crystal and Integrated Optics fields which will ensure that all the needed workforce and knowledge is available to undertake an ambitious project.

We seek resources to create both well-proven fabrication routes and a profound knowledge and understanding of the phenomena involved. Taking as examples microelectronics or integrated optics, it is clear that they could take off only when enough basic understanding was created which allowed efficient fabrication and modelling tools to be developped, which in turn allowed to design and fabricate the many circuits required for evaluations. We intend to continuously develop and validate such tools along the project and this is a reason for the project size infinancial resources as there are a number of sizeable efforts to pursue, at a demanding fabrication level, but also in modelling and characterisation. This is also the reason for the range of skills assembled in PCIC, width which could not be reached in a single, or even two countries of the EU.

It is more and more clear that while european industries enjoy a good competitive position today in telecommunications technologies, these are undergoing fast and major evolutions. If we wish to maintain, or better expand the european industry position in the future, we need to explore today the new technologies which could lead to breakthroughs.

The optical communication area features among the faster growth rates ever in optoelectronics. While the per-bit cost of fiber-launched data has plumetted to such a level that tera-bit/s flow rate will soon be "cheap", the bottleneck has creeped gradually to the switching and routing level. The treatment of electrically converted optical signal at switches has a constant cost for several years (Fig.1, part C3), indicating that all-optical routing is an unescapable prospect in optical-fiber data transfer to match the increased traffic.

The PCIC project aims at solving this major bottleneck of high-throughput optical communications, allowing the direct optical manipulation of optical data streams. Many options, hybrid or monolithic, are already being pursued to manipulate optical beams directly, in some cases with great efficiency and speed. However, today's solutions of integrated optics, although based on excellent engineering, are bulky and costly, in a number of ways in a unavoidable manner due to the physical basis of the concepts being used. These are the main hurdles to the widespread use of integrated optical circuits (IOCs). Taking as an example of issue the large size of integrated optical circuits, it is mainly due to the weakly confining structures used. This could be avoided by using stronger optical confinement, made possible by structures with high index contrast, but in that case losses become prohibitive (unless extremely precise fabrication techniques are used, which is beyond today's fabrication capabilities in production). At the same time, the fabrication routes for such structures are quite complex, involving several critical steps such as regrowths.

Having these challenges in mind, PCs offer two novel aspects which will impact positively the cost and size of IOCs :

This ensuing lower cost will ease the implementation of intelligent all-optical networks towards the rapidly growing category of end customers for dealing with Mbit/s to Gbit/s data rates.

At the end of the project, both the industrial and other partners will have a clear idea of the short and longer term potential of PCICs. It will be a relatively small matter for the industries to carry the most promising candidates all the way to development, and then to comercialisation. In addition, this project is likely to generate significant intellectual property and this can be exploited by the partners under agreement.

It should be emphasized that the partners have a good track record in the dissemination and use of their results. Their previous EU projects activities resulted in numerous talks at international conferences, many of them (>40) invited, in the organization of two summerschools and of a workshop in Europe (in Erice and Cargese) which led to an important development of this field of research in Europe as many European researchers participated, many of them in other consortia, and finally in the formation and coordination of COST 268, the perhaps broadest pan-european forum on wavelength-scale photonics, including photonic crystals. Some research efforts on novel photonic devices are already being commercialized (see the specification sheet for microcavity light-emitting diodes on MITEL's web page : http://www.mitelsemi.com/products/ ). These efforts for dissemination and industrial transfer will of course be central in PCIC. The transfer of the knowledge to industrial applications and to new materials systems of industrial interest represents more than two-thirds of the effort. The two industrial partners will indeed evaluate in various manners the possibility to commercially produce and market PCICs. At this point it is essential to remind that PCICs require fabrication facilities which are for a vast majority of materials studied under PCIC fully compatible with existing production facilities.

The transfer to production is seen at this stage as depending on the balance between PCIC performance, costs, and user's needs, all of which are to be evaluated by the industrial partners as the program develops. More exact statements about this issue will be made along the program.

The participating laboratories have also additional ties with other industrial partners, be it through other programs in ESPRIT, ACTS, COST or IST.

At a more scientific level, the partners will disseminate their results through scientific publications, participation at conferences, and could, if the commission would like it to happen, and if funding could be available, organize during the course of PCIC a summerschool and aworkshop, in connection with other European programs, as was done in other european projects.

We do not foresee at this stage any problem with intellectual property rights (IPR). IPR may arise from the design of specific devices or from specific processing routes which should belong to the partner who makes the effort as it would rely on know-how unrelated to the program. Therefore, we consider at this stage the work undertaken under PCIC to be open, which should facilitate its results dissemination. However, whenever IPR issues would occur, the consortium will decide by its management procedures whether to keep it within all partners, or to limit it to selected partners, or to keep an open policy.