Phase-one goals for the large-scale CMOS photonics project, HELIOS, have been met or exceeded by the 19 partners involved, according to French organisation CEA-Leti, coordinator of the pan-European consortium.
Launched by the European Commission in 2008, the €8.5 million project is designed to develop microelectronics fabrication processes for integrating photonics with CMOS circuits and to make the technology available to a wide variety of users. Participants include major European CMOS photonics and electronics research centres and companies and potential users of the technology.
'The idea of CMOS photonics is to fabricate photonic devices using the same fabrication processes as used in microelectronics,' explained Laurent Fulbert, photonics programme manager at Leti. Speaking with Electro Optics, he said that by being able to employ standard industrial processes, the manufacturing costs can be reduced and the fabrication yields increased. Also, by co-integrating optics and electronics on the same chip, high-functionality, high-performance and highly integrated devices can be fabricated.
First-phase achievements of HELIOS have concentrated on light photodetection and light coupling/routing. 'We have achieved some very good results,' Fulbert commented. For example, an integrated germanium on silicon photodetector was developed with a bandwidth of 90GHz – 'that's a world record', he said. Also, a high-efficiency coupling structure used to couple light from the chip to an optical fibre has been demonstrated.
The four-year project includes the development of essential building blocks, such as efficient sources (silicon-based and heterogeneous integration of III-V on silicon), fast modulators and, more long term, the combination and packaging of these building blocks for the demonstration of complex functions to address a variety of industrial needs.
The main application area of silicon photonics is optical communications, but optical sensing and biological applications could also benefit from the low-cost manufacturing processes – for example, disposable biosensors could be produced using this technology. 'Most of the short- or medium-term applications are in telecommunications – that's where the market is now,' commented Fulbert, 'but all the building blocks we're developing could be used for biosensing too.'
Optical fibres can transmit higher bandwidths than copper wire, which is why it holds so much potential for use in telecommunications, but currently the cost of optical components cannot compete with copper wire. One way to cut costs is to integrate the optical components and use volume microelectronics fabrication processes.
Fulbert noted three advantages to integrating electronics and photonics: the first being the low-cost, high-yield nature of using microelectronics fabrication processes; secondly, integration will drastically reduce the packaging, alignment and assembly costs of the device. Most of the photonics components that exist today, for example the transceiver for optical communications, are made of several discrete components, such as a laser diode, a lens, an isolator, a photodetector, all of which have to be aligned very accurately and packaged. Integrating all of this functionality on the same chip, said Fulbert, will remove the alignment cost and produce a device that is much more compact and efficient. Finally, Fulbert commented that the performance of photonics devices can be improved by adding electronic feedback loops or other electronics technology.
'Europe has a well-established photonics-components industry and it is strategically important for us to maintain photonic chip design and chip-integrating functions that provide new opportunities for our microelectronics companies and enable us to compete with other countries,' said Laurent Malier, CEO of Leti. Large multinationals like Intel and IBM are conducting research in the field along with a few start-ups, mainly in the US, such as Luxtera.
Advances in CMOS photonics will also move the emphasis from device component to architecture. Industrial and RTD efforts then could be focused on new products or new functionalities rather than on the technology level.
According to Fulbert, the aims for 2010 are to continue to improve the different building blocks involved – results on the laser source should be available, a 10Gb/s modulator will be developed along with initial results from a 40Gb/s modulator, and the first prototype of an optical transceiver for communication will be developed.