HERMES: High dEnsity Silicon GeRManium intEgrated photonicS

HERMES is aimed at realising a Ge and GeSi material platform that will be aimed primarily at sustaining optical interconnect circuits to meet the density, data rate and power consumption requirements for the continuation of Moore's law beyond 2020. The ITRS roadmap shows a saturation of the number of electrical pins required for input/output on microprocessors beyond 2020 to about 3000 with current technology. This saturation with an ever increasing latency and a limited on-chip clock speed is a bottle neck that high density optical interconnects have to alleviate. To meet the ITRS 2020 goals the target is clear, with over 100 Tb/s off chip IO capability and power consumption for an entire optical link on the order of 100fJ/bit.
This work proposes a solution to this problem and provides a novel means of fabrication to go beyond the capabilities of standard planar silicon photonics circuits. To do so we aim to develop a multilayer optical platform based on localised Germanium/Silicon compounds on insulator compatible with the fabrication of micrometre sized cavity based structures enabling devices such as modulators and detectors. The growth of laser sources based on III/V materials or doped Germanium could also be envisioned but this is beyond the scope of this proposal. The proposed platform will establish a means to fabricate and demonstrate micrometre scale optical devices fit to tackle the 3 dimensional, high density, low voltage and low capacitance requirements needed for very large scale optical integration necessary for optical on chip interconnects.

The results will have a scientific impact in a number of fields and over a range of different timescales. Most immediate will be the development and demonstration of a platform that will be able to sustain integrated localised Ge or SiGe on grown or deposited oxide of variable thichness. This will enable the use of standard silicon wafer and remove the need for expensive Silicon on Insulator substrates which will therefore make the fabrication of optical devices more cost effective. The technique is also CMOS compatible and has the potential to enable the integration on the same platform of nanocavities devices by locally growing Germanium and III/V lasers, detectors as well as modulators based on the frantz Keldish effect or even QCSE devices.

These will be of benefit to a number of UK manufacturers, particularly for Echerkon Technologies Ltd (who has an established relation with the University of Southampton) and IQE whos is one of the leading global supplier of advanced semiconductor wafers and provide advanced crystal growth technology (epitaxy) to manufacture semiconductor wafers ('epi-wafers') to the major chip manufacturing companies.
Echerkon Technologies Ltd has developed a deposition platform which could enhance the cost advantages, by adapting the company's core technology for the production at research and commercial scale, of the Silicon and Germanium containing films. The combination of the device and production development using UK indigenous innovation will make a positive impact to expand the high-tech manufacturing goals, which have been widely recognised as a vital component for the future economic growth of the UK.

The team recognises that combining innovative device development with a mix of novel and demonstrated deposition and growth methods will deliver a significant leap, it is understood that additional risks have to be mitigated. A parallel device fabrication based on more matured methods will be used as a safer fall back route should there be unexpected complication or delays in the high innovation pathway.
This project is expected to be a stepping stone that will contribute to improving quality of life for consumers and to wealth creation, such as low-cost and complex Si chips for next-generation computers and higher-capacity communication systems. The impact on society will be through the enhanced performance of ICT platforms for the next 20 years and beyond, for the digital economy and the reduction in greenhouse gases and reduced energy usage of these platforms. This will be enabled by the development of CMOS compatible laser integrated with Silicon and Germanium photonics and electronics. Direct benefits include: the ability to fabricate fully integrated photonic circuits on Silicon and germanium on insulator, providing a solution for limited interconnectivity in future generations of microprocessors; and enabling freestanding remote sensors combined with processing electronics. This will provide improved performance and functionality and open up a number of new applications.
The project offers an excellent training opportunity for students both in terms of technical training, research skills, experience and exposure to commercial entities such as Echerkon Technologies Ltd and IQE. We anticipate that immediate commercial exploitation will be through two main routes. First, through licensing of intellectual property and a programme of knowledge transfer to our project partners and to other industrial collaborators. Second, where substantial development is required to establish the value of an innovation, through spin-out activity. Intellectual property arising from the research will be managed with reference to an IP agreement formed at the beginning of the project.