Ebook: Silicon-Based Microphotonics: from Basics to Applications

The Enrico Fermi School on “Si-Based Microphotonics: from Basics to Applications” was a 2-week school which occurred on the second half of July 1998. It brought together 15 lecturers and 50 students from all over the world. All participants were engaged in a common goal: that of understanding and of sharing their ideas to develop an Si-based optoelectronics.

Introductory and more specialistic lectures were presented. Different Si-based luminescing materials were discussed, among them, porous Silicon, rare-earth-doped silicon, Si nanocrystals, Silicides, Si-based multilayers and silicon/germanium alloy or superlattice structures. The different devices needed for an all-Si-based optoelectronics were treated, spanning from light sources to waveguides, from amplifiers and modulators to detectors. During the School both the very basic theoretical treatments as well as applications to real prototype devices and integration in an optical integrated circuit were presented. The enormous progresses made in the last few years in this fast developing field were so evident that further breakthroughs are predicted for the near future. This was particularly clear since many presentations gave a historical overview on how improvements proceeded in this last period and the gap between the initial stages and the latest results was evident.

Among the discussed new results, we can mention the stable and efficient emission in Si-rich silicon dioxide devices, the 10 MHz room temperature modulation in Er-doped light-emitting diodes, the narrow and directional emission in porous Si microcavities, the very intense enhanced emission in rare-earth-doped nanocrystals, the gain in Erdoped waveguide amplifiers, etc. An Si-based laser is not out of sight. Many still unresolved problems have been also underlined: the problem of electrical transport in lowdimensional Si systems, the possibility of gain in Si-based systems, the low modulation speed of Si-based LEDs, etc.

During the School attendees had also the opportunity to present results in specially devoted sessions and they could discuss them with the lecturers. The economics and impact of this field was also taken into account in a special presentation by P. Malinverni of the European Commission, where the point of view of the European Community was given.

The atmosphere of the School was one of enthusiasm. This was also true thanks to the marvellous location in Villa Monastero and to the kindness ;1nd efficiency of the Italian Physical Society staff. In particular we would like to thank E. Mazzi and all of the staff in the secretariat. The generous support from CNR, the ESPRIT projects SMILE and SCOOP are gratefully acknowledged. These Proceedings collect the lectures reported during the School. They give a state-of-the-art picture of Si-based optoelectronics and will represent an important tool for all those researchers intending to work in this fascinating field.

***

Silicon is the leading semiconductor in the microelectronics industry. In fact, due to its mature technology and to the continuous improvements in the scale integration, this semiconductor is able to satisfy the increasing demand for higher complexity integrated circuits. There is no doubt that the silicon technology will dominate the semiconductor market for at least two more decades. At the same time optoelectronics, especially optocommunication, has entered a long term growth phase. Due to its indirect bandgap and to the absence of linear electro-optic effects, Si has been considered unsuitable for optoelectronic applications which remain the domain of III-V semiconductors and glass fibers. Several applications, such as optical interconnections at the chip-to-chip level, require integration of electrical and optical functions on the same device. One way to approach these requirements is to combine the optical properties of III-V's with the electronic performances of crystalline silicon. This hybrid integration is realized by attaching distinct optical components onto Si electronic circuits through a solder bonding technique. This requires a precise alignment between the different components and results in high packaging costs. Alternatively, monolithic growth of III-V semiconductor optical devices on the Si circuit has been explored. This approach guarantees a better positioning accuracy and potentially has a small additional cost. However, the problems associated with the compatibility between the III-V optical devices and the VLSI processing are still unsolved. Finally, it should be noted that the manufacturing of these optoelectronic devices requires a simultaneous know how of processing of both Si and III-V semiconductors.

Therefore, any successful step to realize discrete and integrated optoelectronic functions directly in silicon has very favorable economic perspectives and industrial relevance. This would allow to use the low cost and mature VLSI technology to fabricate optoelectronic devices and would attract the attention of Si device industries.

The achievement of integrated optoelectronics in Si requires fabrication and optimization of the single optical components which comprise light sources (LED and/or lasers), devices for signal handling (waveguides, modulators, amplifiers, etc.) and detectors. Many new experimental and theoretical contributions recently came from academia, research laboratories and industry, demonstrating that Si-based optoelectronics is an active rapidly developing field where inputs from both materials scientists as well as engineers are needed to solve the physical problems and to transform good ideas into real working devices.

The main limiting step towards an Si-based optoelectronics has been the absence of efficient light sources. Recently a strong effort has been devoted to study all those processes able to circumvent the physical inability of silicon to emit efficiently light. Since the discovery of light emission from porous silicon made in 1990 by Leigh Canham, a lot of work has been devoted in studying silicon nanostructures. These comprehend not only porous silicon but also nanocrystals produced by several techniques, as well as siliconinsulator multilayers. The initial problems related to the instability of the luminescence yield have finally been solved and today reliable, stable structures, compatible with the silicon technology have been fabricated. In particular, a group at the Rochester University has now produced silicon-rich silicon dioxide electroluminescent devices integrated with silicon microelectronic circuitry. Alternative approaches comprehend the doping of silicon with rare earths. In this case the luminescence is due to an internal 41 shell transition of the rare earth ion excited through electron-hole recombinations within the silicon matrix. The initial problems related to erbium incorporation and luminescence quenching have been now understood and can be maintained under control. This led to the fabrication of Er:Si devices operating at room temperature, with efficiencies of 0.1 and modulation speed of 10 MHz. The last approach to appear in the scientific arena is that of iron disilicide. In its beta phase, this silicide is semicoducting with a direct bandgap and its capabilities in competing with other approaches are under investigation.

Also other elements to build an Si-based optoelectronics are of major importance. Waveguides are now built directly within silicon with low losses. Moreover waveguide amplifiers in several materials compatible with silicon processing have now been fabricated and present net optical gain. For instance, Er doping of several waveguide materials such as Al₂O₃, sodalime silica glasses and polymers holds high potentialities for an Si-based optoelectronics since these materials with excellent optical properties can be deposited directly on Si. Also in the case of silicon processing compatible modulators and detectors the progresses are enormous. For instance, fast and efficient detectors in the visible region have been fabricated by metal-semiconductor-metal structures using a buried epitaxial cobalt disilicide layer and detectors in the infrared region are not out of sight by using epitaxial SiGe layers.

Indeed the evolution in Si-based optoelectronics has been extremely fast in these last few years and it is predicted that this growth will still continue in the near future. The integration of Si-based optical functions with microelectronic circuitry is no longer a dream but reality.

The present book is a collection of reviews on the different aspects of Si-based optoelectronics written by some of the major experts in the field. We think it gives a fascinating picture of the state-of-the-art in Si microphotonics and a perspective on what we can expect in the near future. For these reasons we hope this book might be useful not only to graduate students but also to all those researchers involved in this field.

O. Bisi, S.U. Campisano, L. Pavesi and F. Priolo