Special issue on silicon photonics

The technology now known as silicon photonics can be traced back to the pioneering work of Soref in the mid-1980s (see, for example, Soref R A and Lorenzo J P 1985 Electron. Lett. 21 953). However, the nature of the research conducted today, whilst it builds upon that early work, is unrecognizable in terms of technology metrics such as device efficiency, device data rate and device dimensions, and even in targeted applications areas. Today silicon photonics is still evolving, and is enjoying a period of unprecedented attention in terms of research focus. This has resulted in orders-of-magnitude improvement in device performance over the last few years to levels many thought were impossible. However, despite the existence of the research field for more than two decades, silicon is still regarded as a 'new' optical material, one that is being manipulated and modified to satisfy the requirements of a range of applications. This is somewhat ironic since silicon is one of the best known and most thoroughly studied materials, thanks to the electronics industry that has made silicon its material of choice. The principal reasons for the lack of study of this 'late developer' are that (i) silicon is an indirect bandgap material and (ii) it does not exhibit a linear electro-optic (Pockels) effect. The former condition means that it is difficult to make a laser in silicon based on the intrinsic performance of the material, and consequently, in recent years, researchers have attempted to modify the material to artificially engineer the conditions for lasing to be viable (see, for example, the review text, Jalali B et al 2008 Silicon Lasers in Silicon Photonics: The State of the Art ed G T Reed (New York: Wiley)). The latter condition means that optical modulators are intrinsically less efficient in silicon than in some other materials, particularly when targeting the popular telecommunications wavelengths around 1.55 μm. Therefore researchers have sought alternative mechanisms for modulation in silicon that have yielded increasingly impressive results (see, for example, Liao L et al 2007 Electron. Lett. 43 issue 22). The convergence of computing and communications and the resultant demand for increased bandwidth has been one of the factors influencing the upsurge of interest in silicon, together with the requirement for photonic and electronic integration, all to be realized at low cost. Thus emerging applications such as short-reach communications links for optical interconnect and fibre to the home (FTTH) (as well as a multitude of other applications) are frequently offered as examples of where silicon photonics will have a significant, perhaps a revolutionary, impact. One of the major conclusions of the joint MIT–industry Communication Technology Roadmap (http://mph-roadmap.mit.edu/index.php), was that 'Photonics technology will be driven by electronic–photonic synergy and short (<1 km) reach interconnection. This direction will ignite a major shift in leadership of the optical component industry from information transmission (telecom) to information processing (computing imaging).' Thus the case is made for low-cost implementation, making silicon a prime candidate, particularly if true electronic/photonic integration is to be realized. Despite the limitations of silicon as an optical material, the intrinsic advantages of the most popular silicon optical platform, silicon-on-insulator (SOI), should not be overlooked. The very high confinement nature of this technology platform brings a host of advantages, including the possibility to miniaturize devices and circuits, to reduce power consumption, optical loss and cost, to increase yield, and to be compatible with CMOS-based intelligence. Thus the limitations of silicon as an optical material can be offset against the very significant advantages, to both commercial as well as technological success. Of course, there is still much to do, hence the increasing global investment in silicon technology and the massive increase in research activity in silicon photonics since the early work in the 1980s. Only time will tell if silicon can realize its potential to satisfy the ever-increasing array of applications. However, the indications are positive, and the contributors to this cause employ increasingly impressive levels of intellectual and technological capability to realize the desired goals. It is an interesting time to be involved in slicon photonics, and it will be equally fascinating to watch the evolution of the technology in the future. Whatever happens, silicon will make the transition from being regarded as purely an electronic material to recognition as an optoelectronic material. The evidence for this is represented in the collection of papers that form this special issue of Semiconductor Science and Technology. This special issue is, in turn, representative of the rapidly increasing body of literature that represents the field of silicon photonics. In a field of ...