Silicon has long been established as the material of choice for the microelectronics industry. This is not yet true in photonics, where the limited degrees of freedom in material design combined with the indirect bandgap are a major constraint. Recent developments, especially those enabled by nanoscale engineering of the electronic and photonic properties, are starting to change the picture, and some silicon nanostructures now approach or even exceed the performance of equivalent direct-bandgap materials. Focusing on two application areas, namely communications and photovoltaics, we review recent progress in silicon nanocrystals, nanowires and photonic crystals as key examples of functional nanostructures. We assess the state of the art in each field and highlight the challenges that need to be overcome to make silicon a truly high-performing photonic material.

Colloidal semiconductor quantum dots (QDs) constitute a perfect material for ink-jet printable large area displays, photovoltaics, light-emitting diode, bio-imaging luminescent markers and many other applications. For this purpose, efficient light emission/ absorption and spectral tunability are necessary conditions. These are currently fulfilled by the direct bandgap materials. Si-QDs could offer the solution to major hurdles posed by these materials, namely, toxicity (e.g., Cd-, Pb-or As-based QDs), scarcity (e.g., QD with In, Se, Te) and/or instability. Here we show that by combining quantum confinement with dedicated surface engineering, the biggest drawback of Si-the indirect bandgap nature-can be overcome, and a 'direct bandgap' variety of Si-QDs is created. We demonstrate this transformation on chemically synthesized Si-QDs using state-of-the-art optical spectroscopy and theoretical modelling. The carbon surface termination gives rise to drastic modification in electron and hole wavefunctions and radiative transitions between the lowest excited states of electron and hole attain 'direct bandgap-like' (phonon-less) character. This results in efficient fast emission, tunable within the visible spectral range by QD size. These findings are fully justified within a tight-binding theoretical model. When the C surface termination is replaced by oxygen, the emission is converted into the well-known red luminescence, with microsecond decay and limited spectral tunability. In that way, the 'direct bandgap' Si-QDs convert into the 'traditional' indirect bandgap form, thoroughly investigated in the past.

Crystalline silicon is the most important semiconductor material in the electronics industry. However, silicon has poor optical properties because of its indirect bandgap, which prevents the efficient emission and absorption of light. The energy structure of silicon can be manipulated through quantum confinement effects, and the excitonic emission from silicon nanocrystals increases in intensity and shifts to shorter wavelengths (a blueshift) as the size of the nanocrystals is reduced. Here we report experimental evidence for a short-lived visible band in the photoluminescence spectrum of silicon nanocrystals that increases in intensity and shifts to longer wavelengths (a redshift) with smaller nanocrystal sizes. This higher intensity indicates an increased quantum efficiency, which for 2.5-nm-diameter nanocrystals is enhanced by three orders of magnitude compared to bulk silicon. We assign this band to the radiative recombination of non-equilibrium electron-hole pairs in a process that does not involve phonons.

No abstract

Carrier multiplication by generation of two or more electron-hole pairs following the absorption of a single photon may lead to improved photovoltaic efficiencies and has been observed in nanocrystals made from a variety of semiconductors, including silicon. However, with few exceptions, these reports have been based on indirect ultrafast techniques. Here, we present evidence of carrier multiplication in closely spaced silicon nanocrystals contained in a silicon dioxide matrix by measuring enhanced photoluminescence quantum yield. As the photon energy increases, the quantum yield is expected to remain constant, or to decrease as a result of new trapping and recombination channels being activated. Instead, we observe a step-like increase in quantum yield for larger photon energies that is characteristic of carrier multiplication. Modelling suggests that carrier multiplication is occurring with high efficiency and close to the energy conservation limit.