Abstract So far, the semiconductor nanowire research area has mainly delivered results on growth procedures and related material properties. As the development lately has been successful in producing novel nanowire-based structures for optical or electronic applications, the time is ripe to review the device work that has been done and in some cases has produced devices ready for the market. In this chapter, we shall review the specific area of nanowire-based LEDs (NW-LEDs) for visible light, including the application area of “solid state lighting” (SSL). A brief review of the progress in the area of visible light LEDs over the last half century is presented, this also mentions some of the progress made in the planar technology so far. The most successful way of producing white light is still based on the use of phosphors, just like in the present compact fluorescence lamps (CFLs). The reason for this is the high efficiency (external quantum efficiency u003e 80%) possible at low currents in the violet planar InGaN-based LEDs used to excite the phosphors. These LEDs are presently mainly produced on foreign substrates, leading to a high dislocation density (DD) and a sizeable droop at high injection currents (25–40%). This droop and the down conversion energy loss in the phosphors (20–25%) has motivated the interest for a phosphor-less white light source based on direct mixing of light of different wavelength (such as red, green, and blue; RGB). To be competitive, this solution must be based on highly efficient LEDs for all RGB (red, green, and blue) colors. Since NW-LED structures can be produced basically free of structural defects (even if grown on a foreign substrate), the idea of using the RGB mixing concept for the production of white light sources with an ultimately higher efficiency than for the phosphor-based lamps is a major technical target for a second generation of light sources in the SSL field. Basic concepts behind the design and optical properties of NW-LED structures are discussed in this chapter, with emphasis on the present developments of III-nitride-based structures. The growth procedure relevant for such NW-LED structures is reported in some detail, specifically the core–shell configuration readily produced with metalorganic vapor phase epitaxy (MOVPE). The first generation processing technology for NW-LED structures is briefly described; this is naturally quite different from the established routines for planar LED chips. Experimental data for nitride-based NW-LEDs for blue, green, and even longer wavelengths are given in terms of radiative efficiencies, light outcoupling, droop, and long-term reliability. The experience so far is that for these NW-based emitters, efficiencies can be obtained that are close to those for the corresponding planar LEDs. There are still problems with the reproducibility of the radiative output, as well as a significant droop that would not be expected for m -plane emitters. More work is needed to pinpoint the cause of these problems. Finally, we briefly discuss various applications (also other than white lamps) where the NW-LEDs may have a specific advantage.

In 2014, the Nobel Prize in physics was awarded for the discovery of the gallium-nitride-based blue-light LED and its use for new, efficient LED-based white lamps. The development efforts in this field, however, do not stop there. The active layers in this type of diode take up less than 1 m. This means that most of the material in planar LEDs now on the market, which are hundreds of micrometers thick, is not used. This provides a motivation to continue research on efficient LEDs with nanoscale dimensions.1 In the last few years, it has been shown that an efficient LED can be made as a 3D nanostructure, in which the active parts take up less than 1 m. This makes more efficient use of the materials involved and, at the same time, excludes structural defects that limit the LED performance from the active device volume. Nanometer-sized LED structures can take many different shapes, such as thin elongated wires, pyramids, cubes, or platelets. To grow each of these small objects, different sections— including contact layers and an active region where the light is produced—are used. Basic material properties, however, put constraints on what can be grown, which affects the choice of shape for the nano-LED. For visible-light LEDs, the III nitrides— gallium nitride (GaN), aluminum nitride, indium nitride, and their alloys—are very suitable since they offer bandgaps in the visible range of photon energies. We have recently summarized the current status of the development of nanowire-based LEDs.1 These devices are mainly shaped like nanorods with hexagonal cross sections, where the different layers are grown in a concentric manner around a core section (see Figure 1). The active region that produces the light consists of indium gallium nitride (InGaN) quantum wells (QWs) on m-plane facets, where ’m’ refers to a specific crystallography orientation. Such QWs do not suffer from any polarization fields since the m-plane is non-polar. All other layers in the structure—the core, underlayer, QW barrier layers, p-layer and p-contact layer—are grown from GaN. A disadvantage with this design is that emission wavelengths longer than green are Figure 1. Top left: Electroluminescence from the first blue nanowire LED fabricated by Glo AB in April 2007. Bottom left: Blue, green, yellow, and red examples from current state-of-the-art nanowire LEDs, grown by Glo-USA Inc. in 2013. Center: Side-view of a scanning electron microscope (SEM) image, showing nanowire LEDs monolithically grown on a wafer of gallium nitride on silicon. The edge of the top contact layer is visible, and both the contacted and non-contacted LEDs can be seen. Right: SEM image illustrating the air-bridge process. Images provided courtesy of QuNano AB and Glo AB.1 p-GaN: p-Type gallium nitride. p-AlGaN: p-Type aluminum gallium nitride.

III-nitride nanowire LEDs grown by MOCVD are demonstrated with external quantum efficiency exceeding 10% at 20 A/cm2 with a peak wavelength of 520 nm. Advantages of this device structure over planar LEDs will be described. Article not available.

We study mechanisms which are thought to contribute to efficiency droop in III-nitrides. We first observe droop in a photoluminescence (PL) experiment on bulk GaN, which confirms the existence of a bulk contribution to droop, unrelated to piezoelectric fields or alloy fluctuations. We then perform biased-PL on a series of InGaN light-emitting diodes to estimate the potential impact of carrier leakage on PL experiments. We conclude that carrier leakage is only significant at very low pump densities and does not contribute to droop, thus validating the use of PL to characterize droop.