What a bright idea

Thomas Alva Edison’s incandescent light bulb, patented in 1879, has served us well for more than a century. But, in the modern eco-conscious environment, it is fantastically wasteful – turning 98% of the electricity into heat rather than light.

Energy-saving compact fluorescent lamps (CFLs), which became available in the 1980s, are five-fold more efficient, but they depend on mercury vapour and have a limited lifetime. When they start flickering erratically and need to be replaced, they become a serious disposal problem.

Soon after the CFLs came the discovery of efficient blue LEDs, which made it possible to produce white light with lamps combining LEDs of different colours. Nobel prizewinner Shuji Nakamura, of Nichia Chemicals at Tokushima, Japan, produced the first high performance blue LED in 1994, based on InGaN materials.

White light lamps combining differently coloured LEDs use less than 10% of the electricity of an incandescent bulb with the same light output.

Around 2004, it became possible to produce white light using only blue LEDs, with the help of the luminescent material or phosphor YAG:Ce (Y₃Al₅O₁₂:Ce), which converts part of the light to a wavelength that, with the remaining blue light, creates the impression of white light. Other phosphors producing better approximations to the spectral distribution of white light have since been developed.

LEDs have already conquered many applications from small portable lights to headlights in cars, so are they going to be the perfect lamps to light our homes and offices?

LEDs work perfectly well in small battery-driven torches or bicycle lights. In fact, it is their high energy efficiency that turned bicycle lights from clunky fist-sized boxes into thumb-sized clip-on ones. For applications in bigger lamps, however, they have a few problems.

The light of a diode has a very narrow wavelength band. This contrasts with the broad black body radiation spectrum of sunlight our eyes evolved for, and which is mimicked reasonably well by traditional incandescent light bulbs. To achieve a broader spectral distribution, LED lamps need a layer of a fluorescent phosphor, which is excited by the wavelength the LEDs produce and releases the energy as a more suitable kind of light.

A recent discovery from a collaboration of several European research groups may offer a new way to make a much cheaper phosphor. Teams led by Marten Roeffaers and Johan Hofkens from the KU Leuven, Belgium, and Paolo Samori at the University of Strasbourg, France, have found a way to stabilise highly luminescent silver clusters by embedding them in a zeolite matrix. The clusters of pyramidal shape only are formed in situ when a zeolite doped with silver ions is heated. They consist of only four silver atoms, a cluster size otherwise prone to aggregating into larger particles.¹

‘This pyramid shape is what we need to obtain the best fluorescence properties. Heating up the silver ions in the zeolite framework makes them adopt this shape,’ Hofkens explains. ‘Because they are “trapped” in the cages of the zeolites, they can only form a pyramid with up to four silver atoms. That is exactly the shape and size in which the silver cluster emits the largest amount of fluorescent light, with an efficiency close to 100%.’

Zeolites are very cheap and only need minute amounts of silver to work as phosphors, so this new fluorescent material is cheaper than the existing ones that often rely on rare earth such as yttrium.

Some of the problems impeding broader utility of LEDs are simply cultural. They are completely different from the lamps in our household fixtures as they run on DC and relatively low current – they lose efficiency or ‘droop’ above a certain threshold current. To build a strong lamp, lots of diodes are needed running at small currents, along with a converter to switch from AC into DC.

To replace a conventional incandescent lamp, the limitation of the light output together with the directional focus of the light from an LED means that multiple LEDs are usually required. Further limitations also arise from the slight tendency to flicker and from the fact that LEDs emit very intense light from a small spot. Both of these characteristics can strain or damage human eyes, so mitigating such problems is a research priority.

LEDs might be in a stronger position if one were to design electrical lighting from scratch, using their properties to optimal effect. As diodes are very small and those made of organic materials are also flexible, there is no reason for the light to come from a lamp with a large clunky light bulb. LEDs could be integrated into fabrics, building materials or windows and fed by weak, off-grid power sources including small solar cells or people’s movements. In such alternative scenarios - developed, for example, for applications in remote areas - LEDs are already the supreme champions. They come across as difficult only in the context of traditional lighting technology in a fully wired household, where they have to pretend they are light bulbs.

Next generation

As incandescent light bulbs are being phased out, CFLs are developing an image problem, and LEDs and customers’ expectations are still cautiously approaching each other, there is still the possibility that some other technology will emerge to light up the homes of the future.

One promising alternative approach is the superluminescent light-emitting diode (SLED), which avoids the efficiency droop as well as other technical problems of the LED. This device is built like a laser diode – as in CD and DVD players – and uses some amplification by stimulated emission, but it deliberately lets light leak out of the chamber before lasing sets in. Invented in the 1980s, SLEDs are already used in highly specialised applications such as optical gyroscopes. They may also have a future for mass-market applications, as they appear to combine the energy-efficient and user-friendly properties of LEDs with the droop-resistance of laser diodes.

The group of Boon Ooi at the King Abdullah University of Science and Technology (KAUST) in Saudi Arabia, in collaboration with researchers at the University of California at Santa Barbara, US, including blue-LED inventor Shuji Nakamura, has recently demonstrated the fabrication of such a high-performance diode with InGaN/GaN quantum wells using metal-organic chemical vapour deposition.²

The resulting high-powered blue light can be converted to white light with conventional phosphors such as YAG, just like in a blue LED. The resulting white light is unaffected by the drooping problem found in LEDs as well as the spatial unevenness or ‘speckle’ of laser diode based lights. Therefore, the authors conclude that this novel device will be suitable both for optical communications and for general lighting purposes.

A completely different approach is based not on diodes but capacitors, with a luminescent layer and a dielectric inserted between the two electrodes, one of which needs to be transparent to let the light out. These so-called Field-Induced Polymer Electro Luminescence (FIPEL) devices come in different designs, but, as David Carroll, from Wake-Forest University at Winston-Salem, North Carolina, US, explains: ‘The defining characteristic is that the majority of current that occurs in the emitter is generated through electric polarisation. It is a polarisation current and not directly injected carriers that creates the light.’ Carroll’s group has recently published a methods paper explaining why such devices can produce very high power efficiencies.³

In two recent papers, Carroll’s group and colleagues from the Chinese Academy of Sciences at Changchun have demonstrated a novel dielectric material based on polymers processed from solution⁴and the use of the phosphor YAG:Ce with a FIPEL light source to produce white light.⁵

Could they be the next generation energy-saving lights? In the lab, their power efficiency depends very sensitively on experimental conditions. ‘The devices are resonant, meaning they must be driven at a specific [AC] frequency to make them power efficient. Being off of this resonance only slightly can cost a great deal in overall power consumption,’ Carroll explains. In contrast to the droop effect in LEDs, they get more efficient at higher power. In recent data being prepared for publication, Carroll’s group demonstrates that under optimal conditions the power efficiency of experimental FIPEL sources exceeds that of LEDs, making them potentially even better energy savers than all other lights that exist so far.

Carroll and colleagues are now working with a company hoping to commercialise their technology. Their aim is to have demo versions ready to be shown at conferences and trade fairs later in 2016, and a consumer version within the next two years. Challenges that remain include the long-term stability of the organic material and the adaptation of the power supply. While FIPEL uses AC like most household electrics, it needs a specific, much higher frequency and very low power, so the converters to provide the exact right power at low cost still have to be developed.

New light

Before Edison came up with the carbonised bamboo filament that made his light bulb commercially viable, many other approaches to electric lighting had been tried throughout the 19th century. Similarly, we are now once more in the situation where many things have to be tried, and it is far from clear which solution will establish itself as the true successor to the light bulb.

Thinking of lighting in terms of the traditional lamp plugged into the mains may be an impediment to progress. The low power needed by LEDs and laser diodes, along with the ability to pipe light through fibre optic cables, create opportunities for dramatically different solutions.

UCSB’s Steven DenBaars, who was also involved with the Saudi Arabian project on superluminescent diodes described above, has proposed a radically new concept of lighting based on laser diodes. These do not suffer from drooping and can produce much larger quantities of light than LEDs. They are typically so bright they damage the eyes.

However, DenBaars envisaged a central light source based on laser diodes, from which small portions of light could be distributed through light guides or even through space to wherever it is needed.⁶

Far from replacing the old light bulbs, DenBaars suggests replacing the entire fittings and wiring with something more ethereal based purely on light. Now that sounds more like 21st century lighting technology.