Light revolution

Isamu Akasaki, Hiroshi Amano and Shuji Nakamura jointly win the Nobel Prize in Physics for developing the world’s first blue LED, which has unleashed a revolution in lighting up the world.

Shuji Nakamura demonstrating a blue LED light in California on October 7, the day the Prize was announced.

“ LED there be light .”

THE above could well be the words of Isamu Akasaki, Hiroshi Amano and Shuji Nakamura, whose path-breaking invention of blue light-emitting diodes (LEDs) enabled the creation of white light in a new way and opened the way for the development of a new energy-efficient and environment-friendly light source. They succeeded where several others before them had tried and failed and scientists around the world had almost given up hope of success.

In the early 1990s, when this trio of Japanese scientists succeeded in producing bright blue light from gallium nitride (GaN)-based semiconductors, it brought about a revolution in lighting technology. LEDs that produced green and red light had been around for several years, but LEDs that could produce the remaining primary colour, blue, had posed a serious technological challenge for nearly three decades despite the considerable efforts of both the scientific community and industry. Only a combination of the three primary colours—red, green and blue—can result in the white light needed for lighting purposes.

The press release of the Royal Swedish Academy of Sciences noted: “Incandescent light bulbs lit the 20th century; the 21st century will be lit by LED lamps.” This transformation is already evident, and LED-based lighting systems are now ubiquitous. The older incandescent and fluorescent light sources are fast giving way to the LED-based white light sources in homes, shops and offices though cost still remains a major factor, and LEDs already form the basis for most of the electronic devices we use—mobile phones, tablets, laptops, computer monitors, television sets, car dashboards and luminous advertising billboards—in which they are the backlighting sources that shine light on their liquid crystal display (LCD) systems.

“In the spirit of Alfred Nobel”, the citation for the award said, the three have been awarded the 2014 Nobel Prize in Physics for “an invention of greatest benefit to mankind”. The total prize money of Swedish kronor (SEK) eight million ($1.12 million) will be shared equally by the three laureates. While Akasaki had worked with his PhD student Amano at Nagoya University to develop the device, Nakamura had worked alone at Nichia Chemicals, a small company located in Tokushima on the island of Shikoku. Today, Akasaki, 85, is a professor at Meijo University, Nagoya, and a distinguished professor at Nagoya University. Amano, 54, is also a professor at Nagoya University. Nakamura, 60, on the other hand, is an American citizen now and currently a professor at the University of California Santa Barbara.

An LED is basically a semiconductor light source that consists of a number of layered semiconductor materials. In an LED, electricity is directly converted into light quanta (photons) by the phenomenon of electroluminescence. (Electroluminescence is, in fact, generally observed only in semiconductor materials.) This results in increased efficiency of the light generation compared with other light sources, where most of the electrical input gets converted into heat and only a small amount into light. In incandescent light bulbs and halogen lamps, electric current heats up a filament to a degree that it begins to glow and emit light. In fluorescent lamps (including compact fluorescent lamps, or CFLs), the current produces a gas discharge that results in both heat and light. Fluorescent lamps also contain the undesirable toxic element mercury whose atoms get excited by the discharge and emit light.

LEDs require less power to emit light than familiar light sources. They are also constantly being improved to obtain greater efficiency. Efficiency for lamps is measured in terms of luminous flux (measured in units called lumens) per watt of electrical power input. R&D efforts in this direction have achieved a record output of 300 lumens/watt (which represents about 50 per cent efficiency in converting electricity into light) for LEDs compared with 16 lm/W for ordinary incandescent bulbs and 70 lm/W for fluorescent lamps (Figure 1). Also, while ordinary light bulbs can last for about 1,000 hours and fluorescent lamps for about 10,000 hours, LEDs can last for as long as 100,000 hours. As the press release points out, about one-fourth of world’s electricity consumption is for lighting, so energy-efficient LEDs will contribute greatly towards saving the world’s energy and material resources.

Electroluminescence was first observed in 1907 in the semiconductor material silicon carbide (SiC), which is also known as carborundum, by Henry J. Round, an associate of the Italian inventor Guglielmo Marconi. At low voltages, the material glowed yellow and changed colour as the applied voltage was increased. The phenomenon was studied in greater detail by the Soviet device physicist Oleg Losev (1903-42), who published several articles in reputed journals on light emission from SiC. But all this happened when the modern theory of electronic structure of solid state materials based on quantum theory was yet to be formulated.

An LED is basically what is known as a p-n junction in the parlance of electronics. A p-n junction forms the elementary “building block” of most electronic devices such as diodes, transistors, solar cells and integrated circuits. They are the active sites where the quantum electronic action of the device happens. For example, the common transistor is a “bipolar junction” device consisting of two p-n junctions in series, in the form n-p-n or p-n-p.

A p-n junction is the boundary, or interface, between two types of semiconducting materials, known as p-type and n-type, in a single crystal of the semiconductor. The n-type layer has more negative electrons than what the available energy states that follow from quantum theory would actually allow, and the p-type layer has an insufficient number of electrons, which is also referred to as a layer with excess positive holes. The p-type and n-type semiconductors are produced by “doping” the basic semiconductor material, such as silicon (Si) or germanium (Ge), with appropriate atomic impurities to increase the concentration of the charge carriers (electrons or holes), the type being determined by what impurity is added.

The physics of semiconductors and p-n junctions was understood in the 1940s and led to the invention of the basic transistor device by John Bardeen, Walter Houser Brattain and William Shockley in 1947. In 1951, Kurt Lehovec and associates were able to explain that electroluminescence occurred in a p-n junction because of a process called radiative recombination (Figure 2). When the junction is biased with external forward voltage (positive battery terminal connected to the p-layer and negative terminal connected to the n-layer), the charge carriers in the material—the excess negative electrons from the n-type layer and the positive holes from the p-type layer—are driven to combine at the interface layer to give out energy in the form of light.

The basic concept underlying the physics of semiconductors is that of “bandgap”, which is the energy range in a solid where no electron states can exist. It is the energy difference (in electronvolts, or eV) between the top of the valence band (the highest energy which the electrons bound in atoms can have at the temperature of absolute zero) and the bottom of the conduction band (the minimum energy required to free an electron that is bound to an atom and enable it to move about freely and conduct electricity). The energy of the photon emitted through radiative recombination of electrons and holes across a p-n junction should thus be equal to the bandgap value.

However, the photon energy observed by Lehovec’s group was less than the energy gap of SiC. For efficient recombination, it is important that the semiconductor material chosen has what is called a direct bandgap. Recombination in materials with indirect bandgaps occurs via an intermediate state that involves the vibrational energy of the material lattice as well, called phonon-assisted recombination, which results in some energy loss and thus limits the device efficiency. The quantum efficiency of an LED is the ratio of the number of photons emitted to the number of electrons passing through the interface at a given time.

In the 1950s, many p-n junction diodes were developed which demonstrated that electroluminescence was due to the injection of charge carriers. These included semiconductors from Group IVA (of the periodic table), such as Si and Ge, and binary Group III-V compounds, which are obtained by combining Group III elements, such as aluminium (Al), gallium (Ga) and indium (In), with Group V elements, such as nitrogen (N), phosphorus (P), arsenic (As) and antimony (Sb). The most important such compounds are gallium arsenide (GaAs), indium phosphide (InP), gallium phosphide (GaP) and GaN.

In the late 1950s, techniques to make efficient p-n junction diodes with GaAs were being pursued seriously. GaAs, being a direct bandgap semiconductor material, enabled the recombination of electrons and holes without the involvement of the lattice. Its bandgap is 1.4 eV, which corresponds to light in the infrared region of the electromagnetic spectrum, and in 1962, infrared light emission from p-n junction was achieved, resulting in the development of an infrared LED.

Around the same time, different researchers were making use of another III-V semiconductor, GaP—which, however, is only an indirect bandgap material—to develop LEDs that emitted light of different wavelengths ranging from red to green. These developments resulted in many companies manufacturing red and green LEDs based on GaP. LEDs of different wavelengths around red colour were also developed using mixed crystals of Ga, As and P (with different proportions of As and P) to reduce the wavelength from the infrared of GaAs to reach the visible range. These began to be used, for example, in digital watches and calculators or as on/off indicators in different appliances.

But it was clear that an LED with the emission of short-wavelength photons, a blue LED, was needed to produce white light. Many laboratories were on the task but did not succeed. It was known that GaN, with its direct bandgap of 3.4 eV corresponding to ultraviolet (UV) wavelengths, was the key material to produce a blue LED, and already in the late 1950s, Philips Research Laboratories was exploring the possibility of a new lighting technology based on GaN.

GaN is a III-V semiconductor material whose crystals can be grown on a substrate of aluminium oxide (Al ₂ O ₃ ) and SiC; GaN can be doped with silicon to make it n-type or with magnesium (Mg) and zinc (Zn) to make it p-type but doping was found to disrupt the crystal growth process, resulting in a fragile material. Only very small powder-like GaN crystals could be grown, which were not good enough for making p-n junctions. So researchers at Philips turned their focus back on GaP.

There was improvement in the growth of GaN crystals with the development in the 1960s of a more efficient technique called Hybrid Vapour Phase Epitaxy (HVPE), which laboratories in the United States, Europe and Japan began to use with the aim of developing blue LEDs with GaN, but material problems persisted and were extremely difficult to overcome. An appropriate substrate to produce a good growth of GaN crystals could not be found. More importantly, it seemed virtually impossible to produce p-type layers by doping the material. Thus, the major problems that needed to be overcome were producing deformation-free crystals and achieving high concentration p-doping.

The newer crystal growth techniques of Molecular Beam Epitaxy (MBE) and Metalorganic Vapour Phase Epitaxy (MOVPE) came to be developed in the 1970s, and efforts were made to use them to grow GaN. Akasaki had been studying the material GaN as early as 1974 while working at Matsushita Research Institute in Tokyo. In 1981, after he became a professor at Nagoya University and took Amano under his wing, he continued his research on GaN with Amano.

In 1986, Akasaki and Amano were the first to succeed in producing high-quality GaN crystals. Their breakthrough was the result of a series of experiments which began with a hand-operated MOVPE device. Not seeing any significant results in his first attempts, Akasaki went back to the “buffer layer technique” developed during his Matsushita days. Their method involved first nucleating a thin layer (30 nanometres) of polycrystalline aluminium nitride (AlN) on a sapphire substrate at a low temperature (500 °C) and growing GaN crystals on it by heating up the sapphire-AlN substrate to the growth temperature of GaN (1,000 °C). The major problem that remained, however, was producing p-type GaN in a controlled manner.

In the late 1980s, Akasaki, Amano and their associates observed that when Zn-doped GaN was studied with a scanning electron microscope it emitted more light. This suggested that the electron beam from the microscope was making the p-type layer more efficient. Similarly, they found that when Mg-doped GaN was irradiated with electrons, it resulted in better p-doping properties. This was an important breakthrough, and in 1992, they were able to make the first LED that emitted blue light.

Nakamura had started to work on blue LEDs only in 1988, much later than Akasaki and Amano, and GaN was his material of choice too, but he adopted a slightly different approach. Instead of growing GaN on a thin layer of AlN, he first grew a thin layer of GaN at a low temperature and subsequent layers of GaN at a higher temperature. He too succeeded in developing a blue LED based on GaN, two years after Akasaki and Amano.

Nakamura could also explain why Akasaki and Amano had succeeded in developing the p-type layer: dopants such as Zn and Mg formed complexes with hydrogen in the material and became passive. The electron beam dissociated these complexes and activated the doping. Nakamura also showed that even simple thermal treatment (annealing) led to the effective activation of p-doping with Mg.

It was known from earlier developments with infrared LEDs and laser diodes that layered semiconductor structures such as heterojunctions (interfaces between dissimilar semiconductor crystals) and quantum wells (thin semiconductor layers in which charge carriers are restricted by quantum effects to move in two dimensions only instead of three) were key to achieving high efficiency. In these, electrons and holes are restricted to a small volume so that recombination can occur more efficiently and with minimal losses. While Akasaki and company developed structures based on AlGaN/GaN, Nakamura exploited combinations of InGaN/GaN and InGaN/AlGaN and succeeded in developing heterojunctions, quantum wells and multiple quantum wells (Figure 3). In 1994, Nakamura succeeded in achieving a quantum efficiency of 2.7 per cent using an InGaN/AlGaN-based double heterojunction.

With the achievements of both these groups, the gates were opened for the development of efficient blue LEDs and their application on a commercial scale. The teams have continued to work on blue LEDs, aiming for higher efficiencies, increased versatility and wider applications. In 1995-96, both groups succeeded in inventing the blue laser in which the blue LED, of the size of a grain of sand, forms a crucial component. While a blue LED only produces dispersed blue light, a blue laser emits a sharp focussed beam. Since blue light has a very short wavelength, it can be packed much tighter than infrared light and therefore can store four times more information. This increased storage capacity of blue (and UV) lasers has resulted in the development of “blu-ray” discs, which can store high-definition video and have longer playback times, and better laser printers.

White light LEDs for lighting can be produced in two different ways. The way that is currently employed involves using a blue LED to excite a phosphor-coated lamp so that it shines in red and green. When all three colours combine, white light is produced. The other way, which is somewhat in the future, is to construct a lamp out of three LEDs (red, green and blue) and let the eye do the work of perceiving it as white light. This technology will also result in flexible light sources with dynamic colour tuning and control by computers so that colours and intensity can be varied as needed over a vast range.

Today’s efficient GaN-based LEDs are the result of a long series of breakthroughs, particularly by this year’s Nobel laureates, in materials physics and crystal growth, in device physics with advanced heterostructure designs and in optical physics to optimise the light output. Future applications of the Nobel Prize-winning invention may include the use of UV-emitting AlGaN/GaN-based LEDs for water purification. LED-based lighting also holds great promise for bringing power to the 1.5 billion poor of the world who lack access to grid power as the LED’s low power requirement can be met through cheap local or even rooftop solar power. Invented barely 20 years ago, LEDs are already unleashing a revolution in lighting up the world.