Nakamura develops high-brightness blue LED whole process (3)

Nikkei BP Society special report

Nakamura develops high-brightness blue LED whole process (2)

Nakamura develops high-brightness blue LED whole process (1)

   

Although GaN light emitting diodes emit light, the light is quite dark. Nakamura decided to change from a pn junction structure to a double Hetero structure. After that, the development speed is rapidly increasing. The development was also smooth, and impurities were incorporated into the double heterostructure to form a luminescent center, achieving a brightness of 1 cd and finally being put into production. Applications such as signal lights have also begun to appear.

 

 

Finally came to the final adjustment stage before the production. This stage is mainly to increase the crystallographic grade of the GaN film and the InGaN film to improve the brightness. At the same time, improve mass production technology and improve the yield. Nakamura cut off all external contacts and locked himself in the lab.

 

After such efforts, in October 1993, less than one year after the successful trial production of InGaN double-heterostructure light-emitting diodes, the conditions for productization were basically met. The brightness reached 1 cd. It was about 100 times that of commercially available blue light-emitting diodes using SiC.

 

Figure 3: High-brightness blue light-emitting diodes with a production of 1 cd . Nakamura displays a display panel made of the blue light-emitting diode developed this time .

 

Breaking the silence of nearly a year, the blue LED finally came to the fore (Figure 3). The product release date is scheduled for November 30th. Prior to this, Nakamura and his boss took a proud blue light-emitting diode and visited a major university and research institution in Japan. Among them, the greatest blessing to this success is the president of Nishizawa Ryuichi of Tohoku University in Japan (Figure 4). On the spot, President Nishizawa proposed to give the doctor a title to Dr. 4).

 

Note 4) At that time, Nakamura had applied for the title of doctoral degree to his alma mater, Tokushima University, Japan. Nakamura said that he had to cut off the good intentions of President Nishizawa.

 

As a result, Nakamura gained confidence in the results of the chest release. Finally arrived the day before the product release. Although it is a product release, it does not rent a first-class hotel like a big manufacturer, and holds a grand press conference. Instead, the contents of the results were disclosed to the director of the Japan Economic News Agency. Nakamura said that because it was a shocking result of "the brightness reached 100 times that of the previous products," the director did not believe it at first. Although a small press conference was held with only internal staff, the news was published on the first edition of Nikkei Industry News on November 30.

 

　　 There are still many goals for the next goal.

 

Since this day, Nichia Chemical Industry Co., Ltd., interview requests from the media, and consultations from users and other companies in the same industry have swarmed. You can receive 40 to 50 calls per day. This situation lasted for more than a week. "Is such an amazing achievement?" The president also panicked. After the phone tide, it was another wave of visits. With various proposals such as technical cooperation and funding, people visiting the chemical industry in Nichia are in constant stream.

 

Figure 4: Appreciated by President Nishizawa Junichi, after seeing the high-brightness blue LED, President Nishizawa of Tohoku University of Japan immediately wrote an inscription.

 

The president rejected these proposals one by one. Nakamura’s belief that “people are not for people, I am for it” has finally achieved results that transcend large enterprises. The president also adheres to his belief that he is "not dependent on others." Previously, the company has been relying on its own strength for research and development. The president decided to continue to be self-reliant in the future.

 

After throwing those cockroaches into the brain, Nakamura continued to conduct research. Although blue light-emitting diodes have produced products, semiconductor lasers as research and development targets have not yet been completed.

 

In addition, after the completion of the blue LED, consumer demand for green and blue-green LEDs has become more intense. If you combine the completed blue LED with the already-productive red and green LEDs, you can create a full-color display (Figure 5). However, the brightness of the green diode is lower than the high brightness of the blue and red diodes. A green LED with higher brightness is required.

 

Figure 5: Large full-color display with LEDs available

The photo was developed by Kinki Nippon Railway and was installed in the central hall of the Uehonmachi Train Station in Japan in January 1995. The number of pixels is 320 × 240.

In addition, red, yellow and blue three-color light-emitting diodes have been manufactured so far, and the door to the signal light is thus opened. However, the color of the Japanese green signal light is blue-green. Therefore, it is also necessary to develop a light-emitting diode that matches this.

 

Strive to achieve inter-band illumination

 

Nakamura first completed the blue-green LED. At present, the signal lamp using such a light-emitting diode has been available and has been put to practical use (Fig. 6). The rest are lasers and high-brightness green LEDs.

 

Figure 6: Blue-green LEDs used on signal lights

At that time, experimental setup was carried out in Aichi Prefecture and Tokushima Prefecture, Japan. A signal lamp using a light-emitting diode can prevent misidentification because it is colorless when it is not illuminated.

There are two challenges in creating a blue semiconductor laser using GaN materials. The first problem is the need to achieve the inter-band illumination required for semiconductor lasers 5). In the case of sandwiching the energy gap of InGaN, blue can be obtained by band-to-band illumination. By further clamping the energy gap, green light-emitting diodes and semiconductor lasers can also be realized (6).

 

Note 5) Light-emitting diodes produced today emit ultraviolet light if they emit light between the bands. Therefore, blue light emission has been achieved by introducing the light-emitting center into the light-emitting layer. However, semiconductor lasers cannot be fabricated by this method. However, if the In concentration of the light-emitting layer InGaN can be increased and the energy gap of InGaN is sandwiched, blue can be obtained by band-to-band illumination. If the In concentration can be further increased, the energy gap can be clamped accordingly, thereby realizing blue-green and green light-emitting diodes and semiconductor lasers.

 

Note 6) In InGaN, the energy gap can be changed by changing the ratio of In to Ga. If Ga is gradually increased, the energy gap can be expanded to a maximum of 6.3 eV, and if the In is gradually increased, the energy gap can be reduced to a minimum of 2.0 eV. By changing the energy gap in this way, a light-emitting diode of any light color can be produced in the wavelength range of red to ultraviolet light. However, the more In is added, the more difficult it is to produce an InGaN film having better crystallinity.

 

Another difficulty is to fabricate the construction necessary for laser oscillation on GaN materials. Current semiconductor lasers employ a configuration in which a light-emitting layer of a double-heterostructure light-emitting diode is mirror-covered, thereby confining light and resonating light. The cleavage plane* of the crystal film is used as a mirror surface. However, GaN cannot be opened. It is necessary to make resonance surfaces by other means.

 

*劈面面 = smooth surface where the crystal is easily split and formed in a certain direction. Making such a split surface is called splitting. When a material such as a semiconductor single crystal is cut, a split method is used. However, depending on the type of crystal, there are points that are easy to open and cannot be opened.

The light between the bands is expected to be realized. Blue LEDs and green LEDs that are brighter than current blue LEDs will be manufactured in the near future.

 

In the remaining problem, the resonance surface, the technology of forming a resonance surface by etching is currently being accelerated. Nakamura confidently stated that it successfully achieved room temperature oscillations during the year of 1995.

 

　　 The people around me finally understand themselves.

 

Everything is going smoothly. Without being recognized by the society, Nakamura firmly believes that it can succeed, and therefore conducts research with one mind. The hard work he has done has produced fruitful results. Since then, the situation has completely changed. The invitation to invite him to give a speech has continued. His life has been hiding in the research room all the time, and has become flying to speak.

 

The research environment has also changed a lot. Previously, I was alone in research, and now the number of researchers has increased to five. Nakamura began to take the initiative to increase the proportion of management work in the work content. Although he said "Because I am a personally intimate personality, so there is still some loneliness", but seeing the growth of young technicians who have been able to stand alone, Nakamura has raised his eyes with satisfaction.

 

"But the biggest change is the recognition of the people around. In the past, although the LEDs were made and the light was emitted, the people around you still don't believe in you. Is it a waste of money, is it really successful? Can you make money? This is everywhere. Now this situation has changed. Everyone has begun to respect me." In this way, Nakamura smiled embarrassedly.

 

Edited later: Scholars' different views on R&D reports

 

Recently, many years of unresolved blue light-emitting diodes (LEDs) have finally been commercialized. As a material researcher, the author is naturally delighted with this achievement. The long series of "Nikkei Electronics" reported in detail the achievements of the Nichia Chemical Industry, and focused on the core figure Nakamura Shuji conducted serial reports.

 

After reading these reports, the author feels that the article's understanding of the development process of GaN LED is quite different from that of his own. Therefore, I would like to take this opportunity to review the development history and explain the opinions of a researcher.

 

GaN LED development history

 

Figure 1 summarizes 758 papers in the INSPEC-A database (about 2.9 million articles) of the British Institute of Electrical Engineers (IEE) with GaN as the key word. Of course, this inclusion does not necessarily cover all the papers, and some papers may not have much to do with LEDs, but you can still see the development trend of the research.

 

Figure 1: Changes in the number of GaN-related papers in different years

As can be seen from the figure, the first paper was published in 1968, and the paper on crystal growth in 1969 was published. Related research reached a peak in the 1970s, and the number of papers was basically fixed in the 1980s, and it began to surge after the 1990s.

 

Manufacturing practical high-brightness LEDs requires a number of important technologies. Table 1 lists the initial implementation dates and inventor names for these technologies. Looking back at the above historical process in comparison with Figure 1, an interesting phenomenon can be found. Driven by the success of material synthesis in 1969, research activities ignited a research climax in the 1970s. The original LED and photoexcited lasers were implemented in the early 1970s. However, due to the existence of major defects such as failure to achieve pn junction type and insufficient luminous intensity, the number of papers tends to decrease.

 

Table 1: Main research results of GaN-based semiconductors

However, the number of papers has surged again since entering 1991. It is estimated that this was affected by the realization of the p-type layer in 1989. Then, in 1992, high-efficiency LEDs were realized. In 1993, blue LEDs were realized, and research activities became more active. This year, the number of papers will increase in one step. The number of papers began to rise rapidly in 1991, and the opportunity was derived from the implementation of the p-type layer. At that time, as other difficulties were solved, the last obstacle of the p-type layer was removed, so a large number of researchers immediately invested in the research.

 

It can be seen from the retrospective history that the research team of Professor Akasaki Akira (now a professor at the famous city university) of Nagoya University has contributed to the study of the basic structure of LEDs. After that, Nakamura Shuji of Nichia Chemical Industry made a huge impact on the development of practical LEDs. contribution. Judging from this historical process, the report of the Nikkei Electronics only covers the latter, which makes people feel biased.

 

The start-up process of Nichia GaN research

 

In 1986, the author went to the University of Florida to study MOCVD (Metal Organic Chemical Vapor Deposition). At the time, I was disturbed by the research prospects for GaAs to achieve crystal growth on Si wafers. Therefore, a wide bandgap semiconductor was investigated. After investigating II-VI semiconductors and chalcopyrite semiconductors, the author concluded in 1987 that GaN-based semiconductors are most promising. There are two main reasons. First, compared with II-VI semiconductors, although the number of researchers is small, the results are more; second, the problems are relatively clear.

 

At that time, the author believed that the problems of GaN-based semiconductors were: (1) a large number of N-holes acted as donors, and it was difficult to realize p-type. (2) Since the growth conditions of AIN, GaN, and InN were greatly different, the crystals mixed together Growing is quite difficult.

 

When the author decided to go to the University of Tokushima in 1988, Nichia Chemical Industries launched research on GaN LEDs. At that time, the author considered that if the hole of the group V atom (nitrogen) is the source of the problem, then the problem can be solved by filling with the group V atom, so by adding P (phosphorus), As (ç ’), or C (carbon). When it comes to landfill, it is expected to form a receptor. The author proposed this idea to Nichia Chemical Industry, which started the R&D attempt. In addition, in order to reduce the growth temperature, the author also proposed a solution to replace ammonia with organic nitrogen compounds, and the company also tried.

 

In order to replace the ternary mixed crystal with a binary multilayer film, in 1988, I designed a special MOCVD device that can form a multilayer film. After that, Nakamura developed the Two-Flow type MOCVD device by improving the device.

 

Reporting attitude on research results

Research and development are like two wheels of a car. No one can do it. People's eyes are always easy to focus on directly creating product development, and often ignore the importance of research as another wheel. No matter what kind of product development, it is impossible for one person to complete all the work. The usual practice is based on the previous achievements, and it is clear what is their own creativity and development.

  The author believes that the media should not only report and evaluate the results, but should focus on the process of technology development and the performance of predecessors. This is not only a way to focus on basic research and originality, but also a mission of technology magazines. (End of the article, special contributor: Sakai Shiro, Tokushima University)