The Parallel Path in Nagoya

Nakamura returned to Tokushima after the 1993 announcement, but the race had already been run on two tracks. In Tokyo, on a cold laboratory bench, a crystal had glowed with a pale blue light in the hands of a young researcher years earlier, a feat the world would only later recognize.

In the spring of 1989, at a conference hall in Japan, Professor Isamu Akasaki stood before an audience of semiconductor specialists and presented a finding that should have ended the gallium nitride problem. His research group at Nagoya University had achieved what every major laboratory had declared unachievable. They had made p-type gallium nitride. The material that had resisted every attempt to conduct positive charge carriers, the stubborn insulator that had killed RCA’s program and driven IBM from the field, now conducted electricity. Akasaki and his graduate student Hiroshi Amano had done it by bombarding magnesium-doped gallium nitride with an electron beam. The treatment transformed the material. The low-temperature luminescence that had baffled researchers for decades now showed the signature of successful doping.

The audience listened. They took notes. Then they returned to their laboratories and their zinc selenide programs. Akasaki’s 1989 paper, published in the Japanese Journal of Applied Physics, should have been the starting gun for a race. Instead, it became a citation in other people’s abandonment papers. The field had moved on. The consensus had calcified. A university professor and his graduate student working in a university lab could not overturn what RCA and IBM and the Japanese Ministry of International Trade and Industry had already decided: gallium nitride was a dead end.

The Nagoya group kept working. They had been pursuing gallium nitride since 1982, seven years before that paper, years spent building equipment that did not exist and running growths that did not work. They inhabited a different world from the one Nakamura occupied at Nichia. Their environment had no profit targets and no product deadlines. It had publication records and peer review and the slow accumulation of evidence that academic science requires. Their world also had no money. Akasaki funded his gallium nitride program through a patchwork of small grants, each one a negotiation, each renewal a gamble. The Japanese government had poured millions into zinc selenide through its national research programs. Gallium nitride researchers survived on scraps.

Akasaki had arrived at Nagoya University in 1981 after a career at Matsushita Research Institute, where he had first encountered the blue semiconductor problem. He knew what the industry wanted. He had seen the corporate laboratories try and fail. He had watched the consensus form around zinc selenide and gallium arsenide phosphide, the materials that could be grown, doped, and processed with existing equipment. Gallium nitride was the material he chose anyway. The choice was calculated, not romantic. The material’s bandgap was right. The theoretical efficiency was right. The only thing wrong was the practice, and practice could be changed.

Hiroshi Amano joined the laboratory as a doctoral student in 1983. He was twenty-two years old. Akasaki handed him the gallium nitride problem and told him to figure out how to grow single-crystal films. The student built his own metalorganic chemical vapor deposition reactor from spare parts and scavenged equipment. He heated substrates and flowed gases and watched the deposits form. Most of the deposits were garbage. The gallium nitride cracked, delaminated, or refused to nucleate on the sapphire substrates. The reactor ran for months before Amano produced anything worth measuring.

The Nagoya laboratory faced the same technical wall that had stopped everyone else. Sapphire and gallium nitride did not like each other. Their atomic lattices mismatched by thirteen percent, a huge gap in crystallographic terms. When gallium nitride tried to grow on sapphire, the atoms could not align. The resulting films were full of defects, polycrystalline messes that scattered light and trapped charges. Researchers had tried buffer layers of aluminum nitride, grown at low temperatures, to bridge the mismatch. The technique worked for gallium arsenide on silicon. It failed for gallium nitride on sapphire.

Amano tried aluminum nitride buffers anyway. He tried different temperatures, different thicknesses, different gas flows. He ran the same growth hundreds of times, changing one variable each run, documenting every failure in his laboratory notebook. The notebook filled with negative results. The films remained unusable. Then, in 1985, Amano tried something different. He grew an aluminum nitride buffer layer at low temperature—five hundred degrees Celsius instead of the thousand degrees required for crystalline growth—and then raised the temperature to grow the gallium nitride on top. The buffer was amorphous, not crystalline. It should not have worked. But when Amano examined his film under the microscope, he saw something no one had seen before. The gallium nitride had grown as a single crystal.

Akasaki examined the film. The X-ray diffraction pattern showed a single peak, narrow and sharp. The electrical measurements showed n-type conductivity without intentional doping. The film was not perfect, but it was usable. For the first time, gallium nitride had been grown on sapphire with quality approaching what the field demanded. Amano wrote up the result. Akasaki submitted it to the Japanese Journal of Applied Physics. The paper appeared in 1986.

The field did not notice. Applied Physics Letters rejected the submission. Reviewers questioned whether the result was real. The technique seemed to contradict what everyone knew about epitaxial growth. The paper eventually found a home in a Japanese journal, but the international audience never saw it. The consensus that gallium nitride was impossible remained undisturbed.

The Nagoya group pressed forward. They had solved the growth problem, or at least solved it well enough to move forward. The next barrier was p-type doping. Gallium nitride naturally conducted electrons, n-type, because of defects in the crystal structure. To make an LED, the material needed to conduct holes as well, positive charge carriers that would recombine with electrons at the junction and emit light. Every attempt to make p-type gallium nitride had failed. Magnesium, the dopant that worked for other III-V semiconductors, produced insulating material when introduced into gallium nitride. The prevailing theory held that magnesium atoms formed complexes with hydrogen during growth, complexes that trapped the charge carriers.

Amano grew magnesium-doped films and tested them. They were insulators. He tried different magnesium concentrations, different growth temperatures, different post-growth annealing treatments. Nothing worked. Then, in 1988, he placed a magnesium-doped sample in a scanning electron microscope and bombarded it with the electron beam. He was not trying to dope the material. He was trying to measure its cathodoluminescence, the light it emitted when struck by electrons. The measurement was routine, a way to characterize defects in the crystal. But when Amano looked at his electrical measurements afterward, he saw something unexpected. The region he had scanned had become conductive. The electron beam had changed something.

Amano reported the result to Akasaki. They repeated the experiment. They varied the beam energy, the exposure time, the magnesium concentration. Every time, the same pattern emerged. Magnesium-doped gallium nitride, bombarded by electrons, became p-type. The resistance dropped by orders of magnitude. The material conducted holes. Akasaki and Amano published the result in 1989. They did not understand the mechanism. They did not need to understand it to know they had broken the doping barrier.

The mechanism emerged later. The electron beam stimulated the desorption of hydrogen atoms from the magnesium-hydrogen complexes. Without hydrogen, the magnesium could accept electrons and create holes. The same effect could be achieved by thermal annealing in nitrogen gas, a much simpler process. Nakamura would use thermal annealing for his own p-type material. But the discovery belonged to Nagoya. They had shown that p-type gallium nitride was possible. They had published the method. They had done it first.

Nakamura read those papers. He had access to the Japanese journals. He had the time and the motivation to study them carefully. When he encountered the doping problem in his own reactor, he knew where to look. The Nichia engineer did not have to discover p-type doping. He had to implement it, improve it, scale it for production. The discovery was already published. The credit for the discovery already belonged to someone else.

The Nagoya work and the Tokushima work converged on a critical technical element: the reactor design. Both groups built two-flow MOCVD reactors, independently, for the same reason. The conventional single-flow design could not deliver the gases uniformly across the substrate. The gallium nitride films grew with uneven thickness and composition. The two-flow design introduced a second gas stream perpendicular to the substrate surface, pushing the reactants down onto the growing film. Film quality improved. Growth rates stabilized. The reactor that Nakamura built in his Tokushima laboratory and the reactor that Amano built in his Nagoya laboratory shared the same fundamental architecture.

The convergence was not coincidence. Both groups had studied the same failed attempts. Both had reached the same conclusions about what needed to change. Both had built their reactors from scratch because no commercial equipment existed for gallium nitride growth. The two-flow reactor emerged from the material itself, discovered independently by researchers who had studied the problem deeply enough to see past the conventional wisdom.

But the Nagoya group did not stop at the reactor and the doping. They built the first working gallium nitride LED. In 1991, Akasaki and Amano published a paper describing a p-n junction diode that emitted blue light. The device was dim. The efficiency was below one percent. But it worked. The junction conducted current. The electrons and holes recombined. The light came out. They had built a blue LED from gallium nitride, and they had published the result two years before Nakamura’s announcement.

The difference between the Nagoya LED and the Nichia LED was brightness. Akasaki and Amano’s device produced light that could be seen in a dark room. Nakamura’s device produced light that could be seen in a lit room. The difference mattered for applications. The difference mattered for commerce. But the difference did not change the fundamental achievement. Both groups had solved the same problems. Both had grown single-crystal gallium nitride on sapphire. Both had achieved p-type doping. Both had built working diodes. The Nagoya group had done it first. The Nichia group had done it brighter.

The institutional contexts could not have been more different. Nagoya University was a research university with limited resources and no stake in commercial outcomes. Akasaki published because publication was the currency of academic science. He shared his methods because sharing was how science advanced. His patents were defensive, filed to protect the university’s intellectual property, not to generate revenue. Nichia Chemical Industries was a chemical company with a stake in commercial outcomes and a culture of secrecy. Nakamura published nothing until the product was ready. He shared no methods until the patents were filed. His work was proprietary, shielded from competitors, hidden even from other divisions of his own company.

Both approaches produced blue LEDs. Both approaches were necessary. The academic work established the science. The industrial work scaled the manufacturing. The Nobel Prize in Physics for 2014 recognized both contributions. Akasaki, Amano, and Nakamura shared the award. The prize recognized a multi-stranded effort, convergent and parallel, spread across institutions and decades, not a single moment by a single genius.

The parallel path through Nagoya complicates every story about the blue LED. The lone genius narrative, the story of one man against the impossible consensus, captures part of the truth. Nakamura did work alone in his laboratory, building his own equipment, fighting his own management, pursuing a material that everyone else had abandoned. But he did not work in a vacuum. He built on published results. He adapted techniques that others had invented. He stood on a foundation that Akasaki and Amano had laid.

The Nagoya group presented their results at conferences. They published in journals. They filed patents. The Japanese Patent Office granted Akasaki and Amano a patent for their low-temperature buffer layer technique in 1991. They did everything that academic scientists are supposed to do. They documented their work and shared it with the field. The field did not care. The major laboratories had already committed to other materials. The funding agencies had already decided that gallium nitride was not worth pursuing.

Akasaki and Amano never sued anyone. They continued their research. They published more papers. They trained more students. They watched the blue LED industry grow from nothing to billions of dollars, knowing that their work had made it possible. The satisfaction of scientific discovery, the recognition of peers, the Nobel Prize—these were their rewards. They did not ask for a share of the profits. They did not hire lawyers. They did what academics do: they published and they taught.

The Nagoya laboratory continued to produce results after the blue LED breakthrough. Akasaki moved to Meijo University in 1992 but maintained his research program. Amano completed his doctorate and eventually joined the faculty. They developed ultraviolet LEDs, laser diodes, and other gallium nitride devices. They built on their own foundation. The science continued.

But the commercial success belonged to Nichia. The company that had taken the risk on Nakamura, that had funded his reactor and his failures, that had kept his work secret until the product was ready, reaped the rewards. The Nagoya group had published their results in academic journals. Anyone could read them. Anyone could use them. Nichia did. The company did not steal the science. It applied the science, improved it, and manufactured it at scale. The distinction between discovery and invention, between science and engineering, between publication and product, would become central to the disputes that followed.

The Nagoya path and the Tokushima path never quite merged. The two groups knew of each other’s work. Nakamura cited Akasaki’s papers in his patent applications. Akasaki followed Nakamura’s progress through the industry grapevine. But they did not collaborate. They did not share data. They did not coordinate their research programs. They worked in parallel, separated by institutional walls and cultural expectations. The academic scientist published. The industrial engineer kept secrets. The gap between them was the gap between two worlds of research.

That gap would matter when the Nobel Prize was announced. The prize recognized three laureates, not one. The Nobel committee understood what the industry consensus had missed: the blue LED required multiple contributions, multiple approaches, multiple failures and successes spread across years and institutions. The impossible material had yielded to sustained pressure from more than one direction.

The Nagoya group had broken the impossible loop before Nakamura did. They had shown that gallium nitride could be grown, doped, and fabricated into devices. They had published the evidence. The consensus should have cracked in 1989, when Akasaki and Amano demonstrated p-type material. It should have cracked in 1991, when they published their LED results. But consensus is sticky. The field had invested too much in zinc selenide to pivot on the basis of a few Japanese papers. The major laboratories had their programs, their budgets, their careers committed to the wrong material. They could not afford to be wrong, so they ignored the evidence that they were.

Nakamura could afford to be right because he had nothing to lose. He was a single engineer at a small chemical company with no reputation in semiconductor research. No one had invested in his program. No careers depended on his failure. He could pursue gallium nitride because no one expected anything from him. The Nagoya group had the same freedom, though for different reasons. They were academics. Their job was to explore, not to deliver products. They could work on impossible materials because no one expected them to make money.

The publication record tells the story that the patent record obscures. Akasaki and Amano’s papers are there in the journals, dated and peer-reviewed. The low-temperature buffer layer technique. The electron-beam activation of p-type doping. The first gallium nitride LED. Each paper a milestone. Each milestone achieved before Nakamura’s announcement. The science was not secret. The science was published. The industry simply chose not to read.

The Nagoya work also demonstrates something about the institutional conditions for breakthrough research. Akasaki had no product deadlines. Amano had no profit targets. They could pursue failures until they became successes. They could publish negative results. They could share their methods with competitors. The academic environment gave them the freedom to explore, but it also denied them the resources to scale. They could prove that gallium nitride worked. They could not manufacture it by the millions.

Nichia had the resources but not the freedom. Nakamura worked under constant pressure to produce results. His reactor was a proprietary secret. His methods were hidden from competitors. His failures were not shared. The industrial environment gave him the resources to scale, but it also imposed constraints on what he could publish and when. The two environments complemented each other. The science needed the academy. The product needed the industry.

The Nagoya laboratory’s contribution to the blue LED story was foundational. They proved that the material could be grown. They proved that it could be doped. They proved that it could emit light. They published the proofs. They filed the patents. They trained the students. The foundation they built was there for Nakamura to stand on. The Nobel committee recognized that foundation when they divided the prize three ways.

The parallel path through Nagoya remained less visible than the Tokushima path, but it was equally important. The academic science established what was possible. The industrial engineering made it profitable. Both were necessary. Both deserve recognition. The Nagoya group received their recognition through the Nobel Prize and through the historical record. They did not receive the profits. They did not ask for them.

The Nagoya work on gallium nitride began before Nakamura’s and continued after it. Akasaki and Amano’s first papers appeared in the mid-1980s. Their LED results appeared in 1991. Their work on ultraviolet devices and laser diodes continued through the 1990s and beyond. The blue LED was a milestone along a longer path, not the end of their research.

The geography matters. The blue LED emerged from Japan, from two laboratories that were four hundred kilometers apart. Nagoya and Tokushima are not the centers of the Japanese electronics industry. Tokyo and Osaka are the centers. The major corporate laboratories, the funding agencies, the research consortia—all were in the centers. The breakthrough came from the periphery. Akasaki at a regional university. Nakamura at a chemical company on Shikoku. The consensus formed in the centers. The breakthroughs came from the edges.

The Nagoya group’s work was always public. They published because that was their job. The Tokushima group’s work was always private. They kept secrets because that was their job. The different incentives produced different behaviors, but the underlying science was the same. Both groups faced the same material, the same obstacles, the same consensus. Both groups overcame them. The public record and the private patents tell the same story from different angles.

The parallel path through Nagoya establishes the historical record that makes sense of what came after. The Nobel Prize. The lawsuits. The disputes over credit and compensation. All of these later developments depend on understanding who did what and when. The Nagoya papers provide that timeline. They show that the foundational science was in place before Nakamura’s announcement. They show that the breakthrough was not a single event but a process. They show that credit for the blue LED belongs to more than one person and more than one institution.

The Nagoya laboratory’s contribution was essential. Without their work, Nakamura would have had to discover p-type doping himself. He would have had to develop the buffer layer technique himself. He would have had to prove that the material could work before he could convince anyone to fund his reactor. The Nagoya papers shortened his path. They showed what was possible. They gave him a starting point. The parallel path was a precondition, not a competition.

The scientific record showed that multiple researchers had broken the same consensus, solved the same problems, achieved the same results. The Nagoya group had published first. The Tokushima group had commercialized first. The Nobel committee had recognized both. The courts would have to untangle what that recognition meant for patents, for profits, for the question of who owned an invention that had multiple parents.

The parallel paths had converged. The conflict over credit was about to begin.