Executive Director, Japan Society for the Promotion of Science (JSPS)
Professor Emeritus, Tokyo Institute of Technology
Birth of surface emitting laser
—An initiator of nanotechnology in photonics—
A surface emitting laser can emit light vertically from the surface of a semiconductor substrate, and it is characterized by low power consumption, a long lifetime, and a monolithic process to fabricate. Since 1999, many institutes have been conducting R&D in order to commercialize the technology. The idea of the surface emitting laser was first proposed by Prof. Iga in 1977.
In the mid 70’s, when it was shown that transmission loss in optical fibers could be reduced by the use of near infrared light with a wavelength of longer than 1.3 mm, research involving the development of a long-wavelength laser which was suitable for optical transmission began worldwide. Prof. Iga and Prof. Yasuharu Suematsu, who was a professor at Tokyo Institute of Technology at the time, developed a long wavelength semiconductor laser with stripe-geometry that emits light parallel to the surface of the substrate. Its performance was one of the best in the world. However, Prof. Iga was not satisfied with it. “To build a stripe geometry laser, a semiconductor substrate with multi-layered thin films has to be cleaved 300 mm in width by a blade. A laser cannot be tested until after the stripe-geometry is formed by cleaving. Since cleaving substrates has to be done manually, they are unsuitable for mass production. As well, they cannot be formed with two-dimensional arrayed configuration. I wondered how a laser could be built without having to cleave the substrate. I thought about it day and night and finally came up with the idea of changing the light emitting direction from horizontal to vertical, “ says Prof. Iga. That was the moment when the idea of “Surface Emitting Laser” was born. Thus, a laser can be built without the cleaving process by placing reflectors above and below the semiconductor thin films. This method makes large-scale integration possible as well. However, insufficient light amplification for laser oscillation became a problem with surface emitting lasers because the thin active layer caused extremely short resonator lengths. To solve the problem, he had to increase the reflectivity of the reflectors to near about 100%.
Prof. Iga and his group began verifying the concept of the surface emitting laser. When the structure with an active layer consisting of GaInAsP was cooled with liquid nitrogen and a pulsed current was applied, it instantly emitted light. However, the threshold current required for laser oscillation was about 1A, which is 1000 times higher than the current presently used in the device. The device had to be cooled to liquid nitrogen temperature; otherwise, it would break. “It might be impossible, in principle, to obtain laser oscillation at a practical level with surface emitting lasers because of its extremely short resonant length or maybe the theory was correct but there might be a problem with the technology,” thought Prof. Iga. Although he continued research on developing crystal growth technology as well as the theoretical studies, his research was not going well because there was only a liquid phase epitaxial method for the crystal growth at that time, and it was unsuitable for nanoscale processing. Later, along with the developments in transistor technology, a metal organic chemical vapor deposition (MOCVD) was developed. In MOCVD, a metal organic material is transformed into a gas and the gas flows over the crystal growth substrate. Nanoscale structure control became possible with this process. He built his own MOCVD for his research, and in 1988, he prepared the first room temperature surface emitting laser with continuous oscillation together with his colleague, Prof. Fumio Koyama of Tokyo Institute of Technology. The threshold current was 20 to 30 mA and the laser output was 1 to 2 mW.
Continuous oscillation at room temperature attracted worldwide attention and research concerning surface emitting lasers accelerated. In the ’90s, they were used in transmitters for high-speed LAN. In the United States, national projects on massively parallel optical transmissions and optical interconnections using surface emitting lasers started, and since about 1999, surface emitting lasers have been used for high-speed networks. Prof. Iga says, “In 1999, during which surface emitting lasers were first used in the United States, we were already on the fourth generation of the laser. I verified its principles from the late 70’s to the early 80’s, improved and developed the device for practical use during the ’80s, and turned various ideas such as wavelength conversion into reality in the ’90s. “Tunable lasers were developed by using the reflectors that are controlled by MEMS technology. The nanofabrication technology used for surface emitting lasers is the same as that used for LSI. Nanotechnology including the application of MEMS is under development,” says Prof. Iga. During the Clinton administration, this laser technology, developed at Tokyo Institute of Technology, was mentioned on the Website of the National Nanotechnology Initiative (NNI). The nanotechnology that Prof. Iga developed will bring about further developments in surface emitting lasers.
