The Vertical Cavity Surface Emitting Laser (VCSEL) is now essential in supporting modern society. How was this unique invention, overcoming a number of significant challenges, brought into practical application? Prof. Kenichi Iga, the “father of VCSEL”, and selected four researchers who worked with him on its development discuss the original research, how the technology was developed and what the future possibilities are.

Soda: 　I think that it was in 1978, the first year of my master’s course, when I was assigned to Prof. Iga’s group. He was an associate professor at that time, and so we did not have enough research funds. This meant that we had to make all the devices by ourselves from scratch. My work moved forward under the supervision of a student who was just a year above me. It was just around the time when Prof. Iga was invited to join Bell Research Laboratory, so I remember feeling worried that I might not be able to finish my master’s degree. To continue my research, I had to visit Prof. Yasuharu Suematsu in his laboratory on the Ookayama campus for supervision.

Iga: 　In 1973, I had my own laboratory and started research into semiconductor laser. Optical communications started to develop in 1976, and my target was to create a single wavelength semiconductor laser which is crucial for future highspeed optical fiber transmission. Starting from this, Chiyuki Kitahara, one of my post graduate students, said “Let’s just start with an LED and try to develop a surface emitting structure.” He was successful and presented the LED device at the Spring Meeting of the Japan Society of Applied Physics in 1978. Adopting Kitahara’s method, Soda successfully demonstrated the world’s first current injection laser oscillation in 1979.

Soda: 　I used an etching technique to create the first structure, but then I realized that it would not oscillate as a laser unless I had a perfect mirror. Prof. Iga told me to polish my own mirror. Because my major was in electrical engineering, I looked on polishing mirrors as a mechanical task, and so I was reluctant to get involved in mirror polishing. However, Prof. Iga just said “polish it until you can see your face in it!” So, I got to work, frantically!

Iga: 　Ah yes, I remember that. When I was working as a Post Doc at quartz oscillator lab, the polishing was one of important tool and I felt it was common for lab. members.

Soda: 　After repeated trial and error, I found a way to produce an unblemished mirror, and this allowed me to generate something resembling an “oscillation.” Employing a variety of etching techniques, I gradually improved the process and it evolved to a level where we could produce a chip. Eventually, we were able to demonstrate stable oscillation at liquid nitrogen temperatures. In the process, I managed to complete both my master's and doctoral degrees!

Iga: 　When I left for Bell Laboratories, my students seemed anxious. However, I thought they would be fine. This was because I had had a similar experience in the third year of my doctoral program when my mentor, Prof. Suematsu, went off to study abroad and left me to commiserate with my fellow students that we had lost our professor just at the most important time. But we managed to get though our situation somehow. In 1988, Fumio Koyama achieved room‐temperature continuous oscillation, and in 1989, Jack Jewell and other researchers at Bell Laboratories published a paper on microcavities. These were decisive years for the VCSEL.

Koyama: 　I became fully involved with VCSEL research after joining the laboratory as a research associate in 1985. Our objective was to achieve room-temperature continuous-wave operation of a VCSEL, which at the time was considered a major challenge. During the hot and humid summer of 1988, we explored every possible means of improving thermal management, including cooling the device in methanol and attaching it to heat sinks using low-melting-point gallium. After extensive trial and error, these efforts finally led to the first demonstration of room-temperature continuouswave operation of a VCSEL.

Koyama: 　Yes. The research results were first released as a postdeadline session at the international conference on OEC (Optoelectronics Conference) established by Prof. Iga. The paper was subsequently accepted for publication at an unprecedented speed by the Institute of Electronics, Information and Communication Engineers, and rapidly published in their newsletter. We went on to compile detailed data, which we published in Applied Physics Letters. This was followed by publication of the conclusive paper by Jack Jewell et al. at Bell Laboratories the following year. These papers suddenly opened up the field of VCSELs.

Iga: 　The timing of publication of the research and the academic society and journals that accepted papers for publication were critical. At that time, I was moved by the vital role played by Japan’s academic society.

Uenohara: 　: I wanted to study fiber-optic communications so in 1986, in my fourth year as an undergraduate, I applied for a position in Prof. Suematsu’s research group. However, there were so many applicants that we had to play paper-rockscissors shoot and I lost! Then I was told about Prof. Iga’s group, which was conducting semiconductor laser research on the Yokohama campus. In that group, as a result of asking Prof. Iga about the possibility of employing a quantum well to improve the performance of the VCSEL, I was given the research goal to investigate the possibility of achieving room‐temperature continuous oscillation with a MOCVD-grown GaAs-based quantum well VCSEL. Dr. Koyama already had the equipment needed and he supervised each step of my project. Despite my insight not being deep enough sometimes, I managed to achieve continuous oscillation at liquid nitrogen temperatures and so completed my master’s degree. The aspects that I missed at that time are now incorporated into the instructions given to current students as improvements.

Miyamoto: 　I also joined Prof. Iga’s group after losing at paper-rock-scissors shoot (laugh). In my fourth year as an undergraduate, I joined the team researching the use of Chemical Beam Epitaxy (CBE) to grow semiconductor crystals, where I received a very thorough training. I went through a worrying period when the laser refused to work at all, but after a number of attempts, I had a breakthrough when I noticed a faint oscillation at liquid nitrogen temperatures. I remember that Prof. Iga confirmed what I had noticed and said, “Well done!” I managed to get my degree, and just two weeks before my final interview with a company I had applied to join, Prof. Iga asked me to stay in his laboratory. He thought he could probably receive funding for his research through the new national Center of Excellence program. I decided to stay at the university, as I realized that I would not be given this kind of opportunity again.

Uenohara: 　From time to time, Prof. Iga would give us a research topic saying, “Would you be interested in such-and-such?” However, how we would proceed with the topic he left up to our own initiative. When I took up a teaching position, I found that if the supervisor is too strict about the detailed conditions, he runs the risk of obliging his students to generate a “model answer.” I want my students to think for themselves. This is one aspect of Prof. Iga’s teaching style that I inherited.

Soda: 　Some of Prof. Iga’s words that left an impression on me include “Just because there is no device to test it yet, that is no excuse for doing nothing” and “Don’t think too much, just try it.” This advice stayed with me even after I started working for a company, allowing me to look at situations from different angles. For example, “If there is no apparatus, I can borrow or make one,” or “It is better to try something and fail because then I can get feedback and a clue for the next attempt.” This allowed me to be proactive in taking up challenges.

Koyama: 　As Uenohara says, Prof. Iga’s approach was to present a topic but not to poke his nose into the details. These days, large projects have to have numerical targets that match the research funds, which obliges the research to be evaluated by how closely it achieves these targets. This makes me feel like poking my nose into the details even though I feel I shouldn't. Prof. Iga also took a very open-minded approach to research communities, an approach that has been respected by researchers of both industrial and academic institutes overseas. I would like to maintain this attitude.

Iga:　These days, academic circles tend to value people who have good presentation skills. However, some talented students are better at working quietly and steadily. Universities should be a place where this kind of talent is recognized and cultivated.

Koyama: 　VCSELs are now used in optical interconnects within data centers. Parallel processing is increasingly widely spread, and speeds are rising, gradually opening up the way to optical interconnects for massively parallel processing. At the same time, device research in universities is shrinking globally and shifting toward the private sector.

Iga: 　In data centers, the Digital Signal Processors (DSPs) consume massive amounts of power. They are not making the best use of the “ecological VCSELs.” If we could transmit signals over one kilometer through optical fiber using only lasers, this would eliminate the need for DSPs and amplifiers. This should significantly reduce the overall power consumption of the data center.

Miyamoto: 　I am currently working on wireless optical power transmission, which is a method for transferring power over distances using high-power VCSEL arrays. If we could produce a single laser beam from tiny laser arrays, the potential of VCSELs would be expandable even further. Arrayed VCSEL products are already available for efficient target heating, such as for automobile painting and the heat processing of metal parts. Further, using VCSELs as an energy source—such as for laser-driven nuclear fusion—has also been discussed. This would require a massive number of laser arrays, but the principle of the VCSEL would work well in such situations. Also, the concept of the Talbot Cavity* which Prof. Iga suggested during the 80s, has now reached the experimental stage overseas.

* The Talbot Cavity is a mechanism to combine beams from VCSEL arrays into a single coherent beam using a mirror resonator.

Uenohara: 　In long-distance fiber-optic communication, optical signals are converted to electronic signals and then converted back to optical signals after processing by DSPs and other devices. This obviously pushes the power consumption up. If optical signals could be transmitted directly over different networks using combinations of optical switches, the power consumption could be significantly reduced. However, I do not think that everything has to be done optically. One future challenge is how to determine a rational overall network design that allocates suitable roles to both optics and electronics.

Soda: 　Users sometimes request improvements such as, “We want to double our data center throughput” or “We want to double the laser output.” These are technically possible; however, they will also increase the power consumption. If we carry on at this pace, every data center will eventually require a nuclear power station next to it! We must keep redesigning entire systems to reduce their power consumption. I believe the VCSEL offers one such critical option.

Iga: 　The power issue is indeed a bottleneck in AI development. We have to work hard to clarify the most suitable areas for optics and electronics. I have the impression that Japan is reluctant to invest in networks after the severe economic blow they experienced in the Telecoms Crash, but we are already in the race to determine which devices will be used in the de-facto-standard technology. I believe that there a lot more opportunities and many different ways in which the VCSEL can be exploited.