Quantum-dot lasers from Japan’s QD Laser will make high-speed “fiber to the home” networks simpler, cheaper, and more power-efficient
This is part of IEEE Spectrum’s SPECIAL REPORT: WINNERS & LOSERS 2009, The Year’s Best and Worst of Technology.
Quantum Dot-Com: Japanese start-up QD Laser’s Yasuhiko Arakawa (left) and Mitsuru Sugawara oversaw the 15-year effort to commercialize a temperature-stable semiconductor laser. Photo: Bruce Osborn; Photo-illustration: Bryan Christie Design
Suppose you had a dog whose personality fluctuated with the weather. On cool, crisp mornings, he’s a champ, fetching, rolling over, and shaking hands at your slightest command. But as the sun climbs higher and the day warms up, he becomes less and less responsive, and you have to ply him with doggy treats to get him to obey. And during heat waves? Forget about it—he barely plays dead unless you double or triple his kibble ration.
While you could excuse such behavior in Fido, something remarkably similar goes on all the time with the semiconductor lasers used in CD and DVD players and in optical communications. These tiny devices are incredibly sensitive to heat. Even a small rise in temperature causes the electrons within to move around faster and migrate out of the laser’s active layer—the thin slice of semiconducting material where the electrons recombine with positively charged holes to make light. As a result, the laser’s light output fluctuates, and it needs stronger and stronger electrical currents to keep lasing. At 85 °C, the device might need two or three times as much current to produce the same amount of light as at 25 °C.
To get around that shortcoming, developers of semiconductor lasers must either cool them or introduce extra circuitry that maintains the device’s output even as the temperature fluctuates. But those workarounds increase both the cost of making the lasers and the power they consume. Ever since this problem came to light, researchers have been hunting for a semiconductor laser that is inherently stable.
One promising technology, first proposed 27 years ago, is the quantum-dot laser. Such a device tightly confines the electrons and holes within many nanoscale blobs, or dots, of semiconducting material. With enough dots—millions or billions, that is—lasing will occur and steady output maintained, regardless of external temperature. While researchers can now grow these devices using standard molecular-beam epitaxy equipment, mass-producing them has been very tricky.
The Japanese start-up QD Laser, of Tokyo, a joint venture of Fujitsu and Mitsui Venture Capital Corp., has finally succeeded. Its quantum-dot lasers use inexpensive substrates made from gallium arsenide (GaAs) and boast an industry-leading density of 60 billion dots per square centimeter [see images, “Collecting the Dots”]. Compared with the conventional indium-phosphide lasers now used in optical networks, QD Laser’s devices will consume just half the power while transmitting up to 10 gigabits of data per second at a wavelength of 1.3 micrometers. Best of all, they will generate the same output at any temperature from –40 to 100 °C.
To mass-produce the GaAs laser chips, QD Laser has partnered with one of Japan’s leading consumer-electronics firms, which will use the same production lines on which it currently cranks out conventional red lasers for DVD and CD players, video-game consoles, and other products. (QD Laser says it will reveal the name of its partner later this year.) The initial shipments of laser chips are destined for an unnamed optical equipment vendor, which sometime this spring will begin offering the world’s first optical transceivers incorporating a quantum-dot laser. Fujitsu will almost certainly buy the transceivers for use in optical LANs and fiber-to-the-home networks.
The quantum-dot laser has long been envisioned as a successor to the quantum-well laser, itself an improvement on earlier laser designs because it confined the injected electrons to an extremely thin layer—no more than tens of nanometers thick—of active material. That way, it required less current to induce lasing. But like the “bulk” semiconductor lasers it superseded, the quantum-well laser is sensitive to temperature.
In the active layer of a bulk semiconductor laser, which you can picture as a fat, rectangular slab, the electrons and holes move in three dimensions, and that makes their interactions hard to control. In a quantum-well laser, they can move in only two dimensions, but electrostatic fields tend to build up, pulling the electrons away from the holes. In both cases, an increase in temperature makes the electrons more unruly.
Researchers began looking at ways to confine the electrons even further. In 1980, Yasuhiko Arakawa, a 28-year-old associate professor at the University of Tokyo, had an epiphany. “I thought, if we fix the position of each electron by confining it in a small box, the energy distribution will not be affected by temperature,” Arakawa recalled in a recent interview at his office at the University of Tokyo. Each “box” would be a semiconducting nanosize crystal into which electrons and holes would be injected. The box would effectively prevent the electrons and holes from being thermally excited to higher energy states.
He presented his quantum-box laser idea at the annual meeting of the Japanese Society of Applied Physics in March 1981. Then, collaborating with another professor, Hiroyuki Sakaki, he published a paper on the topic in the 1 June 1982 issue of Applied Physics Letters. The two researchers followed up with a series of experiments in which they confined electrons using 30-tesla magnets and demonstrated that the devices worked the same over a wide temperature range. “But I thought it would be impossible to fabricate such nanostructures until the 21st century,” Arakawa says.
The quantum-box laser concept didn’t exactly set the world on fire. Some people found it interesting but not particularly useful, while others concluded that the boxes would be structurally unstable. His early work “attracted almost no one to the field,” says Arakawa, now an IEEE Fellow. Today, he adds, thousands of researchers worldwide are working to advance the field.
Just three years after Arakawa and Sakaki’s paper, a research group at France’s Centre National d’Etudes des Télécommunications (CNET) noticed a strange phenomenon in the “superlattices” they were trying to build out of extremely thin alternating layers of indium arsenide and gallium arsenide. Studying their handiwork under an electron microscope, they noticed that some of the indium arsenide had formed tiny regular blobs atop the underlying layer of gallium arsenide. Each blob, it turned out, was a quantum dot. The French team didn’t actually produce lasing from their weird structure, but it was a start.
In 1994, a team at the Tokyo Institute of Technology and a collaboration of the Technical University of Berlin, Russia’s Ioffe Physico-Technical Institute, and the Max Planck Institute of Microstructure Physics independently demonstrated the first quantum-dot lasers. (At that point, the quantum-dot versus quantum-box terminology was still in flux, with the German-Russian team using the former term and the Japanese using the latter. Eventually, Arakawa says, the world settled on quantum dot. “Now even I call them quantum dots,” he says.)
But it’s one thing to create an experimental device in the lab and another thing to mass-produce a laser that operates reliably, can be manufactured cheaply, and performs a useful function.
QD Laser’s president and CEO, Mitsuru Sugawara, and his colleagues began chipping away at the problem of commercialization in 1994. Sugawara was then a research physicist at Fujitsu, aiming to develop a temperature-stable laser that emitted at 1.3 µm, the best wavelength for optical communications. “We weren’t interested in quantum dots per se,” Sugawara recalled in an interview last fall.
Like the CNET group, he and his team had been working on superlattices when they noticed quantum dots forming spontaneously, Sugawara says, “like water beading up on a waxed car.” After realizing what they’d done, they set to work on building a laser. “We knew that to produce lasing, we had to increase the density of the dots, so we started to study how to grow them intentionally,” he says. Five years later, in 1999, they demonstrated their first quantum-dot laser with a wavelength of 1.3 µm.
In a perfect world, the Fujitsu group would have continued to make steady progress, and a commercial quantum-dot laser would have hit the market years ago. In the real world, the IT bubble burst, and corporate priorities shifted. “My boss told me that if we didn’t stop our research [on quantum dots], he’d be fired,” Sugawara says.
Eager to keep Japanese R&D on quantum-dot lasers alive, Arakawa stepped in. By then his pioneering work on nanostructure devices had made him quite influential in Japan’s scientific circles. In 2001 he persuaded the Japanese government to include quantum-dot research in a national project on photonic networking. Fujitsu participated, along with Hitachi, Mitsubishi, NEC, and a number of other Japanese companies.
The Fujitsu group resumed its efforts to increase the dot density, mainly by stacking the quantum-dot layers. In 2004, they built a stack of 10 layers containing 30 billion dots per layer and capable of transmitting data at 10 Gb/s.
“At that point, we could think about starting up a venture company,” Sugawara says. Though it had nurtured the early stages of research, Fujitsu wasn’t the best place to commercialize the results, he says. The company’s main business is building high-end servers and optical networking systems for government and business customers. It has no expertise in the commodity chip-making methods that Sugawara envisioned using for the quantum-dot lasers.
In April 2006, Fujitsu and Mitsui Venture Capital formed QD Laser, providing the start-up with an initial US $2 million. Fujitsu agreed to let QD Laser use its 40 or so patents on quantum-dot technology; Arakawa signed on as the company’s technical advisor.
Although QD Laser’s official headquarters are in a central Tokyo high-rise, most of the company’s staff, including Sugawara, are based at Fujitsu’s facility in Atsugi, about 45 kilometers southwest of Tokyo, and research goes on there and at Arakawa’s labs at the University of Tokyo. There are currently 30 scientists and engineers involved, including five at the University of Tokyo.
After its founding , the start-up continued to work on boosting the lasers’ dot density. “We thought we could keep adding more layers, but we realized that wasn’t enough,” Sugawara says. Using proprietary techniques, researchers at QD Laser and Tokyo University eventually succeeded in doubling the dot density, from 30 billion dots per square centimeter to 60 billion. Sugawara brings out two atomic-force microscope images of quantum dots. The first shows a sparsely dotted surface. “Everyone can make this density,” he says. Then, pointing to the second image, which is crowded with dots, he says, “but only we can make this.”
QD Laser isn’t the first company to bring a quantum-dot laser to market. That distinction belongs to Innolume, a start-up based in Dortmund, Germany, and Santa Clara, Calif. Since 2007 it has sold quantum-dot “comb” lasers, which can emit tens to hundreds of colors over a range of wavelengths. The devices are potentially suitable for optical computing, laser television, and biomedical applications. But Innolume has yet to find a wide market for its products.
QD Laser will do better because its corporate backers have the muscle to see that it does. Fujitsu has already agreed to replace the standard indium-phosphide lasers in its optical networking systems with QD Laser’s gallium-arsenide lasers. But even Fujitsu had to be convinced that the new devices would be as reliable as existing lasers. “The communications market is very conservative,” Sugawara notes. To make its products more palatable to optical equipment makers like Fujitsu, his company spent months tailoring the quantum-dot laser’s output power and performance so that they matched those of a conventional laser. The resulting laser can seamlessly replace an indium-phosphide laser in an optical transceiver, with no significant redesign required.
With telecom giant Nippon Telegraph and Telephone Corp. adding 3 million fiber-to-the-home connections each year, Sugawara thinks his company could claim 5 to 10 percent of the Japanese market by 2011. QD Laser is also working on lasers for long-distance communications of up to 20 kilometers.
At press time, the company was wrapping up reliability tests and planned to begin selling in the spring. Even as it tries to line up more optical equipment customers, QD Laser wants to branch out into the consumer-electronics market, which buys 100 times as many lasers, or about 2 billion devices a year. That’s why the partnership with the Japanese consumer electronics maker holds particular promise.
Back in 2006, shortly after his company was founded, Sugawara visited four of the major Japanese consumer-electronics makers to gauge their interest in quantum-dot lasers. Three said no thanks. But the fourth, Sugawara recalls, told him, “We’ve been waiting for you.” The partnership is unusual in Japan, he adds, where there’s little overlap between the optical-communications sector and the consumer-electronics makers. “We’re one of the first companies to bridge the gap,” he says.
For two years, QD Laser engineers worked closely with the consumer electronics firm to refine the fabrication process for the laser chips. QD Laser grows the 3-inch gallium-arsenide wafers in-house and then ships them to its partner, which can print about 50 000 chips on each wafer. Each 0.3-square-millimeter chip consists of a substrate of n-doped gallium arsenide, followed by a layer of n -doped aluminum gallium arsenide, the quantum-dot layer, and then layers of p -doped AlGaAs and GaAs. The company packages each chip in a can about 2 cm long. “Even though we’re a small company, we can do mass production,” Sugawara says.
QD Laser’s partner would like to start incorporating quantum-dot lasers into its CD and DVD players and other products. By varying the size and concentration of the quantum dots, you can generate different wavelengths of light. To produce red light at 650 nm, for example, you could start with a 1300-nm quantum-dot laser and then pass it through a frequency doubler, which halves the wavelength. To make green light, you similarly start with a 1064-nm laser and double the frequency to get a 532-nm wavelength.
Quantum-dot lasers could also be used in laser TV sets, medical devices, and tiny portable projectors that fit in your cellphone. In the next couple of decades, Arakawa says, we’ll see quantum dots showing up in quantum computers and other IT devices [for more on quantum computing, see “Dot to Dot Design,” IEEE Spectrum, September 2007].
But why stop there? Quantum-dot researchers have been looking at ways to use quantum dots in biochemical sensors, solar cells, and other technologies. It’s a future Arakawa modestly refers to as “quantum dots for everything.”
For more articles, go to Winners & Losers 2009 Special Report.
Snapshot: A Laser That’s Right On the Dot
Goal: To commercialize a reliable and inexpensive semiconductor laser that’s also immune to temperature changes.
Why it’s a winner: These high-speed, low-power, temperature-stable lasers are equally applicable to optical networking and consumer electronics.
Who: QD Laser, a joint venture of Fujitsu and Mitsui Venture Capital Corp., and University of Tokyo
Where: Tokyo and Atsugi, Japan
Staff: 30 scientists and engineers
Budget: US $14 million
When: Spring 2009