After operating in stealth mode for almost a year, Princeton Lightwave has now revealed an impressive range of high-power InP-based lasers, writes Peter Gwynne.

For several months after its founding in May 2000, Princeton Lightwave Inc (PLI) hid its light under a bushel. Then, starting in March this year, the company from Cranbury, New Jersey started to take that light, and the optical products related to it, into the open. The company has announced a series of new high-performance active components based on indium phosphide, designed for applications in long haul and metro networks. "We re looking at a number of different areas, including modulators and semiconductor optical amplifiers," says Pamela York, vice-president of corporate development at PLI. "At present, we re focusing on achieving the highest power levels for pump lasers. We are selling to telecom companies and their system suppliers, to component manufacturers, and to chip-on-carrier customers. PLI is committed to being an industry leader." A start-up with products Despite its low profile, the company has been selling its WaveRider broadband gain chip since its inception. The InP laser chip, which is supplied as a chip-on-carrier subassembly (see ), is designed to be incorporated into customers external cavity lasers (ECLs) to form broadly tunable source lasers. ECLs, in which wavelength tuning is provided by adjusting the external cavity in a variety of ways, are commercially available from companies such as Iolon, Blue Sky Research and New Focus (see Compound Semiconductor April 2001, p24). PLI says that the WaveRider product permits customers to reduce the size of their products and the complexity of their manufacturing process. According to York, PLI had revenues of "several million dollars" last year. The WaveRider was officially announced in March, along with two other new families of products. The WaveHarp is an advanced distributed feedback laser that emits up to several hundred mW of power in a narrow spectral band. By integrating the grating on the laser chip, this approach eliminates the need for an external fiber Bragg grating to achieve wavelength stabilization, thereby reducing costs. At the CLEO meeting in May, PLI claimed a record output power of 340 mW for a 1550 nm DFB laser coupled into a singlemode optical fiber. WavePower pump lasers, meanwhile, are available at 14xx nanometers for Raman amplifiers and 1480 nm for EDFAs. The pump power of up to 500 mW in a module provides higher amplifier gain and more robust amplifiers. The WavePower platform provides PLI with a route into an extremely promising market that, by giving customers longer transmission distances without optical regeneration, promises more bandwidth at lower network costs. Indeed, the company recently announced it had become the first to reach the one Watt power level from a single narrow-stripe pump laser chip. "Last year s record was 700 mW," points out senior analyst Vladimir Kozlov of RHK, a company that analyzes the telecommunications industry. "They aim to provide existing products at lower cost with higher power," adds Jay Patel, a senior analyst at The Yankee Group. "If delivered, the expertise they claim could be pretty significant." The market for such pump lasers is also significant. "This will be one of the fastest growing segments in the next few years," Kozlov says. "PLI is really targeting the sweet spot, and the product should be substantially better than what s available on the market today." Roots at Sarnoff That market potential owes much to PLI s roots in the Sarnoff Corporation, a wholly owned subsidiary of SRI International that is the part of the old RCA Laboratories not bought by GE. Sarnoff produces, incubates and commercializes electronic, biomedical and information technologies. "Sarnoff was one of two federally funded organizations in the United States being paid to develop ultra-high power lasers," recalls York, who spent time on Sarnoff s staff. "Our high-power laser technology seemed a good opportunity to spin off. We had really tremendous results with high-power lasers around 1480 nm, and few companies seemed able to deliver the power levels we can provide." That promise persuaded Sarnoff to spin off PLI as a privately held venture-backed company. Moving out with it were York, four other co-founders, and plenty of intellectual property. "We are unusual for a start-up in that we have an extremely broad background in technology developed at Sarnoff," says John Connolly, a co-founder who is now PLI s vice president of engineering. "About 27 patents moved into the company from Sarnoff." The breadth of the start-up company s technology stemmed from an exclusive license from Sarnoff for active components. The license comprises developments in lasers, detectors, modulators, semiconductor optical amplifiers and other active components, as well as advanced device and fabrication technology for an InP-based integration platform. Facilities and personnel The new company had other assets, in terms of people and infrastructure, on which it has continued to build in the past year. Its vice-president of research, for example, is Dmitri Garbuzov, a renowned laser physicist who trained at St Petersburg s Ioffee Physico-Technical Institute with Zhores Alferov, winner of a share of last year s Nobel prize in physics. Raymond Menna, who oversees materials growth, trained under another Nobel laureate, Daniel Tsui, who shared the physics prize in 1998. Menna has five crystal growers at his disposal. Altogether, PLI s staff of 50 includes a 25-person engineering team, of which 60% have PhDs. Nancy Morris, the fifth co-founder of PLI, has over 20 years experience in active optical components and materials technology gained at Laser Diode Inc and Sarnoff Corp. Pam York also notes that PLI has excellent facilities. The company s 90 000 sq. ft of space, expandable to 150 000 sq. ft, includes state-of-the-art automation and modular tooling for batch manufacturing and rapid retooling for new product lines. "We have two reactors, in which we can do 500 wafer starts per week," says York, who was a crystal grower. Technology choices Initially, the leadership of the new company faced the issue of which technology to pursue, and decided to focus on indium phosphide. "Every wavelength that you need for telecommunications can be done in InP," explains Connolly. "It enables the merger of active and passive technology that will be the key when moving forward with optical integrated chips." Another early decision involved the wavelength on which the firm should concentrate. The question was this: should PLI work in the 980 or 1480 nm range? "Even though we had the world s power record at 980 nm, that market was fairly crowded, and the engineering had already been done," explains Connolly. "We felt we could really improve the power capability in the 1480 nm wavelength range." In fact, high output power is a key feature of PLI s products. "The primary focus that we took to start the company was the power capability that we always prided ourselves on at Sarnoff," says Connolly. "As you move power levels up, the number of dollars per photon will fall, driving down the cost of the network." PLI also made the conscious business decision to enter the marketplace at the high end. "We ll be able to compete across numerous market segments," Connolly says. "Our technology allows us to compete at the high end of the market, while our automated chip-on-carrier line provides the opportunity to go downmarket." Market opportunities The commercial space in which PLI operates is already becoming crowded. Potential rivals include Furukawa Electric (currently the market leader for 14xx nm pumps), Corning, JDS Uniphase and Nortel Networks. However, the market shows signs of expanding rapidly. According to RHK, the markets for 1480 nm and 14xx nm pump lasers and for tunable source lasers will both grow very rapidly in the next three years. PLI seems ready for the challenge. "The company has been in a stealth mode," says Patel. "But it has a market message now. This is the right timing for the company." Future growth What comes next? The company expects to triple its head count by the end of the year, and also has high ambitions for its technology and products. "We re really pushing integration, both hybrid and monolithic," says York. "We see hybrid integration really coming on strong over the next two years. We think monolithic integration is several years out, although we re pushing it as fast as we can; it will make a huge difference." "We really have a cost advantage with our manufacturing facilities," York continues. "With low cost and high performance we have the ability to be a major player. We have the capacity to deliver tens of millions of chips, and hundreds of thousands of modules."