Intense Photonics is into monolithic integration. In other words, it’s aiming to make optical components comprising multiple devices on the same piece of substrate, using III-V semiconductor materials such as gallium arsenide and indium phosphide.
Up until about a year ago, this concept was considered incredibly hot. But the challenges of mass producing monolithic chips have now struck home with the failure of earlier startups in this field like Nanovation Technologies Inc. and Zenastra Photonics Inc. (see Nanovation Files for Chapter 11 and Zenastra Photonics: RIP).
As a result, Intense Photonics missed the boat on getting a huge amount of funding, eventually raising a modest $11 million last summer from its seed-round investors, 3i Group PLC and Ireland’s ACT Venture Capital (see Intensive Care).
Intense Photonics reckons its particular way of making monolithic chips is amenable to mass production, and that it will now be able to prove this with the fabrication plant it’s opening today. The process it’s using, called "quantum well intermixing," enables the company to grow a crystal and then use a diffusion process to modify the electrical and optical properties of parts of it to create different devices.
Intense Photonics’ first product, an array of up to ten 980nm pump lasers on a single gallium arsenide chip, is made from a combination of three basic building blocks formed using this process, according to professor John Marsh, chief research officer at Intense. These are gain sections, passive waveguides, and non-absorbing mirrors. Maximum power output is 200 milliwatts per laser, Marsh adds.
Intense Photonics’ way of making the non-absorbing mirrors helps it relax manufacturing tolerances and improve yields, according to Marsh. That’s particularly important when making arrays of pump lasers, because it only takes one dud laser to render the whole array un-saleable. This is the main reason why other manufacturers of pump lasers haven’t made arrays yet -- because yields making single lasers are low enough as it is.
Marsh says Intense Photonics chose to make a pump laser array first, because it was based on gallium arsenide, which folk are more familiar with than indium phosphide. It’s also a device that can simply replace single pump lasers, he says, and it doesn’t require system engineers to design it into a product.
The main application for the pump laser array is likely to be in waveguide amplifier arrays developed by the likes of Teem Photonics, Cisilias A/S, Northstar Photonics Inc., Redfern Integrated Optics, and Symmorphix Inc. (see Symmorphix Joins the Amp Camp and Which Amp is the Champ?).
All of these outfits are developing the semiconductor equivalent of Erbium Doped-Fiber Amplifiers (EDFAs) targeting lower power, lower cost applications in metro equipment. Instead of using fibers doped with erbium, they use waveguides doped with erbium, the waveguides being microscopic channels within chips. In both cases, the light passing through the fiber or waveguide is re-energized by pumping light into the surrounding material, using a pump laser.
The idea of having up to 10 pump lasers on a single chip, pumping light into an array of up to 10 waveguide amplifiers on another single chip, appeals enormously to James McKenzie, VP of marketing and business development at Teem Photonics. “This is what waveguide amplifiers need, to realize their full potential,” he says.
Right now, pump lasers come with fibers attached to them, which means that Teem not only has to buy several lasers for each waveguide array but is also faced with some time-consuming, complicated tasks in aligning and fusing lots of fibers. All of this could be eliminated with a pump array, which could be linked to Teem’s waveguide array in an automated production process, says McKenzie.
Not surprisingly, developers of EDFAs are less impressed with Intense Photonics’ developments, because they can’t reap the benefits of eliminating fiber connections. With EDFA arrays, it probably makes more sense to use a really powerful pump laser to re-energize light in multiple fibers, according to Anatoly Grudinin, technical director at Southampton Photonics Inc., a startup making EDFA arrays (see Startup Claims First Multi-Port Amplifier). “If [Intense Photonics] can make a pump laser array really cost effective, it would be interesting,” he says. “But I doubt whether they can.”
Marsh says Intense Photonics has already sold some prototypes of its pump laser array and expects to go into volume production by the second half of next year.
This appears to put it ahead of other companies developing pump lasers. Gemfire Corp. has an array of eight pump lasers, but the lasers aren’t all on the same chip (see Gemfire Announces Products). Other startups that have the potential to make pump laser arrays but haven’t announced anything yet include Alfalight Inc. (see Alfalight Closes $28M 2nd Round) and Novalux Inc. (see Novalux Details Laser Advance). The incumbent players in the 980nm pump laser market are JDS Uniphase Inc. (Nasdaq: JDSU; Toronto: JDU) and Nortel Networks Corp. (NYSE/Toronto: NT) (see Nortel Buys JDSU Plant for $2.5B).
Marsh points out that Intense Photonics can make a wide variety of components with its technology. It’s also developing a 2x2 crosspoint switch that has zero losses (it incorporates amplifiers) and “sub-nanosecond” switching speeds, enabling it to be used for optical packet switching (see All-Optical Switching Tutorial, Part 1). The basic building blocks in this widget include waveguides (incorporating splitters and beam combiners), modulators, and gain sections. “It may not be our next product,” notes Marsh. Its main goal is “to show the power of our integration technology.”
— Peter Heywood, Founding Editor, Light Reading
http://www.lightreading.com Want to know more? This very topic is the subject of a session at