Sparkolor Secrets Surface
Now the startup has gone a little further. It’s planning to unveil details of its tunable laser tomorrow, and in preparation for this, it’s briefed Light Reading on key aspects of its technology. In addition, we've dug up a patent that sheds more light on Sparkolor's secrets.
Sparkolor, it turns out, is doing more than playing catchup with other makers of widely tunable lasers based on distributed Bragg reflector (DBR) technology, like Agility Communications Inc. and the Altitun business unit of ADC Telecommunications Inc. (Nasdaq: ADCT), as we'd originally thought (see Sparkolor Plays Catch Up).
It’s different in two respects. First, Sparkolor is aiming to make a wider range of widgets. "Our first product is a tunable laser,” says Steffen Koehler, Sparkolor’s VP of marketing,. “But then we'll move on to make OADMs [optical add/drop multiplexers], optical regenerators, wavelength converters, and stuff like that."
Second, Sparkolor has a technology platform that's more amenable to high-yield production, Koehler claims. Agility and Altitun are making monolithic devices -- creating the complete laser structure out of a single piece of indium phosphide. This is a very complicated process with very low yields, which makes it, in Koehler's view, too expensive.
Sparkolor's laser works in a similar way to the multisection DBR devices from Agility and Altitun, Koehler says (see Tune In!). All are made using wafer-level processing. But rather than being made entirely out of indium phosphide, Sparkolor's laser is a hybrid device -- the gain (light-producing) region is indium phosphide, and the grating region where the wavelengths are defined is made of glass (silica).
Put simply, Sparkolor starts off with a wafer of silica-on-silicon, and embeds little blocks of indium phosphide where required. Incidentally, silica-on-silicon wafers are the most common starting point for making Arrayed Waveguide Gratings (AWGs), and Koehler himself is a former executive of AWG manufacturer Kymata Ltd., now part of Alcatel Optronics (Nasdaq: ALAO; Paris: CGO.PA) (see Ex-Kymata Exec Joins Sparkolor).
Combining the two materials allows Sparkolor to get better performance from its laser, since it doesn't have to compromise on material properties, Koehler says. "Indium phosphide is a lousy material for tuning. When you tune with carrier injection [varying an electric current], the power fluctuates all over the place." To avoid this problem, Sparkolor intends to use thermo-optic tuning.
Koehler wants to keep the remaining details of the laser structure and operation confidential, however. "One piece of magic is how we combine the different materials in a high-yield way. The other piece of magic is the tuning process."
Clues to those "pieces of magic" come from a recently published U.S. patent (6,243,517) filed by Sparkolor's founder David Deacon. This appears to describe a tuning mechanism based on a "grating-assisted coupler" -- the same physical mechanism used in the coarse tuning of Altitun's laser, according to Rob Plastow, Altitun’s former CTO. (For details of how this works, see Tune In!, page 7.)
As noted, Altitun's laser is all made in indium phosphide, which is tuned by injecting a current. In Sparkolor's case, the grating is in glass, so the tuning mechanism is temperature.
Plastow sees potential downsides in the hybrid design. "The only lever is temperature, which is slow," he points out. "And parts of the structure must be made temperature invariant… while adjacent parts are temperature tuned."
With any device that incorporates more than one material, thermal stability is paramount, he adds. Any change in temperature will alter the properties of the laser. Glass is problematic in this respect because its refractive index changes greatly with temperature -- the very feature that makes it suitable for temperature tuning in the first place.
Sparkolor's patent suggests this can be overcome by surrounding the waveguide core with a polymer that has an opposite refractive index change, to exactly counteract the refractive index change in the glass core. "This sounds tricky, and is not, as far as I know, used in any production devices -- but I may be wrong," Plastow notes.
Another tricky area is the transistion from the indium phosphide chip to the glass, which must have extremely low reflections. "Other difficulties are shared with other external cavity hybrid devices," says Plastow. "Very low reflections are needed, mechanical and temperature stability is crucial, and manufacturing cost will depend on how much active alignment or adjustment is needed."
Sparkolor's patent goes on to describe more complicated embodiments of the laser technology and points to how Sparkolor might use its ideas to build more advanced devices in the future.
It’s worth noting that Sparkolor has some formidable technical and financial credentials. Its founder, Deacon, also founded Gemfire Corp., another hot integrated optics startup (see Gemfire Comes Out Blazing). Sparkolor’s investors include New Enterprise Associates (NEA) and Optical Capital Group, the venture capital company founded by David Huber, CEO of Corvis Corp. (Nasdaq: CORV).
— Pauline Rigby, Senior Editor, Light Reading