OTTAWA -- A technology once considered to have a lot of promise for making DWDM components -- the echelle grating -- could be about to make a comeback thanks to the efforts of two startups in this city, Optenia Inc. and MetroPhotonics Inc..

The two startups say they’ve conquered the challenges of making integrated echelle gratings, challenges which caused development work to grind to a halt several years ago. As a result, they think they can give manufacturers of arrayed waveguide gratings (AWGs) a run for their money.

The proof will come in a few weeks time, when Optenia, which was spun out of Mitel Semiconductor Corp. (NYSE/Toronto/London: MLT) last January (see Mitel Leads Startup Launch), expects to ship product samples to customers for evaluation.

Optenia’s CEO, Moris Simson, says his company will be able to develop customized DWDM (dense wavelength-division multiplexing) components faster than AWG suppliers can. They’ll also be significantly smaller and will be able to separate wavelengths with greater precision.

"We'll be able to give customers DWDM muxes with specified numbers of channels within six weeks [of placing an order]," Simson proclaims. "We have incredible miniaturization, flexibility, and turnaround time with our techniques."

Echelle gratings actually predate AWGs, which were invented in 1986 by Meint Smit from the Technical University of Delft in the Netherlands. Smit came up with the idea of AWGs when folk ran into problems making integrated echelle gratings -- ones created on a single piece of semiconductor material, as opposed to bulk echelle gratings, which are made from a collection of mirrors and lenses.

Integrated echelle gratings and AWGs have a lot in common. Both use interference effects to break light out into its constituent wavelengths. In an AWG, curved waveguides of incrementally increasing lengths create small delays or "phase shifts" in the light and, hence, interference (see Arrayed Waveguide Gratings (AWGs). A stepped mirror achieves the same effect in an echelle grating.

The mirror is formed by a smooth, vertical surface etched deep into the wafer. Only recently have etching techniques been good enough to achieve the required accuracy. And now that they have, it is possible that echelle gratings could outperform AWGs, according to Simson.

For a start, Echelle gratings are more compact than AWGs. The stepped mirror folds light back on itself, so the outputs are located adjacent to the inputs. Simson says Optenia's getting ready to sample a 40-channel device that's less than an inch square. Up to thirty-six dies can be cut from a single wafer.

It's also easier to manufacture more complicated gratings. When the number of channels increase or the channel spacing decreases, higher resolution is needed in the grating element. It's a lot easier to put more facets in a stepped mirror than it is to add extra waveguides to an array.

As noted, Optenia isn’t the only startup developing integrated echelle gratings. MetroPhotonics expects to ship its first silica-on-silicon based products later this year, according to its CTO Emil Koteles.

It's no coincidence that both startups are located in Ottawa. Both originated from research carried out at Canada's government-backed National Research Council (NRC). And the story of how the two companies came into existence is an interesting one.

About two years ago, when he was project leader for etched grating demultiplexers at the NRC, Koteles wrote a business plan to develop echelle gratings on glass. It won the NRC a development contract with Mitel. "That document was the genesis of Optenia," he says.

Koteles thought that echelle gratings in indium phosphide were an even more interesting prospect. But he couldn't interest Mitel in his ideas, so he left and set up MetroPhotonics in April 2000 (see MetroPhotonics Raises $62.5M CDN).

Now both startups license the same set of patents from the NRC, says Koteles. Optenia has permission to use the technology in glass-based devices, while MetroPhotonics may use it for indium phosphide-based products.

Then about six months ago, MetroPhotonics also decided to work in glass. "We figure that glass-based products will find quicker acceptance than indium phosphide ones," says Koteles. As noted, Optenia will be first with product samples. Koteles says to bear in mind that MetroPhotonics is only six months into development of this material, whereas Optenia -- though not officially incorporated until the start of 2001 -- has been working in glass for about two years.

He adds that MetroPhotonics expects to release products based on indium-phosphide before any glass-based devices. "In the longer term, the fact that we can integrate active devices into indium phosphide will be a big advantage."

— Mary Jander, Senior Editor, and Pauline Rigby, Senior Editor, Light Reading http://www.lightreading.com