STUTTGART, GERMANY---There was no advance warning that the first convergence between Semicon Europa and the colocated baking/cooking trade show would pop out of the oven at the MEMS/MST Industry Forum.
During Duccio Sassano's presentation, the Olivetti i-Jet R&D/product integration manager described several of the "not-conventional applications in ink jet" projects that his team has been working on. Turns out one niche market is in the food biz, with the use of edible inks. He showed, on the conference-room screen, an image of a beautiful baked item with a full-color, high-resolution image of a red Formula One (I think, Ferrari perhaps) racing machine, inkjet-printed on top of the pastry. Duccio said a 5-picoliter printhead was used to "print" the frosty icing image. One more diet-killing app from the microelectronics space.
Speaking of killer apps, Jean-Christophe Eloy, GM (general manager or "grand maitre"?) of Yole Developpement, zeroed in on a potentially very potent MEMS device category---what he calls "compact light engines." These widgets combine newish micromirror technologies with solid-state lighting (eg, LEDs), facilitating display-quality projections from handheld devices on any type of compatible surface. His characterization of the emerging sector as a "strong growth market" was an understatement, since it's only a few million dollars now but is forecasted to be $500 million-plus by 2011. That's practically zero to 500 mill in five years. Now that's acceleration!
There's another MEMS that's already showing phenomenal unit and market growth from a standing start a few years back---silicon MEMS microphones. Akustica's Troy Chase (standing in for company chair and CTO, Kaigham Gabriel) showed some jaw-dropping numbers. The fabless firm shipped its first million parts by June 30, some 15 months after the first orders went out the door. OK, not bad. But the second million will ship any day now, a mere quarter or so after the first milestone. Not downplaying the heavy lifting that went into the development and productization of their mic products, Troy quipped: "We would like to encourage you not to enter the market...because it is so difficult [to make these things]."
Sometimes difficulties can be overcome with process licensing, and for good money too (if you're the licensor). Just ask MEMS maker Bosch, whose scientists invented the deep-reactive-ion etch (DRIE) process sharing the company name, a process that has been licensed to more microelectromechanical mavens and wannabes than you can shake an oscillator at. Bosch's Juri Marek praised Andrea Urban and Franz Lamer, the two Boschians (Boschites? Boscheviks?) who invented the ubiquitous etch process and recently won the European inventors of the year award from the EU Commission and European Patent Office.
In a later presentation, Jean-Marc Gruffat of Bosch process licensee Alcatel Micro Machining, discussed various enhancements that the French equipment company has made to the German invention. (Please note subliminal historical context there.) He noted that DRIE silicon etch "allows any shape, a variety of basic profiles, any size features." Among other improvements, Alcatel has managed to reduce the surface roughness "scalloping effect" to a smooth 14-nm RMS. It has also come up with a super-high-aspect-ratio process (acronymed "SHARP") that has produced ARs of up to 100 through the insertion of oxygen into the Bosch sequence.
One set of data points were not overly impressive, though. DRIE etch rates aren't exactly throughput-friendly, with wafer-per-hour speeds hovering in the low single-digits (as in one or two). Jean-Marc said Alcatel researchers have increased the rates by 40% and more. But if you run the numbers of the improvement in terms of throughput, even with the champion data and a four-module tool, the wafers-out sums still linger in the lowish double digits. Not exactly true high-volume production numbers, but hey, they're working on it, vingt-quatre sur sept.
Finally, I've one minor beef with an otherwise-solid day of presentations. Where were the metrology papers? Among 18 speakers, at least one from a metrology equipment company or a MEMS device maker with a metrological perspective would have been welcome. What needs to be measured more accurately and/or faster, and why? What the most challenging things to measure in the various MEMS processes? What are the metrology special needs for various types of MEMS, such as gyros, mics, or micromirrors? And that's just a few questions I would have liked addressed.
Perhaps some members of the MEMS community haven't seen the memo yet: Metrology is a critical value-add to any complex submicron or nanoscale chiplike process.
|
Blogs 
