Last Updated: September 23, 2008 11:54 AM

• Global MEMS/Microsystems markets and opportunities
  
• MEMS: On the Defensive
  
• MEMS at SEMICON West
  

Global MEMS/Microsystems markets and opportunities

Burgeoning demand for a host of new products enabled by MEMS devices means the sector will likely see a healthy increase this year despite any general semiconductor slowdown, and will remain on track to maintain its 17% average growth for the next five years.

With hit consumer applications like the Wii and iPhone in high-volume production -- and sparking a jump in interest from other consumer gear designers in similar new MEMS-based features -- MEMS device production hit 2 billion units in 2007, and looks to jump 25% to 2.5B in 2008. But these consumer market volumes bring consumer price pressures as well, holding total MEMS revenue growth to 14% this year.

As an increasing number of these new consumer designs are ramped to volume production, revenue growth will accelerate to 18%-19% in 2010-2012, even with the price pressure. By 2012, MEMS makers will be shipping 8.1B units/year worth some $15.5B, and nearly half that market will be consumer devices. Major market drivers will include silicon microphones, micro displays, RF MEMS, and even microfluidics for biomedical applications. RF MEMS and silicon microphones alone will account for more than 45% of unit demand from 2011.

Taking advantage of mature semiconductor process technology means MEMS makers can get by with spending a remarkably small 4% of revenues on equipment, and a similar 4%-5% on materials. The $8.0B in MEMS device sales this year is expected to generate $320M in demand for equipment and $380M in material sales. Yole Développement revised its forecasting methodology this year to better take into account how much actual usage of each tool was required for specific MEMS device structures, resulting in lower but more realistic figures than in some past projections.

MEMS equipment sales can also expect healthy 17% average annual growth through the next five years as it keeps pace with strong device market growth. The equipment market will slow to 3% growth this year, but then accelerate sharply as consumer devices now in development move to high-volume production. Equipment sales should see 9% growth next year, then see demand jump 26%-28% in 2010 and 2011.

With MEMS devices reaching real volumes for things like silicon microphones for cell phones, gyroscopes for game controllers, and digital micromirrors for displays, traditional semiconductor equipment suppliers including Tegal, Lam Research, and ASML are starting to pay more attention to this market. At the same time, MEMS tool suppliers will have to start paying more attention to service/support and reliability issues to compete with the big IC suppliers. Conversely, MEMS toolmakers are penetrating semiconductor niches requiring deep etching or wafer bonding, such as through-silicon vias, advanced packaging, and photovoltaics. The severe cost pressures of consumer markets, meanwhile, are driving device makers to smaller geometries, creating requirements -- and opportunities -- for new kinds of etch and clean technologies.

More closely linked to total wafer starts, materials for MEMS can expect a more stable growth pattern, with 12% CAGR over the five-year period to $603M. Materials sales should increase at close to that long-term rate in 2008, with about 11% growth to $380M.

Application trends

MEMS' mainstay automotive market will only see modest 3.5% average growth 2007-2012, but a number of consumer-oriented markets are poised for big growth, paced by 30% average annual growth for wireless telecommunications applications, 19% for biomedical applications, and 16% for other consumer gear. Consumer applications -- including inkjet heads, inertial MEMS, micro-displays, and emerging devices such as energy harvesting and autofocus systems -- will account for more than 40% of the total MEMS device market by 2012.

New MEMS devices hitting the market in 2007 included auto-focus systems, oscillators, and dual-axis gyroscopes. Gyroscopes for cell phones and low-cost micro-mirrors for pico projectors currently present strong emerging business opportunities for MEMS suppliers. Best five-year growth prospects for specific products will be RF MEMS and -- somewhat surprisingly -- microfluidics for drug delivery at >40% per year.

Silicon microphones will continue to see fast growth (32%), as will microfluidic chips for diagnostics (25%), micro tips and probes (22%), and microdisplays (21%). Defense markets will also average strong 21% five-year growth, from demand for things like high-value inertial MEMS devices for munitions guidance systems.

More analysis of this report from SEMI and Yole Développement appears in the September-October issue of Small Times.

by Sarah Fister Gale, Contributing Editor

The military is a hot new client for innovative MEMS and nanotech product manufacturers. From tiny explosive devices, to sensors for navigation, or video cameras that can survive dessert heat and torrential rains, security and defense applications offer a brave new marketplace for developers ? if they can find ways to produce customizable, highly-regulated, extremely sensitive low cost devices to meet the critical eye of military shoppers.

"There is a lot of potential for MEMS in this segment," says Bob Scannel, business development manager for Analog Devices Incorporated the global manufacturer of accelerometers and gyroscopes, headquartered in Norwood, MA. He points out that there are many potential applications for core MEMS technology in defense and homeland security devices, particularly in motion sensing for security, quality assurance, and predictive maintenance. Because MEMS solutions are smaller, lighter weight, more precise, and more durable than conventional technology, they deliver greater flexibility in the design of defense solutions which is appealing to the military buyer.

"MEMS Sensors can help detect early changes in vibration on air craft to avoid catastrophic failure; they can give gunners better feedback to improve their aim; and they can shut down or destroy valuable equipment in the field if it is tampered with," he says. "They offer a lot of potential added value to existing solutions."

However, the military is not an easy sell, warns Mike Elconin, client manager for the Center for Commercialization of Advanced Technology (CCAT) a consortium in San Diego, Calif., that helps developers speed the commercialization process for technologies dedicated to military, homeland security, and first responder operations. Elconin has worked with several MEMS developers who found a niche in the world of defense applications. "A lot of technology developers and small companies have a hard time finding their way into the military and homeland defense," he says.

At the same time however, Elconin points out that most technology used by the government is not designed in government labs. "The government doesn't build, it buys," he says. That means the companies that can find their way through the bureaucratic red tape have an opportunity to establish lucrative military relationships.

Customization and Scale

While the military's needs are vast, there are some consistent goals it strives for when it invests in new technology. Namely that it be small, cheap, sensitive and customizable, which can all be accomplished with MEMS technology, says Richard Waters, founder and chief technology officer for Lumedyne, a MEMS sensor designer in San Diego. "When you make things smaller the speed goes up and the power consumption goes down."

CCAT is currently working with Lumedyne Technologies on development of a lower cost optical MEMS displacement sensing technology for use in accelerometers for aerial navigation. The Lumedyne accelerometer measures displacements as small as 10 Femto-meters (.000,000,000,000,010 meters). Waters notes that the sensor technology can "fill the gap between GPS updates," or when satellite signals are jammed or lost.

Elconin has found that companies, such as Lumedyne that design MEMS sensors have had particular luck in meeting the military's ongoing needs. "Their navigation systems demand high levels of sensitivity which requires exquisitely sensitive accelerometers," he says. "If you are tracking an anti aircraft missile traveling at Mach 3, even the slightest adjustment in vertical acceleration could lead to significant divergence in where it ends up, and GPS systems are not fast enough or accurate enough to account for that."

But the process to develop such a technology for commercial military applications is tricky. "It's a long way from having a technology that theoretically works in the lab to one that works on missiles," Elconin says. "You have to build your proof of concept and go from there."

Elconin also notes that many of the 'MEMS for military' success stories begin with a product designed for an entirely different industry. The technology has to have a commercial scale market to be cost effective, and it's not always going to find that solely in military applications. "If you have to design a special tool that will only be used on a few hundred fighter jets it's going to be really expensive," he says. "But if you can develop a product that has large scale commercial applications and get that product into production mode first, then you can find your way back to the military."

Waters agrees. "Dual use is always important," he says, noting that the military likes to take advantage of reductions in start up costs by implementing technologies designed for other purposes. "MEMS development is heavy on the front end. It's cost prohibitive without commercial applications to ramp it up."

MEMS-based accelerometers, for example, are already a well-established technology for automotive safety systems for vehicle stability control and automobile navigation, and are produced on a mass scale, giving developers the chance to tweak an existing technology for new defense applications.

But the tweaking is not a small affair. Military applications are highly customized and developers need to meet exacting requirements to make the sale. Waters notes that proof of reliability for military components is the "number one hurdle" for any developer trying to sway skeptical military buyers. "You have to be able to provide reliability data, and that's not always easy to prove," he says.

Scannel adds that most military buyers want precalibrated systems that have been factory tested. "They don't want to invest in calibration testing. They want that done for them," he says, noting that developers who can precalibrate products may have a competitive edge in the sales process.

Partnerships Take Many Forms

All of this takes a lot of up front investment on the part of developers which can create burdensome obstacles. Some companies turn to groups like CCAT, which helps innovative companies transform ideas into commercial products through guidance and grant programs.

Other smaller companies have found success by merging with or selling themselves to more established companies that have the deep pockets and connections to get their technology through the door. Raytheon Company for example, recently acquired the robotics technologies and capabilities of Sarcos, a Salt Lake City-based company that researches and develops MEMS technology. Raytheon's Integrated Defense Systems provides solutions to the U.S. Missile Defense Agency, the U.S. Armed Forces and the Department of Homeland Security.

"Joining with Raytheon will help to move our technology from research and development to execution," said Dr. Stephen Jacobsen, president of Sarcos, of the sale.

While other developers partner directly with military groups to achieve commercial success. A team of scientists from the Georgia Tech Research Institute (GTRI) for example, is working with the Indian Head Division of the Naval Surface Warfare Center to develop highly-uniform copper structures that can be incorporated into integrated circuits then chemically converted into nano-scale explosives for military munitions.

The tiny copper structures, which have pores at both the nanometer and micron size scales, could play a key role in the next generation of detonators used to improve the reliability, reduce the size and lower the cost of certain military munitions.

Key to the design, says Jason Nadler, GTRI team member and research engineer, is that it creates a stable foundation for explosives. "The explosive material is so volatile it's too dangerous to study under a microscope so we have to understand the precursor material first," he says. "If we can control that we can better understand the rest of the processes."

The copper structure, a precursor material for explosive compounds used in military detonators, is being used to improve U.S. Navy detonator devices. (Source: Georgia Tech Research Institute)

Nadler uses a variety of templates, including microspheres and woven fabrics, to create regular patterns in copper oxide paste whose viscosity is controlled by the addition of polymers. He then thermochemically removes the template and converts the resulting copper oxide structures to pure metal, retaining the patterns imparted by the template. The size of the pores can be controlled by using different templates and by varying the processing conditions. Based on feedback from the Navy scientists, Nadler can tweak the structures to help optimize the overall device - known as a fuze - which controls when and where a munition will explode.

"Practical implementation of this technology will enable the military to reduce the quantity of sensitive primary explosives in each weapon by at least two orders of magnitude," said Gerald R. Laib, senior explosives applications scientist at Indian Head and inventor of the MEMS Fuze concept. "This development will also vastly reduce the use of toxic heavy metals and waste products, and increase the safety of weapon production by removing the need for handling bulk quantities of sensitive primary explosives."

The next step will be for Indian Head to integrate all the components of the fuze into the smallest possible package - and then begin producing the device in large quantities. As with the other successful military applications of MEMS, the copper materials can be mass-produced like computer chips with several hundred copper reservoirs produced on a single wafer.

And while Nadler believes the team will find other commercial applications for the copper material outside the military, he prefers working in the world of defense. "The nice thing about developing technology for military projects is that they are more exotic. You get to start without preconceived notions or requirements and explore ideas that no-one has explored before," he says. But, he adds that the military won't come looking for these solutions. "It's up to the MEMS industry to prove that with the right support and resources we can do anything they need us to do."

by Paula Doe, Contributing Editor

Packaging and test equipment for MEMS could be as much as a $2 billion potential opportunity, as the devices move into high-volume consumer applications.
Booming demand for MEMS systems in consumer applications means the MEMS market is seeing unit growth of more than 25% a year, reports Yole Developpement. But those high volume consumer markets come with consumer product price pressures as well, notes managing director Jean-Christophe Eloy. "That means overall MEMS prices will likely fall 6%-7% each year," says Eloy.

Most opportunity for cost reduction looks to be from packaging and test, which Eloy reports still accounts for 40%-50% of the cost of a typical MEMS device, and is particularly problematic for consumer devices. Cost pressures are driving a move from ceramic to plastic packaging, to wider use of wafer-level packaging, and away from wafer stacking to reduce the number of wafers used. The market remains extremely fragmented, with most device makers still making their own custom test equipment. But it's a potentially significant business. With at least 30% of the revenue of the projected $7.6 billion MEMS market this year going to packaging and testing expenses, that means a potential $2 billion opportunity.

One option to significantly cut packaging costs is wafer level test, so only the known good die are packaged. But physical testing of MEMS devices' proper response to sound, pressure, vibration and the like before packaging has typically been problematic. The MEMUNITY international network of test engineers and scientists is working on developing and sharing solutions to this problem, based on using a wafer prober to do the loading, aligning, contacting and mapping, and adding modules to apply temperature or motion or sound or pressure. "When we began to offer MEMS test, customers kept asking about applying non-electrical stimuli and measuring non-electrical results," says Frank-Michael Werner, business manager, SUSS MicroTec Test Systems GmbH (Dresden,, Germany). "We found there was nothing available on the market." So SUSS and the test house DELTA (H?rsholm, Denmark) got EU funding to develop a prober-based modular MEMS test system. Afterwards, they set up a user group as a way to make the results available to the MEMS community. Now the organization, open to all, holds regular seminars in Europe to exchange information and continue development.

"More and more things are coming together," says Werner, noting the group now has field-installed test set-ups for pressure sensors, silicon microphones, inertial sensors, micro mirrors and micro bolometers. Recent developments from the MEMUNITY community include a membrane-test system that applies an electrostatic stimulus to the device under test and measures the resulting motion with a laser-Doppler vibrometer, and cryogenic and vacuum test systems, such as the high-vacuum wafer-level test system at IMEC used for reliability testing of RF-MEMS devices.

Another option for cutting test costs is higher throughput, where SPEA (Volpiano, Italy) says it can get significant improvement by integrating its high speed pick and place system with electrical test and with physical test --with a multi-axis rate table for accelerometers, or an acoustic chamber for microphonesfor an integrated MEMS test cell.. "We've put the rate table in our pick and place machine," says Dave Webb. "So it can move in ways that resemble real life." He says this increases throughput by true parallel, multisite batch testing of devices.. Webb says the equipment is in use at big MEMS makers in Europe, where the Italian company is more well known for its MEMS and semiconductor test offerings. Though mostly known in the US for its board testers, SPEA is expanding its semiconductor and MEMS business here as well, through SPEA America offices outside Dallas, in Tyler, Texas.

Costs can also be reduced by reducing breakage and contamination of fragile MEMS structures that have to be open to the environment by dicing the wafers without saws or slurry, proposes Disco Corp. Instead the company focuses a laser beam inside the wafer. The laser doesn't have enough energy to effect the wafer as it passes through, but at the small spot where it's focused, it vaporizes the silicon, and the vapor expands to create a stress that breaks the wafer in a line along the cleavage plane. Then simply stretching the tape on the taped wafer gently separates the die.

The majority of silicon microphone makers are now testing or using the technology, says Scott Sullivan, chief technologist of Disco HI-TEC America. The current alternative, he notes, is conventionally dicing the taped wafer with a saw, and then manually pulling the dice apart. "As wafers get thinner and thinner, there are more and more challenges to using a traditional dicing saw, and this may be a solution," says Sullivan.

MEMS events at SEMICON West

These and other key industry players will update show goers on significant recent developments in the MEMS market, the progress of new volume consumer applications, and the impact on equipment and materials suppliers Tuesday at the Emerging Markets TechXpot in Moscone West.
http://semiconwest.org/Visitors/CTR_023121?parentId=&parent=yes&linkval=Emerging%20Markets