MEMS have a long history in Canada, and arguably a bright spot in the country’s broader semiconductor landscape. But the sector needs more commercialization and fabrication capabilities if it is to be a global leader.
While the country is a long way off from any large-scale semiconductor manufacturing, MEMS are a much simpler technology requiring modest fabrication facilities. While MEMS may not be the bleeding edge of Canada’s semiconductor industry, the long history means there is a great deal of expertise that has been built up with well-established companies, as well as innovation by startups and university research facilities.
Canada’s inability to keep up with the broader semiconductor industry is arguably the reason why the country has developed its MEMS capabilities.
“The university labs could not keep up with semiconductor investments in the 80s and 90s and so they were forced to find new technologies they could afford,” Gord Harling, president and CEO of CMC Microsystems, told EE Times in an interview. “Thankfully, silicon photonics and microelectromechanical systems don’t follow the same trends of shrinking design rules as other semiconductors.”
Harling said much of the semiconductor equipment those labs had was perfectly suitable and it does not need to be replaced as frequently. “MEMS and photonics have much more to do with brain power and clever design and processing and much less on capital equipment,” Harling said. “The difficulty with MEMS is that the process tends to be optimized for each design, so it is difficult to get the benefit of large volumes unless you pick and choose your customers.”
MEMs are one of key areas identified as being strategic to Canada’s semiconductor program. The recently funded FABrIC program includes a commitment to create baseline or platform processes that people can use to create designs for MEMS or photonics. Harling said it is not a perfect solution, but enough to demonstrate a minimum viable product to an investor or a potential client. Once the product starts to move into volume production, he said, the company that developed it can then engage directly with the factory to optimize their design and process.
Through the FABrIC program, CMC Microsystems is engaging with Applied Nanotools in Edmonton, Canadian, Photonic Fabrication Centre in Ottawa, C2MI and Teledyne in Bromont, 3IT, 4D labs and others to develop platform processes with process descriptions or process design kits (PDKs), design examples, training and other tools.
CMC facilitates the fabrication of 80 or 90 prototypes each year in MEMS and about 180 in photonics, Harling said. “This is far more than other countries and we have an excellent technology base.”
Despite this strong technology base and long history of academic R&D in MEMs, Canada may not be getting full credit for its innovation as it often ends up being used in applications developed outside of the country.
Sazzadur Chowdhury, professor at the University of Windsor (which is home to the Microelectromechanical Systems MEMS Research Lab) said much of Canadian research activity around MEMS is done in conjunction with U.S. companies. Patents often get filed in the U.S., which means it is difficult to estimate how much Canadian MEMS technology is actually being used in commercial electronics like Apple’s iPhone. He noted that according to Statistics Canada, Canada has seen sharp decline in the number of patents in electrical engineering over the past two decades.
Conducting R&D is expensive, and Chowdhury noted that Canada’s largest MEMS player, Teledyne Technologies, with its MEMS foundry in Bromont, Quebec, is not a Canadian company. (Its presence in Edmonton, Alberta, is the result of acquiring Canadian MEMs pioneer Microlayne in September 2019.)
Canada’s investment in MEMS and semiconductors pales in comparison to the U.S., Chowdhury added. “Compare $52 billion from CHIPS Act with Canada’s $240 million during the same time frame. That tells the whole story!” The European Union currently has the European Chips Act, he added.
Chowdhury said Canada needs to have its own equivalent of the CHIPs Act if it is to develop its semiconductor capabilities, including MEMs, which have wide scale use, including electric vehicles, satellite, cellphones, TVs and microwaves.
He said there needs to be more funding for Canadian universities to support research and furnish them with the necessary equipment, as well as commercialization capabilities. “We need to have our own commercialization ecosystem or commercialization network,” Chowdhury said. “That commercialization network is absent. All the microelectronics campaign they have to add in Canada.”
Douglas Buchanan, professor at the University of Manitoba Price Faculty of Engineering, echoes those sentiments. He said the university is trying figure how to get engineers together with businesspeople, but right now most of the research that is done in Canada at universities gets adopted outside of Canada—and mostly the U.S. because it is next door.
He said focusing on MEMS makes sense for Canada because it is never going to be able to do CMOS or DRAM-level manufacturing because it is just too expensive. “We’re not going to be doing DRAM here anytime soon. The capital investment is tens of billions of dollars.”
Buchanan added that MEMs are in micron range, not nanometers. “You’re talking three orders of magnitude difference in terms of dimension,” he said. “It’s just easier to do and therefore it will be less costly.”
Buchanan, who worked at Watson Labs for 16 years with people from around the world including Nobel Prize physicists, said the R&D that is in Canadian universities is as good as anywhere in the world. “Research that’s done at Canadian universities is top notch. What happens is what we don’t have are enough people who are willing to create a startup and try and get things off the ground,” he said. “It’s a bit of a different mindset.”
Canada does have some spinoff activity. MEMS Vision is a spinoff of Montreal’s McGill University with its headquarters in that city, as well as a design center in Egypt for a total headcount of more than 50 people after being in business for about 15 years.
The company’s focus is on environmental sensors and ultrasound transducers. Hani Tawfik, lead product manager at MEMS Vision, told EE Times in an interview that the company designs the MEMS as well as the integrated circuits, and is set release a fine particle sensor in 2025 or early 2026 that is the first of its kind.
He said the company has developed a portfolio of MEMS processes that have been developed from scratch. MEMS Vision began by focusing on IP development and released its first sensor in 2019.
MEMS Vision continues to collaborate with McGill and the University of Quebec, Tawfik said, which provide access to clean rooms and manufacturing equipment for prototype devices. “Access to the universities has been very helpful.”
He said this ecosystem of clean rooms is one of the strengths of Canada’s MEMS sector because MEMS are very customized. “If you don’t have access to a clean room where you could innovate on the process that you’re building, you’re going to have a hard time building a competitive device.”
Tawfik said there is also a strong pool of educated talent. “This is a big positive in Canada.”
What the country lacks is collaboration across the industry, although it is getting better through initiatives like Canada’s Semiconductor Council, he said.
Canada also needs more robust fabrication capabilities, Tawfik added, as companies likes MEMS Vision often go out of country to fab in Southeast Asia.
Overall, MEMS in Canada needs more investment if it is to compete with the rest of the world, he said, including incentives to manufacturing in-country—which is going to become more critical as more countries cut ties with China and manufacturing in Taiwan becomes riskier.
Tawfik said Canada needs to take advantage of these geopolitical shifts if it wants to be a MEMS leader in North America. “There is that opportunity now, but it needs investments.”
From EETimes