Where MEMS Can Boldly Go Now

They’re not just for smartphones and wearables.


MEMS chips are being designed to go into the human body as biosensors, which will require unique packaging. And as demand grows for assisted and automated driving, MEMS devices also are finding new use cases in automotive electronics, their chief market segment prior to the millennium.

Pressure sensors, such as those that monitor the air pressure in tires, remain the biggest type of MEMS sensors in unit volume, although they play an especially critical function for gigantic mining trucks, where the tires can cost $100,000 apiece and must be replaced every three months.

Yole Développement forecasts the MEMS packaging market will increase from $2.56 billion last year to $6.46 billion in 2022, for a compound annual growth rate of 16.7%. That CAGR is greater than Yole’s estimate of a 14.1% growth rate for MEMS devices from 2016 to 2022.

At the MEMS & Sensors Executive Congress in San Jose, Calif., there was talk of bioMEMS and biosensors, piezoelectric MEMS that can run almost without batteries, and MEMS being made on paper and plastic, as research funding dwindles for silicon-based MEMS technology.

Five companies took part in a technology showcase of their products at the conference. Demonstrating its optical gas sensor for attendees, eLichens showed its Berries Smart Sensor Series, an air quality sensor in a system-in-package containing its proprietary infrared MEMS emitter and detectors, a patented optical sampling chamber, and signal processing – all in a device about the size of a quarter.

Fig. 1: eLichens’ gas sensor. Source: eLichens

The winner of the showcase, as voted by the conference attendees, was Menlo Micro’s Menlo Digital-Micro-Switch technology, which shrinks MEMS switches to less than the width of a human hair. The company’s switches are contained in a land grid array package measuring 6 mm by 6 mm by 1.3 mm.

Hundreds of these microswitches go into one device, capable of operating at more than 25 watts and handling radio-frequency input voltages of more than 200 volts.

The device grew out of long-term industrial research and development at General Electric, spinning out of the GE Global Research Center. GE Ventures is one of the investors in Menlo Micro, along with Corning, Microsemi, and Paladin Capital Group.

Menlo Micro says the device can be used in automated test and measurement systems, broadband power amplifier impedance matching, electronically steerable antennas and phase shifters, and high-power tunable resonators and filters for military/aerospace, scientific and medical, and wireless infrastructure applications.

Fig. 2: Menlo Micro’s MEMS device. Source: Menlo Micro

Proteus Digital Health showed off its Proteus Ingestible Sensor, which is taken with medication and is accompanied by the Proteus Wearable Sensor Patch. A mobile device application uses the Bluetooth wireless protocol to download data from the patch.

The sensor and the patch keep close tabs on when a patient takes prescribed medications, checking to see if they are taken at the right time and in the correct combination.

Andrew Kelly, IC and systems architect at Cactus Semiconductor, and Jonathan Withrington, IC design manager for Proteus Digital Health, jointly gave the presentation. Cactus acted as the foundry that fabricated the Proteus Ingestible Sensor.

The ASIC that went into the digestible sensor had to meet a die size requirement of less than 1 square millimeter. The ASIC die was sandwiched between layers of copper and magnesium through metal deposition. The tiny chip reacts with gastric fluid in the stomach, forming an “inside-out” battery to power the device. An insulator disk increases the electric field from the chip, according to Cactus.

One-time programmable memory is incorporated in the ASIC, and each ingestible sensor has a serial number to qualify as an FDA-approved medical device. Wafer-scale processing was used in making the ASICs. Each wafer contained 25,000 die, and each lot comprised 600,000 die.

“First and foremost when you think about something you have to ingest is how in the world are you going to power the device,” said Kelly. “It’s a tiny little device that has to fit inside of a pill. You can’t use a traditional battery. You can’t use wireless charging because of the sizes of inductors and the signal strength you get. Of course, once you figure that out, whatever you come up with has to be safe and nontoxic. So those are serious challenges. Once you get past that, you still have a giant problem with the signal propagation. You’re going to have a tiny little pill inside your stomach to be sensed by something on the outside of your body. So the contents of your stomach, the size of the body – so many variables, and lots of attenuation in that signal.”

On top of that is the need to program these devices.

“We’re trying to put one of these in every single pill, so you need an astronomical number of unique serial numbers,” he said. “And then there is data integrity. If the data isn’t correct, all of this really means nothing. The whole idea here is to get accurate, objective data, and then make some judgments on it. There are more challenges. You get into the reliability. It’s basically a medical device, and we have to be confident that every single pill is reliable and works, and actually gives us the answer we need. It has to be highly manufacturable — we need a good supply continuity — because we’re going to be in very, very high-volume production and we want it to continue. We can’t have any interruptions in that. And, as always in my world, cost has to be as low as possible. But realistically, the cost has to be very insignificant relative to the cost of the drug itself. Otherwise, it doesn’t make sense. Some of these things we addressed specifically in the second round of the ASIC [design], where we added some functionality and performance. And of course, we were not allowed to increase the die size at all.”

Kelly said it all boils down to three challenges. “I’ve got a very low-voltage, high-impedance supply, so very non-ideal supply from an electronics point of view. I need a really high data integrity communication link. And it has to be tiny.”

Mitchell Lerner, vice president of production for Nanomedical Diagnostics, described how his company developed its system for characterizing kinetic binding of biomolecules. The big challenge was incorporating graphene as a material on its 200-millimeter wafers. Medford, Ore.-based Rogue Valley Microdevices served as the MEMS foundry for the wafers.

Rogue Valley Microdevices would fabricate the wafers, followed by lift-off photolithography, electron-beam processing, and a cleaning step before employing chemical vapor deposition to grow the graphene on an industrial scale. This would be followed by eight more process steps. Optical inspection would determine the graphene quality control.

With enough high-quality die, Samtec would do the chip packaging, Concisys would do the printed circuit board assembly, and Nanomedical would do the final assembly.

The process allows for 50 wafers per month to be produced, and there are plans to scale that to 200 wafers per month, Lerner said.

Looking beyond biosensors, Nanomedical wants to improve thermal stability and heat transfer in batteries to increase their lifespan; functionalize the sensor chip system to detect electrolytes and metabolites such as cortisol, a biomarker for stress; combine the product with fitness wear or conventional electronic systems to extract biometric data for analysis of a person’s physiological status; and replace indium tin oxide as a capacitive touch-screen element.

Growing interest in piezo
Alissa Fitzgerald, founder, CEO, and managing member of A. M. Fitzgerald & Associates, identified multiple emerging technologies in MEMS and sensors. There are sensors based on thin-film bulk acoustic resonators and surface acoustic waves; near-zero power or event-driven sensors; and piezoelectric sensors, she said. There are also CMOS+ sensors, novel piezoelectric materials, and sensors made with paper or plastic substrates.

Jérémie Bouchaud, director and senior principal analyst for MEMS and sensors at IHS Markit, also pointed to piezoelectric MEMS as an up-and-coming technology, although it is not as widely used as capacitive sensors at present.

Matthew Crowley, CEO of Vesper Technologies, touted the future of piezoelectric MEMS microphones. “Voice is the next technology megatrend,” he said. Unlike capacitive MEMS, which are hermetically sealed, piezoelectric MEMS have a diaphragm and a backplane, working with linear transduction, he noted. The ZeroPower Sensing enabled by piezoelectric properties will “be good for IoT devices,” Crowley stated.

Lars Reger, automotive chief technology officer for the Automotive Business Unit of NXP Semiconductors, went back to the future for auto applications of MEMS. The three megatrends in automotive electronics are connectivity, electrification, and autonomy, he said.

He pointed to the “Vision Zero” program initiated in Sweden two decades ago that sought to reduce auto fatalities through conventional means, such as stricter law enforcement, re-engineering roadways, and urging drivers to slow down. The worldwide community has an opportunity to approach “Vison Zero” through advanced electronics technology with radio-frequency MEMS and magnetic sensors, along with microcontrollers, drastically reducing the number of auto accidents and fatalities – perhaps not to zero, but a much lower number than at present, since most accidents are caused by their human drivers.

Functional safety and functional security must be taken into full consideration to realize that vision, along with the resolution of the obvious ethical dilemmas posed by autonomous driving technology, according to Reger. “To achieve the failure-free model, you need the best sensors,” he said.

Related Stories
MEMS Market Shifting
More precision, new materials and a much smaller universe of companies help to make this market attractive again.
MEMS: Improving Cost And Yield
Second in a series: New packaging options could help boost profitability, but testing and thermal issues remain problematic.
MEMS: A Tale Of Two Tough Markets
First in a series: Challenges increase across the board, but for different reasons. Low profits temper investments.
How Testing MEMS, Sensors Is Different
These devices require more than an electrical input and output.
What’s Up With MEMS?
New technologies and different ways of integrating more intelligence are generating new opportunities in the MEMS market.

  • mirk

    Great article.