Source: Fundamentals of Microelectromechanical Systems (MEMS), 1st Edition

ISBN: 9781264257584
Authors: Eun Sok Kim Ph.D.

D. Introduction
Commercial applications of microelectromechanical systems (MEMS) have ever been growing for many decades, starting with
silicon pressure sensors. Some of the commercially successful MEMS include accelerometers, gyroscopes, digital mirror array
of 0.3–1 million movable mirrors, microbolometers for room temperature infrared (IR) sensing or for night vision. The nighttime
vision without ultra low temperature for a camera is possible because of excellent thermal isolation made possible through
MEMS structure, while the large number of movable mirrors is possible due to the unique features of MEMS fabrication
process. There is currently no other technology that will allow an array of 0.3–1 million movable micromirrors on a single chip.

More recently, MEMS microphones have been replacing conventional electret condenser microphones (ECMs), largely due to
the MEMS microphones' ability to withstand the soldering temperature of about 250°C which allows them to be mounted on
printed circuit boards (PCBs) through robotic pick-and-placement. Another MEMS commercialization success has lately been
film bulk acoustic-wave resonators (FBARs) for RF front-end filters for cell phones, as FBARs offer high-quality resonant
characteristics due to air-backing on top and bottom of a piezoelectric film, made possible by MEMS fabrication processing.
The growth of MEMS commercialization has lately been 30% annually, and as of 2020, MEMS market is about a tenth of
integrated circuits (IC) market.

D.1. Next Big Thing

Over the past seven decades, there have been some major technology advancements, starting with mainframe computer and
then minicomputer, followed by notebook tablet and smartphones. These are the unique devices that have made the economy
grow and also impacted people's lives greatly. So, the question in 2021 is what is the next big thing. Internet of things (IoT) has
been hailed as the potentially next big thing, as smart wearables such as smart watches, glasses, and clothes can monitor or
track people's health and activity. Smart home with smart appliances, smart car with collision avoidance or driving assistance,
and smart city with smart parking are the areas where IoT will be needed. In all these cases, though, what are critically needed
is the ability to sense and/or actuate and to compute at the sensor/actuator nodes (or the edges of cloud computing).

One notable example of IoT is a wearable activity tracker, such as FitBit. It contains a microcontroller, MEMS inertial sensors to
sense the wearer's motion, light-emitting diode (LED) and photodetector, analog-to-digital converter, RF communication circuits
and antenna, and battery. It is an intelligent edge device that processes sensed signal and communicate the data wirelessly to
the cloud.

Edge devices may include actuators and/or biomedical platform, in addition to sensors. The number of the expected edge
devices containing sensors and actuators is astronomically large. In comparison, the number of personal computers or laptops
is limited, while the number of smartphones is 10 times more. With edge devices, though, the number quickly reaches hundreds
of billions or even trillion. Thus, the manufacturing costs for edge devices must be very low, likely below $1 per unit. There is
currently no technology, other than MEMS technology, that will allow such a massive number of edge devices to be
manufactured at such a low cost.

D.2. Current MEMS as Edge Devices

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
D.2. Current MEMS as Edge Devices
Already there are many edge sensors. For example, a cell phone has a microphone, temperature sensor, accelerometer,
gyroscope, absolute pressure sensor for measuring ambient pressure at different altitudes, and magnetometer, most of which
are MEMS sensors. Humidity sensor, CO2 sensors, gesture sensor based on ultrasonic sensors, proximity sensor, blood
glucose sensor, blood pressure sensor, autofocusing based on MEMS, etc. (again most of which are MEMS devices) are
already used in smartphones or will likely be in smartphones in near future.

D.3. Emerging MEMS Applications

In addition to IoT, emerging applications of MEMS include sensors and actuators for wearable technology, augmented reality,
or virtual reality; MEMS microspeakers which are now in the market; microfluidic systems and bio-MEMS; and MEMS RF
switches for antenna tuning for 5G communication. In case of 5G communication that uses so many different frequency bands,
antenna will have to perform well for all those different frequencies, and an array of antennas is used for 5G wireless
transceivers. Thus, impedance matching to various antennas is needed. One way to present different capacitances to antennas
for impedance matching is to use a MEMS tunable capacitor or MEMS RF switches in combination with a bank of capacitors to
present different values of capacitance through turning on/off the switches. Another emerging applications of MEMS are in
automobiles, as the auto industry is transforming with driving assistance or self-driving features. There are already many MEMS
sensors in cars such as pressure sensors, accelerometers, etc. However, radar and imaging sensors aided by or with MEMS
technology are likely of high demand in future cars.

D.4. MEMS Industry

Among various MEMS manufacturers, Broadcom became the number one in 2017, over Bosch which sells many different kinds
of MEMS sensors, due to the exploding number of RF front-end filters for smartphones. In 2020, Broadcom is still the largest
MEMS manufacturer in revenue, while Bosch is closely the second. There are many other MEMS companies, almost all of which
are growing in revenue and the number of the MEMS units that they are selling. Most of the current MEMS companies make
their MEMS at their facilities. However, there are also many MEMS foundries which fabricate MEMS based on the designs done
by MEMS companies, similar to silicon foundries. However, unlike silicon foundries (the number of which is rapidly shrinking),
MEMS foundries are many in numbers, because vast majority of MEMS requires quite different fabrication processes. There is
no one standard fabrication process that fits all kinds of MEMS. For example, some MEMS use electrostatic transduction, while
others use electromagnetic or piezoelectric transductions that require quite a different set of materials and fabrication
processes. Some MEMS are manufactured through wet micromachining and some others through dry micromachining. Thus, it
is unlikely that a few foundries can dominate, but rather, there will continue to be many foundries for MEMS design houses.

D.5. MEMS Advantages

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
D.5. MEMS Advantages
Microfabrication process of MEMS is based on silicon microfabrication technology that processes a batch of wafers in a
fabrication run. In a fabrication of one batch of say, about 25 wafers, hundred or thousand sensors on each wafer can be
processed for the 25 wafers. Consequently, at the completion of such a batch processing, there will be 2,500–25,000 sensors.
This mass manufacturability due to batch processing is one of the key advantages of MEMS, because the manufacturing costs
can be reduced tremendously. For IoT, trillion sensors and transducers are expected to be deployed around the world. Trillion
devices will be possible, only if each device costs very little, and MEMS technology is currently the only one that can offer such
possibility.

Another advantage of MEMS is that sensor or actuator can be integrated with IC on a single chip. By having a sensor and IC on
a single chip, parasitic capacitance between the sensor and the IC can precisely be characterized, and the sensor's minimum
detectable signal level can be improved greatly. Also, with IC already integrated on a single chip with a sensor, the integrated
sensor can be small and inexpensive. In other words, through integration of MEMS with complementary metal—oxide
semiconductor (CMOS) on a single chip or in a single package, the size, weight, power, and cost (SWaP-C) can be reduced. One
of the commercially successful MEMS is an accelerometer integrated with signal-processing and temperature-compensating
circuits on a single chip.

D.6. Learning MEMS

Though MEMS is an acronym for a microelectromechanical systems, it covers also microelectrothermal, microelectrochemical,
microfluidic systems, and bio-MEMS. Thus, there are much more than just electromechanical expertise that is needed for an
engineer to thrive in MEMS. Multidisciplinary knowledge is needed, and that is good for an engineer to take a lifetime to learn
and keep on innovating. I myself majored in electrical engineering during my undergraduate years, and when I started my PhD
research in MEMS, more specifically a resonant silicon pressure sensor (as MEMS had not yet been coined in mid-80s when I
started the research), I had pretty good knowledge on microfabrication and circuits but quite limited knowledge on mechanics.
Thus, I had to learn classical mechanics, particularly the statics and dynamics of plates. Ever since then, I have kept on learning
new knowledge and technologies on many other areas such as acoustic, piezoelectric, thermal, and biological principles. This
kind of multidisciplinary knowledge is what makes the job interesting for many decades and also what provides engineers with
unique capability to innovate. With MEMS, it is not only depth but also breadth of knowledge and skills. As we keep learning
new scientific knowledge, new technologies, new scientific disciplines, etc., we see more and further, allowing us to contribute
to humanity through innovative and useful MEMS.

Another aspect of MEMS that excites me is that it is open-ended, with so much application and innovation wide open for well-
educated and motivated MEMS engineers. Engineers and scientists with multidisciplinary knowledge and good ideas will be
able to benefit the world greatly through many new MEMS-based devices and systems.

D.7. Impact of MEMS

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
D.7. Impact of MEMS
Impact of MEMS has been already great, but will be much more than ever, since MEMS technology not only enables product
differentiation but also allows new product capability for transducers. Already many conventional transducers such as electret
condenser microphone, acceleration switch, cooled IR camera, etc. have been replaced by MEMS counterparts, and many more
conventional transducers will likely be replaced by MEMS counterparts. More excitingly, though, many brand new transducers or
systems based on MEMS (e.g., single-cell gene sequencing) will be invented and used.

It is likely that MEMS will be the key technology for the "next big thing," but how big the impact of MEMS will depend on the
people who work on MEMS, and create MEMS devices and their applications. In other words, innovative engineers and
scientists will be the key on how strong MEMS' future impact on human lives will be.

So, let us delve into learning microfabrication and MEMS processing technologies, transduction principles, RF and optical
MEMS, mechanics and inertial sensors, film properties and issues as well as silicon pressure sensors and MEMS microphone,
microfluidic systems and bio-MEMS, vibrational energy harvesting, and interface analog and nonlinear circuits for MEMS.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.