|
Preface |
Microelectromechanical systems (MEMS) technology is indispensable to mankind’s ongoing pursuit of an intelligent future and our exploration of the microcosm. MEMS is an advanced technology based on the mature semiconductor process that enables miniaturization by properly integrating components and systems with electronic and material functions through the flexible use of various materials and diverse professional knowledge. The MEMS industry has been in development for over 50 years. Until 2006, the focus of MEMS applications lay mostly in the automotive, industrial and IT fields. It wasn’t until two well-known companies, Nintendo and Apple, took the innovative step of incorporating MEMS motion sensing components into consumer electronics to achieve resounding success that MEMS became a household term. Nowadays, it has become an integral part of people’s daily lives.
Many MEMS products have already been successfully commercialized. These include pressure sensors, inkjet chips, digital micro-mirrors, accelerometers, gyroscopes, micro-microphones, room temperature infrared sensors, gas sensors, certain optical communication and radio frequency components and more. According to estimates made by market research institutions, the size of the global MEMS market will reach 31 billion USD in 2023, and its compound annual growth rate (CAGR) between 2018 and 2023 will be about 17.5%. At present, the main industry applications of MEMS are in the two major fields of consumer electronics and automotive electronics. Their market shares are 60% and 20% respectively. The remaining 20% includes telecom, medical, industrial and aerospace applications, etc..
|
MEMS is now recognized as the most forward-looking field of research in the world. It is also one of the most shocking of the 21st century’s star industries. In recent years, the rapid development of 5G and the Internet of Things around the world has also greatly increased the demand for MEMS technology in fields such as smart phones, wearable electronic devices, drones, self-driving vehicles, Industry 4.0 and smart homes. It is expected that this increase will soon usher in an era of explosive growth for the MEMS industry. For this issue of the “New Technology Channel | Collaboration Column”, MA-tek has specially invited Professor Weileun Fang, a top scholar in the field of MEMS research, to write a comprehensive introduction to the application and development of MEMS technology. This article also details the many contributions that his research team has made over the years to the technical development of the two major MEMS components “micro-speakers” and “micro-scanning-mirrors”. The goal is to share with readers the progress being made in the academic research in this important field of science and technology. |
|
Hongren Chen, Director of the MA-tek Technology R&D Center 11/28/2022
More than Moore MEMS Technology! The Miniaturization of Sensors and Actuators
Professor Weileun Fang, Department of Power Mechanical Engineering, National Tsing Hua University, NanoVision Institute
Department of Mechanical Engineering: Shiqi Liu, Songcheng Luo
NanoVision Institute: Haoqian Zhang, Shuwei Zhang
(This article was provided by Professor Weileun Fang; Edited by MA-tek)
-
The microelectromechanical system (MEMS) is a technology for making tiny mechanical structures on silicon wafers through the use of semiconductor-related processes, such as Photolithography, Thin Film Deposition, Doping, and Etching. It can be further integrated with microelectronic components to construct complete electromechanical systems, thus realizing miniaturized structures, sensors and actuators that can be applied to a diverse array of biological, optical, mechanical and electrical fields.
The microelectromechanical system was brought about by the miniaturization and integration of the semiconductor process and offers many advantages, including Batch Fabrication. It enables related products to have a range of competitive advantages, including small sizes, low power-consumption, and low prices. It has also become a key technology for many emerging industries, such as the Internet of Things, Smart-X, and the Metaverse. The scope of and demand for its applications are both rising sharply. Figure 1 shows the French market research company Yole’s prediction of the future trend of MEMS. Due to the nature of mechanical structures, MEMS does not need to follow in the footsteps of Moore’s Law in its pursuit of component size reduction. Conversely, through diversification and what some now refer to as the More than Moore approach, MEMS has the ability to expand the influence and applications of semiconductor manufacturing processes. As such, it has important strategic value and deserves the attention of all relevant domestic players.
Figure 1. MEMS Future Trend of MEMS Sensors[1]。 |
|
MEMS Applications – From Smart Living and the Internet of Things to the Metaverse |
|
The applications of MEMS have been summarized in Figure 2. They can be roughly divided into: 1.Microstructures, such as: Probe cards, probes for atomic force microscopes (AFM), and microfluidic channels for biomedical chips, etc. [2-4] 2.Microsensors, such as: Microphones, pressure gages, accelerometers, gyroscopes, etc. [5-7] 3.Microactuators, such as: The heaters and piezoelectric actuators for inkjet printing nozzles and the mirror actuators for Texas Instruments’ (TI) single-gun projector’s light projection chip, etc. [8-10] At present, most MEMS products are sensors. Early on, they were used mainly for improving automobile safety. Related products included the pressure gages for monitoring tire pressure and the accelerometers for activating airbags. Later, MEMS expanded into commercial electronic products. These included the microphones for hearing aids and the inertial sensors for somatosensory game consoles, etc.. Then smart wearable devices and smart phones made even more extensive use of MEMS sensors, taking advantage of their small sizes and low power-consumption to pack microphones, accelerometers, gyroscopes, magnetometers, pressure gages, hygrometers, gas sensors and more into limited spaces in order to provide consumers with more diverse human-computer interactions, better operating experiences and more comprehensive information. |
 Figure 2. MEMS Application Fields [43] |
|
In recent years, the rise of concepts like the Internet of Things, big data, artificial intelligence, and the Metaverse have brought about a rapid increase in the range of and need for miniature sensors and actuators. From a system perspective, microactuators in particular are indispensable. When invited by the author of this article (Professor Fang) to give a speech at the 2017 International Symposium of Transducers hosted by Professor Fang, Dr. Kurt Petersen, recipient of the 2019 IEEE Medal of Honor [Note 1], predicted that microactuators would be of great importance in the future. **Note 1: This is the highest honor awarded by the IEEE, with only one recipient chosen each year. |
 Figure 3. Schematic Diagram of the Operating Principle Behind HP’s Thermally Actuated Inkjet Printhead [11]
Figure 4. The Micro Mirror Array Used in the Digital Light Processing Technology Developed by Texas Instruments [12-13] |
|
Microactuators and Piezoelectric Films |
MEMS is mass produced and integrated on wafers using semiconductor processes. Therefore, whether it’s in terms of appearance or driving principles, there are very significant differences between microactuators and traditional mechanical actuators, such as internal combustion engines and motors. In the current literature, the four most common types of actuators include: Electrostatic, Electromagnetic, Electrothermal, and Piezoelectric. The aforementioned HP inkjet printheads are electrothermal, while TI’s light projection chips are electrostatic actuators. Due to the better compatibility between electrothermal and electrostatic actuators and semiconductor processes, many studies have made use of these two actuation methods.
|
However, there are problems that need to be solved for each type of actuator. Electrostatic actuators must address issues of free space, electrostatic attraction (Pull-In) effects, and driving voltages, etc. [15]. Electromagnetic actuators often require the assembly of related electromagnetic components, and their operation processes need to take heating and energy-consumption into consideration [16]. Energy consumption and reliability are the hidden concerns of electrothermal actuators, whereas piezoelectric actuators are limited by the difficulty of obtaining piezoelectric films that are compatible with semiconductor processes as well as the stability of piezoelectric films [17-18]. The excellent actuation capabilities and wide application potential of piezoelectric materials, however, have led many important research and commercial institutions to actively invest in the development of piezoelectric thin film process technology in recent years. As such, there have been breakthroughs that have made piezo actuators a much anticipated component.
In short, piezoelectric materials, such as aluminium nitride (AlN) and lead zirconate titanate (Pb(Zr1-xTix)O3, PZT), have the unique characteristic of being able to convert mechanical energy into electrical energy and vice versa. This is known as the piezoelectric effect. According to the energy conversion method, they can be further divided into positive piezoelectric effects and inverse piezoelectric effects. The aforementioned inverse piezoelectric effect converts the imput electrical signal into a linear deformation of the piezoelectric material (converting electrical energy into mechanical energy). Using a suspended micromechanical structure design with multiple film stacks like the one shown in Figure 5 will enable the production of a significant displacement output at the tip of the suspended micromechanical structure. This design can be used as a piezoelectric actuator.
As you can see in Figure 6, the composition of the piezoelectric actuator is actually quite simple. It is a stacked structure consisting mainly of the structural layer, the electrode layer, and the piezoelectric film. The shape and material of the piezoelectric film will affect the characteristics of the piezoelectric actuator. The actuation capability of a piezoelectric film usually depends on the material’s piezoelectric coefficient (electromechanical conversion capability). As shown in Table 1, among the many piezoelectric films, PZT in particular has an excellent piezoelectric coefficient, making it a highly anticipated piezoelectric actuation material. |
 Figure 5. Schematic Diagram of a Suspended Micromachine Structure Piezoelectric Actuator [43] Figure 6. Film Stack Structure of a Piezoelectric Actuator [43]。
Table 1. Piezoelectric Coefficients of Common Piezoelectric Materials |
|
In recent years, the microactuators developed using PZT piezoelectric films that have attracted the most attention are Microspeakers and Micro Scanning Mirrors, which are key components in self-driving cars and the Metaverse. |
|
Microspeakers – A More Immersive Auditory Experience |
|
The demand for True Wireless Stereo (TWS), Smart Glasses, and Metaverse Augmented Reality (AR) and Virtual Reality (VR) products is still on the rise. As such, the auditory components that are key to creating the sense of presence have attracted a great deal of attention.
The sound field can be divided into the free field of the far sound field and the diffuse field of the near sound field according to the application situation, as shown in Figure 7. At present, the direction of the research and development of microspeakers is focused mainly on the diffuse field of the in-ear type of near sound field. In order to allow users to hear more complete and immersive sounds, multi-unit speakers are coupled with a frequency divider that enables a wider range of frequency responses, as shown in Figure 8. Take the TWS, for example. It uses the combination of the magnetically actuated Dynamic Driver and Balanced Armature or Piezoelectric Ceramics to realize a multi-unit speaker. |
Figure 7. Schematic Diagram of the Free Field of Far Sound Field and the Diffuse Field of the Near Sound Field [43]
Figure 8. Schematic Diagram of a Multi-Unit Speaker with a Frequency Divider [43] |
When it comes to the application of TWS, the space inside an earphone is limited (especially in terms of thickness). The three types of speakers mentioned above rely on traditional manufacturing. As such, it is difficult to further reduce the physical volume of the speakers, and the manufacturing limitations are reflected in the acoustic response. This led to the emergence of the microspeaker developed using microelectromechanical technology. For these speakers, increasing the piezoelectric coefficient of the piezoelectric film material has led to the ongoing development of more and more complete manufacturing methods, which allow for reduced sizes (thickness below 1.5mm [19-21]) and lower driving voltages. However, sound pressure is proportional to the total out-of-plane displacement of the diaphragm. In order to meet the sound pressure requirements and maintain performance while reducing the diaphragm size, it becomes necessary to rely on the structural design.
With the advantages provided by microelectromechanical technology, breakthroughs have been made regarding the consistency of products, enabling algorithms to more accurately perform Noise Cancellation. Microspeakers also have the following design considerations: the Frequency Response, Total Harmonic Distortion (THD), and Power Consumption. The frequency response determines the integrity of the sound, the total harmonic distortion is how much the signal differs from the original signal, and the power consumption describes the energy consumed per unit of time, which is especially important for the TWS’ hours of usage.
Take, for example, the components developed by Professor Fang’s laboratory. The three different microspeaker diaphragm designs include the enclosed diaphragm shown in Figure 9 [22], the partially enclosed diaphragm shown in Figure 10 [22], and the cantilever diaphragm shown in Figure 11 [23]. First is the enclosed diaphragm, where the diaphragm is surrounded by fixed ends, and there are no structural gaps. Both the structural design and the manufacturing process are simple. However, its performance is easily affected by residual stress in the film, so the diaphragm rigidity is often greater than that of the original design. As such, its high frequency sound performance is better.
|
Figure 9. Enclosed Diaphragm [22] |
Figure 10. Partially Enclosed Diaphragm [22] |
|
The second type is a partially enclosed diaphragm, which uses the structural design to reduce the rigidity in the plane direction. In addition to increasing the low frequency sound pressure, it is also possible with this design to release part of the residual stress through initial deformation. This diaphragm’s mode of vibration is called the piston mode, and it has the advantage of a greater average out-of-plane displacement. It has also been mentioned in the literature that, compared to the piston mode of the enclosed diaphragm, this design has better output sound quality [24] and consistency. Finally, we have the cantilever diaphragm, a design consisting of a single or multiple cantilevered structures similar to the partially enclosed diaphragm. In this design, however, one end is free. Unlike the previous two designs, it can completely release the residual stress, and its consistency is high. |
Figure 11. Cantilever Diaphragm [23] |
|
By combining the aforementioned multi-unit speakers with the frequency division principle, microelectromechanical technology allows the Woofer and the Tweeter to be placed on the same chip, broadening the bandwidth [23, 25]. Reverse drive technology can then be used to optimize the sound pressure and bandwidth of the multi-driver speakers on a single chip, as shown in Figure 12 [23]. Therefore, the MEMS piezoelectric microspeaker structure can, by integrating multi-unit speakers into a single chip, effectively reduce the size of speakers. It can also achieve a wider frequency response and a faster mechanical response.
In other words, the input electrical signal causes the structure to vibrate and generate sound pressure, which is then transmitted to the human ear. The amount of Group Delay during the electromechanical-acoustic conversion process is low [20]. Not only that, but MEMS speakers consume only about half the power that traditional speakers do. In short, the MEMS speakers allow users to hear a more complete sound, and, by combining that with a lower group delay and longer use time, they can significantly improve the listening experience. |
|
Figure 12. Measurement Results of Multi-Unit Speakers with Frequency Divider [23]
So far, only a handful of piezoelectric microspeaker products have been released on the market, which is still in its infancy. In 2018, Usound published the world’s first piezoelectric microspeaker, Ganymede [20], which is shown in Figure 13. This injected new life into the acoustics industry. In 2020, xMEMS published the piezoelectric MEMS microspeaker Montara [21], which makes use of wafer-level MEMS processes (Figure 14). It offers advantages both in terms of size and production, providing more options for the future speaker market. Overall, the existing products have a lot of room for improvement. However, the infinite potential of piezoelectric microspeakers will surely reshape the ecology that has thus far been dominated by traditional speakers.
|
Figure 13. Usound Ganymede [20] |
Figure 14. xMEMS Montara [21] |
|
Micro Scanning Mirrors –Experience the Visual Integration of the Virtual and the Real |
Approximately thirty years ago, Texas Instruments provided the best demonstration of the potential applications of MEMS actuators through its optical projection chips and, by doing so, opened up a whole new space for imagination [8]. Later, around the turn of the millennium, when optical communications were booming, optical switches based on micro scanning mirrors [26] became a popular key component. As the structures of industries such as display and communications continued to be updated and transformed, however, the application of micro scanning mirrors also saw its ups and downs. In recent years, with the emergence of applications such as smart vehicles and the Metaverse, micro scanning mirrors have stepped once again into the spotlight alongside of emerging piezoelectric technologies. It is believed that, under the influence of the Technology Push and the Marketing Pull, micro scanning mirrors will bring about a whole new landscape.
Figure 15 shows a typical micro scanning mirror. The main microstructures include springs, mirrors and actuators. The activation principle is to use the aforementioned electrostatic, thermoelectric, electromagnetic or piezoelectric actuators to drive the suspended micro mirror and spring, and then perform periodic reciprocating torsion at a specific frequency. At this time, the beam of an external laser light source is directed on the micro mirror. By twisting the micro mirror back and forth, the reflected laser spot can be scanned into a one-dimensional line. With a more complete micro scanning mirror structure or system, as shown in Figure 16, it is even possible to scan the laser light spot into a two-dimensional picture, thus enabling the realization of many applications.
|
|
|
Figure 15. Schematic Diagram of a Typical Micro Scanning Mirror [43]
Figure 16. Schematic Diagram of Combined Fast and Slow Access Micro Scanning Mirrors’ Two-Dimensional Scanning Imaging [43]
Generally speaking, related applications can be simply classified as:
(1) Imaging: The direct use of laser scanning images as projection imaging applications such as micro projectors [27]
(2) Sensing: Applications that use the reflection of a laser-scanned light spot for sensing, such as barcodes and LiDARs [28]
As described above in the explanation of microactuators, most of the micro scanning mirrors published in the early days were driven by electrostatic means due to the compatibility and maturity of the manufacturing process. However, the high driving voltage made it easy to produce the Pull-In effect and other problems. The technology’s applications were therefore limited, especially in the vehicle environment, which has harsh operating conditions and high reliability requirements. More recently, with the gradual maturing and popularization of piezoelectric thin film technology, piezoelectric actuated micro scanning mirrors have come to be considered the key solution for dealing with automotive environments.
|
Micro Scanning Mirror Applications—LiDAR |
In recent years, with the rising trend of smart vehicles around the world, major, internationally-renowned semiconductor manufacturers have been paying close attention to the development of automotive sensors. One of these sensors is the LiDAR, an active telemetry technology that scans the surrounding environment using laser light [29]. It has a long sensing distance and high image resolution. What’s more, its detection capabilities remain good even in dark environments and weather conditions like rain and fog. This makes it a key technology for the realization of self-driving cars.
At present, there are many competing technologies that can enable automotive LiDARs [30-32]. However, in response to the trend of miniaturization, LiDAR systems implemented using micro scanning mirrors have gradually gained more attention. In contrast to traditional mechanical LiDARs, which use rotating motors to steer laser lights, LiDAR systems based on micro scanning mirrors have the advantages of being small, having low energy-consumption and being low cost. As such, they are quite competitive in the field of automotive LiDARs.
|
Micro Scanning Mirror Applications—Automotive Head Up Display |
In recent years, the Metaverse and related topics have become increasingly popular, and related business opportunities for the automobile industry have also received a great deal of attention. Many car manufacturers have tried to apply Augmented Reality (AR) and Virtual Reality (VR) technologies to their automobiles to improve the driving experience. One of these applications is the Augmented Reality Head Up Display (AR HUD), which can present images that interact with the external environment, as shown in Figure 17 [33], thus allowing the incorporation of Metaverse applications into automobiles and opening up new business opportunities.
Micro scanning mirror components are used to control the light paths. Combined with RGB primary color lasers like those shown in Figure 18, they can help realize projection imaging technology. This is known as Laser Beam Scanning (LBS) [34]. Current Head Up Displays mostly use Thin Film Transistor Liquid Crystal Displays (TFT-LCD) to present images. Although the technology is mature and the cost is low, its brightness is inadequate, making the images difficult to see in strong ambient light. LBS technology, on the other hand, uses laser light, so it has the advantage of high brightness. As such, it is very suitable for application in vehicle Head Up Displays. It is a technology with a promising future.
|
Figure 17. AR Head Up Display [33] |
Figure 18. LBS Imaging Technology Combined with RGB Primary Color Lasers [34] |
|
By adjusting mechanical properties such as structural rigidity and vibration mode, the distributions of piezoelectric film and driving electrodes, etc., to achieve better drive components and transmission structure designs, it is possible to obtain better scanning frequencies, scanning angles, and other such performance indicators. |
|
Micro scanning mirror components can be divided into resonant and non-resonant drivers based on their operation characteristics [35]. Based on scanning mode, they can be divided into one-dimensional and two-dimensional scanning mirrors [36]. From an application perspective, there are three key things to consider when designing micro scanning mirrors. These are the scanning frequency, scanning angle, and mirror size. The scanning frequency determines the pixel density and update rate of laser sensing and imaging. The scanning angle defines the field of view of laser sensing and the screen size of optical images. The size of the mirror is positively correlated with the maximum distance and imaging resolution of laser sensing [37].
The challenge for micro scanning mirror design comes from the mutual Trade Off between the three performance indicators of scanning frequency, scanning angle and mirror size. For example, as the size of the mirror increases, its moment of inertia also increases, resulting in a decrease in the scanning angle and resonant frequency. On the other hand, reducing the spring rigidity increases the scanning angle, but the resonant frequency decreases. As you can see, it can be difficult to improve all three indicators at the same time. In view of this, design priorities need to be determined according to the application’s specifications. Therefore, the structural designs of existing piezoelectric actuators are many and varied, as shown in Figure 19 [38-41]. |
Figure 19. Multiple Structural Designs for Existing Piezoelectric Actuated Micro Scanning Mirrors [38-41] |
Speaking of design considerations, let us examine the micro scanning mirror developed by this article’s (Professor Fang’s) laboratory. Figure 20 shows the design of a micro scanning mirror driven by a cantilever piezoelectric actuator. The four sets of bending actuators around the central mirror are the drivers. The input of the driving voltage displaces the end of the cantilever, thereby driving the deflection of the mirror. This design extends the effective length of the actuator using a bent cantilever structure. This enables these actuators to achieve greater end displacement under the same driving conditions, further increasing the angle of rotation of the mirror.
Figure 21 shows a different micro scanning mirror design that uses bilaterally symmetrical half-moon piezoelectric actuators. The twisted spring structure connecting the mirror and the piezoelectric actuator is used to generate vibration energy. The vibration energy is transmitted to the mirror to make it twist in the direction of the plane. The twisted spring’s design has a significant effect on the transfer of vibration energy. Let us look at a more extreme case. When, for example, the spring design’s rigidity is too low, meaning that the spring is too soft, all the vibration energy will be absorbed by the spring itself. In this situation, the energy will not be transferred to the mirror even when the spring is greatly deformed. When the spring design’s rigidity is too high, on the other hand, meaning that the spring is too stiff, the strong link between the mirror and the actuator will limit the movement of the actuator, making it vibrate less effectively.
Figure 20. Mirror Component with Bending Actuator [43] |
Figure 21. Mirror Component with Half-Moon Actuator [43] |
The piezoelectric micro scanning mirror is still in the research and development stage. In addition to the influence that the design of the mechanical structure will have on its performance, the micro scanning mirror will need to withstand long-term, high –frequency operation. As such, the reliability and stability of piezoelectric films are two important keys to reaching commercialization. Many scholars have, therefore, focused their research on the effects of conditions such as environmental temperature and humidity and component cycle testing [42] in the hopes of ensuring the stable operation of piezoelectric components under various conditions. Overall, the piezoelectric micro scanning mirror has excellent optical scanning characteristics and market advantages that have led both industry and academic research teams from numerous countries to invest in its research. Their efforts have accelerated the development of piezoelectric micro scanning mirrors. Assisted by smart vehicles and the Metaverse, the micro scanning mirror has found a stage for itself. Will this component shine in the future? We will have to wait and see.
|
Conclusion |
Microelectromechanical systems realized through semiconductor-related processes have already been widely used in commercial applications. They also have important strategic value and the potential to further expand the influence and competitiveness of our semiconductor industry. As such, this technology deserves the attention of all those in relevant domestic industries. More recently, it is not only micro sensors but also micro actuators in general and piezoelectric thin film micro actuators in particular that have attracted a great deal of attention. This paper presented two promising piezoelectric actuator components – micro speakers and micro scanning mirrors, two actuators found in smart vehicles (including cars, drones, and moving vans, etc.), the Metaverse, and a variety of emerging industries. Its potential has inspired many research teams and enterprises to invest in its development, and related components with excellent performance have come out one after another. Relevant packaging and testing technologies have also become more and more complete. It is hoped that the introduction provided by this article will encourage domestic industries and talents to deploy and explore related technologies as soon as possible.
In recent years, the industry’s optimistic outlook on the potential of piezoelectric thin film sensor components has inspired many instrument suppliers and wafer manufacturers to invest in the development of related key equipment and materials. It is anticipated that this will accelerate the commercialization of piezoelectric thin film sensor components. On the other hand, there are still many ongoing discussions around the mechanical properties of piezoelectric films and their influence on the reliability of related applications. These are all challenges that need to be overcome in the future for successful commercialization. We also hope that domestic industry and academic research institutions will establish testing technologies for these materials and components as soon as possible to bolster our competitiveness in the field of piezoelectric thin film sensor components.
|
Postscript by the MA-tek Editorial Team |
Microelectromechanical system (MEMS) technology began in the 70s, and it has thus far experienced four major surges in industrial development. The first industry boom occurred in the early to mid 80s. At that time, the research on silicon’s piezoresistive properties and silicon material processing was gradually maturing. Silicon piezoresistive pressure sensors and capacitive accelerometers were developed and became widely used in the automotive and industrial fields. The second wave of development arrived in the 1990s during the boom of the personal computer and information industries.
In 1988, Hewlett Packard (HP) began marketing its first thermal-bubble MEMS inkjet chip product. Then Texas Instruments (TI) developed its Digital Micromirrors Device (DMD) in 1996 and successfully applied it in commercial projectors and displays. These two important MEMS products had a profound impact on the entire world. Even now, these two products maintain global market shares of more than 50%, and HP and TI are still the leading players in terms of MEMS market revenue.
In the mid to late 1990s, the rapid rise of the global optical communications industry led to the third MEMS market boom. The growth of the optical communications industry reached its peak around the year 2000. However, because the actual market demand for global broadband communications turned out to be far below expectations, the industry quickly bubbled and entered into a depression, and the development of related MEMS technologies declined with it. The optical communications industry did not really begin to recover until 2004, but the MEMS market lost its previous momentum. It wasn’t until 2006, when Nintendo applied MEMS technology in its Wii gaming console to realize somatosensory control that the fourth wave of application development kicked off in the consumer electronics market.
Many consumer electronics devices (especially smart phones) make use of MEMS components such as accelerometers, gyroscopes, electronic compasses, pressure sensors and miniature microphones to realize system functions such as somatosensory control, positioning navigation, system protection, and voice recognition, etc.. There are currently 1.5 to 2 billion smart phones in the world, and each phone uses more than 10 sensors. What’s more, the demand for such applications in this consumer electronics market with its huge business opportunities is still expanding. However, when this wave of market development reaches saturation in a few years, we will have to ask, what will be the highlights of the next wave of the MEMS industry?
At the 2014 annual meeting of the Taiwan Semiconductor Industry Association (TSIA), Moris Zhang gave a speech on the “Next Big Thing”. He said that the next Big Thing will be the Internet of Things (IoT). IoT business opportunities are expected to emerge in 5 to 10 years, and the beautiful world that IoT promises to build has opened up new room for the imagination. Human beings will use a giant intelligent network infrastructure shared by hundreds of millions of people around the world to carry out comprehensive community communication and information sharing, and, by integrating physical objects with virtual analysis, we will be able to achieve ubiquitous detection, identification, control and service. According to estimates made by McKinsey & Company, an internationally renowned consulting firm, the global IoT market will reach 6.2 trillion USD by 2025. These lucrative business opportunities have drawn many international giants, including Apple, Google, IBM, and Samsung, etc. to begin competing for strategic positions and development.
The overall IoT architecture is basically comprised of the sensing layer, the network layer, and the application layer, and its three key technology categories are sensors, network communications and big data analysis. Sensor technology in particular is the core of the IoT smart infrastructure. In the future, IoT will be realized through hundreds of millions of intelligent sensing devices distributed in and outside of the body and throughout the surrounding environment, which will provide long-term collection and recording of all target information before using advanced analysis technologies to process this huge amount of data. By doing so, it will continuously create predictive algorithms and intelligent automation systems to achieve goals such as improving energy efficiency, providing quality medical services, increasing living comforts and convenience, reducing safety risks and improving productivity.
According to market research, in 2007, there were already ten million sensors connecting various devices to the internet. By 2013, connected sensing devices had rocketed to 3.5 billion. It is predicted that the number of IoT connected sensors will exceed 100 trillion by 2030. Through its use of the advantages of the mature semiconductor wafer manufacturing process, MEMS sensors have high sensitivity, a uniform performance, a low cost, batch production and other features. This makes them suitable for popularization and application in a wide range of consumer electronic industries, and they are bound to become a revenue bright spot in the future IoT sensing application market.
This surge in demand for IoT applications is expected to usher in the next wave of MEMS development. According to market research, between 2018 and 2023, the global IoT market is expected to grow from a revenue of 5.28 billion USD to 22.48 billion USD, achieving a compound annual growth rate (CAGR) of 33.6%, and the market size is expected to increase by more than 300% over the next 10 years. If we wish to distinguish between the points in time during which the demand for IoT sensors might skyrocket, the first surge of MEMS industrial application will likely still be in the mobile device and smart wearable products markets. The second will be the Internet of Vehicles. Then we will really enter the age of infrastructure applications such as the industrial Internet of Things (IIOT) and smart cities with their many and diverse technologies and needs.
The conditions of the environment in which IoT applications must function are harsh, and system deployment is becoming more and more complex. As such, the number of technical difficulties is bound to increase. Critical issues such as network bandwidth, latency, energy consumption, system reliability and more must be overcome one by one as we move into the era of the Internet of Everything. In order to meet various smart interconnection needs, new and more innovative MEMS sensor sensor components will continue to emerge. Integration capabilities will also have to change dramatically. The three major system integration trends that are expected to emerge include Edge Computing with embedded artificial intelligence (AI), built-in self-testing (BIST), and energy harvesting.
Existing IoT systems are still limited to simple monitoring and analysis functions. When AI technology matures, however, it will be able to give IoT systems more efficient object recognition, voice communication, decision-making and judgment capabilities, enabling the upgrade and evolution into the “Artificial Intelligence of Things (AIoT)”. Edge computing is an essential component of IoT systems. Look, for instance, at the most anticipated automated driving system, which relies on edge computing to improve the speed and efficiency of information feedback. The operation mechanism mainly involves collecting data on the surrounding environment via a large number of MEMS sensors equipped to the vehicle body then using the edge computing system, with its high-efficiency computing, to process the information and either send the calculations back to the terminal device or upload them to the cloud as required, thus achieving the best communication and response speed at the same time. This will improve the reliability and safety of automated driving.
Based on observations of the market, an industry ecosystem where MEMS sensors are combined with AI technology is gradually taking shape. Take, for example, the Bosch Sensortec’s release of a wearable device at CES 2021 that used MEMS inertial sensors with embedded AI. Other major MEMS players, including STMicroelectronics, Infineon, ADI, Knowles, and Vesper Technologies, and many emerging AI systems companies, such as Syntiant, Cartesiam and Aspinity, have all taken great strides into the market competition and launched related application products. The high stability of the sensing chip is also indispensible for IoT applications, especially when they are used in areas such as autonomous driving or bio-sensing. Such sensors need to be able to maintain stable operations for long periods of time because they can cause inestimable harm or loss the moment they go wrong. Therefore, the ultra high reliability requirements of IoT sensors will be a major technical challenge in the future. In view of this, the built-in self-testing system will also be necessary for ensuring the security of IoT applications.
MEMS energy harvesting technology is the ultimate solution to the problem of numerous and widely distributed micro sensing chip groups and their power supply management in IoT system applications in the future. In the past, the limited physical size of MEMS energy harvesting components meant that they could usually only generate a small amount of power (less than 1mW). This made them difficult to use. In recent years, however, advancements in nano and piezoelectric material technology have enabled considerable progress to be made in key energy harvesting component technologies. Many well-known companies have spotted the huge business opportunities presented by sensing applications in the IoT market and have already begun laying the groundwork for their futures in this promising field of technology, even going so far as to preemptively launch their own pilot products. The U.S. IoT systems company Wireless Sensor Solutions LLC, for instance, acquired the startup company MicroGen Systems in 2017 to quickly obtain its MEMS piezoelectric energy harvesting technology, which converts environmental mechanical vibrations into electrical energy, in order to create market-leading wireless sensor node products. Then there’s the Japanese Fujitsu company, which developed a composite MEMS energy harvesting component able to simultaneously harvest energy from the two different sources of light and heat and actually applied it to the ultra-low-power-consumption circuits of several of their IC products. According to Fujitsu, they plan to further introduce this technology into households and buildings in the future to realize energy management as well as the smart detection applications of wireless sensor networks for farms.
Many international manufacturers, including ADI, Atmosic, EnOcean, Metis Microsystems, ONiO, Powercast, Renesas Electronics, STMicroelectronics and Texas Instruments, etc., have successively launched different energy harvesting technology solutions. It is foreseeable that, in the future, MEMS micro chips that can operate normally without the need for additional energy will become widely used in human medical implants or in the surrounding environment, especially in wireless monitoring networks that require a large number of sensors, such as ecological environment monitoring systems, disaster early warning systems, and smart living spaces, etc… By harvesting free energy from surrounding sources such as solar heat, human body temperature, the walking motion, radio waves, industrial equipment, vibrations in the roads and more, we will eliminate the need to manage and troubleshoot annoying chip power issues. As MEMS energy harvesting technologies mature, it is believed that we will see the commercialization of chips that combine wireless transmission functions with smart sensors that do not require power supplies. They will spread throughout our living environment, becoming one of the most explosive innovative applications in the MEMS field.
This article focused on providing a comprehensive introduction to key MEMS component technologies and their applications. The author, Professor Weileun Fang, is a leading scholar in the field of MEMS research who has been teaching at National Tsing Hua University for 24 years. His research centers mainly on MEMS sensors and micro systems, and he has published almost 500 papers in international journals and conferences. His outstanding contributions have won Professor Fang many major academic awards both at home and abroad, including three Outstanding Research Awards from the Ministry of Science and Technology, the IEEE Fellow (he was the first Taiwanese scholar elected as a Fellow by the American IEEE Sensors Council), the IOP Fellow (UK) and more. He has also served as the editor-in-chief of top SCI journals in the field of microelectromechanical system and sensor technologies. In addition, Professor Fang actively assists the development of the domestic industry on a regular basis and has cultivated nearly 50 doctoral-level talents who have joined the industry. He has also solved many technical problems in the industry through industry-university cooperation and successfully obtained more than 130 certified invention patents, making remarkable contributions to Taiwan’s industrial innovation and internationalization in the fields of MEMS sensors and IoT. MA-tek is honored to be able to work with Professor Fang this year on an industry-university collaborative project, providing the complete analysis services that his team needs in their research of MEMS piezoelectric sensor components. MA-tek has the comprehensive testing equipment and professional technical experience to fully meet all the various analysis and testing needs for electronic materials, manufacturing processes and packaging.
Reference:
[1]http://www.yole.fr/.
[2]Z. Yanwei, Z. Yongxia, and R. B. Marcus, "Thermally actuated microprobes for a new wafer probe card," Journal of Microelectromechanical Systems, vol. 8, no. 1, pp. 43-49, 1999.
[3]J. Brugger, R. A. Buser, and N. F. d. Rooij, "Micromachined atomic force microprobe with integrated capacitive read-out," Journal of Micromechanics and Microengineering, vol. 2, no. 3, pp. 218-220, 1992.
[4]B. He, N. Tait, and F. E. Regnier, “Fabrication of nanocolumns for liquid chromatography,” Anal. Chem., vol. 70, pp. 3790–3797, 1998.
[5]S. C. Lo, C. K. Chan, M. Wu and W. Fang, “Design and Implementation of Differential MEMS Microphones Using the Two Polysilicon Processes for SNR Enhancement,” J. Micromech. Microeng., 30 055006, 2020.
[6]Y. -C. Liu, M. -H. Tsai, S. -S. Li and W. Fang, "A Fully-differential, Multiplex-Sensing Interface Circuit Monolithically Integrated With Tri-axis Pure Oxide Capacitive CMOS-MEMS Accelerometers," TRANSDUCERS & EUROSENSORS XXVII, pp. 610-613, 2013.
[7]N. M. Nguyen, C. Y. Chang, G. Pillai and S. S. Li, "Design of Piezoelectric MEMS Bulk Acoustic Wave Mode-Matched Gyroscopes Based on Support Transducer," MEMS, pp. 338-341, 2021.
[8]P. F. V. Kessel, L. J. Hornbeck, R. E. Meier, and M. R. Douglass, "A MEMS-based projection display," Proceedings of the IEEE, vol. 86, no. 8, pp. 1687-1704, 1998.
[9]M. Wu and W. Fang, "Design and fabrication of MEMS devices using the integration of MUMPs, trench-refilled molding, DRIE and bulk silicon etching processes," Journal of Micromechanics and Microengineering, vol. 15, no. 3, pp. 535-542, 2004.
[10]A. Vergara, T. Tsukamoto, W. Fang, and S. Tanaka, "Integration of buried piezoresistive sensors and PZT thin film for dynamic and static position sensing of MEMS actuator," Journal of Micromechanics and Microengineering, vol. 30, no. 11, p. 115020, 2020.
[11]J. Stasiak, S. Richards, and S. Angelos, “Hewlett Packard‘s inkjet MEMS technology: past, present, and future,” in Proc.SPIE, vol. 7318, p. 73180U, 2009.
[12]https://www.ti.com/dlp-chip/overview.html
[13]D. Michael, “DMD reliability: a MEMS success story,” in Proc. SPIE, vol. 4980, pp. 1-11, 2003.
[14]H. Wang, L. Zhou, X. Zhang, and H. Xie, “Thermal Reliability Study of an Electrothermal MEMS Mirror,” IEEE Transactions on Device and Materials Reliability, vol. 18, no. 3, pp. 422-428, 2018.
[15]M. Wu, H. Lin, and W. Fang, “A Poly-Si-Based Vertical Comb-Drive Two-Axis Gimbaled Scanner for Optical Applications,” IEEE Photonics Technology Letters, vol. 18, no. 20, pp. 2111-2113, 2006.
[16]J. J. Bernstein, W. P. Taylor, J. D. Brazzle, C. J. Corcoran, G. Kirkos, J. E. Odhner, A. Pareek, M. Waelti, and M. Zai, “Electromagnetically actuated mirror arrays for use in 3-D optical switching applications,” J. Microelectromech. Syst., vol. 13, no. 3, pp. 526–535, 2004.
[17]J. Wang, C. Salm, E. Houwman, M. Nguyen, and J. Schmitz, “Humidity and polarity influence on MIM PZT capacitor degradation and breakdown,” IEEE International Integrated Reliability Workshop (IIRW), pp. 65-68, 2016.
[18]M. F. Bevilacqua, V. Casuscelli, S. Costantini, P. Ferrari, and I. Pedaci, "High Temperature Operating Lifetime Test on piezo-MEMS devices," the 6th IEEE International Forum on Research and Technology for Society and Industry (RTSI), pp. 452-456, 2021.
[19]https://www.knowles.com/subdepartment/dpt-ba-receivers/subdpt-hearing-instruments-receivers
[20]https://www.usound.com/products/
[21]https://xmems.com/products/
[22]H. H. Cheng, S. C. Lo, Z. R. Huang, Y. J. Wang, M. Wu, and W. Fang, “On the Design of Piezoelectric MEMS Microspeaker for the Sound Pressure Level Enhancement,” Sensors and Actuators A: Physical, vol. 306, 2020.
[23]Y. J. Wang, S. C. Lo, M. L. Hsieh, S. D. Wang, Y. C. Chen, M. Wu, W. Fang, “Multi-Way In-Phase/Out-of-Phase Driving Cantilever Array for Performance Enhancement of PZT MEMS Microspeaker,” IEEE 34th International Conference on Micro Electro Mechanical Systems (MEMS), pp. 83-84, 2021.
[24]I. Shahosseini, E. Lefeuvre, J. Moulin, E. Martincic, M. Woytasik, and G. Lemarquand, “Optimization and Microfabrication of High Performance Silicon-Based MEMS Microspeaker,” IEEE Sensors Journal, vol. 13, no. 1, pp. 273-284, 2013.
[25]F. Stoppel, C. Eisermann, S. Gu-Stoppel, D. Kaden, T. Giese and B. Wagner, “Novel Membrane-less Two-way MEMS Loudspeaker Based on Piezoelectric Dual-concentric Actuators,” 19th International Conference on Solid-State Sensors, Actuators and Microsystems (TRANSDUCERS), pp. 2047-2050, 2017.
[26]V. A. Aksyuk, F. Pardo, D. Carr, D. Greywall, H. B. Chan, M. E. Simon, A. Gasparyan, H. Shea, V. Lifton, C. Bolle, S. Arney, R. Frahm, M. Paczkowski, M. Haueis, R. Ryf, D. T. Neilson, C. R. Giles, and D. Bishop, "Beam-Steering Micromirrors for Large Optical Cross-Connects," J. Lightwave Technol., vol. 21, no. 3, p. 634, 2003.
[27]https://www.projectorcentral.com/Microvision-SHOWWX_Laser_Pico.htm
[28]https://velodynelidar.com/what-is-lidar/
[29]H. W. Yoo et al., "MEMS-based lidar for autonomous driving," e & i Elektrotechnik und Informationstechnik, vol. 135, no. 6, pp. 408-415, 2018.
[30]https://velodynelidar.com/products/alpha-prime/.
[31]https://www.continental-automotive.com/.
[32]https://quanergy.com/products/s3/.
[33]https://www.fic.com.tw/zh-hant/automotive/ar-hud/
[34]https://www.st.com/content/st_com/en/about/innovation---technology/laser-beam-scanning.html
[35]S. Holmström, U. Baran, and H. Urey, "MEMS Laser Scanners: A Review," Journal of Microelectromechanical Systems, vol. 23, no. 2, pp. 259-275, 2014.
[36]R. Konsta, M. Dmitry, M. E. Y. Oili, and K. Jukka, "Resonating AlN-thin film MEMS mirror with digital control," Journal of Optical Microsystems, vol. 2, no. 1, p. 011006, 2022.
[37]K. Hwang, Y.-H. Seo, J. Ahn, P. Kim, and K.-H. Jeong, "Frequency selection rule for high definition and high frame rate Lissajous scanning," Scientific Reports, vol. 7, no. 1, pp. 14075, 2017.
[38]T. Masanao, A. Masahiro, Y. Yoshiaki, and T. Hiroshi, "A two-axis piezoelectric tilting micromirror with a newly developed PZT-meandering actuator," 2007 IEEE 20th International Conference on Micro Electro Mechanical Systems (MEMS), pp. 699-702, 2007.
[39]U. Baran, D. Brown, S. Holmstrom, D. Balma, W. O. Davis, P. Muralt, and H. Urey, "Resonant PZT MEMS Scanner for High-Resolution Displays," Journal of Microelectromechanical Systems, vol. 21, no. 6, pp. 1303-1310, 2012.
[40]K. Meinel, C. Stoeckel, M. Melzer, S. Zimmermann, R. Forke, K. Hiller, and T. Otto, "Piezoelectric Scanning Micromirror With Built-In Sensors Based on Thin Film Aluminum Nitride," IEEE Sensors Journal, vol. 21, no. 8, pp. 9682-9689, 2021.
[41]G. Mendicino, M. Merli, R. Carminati, N. Boni, A. Opreni, and A. Frangi, "Electro-mechanical validation of a resonant MEMS mirror with PZT actuation and PZR sensing," Proc.SPIE, vol. 11697,2021.
[42]R. P. Dahl-Hansen, F. Tyholdt, J. Gjessing, A. Vogl, P. Wittendorp, J. Vedum, and T. Tybell, "On the Effect of Water-Induced Degradation of Thin-Film Piezoelectric Microelectromechanical Systems," Journal of Microelectromechanical Systems, vol. 30, no. 1, pp. 105-115, 2021.
[43]方維倫教授實驗室
