Son Byung-gi
Modern society, going through a great transition driven by the innovation of science and technology, is rapidly entering an information society. Various signals or pieces of information are sensed or collected by sensors, quickly carried over all sorts of communication means, processed fast and accurately by computers, and provided as the right information at the right time. Recently, the acquisition, processing, transmission, and use of information have expanded explosively — and whoever secures more accurate information first becomes the winner.
Today, sensor technology is used widely — from basic science research to industrial process control and military intelligence gathering, all the way down to everyday home life. Sensor technology is highly interdisciplinary, technologically composite, high-value-added, and has enormous ripple effects. That interdisciplinary and composite nature slowed its innovation, while its high value and ripple effect raised its protective walls and deterred technology transfer. The US, Japan, and the EU — which together hold over 90% of the world's sensor technology — are rushing massive investment into sensor development while pushing strong technology-protection policies. That's because sensor technology is the core of precision measurement and automation, and a pivot of high-level system technology. Thanks to these traits, the market is growing fast and it has become a sharp focus of international technology competition.
With international technology-protection barriers rising ever higher, national self-development is inevitable — but because sensor technology is extremely high-value-added and highly interdisciplinary, a leading role by Korean physics is urgently needed. In particular, leadership in 'device physics' for developing high-quality sensors is absolutely necessary. Under the striking impact of modern cutting-edge science and technology, sensor technology is clearly trending toward miniaturization, multi-dimensionality, multi-functionality, intelligence, and systemization.
Here I will summarize sensor technology trends and FET (field-effect transistor) type microsensors, which are cutting-edge semiconductor sensors.
Sensors, Transducers, and Actuators
The word 'sensor' comes from the Latin 'sens(-us)'. It first appeared in McGraw-Hill's 1967 'English-German Technical and Engineering Dictionary' (2nd ed.) without a definition. In 1974, a definition was added and 'sensor' was entered in the same publisher's 'Dictionary of Scientific and Technical Terms' (1st ed.).
The concept of sensor or the academic system of sensor engineering is still not perfectly established. This is because sensor technology is highly interdisciplinary and composite. In a broad sense, a sensor is any means of detecting external stimuli or various signals. In the narrower, common sense, a sensor is defined as 'a device that selectively captures the quantity of a target to be detected and converts and outputs it as a useful signal (mainly electrical).' A sensor is a device that captures primary information — a detector or sensing element — while a means of converting the captured signal into a corresponding useful signal is a converter or transducer. Sensor and transducer clearly differ in meaning but are often used interchangeably; because devices fusing sensor and transducer functions are commonly developed, the meaning of sensor has become more inclusive recently. Here, 'conversion' is usually received in a limited way — conversion into electrical (rarely optical) signals.
Actuator is another word widely used alongside sensor and transducer. An actuator is a device or means that performs an action in response to the refined signal output by a signal processor. That is, an actuator operates in response to an input signal, or executes actions per command signals. These detectors, converters, and actuators are also broadly accepted in the wider senses of sensing machines, converting machines, and actuating machines.
Humans, however, have actually used sensors long before the word existed — detecting direction with a compass, measuring temperature with a thermometer. Today, through sensors, we can detect infrared, ultraviolet, ultrasound, X-rays, and radioactivity that we cannot see or hear. In a broad sense, we use electron microscopes to magnify things 10 million times and radio telescopes to observe quasars 10 billion light-years away. Radar lets us track moving objects and anticipate enemy aircraft. Sensor technology also detects various biological signals and is being developed as a means to preempt disasters.
A sensor must basically have excellent sensitivity, stability, reversibility, and selectivity — and it must also be good in functionality, applicability, standardization, producibility, storability, and economy. When detecting a specific ion in solution or a specific gas in gas, other coexisting ions or gases' interference must be blocked, and only the target ion or gas must be selectively detected. After operating, a sensor must return to its original state immediately for the next operation, although recovery generally takes some time. This relates to the sensor's hysteresis or memory characteristics, which are also tied to its response speed.
Sensor Technology
Sensor technology gives machines sensing functions and extends the human senses. A machine equipped with sensing functions will inevitably be expensive. Therefore, merging sensor technology with machines creates huge added value. Through sensors, we can also detect things human senses cannot — thereby extending the human senses.
Sensor technology is the core of measurement and automation, and the pivot of high-level system technology. All measurements are enabled by sensors, and without precision measurement technology there can be no high-level control or high-level automation. So sensor technology is the core of measurement and automation. The pivotal technology of the 21st century will be high-level system technology. This is achieved by a harmonious combination of communication, computer, and control technology — but control technology lags far behind computer and communication technology. Without innovation in control, high-level system technology is hard to expect. The main reason control technology lags is that sensor technology lags relatively. Therefore, whether we innovate sensor technology is directly tied to whether we achieve high-level system technology. Because sensor technology is so pivotal, it is a focus of international competition, and technology transfer is extremely difficult — so national self-development is unavoidable.
Sensor technology is high-value-added and its market is growing fast. Securing sensor capability is the road to securing high-tech capability and a bridgehead to producing high-value products. The higher the product grade, the higher the proportion of sensor value in it, and sensor quality sets the price reference of the product. As we enter the information society, the sensor market grows; the world sensor market is currently growing almost exponentially.
Sensor technology is low-volume, many-varieties, but its ripple effect is huge. Sensors are truly diverse. Sensor technology is used broadly — from homes to industrial sites, from school labs to military operations and space exploration. Its scope can expand indefinitely depending on how it is deployed. So sensor technology has enormous ripple effects.
Sensor technology is rooted in basic science and technology, and unfolds from sensor-element technology to sensor-system technology, and sensor-application technology. Sensor-element technology is broadly split into physical, chemical, and biosensors; sensor-system technology into interface, signal processing, intelligence, and systemization technology; and sensor-application technology into public, livelihood, industrial, and special applications.
In 1984 the world's sensor market was just over $20 billion. In 1990 it reached $64 billion, in 1992 it exceeded $110 billion, and it is projected to top $240 billion by 2000. Looking at 1990's demand by country — out of the $64 billion, the US held 38.7%, Japan 23%, West Germany 13%, and Korea only 1.5%. In 1990 the US and Japan held about 75% of world sensor technology, but by 1996 the EU held about 30%. This shows how much the EU was pouring into pivot-technology development based on sensors. Even as US and Japanese investments rose sharply, the EU's share held so high is remarkable.
Sensor technology progresses by improving sensor characteristics, developing new sensors, or developing new signal-processing and application technologies. Lately, however, driven by the striking impact of dazzling advances in basic science and cutting-edge technologies — new materials, precision manufacturing, semiconductor integrated circuits, artificial intelligence, etc. — sensor technology is innovating at tremendous speed and scale. This innovation is clearly showing a new trajectory. Recently, on top of the previous directions of price reduction, mass production, and standardization, it is exploding again into miniaturization, multi-dimensionality, multi-functionality, intelligence, and systemization.
A sensor senses a state, so inserting a sensor must not disturb the target state. That is, the sensor's influence on the original state must be minimized. Active development is underway of intelligent, multi-dimensional, multi-functional sensors that respond quickly in any environment and present the information users need. Some call such intelligent sensors 'smart sensors'. If the current pace of innovation continues, sensor technology will soon reach a high level, and high-level system technology will flourish. Someday, super-sensor systems similar to — or beyond — human sensory organs are expected to appear.
FET-type Microsensors
FET-type microsensors are a general term for ultra-small sensors fabricated by microfabrication such as semiconductor integrated-circuit processes, which operate on the same principle as the FET (field-effect transistor). Created together with cutting-edge technologies — new materials, precision microfabrication, electronic circuit integration, and AI — they have many advantages for sensor miniaturization, multi-dimensionality, multi-functionality, intelligence, and systemization. FET-type microsensors can integrate many sensor elements on a single chip, achieve multi-dimensionality by arraying the same type, multi-functionality by integrating various types, intelligence by integrating intelligent circuits, and systemization by integrating related circuits and devices together — so they are drawing great attention as cutting-edge sensors.
FET-type sensors began to sprout in the 1970s. The most representative is the ISFET (ion sensitive field-effect transistor), a semiconductor microsensor cleverly combining an ISE (ion selective electrode) and a MISFET, with significant advantages over conventional ISEs (fast response, ultra-small and ultra-light, in-vivo and in-situ measurement, single-chip smart integration, etc.). The H-ISFET, a hydrogen-ion sensor, was the first FET-type sensor and is already commercialized as a pH meter.
Figure 2 shows the cross-section of an ISFET. It is the same as a MOSFET except that the gate electrode is replaced by an ion-sensing membrane, a reference electrode, and the solution being measured. The operating principles of ISFET and MOSFET are very similar. The electrochemical potential difference at the interface between the solution and the sensing membrane changes with the ion concentration in the solution, and this potential change shifts the threshold voltage (Vt), i.e. the effective gate voltage (Vg-Vt), which via the field effect changes channel conductance and thus drain current. Measuring the change in drain current lets you detect the change in a specific ion concentration. By interchanging ion-sensing membranes selective to specific ions, various ISFETs can be developed.
Using an ISFET as the transducer and forming a specific bio-functional membrane on its gate, FET-type biosensors can be made. For example, by forming a GOD (glucose oxidase) immobilized membrane on the gate of an H-ISFET, an FET-type glucose sensor can be obtained. In this case, hydrogen ions are produced by the enzymatic reaction of glucose and GOD, and the H-ISFET senses them, thereby sensing glucose. Because this FET sensor is very small, it can be inserted directly into a blood vessel to measure or monitor blood glucose in situ. By forming various bio-functional membranes that serve as enzyme-immobilized films or receptors selective for specific substrates, many kinds of bio-FETs can be developed.
Figure 3 shows various FET-type sensors developed or under development in the author's laboratory. Research is underway on ISFETs, GASFETs (gas-sensitive FETs), PSFETs (pressure-sensitive FETs), HSFETs (humidity-sensitive FETs), OSFETs (optically-sensitive FETs), TSFETs (thermally-sensitive FETs), as well as branches of ISFETs and GASFETs, and various FET-type biosensors — bio-FETs — based on the H-ISFET. The open ends in the figure indicate even broader possibilities. Like the ISFET branches, other FET-type sensors can also diversify.
For chemical sensors or biosensors, even if the sensor itself can be made ultra-small, a micro-reference electrode must be developed at the same time to use them as microsensors. Also, to make the FET sensor at least somewhat smart, the sensor-signal-processing circuit must be integrated on a single chip; to integrate various sensors, reference electrodes, and signal-processing circuits on a single chip, the IC design and process technology must be developed. Micro-reference electrodes, FET-sensor interface circuits, and wafer-level fabrication-process technology have been secured or are being developed at a substantial level. Research on FET-type sensor technology is being backed by enormous global investment and passion thanks to its high technical potential. So high-level research results will pour out in the near future, and FET-type microsensor technology will lead a brilliant era of high-level sensor technology.
Figure 4 summarizes the evolution of sensor technology. At the current stage, sensors are mostly developed independently and their output signals are conditioned, passed through an A/D converter, and processed by a microprocessor. In the next stage, those steps gradually integrate, and eventually everything from sensor to microprocessor becomes a single integrated whole. Afterward, each step is developed to a higher level and their integrated result evolves into an extremely sophisticated sensor system. Figure 5 illustrates the concept of an intelligent image sensor with a multilayered structure that may be realized through 3D integration technology in the future.
Sensor technology is the core of automatic control and the pivot of the high-level system technology that will be the central technology of the early 21st century. Because it is extremely high-value-added with enormous ripple effects, it is a focal point of sharp technology competition. Protection barriers keep rising, technology transfer is hard to expect, and national self-development is unavoidable. This interdisciplinary, composite sensor technology will achieve efficient innovation only when basic science, applied technology, national understanding, and strong government support converge through collective cooperation.
References
[1] P. Bergveld, IEEE Trans. BME-19, 342 (1972).
[2] Son Byung-gi, Journal of the Institute of Electronics Engineers of Korea 18(5), 22 (1981).
[3] Son Byung-gi, "Sensor Technology and Its Development Trends", Ministry of Science and Technology Technology Trend Report, 1-36, 1991.
[4] H. Abe, M. Esashi and T. Matsuo, IEEE Trans. ED-26, 1939 (1979).
[5] Son Byung-gi et al., "Study on the Development of ISFETs and Semiconductor Gas-Sensing Devices (I, II, III)", Ministry of Science and Technology Special Research Report, 1985, 1986, 1987.
[6] Byung-Ki Sohn et al., "Development of FET type ion sensors for in-vivo measurements," Proc. of the 2nd International Meeting on Chemical Sensors (1983), pp. 513-517.
[7] Son Byung-gi et al., "Development of Semiconductor Biosensors (I, II)", Ministry of Science and Technology Special Research Report, 1989, 1990.
[8] H. Yamasaki, "Approaches to intelligent sensors", Microsensors (IEEE Press, New York, 1991), pp. 11-18.
[9] Son Byung-gi et al., "Development of FET-type Urea/Glucose Sensors", Ministry of Science and Technology Special Research Report, 1991.
[10] Son Byung-gi et al., Journal of the Korean Sensor Society 4(2), 7 (1995).
[11] Byung-Ki Sohn, "Recent progress in FET-type microsensor technology" (invited plenary lecture), Proc. 14th Japanese Sensor Symposium (Kawasaki (Japan), 1996), pp. 7-13.
[12] Byung-Ki Sohn and Chang-Soo Kim, Sensors and Actuators B34, 435 (1996).
[13] B. W. Cho et al., Sensors and Actuators B41, 7 (1997)
[14] Byung-Ki Sohn, "Sensor technology — a new perspective,"Proc. KOSEF's 20th Anniv. Sympos. on Issues of Sci. and Tech. in the 21st Cent. (Seoul, 1997), pp. 321-331.
Professor Son Byung-gi received his PhD from the Department of Physics at Kyungpook National University and has been on faculty there since 1965. He has served as Kyungpook National University's Dean of the Graduate School, Dean of the Engineering College and Director of the Engineering Research Institute, President of the Korean Sensor Society, Vice President of the Institute of Electronics Engineers of Korea, Chair of the Council of Directors of Excellence Research Centers, and Visiting Professor at the University of Arizona. He is currently a professor in the School of Electronics and Electrical Engineering, Director of the Sensor Technology Research Institute (ERC), and a member of the Minister of Science and Technology's General Coordination Committee.(bksohn@ee.kyungpook.ac.kr) |
http://cgi.postech.ac.kr/cgi-bin/dept/ce/nano/spboard/board.cgi?id=board01&action=view&gul=99&page=5&go_cnt=1
