TechBlog

r. Yu-Chong Tai, professor of electrical engineering and bioengineering at the California Institute of Technology in Pasadena, is an electrical engineer whose early work pioneered a new direction that is now called, “microelectromechanical systems” (MEMS). He has published on just about every facet of MEMS, from shear-stress sensors to micromachining to thermal sensors to lab-on-a-chip. His recent research forays are leading him into studies of biological systems at the micro level. According to our Special Topics analysis of MEMS research over the past decade, Dr. Tai’s work ranks at #5, with 27 qualifying papers cited a total of 272 times. In the ISI Essential Science Indicators Web product, Dr. Tai’s record includes 41 papers cited a total of 383 times to date. Dr. Tai points to some of his earlier papers and presentations, which are outside of the range of our database, as very important in the field. Among these is a presentation report (Fan L.S., Tai Y.C., Muller R.S., “IC-processed electrostatic micromotors,” Tech. Digest, IEEE International Electron Device Meeting [IEDM ’88], San Francisco, Calif., Dec. 11-14, 1988, pp.666-669; and Fan L.S., Tai Y.C., Muller R.S., “Integrated movable micromechanical structures for sensors and actuators,” IEEE Trans. On Electron Devices ED-35:724-730, 1988). Dr. Tai is a graduate of National Taiwan University and received his master’s and Ph.D. in electrical engineering and computer sciences from University of California, Berkeley. He took a faculty appoint at the California Institute of Technology in 1989.

Your work is in microelectromechanical systems (MEMS). Could you explain what this field is?

The name MEMS didn’t even exist in the ‘80s while I was in graduate school. My major was integrated circuits (IC) then. I learned solid-state devices and IC technology. So I know how to make these devices. It all started with an interesting question. We knew that the IC industry was really big in the 1980s. People had already invested billions, if not trillions, of dollars in IC technology. The question was: can we do something with the IC technology for applications other than IC? In other words, IC technology is a huge investment, could something else benefit from it? Here, IC is really only electrical devices. What devices, other than electrical devices, could we build? From an academic point of view, this whole world is either electrical or mechanical. For example, even biology and its fundamental science are all electrical or mechanical. Similarly, chemistry is no different.

So back to the question: Can we make something else? Electrical devices have been largely explored. What else hasn’t been explored? A good answer turns out to be mechanical devices. Let’s think about it more. Electrical devices have become smaller, all the way down to the nano domain. If you look at the mechanical world, miniaturization sort of stopped at the millimeter range. Currently, it is difficult even if you want to make medium mechanical devices with medium complexity at about a millimeter. Besides, mechanical devices, including parts, are usually made one at a time, but integrated circuits are made in massively parallel fashion. Why couldn’t we use IC technology to make mechanical devices? With that in mind, the rest just happened naturally. We, including another fellow student, started to go to the laboratory at UC Berkeley, where we started making mechanical devices, including gears, sliders, cranks, springs, cantilevers, and beams. One of our early papers was exactly on that: how to make very small mechanical parts. We called them micromechanical parts. And then, obviously, this caught a lot of people’s eyes. So micromechanical devices started to show up in conferences and meetings, and a lot of discussion was about micromechanical applications. Most responses were extremely positive then. One criticism, however, got my attention. Just making mechanical parts is not enough; we also need devices that can move mechanical parts. Here, I am talking about actuators—devices that could output mechanical force.

“MEMS is really an enabling technology.”

My Ph.D. project was then to build the first functional, electrically spun micromotors. I demonstrated the first micromotor on a chip that spun at 100,000 rpm. From an academic point of view, it completed a research avenue—micromachines. Not only can we create parts; we also can create actuators that move these parts. Here, I have to credit my advisor at Berkeley, Prof. Richard Muller. He was very smart. He knew the importance of this field. He provided all the resources and he pushed for this development.

After that, a lot of people joined us. We started to build more complex devices. We started to think about new applications, such as pressure sensors, flow sensors, acceleration sensors, acoustic sensors, and optical sensors. They’re all applicable by using this micromechanical technology. In the micro range, there are even more applications. For example, also becoming extremely useful these days is micro optics. The MEMS field benefited the CD and DVD industry. People make micro mirrors for fiber communication. Even in biology, we are making micro devices to study cells. We are building labs-on-a-chip that can analyze DNA in a cell.

MEMS represents microelectromechanical systems. I believe that the time that the name, MEMS really popped up was around the late ‘80s and early ‘90s. One thing that made the name even more famous was that Ken Gabriel, currently a Carnegie Mellon University professor, but a Bell Lab researcher then, decided to go to the Defense Advanced Research Projects Agency and create a research program using the name “MEMS.” Since then, MEMS really exploded; and everybody used that name, at least in the U.S. In Europe, people just call this field microsystems (MS). In Japan, people initially used words like micromachines, microrobotics, or even micro mechatronics. In the end, they used the word MEMS as well.

So, MEMS has been growing for the last two decades and is still growing now because MEMS can be used in every field, such as physics, chemistry, biology, medicine, optics, even aerospace. Today, there are just too many MEMS conferences and meetings. Everyone complains that there isn’t enough time to go to all the good meetings.

How did you become involved in MEMS research at Caltech?

In 1989, I got my Ph.D. and became a faculty member in electrical engineering at Caltech. This field grew so well and so fast that everybody who sees MEMS can figure out some great applications. I collaborate with all sorts of people.

Your two most-cited papers published within the last 10 years are both reviews: “Micro-electro-mechanical- system (MEMS) and fluid flows,” (Ho, C.M. and Tai Y.C., Ann. Rev. Fluid Mechanics 30:579-612, 1998) and “MEMS and its applications for flow control,” (Ho, C.M. and Tai Y.C., J. Fluid Eng. 118[3]:437-47, Sept. 1996). Why do you think these papers are so highly cited?

Fluid mechanics has a long history, but still is very important to our world. However, MEMS for fluid mechanics applications is very new. These two review papers are based on almost 10 years of research making MEMS for fluids applications. During the time, my main collaborator, Prof. Chih-Ming Ho at UCLA, and I have worked on various micro fluids sensing and control using MEMS. For example, we made MEMS devices to control turbulence. We demonstrated that even microdevices can fly an airplane. Through the work, we’ve tried to show people that MEMS can enable “distributed” flow sensing and control, a direction important to fluid mechanics but never possible due to the lack of technology. What we try to demonstrate is that instead of one big device, we can do amazing things using a lot of MEMS devices in fluid mechanics. That’s where we think the future of fluid mechanics is going. We should use more and more devices collectively. In other words, MEMS brings a fresh concept into fluid mechanics, which is very new and very real. I think that numerous people, such as those from mechanical, aerospace, and civil engineering, are all interested in fluid sensing or control. MEMS is good news to them. MEMS can enable them to do a lot of new things. I think that’s the main reason behind the citations.

In addition, MEMS is really an enabling technology. It enables many new fluids technologies, and fluids are important to so many other fields. For example, biology has to deal with fluids and is actually going to have a really bright future with MEMS. We use the term “BioMEMS” a lot these days. Similarly, MEMS helps to bring many other fields into micro. We already have machines (e.g., scanning electron microscopy and transmission electron microscopy) to see very small things, but a true desire is also to be able to make very small things. MEMS enables a brand-new micro world. Of course, we should not stop at micro. We want to go further into nano, and nanotechnology is the natural extension of MEMS into a smaller scale. There’s no surprise that people, including me, are talking about NEMS (nanoelectromechanical systems) all the time now.

What are your particular research interests?

It would be BioMEMS for biology and biomedicine. For the first 10 years after joining Caltech in 1989, my research was MEMS for flow sensing and control. This, however, naturally transitions into biofluid control for biology, biotechnology, and even nanotechnology. These days, I focus more on building devices for biology or biotechnology. Again, I found that biology can really use MEMS technology to advance to the next level up. For example, in biology you have to deal with small amounts of fluid very often, so MEMS for microfluidics control is extremely useful. MEMS can help biology with so many devices, such as microvalves, micropumps, flow sensors, mixing chambers, mixers, chemical reaction chambers, etc. People can understand that if we are able to control microfluids, we have overcome many barriers in biological research.

The research I’m doing now is to merge micro, nano, physics, biology, and chemistry all into MEMS. I am seriously developing labs-on-a-chip to simplify the use of biology or chemistry labs. With complete labs-on-a-chip, one can do biochemical analysis cheaper, better, faster, and more accurately. Errors due to human factors can be eliminated. My lab is building handheld instruments based on labs-on-a-chip. I’m working with doctors on early cancer detection. My recent research results include a complete HPLC (high-performance liquid chromatography) system on a chip. We build a chip with a performance comparable to commercial state-of-the-art machines. On the chip, we do gradient generation, mixing, microinjection, separation, and electrospraying of peptides. That chip is 2 cm x 1 cm. It doesn’t consist of a computer yet. If we work with Intel, they can put all the electronics there and the chip can be mass-produced. This will be extremely exciting! That’s where I see the future for genomics, proteomics, homeland security, water safety, early disease diagnostics, etc. A lot of the research is with doctors at the University of Southern California and UCLA. Again, these ideas are all enabled by the simple question: Why couldn’t we use IC technology for something else?

I note that just over a period of a few days, your two review papers and a few of your research papers from the past 10 years showed an increase in numbers of citations. Is the MEMS field that hot? If so, why?

The MEMS field is exploding. I remember when I was a student, a conference may have had less than 100 people attending. Now there are easily more than 1,000 people. The interest is so high that we can have a conference every week! Many conferences have MEMS sessions. The SPIE (International Society for Optical Engineering) has a MEMS session; even many biotechnology conferences have MEMS sessions.

Miniaturization is a common need for many engineering fields. That’s why MEMS is one of the most powerful tools. Look at Apple’s iPod—music is also getting smaller. Many electronics are going into handheld form. I predict the cell phone will not be just for communication: it will be for bio, chemical, and many other functions. Even Microsoft is trying to put MEMS in handheld devices that could enable more functions in PDAs. The demand for MEMS will only grow bigger.

There are three major areas: Asia, Europe, and North America. Right now, the U.S. actually leads in this field, but Asia and Europe are catching up. The biggest patent generator for the last 10 years was MEMS. These days, it’s nanotechnology. Nanotechnology, however, is still in the research stage, while MEMS has already entered product development.

Now that people are working in nanosystems, how will MEMS research be affected?

MEMS is also transitioning into nano. I actually believe that if you couldn’t master the microworld, you cannot really master the nanoworld. Besides, micro and nano have to go hand in hand in most cases. For example, the atomic force microscope is really a MEMS device that enables probing in the nano domain. Lab-on-a-chip is another perfect example. If you want to manipulate nanomolecules, you actually need MEMS devices. Besides, in the nanotechnology field, the most successful results so far probably are the development of nanomaterials. I can see a new opportunity to combine MEMS and nanomaterials for new sensors. Even for carbon nanotubes, a lot of techniques people use to manipulate them are MEMS technologies. So there’s no lack of new nano applications for MEMS to march into. Nanotechnology actually has to count on the maturity of MEMS. Therefore, I predict that MEMS will become even more indispensable if nanotechnology is successful. The future is bright only if nanotechnologists work with MEMS people.

From another angle, it helps to look at what large companies are doing. Almost all companies that are high on nanotechnology, such as IBM and GM, all have a big MEMS group. Another example is Johnson & Johnson, which just formed a big MEMS group.

Is there anything else you would like to say about MEMS?

I actually believe every biologist should know MEMS. I joined Caltech bioengineering (from electrical engineering) two years ago. I have two bioengineering Ph.D. students and an M.D./Ph.D. student with me right now. I found there are more and more fantastic things I can do with MEMS for biological use. For example, one of my students is making special MEMS to separate cancer cells from blood. Another student is making MEMS to do blood-count-on-a-chip. We are also working on devices to measure cholesterol from a single drop of blood.

We are planning to work with stem cells because we’re able to make devices to trap the stem cells. We are able to make microvalves and pumps to study how cells respond to different chemicals. It’s a whole lab-on-a-chip, instead of using tubes and pipettes. We can separate cells and study their differentiation. We can study what is released into a culture medium from a single cell. MEMS can work on a single cell. MEMS can work on a group of cells. MEMS can also work on tissue and organs. There are many other opportunities. The biologists and doctors should really know the power MEMS holds for them.

Think about the biological instrument revolution. There will be labs-on-a-chip performing complete biology protocols to do diagnostics or prognoses for you. The whole sensor can be a MEMS chip. Imagine if those chips are cheap—$1 each!

Are there products on the market that we are used to that incorporate MEMS?

Every car that’s produced these days already uses many MEMS devices. For example, the heart of the air bag deployment, a sensor that senses a collision, is a MEMS accelerometer. That didn’t exist 10 years ago but it’s in our daily lives. Another example is that Ford is putting tire sensors in every tire. They use MEMS pressure sensors with wireless signal transmission capability. Moreover, hospitals use even more MEMS devices, especially for in vivo monitoring, such as pressure or bloodflow.