I was in fluid science, working on turbulence.

It was because of my background in fluid science, and Professor Tai’s in MEMS, that we started to build a field called microfluidics, which is now the platform technology for biotechnology.

In small-scale flows, the force field changes quite a lot. Viscous dissipation becomes very strong. Electro-kinetic forces and surface tension become effective ways to move fluids. In microfluidic devices, the surface-to-volume ratio becomes very large. In other words, the molecules in the fluid see the influence of the wall all the time, so the surface properties are important in determining the performance of microsystems.

That was still in the early stage of MEMS development, and especially microfluidics. The editors of the journal invited Professor Tai and I to write a paper summarizing the state of understanding at the time. We thought it was a very timely topic.

It’s hard to say that I expected it. You never know. We did expect that a lot of people would read it. Actually, that paper is the second most-downloaded article from the Annual Review of Fluid Mechanics website. I checked that a month ago and it had been downloaded more than 3,000 times.

Well, that was still the very early stage of microfluidics. In the paper, you can see that we never mentioned the idea of using surface tension to control fluids, a technique that was pioneered by Professor C.J. Kim here at UCLA. Now it has become a very popular topic. That’s one thing. Another one is that since then we have started to use microfluidics for many of the biomedical applications. We have a lot of projects in the works now with colleagues in life sciences and also at medical and dental schools.

When we go to see a doctor, for instance, the doctor may request a blood or urine test. They never ask for a saliva test. Now we’ve started to work on a very interesting NIH project with colleagues at the dental school to establish saliva as a diagnostic fluid. Saliva is the easiest body fluid to get at; we don’t need to pinch a finger to get drop of blood. We have early indications that we can gain ample information through saliva by using microsensors.

For instance, if we need to identify whether bacteria or viruses exist in the fluid, which can be saliva, blood, urine, or air/water. A microfluidic system enables us to work with minute sample volumes—such as the nanoliter or microliter range. So how do we detect very diluted bacteria or viruses in the fluid? With well-designed microfluidic processors, it is possible to concentrate dilute targets from large amounts of fluid and to obtain DNA/RNA or protein for biomarker sensors to identify the types of pathogens. Due to the large surface-to-volume ratio in microsystems, modification of the surface property by nanoscale molecules is a key step to make the biomarker truly functional.

When we started working on MEMS in the early 1990s, we could see that the most interesting applications would likely be in biomedicine for one reason: the size of MEMS devices are about the size of a cell. The length scale matching is quite important. For example, if we use our fingers to pick up something, it’s not that difficult to do if the subject has a size of around a centimeter, but it’s much harder to pick up something like a strand of hair. Micron size devices and systems all us to have a playground in the biomedical field.

Inertial microsensors made by MEMS technology have replaced traditional automobile airbag sensors since 1998. The digital optical developed by Texas Instruments has millions of micro-mirrors in a square-inch area that is currently used for image projection in data projectors, movie theaters, and televisions. The use of micro-optical mirrors for internet data transmission can be a major application, but unfortunately when the dot-com bubble burst in 2000, many optical MEMS companies went down with it. The insertion of this technology was delayed, but it will come back.

I think the biomedical applications will dominate the progress of MEMS. The applications of using MEMS for medical diagnosis, drug synthesizing, drug testing, and drug screening will all be interesting developments.

In the past few years, we’ve worked mostly on using MEMS for medical diagnosis. Now we’re getting into the cellular level. For this research, we’re working very closely with colleagues in life science. We’re applying MEMS systems to understand how cells work and how we can control the development path of cells. For example, in terms of a cancer cell, we’re looking at how we can make such a cell die. We’re also looking at the question of how we can control the differentiation of stem cells.

Yes, we have. We’ll be able to discuss it after publishing the work. All I can say now is that it’s far beyond our original expectation and truly fascinating.