Perhaps because it shows how to think about a broad class of nonlinear electrokinetic phenomena in simple terms. Subsequent papers have filled in many mathematical details, but the Letter article emphasizes intuitive drawings and scaling arguments. Another reason may be that it makes theoretical predictions that have stimulated further experimental and theoretical research.

To varying extents, all of the above. The paper unifies different effects, such as nonlinear flow around metal colloids and microfluidic pumping by electrode arrays, in a common theoretical framework. It also predicts some new effects, such as "induced-charge electrophoresis" of asymmetric colloidal particles and "induced-charge electro-osmotic flow" around metal micro-structures, which have since been observed in experiments.

Electrokinetics is the study of electrically driven fluid flow or particle motion. Examples include DNA electrophoresis, which separates molecular fragments by size as they are driven through a gel, and capillary electro-osmosis, where a fluid is driven through a long, narrow tube. Such "electrokinetic phenomena" result from electric fields acting on net ionic charge in solution, which is mostly found in thin "double layers" around charged surfaces and molecules.

For over a century, the traditional assumption has been that the surface charge remains fixed, even after the electric field is applied. This describes "linear" motion, proportional to the field strength (since flow = charge x field), but recently a number of nonlinear electrokinetic phenomena have been discovered in different communities.

Our paper clarifies the common "induced-charge" mechanism, where an electric field acts on its own induced charge near a polarizable surface, and predicts how it can lead to new kinds of particle motion and fluid flows. These effects are attractive to exploit in microfluidics since they involve no moving parts, enable flow control at the micron scale, and require only low battery voltages.

I was inspired by Armand Ajdari’s paper, "Pumping liquids using asymmetric electrode arrays" PHYS REV E, Vol. 61 [1]: pp 45-8, 2000. As the name implies, "AC electro-osmosis" (ACEO) seems to require electrodes applying AC voltages in a certain range of frequencies, but I felt that the underlying phenomenon was much more general.

In late 2001, I came up with the fundamental example of an uncharged metal particle in a suddenly applied constant field, which leads to a persistent quadrupolar flow, and started thinking about broken symmetries that might cause net motion.

Although I had worked on electrochemical dynamics, I was relatively new to fluid mechanics, so I found an able collaborator in Todd Squires, then a Harvard Ph.D. student, and now a professor at Santa Barbara. Todd first thought of applications in microfluidics and made many contributions, including coining the phrase "induced-charge electro-osmosis" (ICEO).

While developing the theory, we waited to publish, since we felt there must be some relevant prior work in such an old field. I contacted leading experts in electrokinetics and gave talks about our results for over a year, before we submitted the Letter paper along with our first long article to Journal of Fluid Mechanics.

Four reviews were mostly positive and supported the novelty of the work, but a fifth referee for JFM pointed us to some Russian papers by VA Murtsovkin, AS Dukhin, and collaborators from the 1980s, which described my original example of "ICEO" flow around a metal sphere, along with experimental observations. In spite of being translated into English, these important papers had never been cited outside the Soviet Union, so we determined to draw attention to them from the new perspective of microfluidics. I think this historical aspect contributed to the impact of our work.

In my research in induced-charge electrokinetics, I have been focusing on three aspects: (i) designing ICEO-based microfluidic pumps and mixers, (ii) collaborating with experimentalists (mainly Todd Thorsen in Mechanical Enginering at MIT) to test basic theoretical predictions in the lab, and (iii) extending the theory of ICEO to "large voltages," greatly exceeding the thermal voltage, kT/e = 25 mV. In my opinion, the latter represents a frontier of theoretical physics with broad implications, not only for electrokinetics, but also for electrochemical sensing (e.g., biosensors) and energy storage (e.g., super-capacitors and thin-film batteries).

The standard model in all of these fields assumes a dilute solution of weakly interacting point charges, but this theory predicts its own demise under a large voltage, since a large normal field (V/nm) inevitably leads to crowding of the ions. I am working on putting steric effects (finite ion sizes) and other crowding effects (increased viscosity) into the theory, which seem to explain puzzling experimental observations, such as the loss of ICEO flow at high salt concentration and flow reversal of ACEO pumps at high frequency. These are just the first steps, however, and unraveling the physics of ICEO at large voltages will likely require a combination of theoretical, computational, and experimental methods.

This is a hard question since the societal impact of one’s work, if any, is rarely clear and usually indirect. Nevertheless, I will hazard to foresee possible implications of induced-charge electrokinetics, which should, of course, be taken with a grain of salt.

Electrophoresis was studied for almost a century before it transformed medicine and criminal law through DNA fingerprinting. Induced-charge electrophoresis offers sensitive new techniques for manipulating polarizable particles with AC fields, which may also find unexpected applications.

For example, it could be used to separate metallic colloids based on size and shape for subsequent assembly in photonic devices or novel electronic materials, leading to faster communication or computation. It could also be applied in biotechnology to manipulate (pull, twist, sort, aggregate, etc.) single cells, DNA or proteins, via attached metal particles, such as gold beads or nanobarcodes.

Perhaps greater implications will come from enabling small, battery-operated, microfluidic devices. Current strategies for microfluidic pumping mostly require either bulky external plumbing (to actuate valves or force fluid flow) or a high-voltage power source (for linear electro-osmosis or electrophoresis), so that microfluidic chips end up being controlled by table-top systems in the lab.

My research with Todd Thorsen is supported by the Army through the MIT Institute for Soldier Nanotechnologies to develop an ICEO-based microfluidic platform for portable or implantable devices. Specific goals include a portable lab-on-a-chip to rapidly detect exposure to toxic agents and a programmable drug-infusion skin patch. Clearly, this research could also impact civilian point-of-care diagnostics and personalized medicine.