Imagine a drop of water on a solid, pulled up into a ball by surface tension. The simple application of a voltage polarizes the solid surface and prompts the drop to spread out to minimize the system’s free energy; the greater the voltage the greater the spread. In essence, the liquid–solid interface behaves like a capacitor, and the change in wettability arises from the extra electrostatic energy stored at the charged surface, as outlined in figure 1.
Figure 1. (a) When a drop wets a surface, each interface experiences a surface tension. At equilibrium, the horizontal components of liquid–gas γlg, solid–liquid γsl, and solid–gas γsg tensions must balance, which determines the contact angle θ. (b) When a voltage V is applied between the drop and solid surface, the interface behaves like a parallel-plate capacitor whose capacitance C lowers the solid–liquid interfacial energy by CV2/2, which lowers the contact angle. (Adapted from ref. 3.)
Figure 1. (a) When a drop wets a surface, each interface experiences a surface tension. At equilibrium, the horizontal components of liquid–gas γlg, solid–liquid γsl, and solid–gas γsg tensions must balance, which determines the contact angle θ. (b) When a voltage V is applied between the drop and solid surface, the interface behaves like a parallel-plate capacitor whose capacitance C lowers the solid–liquid interfacial energy by CV2/2, which lowers the contact angle. (Adapted from ref. 3.)
The phenomenon, known as electrowetting, has its roots in Gabriel Lippmann’s 1875 experiments on voltage-induced variations in the capillary action of mercury in contact with an electrolytic fluid. But only recently has electrowetting found wide applicability,1 largely thanks to CNRS scientist Bruno Berge’s realization in the early 1990s that a thin insulating dielectric placed between the electrode and electrolyte would stabilize the effect by preventing oxidation or other electrochemical reactions at the interface.
In the past decade, researchers exploiting the electrowetting effect have developed liquid lenses with voltage-tunable focal lengths for miniature cameras, electronic screens whose pixels wet and dewet on command to alter color or contrast, and other applications. In microfluidics, if the electric field is applied nonuniformly, the gradient in surface energy along a channel can be used to control droplet flow through complex circuits without the need for pumps, mixers, or valves.
Tom Krupenkin and Ashley Taylor at the University of Wisconsin–Madison have now developed an approach that runs the electrowetting process in reverse and converts the mechanical energy of liquid motion into electrical energy.2 Figure 2 illustrates the concept: A conductive droplet and two dielectric-coated electrodes are connected by an electrical circuit that provides a constant bias voltage. As the droplet’s shape or position changes—when it’s squeezed between vibrating plates, sheared along sliding ones, or, as in the figure, slid past fixed ones bordering a fluidic channel—the amount of charge stored at the liquid–solid interface changes.
Figure 2. Reverse electrowetting is a process by which liquid motion may be converted into electrical energy. (a) In the new work,2 a liquid-metal droplet flows into or out of alignment with dielectric-coated electrodes connected to an external bias voltage V, ranging from 2 V to 70 V. As the overlap between the electrodes and the droplet decreases, so does the capacitance, and excess charge flows back through the electrical circuit. (b) The process is scalable, and an applied pressure can force multiple droplets to flow past a series of electrode patches that border a fluidic channel. (c) The University of Wisconsin researchers envision energy-harvesting shoes with fluidic chambers embedded in the heel and forefoot that squirt thousands of droplets through channels during each footstep. (Adapted from ref. 2.)
Figure 2. Reverse electrowetting is a process by which liquid motion may be converted into electrical energy. (a) In the new work,2 a liquid-metal droplet flows into or out of alignment with dielectric-coated electrodes connected to an external bias voltage V, ranging from 2 V to 70 V. As the overlap between the electrodes and the droplet decreases, so does the capacitance, and excess charge flows back through the electrical circuit. (b) The process is scalable, and an applied pressure can force multiple droplets to flow past a series of electrode patches that border a fluidic channel. (c) The University of Wisconsin researchers envision energy-harvesting shoes with fluidic chambers embedded in the heel and forefoot that squirt thousands of droplets through channels during each footstep. (Adapted from ref. 2.)
The droplet thus acts like a variable capacitor. As the areal overlap at the liquid–solid interface decreases, so does its capacitance, and excess charge flows back through the electrical circuit. In the case of a microfluidic channel in which a train of droplets is squirted through one end or the other, the flow of droplets into and out of alignment with the electrodes drives an alternating current and can power an external load. The power is proportional to the change in capacitance and to the square of the bias voltage across the interface.
Energy-harvesting schemes based on variable capacitors are not new, but they’ve historically been limited by the mechanical difficulty of machining macroscopically large electrodes fine enough to nearly close a microscopic gap between them. Capacitance scales directly with electrode area but inversely with gap distance, so the thinner the better, barring dielectric breakdown. “Because the dielectric gap in reverse electrowetting is a solid layer,” says University of Twente physicist Frieder Mugele, “it can be grown nanometers thin and yet remain mechanically stable.” As droplets then flow sideways along the dielectric, they generate huge variations in capacitance on the scale of 20 nF/cm2. “It’s a simple, clever idea, and frankly one I wish I’d had myself.”
Power walking
In one implementation, Krupenkin and Taylor fashioned patches of tantalum oxide–coated electrodes along a fluidic channel a millimeter or so wide. Using a resistor as the load, they measured a few milliwatts from the channel containing up to 22 droplets of mercury. But the power scales nonlinearly with the droplet number. From their model of the process, the researchers calculate that average powers of 1 W or more could easily be generated (at a bias of at least 20 V, admittedly) in a fluidic device with 1000 droplets.
The devices can be small and compact. A 4-m-long train of 1000 droplets, each 1 mm long and spaced by 1 mm, would occupy about 40 cm2, less than a quarter of the area of a typical human footprint. The compression required to squirt the roughly 4-ml volume through channels in a pair of shoes, for example, is about 2 mm—well below the level that might affect a person’s gait.
Conductive droplets don’t behave like ideal capacitors, though, and the envisioned energy harvester has limits. For nanometer-thick dielectric films, even a few volts can generate strong electric fields on the scale of 106 V/cm and trap charges in the dielectric, which inhibits the electrowetting effect. Applying a coat of fluoropolymer to the Ta2O5 dielectric ameliorates the problem, the researchers found, but cannot prevent it.
Nonetheless, Krupenkin and Taylor remain sanguine about embedding their circuits in a pair of shoes to drive mobile electronics such as a cell phone, music player, or emergency flashlight. Fortunately, mercury isn’t the only metal that is liquid at room temperature; galinstan, a nontoxic, liquid-metal alloy of gallium, indium, and tin, has proven equally effective when sealed off in closed channels to avoid oxidation. The two researchers don’t expect the shoes to replace batteries, only to keep them charged and to dramatically extend their lifetime. (For more on the challenges associated with rechargeable batteries, see the article by Héctor Abruña, Yasuyuki Kiya, and Jay Henderson in PHYSICS TODAY, December 2008, page 43.)
So far, they have patented the idea, talked with the US military, and founded a company (InStep NanoPower) but have not yet commercialized the technology.
