EPFL Nanodevice Generates Continuous Electricity from Water Evaporation via Hydrovoltaic Effect

Electricity generation typically requires movement - turbines spinning, photons striking semiconductors, chemical reactions proceeding at electrodes. A device developed at the Ecole Polytechnique Federale de Lausanne does something different: it extracts a continuous electrical current from the thermodynamically unavoidable process of water evaporation, with no moving parts, no fuel input, and no sunlight required.

The technology exploits the hydrovoltaic (HV) effect, a phenomenon in which ion motion through or across the surface of charged nanoscale structures generates a measurable electrical potential when fluid is flowing or evaporating. The EPFL team, based in the Laboratory of Nanoscience for Energy Technology (LNET) within the School of Engineering, has built and characterized a platform that makes this effect continuous rather than transient - a critical step toward any practical energy harvesting application.

The Architecture: Hexagonal Silicon Nanopillars

The device consists of a hexagonal array of silicon nanopillars separated by nanoscale gaps. When water occupies these gaps, the negatively charged silicon surfaces attract positive ions from the water, creating an electric double layer at each pillar interface. As water evaporates from the top of the array, it draws up more water from below through capillary action - a continuous upward flow driven entirely by the thermodynamics of evaporation.

This upward flow of ions along the charged pillar surfaces generates a sustained streaming potential - a voltage that can be harvested as usable electricity. The continuous capillary-driven replenishment of water is what distinguishes this system from earlier hydrovoltaic demonstrations that produced short bursts of current before the device dried out. As long as there is ambient humidity to drive evaporation and a water source to replenish the array, the current continues.

The device has been demonstrated to produce electricity under standard laboratory conditions and at ambient humidity levels consistent with normal indoor and outdoor environments, though the output power density at current scales remains far below what would be needed for most practical applications.

Where Hydrovoltaic Power Fits in the Energy Landscape

The power output of hydrovoltaic devices is small - typically in the microwatt to milliwatt range per square centimeter of device area. This places the technology in a category alongside thermoelectric generators and piezoelectric harvesters: devices suitable for powering low-energy sensors and electronics continuously, without batteries, in environments where conventional energy sources are unavailable or impractical.

Potential applications include environmental monitoring sensors deployed in remote areas, implantable medical devices where battery replacement is dangerous or impossible, and distributed Internet of Things nodes in buildings where humidity is consistently present. The technology would not displace grid electricity or compete with solar panels for building energy supply.

The hydrovoltaic effect has a particular advantage over solar harvesting in enclosed or shaded environments where light is absent or intermittent. Humidity is nearly always present indoors and in sheltered outdoor locations where solar harvesting fails.

The Engineering Challenges Ahead

Several significant obstacles stand between the current laboratory demonstration and deployable devices. Power density needs to increase by orders of magnitude for most useful applications. The silicon nanopillar fabrication process used in the LNET platform relies on cleanroom lithography, which is expensive and not easily scalable to large surface areas or high-volume manufacturing.

Long-term stability is also unproven. Continuous operation in a water-evaporation system raises questions about mineral deposition on nanopillar surfaces as ions concentrate, potential oxidation of silicon surfaces that could alter the surface charge properties, and the behavior of the system at varying humidity and temperature conditions. These are engineering problems rather than fundamental physics obstacles, but they represent real work required before practical devices are feasible.

The LNET team's 2024 platform was explicitly positioned as a research tool for studying the hydrovoltaic effect rather than a prototype device. The 2026 findings extending the work to continuous operation represent a meaningful advance within that research program, advancing understanding of the mechanism and demonstrating sustained performance.

Passive Energy Harvesting as a Scientific Frontier

The broader context for this work is the field of ambient energy harvesting - converting the low-grade thermodynamic gradients and flows present in everyday environments into useful electricity. Water in the atmosphere represents an enormous reservoir of thermodynamic potential that human technology has barely touched. The EPFL work is one of several research threads exploring how nanoscale materials and surface chemistry might change that.

Other groups are pursuing hydrovoltaic effects in carbon nanotube films, graphene oxide membranes, and cellulose-based materials. Each material brings different trade-offs in cost, scalability, power density, and stability. The silicon nanopillar approach has the advantage of building on mature semiconductor fabrication knowledge, potentially allowing integration with other silicon-based electronic components.