A research team led by Yue Gao at Fudan University’s Department of Macromolecular Science has now pursued a radically different route to this problem. Rather than trying to make batteries smaller and denser, the researchers asked a deceptively simple question: could a robot harvest energy directly from the surface it walks on? The world is full of active materials such as aluminum, zinc, silicon, tin, and lead, and these metals and semiconductors are ubiquitous in pipes, buildings, data centers, and industrial equipment. If a robot’s foot could trigger a chemical reaction on contact with such a surface, the robot’s runtime would no longer be capped by the amount of energy it can carry. The environment itself would become the fuel tank, and every step across a suitable substrate would become a step across a power source.

The team’s answer is an open electrochemical power system built as a crosslinked potassium polyacrylate membrane, attached directly to the microrobot’s foot like a layer of engineered skin. This membrane converts chemical energy from the substrate, together with atmospheric moisture and oxygen, into electricity. The design is deliberately open rather than sealed, which is what allows it to draw reactants continuously from the air and the ground instead of relying on a finite internal supply. The work, published in National Science Review, demonstrates a system that can power more than 10,000 walking steps and remain stable after one million steps, a durability figure that addresses one of the most common failure modes of electrochemical interfaces.

Manufacturing the skin is strikingly straightforward. The membrane is made in a single step by polymerizing acrylic acid monomer, crosslinker, initiator, and potassium hydroxide directly on the robot’s foot. A flexible platinum/carbon film serves as the reaction site where oxygen and water participate in the electrochemical process. The resulting skin can be as thin as 135 micrometers and as small as one square millimeter, dimensions compatible with the smallest class of crawling microrobots. On zinc, the membrane delivers a power density of 133 milliwatts per square centimeter, and on aluminum it reaches 103 milliwatts per square centimeter. Those values are roughly an order of magnitude higher than those of typical microbatteries, and the system retains its current capability even when scaled down to 0.0004 cubic centimeters, precisely the regime where conventional microbatteries collapse.

What makes the achievement technically demanding is that the membrane must satisfy several conflicting requirements simultaneously. It has to hold water in dry conditions, remain flexible under repeated bending, stick firmly to surfaces during locomotion, release cleanly when the foot lifts, and sense the properties of the ground it touches. Water retention is handled at the molecular level: the membrane’s carboxylate groups, potassium ions, and hydroxide ions interact strongly with water molecules, anchoring them within the polymer network. Even at 20 percent relative humidity, conditions comparable to the Sahara Desert, the membrane retains most of its water and continues to discharge after 48 hours. This resilience against desiccation is essential for a device whose entire energy conversion scheme depends on the presence of water at the reaction interface.

The membrane is equally robust against temperature extremes. At minus 20 degrees Celsius, potassium ions disrupt the ordered hydrogen-bond network of water within the polymer, preventing the kind of ice-like structuring that would normally shut down ion transport. The result is that the membrane remains ionically conductive at deep subzero temperatures, with its voltage dropping by only about 0.2 volts compared with room temperature performance. The crosslinked structure also tolerates heat, allowing operation at 80 degrees Celsius. Perhaps most practically, the system is recoverable: after water evaporates, simply adding water restores performance. This combination of cold tolerance, heat tolerance, and reversibility suggests a power skin that could operate across the wide environmental range real deployments would demand, from refrigerated facilities to sun-exposed metal infrastructure.

The electricity generation itself follows an elegant contact-driven mechanism. The membrane produces current when it touches an active substrate and stops the moment the foot lifts. During contact, the substrate loses electrons, while on the platinum/carbon side oxygen and water accept those electrons to produce hydroxide ions. The reaction begins in about 0.1 milliseconds, fast enough to keep pace with a walking gait, and terminates as soon as contact ends. Each walking step therefore refreshes the reaction interface, presenting a clean patch of substrate to the membrane. Crucially, the discharge products remain on the substrate rather than accumulating on the membrane itself. This self-cleaning behavior avoids the performance decay that plagues static batteries, whose electrodes gradually accumulate reaction byproducts over their lifetime, and it explains how the system can sustain more than 10,000 powered steps and survive one million steps overall.

Adhesion posed its own engineering puzzle, and the membrane resolves it with a directional asymmetry that mirrors how a foot should behave. Carboxylate groups create electrostatic adhesion to metal substrates, and in the walking direction this adhesion is strong enough to prevent slipping, providing the shear grip a crawling robot needs to propel itself. In the vertical direction, however, the detachment resistance is low, so the robot can lift its foot efficiently without expending excessive energy or damaging the interface. This combination of strong shear adhesion and easy normal detachment is what allows the microrobot to crawl on inclined aluminum surfaces, a capability that would be impossible with a uniformly sticky or uniformly slippery interface. The material effectively behaves like a well-designed shoe sole, gripping where it should and releasing where it must.

Beyond powering locomotion, the membrane doubles as a sensory organ, reading the ground through ions. Different substrate materials drive different electrochemical reaction rates and different levels of hydroxide consumption, which in turn cause directed migration of potassium ions and produce a characteristic potential signal for each material. Surface roughness adds a second channel of information: rough surfaces deform the membrane and alter the ion distribution, generating a distinct signal of their own. The two signal types are naturally decoupled, with material changes producing stable square-wave signals and roughness changes producing millivolt-level spikes, so the robot can distinguish what it is standing on from how textured that surface is. Equipped with an X-Y electrode array, the membrane can even locate where contact occurs on its surface.

This sensing capability transforms the power skin from a passive generator into the foundation of an active energy-seeking strategy. Because the robot can recognize substrate material and roughness in real time, it can identify energy-rich paths and navigate toward them, pursuing the surfaces that yield the most power rather than wandering blindly. In effect, the robot forages for energy the way an animal forages for food, using its skin both to harvest and to map its environment. For a field long constrained by the tyranny of small batteries, the implications are considerable: microrobots that walk across the metal infrastructure of the modern world, drawing power from pipes, panels, and equipment with every step, carrying no fuel at all and stopping only when the ground beneath them runs out of chemistry.

Subject of Research: A crosslinked potassium polyacrylate membrane that harvests electrochemical energy from metal and silicon substrates to power battery-free microrobots

Article Title: No battery needed: Robot skin harvests power as it walks

Article References: No battery needed: Robot skin harvests power as it walks. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: microrobots, energy harvesting, potassium polyacrylate, polymer membrane, electrochemistry, power density, zinc, aluminum, atmospheric moisture, oxygen reduction, substrate sensing, National Science Review