How Geckos Stick: The Strange Physics of Hierarchical Adhesion

A medium-sized gecko weighs about 50 grams and can hang upside down from a polished glass ceiling by a single toe. The animal can release that grip and reattach it in 15 milliseconds while running. It can climb wet, dry, dusty, oily, hot, and cold surfaces. It can hang from glass at 200 degrees Celsius or at room temperature with equal facility. It does not shed material onto the surface, and the same toe pad works after thousands of attachments without renewal.

The mechanism that makes this possible was a contested question in biology and surface physics through most of the 20th century. Suction was ruled out because geckos work in vacuum chambers. Capillary action from secreted fluids was ruled out because gecko toes are dry. Static electricity was ruled out because geckos work on conductive surfaces. Friction was ruled out because the surfaces include polished glass that has effectively no microtexture for friction to grab. The answer, finally established between 2000 and 2002 by Kellar Autumn's lab at Lewis and Clark, is that geckos exploit van der Waals forces — the weakest of all intermolecular attractions — through a hierarchical structural amplification so extreme that it produces more adhesion per unit weight than any synthetic adhesive ever made.

The hierarchy of structures

The gecko foot is not a single adhesive surface. It is a four-level hierarchy. At the top level are the toes, which the animal can curl and uncurl to engage and disengage the adhesive system. Each toe bears between 100 and 600 lamellae, the visible scaly ridges. Each lamella carries roughly a million setae, each seta a hair-like keratin structure 30-130 micrometers long and a few micrometers thick. Each seta terminates in 100-1000 spatulae, each spatula a flattened triangular tip 200 nanometers wide.

The total per gecko is roughly two billion spatulae across the four feet. Each spatula contacts the substrate on a surface roughly 200 nanometers across. The total contact area in nominal use is a few square millimeters, but spread across two billion separate contact points. The hierarchical structure is what bridges the gap between a foot that conforms to a substrate at the toe-curl scale and adhesive contact at the molecular scale.

Why the hierarchy is necessary

Van der Waals forces are individually feeble. The attraction between a single 200-nanometer-wide spatula and a glass surface a fraction of a nanometer away is on the order of nanonewtons. A 50-gram gecko hanging from a ceiling needs about half a newton of upward force, which is roughly half a billion nanonewtons. The math works out exactly: half a billion spatulae in contact, each contributing one nanonewton, supports the gecko. The actual capacity is much higher — geckos can support 40-50 times their body weight before the attachment slips — because not every spatula needs to be loaded to its maximum.

The reason van der Waals forces work for geckos and not for, say, a flat patch of plastic of the same area is the surface roughness. Real surfaces — even polished glass — have roughness at scales from a few nanometers to a few micrometers. A flat patch of plastic only contacts the surface at a small fraction of its nominal area, because every protrusion on the plastic and every depression on the glass prevents intimate contact. The van der Waals force depends on the gap between surfaces to the seventh power; gaps of even a few nanometers reduce the force essentially to zero. The hierarchical structure of the gecko foot lets each spatula find its own intimate contact with whatever local microgeometry it encounters, multiplying the contact area by orders of magnitude.

The proof, 2000-2002

Autumn's group at Lewis and Clark, working with Robert Full's biomechanics lab at Berkeley and Ronald Fearing's robotics lab at Berkeley, did the experimental work that closed the question. The 2000 Nature paper measured the adhesive force from a single seta — about 200 micronewtons, far higher than anyone had predicted from Hamaker theory — and showed that the force scaled correctly with the number of spatulae. The 2002 PNAS paper showed that adhesion worked in vacuum (ruling out suction), worked on hydrophilic and hydrophobic surfaces with roughly equal force (ruling out capillary mechanisms), and worked on conductive surfaces grounded to remove charge (ruling out electrostatics).

The remaining puzzle was how the gecko releases its feet, given that the same van der Waals forces that produce adhesion would, in theory, prevent detachment. The mechanism is shear-angle-dependent: the spatulae adhere when pulled along their long axis but release when peeled at an angle greater than about 30 degrees. The gecko engages adhesion by shifting its weight forward, loading the spatulae in shear; it releases by curling the toe from the tip backward, peeling the spatulae one row at a time. The whole engage-release cycle takes about 15 milliseconds, which is what makes gecko running on vertical surfaces possible.

The synthetic-gecko race

The biological mechanism inspired a wave of materials science work starting in the early 2000s, attempting to fabricate hierarchical adhesives that mimic the gecko structure. The early efforts — directional carbon nanotube arrays, polyurethane microposts, lithographically patterned silicone — produced adhesives that worked on smooth glass but failed on rough or dirty surfaces, did not survive multiple use cycles, or produced adhesion only at small scales that did not extrapolate to load-bearing applications.

The state of the art as of 2026 is closer than it was twenty years ago but has not yet reached gecko performance per unit area. The best published results — from Metin Sitti's group at Max Planck Stuttgart, Kellar Autumn's lab continuing the original work, and several groups at Stanford and Berkeley — produce adhesives that achieve roughly 10-50% of gecko adhesion strength on smooth surfaces and decline rapidly on rough surfaces. The gap is real, and it points at the role of the gecko's biological growth process: the hierarchical structure is grown organically, with all four levels of structure self-assembling rather than being separately fabricated. Reproducing that self-assembly in a synthetic process has turned out to be much harder than reproducing the geometric end-state.

The applications that have emerged commercially are at smaller scales. Wearable health-monitoring patches, robotic grippers for fragile objects, and surgical attachment devices use hierarchical adhesive technology informed by the gecko research. The military and aerospace applications — climbing robots, repositionable load-bearing attachments — remain at the prototype stage.

The deeper observation

The gecko foot is a worked example of one of biology's most general design strategies: when a physical effect is too weak to be useful at one scale, multiply the contact points across a hierarchy of structures until the cumulative effect becomes large. The same strategy underlies the lung's gas exchange surface (a few square meters of alveolar membrane fits in the chest cavity through hierarchical branching of bronchi), the small intestine's absorptive surface (the villi-and-microvilli hierarchy gives a fraction of an acre of effective absorption area), and the brain's neural connectivity (hierarchical branching of dendrites and axons produces 10^14 synaptic contacts in 1.4 kilograms of tissue). The gecko's exploitation of van der Waals forces is striking because the underlying force is so weak; the lesson is that no force is too weak to be useful if you can build a hierarchy that multiplies the contact area enough.