Adhesion

Why do two smooth surfaces need a liquid between them in order to stick?

Press two clean, flat panes of glass together and they part again without complaint. Put a film of water between them and they cling so stubbornly that you will slide them apart rather than pull them. That is odd on its face: we have inserted a substance that is itself weak — water has no strength to speak of — and the joint got stronger. What is the water doing that the direct contact could not?

First separate two words that get used interchangeably. Cohesion is a material holding on to itself; adhesion is two unlike surfaces holding on to each other. A joint can fail either way, and which it did is the first thing an engineer looks for: glue left on both faces means cohesive failure inside the glue, glue stripped cleanly from one face means adhesive failure at the interface.

Now the harder question: what does "touching" mean? We picture two flat things meeting flat. But polish a surface as well as you like and it is still, at small enough scale, a landscape of peaks. When two such landscapes meet, contact happens only where peak meets peak, and the true area of contact is a tiny fraction of the area you can see — often a small percentage or less, even under load. This is not an exotic claim; it is the same insight that explains why friction depends on how hard you press rather than on how big the object is.

Why should tiny contact area matter so much? Because the forces available between surfaces are ferocious but absurdly short-ranged. The van der Waals attraction between molecules falls away so steeply with separation that a gap of a couple of nanometres is effectively infinite. So across almost the whole apparent interface, the two solids are simply not close enough to feel each other. The handful of asperity tips that do touch attract strongly — and contribute almost nothing, because there are so few.

Now the water makes sense. A liquid is not there for its strength; it is there because it can conform. It flows into every valley and brings the two solids into molecular proximity across the whole area rather than at a few peaks. Better still, at the edge of the film the liquid surface curves, and a curved liquid surface implies a pressure difference across it — so the meniscus actively pulls the plates together. And squeezing a thin viscous layer out from between two plates takes time, so pulling them apart quickly meets enormous resistance while sliding them sideways meets very little. That asymmetry is exactly what you feel with the wet glass.

Does intimate contact always need a liquid? No — and the counter-example is the useful one. A gecko's toe carries millions of hairs, each splitting into far finer tips a few hundred nanometres across, compliant enough to drape over the roughness of whatever it stands on. It restores contact area mechanically rather than chemically, and the attraction is van der Waals; the effect persists in vacuum, so it is not capillary action in disguise, though humidity does increase it.

Van der Waals plus capillarity is not the whole list. Adhesion in practice runs on four mechanisms, and most real joints use more than one. Molecular attraction across an intimate interface is one. Mechanical interlocking is another — glue seeping into the pores of wood or paper and hardening there is gripping by shape, not by attraction. Chemical bonding is a third: some adhesives form genuine covalent links to the substrate, which is what silane coupling agents are for. Capillary bridging is the fourth, and it is the transient one — it disappears when the liquid dries or drains.

That explains what a glue actually is, and why it has two jobs rather than one. It must arrive as a liquid so it can wet the surface and reach molecular contact everywhere; then it must harden, so that the joint has cohesive strength of its own. A liquid film gives you the first without the second, which is why wet glass resists pulling but not sliding. It also explains the classic failure: a low-energy surface such as PTFE, or a clean surface contaminated with grease, will not be wetted — the adhesive beads up instead of spreading, contact is never made, and no amount of clamping helps.

One honest caveat on a favourite example. Gauge blocks, lapped flat to within tens of nanometres, can be wrung together and will then hang from one another with real force. The mechanism is genuinely contested: molecular attraction across near-perfect contact, a molecularly thin film of oil or water acting as in the glass-plate case, and atmospheric pressure on the outside faces have all been argued for, and the honest answer is that the contributions have not been cleanly separated.

• Why pressure-sensitive adhesives such as tape work without ever solidifying, and what viscoelasticity has to do with peeling.
• How surface energy is measured by contact angle, and why treating a plastic with plasma makes it glueable.

Every term the collection defines is gathered in the glossary.