A water droplet falls onto a lotus leaf and bounces off without leaving a trace. Nature has perfected this self-cleaning marvel over millions of years, and engineers have long tried to replicate it. But if the droplet hits too fast, it shatters into splashes, losing mass and leaving wet marks on the surface. It was thought that avoiding splashes at high impact speeds was impossible, until a bit of polymer was added to the water.

In the new experiments, droplets of a polyacrylamide solution with a diameter of 2.5 mm fell onto a superhydrophobic surface (the kind from which water bounces off) at speeds up to 3.4 m/s. At Weber numbers (a measure of impact force) above 136, ordinary water splashes, but a viscoelastic liquid, one that flows and springs back at the same time, behaves differently. The droplet stretches into a thin vertical filament, at the end of which a 'balloon' inflates, and this balloon lifts off the surface entirely. Not a single secondary droplet. A balloon droplet that inflates itself and lets go.

The secret lies in the surface's micro-gaps. When the droplet impacts, the liquid momentarily penetrates these pores: the Cassie-Baxter state (where the droplet rests on air pockets) transitions to the Wenzel state (where the liquid fills the micro-gaps). This impregnation enhances adhesion and triggers filament growth. Numerical simulations confirmed that the pressure peak during impact is tens of times higher than the pressure during Worthington jet formation, so impregnation occurs precisely at the moment of contact. From there, inertia and gravity take over: the maximum filament length grows linearly with the Weber number, and the center-of-mass trajectory follows a simple ballistic model, with dissipative forces an order of magnitude smaller than gravity.

The result overturns previous assumptions. It was thought that polymer additives suppress rebound and that splashes are inevitable at high impact speeds. Now it is shown that viscoelastic droplets can rebound completely without mass loss even under very strong impacts. The secret is the balance between elastic stresses and capillary pressure, which prevents the filament from breaking before detachment. This opens the way to designing surfaces and liquids for inkjet printing, pesticide application, and cooling of hot surfaces, anywhere avoiding splashes and material loss is crucial. The next step is to systematically measure the threshold for the transition to the balloon regime as a function of roughness and wettability, and to test other polymers. The balloon droplet, lifting off the surface without a single drop of loss, is a beautiful reminder of how elasticity can turn a destructive impact into a controlled flight.