A thin elastic film placed on a water droplet can lift and encapsulate the liquid, wrapping it completely under the action of surface tension and elastic bending. The film bends, the water rises, and the system closes into a sealed capsule. It looks like a robust process — flexible material meets liquid, surface tension pulls, film wraps.
It isn't robust at all. Encapsulation occurs only within an extraordinarily narrow parameter window where three energy scales — bending stiffness, surface tension, and gravity — intersect at comparable magnitudes. Too stiff, and the film folds dry without grabbing liquid. Too floppy, and the film recoils after initial contact. Only at the knife-edge intersection does the film successfully lift and hold.
Note (2026-11-15; read against the paper's full text): the essay has the failure modes the wrong way round and the set-up wrong. In the paper an equilateral triangular PDMS film is lifted quasi-statically from a water surface (by a wire at its center), not placed on a droplet. The three morphologies depend on side length and thickness: in its representative snapshots folding occurs for a 98 micrometer film with a side length of 20 mm, recoiling for a thicker 132.25 micrometer film with a side length of 12.5 mm, and encapsulation for the 98 micrometer film with a side length of 10 mm. So the essay's 'too stiff, and the film folds dry' and 'too floppy, and the film recoils' do not match the paper's examples, where the thicker (stiffer) film recoils; I found no hit for stiff, floppy or flexible in the text. The paper's criterion is that encapsulation occurs in a narrow region where the elastocapillary, elastogravity and capillary length scales intersect.
The physics requires a triple coincidence of length scales. The elastocapillary length (where bending and surface tension balance) must match the capillary length (where surface tension and gravity balance), which constrains the film thickness and material properties to a narrow band. Outside this band, the system fails in qualitatively different ways depending on which energy dominates.
This is a pattern common in self-assembly: the most interesting behaviors live in the margins between competing physical effects. Phase transitions happen at boundaries. Biological morphogenesis depends on parameter tuning. And elastocapillary encapsulation — relevant to drug delivery, soft robotics, and microfluidics — requires landing in a sliver of parameter space that is small but precisely defined.
The film can wrap the water. But only if you ask it exactly right. The instruction set is narrow; the result, when it works, is complete.