During a jaunt through Florence's Natural History Museum this morning, I spent a long time gazing at its 18th-century collection of butterflies. While many other creatures displayed in the museum looked ancient and faded, the butterfly specimens were just as brilliantly blue, white, neon green, and multi-colored as if they'd been netted and pinned up yesterday.
Why should this be the case? It turns out that many butterfly wings aren't colored with pigments which break down or brush off over time; rather the colors arise from the nano-scale structure of the wings themselves.
If you zoomed in on the surface of such a wing, you would see a forest of identical structures a few hundred nanometers tall and apart. These strange structures, made of chitin like the rest of the wings, are spaced regularly across the surface in a periodic array. When light strikes them, the size, shape, and spacing of these structures determines which wavelengths get scattered, and which are allowed to propagate along the surface. The color of the propagating light is what you see.
Physicists and engineers have been trying to construct 3-D photonic crystals - periodic structures which control the flow of light - for the past few decades. To affect light in the visible range, their crystals must have features with sizes on the order of visible wavelengths: namely, several hundred nanometers. Features that are much bigger or much smaller than the wavelengths of light aren't "felt" by it. Light gets bent and diffracted by objects on its own scale, but engineering arrays of such small objects has been exceedingly difficult for humans.
Not so for nature. Butterflies display a huge range of colors and patterns, all coming from different shapes and arrangements of their nano-scale structures. For example, bright green originates from an array of spiraling gyroids. The gyroid shape is known to mathematicians because it has the minimum possible surface area for a certain set of boundary conditions. And it appears by the million, in impossibly miniature form, on the wings of butterflies who know nothing about boundary conditions.
Engineers at Yale have recently started using butterfly wings as templates for modeling their own photonic crystals. I used to work in a nonlinear optics lab myself. Our interest was in light manipulation, mainly using photonic crystal fibers. My research advisor kept a Morpho butterfly wing in a petri dish on his desk for inspiration.