Soft-matter researchers are always looking for new and better ways to coax colloidal particles into organized structures. An oft-cited goal is to create three-dimensional photonic bandgap crystals, which can control the flow of light by blocking transmission at certain wavelengths. A colloidal crystal’s structure and the refractive index of the particles in it determine whether it has a bandgap, and the size of the particles loosely determines which wavelengths are affected. The micron-scale building blocks in colloidal crystals produce bandgaps in the visible to near-IR, a particularly useful part of the spectrum: Some quantum computers use visible-wavelength lasers to switch qubit states, and many optical communications networks operate at 1.5 μm.
Unfortunately, most colloidal crystal structures don’t have photonic bandgaps. That’s why research has focused on the diamond lattice: Calculations show that a colloidal diamond made from available materials would exhibit a bandgap. But producing a diamond lattice has proved unexpectedly tricky because entropy doesn’t favor its open structure.
Now Mingxin He and David Pine, both at New York University, with collaborators at NYU; CNRS in Pessac, France; and Sungkyunkwan University in Suwon, South Korea, have assembled the elusive structure.
The challenge was finding the right colloidal building block. Each carbon atom in an actual diamond has four bonds at specific angles determined by the atom’s electron orbitals. Most colloidal particles, on the other hand, are isotropic spheres. If you try to assemble them by making their surfaces sticky, the number of bonds per particle isn’t fixed, nor are the bond angles. As illustrated in the left column of the first image, using colloidal particles that each have four sticky patches (blue circles) at the desired bond locations constrains the assembly. But that alone isn’t enough: Two bound particles can be rotated with respect to each other, which prevents the formation of long-range crystalline order.
Adding a geometric constraint solved the problem. Instead of having sticky patches on the surface of the sphere, He and coworkers synthesized nonspherical particles with recessed patches (right column in the first image). The sticky patches, shown in light blue, only get close enough to touch and bind if the particles are properly oriented so that the protruding lobes of one particle sit in the creases of the other. The resulting crystals, shown in the second image, are about 40 μm across on average, and 20–40 μm thick. To have a bandgap, the crystal should be at least 10 unit cells, or about 30 μm, in each dimension.
Although the crystals have the right structure, they still lack a photonic bandgap because they’re made from polymer particles whose refractive index is too low. But the researchers have a plan. They’ve already shown that chemically crosslinking the interparticle bonds makes the crystals stable enough to remove them from the aqueous environment in which they’re formed. Sol–gel chemistry could fill the interstitial voids with titanium dioxide, which has a refractive index around 2.6, or atomic layer deposition could introduce silicon, with refractive index 3.4. The colloidal scaffold could then be removed, leaving behind a diamond-lattice void—which, it turns out, has a similar bandgap. (M. He et al., Nature 524, 585, 2020.)
