Symmetrical crystals with asymmetric light absorption

In a surprising twist, researchers have identified crystals that are symmetrical but nevertheless absorb light as if they were chiral (Science 2025, DOI: 10.1126/science.adr5478).

The tiny crystals of Li₂Co₃(SeO₃)₄ (LCSO) may herald a new class of materials that could be used in chiral lasers and other advanced optics applications. “It helps open a new mindset on how to think about crystals and chiroptics,” says Roel Tempelaar of Northwestern University, who led the theoretical aspects of the research.

The discovery involves circularly polarized light, whose electromagnetic waves twist around their direction of travel like a corkscrew. These corkscrews can be either left- or right-handed—they are mirror images that cannot be superimposed—which makes the light chiral.

Molecules that are chiral tend to absorb one “corkscrew” better than the other, an effect called circular dichroism, which researchers use to investigate the structures of proteins and other molecules. But conventional wisdom says that a centrosymmetric crystal should not display circular dichroism. That’s because the crystal has an inversion center—its atoms can be reflected through this point to create an identical structure—which generally means it is nonchiral.

LCSO rewrites those rules by combining two other optical effects that both involve linear polarization, where the light’s electric waves are all oriented in a single plane. The first effect is linear dichroism, a difference in light absorption that depends on how a sample is oriented relative to the light’s polarization. The second effect is known as linear birefringence, where polarization and orientation determine how a material refracts light. When these two effects interfere with one another, a combination dubbed LD-LB, they can mimic circular dichroism.

The team used theoretical calculations to identify crystal structures with symmetries that would show the LD-LB effect but not classical circular dichroism. Then they scoured the literature for materials with the appropriate symmetry properties and found LCSO. “This was really the first material that we tried out,” Tempelaar says.

When the researchers made crystals of LCSO and shined circularly polarized light on the samples, they saw the differential absorption they had predicted. Crucially, when they flipped the crystals over, the absorption flipped as well. That simply wouldn’t happen with classical circular dichroism, but it is the hallmark of LD-LB.

“It’s a really beautiful piece of work,” says Sascha Feldmann of the Swiss Federal Institute of Technology, Lausanne (EPFL), who studies the effects of chirality in light–matter interactions and was not involved in the research. “It’s always interesting to rewrite the textbooks.”

LCSO is able to host LD-LB because it is not strictly achiral despite being a centrosymmetric crystal. Although it does not have garden-variety 3D chirality—determined by reflecting its structure in a mirror plane—it does possess 2D chirality based on reflecting the structure through a line.

The LD-LB effect has been known since 1969 and was often regarded as an annoying artifact in spectroscopic measurements. More recently, this effect has cropped up in thin films of organic compounds and perovskite materials. “But the fact that we now see this in a single crystal is important,” Feldmann says. Thin films tend to have disorder or distortion in their crystal structures, he says, which produces weak and variable LD-LB signals. In contrast, the LCSO crystals exhibit a very strong and reproducible effect. “This is now something you can really engineer,” Feldmann says.

Tempelaar and his colleagues hope to create other crystals that perform this chiroptical trick, which could be useful in applications such as chiral lasers. It’s already possible to produce circularly polarized laser light by shining normal laser light through certain filters, but building the chirality into the lasing process itself would help to miniaturize the setup. That approach could prove useful in quantum information systems that rely on generating photons with specific spin states, Tempelaar says.