Fiber optic cables can transport information at the speed of light. But the ability to perform computations with that data, without translating it back to electric signals, will require a host of new optical components. Researchers at the University of Utah have developed a programmable, chiroptical heterostructure that could enable computers to store and process information using light instead of electrical pulses. The device can dynamically manipulate the circular polarization of light, allowing the degree of circular polarization to be adjusted in real time to enable more efficient optical computing. “Traditional chiral optics were like carved stone — beautiful but frozen,” said professor Weilu Gao, who led the research. “This made them not useful for applications requiring real-time control, like reconfigurable optical computing or adaptive sensors.” The programmable, wafer-scale, chiroptical heterostructure consists of a stack of multiple different thin films, including a collection of aligned carbon nanotubes with different orientations. Courtesy of the University of Utah. The chiroptical heterostructure comprises stacked 1D carbon nanotubes and phase-change materials made from germanium-antimony-tellurium (GST). Once assembled, the layers can selectively reduce the amount of left- or right-circularly polarized light that passes through them, depending on the state of the phase-change material layer. The carbon nanotubes serve both as the layer that produces chiroptical responses and the Joule heating electrode that electrically programs the phase-change materials. “The carbon nanotubes simultaneously act as chiral optical elements and transparent electrodes for [phase-change material(s)] switching, eliminating the need for separate control components,” researcher Jichao Fan said. An electrical pulse along the carbon nanotube layer introduces heat, which causes the internal structure of the phase-change material layer to transition from amorphous to crystalline. This change modifies the heterostructure’s circular dichroism, giving the structure the ability to absorb different types of circularly polarized light at different strengths. The researchers can control which direction the circularly polarized light twists by modifying the circular dichroism of the heterostructure. Because a phase change can be initiated by an electrical pulse, the heterostructure’s circular dichroism can be adjusted in real time. “By adding circular dichroism as an independent parameter, we create an orthogonal information channel,” Gao said. “Adjusting it does not interfere with other properties like amplitude or wavelength.” Consequently, light could be used for data encoding and storage while other properties of the light are used in parallel. The researchers additionally developed a software infrastructure based on high-performance machine learning frameworks, including differentiable programming and derivative-free optimization, to optimize the tunability of circular dichroism responses in the heterostructure. They used advanced manufacturing techniques and AI-assisted design to assemble the stacked layers of the heterostructure without degrading the optical properties of the individual layers. And they used scalable manufacturing techniques for the nanotube and phase-change material films to allow the heterostructure to be built on the wafer scale. Different 1D nanomaterials, phase-change materials, and electro-optic materials can be used with the heterostructure to explore various chiral phenomena and develop both photonic and optoelectronic devices. The chiroptical heterostructure could serve as a multifunctional, reconfigurable component of optical computing systems and could be used to develop high-performance devices such as quantum light emitters. “We’ve created ‘living’ optical matter that evolves with electrical pulses, thanks to our aligned-carbon-nanotube-phase-change-material heterostructure that merges light manipulation and memory into a single scalable platform,” Fan said. The research was published in Nature Communications (www.doi.org/10.1038/s41467-025-59600-w).