The era of artificial intelligence demands ever-increasing processing speed and energy efficiency. While optical analog computing offers inherent advantages in these regards, most existing optical differentiators are static and lack the reconfigurability required for dynamic, multi-task scenarios.
In a new paper published in Light: Science & Applications , a group led by Professor Yan-Qing Lu and Associate Professor Peng Chen from Nanjing University, China has proposed a reconfigurable ferroelectric chiral nanostructure to dynamically control optical differentiation. Based on ferroelectric liquid crystals (FLCs), they designed a fast-switchable optical differentiator that achieves remarkable performances in edge detection.
The heart of their innovation lies in the FLC superstructure. Under proper electric fields, the natural helix of the FLC is totally suppressed, and the rotational direction of molecules is determined solely by the polarity of the applied electric field. By carefully designing the spatial distribution of the optical axis and using a photopatterning technique, the team created a device that can switch between 1 st /2 nd -order optical differentiation (under positive voltage) and bright-field imaging (under negative voltage). The switching time is as short as 62 microseconds on average (under 2 kHz) — three orders of magnitude faster than conventional nematic liquid crystal devices.
“The FLC differentiator demonstrates outstanding exceptional reliability and durability.” Associate Professor Peng Chen pointed out. “After more than 1.8 million switching cycles, the device maintained identical performance. Even after 200 days, the response curve remained unchanged.” The device also exhibits robust performance against temperature fluctuations and humidity variations across a broad operational bandwidth (490 nm to 630 nm).
“This differentiator is not only suitable for light intensity objects, but also for edge detection of phase objects.” Wen Chen, the Ph.D. candidate supervised by Prof. Peng Chen and the first author of this study, says. The group successfully demonstrated the applicability for both biological cells (e.g., onion epidermal cells) and standard resolution test chart. The fast-switchable differentiator clearly identified edges, while simultaneously providing direct bright-field images when needed.
The FLC heliconical superstructure offers a marvelous platform for advancing optical information processing toward more intelligent and real-time operations. The optionality for either spatial differentiation or bright-field imaging efficiently provides distinct and comprehensive morphological information of target objects. The device can be easily incorporated into conventional imaging systems such as microscopes, and its ultra-fast response makes synchronous observation possible. “The FLC differentiator promises to enable visualization of live objects and accentuation of regional boundaries in biology and diagnostics.” says the group leader Prof. Yan-Qing Lu. “This work explores the potential of ferroelectric nanostructures for real-time image processing and analog computing, and discloses their unprecedented possibilities in the fields of neuromorphic photonics, machine vision and bio-microscopy.”
Light: Science & Applications
Reconfigurable ferroelectric chiral nanostructures enable fast-switchable optical spatial differentiation