(Nanowerk Spotlight) Human vision provides a model for advanced artificial intelligence systems that can perceive and understand visual information. However, conventional electronics rely on combining separate light sensors, memory units, and logic circuits to try to replicate human visual capabilities. This multi-component approach increases size and complexity while limiting how well the artificial system can emulate biological vision.
To overcome these constraints, researchers have sought integrated technologies that absorb light, store visual representations, and process images within single devices. Recently, a team from China developed a new dual-function material called a perovskite that makes significant strides toward this goal. In a single integrated unit, the perovskite can both detect light and memorize visual patterns inspired by human vision. This research tackles the challenge of distilling key aspects of biological sight into a streamlined artificial system.
a) Schematic of the human visual system when the Z shape was observed by the eyes. b) Schematic of the route of light and the corresponding track recording results (50 × 50 µm for each single pixel). c) Schematic of the photoelectric writing and the corresponding data storage results. d) Schematic of 4 × 4 arrays Au/Ag–Cs3Sb2I9–ITO (indium tin oxide films) dual-functional devices. e) The photo imaging result of the device (light intensity: 2.34 mW cm−2). f,g) Data storage and shape memory of the image before and after 60 min. h) The result of the array after reset. i) Photoimaging result of the device after resetting. (Reprinted with permission by Wiley-VCH Verlag)
The human brain integrates visual inputs from the eyes with memory centers that store the information. However, artificially replicating this process is challenging, as it requires the integration of light sensors, memory components, and logic circuits into a single system. The new perovskite device overcomes this barrier by combining both light sensing and memory storage capabilities.
Perovskites are emerging as promising materials for optoelectronic applications due to their superior light absorption properties and ease of fabrication. The researchers leveraged these advantages to create a dual-function device using lead-free cesium antimony iodide perovskite microplates.
One of the device’s roles is detecting light. Imagine the perovskite microplate as a super-efficient sponge, but for light instead of water. It can absorb light from a wide range of colors that we can see with our eyes. When it is used as part of a light-detecting setup, it performs extraordinarily well.
To give you an idea, let’s talk about ‘light responsivity’. Think of this as the device’s ability to react to light. A higher responsivity means it can detect and respond to very low levels of light, making it highly sensitive. This perovskite device shows a light responsivity of 276 mA/W at a certain color of green light, which is a measurement far above many other light-detecting devices.
Now, about ‘detectivity’. This term refers to how well the device can pick up a weak light signal in the presence of background noise – kind of like hearing a whisper in a crowded room. The higher the detectivity, the better it is at detecting weak signals. This device boasts a detectivity of 4.7 × 1011 Jones, again showcasing its superior performance compared to most other light detectors.
The second function is nonvolatile memory storage. The perovskite acts as a memristor, a type of resistive random access memory. By applying a small voltage of just 0.15 V, the resistance state of the perovskite memristor can be switched from high to low. This transition corresponds to writing a memory bit. The device retains the written state for over 7,000 seconds without degradation, enabling reliable long-term visual information storage. The memristive properties allow visual patterns detected by the perovskite to be memorized and later retrieved on demand.
Remarkably, the dual light sensing and memory capabilities are achieved in a single integrated device. The researchers demonstrated a 4×4 array of pixels that could detect a moving light source, capture the illumination pattern, memorize the image, and later retrieve it on demand. This emulates how human vision perceives a scene and stores it in the brain’s memory centers.
The memorized visual patterns could be reliably erased and rewritten, highlighting the reusability of the perovskite hardware. The combination of excellent photodetection performance with versatile visual memory capacity in a single device represents a significant advance over previous technologies.
The researchers explain the operating principles enabling the dual functionality. The perovskite microplate structure and energy band design promote efficient light absorption and charge transport, underlying the strong photodetecting behavior. On the memory side, voltage stimuli mobilize ions within the perovskite to form and rupture conductive filaments. This switches the resistance and programs the memory state.
Compared to conventional resistive random-access memory, the perovskite version demonstrates ultralow power consumption down to 3×10-11 W. The low programming energy, coupled with high-performance light sensing, makes the new technology well-suited for applications like low-power artificial neural networks that aim to mimic biological vision and learning.
Looking forward, the integration of photodetection, memory, and processing into single perovskite devices could enable compact camera-like systems. By eliminating the need to combine separate sensors, memory units, and logic components, the approach provides a route toward highly integrated visual capture and artificial intelligence platforms.
The researchers highlight the lead-free composition as making this technology safer and more sustainable than conventional perovskite formulations containing lead. Transferring the concept to other lead-free perovskites could further expand the scope of accessible wavelengths and tunable properties.
Overall, this dual-function perovskite visual memory technology provides a new blueprint for advanced light sensing and information systems. By replicating key aspects of human vision in an integrated device, it opens exciting possibilities for technologies like electronic eyes, visual neural networks, and other biomimetic applications. The combination of useful performance metrics with versatility and ease of fabrication illustrates the significant potential of perovskites to underpin future optoelectronic advancements.
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.