Integration of Nanoscale and Macroscale Graphene Heterostructures for Flexible and Multilevel Nonvolatile Photoelectronic Memory

The development of optical memory with attractive features such as long-lasting, nonvolatile, high-speed, and low-energy consumption is vitally important in the information age. Owing to these advantages, optical memory has been popular for more 10 years. Recently, flexibility has become desirable for the application of wearable devices and smart artificial intelligence; for conventional optical memory, this is still difficult to achieve. To combine optical memory with soft materials, this study presents a flexible and photoelectronic switchable multilevel memory device with long-lasting nonvolatile properties. On the basis of the integration of nanoscale (graphene nanoflakes) and macroscale graphene heterojunctions, a device achieves switchable memory states up to 196 distinct levels under the illumination of lasers with different wavelengths. The photoelectronic memory device can be written optically and erased by both optical and electric methods. Additionally, the device possesses several unique features including a low working bias of 0.5 V, nonvolatility for over 10 000 s, and mechanical stability for more than 10 000 bending cycles. Notably, in previous studies, polymers with poor mobility were used as a conducting channel, which can greatly limit the amplitude of the light-induced switching ratio and electrical performance. In stark contrast, in our device, the graphene layer with the mobility exceeding several orders of magnitude was used to serve as a conducting channel, enabling one to overcome the existing shortcoming. Our approach therefore not only provides an alternative paradigm for the development of photoelectronic memory but also holds great promise for practical applications due to its compatibility with current technologies.