Polarization-encodable photonic memory cells using next-generation 2D phase-change materials

Integration of phase-change materials (PCMs) created a unique opportunity to implement reconfigurable photonics devices that their performance can be tuned depending on the target application. Conventional PCMs such as Ge-Sb-Te (GST) and Ge-Sb-Se-Te (GSST) rely on melt-quench and high temperature annealing processes to change the organization of the molecules in the materials’ crystal. Such a reorganization leads to different optical, electrical, and thermal properties which can be exploited to implement photonic memory cells that are able to store the data at different resistance or optical transmission levels. Despite the great promise of conventional PCMs for realizing reconfigurable photonic memories, their slow and extremely power-hungry thermal mechanisms make scaling the systems based on such devices challenging. In addition, such materials do not offer a stable multi-level response over a long period of time. To address these shortcomings, the research carried out in this study shows the proof of concept to implement next-generation photonic memory cells based on two-dimensional (2D) birefringence PCMs such as SnSe, which offer anisotropic optical properties that can be switched ferroelectrically. We demonstrate that by leveraging the ultrafast and low-power crystallographic direction change of the material, the optical polarization state of the input optical signal can be changed. This enables the implementation of next-generation high-speed polarization-encodable photonic memory cells for future photonic computing systems. Compared to the conventional PCMs, the proposed SnSe-based photonic memory cells offer an ultrafast switching and low-loss optical response relying on ferroelectric property of SnSe to encode the data on the polarization state of the input optical signal. Such a polarization encoding scheme also reduces memory read-out errors and alleviates the scalability limitations due to the optical insertion loss often seen in optical transmission encoding.