Multi-state silicon photonics waveguides using high-index resistive memory

Most reported optical memory devices remain volatile and fail to retain data after power-off, limiting their applicability in high-density optical storage and computing. To overcome this limitation, we demonstrate a non-volatile, multi-level optical memory by integrating resistive random-access memory (ReRAM) based on high-refractive-index materials (BiFeO₃, TiO_x, and Al₂O₃) onto a silicon photonic waveguide. This configuration enhances light–filament interaction, enabling persistent electro-optical modulation. An elevated film stack (EFS) structure is further introduced to enhance optical and electrical field confinement, thereby promoting stronger interaction between the guided light and conductive filaments. This configuration also improves filament formation uniformity, enabling low-voltage operation and reduced power consumption. Spectral analysis reveals extinction ratios (ERs) of 6.34 dB (76.7%), 11.73 dB (93.2%), and 13.24 dB (95.2%) for logic states L1, L2, and L3, respectively, along with a 2 nm total wavelength shift between the initial(L0) and final states(L3). Here, L1 corresponds to switching a single ReRAM cell, while L2 and L3 represent increasing the filament formations across multiple ReRAM cells. These results confirm strong, stepwise optical contrast across memory binary levels at 0 V, demonstrating reliable non-volatile modulation. To support these findings, synchronized electro-optical measurements using a 1550 nm single-wavelength source show normalized transmission increasing from ∼0.72 (L1) to >0.8 (L2, L3). Additionally, within the 1500–1600 nm band, the integrated spectral energy of the L3 state decreases by 1.31 mW • nm relative to the initial state, further validating filament–light interaction. This work presents a broadband, low-power, and non-volatile optical memory platform with clear multi-level behavior, offering promising potential for photonic logic, neuromorphic computing, and reconfigurable silicon photonic systems.