Design of LSPR optical memory based on angle-resolved scattering measurement

Since the storage density of optical memory cannot meet the growing demand for high-density data storage. To further improve the storage density, this paper proposes a two-dimensional diamond Au nanostructure design. This design uses the angle-resolved scatterometer (ARS) and the local surface plasmon resonance (LSPR) effect to achieve information encoding and readout. Each nanostructured cell in the array is rotated by a set angle relative to the previous cell. As each rotation produces a different reflectivity spectrum. Thus, multi-state information encoding based on the rotation angle can be realized. Meanwhile, LSPR effect of the nanostructure array further enhances the distinguishability of the stored data information. The mapping relationship between the stored data information and the ARS spectrum is trained using fully connected convolutional neural network deep learning, and then the decoding process is realized. The results show that with a rotation step of 1°, even with a 10% manufacturing error, the coding test accuracy can still be maintained at around 97.5%. The storage density reaches 15.8 GB/cm², which is 65.83 times that of standard single-layer Blu-ray storage (0.24 GB/cm²).