AbstractConvolution operation is of great significance in on-chip all-optical signal processing, especially in signal analysis and image processing. It is a basic and important mathematical operation in the realization of all-optical computing. Here, we propose and experimentally implement a dispersionless metalens for dual wavelengths, a 4f optical processing system, and then demonstrate the on-chip nanophotonic convolver based on silicon metasurface with the optimization assistance of inverse design. The characteristic size of the dispersionless metalens device is 8 × 9.4 μm, and the focusing efficiency is up to 79% and 85% at wavelengths of 1000 and 1550 nm, respectively. The feature size of the convolver is 24 × 9.4 μm, and the proposed convolver allows spatial convolution operation on any desired function at dual wavelengths simultaneously. This work provides a potential scheme for the further development of on-chip all-optical computing.

A series of experiments are conducted to examine the purification effect on degraded liquid crystal (LC) by soaking porous metal-organic frameworks in the LC material. The results indicate that the electric properties of the degraded LC are dramatically improved due to the adsorption and removal of moisture and impurity ions, most likely originating from the atmosphere. This investigation leads to the advancement in the feasible recycling of deteriorated LC materials, including LC leftovers from flat-panel production lines, with a simple and cost-effective treatment method.

The solution of matrix eigenvalues has always been a research hotspot in the field of modern numerical analysis, which has important value in practical application of engineering technology and scientific research. Despite the fact that currently existing algorithms for solving the eigenvalues of matrices are well-developed to try to satisfy both in terms of computational accuracy and efficiency, few of them have been able to be realized on photonic platform. The photonic neural network not only has strong judgment in solving inference tasks due to the superior learning ability, but also makes full use of the advantages of photonic computing with ultrahigh speed and ultralow energy consumption. Here, we propose a strategy of an eigenvalue solver for real-value symmetric matrices based on reconfigurable photonic neural networks. The strategy shows the feasibility of solving the eigenvalues of real-value symmetric matrices of n × n matrices with locally connected networks. Experimentally, we demonstrate the task of solving the eigenvalues of 2 × 2, 3 × 3, and 4 × 4 real-value symmetric matrices based on graphene/Si thermo-optical modulated reconfigurable photonic neural networks with saturated absorption nonlinear activation layer. The theoretically predicted test set accuracy of the 2 × 2 matrices is 93.6% with the measured accuracy of 78.8% in the experiment by the standard defined for simplicity of comparison. This work not only provides a feasible solution for the on-chip integrated photonic realization of eigenvalue solving of real-value symmetric matrices, but also lays the foundation for a new generation of intelligent on-chip integrated all-optical computing.