Tensor completion algorithm-aided structural color design.

Area of Science:

• Photonics and Materials Science
• Optical Engineering
• Computational Physics

Background:

• Structural color offers advantages like low loss and environmental friendliness, but its inverse design is challenging.
• Traditional optimization methods (e.g., genetic algorithms) are time-consuming, and deep learning requires extensive data.
• Existing methods struggle with efficiency, especially for systems with limited features or data.

Purpose of the Study:

• To develop a novel, efficient algorithm for the inverse design of structural color.
• To address the limitations of current optimization and deep learning techniques in structural color design.
• To enhance the design efficiency of photonic devices through rapid and accurate data prediction.

Main Methods:

• A tensor completion algorithm is proposed to predict missing datasets from partially obtained data.
• The complex physical problem is reframed as a spatial structure relationship problem linking geometric parameters and spectral data.
• Tensor multilinear data analysis is employed to capture intricate relationships between geometric parameters and spectral data.

Main Results:

• The algorithm demonstrates high reliability in speed and accuracy across diverse structures, dataset sizes, and materials.
• It significantly enhances design efficiency for structural color applications.
• Numerical experiments show diamond cruciform resonators overcome blue-wavelength loss and improve corrosion resistance, achieving wide color gamuts and high-narrow reflection spectra.

Conclusions:

• The tensor completion algorithm provides a viable and efficient solution for inverse design problems in complex physical systems.
• This method offers a novel approach to designing photonic devices, improving upon existing techniques.
• Diamond shows significant promise for optical applications, particularly in overcoming material loss issues and enhancing structural properties.