Machine Learning Driven Window Blinds Inspired Porous Carbon‐Based Flake for Ultra‐Broadband Electromagnetic Wave Absorption

The core challenge in developing lightweight and high-efficiency electromagnetic wave absorbing materials lies in achieving a decoupled improvement in impedance matching and loss performance while reducing thickness. Inspired by the structure of window blinds, this study proposes and designs a Discrete Slat Tunable Electromagnetic Wave Absorption Material (DSTEAM). By incorporating a magneto-electric coupling concept and an artificial intelligence-assisted data-driven optimization strategy, the successfully fabricated DSTEAM exhibits outstanding performance with a reflection loss below -10 dB over an ultra-broadband frequency range of 2.6-40 GHz, while maintaining a thin thickness of only 9.85 mm and an areal density of 0.566 kg/m². The superior performance of DSTEAM is attributed to gradient-induced multiple scattering at discrete sheet interfaces, a synergistic enhancement of localized field strength, and a magneto-electric coupling modulation mechanism. This AI-driven collaborative design strategy offers a novel concept and effective pathway for the development of next-generation lightweight and broadband electromagnetic wave absorption materials.