THz Tunable EMI Shielding via Ultra‐Thin Graphene‐MXene Metasurfaces: Design and Machine Learning‐Based Effectiveness Prediction

ABSTRACT This study proposes the design and performance of an ultra‐thin metasurface composed of MXene and graphene nanomaterials. The aim is to make an electromagnetic interference (EMI) shielding in the terahertz (THz) frequency range. The proposed metasurface combines the tunability and flexibility of graphene with the high conductivity of MXene. This hybrid structure achieves strong shielding performance over a wide frequency range. Shielding efficiency (SE) was optimized using progressive simulations and machine learning (ML) models. The metasurface demonstrated a peak SE of 50 dB at a frequency range of 4.9–6.5 THz. The design provides insensitivity to polarization and excellent wide‐angle shielding. This makes it suitable for flexible and stealth electronic applications. The study uses ML, specifically Extreme Randomized Trees (ERT), to predict SE based on structural parameters. This approach achieves high predictive accuracy (R 2 > 0.9967) with low root mean square error (RMSE < 0.24). It also enables rapid design optimization and performance prediction for metasurfaces in THz EMI shielding applications. The findings demonstrate that the synergy between MXene and graphene provides a robust platform for future THz technologies. This structure offers substantial improvements over conventional metallic and polymer‐based shields.