Machine Learning-Optimized Electromagnetic Wave Absorption in Metal/C Nanocomposites

Area of Science:

• Materials Science
• Nanotechnology
• Computational Materials Science

Background:

• Traditional carbon-based absorbers lack tunability due to complex synthesis parameter interactions.
• Rational design of high-performance electromagnetic wave absorption (EWA) materials is challenging.

Purpose of the Study:

• To optimize EWA performance in Metal/C Nanocomposites using a genetic algorithm (GA).
• To identify key synthesis parameters influencing EWA performance.

Main Methods:

• Simultaneous tuning of five synthesis parameters (carbon precursor, metal type, precursor/metal ratio, carbonization temperature, filler loading) over three GA generations.
• Utilizing Random Forest and XGBoost models to quantify parameter importance.

Main Results:

• Enhanced Absorption Band (EAB) improved from 1.24 GHz to 4.08 GHz on average.
• Minimal reflection loss (RLmin) improved from -20.29 dB to -41.9 dB.
• Champion sample achieved RLmin of -25.9 dB with an EAB of 7.56 GHz.

Conclusions:

• GA-driven optimization significantly enhances EWA performance in Metal/C Nanocomposites.
• Carbon precursor type and filler loading ratio are dominant factors in EWA performance.
• Evolutionary algorithms offer a transferable framework for designing high-performance EWA materials.