Related Experiment Video
Updated: Aug 1, 2025
07:41
Advanced Compositional Analysis of Nanoparticle-polymer Composites Using Direct Fluorescence Imaging
Published on: July 19, 2016
7.8K
Visualization of Deep Convolutional Neural Networks to Investigate Porous Nanocomposites for Electromagnetic
Meng Shi1, Chang-Ping Feng2, You-Lei Tu1
1The State Key Laboratory of Polymer Materials Engineering, Polymer Research Institute of Sichuan University, Chengdu 610065, China.
ACS Applied Materials & Interfaces
|April 25, 2023
Summary
This study introduces a visual deep learning approach to understand how porosity affects electromagnetic interference (EMI) shielding in nanocomposites. The method reveals that pore structure significantly influences shielding mechanisms, guiding the development of advanced EMI materials.
Area of Science:
- Materials Science
- Nanotechnology
- Computational Science
Background:
- Porous structures are key for electromagnetic interference (EMI) shielding materials to reduce reflected electromagnetic waves (EMWs).
- Understanding porosity's impact on EMI shielding mechanisms is challenging due to a lack of direct analysis methods.
- Deep learning models (DCNNs) offer potential but lack interpretability for complex material science investigations.
Purpose of the Study:
- To develop a visual approach combining deep learning and experimental methods to elucidate the EMI shielding mechanisms in porous nanocomposites.
- To investigate the influence of porosity and filler loading on the EMI shielding effectiveness of carbon nanotube/polyvinylidene fluoride (CNTs/PVDF) composites.
- To enhance the interpretability of deep learning models in material science for mechanism studies.
Main Methods:
- Fabrication of CNTs/PVDF nanocomposites with controlled porosity using a salt-leaked cold-pressing powder sintering method.
- Characterization of EMI shielding effectiveness, with a solid sample achieving 105 dB at 30 wt% CNT loading.
- Training a modified deep residual network (ResNet) on scanning electron microscopy (SEM) images and employing Eigen-CAM visualization to analyze pore structure impact.
Main Results:
- The study demonstrates an ultrahigh EMI shielding effectiveness of 105 dB in a solid CNTs/PVDF composite (30 wt% loading).
- Visualizations from the deep learning model intuitively show that pore amount and depth significantly influence EMI shielding mechanisms.
- Shallow pore structures were found to contribute less effectively to electromagnetic wave absorption.
Conclusions:
- The proposed visual deep learning approach provides crucial insights into the relationship between porosity and EMI shielding mechanisms in nanocomposites.
- This methodology is instructive for material mechanism studies and offers potential as a tool for marking porous-like structures.
- The findings guide the rational design of high-performance porous EMI shielding materials.
