Microplastic detection and recognition system enabled by a triboelectric nanogenerator and machine learning...

Area of Science:

• Environmental Science
• Analytical Chemistry
• Materials Science

Background:

• Microplastic pollution poses risks, with toxicity linked to material, size, and concentration.
• Existing detection methods are costly, time-consuming, and have limited databases.
• A need exists for rapid, accurate microplastic identification in various matrices.

Purpose of the Study:

• To develop a simple, rapid device for microplastic detection and identification in water.
• To utilize liquid-solid contact electrification and machine learning for enhanced analysis.
• To provide a cost-effective and efficient alternative to current detection techniques.

Main Methods:

• Designed a device based on liquid-solid contact electrification principles.
• Employed machine learning algorithms, specifically Random Forest (RF), for data analysis.
• Measured open-circuit voltage differences generated by different microplastic types.

Main Results:

• Distinct voltage differences (1.6 V–11.7 V open-circuit, 1.7 V–22.0 V peak) were observed for various microplastics.
• The Random Forest model achieved an average recognition accuracy of 95.24%.
• The device demonstrated effective differentiation and identification of microplastic types.

Conclusions:

• A novel, rapid, and accurate method for microplastic detection and identification has been established.
• This technology offers potential for real-time online monitoring of microplastics in water.
• Significant implications for environmental monitoring, food safety, and medical/health fields are anticipated.