Advanced recycling and identification system for discarded capacitors utilizing laser-induced breakdown spectroscopy...

Area of Science:

• Materials Science
• Environmental Science
• Computer Science

Background:

• Electronic waste (e-waste) is rapidly increasing due to technological advancements.
• Discarded capacitors are a major e-waste component, containing hazardous materials and valuable metals.
• Current manual classification methods for capacitors are inefficient and inaccurate.

Purpose of the Study:

• To enhance the classification of discarded capacitors for improved recycling and resource recovery.
• To introduce a novel, accurate, and efficient method for capacitor identification.
• To calibrate spectral lines of pure niobium for spectroscopic studies.

Main Methods:

• Combination of Laser-Induced Breakdown Spectroscopy (LIBS) and machine learning (Backpropagation Artificial Neural Network - BP-ANN).
• Development and application of the Optimized Feature Extraction Variance Algorithm (OFEVA) for feature extraction.
• Comparison of OFEVA with traditional Principal Component Analysis (PCA) for classification accuracy and efficiency.

Main Results:

• The novel OFEVA algorithm significantly improves classification accuracy compared to PCA.
• The LIBS and machine learning approach enables automatic and accurate identification of discarded capacitors.
• Spectral lines of pure niobium were calibrated for the first time.

Conclusions:

• This innovative approach enhances e-waste capacitor recycling rates and reduces environmental pollution.
• The method provides technical support for resource reuse and contributes to environmental protection.
• The study offers valuable data for future spectroscopic research.