Related Experiment Video
Updated: Jun 5, 2025
08:58
Seedless Growth of Bismuth Nanowire Array via Vacuum Thermal Evaporation
Published on: December 21, 2015
8.4K
Machine learning-motivated trace triethylamine identification by bismuth vanadate/tungsten oxide heterostructures
Wei Ding1, Min Feng2, Ziqi Zhang3
1College of Chemistry and Chemical Engineering, Hexi University, Zhangye 734000, PR China; College of Materials Science and Engineering, Qingdao University, Qingdao 266071, PR China.
Journal of Colloid and Interface Science
|December 13, 2024
Summary
A novel machine learning sensor detects hazardous triethylamine (TEA) gas. This BiVO4/WO3 heterostructure sensor achieves high accuracy and predicts TEA concentration, offering a breakthrough in industrial safety.
Area of Science:
- Materials Science
- Chemical Sensors
- Machine Learning
Background:
- Triethylamine (TEA) is widely used in organic synthesis but poses significant health risks.
- Accurate detection and prediction of TEA levels remain a persistent challenge in industrial settings.
Purpose of the Study:
- To design a machine learning-motivated chemiresistive sensor for detecting and predicting ppm-level triethylamine.
- To develop an intelligent framework for identifying and quantifying TEA in industrial environments.
Main Methods:
- Fabrication of hierarchical BiVO4/WO3 heterostructures using 0D BiVO4 nanoparticles on 3D WO3 architectures.
- Utilized a machine learning classifier for accurate identification and a linear regression model for concentration prediction.
- Extracted feature parameters from sensor responses for intelligent analysis.
Main Results:
- The BiVO4/WO3 sensor exhibited a high response of 21 to TEA at 190°C, significantly outperforming pristine WO3.
- Achieved a low detection limit of 57 ppb with excellent long-term stability, reproducibility, and anti-interference capabilities.
- The machine learning classifier demonstrated 92.3% accuracy in identifying TEA, with successful prediction of unknown concentrations.
Conclusions:
- Hierarchical BiVO4/WO3 heterostructures offer superior sensing performance for triethylamine.
- The integrated machine learning approach provides an effective strategy for intelligent identification and prediction of trace TEA.
- This work advances the understanding of BiVO4-based heterostructures in gas sensing applications and offers practical safety solutions.
