Enhanced selectivity of platinum-modified tungsten oxide gas sensor through multivariate feature extraction and...

Area of Science:

• Materials Science
• Chemical Sensing
• Artificial Intelligence

Background:

• Metal oxide semiconductor (MOS) gas sensors struggle with selectivity for volatile organic compounds (VOCs).
• Accurate detection of specific VOCs in industrial environments remains a challenge.
• Environmental factors like humidity can degrade sensor performance.

Purpose of the Study:

• To develop a highly selective and stable gas sensor for VOC detection.
• To improve the accuracy of gas identification and concentration prediction in complex mixtures.
• To address the limitations of traditional MOS sensors in industrial applications.

Main Methods:

• Fabrication of platinum-modified tungsten oxide (Pt/WO3) composite via in-situ reduction.
• Gas sensing performance evaluation towards triethylamine, ammonia, and isopropanol.
• Signal processing using discrete wavelet transform (DWT) for noise reduction.
• Development of an artificial neural network (ANN) for VOC classification.
• Linear regression for gas concentration prediction.

Main Results:

• The Pt/WO3 sensor exhibited superior sensitivity and stability towards triethylamine compared to pristine WO3.
• The DWT and ANN approach achieved 95.2% accuracy in identifying multiple VOCs.
• Robust humidity resistance and long-term operational stability were demonstrated.
• Accurate prediction of unknown gas concentrations was achieved using a linear regression model.

Conclusions:

• The Pt/WO3 composite offers a promising solution for high-performance gas sensing.
• Intelligent signal processing and machine learning enhance sensor selectivity and accuracy.
• This approach provides a reliable strategy for identifying gas concentrations in mixed VOC environments.