Impedance-Assisted Multivariate Analysis Technique for Enhanced Gas Sensing with 2D Dichalcogenides

Area of Science:

• Materials Science
• Nanotechnology
• Sensor Technology

Background:

• Semiconducting 2D materials offer high sensitivity for gas sensors but suffer from drift, nonlinearity, and cross-sensitivity.
• Conventional resistive sensing struggles to capture complex 2D material interactions, limiting accuracy.

Purpose of the Study:

• To overcome limitations of 2D material gas sensors by employing multifrequency impedance measurements and machine learning.
• To achieve accurate relative humidity (RH) quantification and minimize cross-sensitivity.

Main Methods:

• Utilized MoS2 and WS2-based sensors with multifrequency impedance measurements.
• Applied machine learning models (MLP, 1D-CNN, LSTM) for data processing and RH quantification.
• Evaluated sensor performance for stability, response/recovery times, and cross-sensitivity.

Main Results:

• Multifrequency impedance sensing mitigated baseline drift and enabled precise RH measurements (0-90%).
• MoS2 sensors showed long-term stability, while WS2 sensors exhibited mutually exclusive phase behavior.
• Machine learning-assisted WS2 sensors effectively minimized cross-sensitivity between humidity and CO2.

Conclusions:

• Multifrequency impedance sensing combined with machine learning offers a robust solution for 2D material gas sensor limitations.
• This approach leads to more reliable, stable, and precise gas-sensing technologies, particularly for humidity monitoring.