Smart mid-infrared metasurface microspectrometer gas sensing system.

Area of Science:

• Sensor technology
• Spectroscopy
• Machine learning

Background:

• Traditional gas sensors are single-analyte specific, limiting their application.
• Infrared (IR) spectroscopy offers superior performance but is hindered by size and cost.
• Portable and low-cost gas sensing is crucial for environmental and defense monitoring.

Purpose of the Study:

• To develop a smart, low-cost, portable multigas sensing system.
• To integrate a compact mid-infrared microspectrometer with a machine learning algorithm.
• To demonstrate the system's capability for accurate gas detection.

Main Methods:

• A microspectrometer utilizing a metasurface filter array and a commercial IR camera was developed.
• The system was designed to be compact (~1 cm³), lightweight (~1 g), and consumable-free.
• A machine learning algorithm was trained to analyze spectral data and predict gas composition.

Main Results:

• The system achieved 100% accuracy in detecting carbon dioxide and methane (10-100% concentrations).
• Hazardous gases were detected with 98.4% accuracy, including ammonia at 100 ppm.
• Methyl-ethyl-ketone was detected at its permissible exposure limit (200 ppm).

Conclusions:

• The study demonstrates a viable smart, low-cost multigas sensing platform.
• Combining machine learning with IR spectroscopy enables advanced gas detection capabilities.
• The developed system offers a significant advancement for portable environmental and defense monitoring.