Laser-based selective BTEX sensing using deep neural networks

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Environmental Monitoring

Background:

• Benzene, toluene, ethylbenzene, and xylenes (BTEX) are volatile organic compounds with significant health and environmental impacts.
• Accurate and real-time monitoring of BTEX is crucial for industrial safety and regulatory compliance.

Purpose of the Study:

• To develop a novel mid-infrared laser sensor for selective and simultaneous detection of BTEX compounds.
• To achieve high sensitivity and temporal resolution for real-time BTEX measurements.

Main Methods:

• Utilized a distributed feedback inter-band cascade laser emitting near 3.3 µm for mid-infrared absorption.
• Employed wavelength tuning and deep neural networks to differentiate overlapping BTEX absorbance spectra.
• Validated the sensor using gas mixtures and demonstrated real-time measurements.

Main Results:

• Achieved selective and simultaneous detection of BTEX compounds under ambient conditions.
• Demonstrated real-time measurements with a temporal resolution of 1 second.
• Established minimum detection limits for BTEX in air: benzene (8 ppm), toluene (20 ppm), ethylbenzene (5 ppm), and xylenes (46 ppm).

Conclusions:

• The developed laser sensor provides a reliable method for monitoring BTEX emissions.
• This technology is applicable to industries such as petrochemical, rubber, and paint manufacturing.
• The sensor contributes to mitigating hazardous health effects associated with BTEX exposure.