Nitrogen dioxide (NO₂) is a crucial environmental pollutant. NO₂ generated from automobile exhaust is a critical source of pollution, which has severely affected human health and influenced the quality of urban air and global climate change. The photoacoustic spectroscopy (PAS) based on the photoacoustic effect is a critical technology for trace gas detection owing to its benefits of zero background noise and high detection sensitivity. This study suggests a gas sensor based on PAS with a diode laser at 444 nm as the light source. The NO₂ concentration in the atmosphere is monitored. NO₂ concentrations emitted from various types of automobiles are compared, and the photolysis of NO₂ is investigated. We hope that the findings of the study can offer a significant reference for the improvement of urban air quality and play a critical role in the comprehensive regulation of urban air pollution.

A diode laser with a central wavelength of 444 nm is employed as the excitation source, and an optical power meter behind the photoacoustic cell is employed to monitor the laser power’s stability. The optical chopper can modulate the optical intensity at the photoacoustic cell’s resonance frequency, and a reference signal (a square wave signal) is input to the lock-in amplifier. A microphone with a sensitivity of 45 mV/Pa is installed in the middle of the photoacoustic cell to detect the photoacoustic signal. The detected signal is amplified using a preamplifier and demodulated using a lock-in amplifier, and then collected and processed using the data acquisition card and a computer with LabVIEW software. The vacuum pump, pressure controller, and mass flow controller are employed to regulate the pressure and flow of gas in the photoacoustic cell. A filter is installed in front of the gas inlet to prevent the interference of aerosols.

The photoacoustic signal intensity increases as the pressure and volume fraction of NO₂ increase (Fig. 3). The resonance frequency increases with the increase in pressure since both the gas density and speed of sound increase as the pressure increases (Table 1). The resonance frequency slightly decreases with the increase in the volume fraction of NO₂ since the heat capacity ratio and molar mass of gas change with the volume fraction’s variation (Table 2). There is a good linear relationship between the signal-to-noise ratio and optical power, and the system does not achieve the saturation absorption state [Fig. 4(a)]. When the chopper runs at the photoacoustic cell’s resonance frequency, the air near the window of the photoacoustic cell flows, and the optical power into the photoacoustic cell and the photoacoustic signal intensity slightly change. When the distance between the optical chopper and the photoacoustic cell’s window is 50 mm, the maximum signal is obtained [Fig. 4(b)]. The part of NO₂ is dissolved using water vapor, and the vibrational-translational relaxation rate is also influenced, leading to an inverse ratio between photoacoustic signal and relative humidity [Figs. 4(c) and (d)]. It takes some time for the signal to be stable as the pressure changes (Fig. 5). The Allan variance analysis exhibits that the sensor detection limit is 1×10^-9 with an integration time of 100 s (Fig. 6). The change characteristic of NO₂ concentration in two days from 8: 00 a.m. to 6: 00 p.m. is determined and the trend is consistent with the monitoring data from the Jining Environmental Monitoring Center (Fig. 7). NO₂ concentrations in exhaust from five types of automobiles are detected (Fig. 8). At low speed, it can be deduced that at least 150 g NO₂ will be released if the automobile is driven 100 km and the photolysis rate of NO₂ is ~50% from 12: 00 a.m. to 2: 00 p.m. (Fig. 8).

A gas sensor based on photoacoustic spectroscopy is suggested. The 444-nm blue laser is employed as the light source to achieve NO₂ detection in the atmosphere and automobile exhaust, and NO₂ photolysis is examined. The relationship between pressure and resonance frequency and that between the volume fraction of NO₂ and resonance frequency are investigated. The linear relationship between signal-to-noise ratio and optical power is studied. The influence of the position of an optical chopper and relative humidity on the photoacoustic signal is examined. The detection limit of 1×10^-9 with an integration time of 100 s is obtained using Allan variance analysis. Atmospheric NO₂ is monitored for two days, and the result has a good consistency with the data from the Jining Environmental Monitoring Center. Finally, the photolysis of NO₂ emitted from automobiles is also investigated. This offers high sensitivity sensor for NO₂ detection in practical applications.