NO2 trace gas monitoring in air using off-beam quartz enhanced photoacoustic spectroscopy (QEPAS) and interference studies towards CO2, H2O and acoustic noise

Abstract

Abstract We present the development and characterization as well as comprehensive interference studies of a photoacoustic NO2 trace gas detection system. The system is based on an off-beam quartz enhanced photoacoustic scheme (off-beam QEPAS) and signal generation was initiated by amplitude modulating a low-cost diode laser emitting at 450 nm. The QEPAS sensor element features double-resonant amplification, still it is only ∼ 5 × 5 × 2.5 mm in size. The individual and combined resonance characteristics were investigated and specified to 52 dB amplification, adding up 15 dB acoustic- and 37 dB mechanical-resonance amplification. The linearity of the photoacoustic signal dependency on the analyte concentration was verified from 200 ppbV to 100 ppmV NO2 in synthetic air. The detection limit (3σ) was determined to 1.8 ppbV using a lock-in time constant of 10 s and an averaging time of 20 s. The normalized noise equivalent absorption coefficient was specified to 2.5·10−8 W cm−1 Hz−0.5. The stability of the signal was investigated over time and a slight drift by 1‰ was observed after 30 min without temperature stabilizing the photoacoustic cell (PAC). Noise analysis was performed by means of Allan deviation and the inverse dependency of response time and precision of the system on the lock-in time constant was outlined. We performed interference analyses towards N2, O2, CO2, H2O and acoustic noise, respectively. Although neither spectral interferences nor losses due to slow NO2 VT-relaxation were observed, O2 was identified to cause a 15% signal drop due to VVNO2-O2-relaxation. Changing H2O concentrations were found to cause acoustic detuning, which cannot be compensated by adjusting the frequency of modulation, because of the double-resonant feature of the PAC. However, alternative approaches of compensation were discussed. Finally, we carried out heavy traffic noise simulations and determined the QEPAS setup to be 46 times less susceptible towards ambient noise compared to standard microphone-based photoacoustic setups.