Sine domain spectral transformation combined with Fourier transform: a UV-DOAS-based optical sensor-system for sub-ppb NH3 detection considering interference components

The detection of trace ammonia (NH₃) concentrations in breath gas is essential for facilitating noninvasive diagnosis of kidney disease. However, broadband absorption spectra methods, including ultraviolet differential optical absorption spectroscopy (UV-DOAS), often encounter challenges related to cross-sensitivity and noise interference for trace breath NH₃ detection. Here, a UV-DOAS-based sensor that combines sine domain spectral transformation and Fourier Transform (FT) is proposed to enable trace NH₃ detection in complex breath environments. First, the differential absorption spectra of NH₃ were obtained by UV-DOAS, and the influence of breath components on the spectra of NH₃ was analyzed. Second, sine domain spectral transformation was proposed to enhance the NH₃ features by mapping the absorption characteristics to the sine domain while attenuating interfering components and noise. FT is then introduced to improve the selectivity of NH₃. Laboratory results demonstrate that the sensor can detect trace NH₃ accurately under interfering components, with a detection range from 26.58 ppb to 1955.78 ppb. And the Allan variance analysis indicates that the minimum detection limit of the sensor is 0.26 ppb, which is among the best results in this band. In addition, a real human breath experiment was conducted, and the results indicate that the sensor can effectively detect trace NH₃ in breath, demonstrating excellent reproducibility and stability. The performance of the proposed sensor demonstrates that combining sine domain spectral transformation with FT can effectively improve the detection accuracy and break the limitations of spectral overlapping, providing a novel insight for broadband spectra development.