Recently, a research team led by Professor Zhang Zhirong at the Anhui Institute of Optics and Precision Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences, has made significant progress in the study of baseline reconstruction technology for broad-spectrum absorption gases. The findings, titled "Baseline Reconstruction for Broad-Spectrum Absorption Characteristics of Alkane Mixtures Based on Laser Absorption Spectroscopy," were published in the international academic journal Analytical Chemistry.
Direct absorption spectroscopy is the most widely used quantitative gas analysis method within laser absorption spectroscopy. However, its measurement accuracy heavily relies on the precise acquisition of the incident light intensity baseline-the light signal intensity in the absence of gas absorption. For gases with broad-band absorption characteristics, such as alkanes (e.g., propane, butane), absorption lines are densely packed and continuous, often lacking distinct absorption-free regions. This characteristic renders traditional baseline correction methods-such as the absorption-free environment prediction method or polynomial fitting-ineffective or completely unreliable in real-world industrial dynamic monitoring scenarios, creating a critical bottleneck for high-precision applications of this technology.
Addressing the urgent industrial safety demand for high-precision, real-time monitoring of alkane marker gases-particularly in applications like oil and gas tank leak detection-the research team innovatively proposed a dual-wavelength baseline reconstruction strategy grounded in physical principles. The core concept leverages physical correlations within the optical path rather than relying on complex algorithmic assumptions or extensive data training. The team discovered that within multiple-reflection absorption cells, fluctuations in light intensity caused by factors like temperature changes and optical component jitter exhibit strong correlations across different wavelengths.
Leveraging this physical mechanism, the team successfully established a robust linear model linking a target wavelength (1686 nm, primarily for monitoring propane and butane) with a reference wavelength (1653 nm, primarily for monitoring methane). By continuously monitoring precisely measurable baseline variations in the reference wavelength channel, this model synchronously and accurately reconstructs the unknown baseline at the target wavelength's broad absorption band. This approach resolves the challenge of lacking "anchor points" for baseline calibration in broad-spectrum gases.
Experimental validation demonstrates that under dynamic temperature cycling conditions ranging from -10°C to 30°C, the relative root-mean-square error of this baseline reconstruction method remains below 1.63%. When applied to absorbance calculations for propane, butane, and their mixtures, the reconstructed baseline introduced a maximum relative error of only 1.7%. This work transforms a core measurement challenge in direct absorption spectroscopy into a solution based on measurable physical correlations, offering a new approach for high-precision, real-time online monitoring of broad-spectrum absorbing gases in complex industrial environments such as petrochemical safety production.
