Dual-comb spectroscopy for the characterization of laboratory flames

TL;DR

The paper confronts the challenge of non-invasively measuring unburned methane in flames within the mid-infrared, where calibration-free, high-precision spectroscopy is highly desirable. The authors implement a mid-IR dual-comb spectrometer based on electro-optical comb generators and difference-frequency generation to produce a MIR comb centered at $3427.43$ nm, employing differential detection for calibration-free spectra. They demonstrate a detection limit of $1.1$ ppm for a $1$ m path and map spatial CH$_4$ distributions across a McKenna burner, while also resolving flame instabilities with dominant frequencies near $1$ Hz and $8$–$9$ Hz; two optimized comb configurations are identified and validated. Overall, the EO-DC architecture provides a flexible, sensitive, and transportable tool for advanced flame diagnostics with potential for broader deployment in laboratory and field settings.

Abstract

Optical spectroscopy, in particular dual-comb (DC) spectroscopy, is a critical, non-invasive tool for combustion diagnostics, offering high precision and calibration-free advantages. However, its implementation remains challenging, especially in the mid-infrared region. This work presents the development of a robust DC spectroscopic system based on electro-optical (EO) frequency comb generators and difference frequency generation (DFG), specifically designed for the characterization of laboratory flames. Operating at a center wavelength of 3427.43 nm, the system utilizes a differential detection strategy to enable precise, calibration-free measurements of unburned methane ($\mathrm{CH_{4}}$) concentrations in a McKenna burner. The experimental results demonstrate a detection limit of 1.1 ppm for a 1 m path length and effectively resolve spatial concentration gradients across the combustion region. Furthermore, the system's high temporal resolution allowed for the identification of dynamic combustion instabilities, including self-sustained pulsations and fuel leakage under fuel-lean conditions. These findings validate the proposed EO architecture as a flexible and highly sensitive tool for advanced flame characterization.

Figures (7)

• Figure 1: Experimental setup for dual-comb generation and flame characterization. TEC: temperature controller, EDFA: erbium-doped fiber amplifier, AOM: acousto-optic modulator, PM: phase-modulator, PPLN: DFG module, PD: photodetector.
• Figure 2: Translation platform integrating the optical components and mid-infrared photodetectors for flame measurements. The laser propagation path is indicated for visualization purposes.
• Figure 3: Peak absorption for the $\mathrm{CH_{4}}$ spectral features between 3390 nm and 3440 nm with largest absorption as a function of (a) temperature for a concentration of 0.01 VMR and (b) concentration for a temperature of 1200 K. Simulations were done in both cases using HITEMP for a pressure of 1 atm and a path length of 7 cm.
• Figure 4: Fitted $\mathrm{CH_4}$ concentration (a) and its standard deviation (b) on the simulated 0.01 VMR measurements for the 308 studied comb configurations. Each point corresponds to the average or the standard deviation, respectively, of 100 simulated measurements and fittings. The insets show only the most viable comb configurations and the legend is given by ranges.
• Figure 5: Evolution of $\mathrm{CH_{4}}$ concentration in a 5-second continuous measurement at the burner base ($z=1$ mm) for an equivalence ratio of $\gamma = 0.7$. The chosen 13-tooth configuration was used and the complete measurement was analyzed in small intervals of 0.025 s. The inset shows the corresponding FFT, where the zero-frequency component has been removed.