This paper proposed an optical detection mechanism that utilized orthogonal phase modulation (OPM) and Lissajous mode decoupling within a light-induced thermoelastic spectroscopy (LITES) sensor for the first time, enabling real-time multi-component gas sensing. By establishing a two-dimensional forced vibration model of a quartz tuning fork (QTF) under dual-path OPM, it revealed the vibration mode coupling mechanism induced by non-ideal phase and amplitude conditions, thereby providing a theoretical basis and optimization pathway for suppressing channel crosstalk. In the OPM-LITES sensor, two continuous-wave distributed feedback lasers were modulated by sinusoidal waves with the same frequency ( f ₀ /2, where f ₀ is the resonant frequency of the QTF) but a phase difference of 45°. The combined beam, after wavelength division multiplexing, co-excited the self-designed low-frequency QTF. The piezoelectric signal generated by the QTF underwent orthogonal demodulation via a lock-in amplifier (LIA). The concentration information for the two gases was independently retrieved from the decoupled X -component and Y -component of the LIA. Experimental results demonstrated that the OPM-LITES sensor exhibited excellent linear responses to both methane (CH ₄ ) and acetylene (C ₂ H ₂ ), with average relative systematic errors of 0.54% and 0.79%, and maximum relative errors of 1.72% and 1.38%, respectively. Allan deviation analysis indicated that the minimum detection limits (MDLs) for CH ₄ and C ₂ H ₂ reached 0.32 ppm and 0.29 ppm, with normalized noise-equivalent absorption coefficients of 5.09 × 10 ⁻⁹ cm ⁻¹ ·W·Hz ^−1/2 and 6.18 × 10 ⁻⁹ cm ⁻¹ ·W·Hz ^−1/2 , respectively. This study not only provided a simple and efficient solution for real-time multi-component gas detection but also established a novel theoretical framework for mode decoupling and signal processing in spectroscopic sensing.

Photoacoustic spectroscopy (PAS), quartz-enhanced photoacoustic spectroscopy (QEPAS), and light-induced thermoelastic spectroscopy (LITES) represent indirect absorption spectroscopy techniques for trace gas sensing, whose performance has long been advanced through hardware-oriented enhancement strategies. However, as hardware technologies continue to advance, conventional hardware-based enhancements are increasingly bottlenecked by weak responses, complex cross-interferences, and coupled multiphysics parameters. To transcend these limitations, algorithm-assisted methods, including traditional algorithms, machine learning, deep learning, and intelligent optimization, are being systematically integrated into these spectroscopic systems. This review summarizes recent progress in intelligent indirect absorption spectroscopy from three interconnected dimensions. First, we outline advanced signal processing and spectral reconstruction strategies designed to achieve weak-signal recovery and background noise suppression. Second, the focus shifts to data-driven parameter inversion, showing how multidimensional artificial intelligence models contribute to concentration retrieval, environmental compensation, multicomponent recognition, spectral-overlap decoupling, and front–back-end collaborative waveform coding and demultiplexing. Third, intelligent system optimization is examined, in which surrogate modeling, swarm-intelligence search, physics-guided topology optimization and multi-objective algorithms are employed to improve the design efficiency of the key elements such as photoacoustic resonators and multipass cells (MPCs). Additionally, prospects for future technological developments are also discussed in the concluding section.

In this paper, an all-fiber light-induced thermoelastic spectroscopy (LITES) sensor based on hollow-core anti-resonant fiber (HC-ARF) and self-designed low-frequency quartz tuning fork (QTF) is reported for the first time. By utilizing HC-ARF as both the transmission medium and gas chamber, the laser tail fiber was spatially coupled with the HC-ARF, and the end of the HC-ARF was directly guided onto the QTF surface, resulting in an all-fiber structure. This design eliminated the need for lens combinations, thereby enhancing system stability and reducing cost and size. Additionally, a self-designed rectangular-tip QTF with a low resonant frequency of 8.69 kHz was employed to improve the sensor’s detection performance. Acetylene (C₂H₂), with an absorption line at 6534.37 cm⁻¹ (1.53 μm), was chosen as the target gas. Experimental results clearly demonstrated that the detection performance of the rectangular-tip QTF system was 2.9-fold higher than that of a standard commercial QTF system. Moreover, it exhibited an outstanding linear response to varying C₂H₂ concentrations, indicating its high sensitivity and reliability in detecting C₂H₂. The Allan deviation analysis was used to assess the system’s stability, and the results indicated that the system exhibits excellent long-term stability.

A tunable diode laser absorption spectroscopy (TDLAS) sensor with a highly sensitive dual-component for methane (CH₄) and acetylene (C₂H₂) detection is reported in this paper for the first time. A multi-pass cell (MPC) design model was established employing a vector-based ray-tracing method. A dual-channel MPC with an interlaced dual hexagonal star pattern was designed to improve gas absorption and realize real-time synchronous detection of CH₄ and C₂H₂. During the simultaneous continuous monitoring of CH₄ and C₂H₂, the sensor exhibited an excellent linear response to concentration variations. The minimum detection limit (MDL) for CH₄ reached 132.08 ppb, improving to 77.32 ppb when the average time was increased to 300 s. In the case of C₂H₂, the MDL was measured at 20.19 ppb and further reduced to 3.50 ppb under the same extended average time.

In this paper, a highly sensitive carbon monoxide (CO) light-induced thermoelastic spectroscopy (LITES) sensor based on tapered-fiber focusing and a fiber-coupled multipass cell (MPC) is demonstrated for the first time. An MPC with a fiber-coupled structure and an optical length of 40 m is employed to significantly increase the effective absorption path length, reduce the difficulty of optical alignment, and improve the robustness of the sensor system. By comparing five different beam shaping and focusing schemes at the MPC output, the influence of the excitation spot size on the generation of the LITES signal is thoroughly investigated. Experimental results demonstrate that the system response is significantly enhanced as the spot size is compressed. Owing to the micron-scale focused spot produced by the tapered fiber end face, the tapered-fiber configuration exhibits the strongest sensing response among the five schemes. This configuration achieves the highest signal-to-noise ratio (SNR) and exhibits a good linear response to CO concentration. The minimum detection limit (MDL) of the proposed CO-LITES sensor is 3.86 ppm. Based on Allan deviation analysis, when the system integration time reaches 300 s, the minimum detection limit can be further improved to 304 ppb, demonstrating excellent long-term stability.