An ultra-highly sensitive LITES sensor based on multi-pass cell with ultra-dense spot pattern designed by multi-objective algorithm

Abstract Multi-pass cell (MPC) with long optical path length (OPL) and high ratio of optical path length to volume (RLV) can significantly enhance the detection performance of light-induced thermoelastic spectroscopy (LITES) sensor and facilitate system integration. In this paper, an ultra-highly sensitive LITES sensor based on a MPC with ultra-dense spot pattern designed by multi-objective algorithm of parallel nondominated sorting genetic algorithm II (PNSGA-II) was reported for the first time. Five MPCs featuring different dense spot patterns were generated using this PNSGA-II algorithm. The experimental measured OPLs with an excellent RLV of > 20 cm −2 closely matched the theoretical results, validating the PNSGA-II algorithm and our MPC calculation model as reliable guidance for designing MPCs with superior performance. An acetylene (C 2 H 2 )-LITES sensor system was constructed using the designed MPC with an actual OPL of 80.14 m. A self-designed round-head quartz tuning fork (QTF) with resonant frequency of 9.5 kHz was used as the detector. The C 2 H 2 -LITES sensor demonstrated good linear response to C 2 H 2 concentration. The achieved minimum detection limit (MDL) for the C 2 H 2 -LITES sensor was 4.78 ppb, which has a 3.45-fold enhancement when compared to the standard commercial QTF with resonant frequency of 32.768 kHz. Furthermore, according to Allan deviation analysis, the MDL could improve to 891 ppt at an average time of 200 s, demonstrating its ultra-highly sensitive detection performance.