Research on uncertainty of optical feedback linear cavity ringdown spectroscopy

Abstract

Cavity ring-down spectroscopy (CRDS) is a highly sensitive molecular absorption spectroscopic technology, which has been widely used in mirror reflectance measurement, atmospheric trace gas detection, molecular precision spectroscopy and other fields. It deduces the intracavity absorption by measuring the rapid variation of the ringdown signal. As a result, detector with high linearity, broad bandwidth and low electrical noise is indispensable. Additionally, due to the large noise in laser frequency, low laser to cavity coupling efficiency is resulted. Consequently, the cavity transmission is faint, which deteriorates the detection sensitivity. Optical feedback could address this problem by locking the laser to the cavity longitudinal mode. Then, the laser frequency noise is suppressed and hence better detection sensitivity is expected. Optical feedback CRDS has been widely studied with V-shape cavity. Compared to Fabry-Perot cavity, this cavity geometry is more sensitive to mechanical vibration and possesses lower finesse with an additional mirror. In this paper, optical feedback linear cavity ring-down spectroscopy based on a Fabry-Perot cavity with a finesse of 7800 is presented. The principle of the combination of optical feedback and linear cavity is explained from the perspective of the light phase, which shows the reflection would not generate efficient optical feedback if the feedback phase is properly controlled and laser to cavity locking could be therefore realized. And then, the factors influencing the stability of ring-down signal is analyzed, including the feedback ratio, the trigger voltage for the ringdown event and the distance between the light spot and the detector center. The experimental results show a superior fractional uncertainty of the empty ringdown time of 0.026% could be attained with a low feedback rate (3% FSR), a high ringdown signal trigger threshold (90% cavity mode amplitude) and superposition of the light spot with the detector center. With Allan variance analysis, the white noise response of 1.56×10^-9 cm^-1/ HZ^-1/2 and the detection sensitivity of .29×10^-10 cm^-1 for trace gas detection could be achieved at the integration time of 180 s, corresponding to the lowest CH₄concentration detection of 0.35 ppb at 6046.9cm^-1. This robust spectroscopic technique paves the way for the construction of high sensitive and stable cavity based instrument for trace gas detection.