飞秒-纳秒双脉冲激光诱导击穿光谱(LIBS)技术对合金的定量分析

激光诱导击穿光谱(Laser-induced breakdown spectroscopy, LIBS)技术几乎不受聚变环境中的强磁场影响，是一种最有希望实现托卡马克装置中面向等离子体材料(Plasma facing materials, PFMs)原位在线诊断的技术，已被用于多个托卡马克PFMs壁诊断。然而，LIBS技术对PFMs表面元素的探测限、定量分析以及PFMs的服役状态判定依旧面临很大挑战。采用同轴飞秒-纳秒激光协同技术，建立了飞秒-纳秒双脉冲激光诱导击穿光谱(fs-ns-DP-LIBS)技术，通过高峰值功率、低激光能量的飞秒激光诱导等离子体，再用纳秒激光增强常规单脉冲LIBS技术信号发射强度，进而提升常规单脉冲LIBS的探测灵敏度，同时结合6种合金标准样品，采用fs-ns-DP-LIBS技术对样品中的主要元素进行了定量分析，并进一步结合机器学习方法对6种合金进行种类判别。结果显示：在纳秒单脉冲和飞秒单脉冲LIBS检测中，Ni、Fe和Mo在400～800 nm波段没有观察到明显特征峰，仅观察到Cr的特征峰；在飞秒-纳秒脉冲间2μs延时，NiⅠ498.02 nm、FeⅠ517....

Quantitative Analysis of Alloys using Femtosecond-Nanosecond Dual-Pulse Laser-Induced Breakdown Spectroscopy(LIBS)

Laser Induced Breakdown Spectroscopy (LIBS) technology is minimally affected by strong magnetic fields that exist in tokamak nuclear fusion devices. This makes LIBS as a promising technique for in-situ diagnosis of Plasma Facing Materials (PFMs) in tokamaks. Currently, LIBS has been employed for diagnosing PFMs in several tokamaks. However, the LIBS method still faces significant challenges, including the limitation of detection (LOD) for surface elements of PFMs, quantitative analysis, and determination of the serviceability status of PFMs in tokamaks. In this study, a co-axial femtosecond-nanosecond dual-pulse laser-induced breakdown spectroscopy (Fs-ns-DP-LIBS) technique was developed. In the fs-ns-DP-LIBS technique, a femtosecond laser with high peak power and low laser energy was employed to induce a plasma, and then a nanosecond laser was used to enhance the emission intensity of LIBS signal. This enhancement aimed to improve the LOD of conventional single-pulse LIBS. Moreover, six standard samples were used to quantify the major elements using the fs-ns-DP-LIBS technique, and several machine learning methods were further employed to classify the six standard samples. The results showed that under single nanosecond and single femtosecond LIBS measurements, except for the Cr emission lines, the Ni, Fe, and Mo emission lines were not observed in the wavelength range of 400-800 nm. However, it is worth noting that the Ni I 498.02 nm, Fe I 517.16 nm, Fe I 523.85 nm, Mo I 588.83 nm, and Mo I 603.07 nm emission lines were clearly observed when a 2μs delay time between the femtosecond and nanosecond laser was used for Fs-ns-DP-LIBS. Additionally, the emission intensity of the Cr emission line was significantly enhanced. Compared to nanosecond single-pulse LIBS measurements, the emission intensity of the Cr I 534.58 nm line increased by approximately seven times. The quantitative analysis results showed that the calibration curves generated by the fs-ns-DP-LIBS technique had higher fitting degrees (R2), and the LOD for Cr was improved by approximately 3.5 times. Furthermore, classification studies were conducted using decision trees, nearest neighbors, linear discriminant analysis, and support vector machines to classify the six samples. The results showed that the classification prediction accuracy of linear discriminant analysis and support vector machines exceeded 99%. These findings highlight the potential of the presented method for diagnosing surface elements of PFMs, quantitative analysis, and determination of the serviceability status of PFMs.