As a major component of fission products, neodymium (Nd) is labelled as neutron poison due to its large cross section for thermal neutron capture. Hence, the concentration of Nd needs to be known and kept at a certain level throughout the nuclear power generation process, ensuring the quality of nuclear fuel material and also the safety of fission reactors. The traditional analytical techniques such as X-ray fluorescence spectroscopy (XRF), inductively coupled plasma-optical emission spectrometry (ICP-OES) and inductively coupled plasma-mass spectrometry (ICP-MS) are commonly employed to monitor the concentration of Nd in irradiated fuel, but the challenge of identifying materials in highly radioactive environment must be addressed. Generally, both the sample handling and the operation processes for these traditional techniques are very complicated, costly and time consuming, which could greatly increase the radiation hazards for the operators. At same time, the dissolution step by using these techniques can produce a large amount of radioactive liquid waste, which might cause environmental pollution if that is improperly treated. Therefore, more timely and safe analytical techniques for the special nuclear materials are urgently needed. Laser-induced breakdown spectroscopy (LIBS), a new form of atomic emission spectroscopy (AES), has been widely received and deployed as a promising analytical technique because of its rapid analysis capability, no sample preparation, multielement analysis and stand-off analysis abilities. In LIBS, a pulsed laser beam is focused onto a sample target to generate a high temperature luminous plasma. As the laser-induced plasma cools, the emission of characteristic light from the plasma can be analyzed by a spectrometer for the qualitative and quantitate information about the elemental composition of the sample. In order to explore the feasibility of the application of LIBS in nuclear industry and assess the ability of the technique for nuclear material analysis, in this study, a laboratory LIBS system combining with calibration curve was used to quantitatively analyze the content of Nd in high purity graphite samples. A total of 5 homemade Nd2O3-graphite mixture samples were prepared for LIBS experiment, and the Nd concentrations varied from 0% to 20 %. Based on the analysis of LIBS spectral characteristics of Nd element in graphite matrix, two atomic emission lines (Nd I 489.69 and 492.45 nm) and two first ionic emission lines (Nd II 490.18 and 531.98 nm) were identified and selected as the analysis lines, and subsequently used for the quantitatively analysis. The effects of laser pulse energy parameters on the spectral line intensity and signal to back ratio (SBR) were also studied, and finally the pulse energy of 45 mJ was selected as optimum laser fluence for LIBS experiments. Then, the calibration curve based on the peak intensity, the peak area for each analysis line and the corresponding Nd concentrations were constructed and fitted. The results showed that the calibration curves for Nd element constructed by using the peak intensity and peak area methods have good linear correlations. Using the peak intensity methods, the linear relationships of the regression coefficients (R2) for the four analysis lines were 0.9697, 0.9828, 0.9833 and 0.9733, respectively. Compared with the case of peak intensity, the linear correlation based on peak area was a little bit better, and the value of R2 for the analysis lines were 0.9898, 0.9912, 0.9819 and 0.9823, respectively. The maximum value of R2 of 0.9912 demonstrated that the Nd II 490.18 nm line had the best linear trend with respect to Nd concentration. Meanwhile, for the cases of 489.69, 490.18 and 531.98 nm lines by using the peak intensity method, all of the values of R2 could get approximately 1% increase corresponding to that obtained by the peak intensity method. For each analysis line, the limits of detection (LOD) values were calculated to be 0.13%, 0.12%, 0.09% and 0.06% for the peak intensity cases, all the values were much lower than the concentration of Nd in spent fuel. However, LODs for the peak area cases were 0.53%, 0.53%, 0.45% and 0.28%, respectively, which were significantly greater than that of the peak intensity cases. Analytical predictive skill of LIBS technique was studied further using the calibration curves generated from the peak area cases. Most of the relative errors between LIBS measurement concentration and the actual concentration were within 11% except for the case of Sample 3. For Nd II 490.18 nm line, the best prediction performance for Nd content in Nd2O3-graphite mixture samples was found, and the relative deviations for the samples were 1.16%, 9.35%, 6.77% and 1.87%, respectively. The good prediction performance showed the robustness of LIBS technique. The achievements of this study demonstrated that LIBS spectral analysis was capable of monitoring special elements in the nuclear materials and fuels, and these results strongly supported LIBS technique as a viable analysis method for the popularization and application in nuclear industry. © 2022, Youke Publishing Co., Ltd. All right reserved.
