Towards calibration transfer for quantitative analysis of nuclear materials by LIBS
Résumé
To ensure the quality of their products, manufacturers of nuclear materials currently rely on wet chemistry methods coupled with ICP-MS and ICP-AES. These methods are slow and sometimes hard to implement, which leads the manufacturers to seek new possibilities of online, fast and direct preliminary quality control. Thus, the issue of quantifying metal impurities in both uranium and plutonium matrices by LIBS has been raised. But the lack of calibrated samples and the harsh security constraints hinder any possibility of complete analytical development on these materials. Therefore, we propose a method relying on calibration transfer from several non-nuclear metal matrices to uranium and plutonium matrices. Calibration transfer is widely studied and used in various spectrometry methods in order to overcome slope differences that appear in calibration lines between different measuring instruments, different elements or matrices. Here, the latter case will be studied for each analyte, by measuring experimental data such as ablated mass of the sample (m0), and electron temperature (Te) and density (Ne) of the plasma, normalizations of the LIBS signal of one line will be advanced to try to correct the slopes of the calibration curves and, ideally, end up with only one normalized calibration curve for all matrices. This uniformed calibration will then be used to quantify impurities in uranium and plutonium without having to make a full calibration on those matrices, provided m0, Te and Ne are known.Temperature and density are determined in various metal matrices such as steel, aluminum, copper, nickel and titanium alloys. Since we aim to compare measurements between different matrices, it appears critical to use the same measurement method in all of them in order to avoid biases as much as possible. To achieve this, it is necessary to use lines from minor elements that can be found in most samples. Thus, we choose iron as our main element of interest and use its lines to assess Te and Ne through Boltzmann-Plots and Stark Broadening measurements respectively. Considering those methods are rarely used with minor elements, we strive to validate their precision and robustness in several ways. For Te we checked the coherence between measurements of iron lines and another minor element (titanium) on the one hand, and between measurements of lines with high and low energy levels on the other hand. For Ne we compared measurements obtained with the Stark Broadening method to that given by the Saha and Boltzmann equations.We first show that the Boltzmann-Plots obtained from the large variety of neutral iron lines show a very good linearity, which is consistent with the hypothesis of Local Thermodynamic Equilibrium. Moreover, preliminary measurements of Ne validate the McWhirter criterion. Then, we have demonstrated the possibility of reliably measuring Te in various metallic samples using lines of minor elements (thousands of ppm). This enables us to use a single set of lines to measure this parameter in various matrices with a unified approach. Lastly, the comparison of measured Te in different matrices show weak but nevertheless meaningful variations from one matrix to another, with up to 10% variation between aluminum and copper alloys it enables us to investigate matrix effects in the ablation plasma, which will complement previous results on the ablated mass and forthcoming results on the electron density. In this latter case, the choice of favorable lines of minor elements common to all matrices is quite limited and the development of a unified methodology is more challenging than for the plasma temperature. Preliminary investigations will be presented.
Origine : Fichiers produits par l'(les) auteur(s)