Jean-Baptiste Sirven

Laser-induced breakdown spectroscopy is a preferred technique for fast and direct multi-elemental mapping of samples under ambient pressure, without any limitation on the targeted element. However, LIBS mapping data have two peculiarities: an intrinsically low signal-to-noise ratio due to single-shot measurements, and a high dimensionality due to the high number of spectra acquired for imaging. This is all the truer as lateral resolution gets higher: in this case, the ablation spot diameter is reduced, as well as the ablated mass and the emission signal, while the number of spectra for a given surface increases. Therefore, efficient extraction of physico-chemical information from a noisy and large dataset is a major issue. Multivariate approaches were introduced by several authors as a means to cope with such data, particularly Principal Component Analysis. This technique is useful to analyse correlations between different elements, but it is limited to low signal-to-noise ratios. In this paper, we introduce HyperPCA, a new analysis tool for hyperspectral images based on a sparse representation of the data using Discrete Wavelet Transform and kernel-based sparse PCA to reduce the impact of noise on the data and to consistently extract the spectroscopic signal, with a particular emphasis on LIBS data. The method is first illustrated using simulated LIBS mapping datasets to emphasise its performances with an extremely low shot-to-shot signal-to-noise ratio, and with a variable degree of spectral interference. Comparisons to standard PCA and to traditional univariate data analyses are provided. Finally, it is used to process real data in two cases that clearly illustrate the potential of the proposed algorithm. We show that the method presents advantages both in quantity and quality of the information recovered, thus improving the physico-chemical characterization of analysed surfaces.

La technique LIBS présente des avantages uniques pour réaliser des mesures élémentaires en environnement difficile. Elle est développée en France et dans d’autres pays pour répondre à de nombreux besoins de la filière électronucléaire. Elle nécessite pour cela une instrumentation variée capable de s’adapter aux diverses problématiques analytiques, configurations et contraintes de mise en œuvre propres au secteur nucléaire.

Exposure assessment is a key step in the evaluation of the risk induced by the handling of engineered nanomaterials. It is a very complex task, because several properties of nanoparticles are assumed to have an effect on their hazards. For exposure monitoring at the workplace, real-time onsite measurements are commonly implemented to measure the particles size and number density, whereas the sampled material is subsequently analysed by electron microscopy. A complementary approach would consist in doing onsite chemical analysis of the filter samples, in order to routinely monitor a potential chronic exposure. Laser-induced breakdown spectroscopy (LIBS) has distinctive advantages for this purpose. Therefore, this work aims at evaluating the performances of LIBS to assess the exposure to airborne carbon nanotubes (CNTs) at the workplace. As carbon is a ubiquitous element in the environment, our strategy was to target metal impurities in CNTs, aluminum and iron in our case. Then, we proceeded in three steps. First, we optimized the choice of the filter type to get the lowest detection limit for both elements. Secondly, this filter was used to quantitatively measure deposited CNTs. Eventually, we conducted an onsite measurement campaign in an industrial CNT production plant to evaluate the exposure in a real situation. We demonstrated that we could reach a detection limit for CNTs compliant with the current NIOSH recommendation of 1 μg m⁻³, and that the detected CNTs during the onsite campaign in areas accessible to workers were at an extremely low concentration, several orders of magnitude lower than this recommendation.

Laser‐induced breakdown spectroscopy (LIBS) is an analytical technique allowing the determination of elemental concentrations in a variety of matrices in the solid, liquid, and gaseous phases. Because of the inherent complexity of the signal and to the high dimensionality of experimental data, chemometrics has been more and more applied in LIBS to perform samples identification or quantitative measurements. But multivariate methods can also be used for the description and physical interpretation of the plasma, particularly to exploit the temporal dimension of the LIBS signal, which is usually neglected in spectrochemical measurements.In this work, time‐resolved spectra of a pure aluminum sample were treated with 2 methods, mean field‐independent components analysis and multivariate curve resolution–alternating least squares, applying non‐negativity constraints for scores and components in both cases. Results obtained were compared with reference univariate measurements of the emission of the species of interest (ions, neutral atoms, and molecules). The interpretation of scores and components provided a physical description of phenomena that take place between species in the plasma, like ionic recombination and molecules formation. Overall, mean field‐independent components analysis and multivariate curve resolution–alternating least squares yield equivalent solutions with our dataset. This new approach is very promising for the treatment of time‐resolved data obtained by LIBS.