Com clima diversificado e terras férteis, o Brasil tem vocação natural para o desenvolvimento agropecuário e todas as suas vertentes. Assim, o agronegócio é hoje a principal locomotiva da economia brasileira, representando cerca de um terço do nosso Produto Interno Bruto (PIB). Nesse contexto, o Brasil é o terceiro maior produtor de frutas do planeta, com destaque para a produção de laranjas. Particularmente, o país lidera a produção mundial de suco de laranja e conta com uma participação de 85% nas exportações deste produto. Porém, um dos principais fatores atuais que restringem os lucros e a expansão da citricultura é o controle fitossanitário. Atualmente, dentre as principais doenças podemos destacar o HLB e o cancro cítrico. Ambas, doenças bacterianas que não têm cura comprometem a produção e desenvolvimento da fruta e levam à morte da árvore. Dessa maneira, o monitoramento destas é fundamental para evitar danos aos frutos e a necessidade da erradicação de plantações inteiras. O presente trabalho avaliou o emprego das técnicas de Espectroscopia de Fluorescência Induzida por Laser (LIFS) e Espectroscopia de Emissão Óptica com Plasma Induzido por Laser (LIBS) como forma de diagnóstico destas doenças, se apresentando como uma alternativa às inspeções visuais e ao PCR utilizados atualmente. Para isso, folhas de citros in natura provenientes de plantas sadias, com HLB ou com cancro foram amostradas e medidas com ambos os sistemas. Através do sistema LIFS foi realizado um estudo de precocidade no diagnóstico do HLB. Este sistema fotônico pôde detectar a doença até 21 meses antes do aparecimento dos primeiros sintomas visuais. Foram utilizados classificadores sazonais, criados a partir de Regressão por Mínimos Quadrados Parciais (PLSR) e conjuntos de calibração previamente avaliados. Quanto ao cancro cítrico e o sistema LIFS, taxas de acerto superiores a 90% foram alcançadas nos melhores casos de validação cruzada dos dados. A diferenciação do cancro e do HLB também foi possível pela mesma técnica ao avaliar um conjunto pequeno de dados, que atingiu uma taxa de acerto de 82%. Através dos espectros obtidos pelo sistema LIBS, avaliou-se as variações nutricionais causadas na planta devido à doença. Por meio de tais variações e métodos de PLSR foi construído um modelo para o diagnóstico de HLB alcançando taxas de acerto da ordem de 75%. Em relação ao cancro cítrico e o sistema LIBS, um estudo preliminar foi realizado, e uma taxa de acerto superior a 90% foi atingida. Por fim, os resultados corroboraram com a ideia inicial de se realizar o diagnóstico do HLB e do cancro cítrico através de técnicas fotônicas. Dentre outras vantagens estas permitem uma análise rápida, in loco, sem a necessidade de preparo de amostra. Além disso, para o HLB, as técnicas fotônicas demonstraram menor sensibilidade à distribuição não homogênea da doença, quando comparada com a técnica de referência (qPCR). / The diverse climate and fertile soils make Brazil a country with a natural vocation for agricultural development. Thus, agribusiness is now the main locomotive of the Brazilian economy, accounting for about one-third of our Gross Domestic Product (GDP). In this context, Brazil is the third largest fruit producer in the world, with emphasis on the orange production. Particularly, the country leads the world production of orange juice, and with a stake of 85% in exports of this product. However, one of the main factors that restrict current profits and the expansion of citrus production is phytosanitary control. Currently, among the major diseases we highlight the HLB (Greening) and Citrus Canker, two bacterial diseases that have no cure and affect production and fruit development. Therefore, monitoring is essential to prevent damage to the fruits and the complete eradication of infected orchards. The present study evaluated the use of Laser-Induced Fluorescence Spectroscopy (LIFS) and Laser-Induced Breakdown Spectroscopy (LIBS) techniques as alternative diagnostic methods to visual inspection and PCR technique, currently used. For that, citrus leaves from healthy, HLB or citrus canker infected plants, were sampled and measured with both systems. Through LIFS system, a study of the HLB early diagnosis was done. This photonic system could detect the disease up to 21 months before the tree show the visual symptoms. Seasonal classifiers were used, which were designed from Partial Least Square Regression (PLSR) methods and calibration database previously acquired. Regarding citrus canker and LIFS system, success rates higher than 90% were achieved in the best cases. The differentiation between citrus canker and HLB was also possible using the same technique reaching a success rate around 82%. Through the spectrum obtained by LIBS system, it was evaluated the nutritional variations caused in the plant due to HLB, and based on these data, the diagnosis was done. The average success rate was 75%, which was achieved by PLSR model. Regarding on citrus canker and LIBS system, a preliminary study was carried out and a success rate greater than 90% was achieved. Finally, the results corroborated with the use of photonic techniques for the HLB and citrus canker diagnosis. Among other advantages, they allow rapid analysis, in loco and without the need of sample preparation. In addition, for the HLB, photonics techniques showed lower sensitivity to the non-homogeneous distribution of the disease when compared with the reference technique (qPCR).

LIBS é um método baseado na espectrometria de emissão óptica que emprega a microamostragem por ablação por laser e formação de um microplasma para a determinação dos elementos químicos presentes na amostra. Neste trabalho, o método foi avaliado para a análise de embalagens plásticas produzidas com polietileno de alta densidade e polipropileno. A construção de curvas de calibração com amostras dos polímeros contendo diferentes concentrações de Cd, Cr e Pb e a determinação desses analitos, assim como os efeitos da taxa de repetição do laser, número de pulsos, fluência e diâmetro de focalização foram estudados. O sistema LIBS utilizado foi composto por um laser pulsado de Nd:YAG operando a 064 nm, com pulsos de 5 ns e máxima energia de 360 mJ/pulso. Os sinais de emissão foram coletados por um conjunto de lentes acoplado por fibra óptica ao espectrômetro com montagem Echelle e detector ICCD. Os parâmetros instrumentais foram ajustados em 10 pulsos acumulados, 2 \'mü\'s de atraso e 6 \'mü\'s de integração. O software ESAWIN, o banco de dados de espectros atômicos e iônicos do NIST e algoritmo desenvolvido em ambiente MATLAB® foram utilizados para aquisição e tratamento dos dados. A avaliação da topografia das crateras nas placas dos polímeros foi feita por perfilometria e microscopia eletrônica de varredura, que proporcionaram informações úteis para a caracterização das crateras. As massas removidas das placas de polietileno de alta densidade e de polipropileno foram calculadas a partir dos volumes das crateras obtidos pelas análises perfilométricas. A taxa de repetição e número de pulsos do laser tiveram efeitos marcantes na ablação dos polímeros, e as principais causas foram atribuídas ao grau de cristalinidade, à temperatura de fusão cristalina (Tm) e à temperatura de transição vítrea (Tg). Os resultados desta tese indicaram que diâmetros de focalização da ordem de 600 \'mü\'m e fluências entre 50 e 85 J cm-2 podem ser recomendados para análises de PEAD e PP por LIBS com laser de Nd:YAG@1064 nm (pulsos de 5 ns) e taxa de repetição de 10Hz. Nessas condições, as curvas de calibração obtidas apresentaram boa correlação. Identificaram-se três amostras contaminadas por Cd, Cr e Pb dentre 60 embalagens plásticas analisadas. A comparação dos resultados obtidos por LIBS e ICP OES apresentou concordâncias e algumas diferenças significativas (teste t ao nível de 95% de probabilidade) que foram associadas a efeitos de matriz observados na análise dos polímeros por LIBS / LIBS is an optical emission spectroscopy technique that employs a laser for micro sampling and induction of a plasma for determination of chemical elements in a sample. In this work, LIBS was evaluated for the analysis of plastic packaging produced with high density polyethylene (HDPE) and polypropylene (PP). Plates of these polymers with different concentrations of Cd, Cr and Pb were prepared for evaluating the effects of laser repetition rate, number of pulses, fluence and laser focusing as well as for building the analytical calibration curves. The LIBS system was designed by using a Q-Switch Nd:YAG laser operating at 1064 nm (5 ns, 360 mJ/pulse) and the emission signals were collected by lenses into an optical fiber coupled to a high-resolution echelle spectrometer equipped with ICCD. Instrumental parameters consisted of 10 accumulated laser pulses, 2 \'mü\'s delay time and 6 \'mü\'s integration time gate. Software ESAWIN, NIST atomic database and an algorithm develop in MATLAB® were used for acquisition and data processing. The evaluation of topographical features of craters on the plates of the polymers was carried out by perfilometry and scanning electron microscopy, and provides useful information for the characterization of the craters. The mass removal of HDPE and PP were calculated using craters volume obtained in perfilometric analysis. The repetition rate and number of laser pulses affect the ablation process in polymers. Moreover, the main properties of polymers that affect the ablation are degree of crystallinity, crystalline melting temperature (Tm) and glass transition temperature (Tg). The results indicated that laser focusing of 610 \'mü\'m and fluences between 50 and 85 J cm-2 can be indicated for LIBS analysis of HDPE and PP with Nd:YAG@1064 nm, 5 ns and 10 Hz. Under these conditions the calibration curves obtained presented good correlations and 3 samples containing Cd, Cr and Pb were identified from 60 samples from the local market. LIBS results compared well with ICP OES but some data presented significant differences by applying a t-test at 95 % confidence level, which were mainly associated to matrix effects observed in the analysis of polymers by LIBS

A influência das variáveis associadas ao laser como a fluência, o comprimento de onda, e a duração do pulso, assim como as relacionadas às propriedades das amostras, como a distribuição do tamanho das partículas e as características químicas da matriz, foram avaliadas visando à determinação quantitativa de macro- (P, K, Ca, Mg) e micronutrientes (Cu, Fe, Mn, Zn, B) em pastilhas de materiais vegetais por espectrometria de emissão óptica com plasma induzido por laser (LIBS). Os efeitos da distribuição do tamanho das partículas e da fluência do laser foram investigados a partir da análise de amostras peneiradas (150-20 µm) empregando-se um sistema LIBS, em regime temporal de nanossegundos (ns-LIBS), montado com um laser de Nd:YAG a 1064 nm (pulsos de 5 ns e 360 mJ)e um espectrômetro com óptica Echelle e detector ICCD. Observou-se um aumento na sensibilidade e na precisão das medições com a diminuição do tamanho das partículas usadas no preparo das pastilhas, e uma diminuição dos efeitos de matriz causados por diferenças no tamanho das partículas, quando pastilhas preparadas com partículas < 75 µm foram analisadas com pulsos de 50 J cm-2. Verificou-se uma melhora significativa na exatidão das medições de Mg, Fe, Mn e Zn feitas em um conjunto de pastilhas de folhas de cana-de-açúcar (partículas < 75 µm) empregando-se esta fluência. Em uma segunda etapa, observou-se que variações no comprimento de onda do laser de Nd:YAG (1064, 532, 355 e 266 nm) não influenciaram de maneira significativa a precisão e exatidão das medições em pastilhas de folhas de cana-de-açúcar, obtendo-se correlações lineares entre as intensidades dos sinais de emissão e as correspondentes frações de massa dos analitos. Variações no comprimento de onda do laser de Nd:YAG não afetaram a análise de um conjunto heterogêneo de amostras, composto por pastilhas de folhas de plantas de diferentes espécies, como soja, cana-de-açúcar, milho, citros, café, por exemplo, por ns-LIBS. No entanto, diferentemente do observado para o conjunto de pastilhas de folhas de cana-de-açúcar, observou-se, uma baixa correlação (r cal < 0,90) entre as frações de massa de Ca, Mg, P, Cu, Fe, Mn e Zn determinadas por ns-LIBS e os valores de referência, o que demonstra a baixa robustez de ns-LIBS frente às variações na composição química das matrizes. Posteriormente, analisou-se este conjunto heterogêneo de amostras com sistema LIBS, em regime temporal de femtossegundos (fs-LIBS), montado com laser de Ti:Safira (pulsos de 60 fs e 1,65 mJ) e espectrômetro com óptica Czerny-Turner e ICCD. Demonstrou-se que o uso de pulsos em regime de fs proporcionou calibrações e validação menos dependentes da composição química das matrizes. As frações de massa de Ca, Mg, P, Fe e Mn previstas por fs-LIBS foram estatisticamente concordantes com os valores de referência, independentemente do modelo de calibração usado. Além disso, o uso de calibração multivariada melhorou a capacidade preditiva de ns-LIBS, assemelhando-se à de fs-LIBS. Concluiu-se que o emprego de fs-LIBS foi a estratégia mais robusta e que ofereceu maior flexibilidade à variabilidade matricial / The influence of laser properties, such as fluence, wavelength and pulse duration, as well as sample characteristics, such as particle size distribution and chemical matrix composition, was evaluated aiming at the quantitative determination of macro- (P, K, Ca, Mg) and micronutrients (Cu, Fe, Mn, Zn, B) in pellets of plant materials by laser-induced breakdown spectroscopy (LIBS). Firstly, the effects of particle size distribution and laser fluence on the analysis of pellets (test samples) prepared with sieved samples (from 150 to 20 µm apertures) were investigated. Experiments were carried out with a nanosecond LIBS (ns-LIBS) system by using a Q-switched Nd:YAG laser at 1064 nm (5 ns; 360 mJ) and a spectrometer with Echelle optics and intensified charge-coupled device (ICCD) detector. Results indicated that smaller particles yielded to sensitivities\' enhancement and attained better measurements\' precision. Moreover, matrix effects were reduced by analyzing pellets prepared from < 75 ?m sieved fractions and pulses of 50 J cm-2. In addition, there was a significant improvement on accuracy of Mg, Fe, Mn and Zn measurements in a set of test samples of sugarcane leaves by using this fluence. In a second experiment, variations in the Nd:YAG laser wavelength (1064, 532, 355 and 266 nm) did not affect the analysis of test samples of sugarcane leaves, and provided linear correlations between emission signal intensities and corresponding analytes mass fractions. In addition, variations within Nd:YAG laser wavelength did not affect the analysis of a heterogeneous sample set composed by pellets of leaves from different crops, such as soy, sugarcane, maize, citrus and coffee by ns-LIBS. However, in contrast to previous findings, the univariate calibration models for ns-LIBS presented lower linearity (r cal < 0.90) for Ca, Mg, P, Cu, Fe, Mn and Zn, no matter the laser wavelength used for the analysis. These circumstances reflect the low robustness of ns-LIBS to variations within matrix chemical composition among test samples. Afterwards, test samples from different crops were analyzed by a femtosecond LIBS (fs-LIBS) by using a Ti:Sapphire laser, including a mode-locked oscillator and an ultrafast amplifier (60 fs; 1.65 mJ per pulse), and a spectrometer with Czerny-Turner optics and ICCD. Findings indicated that the pulse duration was a decisive variable for providing accurate quantification of nutrients in different plant species, which present substantial differences in terms of matrix chemical composition. Close agreement between Ca, Mg, P, Fe and Mn mass fractions predicted by fs-LIBS and those determined by ICP OES was evidenced, whatever the modeling approach used. Contrarily, for ns-LIBS analysis of test samples from different crops, only the use of multivariate partial least squares (PLS) regression appears capable for resolving the non-linear transformations of the emission intensities according to the physical mechanisms governing this temporal regime of ablation. Thus, when using multivariate modeling, the figures-of merit reflecting the predictive capabilities of ns-LIBS resemble to those achieved by fs-LIBS. Either way, fs-LIBS is a more robust approach that better offers larger flexibility to the matrix variability

La spectrométrie de plasma induit par laser, plus connue sous le nom de LIBS (l’acronyme du terme en anglais Laser-Induced Breakdown Spectroscopy) est une technique analytique qui permet la détection de l’ensemble des éléments du tableau périodique avec des limites de détection de l’ordre du ppm et ceci sur tous types d’échantillons qu’ils soient liquides, solides ou gazeux. Sa simplicité de mise en œuvre, sa rapidité et sa versatilité en font une technique très attractive avec un fort potentiel en termes d’applications que ce soit pour le contrôle en ligne, l’environnement ou l’exploration spatiale. Son point faible reste cependant son manque de fiabilité dans l’analyse quantitative, en particulier lors de l’étude d’échantillons hétérogènes ou de matrices complexes telles que les matrices organiques. Ce travail de thèse propose une étude des propriétés des plasmas induit par laser sur différentes familles de polymères. Une étude du plasma au temps court (~ns) par ombroscopie est tout d’abord présentée, ceci pour différents paramètres expérimentaux (énergie laser, durée d’impulsion, longueur d’onde). Un diagnostic complet du plasma par spectrométrie d’émission est ensuite détaillé pour différents délais de détection et montre que la mesure des températures des différentes espèces du plasma (atomique, ionique et moléculaire) permet de vérifier, dans certaines conditions, les hypothèses d’homogénéité et de l’équilibre thermodynamique local. Ceci permet alors la mise en place de procédures quantitatives telles que la méthode dite sans calibration (calibration free LIBS) tout en optimisant le rapport signal sur bruit de la mesure LIBS. Dans nos expériences cette optimisation est mise à profit pour l’identification de différentes familles de polymères en utilisant, pour le traitement des données de la spectroscopie LIBS, la méthode chimiométrique des réseaux de neurones artificiels. Les résultats obtenus, très prometteurs, permettent d’envisager l’utilisation de la LIBS pour l’identification en temps réel des matières plastiques sur chaine de tri. Par ailleurs et de manière plus générale, ce travail pourrait constituer une base solide pour aller étudier d’autres matériaux organiques plus complexes tels que des tissus biologiques. / Laser-Induced Breakdown Spectroscopy (LIBS) is an analytical technique that has the potential to detect all the elements present in the periodic table. The limit of detection can go below a few ppm and this regardless of the physical phase of the analyzed sample (solid, liquid or gas). Its simplicity of use, its rapidity to get results and its versatility provide this technique with attractive features. The technique is currently developed for applications in a large number of domains such as online control, spatial explorations and the environment. However the weakness of the LIBS technique, compared to other more conventional ones, is still its difficulty in providing reliable quantitative results, especially for inhomogeneous and complex matrix such as organic or biological materials. The work presented in this thesis includes a study of the properties of plasma induced from different organic materials. First, a study of the plasma induced on the surface of a Nylon sample at short time delays (~ns) was carried out using the time-resolved shadowgraph technique for different experimental parameters (laser energy, pulse duration, wavelength). Then, a complete diagnostics of the plasma was performed using the plasma emission spectroscopy. A detailed analysis of the emission spectra at different detection delays allowed us to determine the evolution of the temperatures of the different species in the plasma (atoms, ions and molecules). The homogeneity and the local thermodynamic equilibrium within the plasma was then experimentally checked and validated. We demonstrated that the optimisation of the signalto- noise ratio and a quantitative procedure, such as the calibration-free LIBS, can be put in place within a properly chosen detection window. In our experiments, such optimised detection configuration was further employed to record LIBS spectra from different families of polymer in order to identify and classify them. For this purpose, the chemometrics procedure of artificial neural networks (ANN) was used to process the recorded LIBS spectroscopic data. The promising results obtained in this thesis makes LIBS stand out as a potentially useful tool for real time identification of plastic materials. Finally, this work can also be considered as a base for the further studies of more complex materials such as biological tissues with LIBS.

La spectroscopie de plasma induit par laser (En anglais LIBS: laser-induced breakdown spectroscopy) est une méthode analytique de spectroscopie d'émission optique qui utilise un plasma induit par laser comme source de vaporisation, d'atomisation et d'excitation. Bien que la LIBS ait démontré sa polyvalence et ses caractéristiques attrayantes dans de nombreux domaines, les aspects quantitatifs de la LIBS sont considérés comme son talon d'Achille. D'un point de vue fondamental, cela peut être dû à la nature complexe du plasma induit par laser comme source d'émission spectroscopique. La caractérisation temporelle et spatiale du plasma induit par laser est considérée comme l'un des points clés pour comprendre les fondements de la technique LIBS. D'autre part, la LIBS est habituellement caractérisée par l'utilisation d'une ablation laser directe, sans traitement préalable de l'échantillon. Cela pourrait être assez limitant en particulier pour certains types de matériaux tels que des poudres ou des liquides. Une préparation adéquate ou un traitement approprié de l'échantillon permettant le dépôt d'un film mince et homogène de l'échantillon sur une surface métallique pourrait grandement augmenter le potentiel de la LIBS en vue d'obtenir de meilleures performances analytiques, et notamment une meilleure sensibilité et un effet de matrice réduit. On parle alors de LIBS assistée par surface car la matrice métallique contribue à une augmentation de la température du plasma. Le présent travail de thèse est donc motivé par deux aspects importants de la technique LIBS: la connaissance du plasma induit par laser comme source d'émission spectroscopique, et de nouvelles méthodes de préparation des échantillons pour améliorer la performance analytique de la LIBS, notamment pour des échantillons comme poudres et liquides visqueux. La première partie de cette thèse (chapitre 2) est consacrée à la caractérisation du plasma induit sur des échantillons de verre, en fonction de la longueur d'onde du laser, infrarouge (IR) ou ultraviolet (UV), et du gaz ambiant, de l'air ou de l'argon. L'imagerie spectroscopique et la spectroscopie d'émission résolue en temps et en espace sont utilisées pour le diagnostic du plasma. La deuxième partie de cette thèse est de développer des méthodes de préparation d'échantillons, déposés sur des surfaces métalliques pour l'analyse LIBS de poudres ainsi que de vins comme exemples de liquide. Au chapitre 3, nous avons appliqué la LIBS pour l'analyse quantitative dans des poudres (exemples de poudres : cellulose, alumine ainsi que de la terre). Au chapitre 4, nous avons appliqué la LIBS pour la classification des vins français selon leurs régions de production. Deux modèles de classification basées sur l'analyse des composants principaux (PCA) et la forêt aléatoire (RF) sont utilisés pour la classification. A l'aide de ces applications, ce travail de thèse démontre l'efficacité de la méthode LIBS assistée par surface pour l'analyse de poudres (cellulose, alumine et sols) et de liquides (vins), avec une limite de détection dans l'ordre de ou sous la ppm et une réduction significative de l'effet de matrice / Laser-induced breakdown spectroscopy (LIBS) is an analytical method with optical emission spectroscopy that uses a laser pulse to vaporize, atomize, and excite a hot plasma as the spectroscopic emission source. Although LIBS has demonstrated its versatility and attractive features in many fields, the quantitative analysis ability of LIBS is considered as its Achilles’ heel. From a fundamental point of view, this can be due to the complex nature of laserinduced plasma as the spectroscopic emission source for LIBS application. The temporal and spatial characterization of laser-induced plasma is considered as one of the key points for the LIBS technique. On the other hand, from the analytical point of view, LIBS is usually characterized by direct laser ablation. This can be however quite limiting, especially for some types of materials such as powders or liquids. Proper sample preparation or treatment allowing the deposition of a thin homogeneous film on a metallic surface could greatly improve the analytical performance of LIBS for these types of materials. Since the metallic surface is expected to contribute to increase the temperature and the density of the plasma and, consequently, to a better overall sensitivity, we call this technique surface-assisted LIBS. The present thesis work is therefore motivated by two basic aspects of LIBS analysis: the need of an improved knowledge of laser-induced plasma as a spectroscopic emission source, and new methods to improve the analytical performance of LIBS, including a higher sensibility and a reduced matrix effect. The first part of this thesis (Chapter 2) is dedicated to an extensive characterization of the plasma induced on glass samples, as a function of the laser wavelength, infrared (IR) or ultraviolet (UV), and the ambient gas, air or argon. Both the spectroscopic imaging and time- and space-resolved emission spectroscopy are used for plasma diagnostics in this work. The second part of this thesis is to develop a surface-assisted LIBS method for the elemental analysis in powders, and in wines as examples of liquids. We applied the surface-assisted LIBS for the quantitative elemental analysis in cellulose powders, alumina powders, and soils (Chapter 3). Special attentions are paid on the figures-of-merit, matrix effects, and normalization approaches in LIBS analysis. We also used the surfaceassisted LIBS for the classification of French wines according to their production regions (Chapter 4). Two classification models based on the principal component analysis (PCA) and random forest (RF) are used for the classification. Through these applications, this thesis work demonstrates the efficiency of the surface-assisted LIBS method for the analysis of powders (cellulose, alumina and soils) and of liquids (wines), with ppm or sub-ppm sensitivities and a reduced matrix effect

The elemental analysis of soil is useful in forensic and environmental sciences. Methods were developed and optimized for two laser-based multi-element analysis techniques: laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and laser-induced breakdown spectroscopy (LIBS). This work represents the first use of a 266 nm laser for forensic soil analysis by LIBS. Sample preparation methods were developed and optimized for a variety of sample types, including pellets for large bulk soil specimens (470 mg) and sediment-laden filters (47 mg), and tape-mounting for small transfer evidence specimens (10 mg). Analytical performance for sediment filter pellets and tape-mounted soils was similar to that achieved with bulk pellets. An inter-laboratory comparison exercise was designed to evaluate the performance of the LA-ICP-MS and LIBS methods, as well as for micro X-ray fluorescence (μXRF), across multiple laboratories. Limits of detection (LODs) were 0.01-23 ppm for LA-ICP-MS, 0.25-574 ppm for LIBS, 16-4400 ppm for µXRF, and well below the levels normally seen in soils. Good intra-laboratory precision (≤ 6 % relative standard deviation (RSD) for LA-ICP-MS; ≤ 8 % for µXRF; ≤ 17 % for LIBS) and inter-laboratory precision (≤ 19 % for LA-ICP-MS; ≤ 25 % for µXRF) were achieved for most elements, which is encouraging for a first inter-laboratory exercise. While LIBS generally has higher LODs and RSDs than LA-ICP-MS, both were capable of generating good quality multi-element data sufficient for discrimination purposes. Multivariate methods using principal components analysis (PCA) and linear discriminant analysis (LDA) were developed for discriminations of soils from different sources. Specimens from different sites that were indistinguishable by color alone were discriminated by elemental analysis. Correct classification rates of 94.5 % or better were achieved in a simulated forensic discrimination of three similar sites for both LIBS and LA-ICP-MS. Results for tape-mounted specimens were nearly identical to those achieved with pellets. Methods were tested on soils from USA, Canada and Tanzania. Within-site heterogeneity was site-specific. Elemental differences were greatest for specimens separated by large distances, even within the same lithology. Elemental profiles can be used to discriminate soils from different locations and narrow down locations even when mineralogy is similar.

This thesis deals with a general overview of laser-induced breakdown spectroscopy - ablation of material and plasma formation when a change of the surrounding environment occurs. The aim is to establish ideal conditions for improving detection, which is difficult for some elements when the measurement takes place in the ambient atmosphere. This is the reason why we change the ambient conditions, the gas (air, helium, argon) and its pressure. Buffer gases (helium, argon) influence the development of the material ablation and quality of generated plasma. That is why we inspect the changes in the signal according to particular atmospheres (signal change in buffer gases compared to the ambient atmosphere). The first part of the thesis presents the principle and physical nature of laser ablation with a brief search of experiments performed so far for similar purposes. Based on this search we determine conclusions about the influence of buffer gasses and pressure on the detection used gasses. In the second part of the work, the controlled experiment is presented, which consisted of the detection of spectral lines of fluorine and potassium, see chapter 5.6 and molecular transitions of calcium fluoride, see chapter 5.10. In a controlled experiment, we established two assumptions. The first assumption is the influence of the inert gas atmosphere, which should significantly improve the detection of fluorine and potassium compared to the air atmosphere. We confirmed this finding only for the argon atmosphere. In contrast, in the helium atmosphere, the detection of none of the elements improved. The second assumption is the effect of reducing the ambient pressure, which in combination with an inert gas should allow the detection of fluorine even at low concentrations in the sample. We consider the measured glass disk in chapter 5.7 to be such a sample. This assumption is not confirmed, as the increased intensity of the fluorine spectral line does not manifest itself in any way when the pressure in the vacuum chamber decreases. Finally, in chapter 7 from the results of individual parts of the experiment, the most suitable conditions for the future detection of the measured substances were proposed: fluorine, potassium and calcium fluoride molecules.

Under high irradiation, a fourth state of matter named plasma can be obtained. Plasmas emit electromagnetic radiation that can be recorded in the form of spectra for spectroscopic elemental analysis. With the advent of lasers in the 1960s, spectroscopists realized that lasers could be used simultaneously as a source of energy and excitation to create plasmas. The use of a laser to ignite a plasma subsequently led to laser-induced breakdown spectroscopy (LIBS), an optical emission spectroscopy capable of analyzing samples in various states (solids, liquids, gases) with minimal sample preparation, rapid feedback, and endowed with in situ capability. In this dissertation, studies of LIBS for multi-elemental analysis and geological applications are reported. LIBS was applied to cosmetic powders for elemental analysis, screening and classification based on the raw material used. Principal component analysis (PCA) and internal standardization were used. The intensity ratios of Mg/Si and Fe/Si observed in talcum powder show that these two ratios could be used as indicators of the potential presence of asbestos. The feasibility of LIBS for the analysis of gasification slags was investigated and results compared with those of inductively-coupled plasma−optical emission spectrometry (ICP-OES). The limits of detection for Al, Ca, Fe, Si and V were determined. The matrix effect was studied using an internal standard and PLS-R. Apart from V, prediction results were closed to those of ICP-OES with accuracy within 10%. Elemental characterization of outcrop geological samples from the Marcellus Shale Formation was also carried out. The matrix effect was substantially reduced. The limits of detection obtained for Si, Al, Ti, Mg, Ca and C were determined. The relative errors of LIBS measurements are in the range of 1.7 to 12.6%. Gate delay and laser pulse energy, have been investigated in view of quantitative analysis of variation of trace elements in a high-pressure environment. Optimization of these parameters permits obtaining underwater plasma emission of calcium with quantitative results on the order of 30 ppm within a certain limit of increased pressure. Monitoring the variation of the trace elements can predict changes in the chemical composition in carbon sequestration reservoir.

La caractérisation de matériaux par spectroscopie optique d’émission d’un plasma induit par laser (LIPS) suscite un intérêt qui ne va que s’amplifiant, et dont les applications se multiplient. L’objectif de ce mémoire est de vérifier l’influence du choix des raies spectrales sur certaines mesures du plasma, soit la densité électronique et la température d’excitation des atomes neutres et ionisés une fois, ainsi que la température d’ionisation. Nos mesures sont intégrées spatialement et résolues temporellement, ce qui est typique des conditions opératoires du LIPS, et nous avons utilisé pour nos travaux des cibles binaires d’aluminium contenant des éléments à l’état de trace (Al-Fe et Al-Mg). Premièrement, nous avons mesuré la densité électronique à l’aide de l’élargissement Stark de raies de plusieurs espèces (Al II, Fe II, Mg II, Fe I, Mg I, Halpha). Nous avons observé que les densités absolues avaient un comportement temporel différent en fonction de l’espèce. Les raies ioniques donnent des densités électroniques systématiquement plus élevées (jusqu’à 50 % à 200 ns après l’allumage du plasma), et décroissent plus rapidement que les densités issues des raies neutres. Par ailleurs, les densités obtenues par les éléments traces Fe et Mg sont moindres que les densités obtenues par l’observation de la raie communément utilisée Al II à 281,618 nm. Nous avons parallèlement étudié la densité électronique déterminée à l’aide de la raie de l’hydrogène Halpha, et la densité électronique ainsi obtenue a un comportement temporel similaire à celle obtenue par la raie Al II à 281,618 nm. Les deux espèces partagent probablement la même distribution spatiale à l’intérieur du plasma. Finalement, nous avons mesuré la température d’excitation du fer (neutre et ionisé, à l’état de trace dans nos cibles), ainsi que la température d’ionisation, à l’aide de diagrammes de Boltzmann et de Saha-Boltzmann, respectivement. À l’instar de travaux antérieurs (Barthélémy et al., 2005), il nous est apparu que les différentes températures convergeaient vers une température unique (considérant nos incertitudes) après 2-3 microsecondes. Les différentes températures mesurées de 0 à 2 microsecondes ne se recoupent pas, ce qui pourrait s’expliquer soit par un écart à l’équilibre thermodynamique local, soit en considérant un plasma inhomogène où la distribution des éléments dans la plume n’est pas similaire d’un élément à l’autre, les espèces énergétiques se retrouvant au cœur du plasma, plus chaud, alors que les espèces de moindre énergie se retrouvant principalement en périphérie. / Interest in the characterization of materials by laser induced plasma spectroscopy (LIPS) is growing with new applications emerging at an ever increasing pace. The purpose of this thesis is to verify the influence of the selection of spectral lines according to measured parameters of the plasma: electron density and excitation (neutral and singly ionized atoms) and ionization temperatures. Our measurements are conducted under typical operating conditions of LIPS: spatially integrated and temporally resolved. We used two binary aluminum targets containing trace elements (Al-Fe and Al-Mg). First, we measured the electron density using Stark broadening of lines from several species (Al II, Fe II, Mg II, Fe I, Mg I, Hα). We observed that the absolute density had a different temporal behavior depending on the species. The ionic lines giving electron densities systematically higher (up to 50 % at 200 ns after plasma ignition), and decreasing faster than densities derived from neutral lines. Densities obtained from trace elements Mg and Fe are lower than densities obtained from the commonly used line Al II at 281.618 nm. In parallel, we studied the space-integrated electron density evolution found from hydrogen Hα line and observed that it has a temporal behavior similar to the density obtained by the Al II line at 281.618 nm. Thus the two species probably share the same spatial distribution within the plasma. Finally, we measured the excitation temperature of iron (neutral and ionized, in trace amount in our targets), and the ionization temperature, using Boltzmann and Saha-Boltzmann plots, respectively. As previously described by Barthélémy et al. (2005), it appears that the different temperatures converge to a single value (considering error bars) after 2-3 microseconds. The different temperatures measured from 0 to 2 microseconds do not overlap, which could be explained by a departure from local thermodynamic equilibrium (Barthélémy et al., 2005), or by considering an inhomogeneous plasma where spatial distribution differs from one species to another, so that high energy species are found from within the plasma’s centre, which is hotter, while the lower energy species are found mainly in the periphery.

L’objectif de cette thèse était d’appliquer des méthodes d’analyse multivariées au traitement des données provenant de la spectroscopie de plasma induit par laser (LIBS) et de la spectroscopie térahertz (THz) dans le but d’accroître les performances analytiques de ces techniques.Les spectres LIBS provenaient de campagnes de mesures directes sur différents sites géologiques. Une approche univariée n’a pas été envisageable à cause d’importants effets de matrices et c’est pour cela qu’on a analysé les données provenant des spectres LIBS par réseaux de neurones artificiels (ANN). Cela a permis de quantifier plusieurs éléments mineurs et majeurs dans les échantillons de sol avec un écart relatif de prédiction inférieur à 20% par rapport aux valeurs de référence, jugé acceptable pour des analyses sur site. Dans certains cas, il a cependant été nécessaire de prendre en compte plusieurs modèles ANN, d’une part pour classer les échantillons de sol en fonction d’un seuil de concentration et de la nature de leur matrice, et d’autre part pour prédire la concentration d’un analyte. Cette approche globale a été démontrée avec succès dans le cas particulier de l’analyse du plomb pour un échantillon de sol inconnu. Enfin, le développement d’un outil de traitement par ANN a fait l’objet d’un transfert industriel.Dans un second temps, nous avons traité des spectres d’absorbance terahertz. Ce spectres provenaient de mesures d’absorbance sur des mélanges ternaires de Fructose-Lactose-acide citrique liés par du polyéthylène et préparés sous forme de pastilles. Une analyse semi-quantitative a été réalisée avec succès par analyse en composantes principales (ACP). Puis les méthodes quantitatives de régression par moindres carrés partiels (PLS) et de réseaux de neurons artificiels (ANN) ont permis de prédire les concentrations de chaque constituant de l’échantillon avec une valeur d’erreur quadratique moyenne inférieure à 0.95 %. Pour chaque méthode de traitement, le choix des données d’entrée et la validation de la méthode ont été discutés en détail. / The aim of this work was the application of multivariate methods to analyze spectral data from laser-induced breakdown spectroscopy (LIBS) and terahertz (THz) spectroscopy to improve the analytical ability of these techniques.In this work, the LIBS data were derived from on-site measurements of soil samples. The common univariate approach was not efficient enough for accurate quantitative analysis and consequently artificial neural networks (ANN) were applied. This allowed quantifying several major and minor elements into soil samples with relative error of prediction lower than 20% compared to reference values. In specific cases, a single ANN model didn’t allow to successfully achieving the quantitative analysis and it was necessary to exploit a series of ANN models, either for classification purpose against a concentration threshold or a matrix type, or for quantification. This complete approach based on a series of ANN models was efficiently applied to the quantitative analysis of unknown soil samples. Based on this work, a module of data treatment by ANN was included into the software Analibs of the IVEA company. The second part of this work was focused on the data treatment of absorbance spectra in the terahertz range. The samples were pressed pellets of mixtures of three products, namely fructose, lactose and citric acid with polyethylene as binder. A very efficient semi-quantitative analysis was conducted by using principal component analysis (PCA). Then, quantitative analyses based on partial least squares regression (PLS) and ANN allowed quantifying the concentrations of each product with a root mean square error (RMSE) lower than 0.95 %. All along this work on data processing, both the selection of input data and the evaluation of each model have been studied in details.