Les propriétés de compression de la transformation PCT permettent de réaliser et d'accélérer le calcul des méthodes multivariées même avec des matrices très larges où le nombre de variables est beaucoup plus grand que le nombre d'objets. Les résultats peuvent être retransformés vers le domaine des variables d'origine pour donner exactement les mêmes résultats qu'avec la méthode standard, sans transformation. Nous avons utilisé la PCT pour accélérer la régression PLS et dans le cadre de l'analyse du produit externe avec des très grandes matrices. La méthode ANOVA-PCA facilite la comparaison de la variance des Facteurs responsables des différences entre les échantillons avec la variance résiduelle. Elle a été utilisée pour étudier par spectroscopie d'infrarouge proche la stabilité des matériaux de référence stockés dans différentes conditions. Nous l'avons aussi modifiée de façon à permettre la prédiction de nouveaux échantillons et pour améliorer ses capacités de trouver des facteurs signifiants et pour mieux comprendre les différentes sources de variabilité résiduelle. En profitant d'une des idées de base de la méthode ANOVA-PCA, le calcul des niveaux, nous avons modifié une méthode d'analyse de tableaux multiples, la CCSWA, pour l'adapter à l'analyse des tableaux de niveaux de Facteurs provenant d'un plan d'expériences. La spectroscopie MIR a été utilisée pour l'analyse d'aliments à l'aide d'ATR chauffante. Nous avons étudié des altérations de différentes huiles alimentaires par chauffage accélérée à différentes températures avec acquisition simultanée des spectres. Le même accessoire a été utilisé dans l'analyse de vin pour étudier les effets de différents traitements technologiques, tels que la micro-oxygénation et l'ajout de copeaux de bois. Après évaporation, sur le crystal, de l'eau et de l'éthanol, les constituants majoritaires, des pics correspondants aux autres composantes du vin, normalement cachées, ont pu donc se révéler.

La thèse propose une méthodologie d’analyse rapide basée sur l’étude de l’image hyperspectrale Raman d’un produit pharmaceutique falsifié sous forme solide afin, d’une part d’identifier les substances présentes, et d’autre part estimer la teneur du principe actif dans l’échantillon sans étalonnage préalable.La présence de produits pharmaceutiques falsifiés est un véritable enjeu de santé publique. Ce type de produits de santé est facilement disponible sur internet, et beaucoup d’exemples montrent leur dangerosité. De nombreuses techniques sont disponibles pour analyser ces produits et ainsi participer à la lutte contre la falsification de médicament. La combinaison de ces techniques analytiques permet une caractérisation approfondie de l’échantillon. Cependant, peu de techniques analytiques procurent l’ensemble des informations chimiques.L’imagerie chimique Raman est une technique qui répond aux exigences requises pour l’analyse de produits falsifiés sous forme solide. En effet, cette technique peu destructive permet de réutiliser l’échantillon pour des analyses complémentaires. L’imagerie chimique Raman combine les trois disciplines de spectroscopie Raman, microscopie et chimiométrie. Cette technique réalise des mesures successives de spectres Raman sur des zones adjacentes couvrant la surface de l’échantillon. Elle intègre donc des informations spatiales et spectrales. Les méthodes chimiométriques dites de résolution (MCR-ALS et DCLS) analysent le jeu de spectres pour extraire des informations qualitatives (détection des spectres purs du mélange) et des informations quantitatives (estimation de la concentration de la substance active). La méthodologie a été optimisée et validée avec des échantillons préparés en laboratoire, puis appliquée à des échantillons réels authentiques et falsifiés. La sensibilité de la méthode qualitative a été démontrée par la détection d’un principe actif antibiotique à la teneur de 0,3% m/m dans un comprimé à visée anabolisante. De plus la méthode a été capable de détecter les substances utilisées pour le traitement de dysfonctions sexuelles (sildénafil, tadalafil, vardénafil, dapoxétine). Malgré une forte émission de fluorescence, la méthode a réussi à discriminer les 3 sels de clopidogrel (hydrogénosulfate, bésilate et chlorhydrate). L’analyse quantitative directe sur des échantillons de Viagra® et de Plavix® a été jugée convenable avec une déviation de la teneur entre -15% et +24%. Cette déviation est considérée acceptable pour évaluer le risque sanitaire pour le patient et alerter les autorités de santé.Dans le cadre de l’analyse des produits falsifiés, il a été démontré que la micro-spectroscopie Raman associée aux méthodes chimiométriques permet de réaliser un « screening spectroscopique » des composants de l’échantillon, d’identifier les substances chimiques, de visualiser leur distribution sur la surface de l’échantillon et d’estimer leur teneur par « quantification directe ». / The thesis proposes a rapid methodology of analysis based on the Raman hyperspectral image study of a solid form falsified pharmaceutical product in order to identify the substances in the sample and to estimate the content of the active ingredient in the sample without prior calibration.The presence of falsified pharmaceuticals is a real public health issue. This type of products is easily available on the internet, and many examples show their dangerousness. Many techniques are available for the analysis of these products and thus participate in the fight against drug falsification. The combination of these analytical techniques allows a comprehensive characterization of the sample. However few analytical techniques provide all the chemical information. Raman chemical imaging is a technique that meets the requirements for the analysis of falsified products in solid form. Indeed, this non-destructive technique makes it possible the reuse of the sample for additional testing. Raman chemical imaging combines the three disciplines of Raman spectroscopy, microscopy and chemometrics. This technique performs successive measurements of Raman spectra on adjacent location covering the surface of the sample. It therefore collects spatial and spectral information. The so-called resolution chemometric methods analyse the set of spectra in order to extract qualitative information (detection of pure spectra in the mixture) and quantitative information (estimate of the concentration of the chemical substance). The methodology was optimized and validated with samples prepared in the laboratory, and then applied to genuine and falsified real samples. The sensitivity of the qualitative method was demonstrated with the detection of an antibiotic active ingredient at a content of 0.3% m/m in an anabolic tablet. Moreover, the method was able to distinguish substances used for the treatment of sexual dysfunctions (sildenafil, tadalafil, vardenafil, dapoxetine). Despite a high fluorescence emission, the method successfully discriminated the 3 salts of clopidogrel (hydrogen sulfate, besylate and hydrochloride). Direct quantitative analysis of samples of Viagra® and Plavix® was found to be appropriate with a deviation between -15% and +24%. This deviation is considered acceptable to assess the health risk to the patient and to alert health authorities.For the analysis of falsified products, it has been demonstrated that Raman micro-spectroscopy combined with chemometric methods allows to perform a "spectroscopic screening" of the components in the sample, to identify chemical substances, to visualize their distribution on the sample surface and to estimate their content by "direct quantification".

The quality of food is an increasingly important matter of concern in Europe. The feeding stuffs poisoning episode that occurred in Belgium in May 1999 pointed out the vulnerability of the food chain and the lack of appropriate monitoring. Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and polychlorobiphenyls (PCBs) were the key contaminants involved. This has triggered new EU legislation, including maximum and action limits for relevant food and feed products as well as requirements for analytical method used to verify compliance. Large monitoring programs to test food and feed have been launched and, in many countries, efforts to monitor dioxins and related compounds strongly increased. To cope with the large number of samples statistically required for monitoring, the recommanded strategy involves the use of screening methods based on low resolution mass spectrometry (LRMS) and/or bio-assays, and high resolution mass spectrometry (HRMS) method, used to bear out their presence. Major analytical challenges had to be met to face with the large number of samples including the authoritys requests on developing screening and alternative methods for monitoring programs of PCDD/Fs and PCBs in food and feed. The first part of this document is devoted to the development of an alternative LRMS-based method for PCDD/Fs and PCBs measurement in food and feed. The second part of this thesis is an answer to the basic questions commonly addressed to all analytical chemists developing method but here in a particular context due to very specific family of compounds involved: How to make sure that my method is able to achieve sufficient accuracy on results? Are there any analytical benchmarks available for validation purposes? How to evaluate measurement uncertainty expressed in toxic equivalent (TEQ) units? How to report results right? How can the proficiency of my laboratory be measured? To ensure the quality of data obtained, laboratories in charge of the food control on PCDD/Fs and dioxin-like PCBs encounter a number of severe problems. One has to mention the lack of sufficient and reliable certified reference materials that are necessary to validate methods, too scarce data available on current analytical method performances, and the absence of quality criteria approach for analytical method. During the course of this work, I contributed to answer these questions to the general analytical effort by providing useful tools and methodologies. At that time, straightforward answers could not be found in the scientific literature. One of the reasons was the scarce data of dioxins in food and feed available. One can also mention the unusual part the dioxins play in chemical analysis. Indeed, the main features that characterize a dioxin measurement are the low levels at which these compounds occur in biological samples (sub parts-per-trillion), levels that are currently not explored by any other applications in chemical analysis in the food sector and therefore the difficulty to cope with precision models available; the reporting of results expressed in total tetrachloro dibenzo-p-dioxin (TCDD) toxic equivalent concentration for compliance assessment with statutory limits; and, what necessarily follows from the decision-making: the statement of the uncertainty interval also expressed in toxic units. To answer the foregoing questions in an international frame, notions such as validation of analytical procedures, fitness for purpose, internal quality control, interlaboratory studies, proficiency testing, measurement uncertainty, traceability had to be introduced. They are all encompassed in the analytical quality assurance management a laboratory should implement. These concepts are strongly connected to statistical techniques. This branch of analytical chemistry that consists in extracting relevant information from data using statistical and mathematical methods adapted to the specific needs for the chemists is called chemometrics. Quality is an essential preoccupation of chemometrics but it cannot be only limited to these aspects. Chemometrics relates also to other topics such as experiments and experimental design methodologies, (new) knowledge about chemical systems. Chemometrics and quality This thesis treats several aspects and new approaches of quality assurance for an ultra-trace contaminant laboratory: external method validation through interlaboratory studies and estimation of repeatability, reproducibility and trueness using simple statistics based on normal distributions (ISO 5725) but also more complex statistical tests for heavily tailed, skewed or even bimodal distributions; the production and the use of a reference material for internal validation and internal quality control (QC) purposes; advanced statistics in quality control chart and multi-level control charts for sensitive detection of bias; proposal of quality criteria for assessment of proficiency of dioxin laboratories; proposal of benchmark precision for internal validation purposes; estimation of measurement uncertainty. The thesis is divided in the following chapters: Chapter 1 is a general introduction for dioxins and related compounds. It consists of a brief introduction to general characterization, mechanism of toxicity, human exposure and European legislation in food and feed. Chapter 2 provides an overview of the analytical procedures for mass spectrometry based methods. It gives a brief summary of the most frequently used techniques to extract and purify PCDDs, PCDFs and dioxin-like PCBs (DL-PCBs) from food and feed matrices. Regarding detection, special attention of the principles of detection and quantification by HRMS in selected ion monitoring mode (SIM) and the quadrupole ion storage low resolution mass spectrometer in MS/MS mode is addressed. Chapter 3 discusses the development and optimization of a large volume injection (LVI)-gas chromatography (GC)-ion trap MS/MS method as an alternative to GC-HRMS for the measurement of PCDD/Fs in food and feed. Instrumental detection limits were lowered by a factor 2 to 3 with the development, in collaboration with the manufacturer, of a system of damping gas pressure inside the trap that improves precursor ions trapping efficiency. We achieved 5:1 signal to noise with the injection of 200 fg of 2,3,7,8 TCDD. With slight adjustments to sample size and final extract volume, we demonstrated on QC samples the good agreement between this method and the reference GC/HRMS method for PCDD/Fs and DL-PCBs in food and feed. In chapter 4, the first European inter-laboratory study on dioxins, furans and dioxin-like PCBs using the HRGC/HRMS method in animal feed samples is described with two main objectives. The first objective was to produce a reference material for internal validation and QC purposes. The second objective was to assess the analytical performances of the GC-HRMS method close to maximum levels as no data were available at that time and to check whether EU directives requirements were met. Chapter 5 is a general discussion on the capability of the state-of-the-art HRGC/HRMS method to provide reliable results at decreasing maximum levels. Levels have to decrease according to EU policy regarding human exposure to those contaminants. In this case, we present the issue from a different angle, i.e. the analytical point of view for the future establishment of target levels. Based on the results of PCDD/Fs and DL-PCBs interlaboratory study in animal feedingstuffs described in chapter 4, we demonstrated for the sum of the 17 PCCD/Fs toxic congeners that reliable results can be easily provided up to a value of 0.17 ng WHO-TEQ/kg. The ability to reliably quantify a minute trace of these contaminants has been pointed out with the aim of providing an analytical benchmark for the future establishment of target dioxin levels in animal feedingstuffs. Hence, both analytical and toxicological aspects should be examined together to set realistic target levels achievable for most dioxin laboratories involved in monitoring programs. One of the central themes of this thesis is the establishment of an empirical relationship between reproducibility standard deviation and the dioxin congeners level in food and feed. Chapter 6 deals with raw data from numerous performances interlaboratory studies of PCDD/Fs and DL-PCBs in food and feed. Striking linear functions in log scale between reproducibility standard deviation and congeners level over a concentration range of 10-8 to 10-14 g per g fresh weight were observed. The data fit very well to a Horwitz-type function of the form sR = 0.153c0.904, where sR and c are dimensionless mass ratios expressed in pg/g on fresh weight, regardless of the nature of the toxic congeners, food and feed matrices, or sample preparation methods. I called this relationship the dioxin function. One of the main features of the dioxin function could be its use as a suitable fitness-for-purpose criterion for dioxins and related compounds in proficiency testing (PT) exercises. We illustrated its use with practical example with data from the largest international PT in this field. Another application is its use as benchmark precision criteria for internal validation. Chapter 7 discusses the role of internal quality control (IQC) to monitor analytical processes. Introducing new QC methods derived from the industrial practice to analytical chemistry, improving data evaluation and allowing to detect shifts or trends, are elements that are difficult to point out with classical approach (Shewhart chart). The importance of ARL (average run length) as a key-criteria of the efficiency of a quality control procedure will be emphasized. The introduction of the multivariate approach of multilevel control with the Hotelling's T2-test will lead to a better detection of random errors than the independently managed conventional Shewhart charts. Moreover, the Exponentially Weighted Moving Average (EWMA) will offer a flexible tool for detecting the inacurracy of a method, especially where small shifts or bias are of interest. All these concepts, recently introduced in clinical chemistry, were applied here for the monitoring of PCDD/Fs and DL-PCBs in food and feed. Chapter 8 introduces the concept of measurement uncertainty (MU). Three top-down approaches for uncertainty estimation are proposed on the example of the GC-HRMS method for PCDD/Fs and DL-PCBs in various food and feed matrices: the approach which combines long-term precision and trueness data to obtain an estimate of MU (Barwick and Ellision method); the approach which uses the reproducibility estimate from interlaboratory-studies as uncertainty estimate; the concept of accuracy profile used in the context of validation and internal quality control to assess MU. Chapter 9 presents a general conclusion