Scattering samples such as tissue or milk are difficult to analyze using spectroscopy. Due to scattering, the path length of light through samples varies. However, diffuse scattering is repetitive, and can therefore be considered to be a self-similar process. This thesis shows that mathematical techniques which characterize self-similarity can significantly improve spectroscopic quantification in absorbing/scattering media. / Often, quantification in scattering samples relies on multivariate analysis of near-infrared (NIR) spectra. Using data from wheat samples and tissue phantoms, it is shown that Haar transform (HT) preprocessing improves quantification. Wavelet domain calibration reduces the need for trial and error preprocessing of spectra, and yields more parsimonious models and lower errors than single wavelength selection. In particular, preprocessing with Haar "son" wavelets gives simple calibrations which are useful to design future simplified data acquisition schemes. / Time-resolved intensity measurements compliment NM spectroscopy by providing information on absorption and scattering coefficients (mua and mu s). Haar wavelets are shown to be useful for constructing simple models to determine optical properties of tissue phantoms from times-of-flight (TOFs) of transmitted or diffusely reflected photons. For TOF distributions collected with a research-grade instrument, three wavelets at frequencies less than 800 MHz accurately quantify mua and mus (less than 8% and 3% error, respectively). Furthermore, when the light source is simplified to a portable nanosecond rise-time laser diode, three wavelets of frequencies less than 400 MHz yield accurate estimates from transmitted photon data. Based on the above results, simplifications to the light source and electronics of photon TOF instruments are suggested. / Fractal correlation dimension calculations further improve the analysis of photon TOF distributions. Compared to HT analysis, models based on the correlation sum use fewer variables to quantify mua and mu s, while maintaining comparable accuracy. A simple method of analysis that is forgiving to instrumental drift is obtained. / Overall, analysis of self-similarity facilitates quantification in highly scattering samples. By using wavelets or correlation dimension calculations to analyze NIR spectra or TOF distributions, simple yet accurate models are obtained to estimate optical properties. Improvements proposed in this thesis could serve to make measurements at the bedside of a patient feasible.

Focal plane array (FPA) FTIR imaging spectroscopy provides unprecedented levels of spatially resolvable chemical information for analysis of samples at the micrometer scale. This study evaluates the quantitative performance characteristics of the individual detector elements comprising the FPA camera, and applies them to making analytical measurements of a custom designed microfluidic multichannel transmission cell. Statistical descriptions are provided for the response distributions among the FPA's detector elements; RMS noise, peak response, and linear regression parameters. It was found that individual detector elements of the FPA allowed for accurate milli-absorbance measurements, however the variability was large when contrasting detector elements due to FPA detector non-uniformity issues. When applied to the microfluidic multichannel sampling system designed for the monitoring of four fluid streams, it was found that the detector elements covering the fluid stream could be averaged to generate a very repeatable response between streams – thus allowing for milli-absorbance measurements of 4 samples simultaneously with the current design. / L'imagerie par spectroscopie IRTF dans la matrice plane focale (MPF) offre des niveaux de résolution spatiale sans précédent des informations chimique dans le domaine spatial pour une analyse des échantillons à l'échelle du micromètre. L'étude actuelle examine l'ensemble des applications de la spectroscopie IRTF (MPF) avec l'utilisation d'un système micro-fluidique multicanaux de transmission de cellules conçut sur mesure comme une approche potentielle d'une analyse quantitative des échantillons liquides à haut débit. Des descriptions statistiques sont fournies selon la répartition des réponses parmi ces éléments individuels du détecteur. La réponse des éléments individuels du détecteur dans la MPF a été démontrée comme étant reproductible dans des unités de milli absorbance et ainsi, la plus importante variabilité de réponse à travers l'ensemble est due aux problèmes de non conformité associés à la MPF. La moyenne des réponses des éléments du détecteur sur lesquels les résultats de chaque canal est imagées dans de bonne reproductibilité inter-canal et ainsi compense de manière satisfaisante la non-uniformité des pixels. Les expériences qui prouvent ce concept impliquant des mesures analytiques sur quatre échantillons visualisés simultanément avec la conception actuel des cellules sont présentées.

Current diagnostic tests are unable to unequivocally diagnose many neurological diseases. Parkinson's disease (PD) is one such neurological disorder that has been linked to oxidative stress. Oxidative stress, the imbalance of oxidants and anti-oxidants, results in modifications in molecular structure, which can be detected by vibrational spectroscopy. In this thesis, Raman spectroscopy was used to measure blood samples that were separated. Measured spectra were analysed for evidence of changes associated with oxidative stress. Such changes were observed and the technique was developed into a diagnostic tool for the detection of oxidative stress in plasma from PD patients. In this work, Fenton reaction (Fe2+ and H2O2) was used to oxidize the biomolecules. Calibration measurements were performed in order to characterize Raman spectra. Characterization involved the assignment of the vibrational bands associated with oxidative stress. Changes were identified by comparing the spectra before and after the Fenton reaction. The next investigations were to determine the origin of the spectral signals. Plasma spectra were measured and separations were performed to remove the plasma components. The measured spectra were then compared to spectra taken prior to the removal of the components. By using this process, spectral contributions of the removed components were identified. Knowing the origin of the spectral signals, separations were applied to blood samples from normal and patients with PD. Raw and salting out spectra from three different groups were measured: a) a group of normal elderly control (NEC) patients, b) a calibration group of PD patients from Stage 1, and c) a test group of mixed PD patients from Stages 1-4. Data analyses included the use of stagewise multilinear regression to the measured spectra with cross-validation. This analysis generated a calibration model based on selected frequencies corresponding to oxidation and other functionalitie / Les tests diagnostiques actuels ne nous permettent pas de diagnostiquer d'une manière non équivoque plusieurs désordres neurologiques. La maladie du Parkinson's (PD) consiste en un désordre neurologique qui a été lié au stress oxydatif. Le stress oxydatif, c'est-à-dire le déséquilibre entre les oxydatifs et les anti-oxydatifs, donne lieu à des modifications de la structure moléculaire pouvant être détectées par la spectroscopie vibrationnelle. Dans ce mémoire, nous avons jumelé la spectroscopie Raman à des séparations d'échantillons de sang. Des spectres mesurés ont été analysés dans le but de trouver de l'évidence des changements associés avec le stress oxydatif. De tels changements ont été observés et la technique a été développée dans un outil diagnostique pour la détection du stress oxydatif dans le plasma des patients du PD. Dans ce travail, la réaction Fenton (Fe2+ et H2O2) a été utilisée pour oxyder les biomolécules. Des mesures de calibration ont été accomplies afin de caractériser les spectres Raman. La caractérisation a impliqué l'affectation des bandes vibrationnelles qui sont associées au stress oxydatif. Des changements ont été identifiés par la comparaison des spectres mesurés avant et après le traitement réactionnel de Fenton. Des enquêtes subséquentes ont été menées afin de déterminer l'origine des signaux spectraux. Les spectres de plasma ont été mesurés et des séparations ont été performés afin d'enlever les composantes du plasma. Les spectres mesurés ont été comparés aux spectres avant l'enlèvement des composantes. Par cette méthode, les contributions spectrales des composantes enlevées ont été identifiées. Avec la compréhension de l'origine des signaux spectraux, des séparations ont été appliquées à des échantillons de sang des personnes avec et sans PD. Nous avons mesuré des spectres crus et relargués de trois groupes : a) un group de contrôle normal des aîn

The development of intelligent components for the automated analysis of samples by inductively coupled plasma (ICP) Spectrometry is presented. An expert system for diagnosing an ICP atomic emission spectrometry (AES) system using a blank solution was developed as a warning system. This expert system was able to warn the system of major malfunctions and was able to identify most problems. Three pattern recognition techniques were compared in their ability to recognize similar geological samples in small databases. Two of these techniques, k-Nearest Neighbours and Bayesian Classification, worked extremely well with over 96% success. The development of an objective function for multi-element optimizations in ICP-AES is presented. Various aspects of the application of a Simplex optimization were explored for the optimization of the ion optics of an ICP-mass spectrometry (MS) system. An algorithm for the automatic selection of internal standards for analytes in difficult samples in ICP-MS is presented.

The effects of buffer additives in capillary electrophoresis (CE) separations of various biologically important molecules were studied. Buffers containing $ beta$-cyclodextrin ($ beta$-CD) and borate were used to separate mixtures of monophosphorylated nucleotide isomers. The electrophoretic mobility of nucleotide isomers was altered as a result of selective complex-forming reactions with $ beta$-CD. In addition, complexation of 5$ sp prime$-nucleotides with borate increased the migration time window and led to better separation. / Mono- and di-phosphorylated isomers of the insulin receptor peptide (IRP), were resolved in buffers containing sodium dodecyl sulfate, (SDS). The effects of SDS concentration and pH on electrophoretic velocity and resolution were investigated. The use of polyacrylamide coated capillaries resulted in increased resolution and decreased analysis time. / The effects of various detergents on the electrophoretic behaviour of plasma apolipoproteins in uncoated and polyacrylamide coated capillaries were studied. Interaction between apolipoproteins and detergents was strongly influenced by the nature of the polar head group of the detergent. / Capillaries coated with various crosslinked hydrophillic polymers were prepared. Coatings based on poly(2-hydroxyethylmethacrylate) crosslinked with dimethylol ethylene urea exhibited the best performance for CE separations of basic proteins.

The design and implementation of a framework for the automated analysis of samples by inductively coupled plasma atomic emission spectroscopy (ICP-AES) is presented. Various components of this framework have been explored and validated. After initial work was undertaken to determine a suitable pattern recognition technique for sample classification, approaches for detecting the presence of a problematic matrix, and for selecting analysis conditions and calibration methodologies were developed. For the small universe of 14 elements used, the system performed well when matrix effects were due to concomitants in the examined universe. The application of generalized regression neural networks (GRNN) to compensate for spectral interferences and matrix effects and for correcting long-term drift was also examined. In post-processing the analysis results, the size of the GRNN training sets were found to be too small to permit accurate correction of matrix effects though accurate drift correction was possible.

A flow injection (FI) preconcentration system has been coupled with a direct sample insertion (DSI) system for inductively coupled plasma atomic emission spectrometry (ICP-AES). The FI system developed uses an ion exchange microcolumn to provide preconcentration factors of approximately for several metals. Improved DSI detection limits are obtained by employing a graphite sample probe of reduced mass. The improvements are element specific and range between 10 and 130. The combination of the two optimized systems results in overall detection limit improvements ranging between 140 and 1200 for the elements tested, Cu, Pb and Zn. Precision of the FI-DSI-ICP system averages 4% rsd for these three elements. The flow injection system is fully automated and under computer control, resulting in a reproducible FI processing time of 6.0 minutes, using 5 mL injection volumes. The system offers considerably potential for further gains in performance, by increasing the injection volumes used and optimizing the interface.

The dehydrogenative coupling reactions of phenylsilane (PhSiH$ sb3$) with the following bridged and unbridged bis(indenyl) complexes: ($ eta sp5$-C$ sb9$H$ sb7) sb2$MMe$ sb2$ (1), ($ rm eta sp5$-$ rm C sb9H sb{11}) sb2MMe sb2$ (2), $ rm (CH sb2) sb2( eta sp5$-$ rm C sb9H sb6) sb2MMe sb2$ (3) and $ rm (CH sb2) sb2( eta sp5$-$ rm C sb9H sb{10}) sb2MMe sb2$ (4) (where M = Ti (a) and Zr (b); and Me = CH$ sb3$), were investigated. Some of these reactions were monitored by $ sp1$H- and $ sp{29}$Si-NMR, and by gel permeation chromatography (GPC). The oligosilanes obtained generally showed a polymodal molecular weight distribution. During the reaction of PhSiH$ sb3$ with 1a, a new dititanium complex was isolated: $( mu$-$ rm eta sp6{:} eta sp5$-$ rm C sb9H sb7)( mu$-$ rm H) sb2 lbrack ( eta sp5$-$ rm C sb9H sb7)Ti rbrack sb2$. The structure of this compound was confirmed by X-ray diffraction analysis, by infrared spectroscopy, and by electron paramagnetic resonance (EPR). Some possible mechanisms for the formation of this complex, along with the origin of its short Ti-Ti distance (2.745 A), were discussed. For the other systems, it was shown that the catalytic activity was strongly dependent on the nature of the metal, and on the structure of the ligands surrounding it. The titanium-based catalysts seemed to be more efficient to produce linear oligosilanes having degrees of polymerization (DP) of 12 or 18, depending on the systems. Cyclic oligomers, having DP's of 6-7, were more specifically obtained with the zirconium-based catalysts. The only system having shown a certain kind of stereocontrol was the one based on 3b.

An infrared detector is constructed for use in High Pressure Liquid Chromatography. The detector uses a helium-neon laser which emits at 3.39 microns in the infrared as the source and a commercial lead selenide semiconductor as the detector. The detector when used in the flow-injection analysis mode can detect all classes of lipid: fatty acids, triglycerides, phospholipids, cholesterols and cholestryl esters. Chromatography using a normal phase column with separation into the lipid classes of cholesterol, cholestryl esters and triglycerides is performed. A purchased serum sample has been prepared and separation and quantitation of the 3 lipid classes is performed.

This thesis presents a quantitative protocol of proton nuclear magnetic resonance (1H NMR) that allows the determination of human amniotic fluid metabolite concentrations, which are then used in a metabonomic study to establish patient health during gestation. 1H NMR free inductive decays (FIDs) of 258 human amniotic fluid samples from a 500MHz spectrometer are acquired. Quantitative analyses methods in both the frequency- and time-domain are carried out and compared. Frequency-domain analysis is accomplished by integration of the metabolite peaks before and after the inclusion of a known standard addition of alanine. Time-domain analysis is accomplished by the direct exponential curve resolution algorithm (DECRA). Both techniques are assessed by applications to calibration biological solutions and a simulated data set. The DECRA method proves to be a more accurate and precise route for quantitative analysis, and is included in the developed protocol. Well-defined peaks of various components are visible in the frequency-domain 1H NMR spectra, including lactate, alanine, acetate, citrate, choline, glycine, and glucose. All are quantified with the proposed protocol. Statistical t-test and notched box and whisker plots are used to compare means of metabolite concentrations for diabetic and normal patients. Glucose, glycine, and choline are all found to correlate with gestational diabetes mellitus early in gestation. With further development, time-domain quantitative 1H NMR has potential to become a robust diagnostic tool for gestational health.