Thesis (Ph. D.)--University of Washington, 1997. / Vita. Includes bibliographical references (leaves [137]-142).

In the last two decades chemometrics has become an essential tool for the experimental biologist and chemist. The level of contribution varies strongly depending on the type of research performed. Therefore, chemometrics may be used to interpret and explain results, to compare experimental data with real-word ‘unseen’ data, to accurately detect certain chemical vapour, to identify cancerous related metabolites, to identify and rank potentially relevant/important variables or simply just for a pictorial interpretation and understanding of the results. Whilst many chemometrics methods are well-established in the area of chemistry and metabolomics many scientists are still using them with what is often referred to as a ‘black-box’ approach, that is without prior knowledge of the methods and well-recognised statistical properties. This lack of knowledge is thanks to the wide availability of powerful computers and – perhaps more notably – up-to-date, easy to use and reliable software. The main aim of this study is to reduce this gap by providing extensive demonstration of several approaches applied at different stages of the data analysis pipeline highlighting the importance of appropriate method selection. The comparisons are based both on chemical and biochemical (metabolomics) data and construct a firm basis for the researchers in terms of understanding of chemometric methods and the influence of parameter selection. Consequently, in this thesis the exploration and comparison of different approaches employed for various statistical steps are investigated. These include pre-treatment steps such as dealing with missing data and scaling. First, different substitution of missing values and their influence on unsupervised and supervised learning have been compared, where it has been shown that metabolites that display skewness in distribution can have a significant impact on the replacement approach. The scaling approaches were compared in terms of effect on classification accuracy for variety of metabolomics data sets. It was shown that the most standard option which is autoscaling is not always the best. In the next step a comparison of various variable selection methods which are commonly used for the analysis of chemical data has been carried out. The results revealed that random forests, with its variable selection techniques, and support vector machines, combined with recursive feature elimination as a variable selection method, displayed the best results in comparison to other approaches. Moreover, in this study a double cross-validation procedure was applied to minimize the consequence of over-fitting. Finally, seven different algorithms and two model validation procedures based on either 10-fold cross-validation or bootstrapping were investigated in order to allow direct comparison between different classification approaches.

The present study was undertaken with the view to evaluate methodologies based on traditional Scheffé experimental designs that study mixtures as a tool for discovery research particularly when seeking new and or improved uses of existing mixtures. For the purpose of this study, the topic of controlling the problematic weed known as Yellow Nutsedge (Cyperus esculentus L. var. esculentus) or “Geel Uintjie”, was selected on a rather ad hoc basis. Yellow Nutsedge is a troublesome perennial weed found in most agricultural countries in the world. Herbicidal control is often difficult because of the weeds’ ability to propagate via tubers, which can remain dormant for a number of years and are also resistant to most synthetic herbicide controls. As a first step the study involved the selection of a group of chemical compounds that would be used in suppressing the germination of Yellow Nutsedge tubers. Treatment with various combinations of these chemical compounds as determined by statistical experimental designs was carried out. A review of the literature, particularly literature concerned with the study of the phenomenon of allelopathy, suggested that various phenolic-D-glucopyranosides could show promise in the suppressing the germination of Yellow Nutsedge tubers. This led to the selection of this group of compounds as the target group of “active” substances for the study. Since the group of phenolic-D-glucopyranosides is quite large, and in order to keep the study to a reasonable size, only four phenolic-D-glucopyranosides were selected namely: 4-nitrophenyl-D-glucopyranoside, 4-chlorophenyl--Dglucopyranoside, arbutin and salicin. This selection was based firstly based on a particular phenolic-D-glucopyranoside being a suspected allelochemical, and secondly the ease of technical synthesis using a catalytic process. In addition to the four selected phenolic-D-glucopyranosides, their aglycones (4,nitrophenol, 4,chlorophenol, hydroquinone and salicyl alcohol) were also included as potential “active” substances in order to discern any potential activity between the phenolic-D-glucopyranosides and the aglycones. iii The selected “active substances” were combined in various combinations according to various mixture designs in such a manner that the sum of the proportions of the various actives in any one mixture was always equal to 1. The mixtures of actives were then used in various germination experiments and three experimental responses were measured namely the germination, average dry mass and length of longest shoot. From the results of these germination studies the canonical form of the polynomial equation describing the variation in each of the three germination responses was calculated and evaluated statistically. These equations were then used to estimate the presence of, and the magnitude of synergism between the various active substances. The results from these screening experiments and their detailed statistical analysis indicated that the response surface model for the germination response contains three synergistic blends (4-nitrophenyl--D-glucopyranoside + arbutin; 4-nitrophenyl--Dglucopyranoside + hydroquinone; and 4-chlorophenyl--D-glucopyranoside + salicin) and one antagonistic blend (4-nitrophenyl--D-glucopyranoside + 4- chlorophenol--D-glucopyranoside). The response surface model for the average dry mass response contains two synergistic blends (4-nitrophenyl--Dglucopyranoside + hydroquinone; and 4-chlorophenol--D-glucopyranoside + salicin) and the same antagonistic blend as for germination response. For both germination and average dry mass responses, the most synergistic blend was found to be the combination of 4-chlorophenyl--D-glucopyranoside and salicin. Two additional tests were conducted and both confirmed the results obtained from the screening designs. These tests involved the identification of the two components responsible for the synergistic activity that resulted in the suppression of the germination of the tubers and growth of the seedlings. The experimental response measuring the longest shoot proved to be erroneous and was excluded from the statistical analysis. In summary, this study has clearly shown that statistically designed experiments based on mixture designs can be used as a powerful tool in identifying and quantifying synergistic (or antagonistic) effects of chemicals on the germination ability of plant seeds.

Indiana University-Purdue University Indianapolis (IUPUI) / Trace evidence is a major field within forensic science. Association of trace evidence samples can be problematic due to sample heterogeneity and a lack of quantitative criteria for comparing spectra or chromatograms. The aim of this study is to evaluate different types of instrumentation for their ability to discriminate among samples of various types of trace evidence. Chemometric analysis, including techniques such as Agglomerative Hierarchical Clustering, Principal Components Analysis, and Discriminant Analysis, was employed to evaluate instrumental data. First, automotive clear coats were analyzed by using microspectrophotometry to collect UV absorption data. In total, 71 samples were analyzed with classification accuracy of 91.61%. An external validation was performed, resulting in a prediction accuracy of 81.11%. Next, fiber dyes were analyzed using UV-Visible microspectrophotometry. While several physical characteristics of cotton fiber can be identified and compared, fiber color is considered to be an excellent source of variation, and thus was examined in this study. Twelve dyes were employed, some being visually indistinguishable. Several different analyses and comparisons were done, including an inter-laboratory comparison and external validations. Lastly, common plastic samples and other polymers were analyzed using pyrolysis-gas chromatography/mass spectrometry, and their pyrolysis products were then analyzed using multivariate statistics. The classification accuracy varied dependent upon the number of classes chosen, but the plastics were grouped based on composition. The polymers were used as an external validation and misclassifications occurred with chlorinated samples all being placed into the category containing PVC.

No description available.

Thesis (Ph. D.)--Virginia Commonwealth University, 2006. / Prepared for: Dept. of Chemistry. Bibliography: leaves 139- 154. Available online via the Internet.

Thesis (Ph. D.)--University of Washington, 2007. / Vita. Includes bibliographical references (p. 179-201).

Made available in DSpace on 2014-06-11T19:29:08Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-12-20Bitstream added on 2014-06-13T18:38:51Z : No. of bitstreams: 1 hatanaka_rr_me_araiq.pdf: 624224 bytes, checksum: 3308488ed0fb0545a198f0b29c9cbd3e (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Os óleos lubrificantes exercem papel fundamental no funcionamento dos motores automotivos, pois, além de reduzir o atrito entre as peças, ainda as protegem contra a corrosão. Entretanto, o desempenho do óleo lubrificante pode ser afetado por contaminantes. Embora existam várias normas e trabalhos relacionados a métodos para quantificação de contaminantes em óleo lubrificante, como por exemplo gasolina e óleo diesel, até o alcance de nossos conhecimentos, não há método descrito para quantificação de etanol em óleo lubrificante usado de motores ciclo Otto. Nesse sentido, o presente trabalho teve como objetivo o desenvolvimento e a validação de um método aplicável em rotina para quantificação de etanol em tal óleo lubrificante usado. Para tanto, foram avaliadas as técnicas combinadas: Headspace - Gas Chromatography - Flame Ionization Detector (HS-GC-FID) e Fourier Transform Infra Red Spectroscopy - Total Attenuated Reflectance - Partial Least Squares (FTIR-ATR-PLS), com e sem tratamento prévio das amostras, sendo que os melhores resultados foram obtidos pela técnica FTIR-ATR-PLS, com pré-tratamento por Extração Líquido/líquido (ELL), otimizada por planejamento fatorial. Na sequencia, foram avaliados dois cristais com diferentes ângulos de incidência para obtenção dos espectros por ATR, sendo que o cristal de ZnSe com θ = 45 º mostrou-se mais o mais adequado. Então, a partir dessas melhores condições experimentais, o método foi validado através das figuras de mérito, as quais apresentaram os seguintes resultados: LD (0,049 %), LQ (0,16 %), exatidão (RMSEP = 0,089 % (m/m) de etanol), repetibilidade (0,05 % (m/m) etanol), ajuste (R2 = 0,9997), seletividade média (0,047), sensibilidade (0,011), inverso da sensibilidade analítica (0,016 % (m/m)-1 de etanol), razão sinal ruído (máximo: 812,4 e mínimo: 200,9) e BIAS. Os resultados obtidos mostram... / The lubricating oils have a crucial role in the operation of automotive engines, not only reduce friction between moving parts, but also protect against corrosion. However, the performance of lubricating oil may be affected by contaminants. Although there are many standards methods and studies related to methods for quantification of contaminants in lubricating oil such as gasoline and diesel oil, to the best our knowledge, there are no described methods for quantification of ethanol in Otto cycle engine used lubricating oil. In that sense, this work aimed the development and validation of an applicable in routine and method to quantify ethanol content in this used lubricating oil. For that were evaluated the combined techniques: Headspace/Gas Chromatography/Flame Ionization Detector (HS-GC-FID) and Fourier Transform Infrared Spectroscopy/Attenuated Total Reflectance/Partial Least Squares (PLS-ATR-FTIR) with and without pretreatment of the samples, and the best results were obtained to the FTIR-ATR-PLS with treatment by ELL (which was optimized by factorial design). Then, two crystals of ATR with differents incident angles were studied to obtain the spectra and the ZnSe crystal with θ = 45 ° was more appropriate for analysis. Then, from the best obtained conditions, the method was validated through the figures of merit, which presented the following results: LD (0.049 %), LQ (0.16 %), accuracy (RMSEP = 0.089 % (w/w) ethanol), repeatability (0,05 % (w/w) ethanol), fit (R2 = 0.9997), mean selectivity (0.047), sensitivity (0.011), inverse analytical sensitivity (0.016% (m / m)-1 of ethanol), signal-to-noise ratio (max: 812.4 and min: 200.9) and BIAS. The results show that the method developed and validated can be implemented, in routine, for quality control laboratories of lubricating oils

Orientador: Ronei Jesus Poppi / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Química / Made available in DSpace on 2018-08-17T03:08:03Z (GMT). No. of bitstreams: 1 Rocha_WericksonFortunatodeCarvalho_D.pdf: 5817007 bytes, checksum: 4581b99bb9fbcd5d186656109eb68573 (MD5) Previous issue date: 2010 / Resumo: O estudo do fenômeno de polimorfismo em fármacos é muito importante na indústria farmacêutica. Dessa forma, tem sido extensivamente investigado, devido ao seu impacto sobre a qualidade e eficácia das formulações farmacêuticas. Neste trabalho, três novas metodologias para o controle de qualidade de polimorfismo em fármacos foram propostas. O desenvolvimento, desses métodos, visa tomar as análises de controle de qualidade na indústria farmacêutica mais prática, rápida e sem a necessidade de consumo de reagentes com alto grau de toxicidade. Na primeira metodologia foi proposta a caracterização da composição polimórfica do fármaco Piroxicam em formulações farmacêuticas utilizando cartas de controle multivariadas baseada no cálculo do sinal analítico líquido e da espectroscopia na região do infravermelho próximo. Nesse caso, foi possível a identificação da presença das diferentes formas polimórficas nas formulações farmacêuticas. Em uma segunda metodologia, no qual também foi utilizado o fármaco Piroxicam, procurou-se realizar a quantificação e a distribuição de duas formas polimórficas através da espectroscopia de imagem na região do infravermelho próximo aliado ao método quimiométrico PLS. Os resultados mostraram que para ambos os modelos desenvolvidos os valores de RMSEP (%(m/m)) estão abaixo de 4%. Também foi possível obter a informação da distribuição espacial dos diferentes polimorfos nas formulações farmacêuticas. Na última metodologia utilizou-se da espectroscopia Raman e das cartas de controle multivariadas para identificação no fármaco Carbamazepina de adulteração em excipiente, transformação de formas polimórficas por humidade e detecção de diferentes formas polimórficas. Foi possível identificar a presença e a transformação dos diferentes polimórficos presentes, além da sua correta identificação / Abstract: The study of the polymorphism phenomenon in drugs is very important in the pharmaceutical industry. Thus, it has been extensively investigated because of its impact on the pharmaceutical formulations quality and effectiveness. In this work, three new methods for quality control of polymorphism in drugs were proposed. The development of these methods, aims to make the analysis of quality control in the pharmaceutical industry more practical, faster and without the need of reagent consumptions with a high degree of toxicity. In the first method it was proposed the characterization of Piroxican polymorphic composition in pharmaceutical formulations using multivariate control charts based on the calculation of the net analytical signal and the near infrared spectroscopy. In that case, it was possible to identify the presence of different polymorphic forms of the drug in the pharmaceutical formulations. In a second method, in which it was also used Piroxicam, there was an attempt to carry out the quantification and distribution of two polymorphic forms through the near-infrared imaging spectroscopy technique and PLS chemometric method. The results showed that for both models developed the RMSEP values (% (m / m)) are below 4%. It was also possible to distinguish the information from local and global distributions of different polymorphics in pharmaceutical formulations. In the last method it was used Raman spectroscopy and multivariate control charts to identify adulteration in excipients, polymorphic forms change by humidity and the detection of different polymorphic forms in the Carbamazepine drug / Doutorado / Quimica Analitica / Doutor em Ciências

Currently, two substantial concepts are gaining importance in the food field of knowledge, namely, authenticity and traceability. Food adulteration is becoming a central issue for many players in the food industry, such as stakeholders, institutions, regulatory bodies, and consumers, due to an increased availability and easy access to a range of different food products, which it is focusing the attention of consumers demand to high-quality and authentic products, with special attention to the place of origin. Highly developed targeted analytical methods and regulated tools are used to assess the authenticity of food products; however, they possess undeniable disadvantages. These methods investigate just one characteristic decided a priori and neglect the product in its entirety, which make them ineffective strategies due to increasingly sophisticated adulterations. The foods considered for this study were selected based on the most common sources of food fraud of the so called “Made in Italy” products. One of the most counterfeited Italian food commodities is extra-virgin olive oil (EVOO). An in-depth inspection of the literature about the assessment of its quality and authenticity has revealed that it is a well and deeply studied issue. Many works with different approaches are already published, considering targeted and untargeted approaches, and chromatography and spectroscopic techniques. Nuclear magnetic resonance (NMR) is also highly used in this field, due to its high level of reproducibility and simple sample preparation. For these reasons, this topic was covered with a thorough scientific review. Grana Padano cheese is always among the leading positions in adulteration rankings, and it was the second investigated food in this study. Grana Padano cheese was firstly discriminated from its competitors and other non-PDO (non-Protected Designation of Origin) labeled cheeses. The counterfeiting issue is especially acute when the cheese is in the shredded form because no trademark logo fire-marked on the crust or other authentication signs can be recognized. Samples of Grana Padano cheese, its Italian competitors and foreign non-PDO samples were analyzed with NMR spectroscopy and multivariate analysis to obtain a statistical model able to discriminate Grana Padano cheese from all the other cheeses. Then, another important aspect about cheese, ripening, was considered. Grana Padano cheese is in fact sold with three different aging steps, that also affect its price. Mislabeling about ripening can possibly occur, and for this reason we have decided to proceed with the ripening investigations, to find possible “biomarkers” of the different aging stages. To accomplish this task, another dataset of authentic PDO Grana Padano samples with different ripening ages was analyzed. Both the aqueous and lipid fraction extracted from the samples were considered in order to obtain a holistic view of the samples during maturation. In the end, bactofugation, an additional production step has been considered. Since the production of Grana Padano cheese is strictly controlled by specification rules, the modification of such rules needs further studies to avoid modifications of the organoleptic and nutritional characteristics of the product. For this reason, the additional centrifugation step to remove microorganisms and spores was studied by the comparison of the profile of “traditional” Grana Padano samples with samples produced with bactofugation step and the results were challenging. The resulting differences were mainly affecting the aqueous fraction, which is the major responsible for the organoleptic properties. These results can be helpful for the Consortium for the discussion whether to include this step in the disciplinary or not. A preliminary study on another food commodity, blueberries, was completed, which investigated the differences among four different varieties and with varying storage times. The aim of this investigation was the application of NMR-based metabolomics to highlight differences among varieties with a fast and simple analysis and try to identify compounds potentially sensitive to storage time, as a tool for breeders to understand possible degradation pathways.