Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-01-04T14:27:39Z No. of bitstreams: 1 thiagodeoliveiramendes.pdf: 10542310 bytes, checksum: 9b116b21c7d753e1a26ffeaeab698db5 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-01-25T15:58:12Z (GMT) No. of bitstreams: 1 thiagodeoliveiramendes.pdf: 10542310 bytes, checksum: 9b116b21c7d753e1a26ffeaeab698db5 (MD5) / Made available in DSpace on 2016-01-25T15:58:12Z (GMT). No. of bitstreams: 1 thiagodeoliveiramendes.pdf: 10542310 bytes, checksum: 9b116b21c7d753e1a26ffeaeab698db5 (MD5) Previous issue date: 2015-08-04 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / O estudo de amostras complexas, como alimentos, fármacos e fluidos biológicos necessita de métodos robustos de análise de dados, uma vez que estas amostras podem conter informações de dezenas a milhares de constituintes. A pesquisa desenvolvida neste trabalho tem como temática principal a aplicação de métodos multivariados de análise para extrair informações a partir dos espectros vibracionais (infravermelho e Raman) ou de massas, de amostras complexas. Técnicas de separação tais como CE, GC e LC são utilizadas como abordagem auxiliar para fornecer informações aos modelos multivariados propostos. Cinco aplicações são discutidas no texto. A primeira faz o uso da PCA sobre cromatogramas de GC de diferentes amostras de óleo de soja e azeite de oliva, buscando realizar uma análise exploratória das duas classes de óleos vegetais. Na sequência são analisadas misturas com adições controladas de óleo de soja em azeite de oliva pelas técnicas de absorção infravermelha e espalhamento Raman, com o objetivo de propor uma regressão PLS que determine o percentual de adição. A segunda aplicação aborda a busca de marcadores vibracionais, infravermelho e Raman, que possam ser utilizados para a quantificação do percentual de gordura em amostras de leite fluido. Neste estudo foi realizada uma análise do perfil de ácidos graxos presentes no leite pela técnica de GC, com o objetivo de selecionar uma molécula representativa de triacilglicerol que foi utilizada para simular espectros vibracionais por modelos teóricos. A partir dos espectros obtidos por simulação computacional foi proposta a atribuição de modos vibracionais presentes na gordura do leite. Foi também proposto um modelo MLR preditor do percentual de gordura em amostras de leite fluido, cujas variáveis vibracionais foram previamente selecionadas a partir de um modelo PLS. Outra aplicação foi a utilização de modelo PLS sobre espectros Raman para a determinação simultânea das concentrações de rifampicina, isoniazida, etambutol e pirazinamida em amostras de comprimidos utilizados no tratamento antituberculose. A quarta aplicação está focada na quantificação do percentual de soro de leite adicionado fraudulentamente em amostras de leite por análise do perfil de ácidos graxos por GC e CE. Uma análise discriminante foi utilizada para selecionar marcadores de ácidos graxos que foram monitorados por regressão MLR. Por fim, é apresentada a elaboração de um tutorial completo de análise univariada e multivariada de dados, desenvolvido em R-software, para a determinação de biomarcadores em experimentos de metabolômica baseados em análises de espectros de massas. Todas estas aplicações têm como interseção o uso de métodos multivariados de análise de dados como ferramenta principal para propor novos marcadores vibracionais/metabolitos, assim como, métodos alternativos para a quantificação de diferentes analitos. / The study of complex samples such as food, drugs, and biological fluids require a robust data analysis methods, since these samples may contain information tens of thousands of constituents. The research developed in this work has as main theme the application of multivariate analysis methods to extract information from the vibrational spectra (infrared and Raman) or mass spectra of complex samples. Separation techniques such as CE, GC and LC are used as an auxiliary approach to providing information to the proposed multivariate models. Five applications are discussed in the text. The first makes use of PCA on GC chromatograms of samples of soybean oil and olive oil, searching for an exploratory analysis of the two classes of vegetable oils. Following are analyzed with mixtures of controlled additions of soybean oil in olive oil, by infrared absorption and Raman scattering techniques, with the aim of proposing a PLS regression to determine the percentage of addition. A second application addresses the search vibrational markers, infrared and Raman, that may be used to quantify the percentage of fat in milk fluid samples. In this study we carried out a profile analysis of fatty acids present in milk by CG technique, for the purpose of selecting a representative molecule of triacylglycerol was used to simulate vibrational spectra by theoretical models. From the spectra obtained by computer simulation was proposed assignment of vibrational modes present in milk fat. It was also proposed an MLR predictor model of the percentage of fat in fluid samples, whose vibrational variables were previously selected from a PLS model. Another application has been the use of PLS model of Raman spectra for the simultaneous determination of concentrations of rifampicin, isoniazid, ethambutol and pyrazinamide in tablet samples used in anti-tuberculosis treatment. The fourth application is focused on quantifying the percentage of whey added fraudulently in milk samples for analysis of the fatty acid profile by GC and CE. A discriminant analysis was used to select fatty acids markers that were monitored by MLR regression. Finally, it shows the preparation of a complete tutorial of univariate and multivariate data analysis, developed in R-software, for the determination of biomarkers in metabolomics experiments based on mass spectra analysis. All these applications have the intersection as the use of multivariate methods of analyzing data as the main tool to propose new vibrational / metabolite markers as well as alternative methods for quantification of various analytes.

Condensed phase membrane introduction mass spectrometry (CP-MIMS) is a direct, in situ analysis technique that is well suited to persistent organic pollutants, pesticides, and other small molecules. In CP-MIMS, neutral analytes permeate a hollow fibre membrane, typically polydimethylsiloxane (PDMS), driven by a concentration gradient. Analytes are subsequently dissolved by a liquid (condensed) solvent acceptor phase that is continuously flowed through the membrane lumen, which finally entrains the analytes to a mass spectrometer for detection. The membrane rejects charged and particulate matrix components, therefore eliminating sample cleanup that is otherwise necessary for conventional (i.e., chromatographic) techniques. However, larger analytes may suffer from relatively lengthy response times and lower sensitivity. A heptane cosolvent was therefore doped into the PDMS membrane, resulting in a polymer inclusion membrane (PIM). Through a system coupling CP-MIMS to electrospray ionization (ESI), the use of a PIM for model compounds resulted in faster response (~3×) and improved sensitivity (~3.5×, parts per trillion level detection limits). While effective for the demonstration of the PIM, pairing ESI with CP-MIMS represents an inherent incongruity: ESI is effective for polar, hydrophilic analytes, whereas CP-MIMS (i.e., PDMS membranes) is effective for hydrophobic analytes. CP-MIMS was therefore coupled with liquid electron ionization (LEI) as a more suitable ionization strategy. In LEI, the post-membrane solvent flow is entrained at nanolitre per minute flowrates to a LEI source, where the liquid is sequentially nebulized, vaporized, and ionized. The CP-MIMS-LEI coupling was optimized for the measurements of polycyclic aromatic hydrocarbon (PAH) isomer classes from aqueous samples, demonstrating low ng/L detection limits and response times (≤1.6 min). CP-MIMS-LEI was also applied to PAH isomer classes from soil samples, demonstrating rapid sample throughput (15 samples/hr) and low μg/kg detection limits, and additionally was quantitatively comparable to conventional techniques. A similar CP-MIMS-LEI system was applied to online monitoring of catalytic oxidation and alkylation reactions, demonstrating quantitative, real-time results for harsh, complex organic reaction mixtures. A significant analytical improvement was conducted by intentionally exploiting the already present liquid acceptor phase as an in situ means of providing liquid chemical ionization (CI) reagents for improved analyte sensitivity and selectivity (i.e., CP-MIMS-LEI/CI). Acetonitrile and diethyl ether were used as a combination acceptor phase/CI reagent system (i.e., proton transfer reagents) for the direct analysis of bis(2-ethylhexyl)phthalate from house dust (6 mg/kg detection limit). CP-MIMS-LEI/CI was then applied to PAHs from soils. Using methanol and dichloromethane combination acceptor phase/CI reagents, CP-MIMS-LEI/CI was shown to quantify and resolve PAH isomers from direct soil analyses via diagnostic PAH adduct ions: [M+CH2Cl+CH3OH-HCl]+ or [M+CHCl2-HCl]+. Using these selective ions, CP-MIMS-LEI/CI was again shown to be rapid (15 soils/hr), sensitive (ng/g detection limits) and quantitatively comparable to gas chromatography-MS for PAH measurements (average percent difference of -9% across 9 PAHs in 8 soil samples). The results across this thesis present a compelling argument for direct, quantitative screening from complex samples using CP-MIMS-LEI/CI, particularly given the simple workflow and short analytical duty cycle. / Graduate