The port of Kaohsiung which was the third largest harbor in the world previously is still the biggest one in Taiwan at present. Because of shipping coming from all over the world persistently and there are several shipyards and fishing ports in harbor, the harbor area is seriously polluted by organotin compound. Nevertheless, it is not understood whether the pollution of organotin affect the marine ecosystem around coastal areas out of harbor. The first goal of this study concerns about organotin contamination and distribution of the sediments and major benthic organisms around coastal areas out of harbor. The second one is to find the adequate indicator organism which fits for organotin contamination. Five major benthic organisms including Distorsio reticularis, Niotha conoidalis, Metapenaeopsis palmensis, Portunus argentatus and Portunus hastatoides were collected and investigated in this study. The organotin concentration of sediments in all stations was not detected totally. The organotin accumulations in Metapenaeopsis palmensis, Portunus argentatus and Portunus hastatoide are not obvious. In Portunus argentatus and Portunus hastatoide, both the contents of tributyltin (TBT) were not detected. In Metapenaeopsis palmensis, the concentration of TBT is between ND¡ã3.4 ng/g as tin. In Distorsio reticularis and Niotha conoidalis, organotin is measured in all station; monobutyltin (MBT) is the main composition of butyltins; TBT is less than 10 ng/g as tin. The concentration of butyltins in Niotha conoidalis decreased with the increased wet weight. There was no significant variation for organotin concentrations in Niotha conoidalis collected from each stations and seasons. According to this study, the content of organotins in sediment around coastal areas out of harbor is obviously less than other coastal areas of large-scale harbors (Lisbon, Sines, Sagres, Huelva, Cadiz and Trafalgar). The organotin contamination around coastal areas out of harbor is not serious. Base on amounts, distributions and contents of organotin in organisms, Distorsio reticularis and Niotha conoidalis are better bethic indicator organisms than Metapenaeopsis palmensis, Portunus argentatus and Portunus hastatoides. Niotha conoidalis distributes over the depth less than 30 m, and therefore is a good indicator organism for this area. The facts that Distorsio reticularis distributes over the depth more than 30 m, suggest that it is a proper indicator organism for this area although lesser in population.

Cette thèse porte sur le comportement d’un composé prioritaire au titre de la Directive Cadre sur l’Eau au niveau européen : le tributylétain (TBT) et ses métabolites (dibutylétain (DBT) et monobutylétain (MBT)). On s’intéresse plus particulièrement aux eaux usées domestiques lors de leur transit dans deux stations d’épuration par lagunage naturel (STEP), traitement adapté aux effluents provenant des zones faiblement urbanisées. Ce travail de thèse vient combler le manque de résultats scientifiques sur (1) le niveau des concentrations en butylétains dans les effluents domestiques des zones peu urbanisées en entrée des stations par lagunage et en sortie après traitement ; (2) l’efficacité du traitement par lagunage à éliminer les butylétains ; (3)l’identification des processus qui contrôlent le devenir de ces éléments au cours du traitement.Les butylétains sont mesurés dans les matrices particulaires, dissoutes et dans les boues, par couplage GC-ICP-MS, dans toutes les étapes du traitement par lagunage de deux communes rurales du département de l’Hérault (34,France) : Gigean (6000 eq-hab.) et Montbazin (4500 eq-hab.). On montre que les trois espèces butylées sont systématiquement présentes dans tous les échantillons prélevés sur les deux STEP. Les butylétains arrivent principalement sous forme particulaire dans les eaux brutes en entrée des stations. Au cours du traitement, les processus de sorption sur les particules en suspension et la décantation de ces dernières au fond de chaque bassin sont les principaux facteurs permettant l’élimination des butylétains. Les premiers bassins anaérobies, de grande profondeur et de long temps de séjour, permettent l’élimination de plus de 60% à Gigean (93% à Montbazin) des butylétains. Les boues de fond de chaque bassin montrent des concentrations élevées en butylétains. La remise en suspension des butylétains est notée dans les derniers bassins de finition, peu profonds, ce qui peut être due à la re-stratification thermique. En sortie de station, malgré les abattements importants, des concentrations non négligeables en butylétains (12 à 228 et 1.8 à 15 ng(Sn).L-1 respectivement pour Montbazin et Gigean) sont rejetées dans le milieu récepteur. Des expériences de sorption basées sur la mise en contact de butylétains avec des boues de fond de bassin diluées ont été effectuées selon plusieurs approches expérimentales. On a ainsi montré que (1) la distribution des espèces MBT, DBT, TBT, entre les phases solide et liquide, est variable d’un test à l’autre selon les conditions opératoires ; (2) l’affinité entre les butylétains et la boue est importante, quelle que soit sa provenance. 98% des butylétains se retrouvent dans la phase solide, du fait de leur hydrophobicité. Le coefficient de distribution Kd entre le TBT et les boues est plus élevé que celui obtenu pour d’autres type de phases solides (75000 L.kg- 1) ; (3) le processus de sorption est rapide. De plus des processus de sorption, désorption et déalkylation du TBT, expliquant les variations des concentrations en MBT et DBT dans la phase solide ont été identifiés. La désorption du TBT est rapide. Il se dégrade dans la phase liquide en DBT dont une partie se transforme en MBT. La sorption du MBT sur les boues a été systématiquement mesurée. Le phénomène de sorption n’est donc pas complètement réversible au cours du temps. Ces résultats contribuent à expliquer les dysfonctionnements observés dans les stations de traitement par lagunage, par exemple lors d’apports directs d’eau de pluie.Cette thèse démontre la présence systématique des butylétains dans les stations d’épuration par lagunage naturel. Les données acquises in-situ couplées à des essais en laboratoire permettent de comprendre le comportement de ces composés lors de ce traitement : la fraction solide joue un rôle clé dans les processus de sorption-désorption-(bio)dégradation des butylétains. / This thesis deals with the behavior of a priority compound on the basis of the framework Directive on water at European level: tributyltin (TBT) ) and its metabolites (dibutyltin (DBT) and monobutyltin (MBT)). The study is focused on two waste stabilization pond treatment plants (WSP), which is an appropriate treatment to domestic effluents from sparsely urbanized areas.This thesis comes to fill the lack of scientific results on (1) the level of butyltin concentrations in domestic effluents from sparsely urbanized areas, in the input of the WSP and the output after treatment; (2) the efficiency of butyltins elimination by the WSP treatment; (3) the identification of the processes that control the fate of these elements during the treatment.Butyltins are measured in particulate and dissolved matrices and in sludge, by GC-ICP-MS, in all stages of the treatment by lagooning of two villages of the Department of Hérault (34, France): Gigean (6000 p.e.) and Montbazin (4500 p.e.).It is shown that the three butyltin species are consistently present in all samples taken from the two WSP. Butyltins arrive primarily in particulate form in the raw water. During treatment, the processes of sorption onto particulate matter and the settling of the latter at the bottom of each basin are the main factors for the elimination of butyltins. The first anaerobic ponds, characterized by deep depth and long residence times, allow the elimination of more than 60% Gigean (93% at Montbazin) of butyltins. Bottom sludge of each basin show high butyltins concentrations. The resuspension of butyltins is noted in the last maturation ponds, characterized by shallow depth, which may be due to re-thermal stratification. At the outlet, despite significant butyltin removal, concentrations in butyltins remain high (12 to 228 and 1.8 to 15 ng (Sn). L-1, respectively for Montbazin and Gigean).Laboratory batch experiments were conducted to study sorption processes following several experimental protocols. It was thus shown that: (1) the distribution of the butyltin species between solid and liquid phases is different from one test to another according to the operating conditions; (2) the affinity between butyltins and bottom sludge is important, regardless of its provenance. 98% of the butyltins are found in the solid phase, due to their hydrophobicity. The TBT sludge-water partition coefficient is higher than that obtained for other type of solid phases (75000 L.kg-1); (3) the process of sorption is fast.Moreover processes of sorption, desorption and dealkylation of TBT, explaining changes in concentrations of MBT and DBT in the solid phase were identified. Desorption of TBT is fast. It degrades in the liquid phase in DBT which part turns into MBT. The sorption of MBT on sludge has been systematically measured. Therefore, the sorption phenomenon is not completely reversible over time. These results help to explain the dysfunctions observed in WSP, for example after direct inputs of rainwater.This thesis shows the systematic presence of butyltin in waste stabilization pond treatment plants. In-situ data coupled with laboratory tests allow to understand the behaviour of these compounds during this treatment: the solid fraction plays a key role in the processes of sorption-desorption-(bio) degradation of butyltins.

[pt] Neste trabalho, métodos baseados em diferentes abordagens em eletroforese capilar (CE) foram propostos. No caso da determinação de compostos organoestanhos ou OTs (difenilestanho e monofenilestanho) em fluidos biológicos, foi usada abordagem de eletroforese capilar por zona (CZE) hifenada com a espectrometria de massas (do tipo quadrupolo) com fonte de plasma indutivamente acoplado (CE-ICP-MS). As condições de análise foram estudadas no modo univariado visando otimizar a composição da solução eletrolítica (tampão acetato 5,0 mmol L(-1), pH 2,8) e obter os parâmetros instrumentais (45 graus Celsius, mais 30 kV e 30 s de tempo de introdução hidrodinâmica de amostra). A solução de complementação foi uma solução aquosa 5,0 mmol L(-1) de NH4NO3 contendo 10 por cento metanol em volume e 1,0 Mg L(-1) de Cspositivo, com pH ajustado para 2,8 com tampão acetato. A vazão dessa solução foi mantida em 40 ML min(-1). Os OTs foram diluídos em solução de metanol:tampão acetato de sódio 50:50 por cento v/v ou apenas em tampão acetato de sódio pH 2,8. As condições de detecção do ICP-MS foram ajustadas em 1200 W, 15 L min(-1) de vazão de argônio para formação do plasma, 1 L min(-1) de vazão de argônio auxiliar. A vazão de argônio do nebulizador foi ajustada diariamente. Os isótopos de estanho 120Sn e 118Sn foram monitorados, assim como o 133Cspositivo para controlar a eficiência e estabilidade do processo de nebulização. A resposta linear do método ficou entre 0,050 a 2,0 mg L(-1) de Sn (0,42 a 17 Mmol L-(1)). Os limites de detecção (LOD) e de quantificação (LOQ) em termos de Sn foram de 15 Mg L(-1) (0,13 Mmol L(-1)) e 50 Mg L-1 (0,42 Mmol L(-1)), calculados utilizando a menor concentração dos picos dos analitos que podem ser diferenciados do sinal de fundo. A repetibilidade para o tempo de migração e área dos picos ficou próximo a 5 por cento. O método foi aplicado na análise de urina, sangue total e plasma fortificados com os OTs. Recuperações entre 75 e 95 por cento foram obtidas. No caso da determinação de sete pesticidas da classe das estrobilurinas (azoxistrobina, 9 dimoxistrobina, fluoxastrobina, picoxistrobina, piraclostrobina, trifloxistrobina e kresoxim-metil) em sopas infantis, foi usada a cromatografia eletrocinética capilar micelar (MEKC) com detecção fotométrica (no UV) com capilar de caminho óptico estendido. Um estudo multivariado, usando um planejamento 33 Box Behnken, indicou que a melhor separação para os pesticidas foi com solução aquosa de eletrólito composto por tampão tetraborato de sódio (5,1 mmol L(-1), pH 9,0) contendo 51 mmol L(-1) de dodecil sulfato de sódio (SDS) e 24 por cento acetonitrila (ACN) em volume. As condições instrumentais foram 25 C e mais 30 kV de diferença de potencial aplicada, 45 s de tempo de introdução hidrodinâmica de amostra e detecção em 210 nm. Para aumentar o poder de detecção, foi usada a concentração dos analitos no capilar. Para tal, as soluções de padrões e amostras foram dissolvidas em solução tampão tetraborato de sódio 45 mmol L(-1): acetonitrila 80:20 por cento v/v. As curvas analíticas apresentaram comportamento linear e os valores de LOD ficaram entre 7,0 Mg L(-1) ou 18 nmol L(-1) (piraclostrobina) a 15 Mg L-1 ou 33 nmol L(-1) (fluoxastrobina). Os valores de LOQ ficaram entre 21 Mg L(-1) ou 54 nmol L(-1) (piraclostrobina) a 45 Mg L(-1) ou 98 nmol L(-1) (fluoxastrobina). A repetibilidade ficou entre 1,7 a 7,9 por cento para a área de pico e entre 0,25 a 0,71 por cento para o tempo de migração. A precisão intermediária, avaliada com análises realizadas em diferentes dias, apresentou valores entre 1,3 a 5,3 por cento para a área de pico e entre 0,06 a 0,90 por cento para o tempo de migração. O método foi aplicado na análise de sopas prontas infantis fortificadas com as estrobilurinas. Os pesticidas foram extraídos aplicando o método QuEChERS com ajuste de pH com tampão acetato e limpeza com extração em fase sólida dispersiva. Os resultados das análises obtidos com um método cromatográfico adaptado da literatura foram estatisticamente iguais aos alcançados com o método proposto. A CZE foi o modo de separação escolhido para mostrar o potencial da determinação indireta de aminoglicosídeos com medição de fluorescência de pontos quânticos (excitação com laser de diodo em 410 nm) amplificada na presença dos analitos. A fotoluminescência dos pontos quânticos (nanopartículas de CdTe modificados com ácido tioglicólico monodispersas em solução) foi mais intensa em solução tampão (pH 8,0) contendo entre 5 e 10 mmol L(-1) de tetraborato de sódio. A interação no capilar entre aminoglicosídeos (neomicina e canamicina) e os pontos quânticos provocou aumento de fotoluminescência dependente do pH do meio (indício de interação de natureza 10 eletrostática). Alguns parâmetros de mérito foram avaliados com uma faixa linear curta (0,1 a 1,0 mol L(-1) para canamicina e 0,03 a 0,5 mol L(-1) para neomicina). Os valores mínimos detectados de 0,1 Mmol L(-1) ou 58 Mg L(-1) (canamicina) e 0,03 Mmol L(-1) ou 27 Mg L(-1) (neomicina) mostram que essa é uma abordagem interessante para a determinação sensível de aminoglicosídeos. / [en] In this work, analytical methods based on different approaches using capillary electrophoresis (CE) have been proposed. For the determination of organotins or OTs (diphenyltin and monophenyltin) in biological fluids, the separation using capillary zone electrophoresis (CZE) was applied using tandem with inductively coupled plasma mass spectrometry (CE-ICP-MS). The conditions for the analysis were optimized in an unvaried way aiming to find the conditions for the electrolyte solution (acetate buffer, 5.0 mmol L (-1), pH 2.8) and the employed instrumental parameters (45C, 30 kV and 30 s of the time for hydrodynamic introduction of the sample). A complementary solution was composed by NH4NO3 5.0 mmol L(-1), 10 por cento v/v of methanol and 1.0 g L(-1) of Cspositive in acetate buffer with pH adjusted to 2.8. The flow of this solution was set to 40 ML min(-1). The OTs were diluted either in a methanol: acetate buffer 50:50 por cento v/v solution or only in sodium acetate buffer pH 2.8. The conditions for detection by ICP-MS were set to 1200 W, 15 L min(-1) for the Ar plasma flow and 1,0 L min(-1) for the auxiliary Ar. The nebulizer Ar flow was adjusted daily. The monitored tin isotopes were 120Sn 118Sn. The isotope 133Cs was also monitored in order to control the efficiency and stability of the nebulization. The method presented a linear response between 0.05 and 2.0 mg L(-1) (0.42 a 17 Mmol L(-1)) for Sn. The value for the limits of detection (LOD) and for the limits of quantification (LOQ) for Sn were 15 Mg L(-1) (0.13 Mmol L-1) e 50 Mg L(-1) (0.42 Mmol L(-1)), calculated based on the lowest concentration of the analyte peaks that can be differentiated from the background signal. The repeatability for migration time and peak area was approximately 5 per cent. The method was applied in the analysis of organotin fortified blood and urine samples with recoveries between 75 and 95 per cent. In the case of the determination of seven strobilurin class pesticides (azoxystrobin, dimoxystrobin, fluoxastrobin, picoxystrobin, pyraclostrobin, trifloxystrobin and kresoxim-methyl) in baby food (vegetable and fruit soups), 12 the micellar electrokinetic capillary chromatography (MEKC) was used using photometric detection (UV) in a capillary with extended optical path. A multivariate study, with 33 Box Behnken design, indicated the best composition for the electrolytic solution to separate the seven pesticides: a sodium tetraborate buffer (5.1 mmol L(-1), pH 9.0) solution containing 51 mmol L(-1) sodium dodecyl sulfate and acetonitrile (24 por cento in volume). The instrumental conditions were 25C, 30 kV of applied voltage, 45 s for hydrodynamic introduction of the sample and detection at 210 nm. To increase the detection power, the concentration of the analytes into the capillary was used by using the Normal Stacking Mode. For this purpose, the solutions of standards and samples were prepared in 45 mmol L(-1) sodium tetraborate buffer solution: acetonitrile 80:20 por cento v/v. The analytical curves presented a linear behavior and the LOD values were between 7.0 Mg L (-1) or 18 nmol L(-1) (pyraclostrobin) to 15 Mg L(-1) or 33 nmol L(-1) (fluoxastrobin). The LOQ values were between 21 Mg L(-1) or 54 nmol L(-1) (pyraclostrobin) a 45 Mg L(-1) ou 98 nmol L(-1) (fluoxastrobin). The repeatability was between 1.7 to 7.9 por cento for the peak area and between 0.25 to 0.71 por cento for the migration time. The intermediate precision, evaluated by the analysis performed in different days were between 1.3 to 5.3 por cento for the peak area and between 0.06 and 0.90 per cent for the migration time. The method was applied in the analysis of baby food spiked with strobilurin. Pesticides were extracted using the QuEChERS method with pH adjustment with acetate buffer and clean-up using the dispersive solid phase extraction. The analysis results were statistically identical to those obtained with a chromatographic method adapted from the literature. The CZE separation mode was chosen to evaluate the potential of the indirect determination of aminoglycosides through the amplified photoluminescence from quantum dots (excitation laser diode 410 nm) in the presence of the analytes. The photoluminescence from quantum dots (monodispersed CdTe nanoparticles modified with thioglycolic acid) was more intense in buffer solution (pH 8.0) containing between 5 and 10 mmol L(-1) sodium tetraborate. The interaction between aminoglycosides (kanamycin and neomycin) and quantum dots inside the capillary caused the increasing of fluorescence in a pH-dependent way (indicating the electrostatic nature for the interaction). A few figures of merit were evaluated with a short linear range (0.1) to 1.0 Mmol L(-1) for kanamycin and 0.03 to 0.5 Mmol L(-1) for neomycin). The 13 minimum values detected were 0.1 nmol L(-1) or 58 Mg L(-1) (kanamycin) and 0.03 nmol L(-1) or 27 Mg L(-1) (neomycin) showing that the proposed approach can be used to detect aminoglycosides in a relatively sensitive way.