Hannover, Universiẗat, Diss., 1999.

Vid produktion av nya transformatorer adsorberas fukten från luften i den fasta isolationen (presspan) vilket påverkar isolationskapaciteten. Därför genomgår transformatorn en torkningsprocess innan den fylls med olja. Vatteninnehållet i den fasta isolationen mäts kvantitativt genom Karl Fischer (KF)-titrering. KF-titrering används ofta för bestämning av vattenhalten i produkter inom industrin och är en beprövad metod som använts under många år. Bestämning av vattenhalten genom manuell KF-titrering är tidskrävande på grund av omständlig provberedning genom termisk extraktion med metanol. En automatiserad metod för bestämning av vattenhalten i presspan kloss har utvecklats genom KF-ugnsextraktion. En kvantitativ analys av vattenhalten i presspan med KF-coulometri genom ugnsextraktion jämfördes med KF-coulometri genom metanolextraktion. Syftet med projektet var att undersöka möjligheten att utveckla en snabbare och enklare mätmetod för vattenhalten i kaseinlimmad presspankloss (HDLC) och polyesterlimmad presspankloss (HDLP) genom att implementera KF-coulometri med ugnsextraktion. Det initiala steget var att undersöka och bestämma en metod för provberedning och bestämning av driftparametrar (ugnstemperatur, inertgasflöde och extraktionstid). Driftparametrar som bestämdes var en ugnstemperatur på 170 grader Celsius, inertgasflödet på 50 ml/min och extraktionstid på 600 sekunder. Resultatet av undersökningen av olika provtagningsmetoder indikerade att provtagning med skölp eller stämjärn inte uppfyllde kraven för att ersätta provtagning genom borrning. Validering av metoden genomfördes med en certifierad fast standard med känd vattenhalt (1%) från Aquastar (certifierad i ett ISO/IEC 17025 ackrediterat testlaboratorium). Den uppmätta halten med den utvecklade metoden var lägre än den förväntade. En kalibreringsmodell upprättades utifrån standarden med en korrelationskoefficient av 0,9952 som indikerar att kalibreringsmodellen visade god linjäritet. Riktigheten för mätmetoden har utvärderats enligt ISO 5725-4. Metodens riktighet visade sig inte vara godkänd på grund av att avvikelsen översteg gränsvärdet 3%. Resultatet av jämförelse mellan metanolextraktion och ugnsextraktion indikerade att metoden inte var lika bra som metanolextraktion på grund av lägre resultat av vattenhalter i jämförande prov. Metoden är möjlig att tillämpa under villkoren att felen som påverkar analysmetoden identifieras och elimineras. Ytterligare bekräftande experiment behövs för implementering av metoden. / In the production of new transformers, the moisture from the air is adsorbed in the solid insulation and affects the insulation capacity. Therefore, the transformer undergoes a drying process before filling it with oil. Water content in the solid insulation is measured quantitatively by Karl Fischer titration. Karl Fischer titration is widely used for determining water content in products in the industry and is a proven method that has been used for many years. The manual method for determining moisture content in pressboard is time consuming due to long sample preparation with thermal extraction with methanol. An automated method for water content determination in pressboard has been developed by KF oven extraction. A quantitative analysis of the water content of pressboard by KF coulometry through oven extraction was compared with KF coulometry through methanol extraction. The purpose of this study was to investigate the possibility to develop a faster and simpler measurement method for the moisture content in High-Density Casein Laminated Pressboard (HDLC) and High-Density Polyester Laminated Pressboard (HDLP) by implementing oven extraction for KF coulometry. The initial step was to investigate and determine methods for sample preparation and operating parameters (oven temperature, carrier-gas flow and extraction time). Determined operating parameters were oven temperature 170 degrees Celsius, carrier gas flow of 50 ml/min and extraction time of 600 seconds. The result of investigation of different sample preparations indicated that sample selection with gouge or chisel did not meet the demands to replace sample selection by drilling. Validation of the method was carried out with a certified solid standard of known moisture content (1%) from Aquastar (certified in an ISO/IEC 17025 accredited testing laboratory). The measured concentration was lower than the expected concentration. A calibration model was established from the standard with a correlation of 0.9952 indicating the calibration model showed good linearity. The trueness of the measurement method has been evaluated according to ISO 5725-4 confirming the trueness of the method was not approved. The result of the comparison of methanol extraction and oven extraction indicated that the method was not as good as methanol extraction due to lower results of water content in the comparison samples. The method is applicable under the terms of the error is identified and eliminated. Further confirmatory experiments are needed to implement the method.

Cepheid provides instruments and tests for molecular diagnostic testing of bacterial and viral diseases. Their goal is to deliver a better way to improve patient outcomes by enabling access to molecular diagnostic testing everywhere. These test kits are used in warm and humid countries which may challenge the stability of the kits. The test kits containly ophilized beads, which contain enzymes for the polymerase chain reaction, which usually ages first. The ageing of the beads is effected by the initial water content in them since it defines the quality. Therefore, it is of great importance to have a very precise and exact water measurement method. The degree project covers the optimization of the Standard Operating Procedure used today at Cepheid with the research question: Can optimization of the Karl Fischer method improve the precision when measuring the initial moisture content of freeze-dried enzyme beads? The method for measuring the water content was the Karl Fischer titration method and the instrument used was the Mettler Toledo C30S Compact Karl Fischer-coulometer. The different optimized parameters were the sample preparation, extraction process and how the measurement was proceeded. It can be concluded that optimized method provides significantly different results compared with the old method and therefore the new method is better. / Cepheid tillverkar instrument och test för molekylär diagnostisk testning av bakteriella och virala sjukdomar. Målet är att leverera en lösning som leder till en förbättring för patienterna genom att möjliggöra tillgång till molekylär diagnostisk var än den behöver utföras i världen, under stora variationer i mätmiljön. Dessa testkit används i varma och fuktigalä nder vilket kan utmana stabiliteten hos produkten. Testen innehåller lyofiliserade kulor och det är dessa som vanligtvis åldras först. Kulornas hållbarhet påverkas av den ursprungliga vattenhalten, vilket definierar kvalit´en. Därför är det mycket viktigt att ha en mycket exakt metod att mäta fuktinnehållet i kulorna. Examensarbetet omfattar optimering av den standardoperativa proceduren som används idag på Cepheid och fr°ageställningen är: Kan optimering av Karl Fischer-metoden förbättra precisionen vid mätning av det ursprungliga fuktinnehållet i frystorkade enzymkulor?Den metod som används vid mätningen av vattenhalten är Karl Fischer-titreringsmetoden och instrumentet som användes var Mettler Toledo C30S Compact Karl Fischer-coulometer. De olika optimerade parametrarna var: förberedelsen av proven, extraktionsprocessen och själva utförandet av mätningen.Den slutsats som kan dras efter optimeringen är att den nya metoden ger ett mer exakt och bättre resultat än den gamla.

Dans les applications industrielles et lors des opérations de transport du gaz naturel, la présence d'eau sous forme liquide ou en phase vapeur peut entraîner la formation d'hydrates provoquant le colmatage des unités industrielles et des lignes de conduites et il est indispensable de définir précisément les seuils de déshydratation à réaliser, afin d'éviter la formation d'hydrates. Cela est réalisé à l'aide d'un modèle thermodynamique qui prédit la stabilité des hydrates, en fonction de la température, de la pression et de la composition du gaz.Les modèles thermodynamiques classiques, développés uniquement sur la base de données expérimentales de formation d'hydrates en présence d'eau liquide, surestiment fortement la température de dissociation des hydrates en l'absence d'une phase aqueuse.Dans le but de définir un modèle thermodynamique capable de représenter convenablement les équilibres de phases vapeur-hydrate et prédire ainsi la température de dissociation des hydrates que l'on soit en présence ou en l'absence d'eau liquide, nous avons mis au point une méthodologie originale pour la détermination de la température de dissociation des hydrates de corps purs et de mélanges en l'absence d'eau liquide. Cette méthodologie, basée sur le suivi de la teneur en eau de phase vapeur, en fonction de la température par coulométrie Karl Fischer, a permis la détermination de la température de dissociation de plusieurs hydrates simples et mixtes à des teneurs en eau et pressions différentes ainsi que les quantités d'hydrates formées dans ces conditions.Sur la base de ces nouvelles données, nous avons défini un modèle thermodynamique basé sur l'utilisation de l'approche de Dharmawardhana pour le calcul de la fugacité de l'eau dans l'hydrate vide,le potentiel de Kihara pour le calcul de la constante de Langmuir et l'équation d'état CPA (Cubic Plus Association) pour la modélisation des phases fluides. Nous avons montré que l'utilisation de l'équation d'état CPA, capable de prendre en compte l'auto association de l'eau apporte une amélioration très significative.Le développement d'un flash biphasique hydrate-fluide nous a permis de calculer les quantités d'hydrates mixtes formées et de les comparer à nos données expérimentales. / In industrial applications and during natural gas transport, the presence of water under liquid form or within a vapour phase can lead to gas hydrate formation causing the blockage of industrial units and transport lines. Hence, in order to avoid such situation, it is very important to well determine its formation conditions. It is occurred by using a rigorous thermodynamic model. Due to the lack of data in the literature concerning gas hydrates formation in the absence of an aqueous phase,usual thermodynamic models predict correctly gas hydrate dissociation temperature only in the presence of aqueous water. Our purpose is to propose a thermodynamic model with hydrate phase that can predict gas hydrate dissociation temperature in both cases: with and without water liquid phase.At first, using an existing apparatus, we have developed a new experimental protocol in order tomeasure gas hydrate dissociation temperature in the absence of liquid water. It consists in measuring the water content in the vapour phase as a function of the temperature by using a Karl Fischer coulometer. We have measured the dissociation temperature of many simple and mixture hydrates.We have also developed a thermodynamic model that is able to predict correctly gas hydrate dissociation temperature, in the absence and in the presence of liquid water. This model is based onthe use of Dharmawardhana's approach for the calculation of hydrate fugacity in the empty hypothetical hydrate, Kihara potential for the calculation of the Langmuir constant and CPA EoS forfluid phases modelling. We have shown that the use of CPA EoS improves the prediction of gas hydrate dissociation temperature. We have also developed a biphasic flash (hydrate-fluid) allowing the calculation of the formed mixture hydrate amounts. The calculated amounts are in agreement with the experimental ones.

<p>Abstract (in English)</p><p>The task for the diploma work was to develop methods for measuring of moisture content by the Karl Fischer method, in jam, jelly sweets, and fudge.</p><p>The start premises was a for the company whole new equipment, which should be started up. Then programs should be developed for different kind of samples.</p><p>In the task it was as well included to develop methods for dissolving the different kind of samples, since the Karl Fischer method demands completely dissolved sample material.</p>

Vattenhaltsbestämning är av stor vikt i många sammanhang. T.ex. kan vattenhalten påverka utbytet av en kemisk syntes, eller ha negativ inverkan på hållbarheten av läkemedel och livsmedel. Standardmetoden för vattenhaltsbestämning är Karl Fischer-titrering, baserad på antingen volymetri eller coulometri. I den här avhandlingen presenteras nya infallsvinklar för bestämning av mycket låga halter vatten i komplexa provmatriser, som t.ex. tekniska oljor och substanser som interfererar med alkoholbaserade Karl Fischer-reagens. Vattnet avskiljs ofta från oljematrisen före titrering genom förångning. I samband med framtagningen av nya referensmaterial för vatten i olja ifrågasattes förångningsteknikernas effektivitet av National Institute of Standards and Technology (NIST). NIST menade att en fraktion av vattnet bands hårt i oljefasen och att det inte kunde frigöras och detekteras annat än med en modifierad volymetrisk metod där reagenset innehöll minst 65% kloroform. I den här avhandlingen presenteras en alternativ metod som uppfyller det ställda kravet för en fullständig upplösning av oljefasen. Med denna metod visas att det inte finns någon anledning att ifrågasätta förångningsteknikernas effektivitet och att den modifierade metoden som NIST använder ger systematiskt för höga resultat. Fördelar som enklare handhavande, kortare konditioneringstider och att endast ett reagens behövs har gjort att diafragmafri coulometri har blivit allt mer populär. Spårhaltsbestämning med denna teknik ställer dock speciellt höga krav på reagensen eftersom strömtätheten vid katoden är låg. Med anledning av detta testades olika typer av kommersiella reagensblandningar för bestämning av små vattenmängder och kritiska parametrar identifierades. Dekanol visade sig ha en gynnsam effekt på katodreaktionen i reagens modifierade med xylen enligt standardmetodbeskrivningen för bestämning av vatten i oljor. För provtyper som inte går att analysera med alkoholbaserade reagenser presenteras en ny typ baserad på N-metylformamid. Med ett sådant reagens bestämdes vattenhalten i ett reaktivt salt som används i litiumjonbatterier. Liknande alkoholfria reagens undersöktes mer utförligt i en djupare studie som även inkluderade formamid och dimetylformamid. För- och nackdelar med dessa alternativa lösningsmedel diskuteras och möjliga reaktionsförlopp föreslås. Det visade sig att läget på jämvikten mellan svaveldioxid och vätesulfit är en avgörande faktor för att förklara den stora skillnaden i reaktionshastighet i dessa lösningsmedel. / Moisture determination is of great importance in the production and use of many substances. For example, the moisture content can affect the efficiency of a chemical reaction or determine the shelf life of pharmaceuticals or foods. The standard method for moisture determination is Karl Fischer (KF) titration, based on either volumetry or coulometry. This thesis concerns new approaches to trace determination in complex sample matrices and is focused on oils and substances that interfere with alcoholic KF reagents. Moisture is frequently separated from oil matrices before titration by means of evaporation techniques. In connection with the preparation of new reference materials for moisture in oil, the National Institute of Standards and Technology (NIST) questioned the efficiency of such evaporation techniques. NIST claimed that some of the moisture was sequestered in the oil phase and that it could only be released and detected by using a modified volumetric KF method with a reagent containing at least 65% chloroform. In this thesis, an alternative KF method that meets the proposed requirement for a complete dissolution of the oil sample is presented. With this method it is shown that there is no reason to question the efficiency of the evaporation techniques and that the criticized volumetric method used by NIST is biased high. Ever since its introduction diaphragm-free coulometry has gained popularity due to its ease of use, with a single reagent and short conditioning times. Trace determination with this technique sets great demands on the reagent due to the resulting low current densities at the generator cathode. The performance of several commercial reagents is evaluated under such unfavorable conditions and critical titration parameters are identified. It is also shown that decanol has a favorable effect on the cathode process when using reagents modified with xylene according to standard methods for moisture determination in oils. For samples that are incompatible with the alcohol component in ordinary KF reagent a new reagent based on N-methylformamide is presented. It is shown that is works well for determinations of moisture in a conductive salt used in lithium-ion batteries. The concept of alcohol-free KF reagents is taken a step further in a systematic investigation, also including formamide and dimethylformamide. Advantages and disadvantages with these solvents are discussed and possible reaction paths are surveyed. It is shown that the position of the sulfur dioxide/hydrogen sulfite equilibrium is the main explanation for the large differences in the KF reaction rates in these solvents.

Abstract (in English) The task for the diploma work was to develop methods for measuring of moisture content by the Karl Fischer method, in jam, jelly sweets, and fudge. The start premises was a for the company whole new equipment, which should be started up. Then programs should be developed for different kind of samples. In the task it was as well included to develop methods for dissolving the different kind of samples, since the Karl Fischer method demands completely dissolved sample material.

This diploma thesis deals with the monitoring of chemical changes in ground coffee stored in various containers and then it deals with the determination of water, moisture and aromatic profile of coffee in individual containers, namely a can, a paper bag, a bag with a polyethylene liner, a bag from low density polyethylene with aluminium liner (LDPE + Al) and a glass container. At the beginning of the experiment it was found that the water content in the coffee was 4.72 ± 0.10 % and the moisture content was 9.47 ± 0.01 %. After the original sample was measured, the coffee was poured into 5 packages. At the end of the experiment, a can was chosen as the most suitable package where the lowest water content was measured. The water content was determined to be 6.51 ± 0.11 % and the moisture content was 9.93 ± 0.01 %. However, the differences between the other packages, apart from the paper bag, were very small, and thus in general it can be said that the can, the LDPE + Al package, the bag with the PE liner and the glass container are suitable for storing ground coffee. A total of 44 aromatics were identified when determining the aromatic profile in coffee by HS-SPME-GC-MS. The most represented groups of substances were furans and pyrazines. During storage, there was some loss of substances due to 2 mechanisms. The first of them is evaporation and the second one is the oxidation reactions themselves. Most aromatic substances have evaporated / degraded in paper packaging, where this large decrease is related to its barrier properties. A glass container was chosen as the most suitable packaging with the least loss of aromatic substances.

The actual life of a transformer is determined by ageing of the cellulosic insulation such as transformer paper. The presence of moisture in the insulation system decreases the electrical strength of paper and accelerates the aging. It is an efficient way to monitor the moisture content in paper insulation by measuring the dielectric constant of the paper. The moisture dependent permittivity of impregnated transformer paper below 1 MHz has been widely investigated. High frequency (&gt; 1 MHz) dielectric spectroscopy is under requirement since the loss peak information is missing. The impregnated paper is kept in desiccators with saturated salt solutions to get the samples with moisture content from 1 % to 5.5 %. Then they are placed in a coaxial line and the scattering parameters are obtained by modern Vector Network Analyzer. Full wave analysis is utilized to calculate the permittivity from the obtained S-parameters due to its high accuracy. The magnitude of the dielectric spectroscopy below 100 Hz is very dependent on the moisture content while there is a horizontal shift of curves towards higher frequencies depending on the water content over a wide frequency range. The loss peaks appear between 1 MHz and 1 GHz for the impregnated paper with moisture level less than 5.5 %. Due to the limited system accuracy, there is a blank frequency band from 1 MHz to 100 MHz.

A esterificação do sec-butanol com ácido acético, submetida à catálise homogênea, apresenta poucos resultados experimentais de cinética reacional e equilíbrio químico disponíveis na literatura. A presente dissertação visa determinar os parâmetros cinéticos e de equilíbrio químico desta reação de esterificação, empregando modelo cinético de 2ª ordem baseado em atividades. Os experimentos cinéticos foram realizados em reator batelada com auxílio da metodologia Karl Fischer para análise do teor de água e determinação do perfil de frações molares ao longo da reação. A principal dificuldade na modelagem desta esterificação consistiu em representar satisfatoriamente as não-idealidades da mistura em fase líquida, em espacial nas reações cuja fração molar inicial de ácido acético foi superior a 50%. Netas condições admitiu-se a ocorrência de dissociação do ácido acético levando ao surgimento de interações iônicas. O modelo empregado para obtenção das atividades, com base em frações molares, foi o NRTL. Este modelo representou bem as interações moleculares em condição de baixa acidez, mas apresentou desvios em condições de acidez e teores de água mais elevados. O estudo também visou o desenvolvimento de modelos de calibração para um método analítico on-line visando determinar o perfil de frações molares ao longo da reação. O método empregado foi a espectroscopia de infravermelho próximo (NIR), em vista à capacidade deste método analítico em determinar quantitativamente frações molares de água, além dos demais compostos envolvidos. Outro fator determinante ao emprego do NIR é a seu crescente uso no meio industrial, e o presente trabalho visou também contribuir para o uso futuro desta técnica de monitoramento no processo industrial. Modelos de calibração multivariável com conjuntos de calibração interna por validação cruzada se mostraram adequados ao sistema estudado. Calibrações externas resultaram em modelos imprecisos, não sendo possível determinar claramente os fatores responsáveis pela imprecisão. / There is a lack of literature data on the reaction kinetics and equilibrium of the homogeneously catalyzed esterification of sec-butanol with acetic acid. This work aims at obtaining the kinetic parameters and chemical equilibrium parameters of this esterification reaction by using a second order kinetic model based on activities. The kinetic experiments were performed in a batch reactor and the Karl Fischer analytical method was used to determine the molar fraction of water along the reaction progress. The main challenge of modeling this reaction was to represent the highly non-ideal mixture in liquid phase, specially for reactions runs starting with acetic acid mole fraction higher than 50%. In these conditions, it was assumed the occurrence of acetic acid dissociation that leads to ionic interactions. The thermodynamic model used to calculate the activities was the NRTL. This model predicted well the molecular interactions on low acidity conditions but presented deviations when molar fractions of acid and water were higher. The present study also aims to develop calibration models for analytical on-line method to obtain molar fraction profiles along the reaction. Near infrared spectroscopy (NIR) was used due to the possibility of quantitatively analyze water mole fraction, besides the mole fractions of the other components. Another important reason to choose NIR was its crescent use in industry, and the present work intended to contribute towards the future use of this on line monitoring technique in industrial applications. Multivariate calibration models using internal set of calibration and cross-validation seems to be suitable for this system. External set of calibration leads to less accurate models, and the causes of this lack of accuracy were not clearly identified.