La pyrolyse rapide de biomasse lignocellulosique produit des bio-huiles avec des rendements élevés mais ces liquides contiennent de nombreux composés oxygénés, une acidité élevée et de ce fait sont instables. Un hydrotraitement poussé de ces huiles est nécessaire avant qu'elles puissent être utilisées comme combustibles liquides. Afin de réduire le coût et d'améliorer les performances de l'hydrotraitement ultérieur, plusieurs stratégies ont été proposées pour réduire les teneurs en oxygène et en acides, comme l'ajout d'un lit catalytique après la pyrolyse. La conversion catalytique des vapeurs de pyrolyse permet une désoxygénation partielle avant la condensation des vapeurs. Dans ce contexte, nous avons étudié la conversion de molécules modèles, l'acide acétique et le gaïacol, sur différents catalyseurs acides afin d'identifier des phases actives performantes et de comprendre les mécanismes réactionnels. Dans un deuxième temps, un test de pyrolyse semi-continu combiné avec un réacteur catalytique a été conçu et mis en place et les meilleurs catalyseurs acides sélectionnés pour les molécules modèles ont été testés pour la conversion de bois de hêtre. Après conversion, un protocole de séparation des fractions a été appliqué et les fractions gaz, liquide et solide ont été caractérisées par différentes méthodes (?-GC, GC×GC, GPC, RMN…). Une attention particulière a été portée à la quantification des monomères dans la fraction liquide par GC×GC. Les bilans massiques atteignent plus de 90% et des bilans carbones ont été également réalisés dans les différentes fractions. L'ensemble des caractérisations et la comparaison entre pyrolyse avec et sans étape catalytique permet d'améliorer la compréhension du rôle du catalyseur dans la conversion des vapeurs de pyrolyse / Fast pyrolysis of biomass yields bio-oils with high levels of oxygen-containing components, high acidity and low stability. Further upgrading of these oils is necessary before they can be used as liquid fuels. Several low-cost strategies have been proposed for reducing the oxygen and acid contents including the catalytic conversion of pyrolytic vapors before vapor condensation. In this context, a first step in this work was the study of model molecules conversion, acetic acid and guaiacol, on different acid catalyst to understand reaction mechanisms and determine the best catalyst. In a second time, a semi-continuous pyrolysis test combined with a catalytic reactor was built and use to test the best acid catalysts found on model molecules, for the conversion of beech wood chips. A product recovery protocol was implanted to recover all the gas, liquid and solid fractions. Those fractions were characterized in depth by various techniques (?-GC, GC×GC, GPC, RMN…). A special focus was made on the quantification of monomers in the bio-oils by GC×GC. Mass balance reached 90% and carbon balance were established for some experiments. The performed characterizations as well as the comparison between catalytic and thermal experiments allowed to get more insights in the role of the catalyst on pyrolytic vapors conversion

Wildland fire, which includes both planned (prescribed fire) and unplanned (wildfire) fires, is an important component of many ecosystems. Prescribed burning (controlled burning) is used as an effective tool in managing a variety of ecosystems in the United States to reduce accumulation of hazardous fuels, manage wildlife habitats, mimic natural fire occurrence, manage traditional native foods, and provide other ecological and societal benefits. During wildland fires, both live and dead (biomass) plants undergo a two-step thermal degradation process (pyrolysis and combustion) when exposed to high temperatures. Pyrolysis is the thermal decomposition of organic material, which does not require the presence of oxygen. Pyrolysis products may later react with oxygen at high temperatures, and form flames in the presence of an ignition source. In order to improve prescribed fire application, accomplish desired fire effects, and limit potential runaway fires, an improved understanding of the fundamental processes related to the pyrolysis and ignition of heterogeneous fuel beds of live and dead plants is needed.In this research, fast pyrolysis of 14 plant species native to the forests of the southern United States has been studied using a flat-flame burner (FFB) apparatus. The results of fast pyrolysis experiments were then compared to the results of slow pyrolysis experiments. The plant species were selected, which represent a range of common plants in the region where the prescribed burning has been performed. The fast pyrolysis experiments were performed on both live and dead (biomass) plants using three heating modes: (1) convection-only, where the FFB apparatus was operated at a high heating rate of 180 °C s-1 (convective heat flux of 100 kW m-2) and a maximum fuel surface temperature of 750 °C; (2) radiation-only, where the plants were pyrolyzed under a moderate heating rate of 4 °C s-1 (radiative heat flux of 50 kW m-2), and (3) a combination of radiation and convection, where the plants were exposed to both convective and radiative heat transfer mechanisms. During the experiments, pyrolysis products were collected and analyzed using a gas chromatograph equipped with a mass spectrometer (GC-MS) for the analysis of tars and a gas chromatograph equipped with a thermal conductivity detector (GC-TCD) for the analysis of light gases.The results showed that pyrolysis temperature, heating rate, and fuel type, have significant impacts on the yields and the compositions of pyrolysis products. These experiments were part of a large project to determine heat release rates and model reactions that occur during slow and fast pyrolysis of live and dead vegetation. Understanding the reactions that occur during pyrolysis then can be used to develop more accurate models, improve the prediction of the conditions of prescribed burning, and improve the prediction of fire propagation.

The combustion of fossil fuels has for a long time been a problem from an environmental and sustainability point of view, especially when it comes to the emissions of atmospheric carbon dioxide. The environmental concern has for instance shifted the attention towards finding new sustainable alternatives for producing chemicals and fuels, as a substitute to today’s dependence on fossil based crude oil. Catalytic Fast Pyrolysis of biomass is an excellent way to produce valuable chemicals and fuels using renewable resources. However, the process has some drawbacks, for example rapid deactivation of catalysts due to coke formation. Little is known about the characteristics of the formation of catalytic coke from pyrolysis processes, which should be a vital concern in future industrial processes. This thesis is dedicated to investigate the chemical coke characteristics found on zeolitic catalysts. Four zeolites of the type ZSM-5 were chosen for this thesis to deduce any chemical differences in the coke: HZSM-5, Fe/ZSM-5, Ni/ZSM-5, and Fe-Ni/ZSM-5. The coke were characterized by TGA, GC/MS, and FTIR. The results show that Fe/ZSM-5 produced the highest amount of coke compared to the other zeolites, where HZSM-5 had the lowest amount of coke formation. The coke consisted mainly of aromatic and cyclic hydrocarbons, dominated by polycyclic aromatic hydrocarbons. The content of ketones and alcohols in the coke found on HZSM-5 was higher compared to the metal-doped zeolites, while the formation of naphthalenes was lower. The FTIR results also show that coke was mainly comprised of aromatic hydrocarbons. However, traces of alkanes and alkenes reveal that the coke may have a greater variety than the GC/MS analysis suggests. The results show interesting features when metals are introduced to the zeolitic structure, at least when it comes to coke formation. The metal-doping of zeolites certainly seems to alter the chemistry of the catalytic reactions, compared to the parent zeolite. The differences in the chemical characteristics found in the coke are certainly interesting, and it could mean that the chemistry of the bio-oil also varies depending on the metals chosen for the ZSM-5. The new properties that metals introduce to the parent catalyst may open up new possibilities in industrial catalytic processes, and allow industries to take more advantage of the great benefits that biomass has to offer. / Förbränning av fossila bränslen har under lång tid utgjort ett problem ur miljö- och hållbarhetssynpunkt, i synnerhet gällande utsläppen av koldioxid. En större miljömedvetenhet har gett upphov till sökandet efter nya råvaror för att framställa bränslen och kemikalier, utan att förlita sig på fossil råolja. Katalytisk pyrolys av biomassa är ett utmärkt sätt att framställa värdefulla kemikalier från förnybara källor. Processen står dock inför en del tekniska utmaningar, bland annat en snabb deaktivering av använda katalysatorer genom koksning. Målet med detta examensarbete är att undersöka den kemiska sammansättningen av koks, som bildats på zeolitkatalysatorerna. Mer specifikt, att försöka undersöka huruvida den kemiska sammansättningen av koks skiljer sig mellan katalysatorn HZSM-5 och metalldopad HZSM-5. Fyra katalysatorer valdes för detta examensarbete, nämligen HZSM-5, Fe/ZSM-5, Ni/ZSM-5 och Fe-Ni/ZSM-5. Kokset har analyserats genom termogravimetrisk analys (TGA), gaskromatograf kopplad med en masspektrometer (GC/MS), samt Fourier-transform-infraröd-spektroskopi (FTIR). Resultaten visar att Fe/ZSM-5 bildade en större mängd koks jämfört med de andra zeoliterna, varpå HZSM-5 hade lägst halt koks. Utöver detta bestod kokset till största del av aromatiska- och cykliska kolväten, speciellt polycykliska aromatiska kolväten. Innehållet av ketoner och alkoholer i kokset var störst för HZSM-5, medan bildandet av naftalenföreningar ökade för de metalldopade zeoliterna. FTIR-analysen gav även upphov till signaler som är signifikanta för både alkaner och alkener. Därför kan det innebära att kokset innehar en större kemisk variation än vad GC/MS-analysen påvisade. Resultaten visar intressanta egenskaper hos metallmodifierade zeoliter, i synnerhet gällande koksbildning. Det verkar som att de metalldopade zeoliterna påverkar de katalytiska reaktionerna som sker i katalysatorn, jämfört med den obehandlade katalysatorn. Skillnaderna i den kemiska sammansättningen hos kokset för de olika katalysatorerna är definitivt intressant och kan indikera att det även kan föreligga skillnader i den kemiska sammansättningen hos bio-olja, beroende på vilken metall ZSM-5 har behandlas med. De nya egenskaperna som metaller bidrar med till ZSM-5 kan öppna upp nya möjligheter i industriella katalytiska processer, vilket även kan medföra att industrier bättre kan ta tillvara på de fantastiska egenskaper biomassa innehar.

När befolkningsmängden ökar och teknisk och ekonomisk utveckling sker så påverkas även energianvändningen. Detta ställer krav på att energitillförseln är säker, stabil och hållbar. I dag är det fossila bränslen som dominerar globalt sett vilket får konsekvenser för den miljö vi lever i, och dessutom är det en ändlig, ohållbar resurs. Därför behöver dessa ersättas av hållbara alternativa energikällor, vilket också är centralt för miljömål i både Sverige och i den Europeiska Unionen. Förhoppningar finns om att processer som omvandlar lignocellulosa till fasta, flytande och gasformiga drivmedel och bränslen kan bidra till omställningen från fossilt till förnybart. I detta examensarbete som utförts i samarbete med KTH och IVL Svenska Miljöinstitutet har främst en av dessa omvandlingsprocesser undersökts närmare – pyrolys. Pyrolys är en termisk process som omvandlar lignocellulosa under temperaturer mellan cirka 300-650 °C under syrefria förhållanden. Tre faser kan erhållas. En gasfas som kan kondenseras till pyrolysolja, en fast fas som benämns biokol eller kol (beroende på slutanvändning) och en okondenserbar gasfas. Utbytet av produkter och kvalitet på dessa styrs främst av: typ av råvara, typ av reaktor och av vilka processförhållanden som råder. En undersökning av olika pyrolysprocessers status på marknaden har gjorts. Graden av kommersialisering och status i nuläget och hur framtiden kan se ut för både tekniken och produkterna har uppskattats genom litteraturstudier, internetsökningar och intervjuer med utvalda företag och personer med kunskaper inom pyrolys. Rapporten visar att pyrolys inte ännu är en helt kommersiell process, men att den har möjlighet att bli det med rätt förutsättningar. Det är svårt att säga när det sker, då det förutom fortsatt teknisk utveckling, ökad kunskap kring pyrolysprocessen och resultat av demonstrationer beror på olika externa faktorer. Yttre faktorer för kommersialisering av pyrolys i Sverige har identifierats som ökad säkerhet kring politiska styrmedel och beslut kring långsiktiga sådana (osäkerhet och kortsiktiga beslut skrämmer bort investerare), vikten av att etablera en värdekedja för att säkra investeringen, och priser på fossila drivmedel och biomassa som råvara. Processer för produktion av biokol verkar dock ha hunnit längre än de för pyrolysolja och är i ett tidigt stadium av kommersialisering.  Den enda tillämpningen som är fullt kommersiell idag är produktion av träkol och för detta tillämpas ofta traditionella satsvisa processer. Många möjliga användningsområden för produkterna finns där de har potential att reducera koldioxidutsläpp och bidra till en mer hållbar framtid. Standardisering och certifiering av produkter är då viktigt, samt demonstration av användning. Stabilisering och vidare förädling av pyrolysoljan är en annan viktig faktor för kommersialisering. Ännu verkar processer för katalytisk uppgradering inte vara tillräckligt tekniskt eller ekonomiskt utvecklade för att ge en konkurrenskraftig produkt, men forskning pågår kring detta. Integrering av processen ser ut att kunna öka energieffektiviteten, samt bidra till minskade produktionskostnader. / The population growth as well as a rapid technical and economic development globally affects the energy consumption. This requires a secure, stable and sustainable supply of energy. Today fossil fuels dominate globally and this results in environmental problems. Fossil fuels are also a finite, unsustainable resource. Thus, there is a need to replace fossil fuels with sustainable alternative sources of energy. This is also central for environmental goals both in Sweden and in the European Union. There are expectations that processes for the conversion of lignocellulosic biomass to solid, liquid and gaseous fuels can contribute to a transition from fossil to renewable fuels. In this thesis, carried out in collaboration between KTH and IVL Swedish Environmental Research Institute, one of the conversion processes is investigated in detail – pyrolysis. Pyrolysis is a thermal process that converts lignocellulose under anaerobic conditions at temperatures between about 300-650°C. Three phases can be obtained as products. A volatile which can be condensed into pyrolysis oil, a solid which may be termed biochar or charcoal depending on the end use, and a gas phase. The yield and the quality of the products is dependent upon the type of raw material, the type of reactor and the process conditions. An examination of the status of different pyrolysis processes on or on the way to the market has been made. The current degree of commercialization and what the future may look like for both the technology and the products have been assessed through literature studies, internet searches, and interviews with selected companies and individuals with expertise in pyrolysis.   This report reveals that continuous pyrolysis is not yet a fully commercial process, but that it has the opportunity to reach commercialization during the right conditions. It is difficult to say when it occurs, due to various external factors, continued technical development, increased knowledge of the pyrolysis process and results of the current demonstrations. In this report, several critical factors for the commercialization of pyrolysis in Sweden have been identified, e.g. increased stability for policy instruments and that will limit the risk for investments (uncertainty and short-term decisions frightens investors) and the establishment of a value chain for the products, i.e. a stable market. Prices on fossil fuels and biomass feedstock are also important factors. Processes for the production of biochar is in the early stages of commercialization, and seem to have reached further in their development than processes for pyrolysis oil. The only fully commercial application of pyrolysis today is the production of charcoal that commonly is performed in traditional batch-wise processes. There are many possible uses for the products in which they have the potential to reduce carbon emissions and contribute to a more sustainable future. Standardization and certification of products is important, and demonstration of the use. Stabilization and further upgrading of pyrolysis oil is another important factor for commercialization. It seems like processes for catalytic upgrading are not yet sufficiently technically or financially developed to be able to provide a competitive product. Research and development in this area are ongoing. Integration of the process with incumbent industrial processes seems to be able to offer increased energy efficiency and reduced production costs.

Pyrolytic oils produced from poultry litter differ in physico-chemical properties and the chemical composition. The litter is composed of manure and bedding material with traces of spilled feed and feathers. The type and amount of bedding material was varied to investigate its influence on the pyrolysis of layer manure. 400g of each feedstock: manure, wood (pine and oak), and mixtures of manure and wood in proportions (75:25 50:50, and 25:75 w/w %) respectively were subjected to fast pyrolysis at 450oC in a fluidized bed reactor. The total pyrolytic oil yield ranged from 43.3% to 64.5 wt%. The highest bio oil yield and the lowest char yield were obtained from oak wood. The manure oil had the highest HHV of 29.7 MJ/kg, the highest pH (5.89), the lowest density (1.14 g/cm3) and a relatively low viscosity of 130cSt. The oils had relatively high nitrogen content ranging from 5.88wt% to 1.36 wt%; low ash content (approximately <0.07wt %) and low sulfur content (<0.28wt %). FT-IR, 13CNMR, and 1HNMR analysis showed that manure oil was rich in aliphatic hydrocarbon and primary and secondary amides and the addition of wood introduced oxygenated compounds like aliphatic alcohols, phenols, aromatic ethers, and carbonyl/carboxylic groups into the oil. TG/DTG analysis also showed that the thermal decomposition of the oils were different depending on the amount and the type of wood in the manure/wood mixture. The parametric variables used for the mixture of 50% manure and 50% pine wood shavings study were; temperature (400-550°C), nitrogen gas flow rate (12-24 L/min), and feed rate (160-480 g/h). The results showed that the pyrolysis product yields, physical properties and the chemical composition of the oil were influenced by all parameters. Temperature was the most influential factor and its effect on the liquid, char and gas yields were significant. It was evident that depending on the gas flow rate and the feed rate, a maximum oil yield (51.1wt.%) can be achieved between 400-500 oC. Also an increase in temperature significantly increased the oil viscosity and decreased the carbonyl/carboxylic and the primary aliphatic alcohol functional groups in the oil. The study on the influence of wood on the stability of the oils when stored at ambient conditions for 8 months in a 30ml glass bottle showed that the viscosity of the oils increases when stored, however the manure oil was relatively more stable and the oil from the 50/50 mixture for both pine and oak was the least stable. It was found that the stability of the oils from the manure and wood mixtures were dependent on the amount and the type of wood (pine or oak) added to the manure. Also the addition of 10% solvent (methanol/ethanol) to the oil from 50% manure and 50% pine reduced the initial viscosity of the oil and was also beneficial in slowing down the increase in viscosity during storage. / Master of Science

La pyrolyse rapide de différents types de celluloses, lignines, xylanes ainsi que de mélanges synthétiques obtenus à partir de ces composés modèles et aussi de deux biomasses réelles (bois de bouleau et paille de blé), est étudiée dans un four à image. Les vitesses de perte de masse pour les composés modèles étudiés sont très différentes, de manière générale la plus grande vitesse de perte de masse est observée pour la cellulose, suivie du xylane et de la lignine. La lignine se distingue par un temps de début de réaction plus petit que celui observé pour le xylane et la cellulose. Pour un temps donné, le rendement en charbon est plus grand pour la lignine que pour le xylane. La formation de charbon à partir de la cellulose est marginale. La cellulose microgranulaire produit en majorité des vapeurs et très peu de gaz. La cellulose extraite du bois de bouleau montre un comportement intermédiaire entre le comportement de la cellulose microgranulaire et les lignines. Le xylane se distingue par une vitesse de formation de gaz supérieure à celles des vapeurs. Les mélanges synthétiques à partir des trois composés modèles montrent un comportement intermédiaire. Le bois de bouleau est plus réactif que la paille de blé et son comportement peut être comparé à celui de la cellulose extraite du bois de bouleau en ce qui concerne les vitesses de perte de masse et de formation de vapeurs. Une corrélation simple se basant sur l’addition pondérée des vitesses de formation des produits est établie et les résultats obtenus sont comparés aux résultats expérimentaux. Il s’avère que les inorganiques jouent un rôle important lors de la pyrolyse : le comportement de la paille de blé, une biomasse riche en inorganiques, ne peut pas être décrit par de telles corrélations. Des écarts moins significatifs entre les valeurs obtenues par les corrélations théoriques et celles provenant des résultats expérimentaux sont observés pour les mélanges synthétiques et le bois de bouleau. En conclusion, l’hypothèse d’une corrélation simple, appelée loi prévisionnelle, ne semble pas être justifiée sur les vitesses de formation de produits. Un modèle est alors développé qui permet de décrire le comportement pyrolytique à l’aide de courbes non linéaires pour la perte de masse et la production des produits. A l’exception de la paille de blé, de nouvelles lois prévisionnelles additives peuvent alors être déterminées de manière très satisfaisante au regard de la précision des mesures expérimentales / The fast pyrolysis of different types of celluloses, lignins, xylanes as well as different mixtures of these basic compounds and two real biomasses (birch wood and wheat straw) has been studied in an image furnace. The mass loss rates of each of the studied model compounds are very different, but usually higher for cellulose than xylane and higher for xylane than lignins. Lignins begin to react before xylane and cellulose. For a given pyrolysis time, lignins produce more char than xylane. Char formation from cellulose remains insignificant, close to zero. Microgranular cellulose gives mainly rise to vapours and very little gas is formed. Cellulose extracted from birch wood shows an intermediate behaviour between microgranular cellulose and lignins. Production rates of gases are higher than those of vapours for xylan. Regarding the mixtures of these three compounds, an intermediate behaviour can be observed. Birch wood is more reactive than wheat straw. Its pyrolytic behaviour is similar with that of cellulose extracted from birch wood with regard to concerning mass loss rates and vapours formation. A simple additive correlation relying on mass loss rates and products formations rates is elaborated and the results compared to the experimental values. It appears that inorganics play an important role in biomass pyrolysis : the pyrolytic behaviour of wheat straw, a biomass which is rich in inorganics, cannot be described by simple additive correlations. Less significant gaps between expected rates obtained by the correlations and the observed rates obtained by the experiments are found for synthetic mixtures and birch wood. In conclusion, the hypothesis of linear correlations cannot be supported for representing the products formations rates. A complete mathematical model is then developed revealing a non-linear behaviour of the variations of mass losses and products formations. Except for wheat straw, new additive correlations are then determined. The agreement with the experimental results is very satisfying with regard to the measurements accuracies

La pyrolyse primaire rapide de cellulose, lignine, xylane et bois de bouleau (réel, reconstitué, lavé ou non) imprégnés par différents types de catalyseurs (KCl, MgCl2, NiCl2 et ZnCl2) est étudiée expérimentalement dans un four à image. De manière générale, l’effet résultant de l’ajout d’un catalyseur diffère selon le type de biopolymère. Il entraine cependant toujours la diminution de la vitesse de production des liquides et l’augmentation de celle des charbons (effets plus ou moins accentués selon la nature du catalyseur). La composition de phase gazeuse est également modifiée avec augmentation de la fraction volumique en H2. La nature et la morphologie des charbons dépendent également du type de catalyseur. Pour la cellulose, on montre que la quantité et la composition du composé intermédiaire liquide produit lors de la pyrolyse sont également modifiées. Des lois linéaires de comportement pyrolytique établies à partir de la simple superposition des observations faites avec les biopolymères pris isolément n’ont pu être établies avec précision, ce qui implique une phrase de modélisation. Le modèle repose sur l’écriture de bilans de matière et d’énergie au niveau des pastilles de biomasses soumises à une densité de flux de chaleur imposée. L’accord avec les résultats expérimentaux est satisfaisant mais souffre d’une mauvaise connaissance des propriétés physico-chimiques des biomasses. Les écarts proviennent également de la non prise en compte des interactions entre les biopolymères au sein des biomasses réelles. Des actions de recherche dans ces domaines pourraient être recommandées / The fast primary pyrolysis of cellulose, lignin, xylan and birch wood (natural, reconstituted, washed or not) impregnated by different types of catalysts (KCl, MgCl2, NiCl2 and ZnCl2) is experimentally studied in an image furnace. The effect of a catalyst addition differs according to the type of biopolymer. However, the decrease of the liquids and the increase of the char production rates are always observed (the effects are more or less stressed according to the nature of the catalyst). The composition of the gaseous phase is also modified with an increase of the H2 volume fraction. The nature and morphology of chars depend on the type of catalyst. The quantity and composition of the intermediate liquid compound formed during the pyrolysis of cellulose are also modified. It’s difficult to establish linear laws describing the pyrolytic behavior of biomasses from the simple overlapping of the observations made with basic biopolymers, implying the need of a modelling study. The model is based on mass and energy balances written at the level of biomasses pellets subjected to a given heat flux density. The agreement with the experimental results is quite good but suffers from a bad knowledge of the physicochemical properties of the biomasses. It is anticipated that the model could give better results if the influence of the interactions between basic biopolymers within the natural biomasses could be estimated and taken into account. Some researches in these domains could be recommended

Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro / A velocidade do som é uma propriedade que vem sendo cada vez mais utilizada em diferentes áreas tecnológicas. Além disso, a velocidade do som é uma propriedade termodinâmica que está associada a outras propriedades do meio como a compressibilidade isentrópica e isotérmica, entre outras. Neste contexto, muitos estudos foram realizados a fim de obter modelos precisos que possam representar fielmente a velocidade do som, sendo observados desvios absolutos médios entre 0,13 e 24,8%. Neste trabalho, um banco de dados de velocidade do som e massa específica à pressão atmosférica de n-alcanos, alcanos ramificados, n-alcenos, aromáticos, alcoóis, éteres e ésteres, foram compilados da literatura aberta. Utilizando estes dados e baseando-se no modelo de Wada por contribuição de grupo recentemente proposto, foi desenvolvido um novo modelo por contribuição atômica para predizer a velocidade do som de todas as famílias dos compostos investigados neste trabalho. É mostrado que o modelo proposto é capaz de prever a velocidade do som para os compostos destas famílias com desvios próximos da incerteza experimental calculada a partir de diferentes dados da literatura. Este trabalho também discute o efeito da ramificação das cadeias na constante Wada, ressaltando a importância de novas medições para este tipo de compostos. Além disso, observou-se que a literatura necessita de mais dados experimentais de velocidade do som, à pressão atmosférica e diferentes temperaturas para substâncias puras presentes em biodiesel e bio-óleo de pirólise rápida. Neste contexto, o presente trabalho fornece novos dados experimentais de velocidade do som e massa específica de cinco ésteres metílicos de ácidos graxos, também conhecidos como FAMEs, (caprilato de metila, caprato de metila, palmitato de metila, estearato de metila e linoleato de metila), e sete componentes puros presentes em bio-óleo de pirólise à pressão atmosférica, de vários fenóis (fenol, o-, m- e p-cresol), dois éteres fenólicos (2-metoxifenol e eugenol) e um éster fenólico (salicilato de metila), a temperaturas de (288,15-343,15) K. O modelo preditivo de Wada atômico foi utilizado para calcular a velocidade do som dos FAMEs estudados neste trabalho, e os desvios foram comparados com o modelo de Wada por contribuição de grupo. O modelo atômico de Wada foi utilizado para prever a velocidade do som dos componentes puros presentes no bio-óleo de pirólise rápida experimentalmente estudados nesta dissertação. Além disso, os dados de massa específica e velocidade de som foram correlacionados com o modelo de Prigogine-Flory-Patterson (PFP). As propriedades foram bem representadas pelo modelo PFP, no entanto, para a velocidade do som o modelo apresenta desvios sistemáticos na dependência com a temperatura. O desempenho do modelo preditivo de Wada atômico foi considerado satisfatório, devido os desvios observados serem compatíveis ou até menores do que os desvios típicos obtidos na literatura com outros modelos correlativos para o cálculo da velocidade do som de outras substâncias / Speed of sound is a property that is being increasingly used in different technological areas. Furthermore, the speed of sound is a thermodynamic property which is associated with other properties of the medium, such as isentropic and isothermal compressibility, among others. In this context, many studies were carried out to obtain accurate models that can faithfully represent the speed of sound, with average absolute deviations between 0.13 and 24.8%. In this work a database of speed of sound and density at atmospheric pressure for n-alkanes, branched alkanes, n-alkenes, aromatics, alcohols, ethers and esters were collected from the open literature. Using these data a Wada group contribution model recently proposed was used as basis for the development of a new atomic contribution model to predict speed of sound for all families of compounds investigated in this work. It is shown that the proposed model is able to predict the speed of sound for compounds of these families with deviations close to the experimental reproducibility. This work also discusses the effect of branching on the Wadas constant, pointing out the importance of new measurements for this type of compounds. It was also observed that the literature needs more experimental data of speed of sound at atmospheric pressure and different temperature for pure compounds present in biodiesel and fast pyrolysis bio-oil. In this context, this work provides new experimental data of speed of sound and density for five Fatty Acid Methyl Esters, also know FAMEs, (Methyl Caprylate, Methyl Caprate, Methyl Palmitate, Methyl Stearate and Methyl Linoleate), and seven pure components of pyrolysis bio-oil at atmospheric pressure for several phenols (phenol, o-, m- and p-cresol), two phenolic ethers (2-methoxyphenol and eugenol) and one phenolic ester (methyl salicylate) at temperatures ranging from (288.15 to 343.15) K. The predictive atomic Wada model was used to calculate speed of sound of FAMEs studied in this work, and the deviations were compared with group contribution Wada model. An extension of atomic Wada model was used to predict the speed of sound of pure compounds of fast pyrolysis bio-oil experimentally studied in this thesis. Furthermore, data of densities and speed of sound are correlated with the Prigogine-Flory-Patterson (PFP) model. The properties are well described by the PFP model, however the model presents a systematical deviation on the temperature dependency of the speed of sound. The performance of the predictive atomic Wada model was very satisfactory because its deviations are comparable to, or better than, those obtained in the literature with other models

Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro / A velocidade do som é uma propriedade que vem sendo cada vez mais utilizada em diferentes áreas tecnológicas. Além disso, a velocidade do som é uma propriedade termodinâmica que está associada a outras propriedades do meio como a compressibilidade isentrópica e isotérmica, entre outras. Neste contexto, muitos estudos foram realizados a fim de obter modelos precisos que possam representar fielmente a velocidade do som, sendo observados desvios absolutos médios entre 0,13 e 24,8%. Neste trabalho, um banco de dados de velocidade do som e massa específica à pressão atmosférica de n-alcanos, alcanos ramificados, n-alcenos, aromáticos, alcoóis, éteres e ésteres, foram compilados da literatura aberta. Utilizando estes dados e baseando-se no modelo de Wada por contribuição de grupo recentemente proposto, foi desenvolvido um novo modelo por contribuição atômica para predizer a velocidade do som de todas as famílias dos compostos investigados neste trabalho. É mostrado que o modelo proposto é capaz de prever a velocidade do som para os compostos destas famílias com desvios próximos da incerteza experimental calculada a partir de diferentes dados da literatura. Este trabalho também discute o efeito da ramificação das cadeias na constante Wada, ressaltando a importância de novas medições para este tipo de compostos. Além disso, observou-se que a literatura necessita de mais dados experimentais de velocidade do som, à pressão atmosférica e diferentes temperaturas para substâncias puras presentes em biodiesel e bio-óleo de pirólise rápida. Neste contexto, o presente trabalho fornece novos dados experimentais de velocidade do som e massa específica de cinco ésteres metílicos de ácidos graxos, também conhecidos como FAMEs, (caprilato de metila, caprato de metila, palmitato de metila, estearato de metila e linoleato de metila), e sete componentes puros presentes em bio-óleo de pirólise à pressão atmosférica, de vários fenóis (fenol, o-, m- e p-cresol), dois éteres fenólicos (2-metoxifenol e eugenol) e um éster fenólico (salicilato de metila), a temperaturas de (288,15-343,15) K. O modelo preditivo de Wada atômico foi utilizado para calcular a velocidade do som dos FAMEs estudados neste trabalho, e os desvios foram comparados com o modelo de Wada por contribuição de grupo. O modelo atômico de Wada foi utilizado para prever a velocidade do som dos componentes puros presentes no bio-óleo de pirólise rápida experimentalmente estudados nesta dissertação. Além disso, os dados de massa específica e velocidade de som foram correlacionados com o modelo de Prigogine-Flory-Patterson (PFP). As propriedades foram bem representadas pelo modelo PFP, no entanto, para a velocidade do som o modelo apresenta desvios sistemáticos na dependência com a temperatura. O desempenho do modelo preditivo de Wada atômico foi considerado satisfatório, devido os desvios observados serem compatíveis ou até menores do que os desvios típicos obtidos na literatura com outros modelos correlativos para o cálculo da velocidade do som de outras substâncias / Speed of sound is a property that is being increasingly used in different technological areas. Furthermore, the speed of sound is a thermodynamic property which is associated with other properties of the medium, such as isentropic and isothermal compressibility, among others. In this context, many studies were carried out to obtain accurate models that can faithfully represent the speed of sound, with average absolute deviations between 0.13 and 24.8%. In this work a database of speed of sound and density at atmospheric pressure for n-alkanes, branched alkanes, n-alkenes, aromatics, alcohols, ethers and esters were collected from the open literature. Using these data a Wada group contribution model recently proposed was used as basis for the development of a new atomic contribution model to predict speed of sound for all families of compounds investigated in this work. It is shown that the proposed model is able to predict the speed of sound for compounds of these families with deviations close to the experimental reproducibility. This work also discusses the effect of branching on the Wadas constant, pointing out the importance of new measurements for this type of compounds. It was also observed that the literature needs more experimental data of speed of sound at atmospheric pressure and different temperature for pure compounds present in biodiesel and fast pyrolysis bio-oil. In this context, this work provides new experimental data of speed of sound and density for five Fatty Acid Methyl Esters, also know FAMEs, (Methyl Caprylate, Methyl Caprate, Methyl Palmitate, Methyl Stearate and Methyl Linoleate), and seven pure components of pyrolysis bio-oil at atmospheric pressure for several phenols (phenol, o-, m- and p-cresol), two phenolic ethers (2-methoxyphenol and eugenol) and one phenolic ester (methyl salicylate) at temperatures ranging from (288.15 to 343.15) K. The predictive atomic Wada model was used to calculate speed of sound of FAMEs studied in this work, and the deviations were compared with group contribution Wada model. An extension of atomic Wada model was used to predict the speed of sound of pure compounds of fast pyrolysis bio-oil experimentally studied in this thesis. Furthermore, data of densities and speed of sound are correlated with the Prigogine-Flory-Patterson (PFP) model. The properties are well described by the PFP model, however the model presents a systematical deviation on the temperature dependency of the speed of sound. The performance of the predictive atomic Wada model was very satisfactory because its deviations are comparable to, or better than, those obtained in the literature with other models