Phase Equilibria for extraction processes with designer solvents
Abstract
In recent years, there has been an increasing concern about the effects of toxic chemicals in the environment. In response to this concern, there is a growing impetus to develop chemical manufacturing processes which can reduce or eliminate the use or generation of hazardous substances. Within this aspect of green chemistry, in this Thesis, Ionic Liquids (ILs) and Deep Eutectic Solvents (DESs) have been tested as greener alternatives in different chemical processes. In a first stage, the capability of using ILs as solvents for citrus essential oil deterpenation by liquid-liquid extraction was investigated. The liquid-liquid equilibria of ternary systems limonene + linalool + IL were determined. Results were analysed in terms of the solute distribution ratio and selectivity, allowing to draw conclusions about the influence of the structure of the ILs in these thermodynamic parameters. Experimental data were correlated by means of the NRTL and UNIQUAC equations. To evaluate the capability of some ILs to act as surfactants in Enhanced Oil Recovery, two of these salts showing surfactant behaviour were tested. Liquid-liquid equilibria of ternary systems water + IL + ndodecane were determined. Winsor Type III systems were found, and the interfacial tensions between phases were measured to prove the ability of these salts to reduce the water – oil interfacial tension. Physical and transport properties of the phases involved, such as density and viscosity, were also experimentally measured. In a last stage, an introductory study on the processing of lignocellulosic biomass with DESs was carried out. Different renewable DESs were prepared from hydrogen bond donor and hydrogen bond acceptor starting materials. Solubility tests of pine wood and wheat straw in these solvents were carried out. From the thermodynamic studies that were carried out in this Thesis, it turns out that designer solvents can be used to develop greener and/or more efficient chemical processes.
Full text
UNIVERSITYOFSANTIAGODECOMPOSTELA SCHOOLOFENGINEERING Department of Chemical Engineering P PH HA AS SE E E EQ QU UI IL LI IB BR RI IA A F FO OR R E EX XT TR RA AC CT TI IO ON N P PR RO OC CE ES SS SE ES S W WI IT TH H D DE ES SI IG GN NE ER R S SO OL LV VE EN NT TS S A thesis submitted by SARA LAGO GARCÍA DE DIOS for the degree of Doctor in Chemical and Environmental Engineering to the University of Santiago de Compostela Santiago de Compostela, September 2013
iii Department of Chemical Engineering School of Engineering Rúa Lope Gómez de Marzoa s/n 15782 Santiago de Compostela Tel: + 34 881816730 Authorizationforsubmissionbythethesisdirectors (inGalician) Alberto Arce Arce e Ana M. Soto Campos, Catedráticos de Enxeñería Química, INFORMAN: Que a presente memoria, titulada “Phase Equilibria for extraction processes with designer solvents”, que para optar ó grao de Doutor en Enxeñería Química, Programa de Doutoramento de Enxeñería Química e Ambiental, presenta Sara Lago García de Dios, foi realizada baixo a nosa inmediata dirección no Departamento de Enxeñería Química da Universidade de Santiago de Compostela. Considerando que constitúe traballo de Tese, autorizan a súa presentación ao Centro de Posgrao da Universidade de Santiago de Compostela. E para que así conste, asinan o presente informe en Santiago de Compostela, a 01 de setembro de 2013. Asdo.: Alberto Arce Asdo.: Ana Soto
v Abstract In recent years, there has been an increasing concern about the effects of toxic chemicals in the environment. In response to this concern, there is a growing impetus to develop chemical manufacturing processes which can reduce or eliminate the use or generation of hazardous substances. Within this aspect of green chemistry, in this Thesis, Ionic Liquids (ILs) and Deep Eutectic Solvents (DESs) have been tested as greener alternatives in different chemical processes. In a first stage, the capability of using ILs as solvents for citrus essential oil deterpenation by liquid-liquid extraction was investigated. The liquid-liquid equilibria of ternary systems limonene + linalool + IL were determined. Results were analysed in terms of the solute distribution ratio and selectivity, allowing to draw conclusions about the influence of the structure of the ILs in these thermodynamic parameters. Experimental data were correlated by means of the NRTL and UNIQUAC equations. To evaluate the capability of some ILs to act as surfactants in Enhanced Oil Recovery, two of these salts showing surfactant behaviour were tested. Liquid-liquid equilibria of ternary systems water + IL + n-dodecane were determined. Winsor Type III systems were found, and the interfacial tensions between phases were measured to prove the ability of these salts to reduce the water – oil interfacial tension. Physical and transport properties of the phases involved, such as density and viscosity, were also experimentally measured. In a last stage, an introductory study on the processing of lignocellulosic biomass with DESs was carried out. Different renewable DESs were prepared from hydrogen bond donor and hydrogen bond acceptor starting materials. Solubility tests of pine wood and wheat straw in these solvents were carried out. From the thermodynamic studies that were carried out in this Thesis, it turns out that designer solvents can be used to develop greener and/or more efficient chemical processes.
vii Acknowledgements The last four years, I had the privilege of being part of the Group of Phase Equilibria and Separation Processes of the University of Santiago de Compostela working as a PhD student. The present thesis tries to summarise the experimental work carried out in this period. Now that the work is concluded and a new cycle is coming ahead, I would like to thank all the people who helped me one way or another. I would like to thank Professors Alberto Arce and Ana Soto for giving me the great opportunity of being part of their group, teaching and giving me advise in every professional and personal aspect of my life, always making things easier. Also, for treating their PhD “children” as we all were members of a big family. I hope they can feel proud of the final result. I would also like to thank all the people who were part of Laboratory 1.3 during more or less time for sharing such a great time together. Specially I would like to thank María Francisco who taught me all the knacks of the laboratory and was also a good friend. To Alicia and to Iago, for being not only great mates but also good friends and for making research so much fun. To Héctor for having always an answer for every question and for being such a patient and good teacher. I would like also to thank Eva for giving me advise and having very nice times at the coffee break. And last of all I would like to thank Borja, who besides of being the best working mate and friend, he is the best person in the world and I think I would had never reach this point without him. My gratitude is extensive to all the members of Separation Technology Group of Eindhoven University of Technology where I spent three fabulous months learning and enjoying with them. It was my pleasure to work under the supervision of Professor M.C. Kroon, giving me the opportunity of spending this time at her laboratory and teaching me about the existence of Deep Eutectic Solvents. I would like to thank María Francisco again, who once more crossed in my life, this time as a postdoc, teaching me and spending a lot of time to take advantage of the research in such a short time. Thanks to this stay I was able of adding the chapter of biomass fractionation with DESs to this thesis work. I can not forget my lab colleagues. Adriaan who also helped me every day with the biomass processing and was a fantastic colleague, and of course Agustin, who was a great friend and I keep in mind really nice moments with him. Thanks to Somayeh and Lawien for sharing happy lunches. And almost ending I would like to thank all my friends of Santiago de Compostela and A Coruña, for always being there and giving me support no matter what.
viii To Roge because you were always there unconditionally. And finally, I would like to thank my family, Kira and Carlos and my parents as they believed in me from the very beginning and supported me unconditionally along the way. I want to close these lines by expressing my gratitude to the University of Santiago de Compostela, and the “Ministerio de Educación y Ciencia” of the Spanish Government. We acknowledge the Ministry of Science and Innovation of the Spanish Government for financial support through project CTQ2009-10776. Sara Lago Santiago de Compostela, September 2013
Contents ix CONTENTS 1. OBJECTIVES ............................................................................... 3 2. THEORETICAL CONSIDERATIONS ............................................... 9 2.1. DESIGNER SOLVENTS .......................................................... 9 2.1.1. IONIC LIQUIDS ................................................................ 9 2.1.1.1. Definition .................................................................. 9 2.1.1.2. Properties ................................................................ 11 2.1.1.3. Applications ............................................................ 12 2.1.2. DEEP EUTECTICS SOLVENTS ....................................... 14 2.1.2.1. Definition ................................................................ 14 2.1.2.2. Properties ................................................................ 15 2.1.2.3. Applications ............................................................ 16 2.2. SOLVENT EXTRACTION PROCESSES .................................. 18 2.2.1. CITRUS ESSENTIAL OIL DETERPENATION ................... 18 2.2.1.1. Essential oils: definition and composition ................ 18 2.2.1.2. Citrus essential oil: extraction processes ................. 19 2.2.1.3. Deterpenation: definition and methods .................... 20 ILs in essential oil deterpenation ...................................... 22 2.2.2. ENHANCED OIL RECOVERY ......................................... 22 2.2.2.1. General methods for oil recovery ............................. 22 2.2.2.2. Tertiary oil recovery ................................................. 24 2.2.2.3. Surfactant flooding .................................................. 26 Phase behaviour ............................................................... 28 Ionic liquids in surfactant flooding.................................... 29 2.2.3. BIOMASS PROCESSING ................................................ 30 2.2.3.1. Biomass composition .............................................. 31 2.2.3.2. Biomass fractionation ............................................. 32 Ionic Liquids in biomass fractionation .............................. 33 Deep Eutectic Solvents in biomass fractionation ............... 35 3. RESEARCH PROTOCOL ............................................................. 39 3.1. CITRUS ESSENTIAL OIL DETERPENATION ......................... 39 3.1.1. CHEMICALS .................................................................. 39 3.1.1.1. Physical properties of pure compounds ................... 44
2 Theoretical considerations
Theoretical Considerations 9 2. THEORETICAL CONSIDERATIONS 2.1. DESIGNER SOLVENTS Green chemistry research based on new solvents is a topic of interest because many of the solvents commonly used in the chemical industry are considered as unsafe for reasons of environmental protection. They are often used in huge amounts and they are volatile liquids responsible of atmospheric contamination. Promising neoteric solvents are Ionic Liquids (ILs) and Deep Eutectic Solvents (DESs). 2.1.1. IONIC LIQUIDS 2.1.1.1. Definition Even though numerous definitions can be found in literature for the term Ionic Liquid (IL), the most common is the one that defines them as salts with a melting point under 100 ºC [1-3]. Within this family of salts we can find those which are liquid at room temperature, widely used due to their easier handling, these are the so-called RoomTemperature Ionic Liquids (RTILs). Figure 2.1. Schematic structures of some common cations in ILs. Alkyl substituent chains are represented by R, R’, R’’ and R’’’. One of the very first uses of an IL can be dated in the latest 1880’s by Gabriel and Weiner [4], who reported the use of the protic IL ethanol-ammonium nitrate. Even though, the year 1914, in which Walden’s reported [5] the synthesis of ethylammonium nitrate, is nowadays considered the birth of ILs. Advances in research of these salts were slow and not numerous until 1990. That year, Wilkes and Zaworotko [6] arouse the interest on these salts with the report of NR 1,1-dialkylpyrrolidinium tetraalkylammonium tetraalkylphosphonium trialkylsulfonium 1,3-dialkylimidazolium 1-alkylpyridinium
Phase Equilibria for Extraction Processes with Designer Solvents 10 moisture stable ILs created by replacing the aluminum chloride with other anions, such as tetrafluoroborate or hexafluorophosphate. Since this date, the number of scientific publications on this topic has been exponentially increasing over the years. ILs consist exclusively of ions. The low melting point of ILs is due to the poor coordination of the cations and anions which constitute them. ILs are formed by a weekly coordinating anion and a large asymmetrical organic cation. These characteristics of their constituting species tend to weaken the packing, thus reducing the lattice energy of the crystalline form of the salt and lowering this way the melting point of the IL [7]. Figures 2.1 and 2.2 show several cations and anions which constitute the most typical ILs found in bibliography. Figure 2.2. Schematic structures of some common anions in ILs. Alkyl substituent chains are represented by R. Given the current emphasis on environmentally friendly chemistry and the sustainable use of resources, another area of growing interest is the obtaining of ILs that are based on low-toxicity natural products. Choline, for example, can be used as a cationic head-group, while the salts of fatty acids that have been modified by insertion of a few ethylene oxide groups between the alkyl chain and the carboxylate head-group can be liquid even at room temperature [8]. Regarding to the great variety of anions and cations present in nature, the freedom in designing the organic cation (by variation of the side chain length or by varying the substituents on the ring and/or on the chain), and the different possible combinations chloride bromide nitrate tetrafluoroborate hexafluorophosphate bis(trifluoromethylsulfonyl)imide trifluromethanesulfonate alkylsulfatedicyanamide trifluoroacetate
Theoretical Considerations 11 between the cation and anion, there is the possibility of designing millions of different ILs [9]. Besides this, an almost limitless number of systems are theoretically possible by combination of two or more ILs. 2.1.1.2. Properties The current technological revolution based on ILs is driven by their unique properties. Determination of these properties and their trends constitutes an essential point for “choosing” the right IL for a determinate application. Their characteristic strong anion-cation interactions, gives them specific features which commonly used organic solvent or even high temperature molten salts do not show. Their negligible vapour pressure and the huge liquid range are two important consequences of these strong interactions. The consequence for green chemistry is that ILs represent the ultimate non-volatile solvents [10]. Thus, one of the primary driving forces behind research into ILs derived from the benefit of substituting traditional industrial solvents, most of which are volatile organic compounds (VOCs), with ILs which show a practically null vapour pressure [11]. This replacement would prevent the emission of a major source of environmental pollution. The nature of the forces in different ILs may however differ from one to another and mainly control their physical properties. A key advantage derived from this characteristic of ILs over conventional solvents is their tuneability [2,7]. Specific properties for a determinate application can be obtained by subtly modifying the side chain length of the cation or by exchanging the anion. This tailoring allows obtaining ILs to dissolve a determined compound or making them miscible in aqueous or organic phases. Also their chemical and physical properties as viscosity, density, hydrophobicity, melting point, and reactivity, among others, can be modified by their specific design [7]. This feasibility in obtaining desired properties by changing as much the cation as the anion, has led to the introduction of the term “designer solvents” [12], which is an advantage for the development of task-specific ILs. In conclusion, there are plenty of different properties which make ILs attractive compared with volatile solvents [12-14]. Furthermore, these properties can be changed by selecting a determinate cation or anion or changing the side length of the ions which constitute them. Some of these properties can be generalized for the big family of ILs, as their low melting points or negligible vapour pressure. For other properties, these generalizations can’t be applied to all the ILs. For example, it has been established that some characteristics as thermal stability or miscibility depend basically on the anion structure, while viscosity, superficial tension or density are mainly affected by the side chain length of the cation or by its shape or symmetry. Some ILs are hydrophobic while others are hydrophilic. This change in a property can be seen in imidazolium halide ILs,
Phase Equilibria for Extraction Processes with Designer Solvents 12 which show a high miscibility with water, but by just exchanging the halide anion for a bistriflamide, the IL becomes hydrophobic. Depending on the IL a wide range of inorganic, organic or polymeric materials can be dissolved. Some ILs are protic and other aprotic. Some are easily biodegradable and some aren´t. Some are toxic while others are non-toxic. Most of the ILs show high conductivity, thermal and electrochemical stability and they are normally non-flammable [15]. This long list of varied properties, their low vapour pressure, large liquid range and their attractive behaviour over common organic solvents, make these salts an outstanding field of study to find the right structure in a solvent to make it fit in a determinate application. 2.1.1.3. Applications The increasingly interest in the study of ILs and the possibility of designing millions of different ILs by combining different anions and cations, multiplied by their numerous physical and chemical properties, make the number of laboratory and industrial applications of these solvents be practically countless. In this section just a brief overview of all these applications will be listed. Due to the particular properties of ILs above cited, enthusiasm for these salts was initially focused on their use as an alternative to volatile organic solvents in synthesis and catalytic reactions [9,14]. ILs have been employed as reaction media for processes as Heck reaction, addition of thiols to unsaturated ketones, l-proline catalysed aldol reaction, Diels-Alder reactions, Friedel-Crafts alkylation and acylation reactions, hydroformylation reactions, Pdmediated C-C bond formation, alkene polymerization and biotransformations… among many others. In general, it was found that reaction rates and selectivities are as good or better in ILs as in conventional organic solvents. Nevertheless, in the last years, ILs have opened new technological possibilities for a constantly increasing range of application fields [16,17], due to their wide range of physical and chemical properties enhanced by their promise for environmentally friendly applications. ILs in general, present attractive properties to be considered as alternative to common organic solvents in separation processes. These applications are mainly due to their high capacity of solubilisation and selectivity for some compounds instead of others. In this line, ILs are investigated as solvents for liquid-liquid extractions [18] and gas separations (specially CO2 capture) [19]. The long-term stability of a liquid membrane configuration is mainly dependent on the volatility of the liquid membrane, thus these salts are of great interest because they offer improved liquid membrane stability with high permeability and selectivity [20,21]. Electrochemical processes have been another important application area for ILs since their early development. As ILs present
Theoretical Considerations 13 a wide electrochemical window, high conductivity, and vanishingly low vapour pressure, they have been explored for being used in various electronic applications [22], including substituent of organic volatile electrolytes for Li-ion batteries and solar cells [22,23], capacitors and charge storage devices [24], as well as electrodeposition of metals [22,25], among many others. ILs provide amazing functional properties such as dissolution of bio-related materials that are hardly dissolved in common solvents. The dissolution of biopolymers as cellulose or lignin is one of such examples [26,27]. Chloride based ILs dissolve cellulose better than other solvents because of the hydrogen bonding between chloride anions with the hydroxyl groups of the cellulose. The use of ILs allows a simple system for the processing of cellulose and has potential environmental and cost advantages over current processing methodologies. Another interesting feature of ILs is their application as reaction media for different organic and inorganic nanomaterials [28,29]. The theories of the formation and solubility of nanoparticles in ILs have different foundations, for example the salting out effects or cation exchange processes, among others. In addition, ILs can be used for the characterization of nanomaterials. The fact that they can be easily solubilized in ILs facilitates the analysis. Petroleum industry has also put attention in the use of ILs because of their properties and versatility of potential application in this field [30]. ILs may be used to inhibit the aggregation of asphaltenes and paraffins which deposition into the rock reservoir. Clogging pumps, valves and pipes represent a very serious and constant problem (with an enormous economic impact) in oil production and transport [31,32]. Furthermore, the self-organization, micelle formation and surface properties of ILs as well as their use to dehydrate and desalt crude oil have been reported [33,34]. Another interest field of ILs application in the oil industry is the petroleum refining. These solvents have been reported as extraction solvents for desulfurization and denitrogenation of diesel and gasoline fuels by mere liquid-liquid extraction or liquid-liquid extraction coupled with catalytic oxidation [35]. There is a huge amount of research done on ILs. Their potential application in many and diverse fields of research as: gas chromatography stationary phases, lubricants, heat-transfer fluids, additives, absorbents in absorption refrigeration, liquids for thermometers, etc… has been carried out [10]. In spite of this big research effort, the number of processes working at an industrial level is still very reduced [16]. Among them, the BASIL (Biphasic Acid Scavenging using ILs) can be highlighted. This process produces an alkoxy-phenyl-phosphine and generates an IL in situ that facilitates the removal of the waste HCl formed in the reaction. BASF realized therefore the first commercial application of ILs in industry. Eastman Chemical Company has been running a
Phase Equilibria for Extraction Processes with Designer Solvents 14 process for the isomerization of 3,4-epoxybut-1-ene to 2,5dihydrofuran since 1996. The plant is now idle, because the market for the product has declined. The IFP (Institute FranÇais du Pétrole) has modified the Dimersol process (dimerization of alkenes) using chloroaluminate (III) ILs as solvents for these nickel-catalysed dimerization reactions. The Dimersol-Difasol process become much more efficient. QUILL (Queen’s University Ionic Liquid Laboratories) has developed a new technology in partnership with Petronas for scrubbing toxic and corrosive mercury from natural gas streams [36]. The system, HycaPure Hg, has been installed in a full-scale commercial on-shore gas terminal in Malaysia successfully producing sales-quality natural gas for over a year. Many other companies, such as BP, Degussa, Air products, ExxonMobil, IoLiTec, etc. do also have projects involving ILs at advanced stage of development. The main limitation of the use of ILs, even with all of the synthetic advances, is still their cost. Common ILs remain quite expensive, particularly when compared to conventional organic solvents. As a result, a very important area of study in the design of equally stable anions much less expensive, and also less toxic and more biodegradable, is being carried out. 2.1.2. DEEP EUTECTICS SOLVENTS 2.1.2.1. Definition Green solvents are characterised for their low toxicity, higher low solubility in water (low miscibility), easily biodegradability under environmental conditions, high boiling point (not very volatile, low odour, less health problems to workers), and easily to recycle after use [37]. In this line of green chemistry we can find some relatively new solvents which have gained increasing attention over the last years, the so called Deep Eutectic Solvents (DESs). These are ionic fluids generally composed by a hydrogen bond donor and a hydrogen bond acceptor in solid state, self-associated by hydrogen bond interactions, deriving in a eutectic mixture with lower melting point than that of each individual component. These starting materials are usually easily biodegradable, cheap and non-toxic compounds. DESs are generally liquid at temperatures lower than 150 ºC. Although the term DES is standardized and its use generalized in bibliography, Francisco et al. [38] introduced an alternative term: Low Transition Temperature Mixtures (LTTMs). They choose this term for mixtures that do not show eutectic melting points but glass transitions instead. Abbott et al. [39], presented these new solvents based on new renewable sources in 2003, with the mixture of choline chloride and urea, with melting points of 302 ºC and 133 ºC respectively, in a ratio 1:2. The resulting liquid mixture has a melting point of 12 ºC. Nevertheless, the term DESs does not appear until 2004, when the
Theoretical Considerations 15 same author proposes these solvents as an alternative to ILs due to their “greener” characteristics [40]. One of the most interesting advantages of these solvents is their easy and cheap fabrication by simply mixing two or three compounds, avoiding additional volatile organic solvents as synthesis medium. Even more, these precursors are normally cheap, easily available, biodegradable and non-toxic compounds as quaternary ammonium salts (usually choline chloride) combined with different hydrogen bond acceptors as natural amino acids (i.e. alanine or glycine) or carboxylic acids present in fruits and vegetables [38, 41]. Figure 2.3 shows typical structures of halide salts and hydrogen bond donors used for DESs syntheses. Figure 2.3. Typical structures of halide salts and hydrogen bond donors used for DES preparation. 2.1.2.2. Properties DESs show a wide range of properties which make them an attractive new family of solvents for different applications. Due to the similar characteristics of DESs compared to ILs, it is hard to find NOH Cl NH3Cl ZnCl2NO O Cl NCl N Bu Bu Bu Bu Cl P Br H NOH Cl NCl Cl N Et Et Et Et Br NOH Cl H2N O NH2H2N S NH2NNH O H H2N O NN H O N HHN O NH Ph O NH2H3C O NH2F3C O NH2 HO OH HO OH OH O OH H HO OH HO OH O O HO O O HO OH O O HO OH OOH OOH O OH O HO OOH O Halide Salts Hydrogen Bond Donors (ChCl) (EtNH 3 Cl) (TMACl) (AcChCl) (TBACl) (MeP(Ph) 3 Br) (ClChCl) (Et 2 (EtOH)ACl) (TBABr)
Phase Equilibria for Extraction Processes with Designer Solvents 22 ILs in essential oil deterpenation Interesting properties and advantages that ILs present in liquidliquid extraction over conventional solvents [17,18] have drove to the exploration of the use of these innovative solvents in extraction processes for the separation of different valuable compounds of essential oils. Among other applications in liquid-liquid extraction, the suitability of using ILs as solvents for the deterpenation of essential oils has been already proposed for the Group of Separation Processes of the University of Santiago de Compostela [55,63]. In these works, the essential oil was simulated as a mixture of two compounds: the terpene limonene and the oxyterpene linalool. Two ILs based on an imidazolium cation were investigated for the deterpenation of the model of the citrus essential oil: 1-ethyl-3methylimidazolium methanesulfonate ([C2mim][OMs]) and 1-ethyl-3methylimidazolium ethylsulfate ([C2mim][EtSO4]). In the first work [55], the extraction capability of the IL [C2mim][OMs] was compared to that of other conventional organic solvents, obtaining good results for the extraction with the IL over them. Also, the suitability of the IL [C2mim][EtSO4] for the extraction of linalool was studied, obtaining high selectivity values for the solute [63]. Nevertheless, low solubilities of linalool have been found in both cases, meaning large quantities of IL needed for an efficient separation. Therefore, their use on an industrial scale could not be competitive even though the recovery of IL could be practically quantitative. In conclusion, results obtained from the studies done in this field to date invite to be optimistic to the use of ILs as solvents for an efficient citrus essential oil deterpenation by means of liquid-liquid extraction. Nevertheless, an IL showing high solubilities and selectivities is needed to ensure the efficacy of the process. 2.2.2. ENHANCED OIL RECOVERY Energy availability is strongly linked to health and to the level of welfare of the human being. While improving the exploiting efficiency of other sources of energy (renewable, hydrogen, nuclear…), dependence on fossil fuels remains enormous, and their greater exploitation is needed. Petrochemical industry is, moreover, the platform for global growth and development. It supplies to more than 30 derived industrial activities with raw materials. This demand of oil has driven to the search of new methods to recover a bigger part of the oil retained in the wells. With the increase in oil price, new technologies developed for this task have become more interesting over the years and lots of investigations are being carried out in this field. 2.2.2.1. General methods for oil recovery The extraction of natural gas and oil from deep underground reserves consists on a hard task which can be affected by numerous factors of the subterranean environment, i.e. the porosity of the rock
Theoretical Considerations 23 and the viscosity of the deposit. These conditions can impede the free flow of product into the well. Some decades ago, the maximum recovery of oil did not go up to 10% of the total oil contained in the reservoir, leaving the rest underground because the technology did not exist to bring the rest to the surface. Today, advanced technology allows production of about 60% of the available resources from a formation. The overall recovery of oil can be divided in three categories: primary, secondary and tertiary or Enhanced Oil Recovery [64]. Primary recovery occurs because of natural energy, inherent in the oil reservoir, which pushes fluids into the well bore and lifts them to surface. The main source of this energy is the liberation and expansion of dissolved gas, the expansion of the gas cap of an active aquifer, gravity drainage, or a combination of these effects. The primary recovery stage reaches its limit either when the reservoir pressure is so low that the production rates are too low, or when the proportions of gas or water in the production stream are too high. Before this point is achieved, there is a need of an extra-energy source to drive the oil towards the production well. During primary recovery, only a small percentage of the initial hydrocarbons in place is recovered, typically around 10-15 % of oil reservoir [65]. Figure 2.6. Displacement of oil through the reservoir rocks by water-flooding (five-spot pattern) [64]. Secondary recovery methods rely on the supply of external energy into the reservoir in the form of injecting fluids to increase reservoir pressure, hence replacing or increasing the natural reservoir drive with an artificial alternative. Sometimes pumps are used to bring the oil to the surface. Other secondary recovery techniques increase the reservoir’s pressure by water
Phase Equilibria for Extraction Processes with Designer Solvents 24 injection, natural gas reinjection and gas lift, which injects air, carbon dioxide or some other gas into the bottom of an active well, reducing the overall density of fluid in the wellbore. The most commonly used technique, water-flooding, drags out oil mobilised by viscous forces. Ultimately secondary recovery techniques reach a state where the oil production is too low to maintain profitability. Typical recovery factor from water-flooding operations is about 25-30 %, with variations depending on the properties of oil and the characteristics of the reservoir rock [65]. Figure 2.6 shows an example of the mechanism and functioning of a water-flooding process within the reservoir [64]. Primary and secondary oil extraction methods typically recover one third of the original oil in place. This is due to their limitations regarding to two factors [65]: At a pore scale, crude oil is trapped by capillary forces inside the pores as discontinuous globules. At a reservoir scale, there are zones where the injected fluid while secondary extraction does not penetrate because of low permeability of the reservoir, preferential paths, or the non favourable geometry of the wells. Tertiary oil recovery (also known as Enhanced Oil Recovery, hereinafter EOR) consists on the extraction of the residual oil which remains in the reservoir after primary and secondary recovery methods. EOR occurs when media of increasing fluid mobility within the reservoir are introduced in addition to secondary techniques. A deeper study of these EOR methods will be carried out in the following section. 2.2.2.2. Tertiary oil recovery The worldwide diminution of easy-to-access oil, along with an increase in the demand for oil and its high price, has rendered full exploitation of reservoirs (including the recovery of the hard-toaccess oil) necessary. EOR methods have been proven to be effective in recovering the remaining oil from reservoirs that have lost drive during the application of primary and secondary recovery methods [66]. Remaining oil left after primary and secondary recovery operations is usually distributed in pores in the reservoir, where the oil is trapped, mainly due to capillary and viscous forces. Tertiary recovery begins when secondary oil recovery isn’t enough to continue adequate extraction, but only when the oil can still be extracted profitably. This depends on the cost of the extraction method and the current price of crude oil. When prices are high, previously unprofitable wells are brought back into use and when they are low, extraction is curtailed. The three major EOR methods currently used can be classified in three categories [66]: Chemical flooding, which can be divided into three main categories: surfactant flooding, polymer flooding and alkaline flooding.
Theoretical Considerations 25 Miscible flooding: Carbon dioxide, nitrogen and hydrocarbons are commonly used. Thermal recovery. The thermal processes include: steam drive, cyclic steam injection and in situ combustion. All these techniques employ one or more of the following mechanisms for improving the extraction of oil: a) Increasing the mobility of the displacement medium by increasing the viscosity if the water, decreasing the viscosity of the oil, or both. b) Extraction of the oil with a solvent. c) Decrease of the oil/water interfacial tension. Thermal recovery is normally applied in reservoirs which contain heavy oil. It mainly consists on the injection of steam into a petroleum reservoir or propagation of a combustion zone through a reservoir by air or oxygen-enriched air injection. The principles of functioning of this process are illustrated in Figure 2.7. This high increase of temperature reduces oil viscosity within the reservoir facilitating its drive towards the production wells. Figure 2.7. Thermal recovery by steam injection into the reservoir. The fastest growing EOR process is miscible flooding. It consists on the injection into a petroleum reservoir of a material that is miscible, or can become miscible, with the oil in the reservoir. This miscibility enhances the mobility of the oil trapped in the porous rocks. Miscible flooding with CO2 is one of the most widely used EOR processes. CO2 is highly soluble in oil, swelling the oil and reducing the viscosity. This process is illustrated in Figure 2.8. The main problem with miscible gas flooding is the adverse mobility ratio caused by the low viscosity of the typical injecting gas compared to oil, which can be as high as one or two orders of magnitude. This can derive in the formation and propagation of viscous fingers thorough the displaced fluid, reducing the contact with some oil zones [67].
Phase Equilibria for Extraction Processes with Designer Solvents 26 Figure 2.8. Miscible flooding using CO2 as miscible gas. An alternative method that does not suffer from the shortcomings of the miscible flooding is chemical EOR. The development of EOR processes based on operations which involve chemicals is greatly promising. The principles of chemical flooding operation are schematized in Figure 2.9. Figure 2.9. Principles of chemical flooding. ASP stands for Alkali-SurfactantPolymer. Chemical flooding consists on the injection of water together with other chemicals (surfactants, alkalis…) into the reservoir. The lowering of the interfacial tension between water and oil is the main driving force that enables the use of such methods. Change in fluid viscosity upon addition of chemicals like polymer mixtures are also observed and present some advantages. In general, there are three types of chemical EOR: surfactant flooding, polymer flooding and alkaline flooding. 2.2.2.3. Surfactant flooding Surfactant flooding (also known as microemulsion flooding) is based on injecting an aqueous solution of surfactant and co-surfactants into the reservoir creating advantageous conditions in order to mobilize the trapped oil. The injection of surfactant reduces the water/oil interfacial tension, enhancing the mobility of the oil
Theoretical Considerations 27 trapped inside the pores of the rocks, letting it be flushed away by the flowing water. Correctly designed surfactant systems together with the crude oil can create microemulsions at the interface between crude oil and water, thus reducing the interfacial tension to ultra low values, which consequently will mobilize the residual oil and results in improved oil recovery. Figure 2.10 illustrates the principles of surfactant flooding. Figure 2.10. Principles of surfactant flooding [64]. There are two possible ways of application of this process. In the first, a relatively low concentration of surfactant (2-4 %) is injected in the bigger pores to reduce the interfacial tension between water and crude oil, enhancing the recovery. If phase formulation is the adequate, interfacial tension between aqueous and oil phases can achieve a minimum. The conditions of obtaining such tension have to do with the optimum formulation. This optimum formulation is associated to the presence of a triphasic system where a microemulsion, an oil excess phase and a water excess phase are coexisting. In the second case, bigger quantities of surfactant (8-12 %) are injected in smaller pores. The micelles solubilize the trapped oil and water into which is called a mobile microemulsion [66]. Mobility control is important in order to obtain an effective process. Surfactant viscosity should be high enough to decrease the water/oil mobility ratio, but not as much to create pumping problems. To avoid phase degradation an aqueous polymer phase is usually introduced into the reservoir after injection of the surfactant slug. The aim of the polymer is increasing the water viscosity, increasing this way the sweep efficiency. This method is called surfactantpolymer flooding. Although surfactant flooding techniques are very promising, its application to date has been limited due to several factors as the high cost of the surfactant and the difficulty in its recovery, as a result of its adsorption onto the bearing forming rocks. Furthermore, the surfactants must remain active at reservoir conditions such as high temperatures, pressures and salinities. For these reasons, there is growing interest in finding new surfactants which properties that
Phase Eq u best fit process Phase b During system w ater/ b propos e of the p as well behavi o equilib r studied types o f T hese r below. Figure 2 W (2) a n constit u regions diagra m general l tie-line concen t I n region s w ith a n is low. u ilibria for Ex the surfa c s . b ehaviour surfactant - can be b rine, oil a e d surfacta n p hase beha v as the ph y o ur of thes e r ium can b e the beha v f ternary p h r epresenta t 2 .11. Winso r W insor type n d a mono u ted by on e and one m s, it was l y very red u is genera l t ration on t h n the type s eparates i n n upper oil e The affinit y traction Proc e c tant EOR - assisted E considere d a nd surfac t n t flooding v iour of the y sical and t e compone n e represen t v iour of th e h ase diagra m t ions are s r ternary di a I and typ e phasic re g e triphasic r monophas i demonstra t u ced and n l ly horizo n h e critical m I diagram , n to a lowe r e xcess pha s y of the s u e sses with Des 28 requireme n OR, once w d as a ant. To a n method, it ternary sy s t ransport p n ts and the i t ed in tern a e se compo n m s at a fix e s howed in a gram types. e II diagra m g ion (1), w r egion (3) s i c region [ t ed that t h n egligible f o n tal and c o m icelle con c , any mixt u aqueous m s e when th e u rfactant to w s igner Solvent s n ts, and w w ithin the r pseudo-te r n alyse the is critical t s tem water roperties o f ir different a ry diagra m n ents defin i e d tempera t Figure 2. 1 A: Surfacta n m s have o n w hile type s urrounde d 70]. Conc e h e inferior b o r this typ e o rresponds c entration o u re within m icroemuls i e concentr a w ards the s h ich opti m r eservoi r , t h r nary sys t effectivene s t o have kn o + surfacta n f each pha s concentra t m s. Winsor i ng three d t ure and pr 1 1 and ex p n t, W: Water , n e biphasic III diagra m d by three b e rning to t b iphasic r e e of syste m to a su r or der [71]. the immi s i on in equi l a tion of su r aqueous p h ise the h e fluid em of s s of a o wledge n t + oil, s e. The t ions in [68,69] ifferent e ssure. p lained O: Oil. region m s are i phasic t ype III gion is m s. This r factant s cibility l ibrium r factant h ase is
Theoretical Considerations 29 bigger than that of the oil. In the monophasic region the surfactant concentration ensures a complete miscibility between oil and water. Type II diagram shows a similar situation to the type I. Nevertheless, oil and water behaviour are “inverted”. The surfactant affinity for the oily phase predominates. Any composition within the binary region separates into an upper oily microemulsion phase in equilibrium with a water excess phase. For the type III diagram, any system with global composition within the biphasic regions separates equally to previous explained systems. For systems with global composition within the triphasic region, the separation results into three phases in equilibrium: a lower water phase and an upper oil phase, which are mainly constituted of water and oil, respectively, and a microemulsion medium phase. In this diagram interactions are equilibrated and the surfactant is in which is called the optimum formulation. This physicochemical situation corresponds to the obtaining of an extremely low interfacial tension, and consequently the practically total elimination of the capillary forces which trap the crude oil in the porous medium. Although real systems are multicomponent and diagrams are more complex, the optimum formulation concept maintains its original definition. Besides this, the number of phases can be visually analysed and compositions can be determined by adequate analytical methods. Ionic liquids in surfactant flooding Within the large family of ILs, some have been found to be surface active agents. Pino et al. [72] have described several dozens of ILs that exhibit characteristics of cationic surfactants (long carbon chain substituents appended to a charged cationic headgroup) and summarize their important applications in analytical chemistry and separation science. The amphiphilicity responsible for a remarkable interfacial and aggregation capability of IL surfactants may be associated with the cation, with the anion, or with both [73]. Properties such as the critical micelle concentration or the aggregation number can be controlled by changing the length and shape of the hydrocarbon tail, the type of polar head group, and/or the nature and size of the counter-ion [74-77]. In the patent by Collins et al. [73] a great variety of ILs which show surfactant characteristics are listed. These ILs show important properties for the formation of stable emulsions and microemulsions. At some point of this patent, authors cite the use of the water-in-oil microemulsions obtained for deploying water soluble or water dispersible oil field or gas field production chemicals as described in EOR. Other researchers have investigated the use of ILs to demulsify water-in-crude oil emulsions to obtain water free oil [33,34,78]. Shang et al. [79] use ILs as co-surfactants complementing the action of classical surfactants in EOR.
Phase Equilibria for Extraction Processes with Designer Solvents 30 Several Winsor type III diagrams have been found in the literature for water + IL + oil systems [80-82] although they are not focused in EOR. The first direct proposal of using ILs as surface active agents in EOR was done by the Group of Separation Processes of the University of Santiago de Compostela, as consequence of this PhD work [83]. We advanced that ILs have the potential to constitute an attractive alternative to the surfactants commonly used in EOR processes, with important advantages over the traditional surfactant or polymer flooding methods, as for example: ILs are often liquid (and essentially non-volatile) over a wide range of temperatures. It is possible, in principle, to “design” an optimum IL for specific reservoir conditions (considering different fluid and rock properties). ILs have a relatively high viscosity, which is important to avoid the formation of digitations (phase degradation) resulting from unfavourable mobility ratios. The use of co-surfactants (usually long chain alcohols, which are volatile compounds and pose an environmental risk) may be avoided. This is a result of the strong cohesive forces in ILs, which enable them to form stable micelles without the need of additional chemicals. Moreover, Murillo-Hernández and Aburto [30] summarized several applications of ILs in the petroleum industry. One of these is the capacity of ILs to solubilize heavy compounds in oil as asphaltenes or paraffins. Several works on this topic show that the viscosity of crude oil decreases considerably after being treated for removal of asphaltenes and similar substances [78,84,85]. Thus, the ability of some ILs to solubilize those heavy compounds may also be advantageous if ILs are used in EOR, where such solubilisation into a distinct phase would contribute to levelling the viscosity of the aqueous and organic phases. After our work, first presented about ILs in EOR, only another recent publication was found on this application [86]. In that work, no phase diagrams are studied, systematic core flooding experiments are carried out and 1-dodecyl-3-methylimidazolium chloride is proposed as possible surfactant in EOR. Limited investigation on this field of application of ILs makes the study of the use of these salts in chemical flooding an interesting and challenging aim. Attractive surface properties of ILs and advantages that these solvents present over commonly used surfactants rend optimistic to obtaining promising results in EOR processes. 2.2.3. BIOMASS PROCESSING Since the last century, coal, natural gas, and petroleum are our main sources of energy and chemicals. It is estimated that they provide
a p c h a n t h a n H g r h u n a p l g o e f 2 . B i wh c o ( e a n c o gl y li g c a t h p o F i [8 gl m e n p proximate l h emicals [ 8 n d researc h h e depletio n n increase ence, ther e r eener ener g u ge amou n a tural biop o l atform, is o al. Other f fective feed . 2.3.1. Bio m i omass is h ere simpl e o mplex sub e .g. photos y n d transfo r o mplex m o l ycerides a n g nin, hemi c a rbon-neut r h e three d e o lymers is r i gure 2.12. S 9]. Cellul o l ucose mol e m ost abun d n zymaticall y l y the 86 % 8 7]. The de p h for their n of this so u of global e n e is an urg e g y sources. n ts of stor o lymers, a n a major s than bein g stock. m ass com p one of the e molecule s stances ac t y nthesis). T r m it into b o lecules s n d others. L c ellulose, a n r al renewa b e pend on t r epresente d S tructure of o se consist s e cules con n d ant bio p y fermen t 3 % of world p endence o greater ex p u rce of ene n ergy dem a e nt need to The answ e ed chemic a n d subsequ e tep towar d g renewabl e p osition most val u s such as H t ivated or c T rees trap a b iomass. T h s uch as L ignocellulo s n d extracti v b le resourc t he materi a d in Figure 2 the three m a s on a sem i n ected pri m olymer o n t ed to y 3 1 energy a n o n fossil f u p loitation i s e rgy in a n e a nd, remai n look for al t e r lies in bi o a l energy. ent enabli n d s sustain a e and abu n u able prod u H 2O, CO2, N c atalysed b y a pproxima t h is energy i lignin, c a se is mainl y v es, which c e [88]. Th e a l source. 2 .12 [89]. ajor biopoly m i -crystallin e m arily by gl y n Earth y ield etha n Theoretica l n d the 96 % u els remai n s needed. N e ar future, a n s as a m a t ernative r e o mass, whi An efficie n n g of fuels a a bility and n dant, bio m u cts from l N 2 are tra n y biochemi c t ely 1% of i s stored i n a rbohydrate y compose d represent a e relative p r T he struct u m ers of the p e polymeric y cosidic bo n and can n ol and l Consideratio n % of orga n n s enormo u N everthele s a ssociated t a jor concer n e newable a n c h possess e n t access t a nd materi a a worldwi d ass is a co l iving bein g sformed in t c al process e solar ener g n the form s, protei n d of cellulo s a n abunda n r oportions u re of the s p lant cell wa l structure n ds. It is t h readily b catalytica l ns n ic u s, s s, to n. n d e s to a ls d e o st g s n to e s g y of n s, s e, n t of s e l ls of h e b e l ly
Research Protocol 39 3. RESEARCH PROTOCOL For a better understanding of the research protocol, it will be divided in the main subsections of this thesis work: citrus essential oil deterpenation, EOR, and biomass processing. 3.1. CITRUS ESSENTIAL OIL DETERPENATION 3.1.1. CHEMICALS Citrus essential oil was simulated as a mixture of its two main compounds, the terpene limonene and the oxyterpene linalool.(R)-(+)- limonene was supplied by Sigma-Aldrich with a nominal purity of 97 wt% , and (±)-linalool was purchased from SAFC and its purity is ≥97 wt%. Both chemicals were used as received, without further purification. The IL 1-ethyl-3-methylimidazolium acetate ([C2mim][OAc]) was purchased from Iolitec with a nominal purity of >95wt% and 1buthyl-3-methylimidazolium acetate ([C4mim][OAc]) with a nominal purity of ≥95 wt% was purchased from Fluka. Both ILs were purified by stirring and heating at 70 ºC for 48 h under high vacuum (< 0.1 mbar) to remove any residual volatile compound present. The improved purity of the ILs was verified by proton and carbon nuclear magnetic resonance spectroscopy (1H NMR and 13C NMR, respectively). The rest of the ILs used as solvents for the deterpenation of citrus essential oil were synthesized in the laboratory as it is explained below. Synthesis of 1-ethyl-3-methylimidazolium bis(trifluoromehylsulfonyl)imide ([C2mim][NTf2]): This reaction was carried out in two stages: first an alkylation of 1-methylimidazolium with bromoethane, and a metathesis second stage with lithium bis(trifluoromethylsulfonyl)imide (Li[NTf2]) [105]. Figure 3.1. Alkylation reaction of 1-methylimidazolium with the haloalkane (R = akyl-group; X = halide) for the obtaining of 1-alkyl-3-methylimidazolium halide.
Phase Equilibria for Extraction Processes with Designer Solvents 40 1-methylimidazole (Aldrich, 99 wt%) was placed in a roundbotton flask with a reflux condenser attached. A 5 wt% molar excess of bromoethane (Aldrich, 99 wt%) was added dropwise under inert atmosphere at 40ºC during 48 h (Figure 3.1). A few milliliters of acetonitrile (Panreac, >99.9 wt%) were added to the reaction product, 1-ethyl-3-methylimidazolium bromide ([C2mim]Br), to avoid its solidification. Sample was then washed several times with ethylacetate (Aldrich, ≥99.5 wt%). The remaining molecular compounds were removed first with a rotary evaporator and, to complete the purification, high vacuum was used (<0.1 mbar) at 70 ºC at least 48 h. The purity of the intermediate product was checked with 1H NMR and 13C NMR. Figure 3.2. Methathesis reaction to synthesize 1-alkyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (R = alkyl-group). Once the purity of the [C2mim]Br was checked, the second stage of the reaction was carried out (Figure 3.2). The IL was dissolved in water and mixed with a 5wt% molar excess of a Li[NTf2] (Solvionic, >99 wt%) water solution. The solution was left reacting during 4 hours with vigorous stirring. The aqueous phase was separated in a decantation funnel and the reaction product (water immiscible) was dissolved in dichloromethane (Fluka, >99.9 wt%). Following, several washing steps were carried out with distillated water until there was no observance of a precipitate by adding silver nitrate to the residual aqueous phase, thus indicating the absence of halide in significant levels. Part of this water and dichloromethane were removed in a rotatory evaporator. Finally, a more exhaustive purification was done by putting the product in high vacuum (<0.1 mbar) during at least 48 h and 70 ºC. A slightly yellow liquid was obtained. The IL purity was verified with 1H NMR and 13C NMR. Synthesis of 1-hexyl-3-methylimidazolium bis(trifluoromehylsulfonyl)imide ([C6mim][NTf2]): The synthesis was carried out following the method above explained (Figures 3.1 and 3.2). Alkylation of the 1methylimidazolium was done with 1-chlorohexane (Aldrich, 99 wt%) and higher temperature (70 ºC). The reaction product was the IL
Research Protocol 41 [C6mim]Cl. The following steps of metathesis and washing were similar to that of [C2mim][NTf2]. Synthesis of 1-decyl-3-methylimidazolium bis(trifluoromehylsulfonyl)imide ([C10mim][NTf2]): The synthesis was analogous to that of the [C6mim][NTf2] (Figures 3.1 and 3.2), but using 1-chlorodecane (Aldrich, 99 wt%) as the initial haloalkane. The final product had a yellow coloration. Synthesis of 1-ethyl-3-methylimidazolium 2-(2methoxyethoxy)ethylsulfate ([C2mim][Me(OEt)2SO4]): The IL 1-ethyl-3-methylimidazolium 2-(2methoxyethoxy)ethylsulfate, [C2mim][Me(OEt)2SO4], was synthesized in the laboratory following the procedure proposed by Himmler et al. [106]. This method consists of two reaction steps: a quaternization of 1-ethylimidazolium with diethylsulfate followed by a transesterification. In the first step, a determinate quantity of 1-methylimidazolium (Sigma-Aldrich, 99 wt%) was placed in a round bottom flask with an excess amount of toluene (Sigma-Aldrich, 99.5 wt%) as solvent. Diethylsulfate (Fluka, ≥99 wt%) in an equimolar proportion was added dropwise under inert atmosphere. The reaction was carried out in an ice bath due to its exothermicity. As showed in Figure 3.3, the formed IL is the 1-ethyl-3-methylimidazolium ethylsulfate ([C2mim][EtSO4]). Temperature was kept under 10 ºC. After 2h, toluene and the reactants in excess were separated by decantation in a funnel. [C2mim][EtSO4] was washed with fresh toluene several times. The process was completed with removal of remaining starting materials, first in a rotary evaporator and later under high vacuum (<0.1 mbar) during 48 h at 70 ºC. The product purity was verified by 1H NMR and 13C NMR. Figure 3.3. Synthesis of 1-ethyl-3-methylimidazolium ethylsulfate ([C2mim][EtSO4]). A 1:2 molar relation of [C2mim][EtSO4] and di(ethylenglycol) methylether (Aldrich, ≥99.0 wt%) was mixed with a small quantity of methane sulfonic acid (Merck, ≥99.0 wt%), in a molar rate of 1:0.1. The reaction, Figure 3.4, was carried out at 70ºC, argon atmosphere
Phase Equilibria for Extraction Processes with Designer Solvents 42 and constant stirring during 5 hours. Afterwards, the ethanol formed as co-product was removed introducing the flask first in a rotatory evaporator from 1 to 2 hours at 70ºC, and later under high vacuum at 70ºC and 12 h. Afterwards, a 1:1 molar quantity of di(ethylenglycol) methylether (in relation to the starting quantity of [C2mim][EtSO4]) was added and left reacting during 30 minutes in the above mentioned conditions. The formed ethanol was removed as explained for the previous stage. The IL was washed 5 times with a twofold excess of diethyl ether (Sigma-Aldrich, 99 wt%, ACS reagent). To remove the residual volatile compounds left in the IL, the flask was introduced first in a rotatory evaporator and later under high vacuum (<0.1 mbar) at 70 ºC. 1H and 13C NMR spectra were obtained to check the purity of the product [C2mim][Me(OEt)2SO4]. Figure 3.4. Synthesis of 1-ethyl-3-methylimidazolium 2-(2methoxyethoxy)ethylsulfate ([C2mim][Me(OEt)2SO4]). Synthesis of 1-methylpyridinium methylsulfate ([C1py][MeSO4]): The IL [C1py][MeSO4] was synthesized by alkylation of pyridine (Riedel-de Haën, ≥99.5 wt%) with dimethylsulfate (Aldrich, ≥99 wt%) [107]. Careful dropwise addition of dimethylsulfate to mildly stirred pyridine, in a round-bottomed flask with reflux condenser attached, under argon atmosphere, was carried out in an ice bath to keep the temperature low due to the exothermic character of the reaction. After melting of the ice and natural warming of the water to room temperature, the flask was placed in an oil bath and heated up to 100 ºC for 8 h. The resulting mixture, with an IL appearance, was placed in a rotary evaporator for removal of unreacted materials or other volatile impurities. This removal of volatiles was completed by heating of the mixture at 70 ºC, with stirring, for 48 h under high vacuum (<0.1 mbar). The chemical identity and the purity of the desired products were assessed by 1H NMR and 13C NMR spectroscopy.
Table 3.1. CAS number, water content (H2O), and experimental and literature values for density (ρ) and refractive index (nD) of the pure components at 25 ºC and atmospheric pressure. H2O (ppm) ρ (g cm-3) nD Compound CAS number Exp. Exp. Lit. Exp. Lit. Limonene 5989-27-5 152 0.83868 0.8383[109] 1.47081 1.4701[109] Linalool 78-70-6 204 0.85683 0.85760[110] 1.45961 1.4601[111] [C2mim][OAc] 143314-17-4 1200 1.09902 1.0993[112] 1.50069 1.50091[112] [C4mim][OAc] 284049-75-8 2087 1.05270 1.0532[113] 1.49372 1.49381[113] [C2mim][Me(OEt)2SO4] 790663-77-3 289 1.23861 1.2367[114] 1.48113 ------ [C2mim][NTf2] 174899-82-2 298 1.51891 1.51845[115] 1.42298 1.4230[115]* [C6mim][NTf2] 382150-50-7 87 1.37209 1.37213[116] 1.42879 1.42958[116] [C10mim][NTf2] 433337-23-6 76 1.27840 1.2780[117] 1.43584 1.4356[117] [C1py][MeSO4] 37943-43-4 149 1.34812 1.34483[118] 1.51353 1.51296[118] [C2py][EtSO4] 2073-48-5 124 1.25283 1.2520[119] 1.50527 1.50525[119] * Interpolated value. Research Protocol 43
Phase Equilibria for Extraction Processes with Designer Solvents 44 Synthesis of 1-ethylpyridinium ethylsulfate ([C2py][EtSO4]): The same method used for the synthesis of [C1py][MeSO4] was applied. In this case the alkylation of the pyridine was done with diethylsulfate (Fluka, ≥99 wt%). The final product was a colourless liquid. 3.1.1.1. Physical properties of pure compounds Some representative physical properties, as density or refractive index, were determined for the used chemicals as a proof of their purity. It is well known that the behaviour of ILs is notably sensible to their water content [108], for this reason their water content was also measured by using the Karl-Fischer titration method. The ILs were stored in desiccators, in order to avoid moisture uptake. The CAS number and water content of the pure compounds used in this work are reported in Table 3.1. Their experimental densities and refractive indices at 25 ºC are also shown and were found to be in good agreement with values previously reported in literature [109-119], thus confirming the purity of the commercial and synthesized chemicals. 3.1.1.2. Recoverability of ILs The ILs used all along the experimental section were recovered in the laboratory. Due to their negligible vapour pressure they can be recovered, to obtain a required purity, by evaporation of the volatile compounds in the sample. For this purpose, a rotatory evaporator was used in a first place, followed by a high vacuum line (<0.1 mbar) at 70 ºC until the obtaining of the desired purity, which was checked by 1H and 13C NMR. The water content as well as their physical properties of the recovered ILs were also verified before their use. 3.1.2. EQUIPMENT Densities were measured in an Anton Paar DMA 5000 densimeter with viscosity correction and with self-control of temperature to 0.01 K using the Peltier effect. The measurement was repeated at least three times for each sample. The uncertainty in the density measurement is 10-5 g·cm-3. Refractive indices were measured in an ATAGO RX-5000 refractometer. The temperature was controlled with an uncertainty of ±0.02 K by means of a HetoTherm thermostat. The measurement was repeated at least three times for each sample. The uncertainty in the refractive index measurement is ±4 × 10-5. Water content of the chemicals used in this work was determined by Karl-Fischer titration with a MetrOhm 737 KF coulometer. The measurement was repeated at least three times for each sample. The uncertainty in the measurements is 5 μg when water content is lower than 103 μg and 0.5 wt% for higher mass of
wa H d e Th t h a n s t l a a m F i c o N M wa p r e x t h o f t h h e v a t h c h o f a n c o b y a ter. The p aën). Specia l e terminatio h ese cells w h ermostatic n uncertai n t irring the c a yer in whi c constant t e m easureme n i gure 3.5. E x T wo a o mposition s M R spectr o a s carried r ecision of 1 Colu m x isting rate h e studied s f these sub h e chromat o e ight of th e a riations i n h e width o f h romatogr a f several v a n d sample i o lumn sho u y means o f p roduct us e l ly design e n of tie-li n w ere conne c bath whic n ty of ±0. 0 c ontent of c h an immi s e mperature n t of the LL E x perimental a nalytical s of equili b o scopy. For out in a M 1 0-4 g. m n chroma t between t h s ubstances stances in o graph co n e peaks, a n the opera t f the peak a ms. This is a riables su c i njection s p u ld be take n f areas un d 4 e d for thes e e d jackete d n es of the c ted to a S h allows t h 0 5 ºC. T he the cell, s e s cible liqui d . Figure 3. 5 E . setup for de t techniques b rium pha s calibratio n ettler Tole d t ography q u h e height o in the ch r the sampl e n ditions ar e high prec i t ional con d during th only possi b c h as ove n p eed. Even m n into acc o d er the pea k 4 5 e titrations d glass ce l Liquid-Li q S electa (Ult r h e control o special d e e ttling, an d d mixture s p 5 shows th e t ermination o were u s s es: gas c h n of both m e d o AE 240 u antitative a r the anal y r omatogra m e (in most o e well-cont r ision can o d itions of t h e run tim e b le if there n temperat u m ore, a po s o unt. The a ks is inde p R e is Hydran a l ls were u q uid Equili b r aTerm 60 0 o f the tem p e sign allo w d taking sa m p lits, while e experime n o f the LLE d a s ed to de h romatogr a e thods, all analytic b a a nalysis is y tic area of m and the c of the case s r olled). Wh e o nly be ob t h e column e needed t o is an exha u u re, efflue n s sible overl o a nalysis of c p endent of e search Proto c a l (Riedeld u sed for t h b rium (LL E 0 0383 mod e p erature wi t w s vigorou s m ple of ea c keeping it a n tal setup f o a ta. t ermine t h a phy and 1 the weighi n a lance with based on t h the peaks c oncentrati o s , only wh e e n using t h t ained if t h barely affe o obtain t h u stive cont r n t flux spe e o ading of t h c ompositio n the effect o c ol d e h e E ). e l) t h s ly c h at f or h e 1 H n g a h e of o n e n h e h e e ct h e r ol e d h e n s o n
Phase Equilibria for Extraction Processes with Designer Solvents 46 the peak shape produced by the previously cited variables. From this point of view, areas are analytic variables more adequate than height of peaks to carry out a quantitative analysis. For this reason, in this work, the compositions of the different compounds in the sample were determined by the areas under the peaks. The equipment used was a HP 6890 Series gas chromatograph (Figure 3.6) using a thermal conductivity detector, an injector, a HP5 capillary column of 30 m × 0.32 mm × 0.25 μm (film thickness), and an empty pre-column (without stationary phase) to avoid the ILs, not retained in the liner, reach the column. Table 3.2 summarises the chromatograph operation conditions for this experimental work. These conditions are the results of an optimisation process to achieve the best analytical precision. Figure 3.6. HP 6890 Series gas chromatograph. Table 3.2. Gas chromatograph operation conditions for the analysis of the ternary systems limonene + linalool + IL. Column Type: HP-5 Flux: Constant flux of 1 mL/min Detector Type: TCD Temperature: 240 ºC Carrier gas He Injector Temperature: 250 ºC Split rate: 20/1 Injection volume: 1μL Oven Temperature program: 70 ºC (3.70 min) 60 ºC/min until 250 ºC (3.10 min)
Research Protocol 47 As alternative to gas chromatography, the 1H NMR spectroscopy combines simplicity and sufficient accuracy for practical purposes. Several satisfactory examples of the use of this analysis method in studying LLE of systems involving ILs can be found in literature [120-122]. Figure 3.7. Mercury 300 with robot sampler spectrometer. In a 1H NMR spectrum, a series of peaks over a baseline are recorded. These peaks are related to the hydrogen atoms of the chemical structures present in the sample, the hydrogens with equivalent ‘chemical vicinity’ accounting for the same peak. Thus, every chemical species in a mixture, with hydrogen atoms in its structure, will generate a series of peaks in the spectrum. In principle, the areas under these peaks are proportional to the number of hydrogens they are representing. This proportionality opens a door to the use of 1H NMR spectroscopy as a quantitative technique for the determination of compositions of mixtures. In the case of ternary systems, it is very important to select a peak not overlapped for each component, to make quantitative this analysis method. The NMR runs were performed at 25 ºC with a relaxation time of 20-30 s in a Varian Mercury 300, with robot sampler, of 7.04 T (300 MHz resonance for 1H). This equipment is shown in Figure 3.7. 3.1.3. PROCEDURE The experimental procedure for the development of this work can be divided in two sections: solubility tests and equilibrium determinations.
Phase Equilibria for Extraction Processes with Designer Solvents 54 Table 3.7. Compositions of the experimental tie-line ends, solute distribution ratio ( ) and selectivity (S) for the LLE of the ternary system limonene + linalool + [C10mim][NTf2] at 25 ºC. The molar fractions of limonene, linalool and IL are represented by x1, x2 and x3, respectively. IL-rich phase Organic-rich phase β S x1 x 2 x3 x 1 x 2 x 3 0.427 0.000 0.573 1.000 0.000 0.000 ----- ----- 0.435 0.033 0.532 0.978 0.022 0.000 1.50 3.37 0.442 0.108 0.450 0.919 0.081 0.000 1.33 2.77 0.440 0.203 0.357 0.812 0.188 0.000 1.08 1.99 0.400 0.325 0.275 0.634 0.363 0.003 0.90 1.42 0.347 0.423 0.230 0.502 0.495 0.003 0.85 1.24 0.283 0.505 0.212 0.392 0.602 0006 0.84 1.16 0.221 0.591 0.188 0.292 0.701 0.007 0.84 1.11 0.187 0.629 0.184 0.239 0.752 0.009 0.84 1.07 0.126 0.708 0.166 0.160 0.828 0.012 0.86 1.09 0.080 0.756 0.164 0.102 0.886 0.012 0.85 1.09 0.032 0.817 0.151 0.039 0.946 0.015 0.86 1.05 0.000 0.851 0.149 0.000 0.984 0.016 0.86 ----- Table 3.8. Compositions of the experimental tie-line ends, solute distribution ratio ( ) and selectivity (S) for the LLE of the ternary system limonene + linalool + [C1py][MeSO4] at 25 ºC. The molar fractions of limonene, linalool and IL are represented by x1, x2 and x3, respectively. IL-rich phase Organic-rich phase β S x1 x 2 x3 x 1 x 2 x 3 0.001 0.000 0.999 1.000 0.000 0.000 --- --- 0.000 0.009 0.991 0.878 0.122 0.000 0.07 ∞ 0.000 0.015 0.985 0.770 0.230 0.000 0.07 ∞ 0.000 0.025 0.975 0.629 0.369 0.002 0.07 ∞ 0.000 0.038 0.962 0.502 0.497 0.001 0.08 ∞ 0.000 0.051 0.949 0.304 0.695 0.001 0.07 ∞ 0.000 0.052 0.948 0.000 0.995 0.005 0.05 ∞
Research Protocol 55 Table 3.9. Compositions of the experimental tie-line ends, solute distribution ratio ( ) and selectivity (S) for the LLE of the ternary system limonene + linalool + [C2py][EtSO4] at 25 ºC. The molar fractions of limonene, linalool and IL are represented by x1, x2 and x3, respectively. IL-rich phase Organic-rich phase β S x1 x 2 x3 x 1 x 2 x 3 0.000 0.000 1.000 1.000 0.000 0.000 --- --- 0.006 0.018 0.976 0.942 0.058 0.000 0.31 48.7 0.009 0.053 0.938 0.872 0.128 0.000 0.41 40.1 0.020 0.097 0.883 0.818 0.181 0.001 0.54 21.9 0.022 0.108 0.870 0.787 0.213 0.000 0.51 18.1 0.031 0.152 0.817 0.740 0.258 0.002 0.59 14.1 0.035 0.181 0.784 0.707 0.290 0.003 0.62 12.6 0.091 0.303 0.606 0.646 0.347 0.007 0.87 6.20 0.215 0.442 0.343 0.556 0.430 0.014 1.03 2.66 0.317 0.487 0.196 0.480 0.477 0.043 1.02 1.55 Table 3.10. Compositions of the experimental tie-line ends, solute distribution ratio ( ) and selectivity (S) for the LLE of the ternary system limonene + linalool + [C2mim][OAc] at 25 ºC. The molar fractions of limonene, linalool and IL are represented by x1, x2 and x3, respectively. IL-rich phase Organic-rich phase β S x1 x 2 x3 x 1 x 2 x 3 0.044 0.000 0.956 1.000 0.000 0.000 --- --- 0.047 0.038 0.915 0.994 0.006 0.000 6.33 134 0.057 0.090 0.853 0.988 0.012 0.000 7.50 130 0.097 0.212 0.691 0.981 0.019 0.000 11.2 113 0.152 0.279 0.569 0.973 0.027 0.000 10.3 66.2 0.338 0.319 0.343 0.963 0.035 0.002 9.11 26.0 0.551 0.229 0.220 0.901 0.070 0.029 3.27 5.35
Phase Equilibria for Extraction Processes with Designer Solvents 56 Table 3.11. Compositions of the experimental tie-line ends, solute distribution ratio ( ) and selectivity (S) for the LLE of the ternary system limonene + linalool + [C4mim][OAc] at 25 ºC. The molar fractions of limonene, linalool and IL are represented by x1, x2 and x3, respectively. IL-rich phase Organic-rich phase β S x1 x 2 x3 x 1 x 2 x 3 0.067 0.000 0.933 1.000 0.000 0.000 --- --- 0.071 0.019 0.910 0.999 0.001 0.000 19.0 267 0.093 0.106 0.801 0.995 0.005 0.000 21.2 227 0.169 0.198 0.633 0.992 0.008 0.000 24.8 145 0.226 0.225 0.549 0.989 0.011 0.000 20.5 89.5 0.479 0.237 0.284 0.959 0.030 0.011 7.90 15.8 0.667 0.163 0.170 0.917 0.052 0.031 3.13 4.31 A first study was carried out to see how the temperature affects to the equilibrium. For this matter, the system limonene + linalool + [C2mim][Me(OEt)2SO4] was studied at three different temperatures: 25, 35 and 45 ºC [123]. Table 3.4 shows linalool solute distribution ratios and selectivities for this system. It can be noticed that β and S have very similar values at the studied temperatures, getting even slightly better results at 25 ºC. At this temperature the results obtained for β and S are higher, especially at low concentrations of linalool. This is a region of particular interest, since the initial concentration of oxyterpenes (linalool) in the essential oil lies within it. Due to the low influence of the temperature in the equilibrium data, and to the fact that an increase of this parameter would result in an increase in the operational costs, it was decided to carry out all the following studies at 25 ºC, which is a recommended thermodynamic value or reference. Figures 3.10 to 3.17 show the representation of the experimental and correlated LLE data in equilateral triangular diagrams. This kind of representation of ternary systems provides a good visualization of the shape and size of the immiscibility region, as well as the slopes of the tie-lines. The corresponding solute distribution ratios (β) and selectivities (S) for each system and their correlated values are also represented in these Figures. According to the classification proposed by Srensen et al. [124], the systems with [C2mim][NTf2], [C6mim][NTf2], [C10mim][NTf2] and [C1py][MeSO4] (Figures 3.11-3.14) correspond to Type II, since they present two immiscible pairs (limonene-IL, and linalool-IL), a completely miscible pair (limonene-linalool), and only one continuum immiscibility domain. In the same classification, the systems with [C2mim][Me(OEt)2SO4], [C2py][EtSO4], [C2mim][OAc] and [C4mim][OAc] (Figures 3.10, 3.15-3.17) correspond to Type I, since they contain only one immiscible pair (limonene-IL).
Research Protocol 57 Figure 3.10. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C2mim][Me(OEt)2SO4] at 25 ºC. Correlated values were obtained with the NRTL model (α = 0.1) with a previously fixed value of β. Figure 3.11. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C2mim][NTf2] at 25 ºC. The correlated values were obtained with the UNIQUAC model with a previously fixed value of β. x linalool (limonene-rich phase) 0.0 0.1 0.2 0.3 0.4 0.5 S 0 10 20 30 40 x linalool (limonene-rich phase) 0.00.10.20.30.40.5 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 [C 2 mim][Me(OEt) 2 SO 4 ] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 0.20 0.25 0.30 0.35 0.40 0.45 xlinalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 S 2 4 6 8 10 [C 2 mim][NTf 2 ] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Phase Equilibria for Extraction Processes with Designer Solvents 58 Figure 3.12. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C6mim][NTf2] at 25 ºC. The correlated values were obtained with the UNIQUAC model with a previously fixed value of β. Figure 3.13. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C10mim][NTf2] at 25 ºC. The correlated values were obtained with the UNIQUAC model with a previously fixed value of β. x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 0.5 0.6 0.7 0.8 0.9 1.0 1.1 xlinalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 S 1 2 3 4 5 6 7 [C 6 mim][NTf 2 ] 0.00.10.20.30.40.50.60.70.80.91.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 [C 10 mim][NTf 2 ] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 0.8 1.0 1.2 1.4 1.6 1.8 x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 S 1 2 3 4
Research Protocol 59 Figure 3.14. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines) and experimental solute distribution ratio (β) as a function of linalool in the organic-rich phase, and its corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C1py][MeSO4] at 25 ºC. The correlated values were obtained with the UNIQUAC model with a previously fixed value of β (NOTE: Selectivities for this system were not plotted because of their infinite values) Figure 3.15. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C2py][EtSO4] at 25 ºC. The correlated values were obtained with the UNIQUAC model with a previously fixed value of β. x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 [C 1 py][MeSO 4 ] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 [C 2 py][EtSO 4 ] 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 x linalool (limonene-rich phase) 0.0 0.1 0.2 0.3 0.4 0.5 0.2 0.4 0.6 0.8 1.0 1.2 xlinalool (limonene-rich phase) 0.0 0.1 0.2 0.3 0.4 0.5 S 0 10 20 30 40 50 60
Phase Equilibria for Extraction Processes with Designer Solvents 60 Figure 3.16. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C2mim][OAc] at 25 ºC. The correlated values were obtained with the NRTL model (α = 0.3) with a previously fixed value of β. Figure 3.17. Triangular diagram representation of the experimental tie-lines (, solid lines), and their corresponding correlated tie-lines (o, dashed lines), experimental solute distribution ratio (β), and selectivity (S) and their corresponding correlated values (dashed line), for the ternary system limonene + linalool + [C4mim][OAc] at 25 ºC. The correlated values were obtained with the NRTL model (α = 0.3) with a previously fixed value of β. [C 2 mim][OAc] 0.00.10.20.30.40.50.60.70.80.91.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 0.08 2 4 6 8 10 12 x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 0.08 S 0 20 40 60 80 100 120 140 160 180 [C 4 mim][OAc] 0.00.10.20.30.40.50.60.70.80.91.0 linalool 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 limonene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 0 5 10 15 20 25 30 x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 S 0 100 200 300
Research Protocol 61 From a thermodynamic point of view, high values (always bigger than one) of solute distribution ratios and selectivities favour the desired separation. A higher value of β means that a smaller amount of solvent is needed to treat a given feed, and a higher value of S implies the need of fewer separation stages to achieve a specific degree of separation. For a better comprehension of the behaviour of ILs as solvents for the extraction of linalool from citrus essential oil, a comparison based on the influence of the cation and anion which constitute the ILs is going to be carried out. The combined study of the [Cnmim][NTf2] (n = 2, 6 or 10) [125] allows an analysis of the influence of the variation of the alkyl substituent chain in the imidazoliumcation on the deterpenation, being one of the most popular tuneable structural features in ILs, besides the own nature of the cation and the anion. Solute distribution ratio and selectivity for the systems limonene + linalool + [Cnmim][NTf2] are plotted in Figure 3.18. Both parameters initially decrease, for all three systems, as the concentration of linalool increases, and they plateau at higher linalool concentrations (with perhaps the only exception of selectivity for the system [C2mim][NTf2], which monotonically decreases over the whole composition range). For a given molar fraction of linalool in the top phase, β increases as the length of the alkyl substituent chain of the IL increases, whereas the selectivity follows the opposite trend. The trends observed can be explained on the basis of segregated domains in the ILs. It is known that polar and nonpolar domains may exist in imidazolium ILs with alkyl substituent [126-128]. The presence of nonpolar domains increases as the length of the alkyl substituent increases. Therefore, as n increases in the [Cnmim][NTf2] ILs, the most hydrophobic limonene will find better accommodation in the larger nonpolar domains of the IL phase. Linalool will also partition more easily to the IL phase with larger hydrophobic domains, resulting in a greater solute distribution ratio, although the overall effect will be a reduction in the selectivity. With an increasing presence of limonene and linalool in the extract, the two phases will get more similar in composition, causing a reduction in the difference of the interactions occurring within one phase and within the other one; thus, solute distribution ratio and selectivity vary to a minor extent when the global concentration of limonene and linalool in the systems is high enough.
Phase Equilibria for Extraction Processes with Designer Solvents 62 Figure 3.18. Solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, for the systems limonene + linalool + [C2mim][NTf2] (), [C6mim][NTf2] () and [C10mim][NTf2] (), at 25 ºC. A similar study, influence of the variation of the alkyl substituent chain in the imidazolium cation on the deterpenation, can be done from results found for systems with [C2mim][OAc] and [C4mim][OAc]. Figure 3.19 shows a comparison of β and S for the ternary systems limonene + linalool + [Cnmim][OAc]. In both cases, solubilities initially increase reaching a maximum at a determined concentration of linalool, and the parameter diminishes at higher concentrations. Selectivities, similar to systems with [NTf2], decrease as the concentration of linalool increases, and they plateau at higher linalool concentrations. At low linalool concentrations (real interest), β and S increase as the length of the alkyl substituent chain of the IL increases, whereas at high linalool concentrations the parameters follow the opposite trend. The much higher solubilities and selectivities found with these ILs in comparison with [Cnmim][NTf2], and the different qualitative behaviour (especially at low concentrations of linalool), is due to the introduction of the strongly interacting [OAc]- responsible of hydrogen bonds with the hydroxyl group of linalool. Figure 3.19. Solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, for the systems limonene + linalool + [C2mim][OAc] () or [C4mim][OAc](), at 25 ºC. x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 0.08 0 5 10 15 20 25 x linalool (limonene-rich phase) 0.00 0.02 0.04 0.06 0.08 S 0 50 100 150 200 250 300 xlinalool (limonene-rich phase) 0.00.20.40.60.81.0 S 0 2 4 6 8 10 x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Research Protocol 63 x linalool (limonene-rich phase) 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 x linalool (limonene-rich phase) 0.0 0.1 0.2 0.3 0.4 0.5 S 0 10 20 30 40 50 Another interesting study is the influence of the cationic ring (pyridinium versus imidazolium) on the deterpenation ability. Figure 3.20 shows a comparison of both pyridinium ILs studied in this work [129] and the previously published 1-ethyl-3-methylimidazolium ethylsulfate ([C2mim][EtSO4]) [63]. The choice of the IL [C2mim][EtSO4] for this comparison is due to the structural analogy between [C2mim][EtSO4] and [C2py][EtSO4]. In the systems with these two ILs, there is an increase in β with an increase of linalool concentration. The values for the system with [C2mim][EtSO4] are higher, and tend to converge with those of the system with [C2py][EtSO4] as the plait points (almost positioned at the same concentration of linalool) are approached. It can also be highlighted, that selectivities in the systems with [C2py][EtSO4] and with [C2mim][EtSO4] follow similar trends; although it seems that, for low concentrations of linalool in the system, the performance of [C2mim][EtSO4] would be better. Therefore, since the selectivities are equivalent or better for the system with [C2mim][EtSO4], and its solute distribution ratios are also somewhat higher up to a practically coincident plait point, it can be stated that, from a thermodynamic perspective, this imidazolium IL would be a preferred solvent than its pyridinium analogous [C2py][EtSO4] for the deterpenation of citrus essential oil. In addition to the effect of the cationic core, a comparison between the performances of the two pyridinium ILs can be done. In Figure 3.20, it can be observed that the system with [C1py][MeSO4] is clearly leading to the lowest values of β, which remain practically constant as the concentration of linalool in the system is varied. It must be noted that the selectivities for the system with [C1py][MeSO4] have not been plotted in Figure 3.20 because of their infinite value. In spite of this very large or infinite selectivity, the use of [C1py][MeSO4] as extracting solvent in the deterpenation of citrus essential would be impractical, due to the large amounts of solvent that would be needed as a result of the low solute distribution ratios. Figure 3.20. Solute distribution ratio (β) and selectivity (S) as a function of linalool in the organic-rich phase, for the systems limonene + linalool + [C1py][MeSO4] (), [C2py][EtSO4] () or [C2mim][EtSO4] (), at 25 ºC (NOTE: Selectivities for the system with [C1py][MeSO4] were not plotted because of their infinite values).
Phase Equilibria for Extraction Processes with Designer Solvents 70 Table 3.13. LLE data correlation for limonene + linalool + IL ternary systems at 25 ºC. Residual function F, and mean error of the solute distribution ratio, ∆β, for each model defining or not the solute distribution ratio at infinite dilution, . IL Model β F ∆β [C2mim][Me(OEt)2EtSO4] NRTL ( = 0.1) ----- 1.6636 5.3 1.9 0.5429 2.9 UNIQUAC ----- 0.4212 3.2 0.83 0.4215 3.4 [C2mim][NTf2] NRTL ( = 0.2) ----- 0.2467 10.6 0.50 0.3058 1.9 UNIQUAC ----- 0.1450 1.9 0.50 0.1460 1.4 [C6mim][NTf2] NRTL ( = 0.2) ----- 0.6896 3.0 1.10 0.6918 2.4 UNIQUAC ----- 0.6045 6.9 1.12 0.5586 2.6 [C10mim][NTf2] NRTL ( = 0.3) ----- 0.7465 3.4 1.75 0.7465 3.5 UNIQUAC ----- 0.4827 25.2 2.00 0.6338 6.2 [C1py][MeSO4] NRTL ----- ----- ----- ----- ----- ----- UNIQUAC ----- 0.0731 9.9 0.05 0.0815 5.5 [C2py][EtSO4] NRTL ----- ----- ----- ----- ----- ----- UNIQUAC ----- 0.6336 9.7 0.33 1.0679 9.0 [C2mim][OAc] NRTL ( = 0.3) ----- 0.7924 84.7 3.57 0.8741 15.8 UNIQUAC ----- 1.9718 595 1.80 1.4989 16.4 [C4mim][OAc] NRTL ( = 0.3) ----- 0.5817 40.0 33.33 0.8335 26.7 UNIQUAC ----- 0.7994 273 40.00 0.8892 28.2
Research Protocol 71 Table 3.14. LLE data correlation for limonene + linalool + IL ternary systems at 25 ºC. Binary interaction parameters for selected model, NRTL (gij, gji) or UNIQUAC (∆uij, ∆uji),with the optimal value of the fixed solute distribution ratio at infinite dilution, . IL Model i-j gij or ∆uij/J·mol-1 gji or ∆uji/J·mol-1 [C2mim][Me(OEt)2SO4] NRTL 1-2 17825 -13306 = 0.1 1-3 39840 2816.2 2-3 15911 11347 [C2mim][NTf2] UNIQUAC 1-2 1791.9 -1131.8 1-3 3821.1 -428.54 2-3 3171.5 -1087.9 [C6mim][NTf2] UNIQUAC 1-2 1718.2 -1134.1 1-3 2636.6 -589.75 2-3 2813.0 -1310.7 [C10mim][NTf2] UNIQUAC 1-2 2793.3 -1771.6 1-3 4419.3 -1567.1 2-3 3088.2 -1686.5 [C1py][MeSO4] UNIQUAC 1-2 -1963.7 4404.9 1-3 2989.1 2115.4 2-3 3797.6 -635.83 [C2py][EtSO4] UNIQUAC 1-2 1178.4 -808.27 1-3 7873.7 -390.24 2-3 -1313.7 2459.8 [C2mim][OAc] NRTL 1-2 3260.0 4528.1 = 0.3 1-3 14732 7518.3 2-3 -3914.4 9349.6 [C4mim][OAc] NRTL 1-2 2509.2 6288.1 = 0.3 1-3 13894 5773.6 2-3 -5350.8 7329.3
Phase Equilibria for Extraction Processes with Designer Solvents 72 3.2. ENHANCED OIL RECOVERY 3.2.1. CHEMICALS Once inside the oil well, the reservoir fluid was simulated as a mixture of oil, water (or brine) and surfactant. In this work the oil was simulated as n-dodecane purchased from Merck with a nominal purity ≥ 99 wt%. In spite of its high purity, the small content of impurities drastically affects the interfacial tension measurements. Prior to its use, it was washed with fresh bidistilled water at least three times and passed several times through a column of alumina (Sigma-Aldrich, activated, basic, Brockman I) until its equilibrated mixture with water yielded and interfacial tension value in line with literature results. To study the effect of adding salt to the mixture, a 4 wt% solution of sodium chloride (Sigma-Aldrich, ≥ 99 wt%) in water was prepared. Any organic impurities contained in the salt were removed by heating up to 450 ºC for 12 h. Two ILs with surfactant characteristics were used in this work; trihexyl(tetradecyl)phosphonium chloride ([P6 6 6 14]Cl) and trihexyl(tetradecyl)phosphonium bis(trifluoromethylsulfonyl)imide ([P6 6 6 14][NTf2]).[P6 6 6 14]Cl was obtained from Cytec Industries Inc. with a purity of 96-97 wt%. Prior to use, this IL was dried under high vacuum (<0.1 mbar) at 70-80 ºC for a minimum of 24 h. Its purity was confirmed by 1H NMR and 13C NMR. Synthesis of trihexyl(tetradecyl)phosphonium bis(trifluoromehylsulfonyl)imide ([P6 6 6 14][NTf2]): The starting material was [P6 6 6 14]Cl obtained from Cytec and previously purified. [P6 6 6 14]Cl was dissolved in dichloromethane and afterwards added to an aqueous solution of Li[NTf2] [139]. The mixture was left stirring during 4 h. Extraction of the organic phase was carried out in a separating funnel. Then, the remaining dichloromethane and water were removed in a rotary evaporator and deep removal of volatiles under high vacuum was achieved at 80 ºC for more than 24 h. Its purity was confirmed by 1H NMR and 13C NMR. The final product was a colourless and viscous liquid product. 3.2.1.1. Physical properties of pure compounds The CAS number, water content, density, viscosity and surface tension of the pure compounds used in the present section of this work are reported in Table 3.15 and compared with available literature data as a proof of their purity [82,109,140-142]. Due to the high hygroscopicity of [P6 6 6 14]Cl, its water content was frequently checked and in case of obtaining higher values of water mass, put into the vacuum line. 3.2.1.2. Recoverability of ILs The ILs used all along this experimental section were recovered in the laboratory for reuse, as it was explained in the 3.1.1.2 section.
T a li t p u Compound Water n-Dodecane [P 6 6 6 14 ]Cl [P 6 6 6 14 ][NTf 2 3 . To t h n e u s c o c o 7 5 t o F i P a t h x we + Th r e M a ble 3.15. t erature val u u re compon e CA S Num b 7732-1 8 112-4 0 2588645 2 ] 4600920 . 2.2. EQUI P o determin e h e differen t e eded. For s ed with a o nnected t o o ntact of t h 5 ºC using a o within 0. 0 i gure 3.23. E T he co m a ckard H P h ermal con d 0.2 mm x ere used fo [P6 6 6 14]Cl h ese cond e spectively. M ettler Tole d CAS numb e u es for dens e nts at 25 ºC H2O (ppm) S b e r Exp 8 -5 ---- 0 -3 74 5 4-9 898 0 3-9 180 P MEN T e the equili b t phases o this reaso total liqui d o a IKA RW h e phases. a Selecta U l 0 2 ºC. E quilibrium m position o P 6890 Se r d uctivity d e 0.33 μm). D r the anal y + n-dodec a itions are For calibr a d o AE 240 7 e r, water c o ity (ρ), visc o and atmos p ρ (g· c Exp 0.99704 0.74527 0.89103 1.06550 b rium com p f the syst e n, speciall y d capacity 16 basic m Temperat u l traTerm 6 0 c ell for EOR o f each ph a r ies gas c tector and D i f ferent c h sis of the s a ne and w a summari s a tion, all t h analytic b 7 3 o ntent ( H2 O o sity (η) an d p heric press u c m -3 ) Lit 0.99705 [109] 0.74518 [109] 0.8899 [140] 0.8826 [141] 1.0661 [140] 1.0501 [141] p ositions a n e m, a larg e y designed of 75 mL m otor was u u re w as ke p 0 00383 the r determinati o a se was an a c hromatogr a a HP-FFAP h romatogra p tudied sys t a ter + [P6 6 6 s ed in T a he weighi n b alance wit h R e O ), and exp e d surface te n u re. η (mPa· s Exp L 0.904 0.8 9 1.356 1.3 7 2014 272 9 215 2 311.4 336 . n d physical e volume o equilibriu m (Figure 3. 2 u sed to ens p t constan r mostatic b o ns. a lysed usi n a ph, equi p capillary c p h operati o t ems in thi s 6 14][NTf2] + a bles 3.16 n g was car r h a precis i e search Proto c e rimental a n n sion(γ) of t h s ) γ L it Exp 9 0 [109] 72.0 7 8 [109] 24.9 9 .1 [140] 2 [82] 32.3 . 74 [140] 30.9 properties o f sample m cells we r 2 3). A stirr e u re a perfe t at 25ºC o b ath accura n g a Hewle t p ped with o lumn (25 m o n conditio n s work: wat e n-dodeca n and 3.1 r ied out in on of 10-4 c ol n d h e (mN·m -1 ) Lit 71.8 [109] 24.9 [109] 33.6 [140] 30.6 [142] 33.08 [141] 30.1 [142] of is re r er e ct or a te t t– a m n s er n e. 7, a g.
Phase Equilibria for Extraction Processes with Designer Solvents 74 When phases were practically pure n-dodecane or water, the IL content determined by chromatography was also contrasted by using ICP (optical) spectroscopy. Similarly, when water content in ndodecane was very low, this concentration was determined by KarlFischer titration with a MetrOhm 737 KF coulometer. Table 3.16. Gas chromatograph operation conditions for the analysis of the ternary system water + [P6 6 6 14]Cl + n-dodecane. Column Type: HP-FFAP Flux: Constant flux of 1 mL/min Detector Type: TCD Temperature: 230 ºC Carrier gas He Injector Temperature: 230 ºC Split rate: 200/1 Injection volume: 2 μL Oven Temperature program: 80 ºC (3 min) 100 ºC/min until 200 ºC (1 min) Table 3.17. Gas chromatograph operation conditions for the analysis of the ternary system water + [P6 6 6 14][NTf2] + n-dodecane. Column Type: HP-FFAP Flux: Constant flux of 1.1 mL/min Detector Type: TCD Temperature: 230 ºC Carrier gas He Injector Temperature: 230 ºC Split rate: 200/1 Injection volume: 1μL Oven Temperature program: 100 ºC (1.90 min) 100 ºC/min until 200 ºC (3.20 min) The existence of micro-emulsions was proved by optical & confocal microscopy, using a spectral confocal microscope (Leica model TCS SP2). As it was explained in the 3.1.2 section, densities were measured in an Anton Paar DMA 5000 densimeter (Figure 3.24) with viscosity correction and with self-control of temperature to 0.01 K using the Peltier effect. The measurement was repeated at least three
Research Protocol 75 times for each sample. The uncertainty in the density measurement is 10-5 g·cm-3. Kinematic viscosity (ν) was determined using a micro Ubbelohde viscosimeter technique. Previous to its use, the Newtonian character of the ILs was checked by using an Anton Paar Physica MCR 301 rheometer. Depending on the viscosity of the samples, two micro Ubbelohde viscosimeters (capillaries I and III) were used to measure the viscosity of the samples. The capillaries were calibrated and credited by the company and verified by measuring the viscosity of pure compounds with known viscosities. Flow time measurement was performed by Lauda Processor Viscosity system PVS1 (Figure 3.24) with a resolution of 0.01 s. The temperature of the viscosimeter was kept constant using a Lauda clear view thermostat D 20 KP with a through-flow cooler DLK 10. Viscosity measurements were repeated at least 3 times for each sample and were found to be reproducible to within ±0.04 s for times less than 100 s and ±0.4 s for longer times. Dynamic viscosity is given by the formula η = νρ = K(t-y)ρ (3.20) where η and ν are the dynamic and kinematic viscosity respectively, ρ is the density of the sample, K is the capillary constant provided by the manufacturer, t is the flow time and y is the kinetic energy correction used if necessary. The uncertainty for the dynamic viscosity is estimated to be ±0.5%. Figure 3.24. Densimeter, viscosimeter and tensiometer. The interfacial tension of the samples was measured using a Krüss K11 tensiometer (Figure 3.24). The Wilhelmy plate method was adopted in which a specially adapted platinum plate with cylindrical shape (Krüs accessory reference PL22; dimensions: 10 mm height x 20 mm base perimeter x 0.1 mm width) was used to carry out reliable measurements with lower amounts of sample than a conventional plate. For the measurement of the interfacial tension sufficient samples were withdrawn from each phase in the equilibrium cells and stored at 25 ºC or 75 ºC, depending on the
Phase Equilibria for Extraction Processes with Designer Solvents 76 experiment. Cylindrical open-top glass vessels with a diameter of 30 mm were used to place the lighter and denser phases during the different steps of the measuring procedure, which was performed according to the instructions of the manufacturer. The sample vessels were placed inside a jacketed oil bath through which refrigerating water from a Julabo F12 cryogenic thermostat was circulated. The actual temperature of the oil bath was determined by a built-in thermometer precise to within 0.1 ºC. Every interfacial tension data point reported in this work is the average of ten consecutive immersion measurements, after the first two immersion measurements were systematically discarded. The method described above has an estimated uncertainty of 0.1 mN·m-1. 3.2.3. PROCEDURE 3.2.3.1. Solubility tests [P6 6 6 14][Cl] and [P6 6 6 14][NTf2] ILs were selected as possible surface active agents due to their long hydrocarbon chains attached to the charged head-group of the cation. To be sure that they were capable to form triphasic systems, samples of water, IL and n-dodecane at several concentrations were prepared in 2 mL vials. Following, each vial was vigorously stirred with a vortex mixer and left to settle down. The appearance at several concentrations, for both ILs, of three phases, confirmed the interest of studying the use of these ILs for EOR and the need of obtaining equilibrium data. 3.2.3.2. Equilibrium The cloud point method cited previously (see section 3.1.3.2) was used to have an approximation of the shape of the solubility curve formed by the system water + IL + n-dodecane. LLE data for the studied systems were determined at 25 ºC and 75 ºC to analyse the effect of the temperature in the water + IL + n-dodecane phases diagram. To perform the measurements, a known composition of the three components (or two in the case of the binary tie-lines) within the biphasic or triphasic region were prepared by weight to ensure that every repeated experiment had the same composition (very important feature to carry out interfacial tension data comparisons). The mixture was added into a specially designed jacked equilibrium cell (Figure 3.23). The content of each cell was vigorously stirred for 2 hours maintaining its temperature constant at 25 ºC or 75 ºC using a Selecta UltraTerm 6000383 thermostatic bath. The cells were then left to settle down for 48 hours for systems with [P6 6 6 14]Cl and 72 h for systems involving the IL [P6 6 6 14][NTf2]. This time ensures a complete separation of the phases. Previously, several tests with different settle down times were performed to ensure that 48 and 72 hours were sufficient to reach equilibrium and complete phase
Research Protocol 77 separation. A sample of each phase was withdrawn to measure the composition and physical properties. For both ternary systems, water and n-dodecane compositions of phases in equilibrium were analysed by gas chromatography using the internal standard method. The IL fraction was determined by difference. For the system water + [P6 6 6 14]Cl + n-dodecane, isopropanol was used as solvent and also acts as internal standard. For the system water + [P6 6 6 14]NTf2 + n-dodecane, acetone was used as solvent and 1-decanol as internal standard. When by means of chromatography, IL compositions less than 0.1% were obtained, ICP (optical) spectroscopy was used to determine the exact composition. Similarly, when water content of organic phase was very low, this concentration was determined by Karl-Fischer titration with a MetrOhm 737 KF coulometer. Samples of the micro-emulsion intermediate phase were taken and dyed with toluidine blue (Schaurlau Chemie) and/or rhodamine B (Sigma-Aldrich). Small samples were placed on a standard 76x26 mm glass slide and covered with a 22x22x0.17 mm coverslip. Both transmitted and reflected light images were captured lightning the sample with an Argon laser at 488 nm. Besides, optical microscope images were obtained for selected samples. 3.2.3.3 Physical properties Density and viscosity were determined for homogeneous phases at 25 ºC and 75 ºC accordingly with temperature of equilibrium determination. Interfacial tension among phases in equilibrium was also determined. Samples were perfectly sealed to avoid losses by evaporation or pickup of moisture, and kept at the desired temperature of measurement to avoid phase separation. Special care was put in samples at 75 ºC. The withdrawn of these samples was done with hot syringes and keeping vials always at this temperature, placing a thermostatic bath next to the measurement apparatus. 3.2.4. RESULTS AND DISCUSSION Two ILs presenting amphiphilic character have been studied as surfactants for EOR by microemulsion flooding. Namely, [P6 6 6 14]Cl and [P6 6 6 14][NTf2]. To determine the suitability of these ILs as a potential replacement for conventional surface active agents in this application, the reservoir fluid was modeled as a mixture of water (or brine), surfactant (IL) and oil (n-dodecane). LLE data for the systems water + [P6 6 6 14]Cl + n-dodecane and water + [P6 6 6 14][NTf2] + n-dodecane at 25 ºC and 75 ºC are reported in Tables 3.18 to 3.21 (uncertainties in compositions determination are also presented) and represented in Figures 3.25 to 3.28. All these ternary systems are Winsor type III, with a triphasic region and two biphasic regions around it, as well as a monophasic domain. Although the Winsor type III diagrams do ‘nominally’
Phase Equilibria for Extraction Processes with Designer Solvents 78 contain a third biphasic region, in most cases it has a negligible size and cannot be detected, as it is the case in the systems studied here. The first biphasic region (marked with circles) implies the coexistence of an excess-water phase and another phase where the IL solubilises a relevant amount of n-dodecane. The other biphasic region (marked with squares) corresponds to an excess-oil phase in equilibrium with another phase where the IL solubilises a relevant amount of water. In the triphasic region (under the dashed lines), there is an IL-rich middle phase coexisting with an excess-water phase (consisting of almost pure water) and an excess n-dodecane phase (consisting of almost pure n-dodecane). This situation is ideal to achieve ultralow interfacial tension values and is favourable for EOR. The long alkyl side chains of the used phosphonium ILs classify them as hydrophobic. Nonetheless, there is a large influence of the anion on the water miscibility, which means that they can be substantially hygroscopic [143,144]. The chloride anion is more hydrophilic (through hydrogen bonding), resulting the water miscibility in [P6 6 6 14]Cl larger than in [P6 6 6 14]NTf2 at the studied temperatures. Solubility data here obtained at 25ºC for water-IL binaries are in good agreement with those found in the literature [144]. No comparative data have been found at 75ºC. On the other hand, the non-polar domains of these ILs have a large affinity for alkanes [145] through van der Waals interactions. The miscibility of the pair [P6 6 6 14]Cl + n-dodecane is slightly larger than for the pair [P6 6 6 14]NTf2 + n-dodecane, being the first completely miscible at 75ºC. Only solubility data for [P6 6 6 14]Cl + n-dodecane at 25ºC have been found in the literature [146], and are in agreement with the data presented here. For both ILs and both pairs, IL + water and IL + n-dodecane, miscibility increases with temperature, accordingly with the Upper Critical Solution Temperature found in the literature for these systems [144,146]. For the system with [P6 6 6 14]Cl, the maximum composition of IL in the water-rich phase (points near the left apex of triphasic system) was 2.1·10-5 at 25 ºC and 4.5·10-4 at 75 ºC, in mole fraction and obtained by ICP (optical) spectroscopy. Maximum water content in ndodecane-rich phases, determined by Karl-Fischer titration, was 4.4·10-4 at 25 ºC and 1.1·10-3 at 75 ºC in mole fraction. In the case of system with [P6 6 6 14]NTf2, for biphasic region 1 it is not possible to use the same nomenclature (upper and lower phases) that with [P6 6 6 14]Cl. This is due to the fact that depending on compositions of phase II, this in some cases is the upper phase but in others is the lower phase. Maximum mole fraction, obtained by ICP (optical) spectroscopy, of IL in water (points near the left apex of triphasic system) was 1.4·10-6 at 25 ºC and 2.5·10-5 at 75 ºC, and maximum IL content in n-dodecane (points near the right apex of triphasic system) at 25 ºC was 4.4·10-4. Maximum water content, determined by Karl-Fischer titration, in n-dodecane was 4.7 ·10-4 at 25 ºC and 9.9·10-4 at 75 ºC.
Research Protocol 79 Table 3.18. Experimental tie-lines for LLE of water (1) + [P6 6 6 14]Cl (2) + ndodecane (3) ternary system at 25 ºC and atmospheric pressure. Biphasic region 1 Upper Phase Lower Phase x1 x2 x3 x1 x2 x3 0.824 0.176 0.000 1.000 0.000 0.000 0.777 0.176 0.048 1.000 0.000 0.000 0.718 0.182 0.010 1.000 0.000 0.000 0.683 0.173 0.144 1.000 0.000 0.000 (x1)=0.006 (x2)=0.004 (x3)=0.003 (x1)=0.001 (x2)=0.001 (x3)=0.001 Biphasic region 2 Upper Phase Lower Phase x1 x2 x3 x1 x2 x3 0.000 0.005 0.995 0.000 0.259 0.741 0.000 0.006 0.994 0.043 0.267 0.690 0.000 0.003 0.997 0.083 0.275 0.642 0.000 0.004 0.996 0.148 0.279 0.573 0.000 0.003 0.997 0.195 0.270 0.535 0.000 0.008 0.992 0.292 0.256 0.452 0.000 0.006 0.994 0.424 0.221 0.355 0.000 0.007 0.993 0.506 0.209 0.285 0.000 0.003 0.997 0.562 0.190 0.249 (x1)=0.001 (x2)=0.001 (x3)=0.001 (x1)=0.006 (x2)=0.004 (x3)=0.003 Triphasic region Upper Phase Medium Phase Lower Phase x1 x2 x3 x1 x2 x3 x1 x2 x3 0.000 0.005 0.995 0.654 0.158 0.188 1.000 0.000 0.000 (x1)=0.001 (x2)=0.001 (x3)=0.001 (x1)=0.004 (x2)=0.001 (x3)=0.003 (x1)=0.001 (x2)=0.001 (x3)=0.001 Figure 3.25. Experimental tie-lines for the water + [P6 6 6 14]Cl + n-dodecane ternary system at 25 ºC. Biphasic region 1 (), biphasic region 2 () and triphasic region ( and dashed lines). water 0.00.10.20.30.40.50.60.70.80.91.0 [P6 6 6 14]Cl 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 n-dodecane 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Phase Equilibria for Extraction Processes with Designer Solvents 86 Table 3.23. Physical properties of involved phases (x1 = water, x2 = [P6 6 6 14]Cl, x3 = n-dodecane): density ( ), viscosity (η) and interfacial tension ( ) at 25 ºC. Phase (g·cm-3) η (mPa·s) (mN·m-1) Experiment 1 x1 =0.890 , x2 = 0.000 , x3 = 0.110 Upper 0.74527 1.356 52.2 Lower 0.99704 0.904 Experiment 2 x1 = 0.970 , x2 = 0.030 , x3 = 0.000 Upper 0.90408 113.3 1.8* Lower 1.00092 0.908 Experiment 3 x1 = 0.950, x2 = 0.025 , x3 = 0.025 Upper 0.86244 49.34 1.4* Lower 1.00047 0.910 Experiment 4 x1 = 0.885, x2 = 0.025 , x3 = 0.090 Upper 0.74578 1.363 0.3 1.4* Middle 0.85795 44.99 Lower 0.99829 0.904 Experiment 5 x1 = 0.118, x2 = 0.088 , x3 = 0.794 Upper 0.74583 1.370 0.4 Lower 0.83619 77.73 Experiment 6 x1 = 0.000 , x2 = 0.100, x3 = 0.900 Upper 0.74605 1.368 0.5 Lower 0.81508 84.20 * Negative values are reported by tensiometer due to light phase is the wetting fluid Table 3.24. Physical properties of involved phases (x1 = brine, x2 = [P6 6 6 14]Cl , x3 = n-dodecane): density ( ), viscosity (η) and interfacial tension ( ) at 25 ºC. Phase (g·cm-3) η (mPa·s) (mN·m-1) Experiment 1 x1 =0.890 , x2 = 0.000 , x3 = 0.110 Upper 0.74526 1.353 53.2 Lower 1.02544 0.957 Experiment 2 x1 = 0.970 , x2 = 0.030 , x3 = 0.000 Upper 0.90356 122.2 2.4* Lower 1.03248 0.971 Experiment 3 x1 = 0.950, x2 = 0.025 , x3 = 0.025 Upper 0.86168 51.36 2.0* Lower 1.03186 0.968 Experiment 4 x1 = 0.885, x2 = 0.025 , x3 = 0.090 Upper 0.74590 1.365 0.5 1.7* Middle 0.85607 46.54 Lower 1.03228 0.978 Experiment 5 x1 = 0.118, x2 = 0.088 , x3 = 0.794 Upper 0.74598 1.368 0.6 Lower 0.91003 78.48 Experiment 6 x1 = 0.000 , x2 = 0.100, x3 = 0.900 Upper 0.74605 1.368 0.5 Lower 0.81508 84.20 * Negative values are reported by tensiometer due to light phase is the wetting fluid
T a x 2 te Experime n x 1 =0.890 Experime n x 1 = 0.970 Experime n x 1 = 0.950 Experime n x 1 = 0.885 Experime n x 1 = 0.118 * we a ble 3.25. = [P 6 6 6 14 ] C nsion ( ) at 7 n t 1 , x 2 = 0.000 n t 2 0 , x 2 = 0.030 n t 3 0 , x 2 = 0.025 n t 4 , x 2 = 0.025 n t 5 , x 2 = 0.088 Negative va e tting fluid Physical C l, x 3 = n-d 7 5 ºC. , x 3 = 0.110 , x 3 = 0.000 , x 3 = 0.025 , x 3 = 0.090 , x 3 = 0.794 lues are re p 8 properties odecane): d e Phas e Uppe Lowe r Uppe Lowe r Uppe Lowe r Uppe Midd l Lowe r Uppe Lowe r p orted by t e 8 7 Figur e samp l prope + [P 6 75ºC. of involve d e nsity ( ), v i e (g· c er 0.70 r 0.97 er 0.86 r 0.97 er 0.82 r 0.97 er 0.70 l e 0.81 r 0.97 er 0.70 r 0.75 e nsiometer d R e e 3.32. Loc a l es used r ties fo r the 6 6 14 ]Cl + n d phases ( i scosity (η) a c m -3 ) η ( 851 0. 6 4 98 0. 3 515 22 4 96 0. 3 689 10 4 91 0. 3 9 29 0. 6 307 6. 9 4 79 0. 3 9 21 0. 6 9 15 5. 4 d ue to light e search Proto c a lisation of t h to meas u system wa t n -dodecane ( x 1 = wat e a nd interfac i m Pa·s) 6 60 4 3 89 .69 1 3 88 .33 1 3 85 6 76 0 0 9 59 3 85 6 81 0 4 12 phase is t h c ol h e u re t er at e r, i al (mN·m -1 ) 4 9.6 1.5* 1.2* 0 .1 0 .4* 0 .1 h e
Exp e x 1 = 0 Exp e x 1 = Exp e x 1 = Exp e x 1 = Exp e x 1 = Phase Eq u Table 3 x 2 = [P 6 6 tension ( e riment 1 0 .890 , x 2 = 0 e riment 2 0.970 , x 2 = e riment 3 0.950, x 2 = 0 e riment 4 0.885, x 2 = 0 e riment 5 0.118, x 2 = 0 * Negat i w etting f u ilibria for Ex 3 .26. Phy s 6 6 14 ]Cl , x 3 ( ) at 75 ºC. 0 .000 , x 3 = 0 0.030 , x 3 = 0 .025 , x 3 = 0 0 .025 , x 3 = 0 0 .088 , x 3 = 0 i ve values a f luid traction Proc e s ical prope r = n-dodeca 0 .110 0.000 0 .025 0 .090 0 .794 a re reported e sses with Des 88 r ties of in n e): density Phase Upper Lower Upper Lower Upper Lower Upper Middle Lower Upper Lower by tensiom s igner Solvent s n volved ph a ( ), viscosit y (g·cm -3 ) 0.70851 1.00971 0.87932 1.00969 0.82634 1.00975 0.70951 0.80199 1.00969 0.70970 0.77313 m eter due to Figure 3 the samp l propertie s w ater + dodecane s a ses (x 1 = y (η) and in t η (mPa·s) 0.660 0.447 21.37 0.445 8.816 0.449 0.683 6.303 0.439 0.687 5.547 light phas e .33. Localis a l es used to m s for the [P 6 6 6 14 ][N T at 25 ºC. brine, t erfacial (mN· m 51.2 1.7* 1.3* 0.09 0.7* 0.1 e is the a tion of m easure system Tf 2 ]+ nm -1 )
Research Protocol 89 Table 3.27. Physical properties of involved phases (x1 = water, x2 = [P6 6 6 14][NTf2], x3 = n-dodecane): density ( ), viscosity (η) and interfacial tension ( ) 25 ºC. Phase (g·cm-3) η (mPa·s) (mN·m-1) Experiment 1 x1 =0.890 , x2 = 0.000 , x3 = 0.110 Upper 0.74527 1.356 52.2 Lower 0.99704 0.904 Experiment 2 x1 = 0.970 , x2 = 0.030 , x3 =0.000 Upper 0.99707 0.900 27.1 Lower 1.06386 293.6 Experiment 3 x1 = 0.950 , x2 = 0.025 , x3 = 0.025 Upper 0.99709 0.897 25.3 Lower 1.00815 104.0 Experiment 4 x1 = 0.885 , x2 = 0.025 , x3 = 0.090 Upper 0.74537 1.354 0.6 23.5 Middle 0.94806 44.49 Lower 0.99708 0.907 Experiment 5 x1 = 0.000, x2 = 0.100 , x3 = 0.900 Upper 0.74530 1.355 0.9 Lower 0.94703 43.78 Table 3.28. Physical properties of involved phases (x1 = brine, x2 = [P6 6 6 14][NTf2], x3 = n-dodecane): density ( ), viscosity (η) and interfacial tension ( ) at 25 ºC. Phase (g·cm-3) η (mPa·s) (mN·m-1) Experiment 1 x1 =0.890 , x2 = 0.000 , x3 = 0.110 Upper 0.74526 1.353 53.2 Lower 1.02544 0.957 Experiment 2 x1 = 0.970 , x2 = 0.030 , x3 =0.000 Upper 1.02548 0.954 26.2 Lower 1.06464 301.9 Experiment 3 x1 = 0.950 , x2 = 0.025 , x3 = 0.025 Upper 1.00667 103.5 24.9 Lower 1.02548 0.958 Experiment 4 x1 = 0.885 , x2 = 0.025 , x3 = 0.090 Upper 0.74525 1.355 0.5 23.1 Middle 0.94567 43.67 Lower 1.02542 0.957 Experiment 5 x1 = 0.000, x2 = 0.100 , x3 = 0.900 Upper 0.74528 1.351 0.9 Lower 0.94703 43.78
Exp e x 1 = 0 Exp e x 1 = Exp e x 1 = Exp e x 1 = Exp e x 1 = Phase Eq u Table 3 [P 6 6 6 14 ] [ tension ( e riment 1 0 .890 , x 2 = 0 e riment 2 0.970 , x 2 = e riment 3 0.950 , x 2 = e riment 4 0.885 , x 2 = e riment 5 0.000, x 2 = 0 u ilibria for Ex 3 .29. Physi c [NT f 2 ], x 3 = ( ) at 75 ºC. 0 .000 , x 3 = 0 0.030 , x 3 = 0 0.025 , x 3 = 0.025 , x 3 = 0 .100 , x 3 = 0 traction Proc e c al properti e n-dodecan 0 .110 0 .000 0.025 0.090 0 .900 e sses with Des 90 F s p + 7 e s of invol v e): density ( Phase Upper Lower Upper Lower Upper Lower Upper Middle Lower Upper Lower s igner Solvent s F igure 3.34 s amples u p roperties f o + [P 6 6 6 14 ][N T 7 5 ºC. v ed phases ( ), viscosit y (g·cm -3 ) 0.70851 0.97498 0.97501 1.02117 0.97498 0.98753 0.71028 0.87426 0.97503 0.71019 0.87212 s . Localisatio n u sed to m o r the syste m Tf 2 ] + n-dod e (x 1 = wate y (η) and in t η (mPa·s) 0.660 0.389 0.391 32.23 0.391 19.53 0.679 5.802 0.394 0.676 6.316 n of the m e a sure m water e cane at r , x 2 = t erfacial (mN· m 49.6 25.6 23.8 0.4 22.2 0.6 m -1 )
Research Protocol 91 Table 3.30. Physical properties of involved phases (x1 = brine, x2 = [P6 6 6 14][NTf2], x3 = n-dodecane): density ( ), viscosity (η) and interfacial tension ( ) at 75 ºC. Phase (g·cm-3) η (mPa·s) (mN·m-1) Experiment 1 x1 =0.890 , x2 = 0.000 , x3 = 0.110 Upper 0.70851 0.660 51.2 Lower 1.00971 0.447 Experiment 2 x1 = 0.970 , x2 = 0.030 , x3 =0.000 Upper 1.00979 0.443 25.0 Lower 1.02519 29.15 Experiment 3 x1 = 0.950 , x2 = 0.025 , x3 = 0.025 Upper 0.97225 15.28 23.7 Lower 1.00975 0.448 Experiment 4 x1 = 0.885 , x2 = 0.025 , x3 = 0.090 Upper 0.71135 0.673 0.3 22.1 Middle 0.87080 5.441 Lower 1.00980 0.447 Experiment 5 x1 = 0.000, x2 = 0.100 , x3 = 0.900 Upper 0.71019 0.676 0.6 Lower 0.87212 6.316 The high water-oil interfacial tension is responsible for the need of adding a surface active agent to reduce it and dissolve oil, sweeping trapped oil out of the rock pore spaces. The interfacial tension obtained for water + n-dodecane system at 25 and 75 ºC is in good agreement with bibliographic data [147,148]. Literature also shows that in presence of NaCl, the interfacial tension of this binary increases [149]. According to Tables 3.23 and 3.25, interfacial tensions of 1.8 mN·m-1at 25 ºC and 1.5 at 75 ºC were found for the water + [P6 6 6 14]Cl binary system. A considerable deviation was found with data reported by Carrera et al. [150], who found an interfacial tension of 6.11 mN·m-1 at 20ºC. This is probably due to the different method used in the measurement. These authors used a pendant drop method, which requires knowledge of the density of the phases in equilibrium, in order to calculate the interfacial tension. In the case of the IL phase, it is well known that this physical property is substantially affected by the presence of impurities; in particular, of the water content, which is not stated by Carrera et al. in their work. No comparative data have been found in bibliography for systems with [P6 6 6 14]NTf2. The interfacial tension water/brine + n-dodecane is reduced by adding any of the ILs. The presence of [P6 6 6 14]Cl at 25 ºC reduces this interfacial tension from 52.2 mN·m-1 (53.2 with brine) down to 1.0 mN·m-1 (1.9 with brine) and at 75 ºC from 49.6 (51.2 with brine) down to 0.3 mN·m-1 (0.6 with brine). The comparative values are given for the interfacial tension measured among the aqueous and ndodecane phase of the triphasic system. In the case of [P6 6 6 14][NTf2] the reduction at 25 ºC is from 52.2 mN·m-1 (53.2 with brine) down to
Phase Equilibria for Extraction Processes with Designer Solvents 92 21.7 mN·m-1 (21.5 with brine) and at 75 ºC from 49.6 mN·m-1 (51.2 with brine) down to 20.1 mN·m-1 (19.9 with brine). The IL [P6 6 6 14]Cl shows a bigger surfactant character than [P6 6 6 14][NTf2]. This can be due to the more polar character of chloride anion against bis(trifluoromehylsulfonyl)imide. In the literature, the interfacial tensions between water and ndodecane in the triphasic systems are not usually measured, because these phases are not in contact at equilibrium (there is a microemulsion as middle phase). This middle phase is a continuous layer containing the surfactant together with the dissolved water and n-dodecane. For this reason, the discussion that follows is established according to phases in contact (as it is usual in bibliography). Figure 3.35 shows the variation with temperature of the microemulsion-oil and microemulsion-water interfacial tension in the triphasic system with water or brine when using IL [P6 6 6 14]Cl as surfactant. The presence of salt increases the value of this property. There is an important decrease of the microemulsion-water/brine interfacial tension with temperature, and a less significative decrease in the case of the microemulsion-oil. Figure 3.35. Effect of temperature on the interfacial tension of watermicroemulsion (■ water, □ brine), microemulsion-oil (▲ water, ∆ brine) of the triphasic system water/brine + [P6 6 6 14]Cl + n-dodecane. The lines are included for visual guidance. A similar analysis can be done through Figure 3.36 for water + [P6 6 6 14]NTf2 + n-dodecane ternary system. Similarly to the other system, there is a decrease of the microemulsion-water/brine and microemulsion-oil interfacial tensions with temperature, being more noticeable for the first. Nonetheless, in this case the influence of the presence of NaCl on the interfacial tensions is negligible. There is little information in bibliography about interactions among ILs and inorganic salts in water. The addition of an electrolyte to aqueous solutions of a surfactant diminishes the solvation of the hydrophilic T (ºC) 20 30 40 50 60 70 80 i (mN·m -1 ) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Research Protocol 93 part of the surfactant. This effect is most noticeable in the case of the most polar IL, the [P6 6 6 14]Cl, justifying the larger effect of the inorganic salt in their interfacial tensions. Figure 3.36. Effect of temperature on the interfacial tension of watermicroemulsion (■ water, □ brine), microemulsion-oil (▲ water, ∆ brine) of the triphasic system water/brine + [P6 6 6 14][NTf2] + n-dodecane. The lines are included for visual guidance. Finally, it should be noted that viscosity (a property that, as expected, decreases with temperature) of aqueous phases, increases when mixed with any of the ILs due to the higher viscosity of these compounds. Regarding to EOR applications, this can be useful to get a more homogeneous front and reduce fingering. These viscosity results, however, must be taken with particular care since it is hard to extrapolate viscous behaviour from the use of n-dodecane in the model systems to a real field case, with real (and much more viscous) oil. T (ºC) 20 30 40 50 60 70 80 i (mN·m -1 ) 0 5 10 15 20 25
Phase Equilibria for Extraction Processes with Designer Solvents 94 3.3. BIOMASS PROCESSING 3.3.1. CHEMICALS For the preparation of the DESs, different starting materials were used as hydrogen bond donors and acceptors. DL-malic acid (M) and glycerol (Gly) were purchased from Merck with a nominal purity of ≥ 99 wt%. Choline chloride (C), proline (P), alanine (A), oxalic acid (O), tetramethylammonium chloride (TMAC) and 2chloroethyltrimethylammonium chloride (CEAC) were obtained from Sigma-Aldrich and their purity was ≥ 98 wt%. Finally, lactic acid (L) was purchased from PURAC Biochem BV at pharmaceutical grade (≥ 99 wt%). All the materials were used without further purification, apart from the choline chloride and lactic acid which were previously dried under vacuum due to their high hygroscopicity. Wheat straw was produced in Spain under controlled conditions. Before usage, the material was milled using a Tomado stick blender set and sieved with a mesh size <1mm. The straw was analysed by the Klason method [151] and it was found to consist of 30 wt% cellulose, 32 wt% hemicellulose and 38 wt% lignin. Pine wood pellets were purchased from Energy Pellets Moerdijk (the Netherlands) and were milled using a Retsch ZM-200 ultracentrifugal mill. The wood powder was sieved with a mesh size < 1mm. Klason analysis [151] indicated the following composition for the wood: 39 wt% cellulose, 26 wt% hemicellulose and 35 wt% lignin. 3.3.2. EQUIPMENT All the weighing was done with a Mettler AX205 balance with a precision of ±0.02 mg. Proton Nuclear Magnetic Resonance (1H-NMR) spectra of DESs were obtained from a Bruker BZH 400/52 with Varian Mercury and the Fourier Transform Infra-Red (FT-IR) spectra on a PerkinElmer Spectrum One. To keep a constant temperature when mixing, a thermostatic oil bath with an IKA ETS-D5 temperature controller (uncertainty of ±0.1 ºC) was used. Phase separation was carried out by means of a MSE Mistral 3000 E-Bench model centrifuge. 3.3.3. PROCEDURE Several DESs which were previously proved as being useful for pure lignin solubilisation [38], and some other hitherto not synthetized, were tested for biomass processing. 3.3.3.1. DESs preparation The preparation of the DESs (Figure 3.37) was done by mixing one hydrogen bond donor with a hydrogen bond acceptor in the solid
s t l a s e t h p r F i r a T a p r H-bon d lactic a malic a oxalyc 3 . F o [ 2 0 . vi a t s t l a h i d o F o p r le t ate, follow e a ctic or oxa e lected te m h e differen t r epared to b i gure 3.37. a tio 2:1 of th e a ble 3.31. C r epare the c o d donor H - a cid c h c h te t te t 22a cid al a c h gl y p r acid c h . 3.3.2. Bio m o llowing th 2 7], differen . 5 g of bio m als. Sam p t mosphere. T he vi a t irring over n a ctic or oxa l i gher visco s o wn. Figu r o llowing, t h r omote sett l ss viscous e d by melti n lic acid an d m peratures d t precurso b e tested a s Preparation e starting m a C ombination o rrespondin g - bond accep t h oline chlori d h oline chlori d t ramethyla m t ramethyla m chloroethylt r chloroethylt r a nine h oline chlori d y cine r oline h oline chlori d m ass treat e same pr o t samples c m ass (pine w p les were a ls were pl a n ight (at le a l ic acid, a n s ity. After t h r e 3.38 s h h ey were l ing down o samples, t h 9 n g them at d at 130 º C d epend on rs. Table s biomass f r of lactic ac a terials. of different H g DESs. t or d e d e m monium ch l m monium ch l r imethylam m r imethylam m d e d e i n g o cedure as c ontaining 1 w ood or wh sealed to a ced in a si l a st 14 hou r n d at 85 ºC h is period o h ows a sc h centrifuge d o f the biom a h e supern a 9 5 65 ºC for t C for the m the stabili t 3.31 sho w r actionatio n c id-chloline l H -bond don o l oride l oride m onium chlo r m onium chlo r that repo r 1 0 g of the h eat straw) w avoid w a licon oil th e r s) at 60 º C for malic a o f time the s h eme of t h d (30 min u a ss suspen a tant (DES w R e t he mixtur e m alic acid m t y and me l w s the di ff n solvents. actate DES, o rs and H-bo Rati o 2:1 9:1 2:1 3:1 r ide 2:1 r ide 5:1 1:1 1:1 1:1 1:3 1:1 r ted by Sp r correspon d w ere prepa r a ter upta k e rmostatic C for sampl e a cid mixtu r s amples we h e biomas s u tes at 3 0 ded in the D w ith dissol v e search Proto c e s containi n m ixtures. T h l ting point ff erent DE S with a mol n d aceptors o D E LC 2 LC 9 LTMA C LTMA C LCEA C LCEA C MA MC MGl y MP OC r onsen et a d ing DES a n r ed in 15 m k e from t h b ath and l e e s containi n r es due to i r e left to co s treatme n 0 00 rpm) t D ES. For t h v ed bioma s c ol n g h e of S s l ar to E S 2 :1 9 :1 C 2:1 C 3:1 C 2:1 C 5:1 1:1 1:1 y 1:1 1:3 1:1 a l. n d m L h e e ft n g i ts o ol n t. to h e s s)
4 Conclusions
Conclusions 105 4. CONCLUSIONS In this thesis, the suitability of using designer solvents to develop greener and/or more efficient chemical processes was proved through thermodynamic studies. The main advantages of the proposed solvents, ILs and DESs, are: the liquid ranges exhibited, their very low vapour pressure (which avoids atmospheric contamination), their thermal and chemical stability, and their tuneability, which allows them to be optimized for a specific application as low toxicity and high biodegradability are sought simultaneously. The main conclusions drawn from the results obtained in the three different sections of this work are summarized below. Citrus essential oil deterpenation The feasibility of ILs to carry out the separation of the mixtures limonene and linalool (characteristic compounds of the citrus essential oil) by solvent extraction, was investigated through the determination of the LLE of ternary systems limonene + linalool + IL. The following ILs were specifically investigated: [C2mim][Me(OEt)2SO4], [C2mim][NTf2], [C6mim][NTf2], [C10mim][NTf2], [C1py][MeSO4], [C2py][EtSO4], [C2mim][OAc] and [C4mim][OAc]. The suitability of each IL to extract the linalool was evaluated by means of the solute distribution ratio (β) and selectivity (S). It is important to highlight that all the conclusions set in this section are established for β and S values at low concentrations of linalool, in accordance with the essential oil composition. From obtained limonene + linalool + [C2mim][Me(OEt)2SO4] LLE data, it can be concluded that temperature has a slight effect on the equilibrium, this means that the separation could be carried out at room temperature. For this reason the other equilibria were determined only at 25ºC. Several conclusions can be drawn from the structure of the studied ILs. The influence of the length of the alkyl substituent in a 1-alkyl-3-methylimidazolium cation depends on the type of anion. In the case of the [NTf2] anion, β increases as the length of the alkyl substituent increases whereas S follows the opposite trend. In the case of the [OAc] anion, both parameters increase with the length of the alkyl substituent. Comparing the results obtained for [C2py][EtSO4] with those of the previously studied [C2mim][EtSO4], it can be confirmed that imidazolium IL shows slightly better results for β than its analogous pyridinium, with S values being very similar.
Phase Equilibria for Extraction Processes with Designer Solvents 106 In any case, the type of anion seems to be the key factor to achieve the high desired values for these thermodynamic parameters. Conclusions can be established through the comparison of results obtained with ILs based in the [C2mim] cation. Several of them showed good selectivities, clearly higher than one, ranking as follows: [OMs]> [OAc]> [EtSO4]> [C2mim][Me(OEt)2SO4]> [NTf2]. Nonetheless, among these ILs the only solubilities clearly higher than one were found with the IL [C2mim][OAc]. This is due to the linalool susceptibility of establishing hydrogen bonding with the IL which has a strong capacity to act as a hydrogen bond acceptor. The final conclusion is that among all the tested ILs and those found in the bibliography, the IL [C4mim][OAc] is the most apt to carry out the deterpenation of citrus essential oil by LLE. Moreover, β and S values obtained with this IL are clearly higher to those obtained with solvents proposed in the bibliography of this application (diethylene glycol and a mixture ethanol and water with 85 wt% in ethanol). The practically null vapour pressure of the IL makes the recoverability of linalool easy by applying high vacuum and low temperatures. In addition, the proposed IL has moderately low viscosity and toxicity, in comparison to other ILs. This proposal avoids the need of a distillation process, which implies two stages, high energetic requirements, and the possible degradation of the oil. The NRTL and UNIQUAC equations were found to provide good correlations of the LLE, in spite of being a model not intended for systems involving electrolytes. Depending on the system, one of the two models provided slightly better results. Nevertheless no convergence was found for the NRTL model with the systems limonene + linalool + alkylpyridinium alkylsulfate IL. For this reason, UNIQUAC would be selected for the data treatment and its incorporation to simulation software. Enhanced oil recovery The possibility of using surfactant ILs to improve the surfactant flooding EOR methods was evaluated by simulating the reservoir fluid as a mixture of water/brine, n-dodecane (oil) and IL. LLE data and interfacial tension among equilibrium phases were determined. Concretely, the ILs [P6 6 6 14]Cl and [P6 6 6 14][NTf2] were investigated for this aim. Both ILs could be used in EOR as in presence of water/brine and oil, they form a Winsor type III system, with a triphasic region consisting of a microemulsion (middle phase) coexisting with excesswater and excess-n-dodecane phases. In this middle phase, the IL solubilises a relevant amount of water and n-dodecane. This is associated to a drastic reduction of the water-oil interfacial tension, as was shown by means of the measurement of this property. Therefore, during the EOR process, these ILs could eliminate the capillary forces that hold the oil in the porous medium, thus improving the oil recovery. The advantages of ILs being liquids at room temperature, and avoiding the addition of any co-surfactant
Conclusions 107 must be highlighted. This last point could be a particularly important improvement over traditional surfactants. The temperature and the presence of salt play an important role when looking for an optimal formulation water-oil-surfactant to be used in surfactant flooding EOR processes. There is a small influence of the temperature on the composition of the apex’s threephase tie triangle. Thus, with any of the ILs, the triphasic system is stable in a wide range of temperatures and in presence of salt. This stability greatly enhances the use of ILs to this application. Moreover, water or brine viscosity increases when mixed with the IL due to the high viscosity of this compound. Regarding EOR applications, a reduced difference among water and oil viscosities reduces the mobility ratio between the heavy oil and the water, resulting in a more homogeneous front and reducing fingering. This results in a more efficient oil recovery. Establishing a comparison among the two proposed ILs, the less amphiphilic character of the IL with the [NTf2] anion makes the microemulsion solubilise practically only n-dodecane, and therefore, the water-oil interfacial tension is slightly decreased. The IL [P6 6 6 14]Cl shows a bigger surfactant character (it reduces the wateroil interfacial tension by 98% at 25ºC and by 99% at 75ºC) and solubilises both water and oil. Concluding, due to the fact that these ILs can form a triphasic system, in a wide temperature range without the need of cosurfactant and in presence of salt, this can encourage further research on the use of ILs as surface active agents in EOR. Among the two studied ILs, the [P6 6 6 14]Cl showed the best properties for this purpose. This is an initial study. The capability of tuning the ILs must be used to increase the oil solubilisation and to get the extremely low interfacial tensions needed in this application, while simultaneously looking for an environmental friendly IL. Further investigation on the influence of pressure, adsorption of ILs in rocks, behaviour of real oil, and many others are needed before envisioning in situ research in the oilfield. Biomass processing The capability of using bio-renewable DESs as solvents for lignocellulosic biomass processing was proved through solubility studies of pine wood and wheat straw. Different hydrogen bond donors (lactic, malic and oxalic acid), and different hydrogen bond acceptors (glycerol, choline chloride, proline, alanine, tetramethylammonium chloride and 2chloroethyltrimethylammonium chloride) were combined to prepare the DESs. The solid-liquid extraction (carried out under mild conditions), allowed the obtaining of a supernatant (DESs-rich phase) of a brownreddish colour associated to the presence of lignin, as it was posteriorly confirmed by FT-IR spectra. Lignin was precipitated from
Phase Equilibria for Extraction Processes with Designer Solvents 108 the supernatant (mixed with ethanol for the most viscous DESs) by water addition. Nonetheless, the method must be improved for DESs which are immiscible with ethanol, because this fact produces the precipitation of the solvent which was detected with the filtered biomass. The tested DESs show high lignin solubility and very poor or negligible cellulose or hemicellulose solubility. Hence, the selectivity of the process is very high, a promising feature for delignification processes. The cellulose which remained in the biomass showed a noticeable change in its crystallinity in most cases. It can be concluded that the best DES to carry out this separation depends on the nature of the biomass (different structures of lignin). Thus, in the case of pine wood, best results were found with LCEAC 5:1, and in the case of wheat straw, best results were found with LTMAC 3:1. To sum up, tested DESs are promising solvents for biomass processing due to the high lignin solubility achieved over cellulose and hemicellulose. Nevertheless, further research needs to be carried out. Some of the parameters to be optimised are: the pretreatment process, new DESs components, the use of additives, other antisolvents, and different temperature or pressure.
LIST OF SYMBOLS
List of Symbols 111 LIST OF SYMBOLS a activity CAS Chemical Abstracts Service DES Deep Eutectic Solvent EOR Enhanced Oil Recovery F residual function Fa objective function (activities) Fb objective function (compositions) G Gibbs energy g binary energy interaction parameter in the NRTL model IL Ionic Liquid i generic species j generic species K capillary constant k generic species k kinetic energy correction LLE Liquid-Liquid Equilibrium LTTM Low Transition Temperature Mixtures l parameter in the UNIQUAC model M number of tie-lines m number of phases NMR Nuclear Magnetic Resonance NRTL Non Random Two-Liquid activity coefficient model n number of carbon atoms in an alkyl substituent chain of the imidazolium and pyridinium cation of an ionic liquid nD refractive index P pressure Pn binary interaction parameter Q empirical constant, in the expression of Fa q area molecular-structure constant in the UNIQUAC model
Phase Equilibria for Extraction Processes with Designer Solvents 118 Comprehensive Membrane Science and Engineering, vol. 2, Elsevier, Oxford, 2010. [21] E. Santos, J. Albo, C. I. Daniel, C. A. M. Portugal, J. G. Crespo, A. Irabien. J. Membr. Sci. 430 (2013) 56. [22] Y. S. Liu and G. B. Pan. Ionic Liquids for the Future Electrochemical Applications. In: A. Kokorin (Ed.). Ionic Liquids: Applications and perspectives, InTech, Croatia, 2011. [23] A. Lewandowski, A. Świderska-Mocek. J. Power Sources 194 (2009) 601. [24] J. D. Stenger-Smith, A. P. Chafin, Jr. Kline, F. Clare, G. S. Ostrom, R. L. Quintana. Ionic Liquids in Charge Storage Devices: Effect of Purification on Performance. T. Sato, S. Marukane, T. Morinaga. Ionic Liquids for the Electric Double Layer Capacitor Applications. In: S. T. Handy (Ed.). Applications of Ionic Liquids in Science and Technology, InTech, Croatia, 2011. [25] A. P. Abbott, K. J. McKenzie, Phys. Chem. Chem. Phys. 8 (2006) 4265. [26] R. P. Swatloski, S. K. Spear, J. D. Holbrey, R. D. Rogers. J. Am. Chem. Soc. 124 (2002) 4974. [27] J. van Spronsen, M. A. Tavares-Cardoso, G. J. Witkamp, W. de Jong, M. C. Kroon. Chem. Eng. Proces: Proces Intens. 50 (2011) 196. [28] A. V. Mudring, T. Alammar, T. Baecker, K. Richter. Nanoparticle Synthesis in Ionic Liquids, ACS Symposium Series vol. 1030, American Chemical Society , Washington D.C., 2009. [29] J.D. Scholten, M. H. G. Prechtl, J. Dupont. Formation of Nanoparticles assisted by Ionic Liquids. In: P. T. Anastas (Ed) Handbook of Green Chemistry, John Wiley and Sons, Weinheim, 2012. [30] J. A. Murillo-Hernández, J. Aburto. Current knoweledge and potential applications of ionic liquids in the petroleum industry. In: A. Kokorin (Ed.). Ionic liquids: applications and perspectives, Intech-Open, 2011. [31] M. Boukherissa, F. Mutelet, A. Modarressi, A. Dicko, D. Dafri, M. Rogalski. Energ. Fuel. 23 (2009) 2557. [32] Y. F. Hu, T. M. Guo. Langmuir 21 (2005) 8168. [33] R. C. B. Lemos, E. B. da Silva, A. dos Santos, R. C. L. Guimaraes, B. M. S. Ferreira, R. A. Guarnieri, C. Dariva, E. Franceschi, A. F. Santos, M. Fortuny. Energ. Fuel. 24 (2010) 4439. [34] D. Guzmám-Lucero, P. Flores, T. Rojo, R. Martínez-Palou. Energ. Fuel. 24 (2010) 3610. [35] P.S. Kulkarni, C.A.M. Afonso, Green Chem. 12 (2010) 1139. [36] M. Abai, M. P. Atkins, K. Y. Cheun, J. D. Holbrey, P. Nockemann, K. R. Seddon, G. Srinivasan, Y. Zou. Process for
References 119 removing metals from hydrocarbons by using ionic liquids. PTC Int. Appl. WO 2012/046057 A2. [37] C. Capello, U. Fischer, K. Hungerbühler. Green Chem. 9 (2007) 927. [38] M. Francisco, A. van den Bruinhorstand, C. Kroon, Green Chem. 14 (2012) 2153. [39] A. P. Abbott, G. Capper, D. L. Davies, R. K. Rasheed, V. Tambyrajah. Chem. Commun. (2003) 70. [40] A. P. Abbott, D. Boothby, G. Capper, D. L. Davies, R. K. Rasheed. J. Am. Chem. Soc. 126 (2004) 9142. [41] C. Ruß, B. König. Green Chem. 14 (2012) 2969. [42] Q. Zhang, K. Oliveira-Vigier, S. Royer, F. Jerome. Chem. Soc. Rev. 41 (2012) 7108. [43] M. Francisco, A. van den Bruinhorst, M. C. Kroon. Angew. Chem. Intl. Ed. 52 (2013) 3074. [44] J. Lawless. The Encyclopedia of Essential Oils, Harper Collins, London, 2002. [45] A. Verzera, A. Trozzi, G. Dugo, G. Di Bella, A. Cotroneo. Flavour Fragr. J. 19 (2004) 544. [46] A. Arce, A. Soto. Citrus Essential oils: Extraction and Deterpenation. In: N. Benkeblia, P. Tennant (Eds.). Tree and Forestry Science and Biotechnology, Special Issue 1, vol. 2, Global Sciences Books, London, 2009. [47] M. Kondo, M. Goto, A. Kodama, T. Hirose. Ind. Eng. Chem. Res. 39 (2000) 4745. [48] J. Owusu-Yaw, R. F.Matthews. J. Food Sci. 51 (1986) 1180. [49] G. R. Stuart, D. Lopes, J. V. Oliveira. J. Am. Oil Chem. Soc. 78 (2001) 1041. [50] A. Arce, A. Marchiaro, O. Rodríguez, A. Soto. Chem. Eng. J. 89 (2002) 223. [51] A. Arce, A. Machiaro, A. Soto. Fluid Phase Equilib. 211 (2003) 129. [52] A. Arce, A. Marchiaro, A. Soto. J. Sol. Chem. 33 (2004) 559. [53] A. Arce, A. Marchiaro, A. Soto. Fluid Phase Equilib. 226 (2004)121. [54] A. Arce, A. Marchiaro, J. M. Martínez-Ageitos, A. Soto. Can. J. Chem. Eng. 83 (2005) 366. [55] A. Arce, A. Marchiaro, O. Rodríguez, A. Soto. AIChE J. 52 (2006) 2089. [56] M. B. Gramajo de Oz, A. M. Cases, H. N. Sólimo. J. Chem. Thermodyn. 40 (2008) 1575. [57] L. M. Sevgili, S. Sahin, S. I. Kirbaslar. J. Chem. Eng. Data 53 (2008) 737.
Phase Equilibria for Extraction Processes with Designer Solvents 120 [58] D. J. Brose, M. B. Chidlaw, D. T. Friesen, E. D. LaChapelle, P. van Eikeren. Biotechnol. Prog. 11 (1995) 214. [59] K. Sakamoto, K. Fujii, A. Inoue, H. Kozuka, H. Ohta. Food Sci. Technol. Res. 9 (2003) 11. [60] S. Díaz, S. Espinosa, E. A. Brignole. J. Supercrit. Fluids 35 (2005) 49. [61] F. Gironi, M. Maschietti. Chem. Eng. Sci. 63 (2008) 651. [62] F. Benvenuti, F. Gironi, L. Lamberti. J. Supercrit. Fluids 20 (2001) 129. [63] A. Arce, A. Pobudkowska, O. Rodríguez, A. Soto. Chem. Eng. J. 133 (2007) 213. [64] E.C. Donaldson, G.V. Chilingarian, T.F. Yen. Enhanced oil recovery, II: processes and operations, Elsevier, Amsterdam 1989. [65] J. L. Salager. Recuperación mejorada del petróleo (Cuaderno FIRP S357-C), Universidad de los Andes, Mérida-Venezuela, 2005. [66] J. G. Speight. Enhanced oil recovery methods for heavy oil and tar sands, GULF Publishing Company, Houston, 2009. [67] L. W. Lake, R. L. Schmidt, P. B. Venuto, Oilfield Rev. 4 (1992) 55. [68] P. A. Winsor. Trans. Faraday Soc. 44 (1948) 376. [69] P. A. Winsor. Solvent Properties of Amphiphilic Compounds, Butherworth& Co., London, 1954. [70] J, L. Salager, J. C. Morgan, R. S. Schechter, W. H. Wade, E. Vasquez. SPE J. 19 (1979) 107. [71] J.L. Salager. Physicochemical Properties of Surfactant-oil-water Mixtures: Phase Behaviour, Microemulsion Formation and Interfacial Tension, Tesis PhD, Univ. Texas, USA, 1977. [72] V. Pino, M. Germán-Hernández, A. Martín-Pérez, J. L. Anderson. Sep. Sci. Technol. 47 (2012) 264. [73] I. R. Collins, M. J. Earle, S. P. Exton, N. V. Plechkova, K. R. Seddon. Ionic liquids and uses thereof, PTC Int. Appl. WO 2006/111712 A2. [74] J. Bowers, C. P. Butts, P. J. Martin, M. C. Vergara-Gutierrez. Langmuir 20 (2004) 2191. [75] M. Blesic, M. H. Marques, N. V. Plechkova, K. R. Seddon, L. P. Rebelo, A. Lopes. Green Chem. 9 (2007) 481. [76] M. Blesic, A. Lopes, E. Melo, Z. Petrovski, N. V. Plechkova, J. N. Canongia-Lopes, K. R. Seddon, L. P. Rebelo. J. Phys. Chem. B. 112 (2008) 8645. [77] M. Blesic, M. Swadzba-Kwasny, J. D. Holbrey, J. N. Canongia Lopes, K. R. Seddon, L. P. Rebelo. Phys. Chem. Chem. Phys. 11 (2009) 4260.
References 121 [78] Z. X. Fan, T. F. Wang, Y. H. He. J. Fuel Chem. Technol. 37 (2009) 690. [79] Y. Z. Shang, W. J. Song, N. Zhao, H. Li, H. I. Liu. You tian huaxue 27 (2010) 306. [80] G. Zhang, X. Chen, Y. Zhao, Y. Xie, H. Qiu. J. Phys. Chem. B. 111 (2007) 11708. [81] J. Liu, Y. Li, J. L. Chai, C. K. Kin, X. Y. Yu, Y. Xia. Tenside Surf. Det. 46 (2009) 306. [82] K. Anderson, H. Rodríguez, K. R. Seddon. Green Chem. 11 (2009) 780. [83] S. Lago, H. Rodríguez, A. Soto, A. Arce, M. Khoshkbarchi. RSC Advanc. 2 (2012) 9392. [84] M. Siskin, M. A. Francisco, R. M. Billimora. Upgrading of petroleum residues, bitumen, or heavy oils by separation of asphaltenes and resins using ionic liquids. PCT Int. Appl. WO 2008/124042 A1 20081016. [85] M. Siti-Qurratu’Aini, I. Mohd-Saaid. A review of ionic liquids used for upgrading of heavy oils, 3rd CUTSE International Conference, Miri, Sarawak, 2011. [86] A. Zeinolabedini-Hezave, S. Dorostkar, S. Ayatollahi, M. Nabipour, B. Hemmateenejad, Colloid Surface A 421 (2013) 63. [87] J.B. Binder, R. T. Raines. J. Am. Chem. Soc. 131 (2009)1979. [88] Z. Zhang, Z. K. Zhao. Bioresource Technol. 101 (2010) 1111. [89] N. Jiang, A. J Ragauskas. Perdeuterated Pyridinium Ionic Liquids for Direct Biomass Dissolution and Characterization, In: A. Kokorin (Ed.). Ionic liquids: applications and perspectives, Intech-Open, 2011. [90] Md. M. Hossain, L. Aldous. Aust. J. Chem. 65 (2012) 1465. [91] E. M. Rubin. Nature 454 (2008) 841. [92] N. Sun, H. Rodríguez, M. Rahman, R. D. Rogers. Chem. Commun. 47 (2011) 1405. [93] Y. Zheng, Z. Pan, R. Zhang. Int. J. Agric. Biol. Eng. 2 (2009) 51. [94] M. Mazza, D. A. Catana, C. Vaca-García, C. Cecutti. Cellulose 16 (2009) 207. [95] J. Holm, U. Lassi. Ionic Liquids in the Pretreatment of Lignocellulosic Biomass, In: A. Kokorin (Ed.). Ionic liquids: applications and perspectives, Intech-Open, 2011. [96] A. Pinkert, K. N. Marsh, S. S. Pang, M. P. Staiger. Chem. Rev. 109 (2009) 6712. [97] P. Maki-Arvela, I. Anugwom, P. Virtanen, R. Sjoeholm, J. P. Mikkola. Ind. Crops Prod. 32 (2010)175. [98] A. Stark. Energ. Environ. Sci. 4 (2011) 19. [99] H. Tadesse, R. Luque. Energy Environ. Sci. 4 (2011) 3913.
Phase Equilibria for Extraction Processes with Designer Solvents 122 [100] M. Mora-Pale, L. Meli, T. V. Doherty, R. J. Linhardt, J. S. Dordick. Biotechnol. Bioeng. 108 (2011) 1229. [101] N. Muhammad, Z. Man, M. A. Bustam-Khalil. Eur. J. Wood Prod. 70 (2012) 125. [102] H. Wang, G. Gurau, R. D. Rogers. Chem. Soc. Rev. 41 (2012) 1519. [103] A. Brandt, J. Graesvik, J. Hallett, T. Welton. Green Chem. 15 (2013) 550. [104] Q. P. Liu, X. D. Hou, N. Li, M. H. Zong. Green Chem. 14 (2012) 304. [105] P. Bonhôte, A. P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel. Inorg. Chem. 35 (1996)1168. [106] S. Himmler, S. Hormann, R. van Hal, P. S. Schulz, P. Wasserscheid. Green Chem. 8 (2006) 887. [107] A. Arce, E. Rodil, A. Soto. J. Chem. Eng. Data 51 (2006) 1453. [108] K. R. Seddon, A. Stark, M. J. Torres. Pure Appl. Chem. 72 (2000) 2275. [109] J. A. Riddick, W. B. Bunger y T. K. Sakano, Organic Solvents – Physical Properties and Methods of Purification, 4ª ed., Wiley, New York, 1986. [110] F. Comelli, S. Ottani. J. Chem. Eng. Data 47 (2002) 93. [111] R. Francesconi, C. Castellari. J. Chem. Eng. Data 46 (2001)1520. [112 ] M. G. Freire, A. R. Teles, M. A. Rocha, B. Schröder, C. M. Neves, P. J. Carvalho, D. V. Evtuguin, L. M. Santos, J. A. Coutinho. J. Chem. Eng. Data 56 (2011) 4813¬. [113] M. Tariq, P.A. Forte, M.F. Costa Gomes, J.N. Canongia Lopes, L.P. Rebelo. J. Chem. Thermodyn. 41 (2009) 790. [114] A. N. Soriano, B. T. Doma Jr., M. H. Li. J. Chem. Thermodyn. 40 (2008) 1654. [115] A. P. Fröba, H. Kremer, A. Leipertz. J. Phys. Chem. B 112 (2008) 12420. [116] A. Muhammad, M. I. Mutalib, C. D. Wilfred, T. Murugesan, A. Shafeeq. J. Chem. Thermodyn. 40 (2008) 1433. [117] K. Shimizu, M. Tariq, M. F. Costa Gomes, L. P. Rebelo, J. N. Canongia Lopes. J. Phys. Chem. B 114 (2010) 5831. [118] E. Gómez, N. Calvar, A. Domínguez, E. A. Macedo. J. Chem. Thermodyn. 42 (2010) 1324. [119] E. Gómez, I. Domínguez, B. González, A. Domínguez. J. Chem. Eng. Data 55 (2010) 5169. [120] A. Arce, M. Earle, H. Rodriguez, K. R. Seddon. J. Phys. Chem. B 111 (2007) 4732. [121] U. Domanska, Z. Zolek-Tryznowska, A. Pobudkowska. J. Chem. Eng. Data 54 (2009) 972.
References 123 [122] R. Anantharaj, T. Banerjee. Fluid Phase Equilibr. 312 (2011) 20. [123] M. Francisco, S. Lago, A. Soto, A. Arce. Fluid Phase Equilibr. 296 (2010) 149. [124] J. M. Sørensen, T. Magnussen, P. Rasmussen, A. Fredenslund. Fluid Phase Equilib. 2 (1979) 297. [125] S. Lago, H. Rodríguez, A. Soto, A. Arce. J. Chem. Eng. Data 56 (2011) 1273 [126] Y. A. Wang, G. A. Voth. J. Am. Chem. Soc. 127 (2005) 12192. [127] J. N. Canongia-Lopes, A. A. Pádua. J. Phys. Chem. B 110 (2006) 3330. [128] J. N. Canongia-Lopes, M. F. Costa-Gomes, A. A. Pádua. J. Phys. Chem. B 110 (2006) 16816. [129] S. Lago, H. Rodríguez, A. Soto, A. Arce. Sep. Sci. Technol. 47 (2012) 292. [130] http://www.merckmillipore.com/chemicals/1-ethyl-3methylimidazolium-methanesulfonate/MDA_CHEM490286/p_sLusHfETyAcAAAE852pCX8kx (Last accessed, July 2013). [131] H. Renon, J. M. Prausnitz. AIChE J. 14 (1968) 135. [132] D. S. Abrams, J. M. Prausnitz. AIChE J. 21 (1975) 116. [133] A. Bondi. Physical Properties of Molecular Crystals, Liquids and Glasses, Wiley, New York, 1968. [134] Z. Lei, J. Zhang, Q. Li, B. Chen. Ind. Eng. Chem. Res. 48 (2009) 2697. [135] R. Kato, J. Gmehling. J. Chem. Thermodyn. 37 (2005) 603. [136] T. Banerjee, M. K. Singh, R. K. Sahoo, A. Khanna, A. Fluid Phase Equilib. 234 (2005) 64. [137] R. S. Santiago, G. R. Santos, M. Aznar. Fluid Phase Equilib. 293 (2010) 66. [138] J. M. Sørensen, W. Arlt. Liquid-Liquid Equilibrium Data Collection, DECHEMA Chemistry Data Series, Frankfurt, 1980. [139] P. Bonhôte, A. P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel. Inorg. Chem. 37 (1998) 166. [140] C. M. Neves, P.J. Carvalho, M.G. Freire, J.A. Coutinho. J. Chem. Thermodyn. 43 (2011) 948-957. [141] P. Kilaru, G. A. Baker, P. Scovazzo. J. Chem. Eng. Data 52 (2007) 2306. [142] H.F. Almeida, J.A. Lopes-da-Silva, M.G. Freire, J.A. Countinho. J. Chem. Thermodyn. 57 (2013) 372. [143] M. G. Freire, C. M. Neves, P. J. Carvalho, R. L. Gardas, A. M. Fernandes, I. M. Marrucho, L. M. Santos, J. A. Coutinho. J. Phys. Chem. B. 111 (2007)13082.
Phase Equilibria for Extraction Processes with Designer Solvents 124 [144] M. G. Freire, P. J. Carvalho, R. L. Gardas, L. M. Santos, I. M. Marrucho, J. A. Coutinho. J. Chem. Eng. Data 53 (2008) 2378. [145] M. Blesic, J. N. Canongia-Lopes, M. F. Costa-Gomes, L. P. Rebelo. Phys. Chem. Chem. Phys. 12 (2010) 9685. [146] A. Makowska, A. Siporska, P. Oracz, J. Szydlowski. J. Chem. Eng. Data 55 (2010) 2829. [147] S. Zeppieri, J. Rodríguez, A. L. López-Ramos. J. Chem. Eng. Data 46 (2001) 1086. [148] D. M. Mitrinovic, A. M. Tikhonov, M. Li, Z. Huang, M. L. Schlossman. Phys. Ver. Lett. 85 (2000) 582. [149] R. Aveyard, S. M. Saleem. J. Chem. Soc. Farad. T 172 (1976) 1609. [150] G. V. Carrera, C. A. Afonso, L. C. Branco. J. Chem. Eng. Data 55 (2010) 609. [151] Y. C. Chung, A. Bakalinsky, M. H. Penner. Appl. Biochem. Biotechnol. 66 (1997) 249.
APENDIX A: PUBLICATIONS
Publications 127 APENDIX A: PUBLICATIONS 1.- M. Francisco, S. Lago, A. Soto and A. Arce Essential oil deterpenation by solvent extraction using 1-ethyl-3methylimidazolium 2-(2-methoxyethoxy) ethylsulphate ionic liquid. Fluid Phase Equilibria, vol. 296, 149-153 (2010) 2.- S. Lago, H. Rodríguez, A. Soto and A. Arce Deterpenation of Citrus Essential Oil by Liquid-Liquid Extraction with 1-Alkyl-3-methyl-imidazolium bis(trifluoromethylsulfonyl)amide ionic liquids Journal of Chemical and Engineering Data, vol. 56, 1273-1281 (2011) 3.- S. Lago, H. Rodríguez, A. Soto and A. Arce Alkylpyridinium Alkylsulfate Ionic Liquids as Solvents for the Deterpenation of Citrus Essential Oil Separation Science and Technology, vol. 47, 292-299 (2012) 4.- S. Lago, H. Rodríguez, A. Soto, A. Arce, and M. Khoshkbarchi. Study of phase behavior and transport properties of trihexyl(tetradecyl)phosphonium chloride ionic liquid for enhanced oil recovery. RSC Advances, vol.2, 9392-9397 (2012) 5.- M. Francisco, A. S. B. González, S. Lago García de Dios, W. Weggemans, M. C. Kroon Comparison of choline-based low transition temperature mixtures (LTTMs) with choline-based ionic liquids and salts: physical properties and vapour-liquid equilibria of mixtures containing water and ethanol. RSC Advances. Major revision July 2013 6.- A. S. B. González, M. Francisco, G. Jimeno, S. Lago García de Dios, M. C. Kroon Liquid – liquid equilibrium data for the systems {LTTM + benzene + hexane} and {LTTM + ethyl acetate + hexane} at different temperatures and atmospheric pressure Fluid Phase Equilibria. Sent to be published June 2013
Phase Equilibria for Extraction Processes with Designer Solvents 134 Los datos de equilibrio para el sistema limoneno + linalool + [C2mim][Me(OEt)2SO4] se determinaron a 25, 35 y 45 ºC, para ver el efecto que la temperatura ejerce en el equilibrio. Dado que los valores de β y S prácticamente no se vieron afectados por la temperatura, obteniéndose incluso ligeramente mejores resultados a temperatura ambiente, el resto de los estudios se realizó a 25 ºC. Se determinó el equilibrio con diferentes LIs con catión imidazólico dialquilsustituido ([Cnmim]), siendo estos radicales un metilo y una cadena lineal hidrocarbonada de longitud variable. Se encontró que la influencia del tamaño de dicha cadena en β y S depende del tipo de anión del LI. Así en los líquidos iónicos con anión [NTf2], al aumentar la longitud de la cadena β aumenta y S disminuye. Sin embargo, con el anión [OAc], al aumentar el tamaño de la cadena, ambos parámetros aumentan para bajas concentraciones de linalool que son las existentes en el aceite esencial previo a su desterpenación. Comparando los resultados obtenidos con los LIs [C2mim][EtSO4], previamente estudiado, y [C2py][EtSO4], se encontraron solubilidades superiores con el primero (de menor coste) y selectividades similares con ambos. El análisis de los datos obtenidos con diferentes LIs del tipo [C2mim]X, siendo X el anión ([Me(OEt)2SO4], [NTf2], [OAc], [OMs] y [EtSO4]), permite establecer que la influencia de este ión no es clara. Con todos ellos, se obtuvieron selectividades claramente superiores a la unidad, pero los coeficientes de distribución de soluto dieron valores favorables únicamente en el caso de los acetatos. La formación de puentes de hidrógeno entre el linalool y el LI resultó ser un parámetro clave en el aumento de esa solubilidad. Los valores de β y S más favorables (obtenidos para el [C4mim][OAc]) se compararon con los valores obtenidos en la bibliografía para los disolventes tradicionales propuestos para este proceso (dietilenglicol y la mezcla etanol-agua con un 85% en masa de etanol). Los resultados mostraron que el líquido iónico es indiscutiblemente el mejor extractor de linalool. Además, su prácticamente nula volatilidad permitiría su recuperación con relativa facilidad. Finalmente, los datos experimentales de equilibrio obtenidos para cada sistema limoneno + linalool + LI se correlacionaron mediante los modelos termodinámicos NRTL y UNIQUAC, obteniendo los correspondientes parámetros binarios de interacción energética, con el fin de modelar el comportamiento de las mezclas consideradas para el diseño de las unidades de separación. Aunque, en general, ambas ecuaciones proporcionaron resultados adecuados, obteniendo resultados ligeramente mejores con un modelo u otro dependiendo del sistema estudiado, no se consiguió convergencia para los sistemas con [Cnpy][RSO4] correlacionados con el modelo NRTL.
Resumen (Summary, in Spanish) 135 Extracción mejorada del petróleo Los métodos tradicionales de extracción primaria y secundaria del petróleo extraen aproximadamente la tercera parte del petróleo de un yacimiento. En los últimos años, la escasez de crudo ha llevado a la necesidad de recuperar más petróleo de yacimientos ya explotados. La extracción terciaria o mejorada del petróleo (“Enhanced Oil Recovery”), consiste en la utilización de medios que aumentan la movilidad del fluido en el pozo conjuntamente con los métodos de extracción secundaria (inyección de fluidos para generar un gradiente de presión), permitiendo una extracción mucho más eficaz en el yacimiento. Dentro de los métodos de extracción mejorada del petróleo, hay que citar: los métodos térmicos, los métodos de desplazamiento mediante fluido miscible (CO2 el más utilizado) y los métodos químicos. Estos últimos se consideran los más eficaces, sin embargo su coste es elevado, y es necesario mejorar su aplicabilidad puesto que estos métodos se ven afectados por múltiples parámetros como: tipo de crudo, temperatura, presión, porosidad del pozo, permeabilidad de los fluidos, naturaleza de los productos químicos, necesidad de aditivos, etc. El método de extracción terciaria mediante drenaje micelar es un método químico que se basa en la inyección de una disolución acuosa de baja concentración de surfactante dentro de la reserva de petróleo, con el fin de reducir la tensión interfacial existente entre el agua y el petróleo contenido en los poros de las rocas. De esta forma se mejora la movilidad del petróleo atrapado dentro de la reserva pudiendo ser arrastrado a la superficie mediante agua. El control de la movilidad es importante para que el proceso sea efectivo, por eso, habitualmente tras la inyección del sistema micelar, se introduce una fase acuosa de polímero, éste aumenta la viscosidad del agua y la eficiencia de barrido evitando la formación de digitaciones. Una aproximación válida para el estudio de sistemas de interés en la extracción del petróleo, a pesar de que éstas son mezclas complejas (agua, sales, surfactante, co-surfactante, componentes del petróleo…), es considerarlos como un sistema pseudo-ternario surfactante – agua – aceite. Así, a temperatura y presión constante, Winsor definió tres tipos de diagramas básicos (Winsor tipo I, tipo II y tipo III). El diagrama tipo III comprende una región trifásica rodeada por tres zonas bifásicas y una región monofásica. En lo que concierne a los sistemas cuya composición global se encuentra en la zona trifásica, se separan en tres fases en equilibrio: una fase acuosa y una fase aceitosa que contienen esencialmente agua y aceite, respectivamente, y también una fase media de densidad intermedia entre las fases anteriores. En el diagrama tipo III, las interacciones se equilibran y el surfactante está en lo que se llama la formulación óptima, ya que dicha situación físico-química corresponde a la existencia de una tensión interfacial extremadamente baja y por
Phase Equilibria for Extraction Processes with Designer Solvents 136 tanto a la eliminación casi total de las fuerzas capilares que atrapan el petróleo en el medio poroso. Experimental El sistema en el interior del yacimiento se simuló como una mezcla de surfactante (LI), aceite (n-dodecano) y agua/agua salada. La capacidad de algunos LIs surfactantes para mejorar los métodos existentes de recuperación mejorada del petróleo, se estudió a través de la determinación del equilibrio líquido-líquido de los sistemas ternarios agua + n-dodecano + LI. Se determinaron también la viscosidad y densidad de las distintas fases obtenidas en las diferentes regiones del equilibrio, así como las tensiones interfaciales existentes entre las fases. Los LIs seleccionados fueron [P6 6 6 14][Cl] y [P6 6 6 14][NTf2], por su carácter surfactante y por ser líquidos a temperatura ambiente. Para cada sistema agua + n-dodecano + LI, el estudio se inició con la determinación de las curvas de solubilidad a la temperatura de trabajo mediante el método del punto de niebla. Posteriormente, se determinó el equilibrio líquido-líquido (a 25 y 75 ºC), de manera similar al procedimiento explicado en el apartado anterior, pero en celdas termostatizadas especialmente diseñadas con una capacidad suficiente para abarcar los volúmenes necesarios de líquido para la determinación de las rectas de reparto, propiedades físicas y tensiones interfaciales. El tiempo de agitación fue de 2h, y el de reposo 48 h para los sistemas con [P6 6 6 14]Cl y 72 h para los sistemas que contienen el LI [P6 6 6 14][NTf2]. Las composiciones de equilibrio de agua y n-dodecano se determinaron mediante cromatografía de gases (HP 6890) utilizando el método del patrón interno. La composición del LI se estableció por diferencia. Cuando la composición del LI obtenido era menor que el 0.1 %, se utilizó el método de espectroscopia ICP (óptico) para determinar las composiciones exactas. De una forma similar, cuando el contenido en agua de la fase orgánica era muy pequeño, éste fue determinado mediante valoración (Karl-Fischer). La densidad y la viscosidad de las fases homogéneas se determinaron a 25 ºC y 75 ºC mediante un densímetro (Anton Paar DMA 5000) con corrección de viscosidad y autocontrol de temperatura y un viscosímetro capilar (Ubbelohde), respectivamente. La medida de las tensiones interfaciales entre las diferentes fases se llevó a cabo en un tensiómetro (Krüss K11) utilizando el método de la placa Wilhelmy. Resultados y discusión A partir de los datos de equilibrio se puede confirmar que en presencia de agua y n-dodecano, ambos LIs forman sistemas de Winsor Tipo III. Se encontró que la fase intermedia es una microemulsión donde el LI, en el caso del [P6 6 6 14]Cl solubiliza una cantidad considerable de n-dodecano y agua, y en el caso del [P6 6 6 14][NTf2] se solubiliza prácticamente sólo n-dodecano.
Resumen (Summary, in Spanish) 137 A partir de la medida de las tensiones interfaciales se confirmó que el sistema trifásico está asociado a una importante reducción de la tensión interfacial, mucho mayor en el caso del LI con el anión Cl, probablemente debido al mayor carácter polar de este anión. Se observó también un aumento de la viscosidad del agua/agua salada al mezclarla con el LI. Esto constituye una ventaja en esta aplicación, ya que el aumento de este parámetro reduce la diferencia de movilidad entre el agua y el aceite, viéndose disminuida la formación de digitaciones, y por tanto se mejora la eficacia del proceso. Se encontró también que con el aumento de temperatura se mantiene el sistema trifásico. La variación en la composición del vértice del triángulo que representa a dicho sistema es muy pequeña con esta variable. Así mismo, en presencia de sal se mantienen las tres fases encontradas. Fraccionamiento de biomasa Los principales recursos energéticos utilizados hoy en día (carbón, petróleo, gas natural, etc.) son limitados y, por lo tanto, pueden agotarse. Esta limitación ha generado, además de una investigación para el mejor aprovechamiento de los recursos existentes, un interés creciente por el desarrollo de nuevas tecnologías para la utilización de fuentes de energía renovables alternativas que, aunque actualmente son poco rentables, tienen la ventaja de ser poco contaminantes e inagotables. La biomasa es la energía solar convertida por la vegetación en materia orgánica. Esa energía puede ser recuperada por combustión directa o transformando la materia orgánica en otros combustibles. La biomasa lignocelulósica está constituida fundamentalmente por celulosa, hemicelulosa y lignina. Las interacciones intermoleculares de estos polímeros son muy difíciles de romper, consistiendo este hecho el mayor reto a la hora de obtener combustibles y productos químicos a partir de la biomasa. Los procesos convencionales de pretratamiento de la biomasa muestran conversiones relativamente altas, sin embargo la eficiencia energética y el impacto medioambiental necesitan ser mejorados. Una investigación puntera hoy en día se centra en el uso de LIs para el fraccionamiento de la biomasa en sus principales biopolímeros, evitando las condiciones tóxicas y peligrosas de los procesos actuales y convirtiéndolos en métodos más sostenibles medioambientalmente. Como alternativa, los DEs en general presentan costes inferiores, menor toxicidad y mayor biodegradabilidad. Sin embargo, los estudios encontrados en la bibliografía sobre el uso de estos disolventes para el fraccionamiento de la biomasa son muy escasos y en su mayoría teóricos.
Phase Equilibria for Extraction Processes with Designer Solvents 138 Experimental El trabajo experimental de esta sección se inicia con la preparación de varios DEs a partir de diferentes precursores. Para ello, se utilizaron distintos dadores de hidrógeno (ácido láctico, málico y oxálico) y aceptores de hidrógeno (glicerol, cloruro de colina, prolina, alanina, cloruro de tetrametilamonio y cloruro de 2cloroetiltrimetilamonio). La capacidad de estos DEs para fraccionar la biomasa se analizó a partir de diferentes ensayos de solubilidad de dos materiales lignocelulósicos: paja de trigo y madera de pino. La biomasa se sometió a un proceso de molienda manual y otro posterior de tamizado. Se mezcló con el disolvente y las muestras se agitaron vigorosamente durante por lo menos 14 h y a 60 ºC para las muestras con ácido láctico u oxálico, y a 85 ºC para las muestras con ácido málico (debido a su alta viscosidad). A continuación, las muestras se dejaron enfriar y posteriormente se centrifugaron para facilitar el precipitado de la biomasa suspendida en el DEs. El sobrenadante fue directamente retirado en los casos en los que fue posible, y mezclado con etanol en el caso de las muestras más viscosas. El sobrenadante fue utilizado para una posterior precipitación del biopolímero disuelto mediante el uso de agua como anti-disolvente. Resultados y discusión Al mezclar el material lignocelulósico con el DEs, el sobrenadante adquirió un color entre rojo y marrón, dependiendo del tipo de biomasa, asociado a la disolución de lignina. La separación de este biopolímero fue comprobada posteriormente mediante espectroscopía FT-IR. Se encontró una alta solubilidad de lignina en los disolventes sintetizados, al mismo tiempo que una solubilidad casi despreciable de celulosa y hemicelulosa. Estos resultados coinciden con estudios previos de la bibliografía sobre la solubilidad de estos biopolímeros en alguno de los DEs ensayados, investigación que dio origen al trabajo realizado en esta sección de la tesis. Por tanto, el proceso presenta una alta selectividad, mostrándose prometedor como proceso de deslignificación de biomasa. Se encontró que, una vez extraída la lignina, la celulosa que permanece en la biomasa sufre un cambio en su cristalinidad. Esto podría ser de interés, por ejemplo, para un posterior proceso de sacarificación. El mejor DES para llevar a cabo esta separación depende de la naturaleza de la biomasa. Sin embargo, en mezclas de ácido láctico y un aceptor de hidrógeno (cloruro de colina, cloruro de tetrametilamonio o cloruro de 2-cloroetiltrimetilamonio) la disolución de lignina aumenta con la proporción del ácido. El método utilizado debe ser optimizado, especialmente en el caso de los DEs más viscosos, puesto que la adición de etanol produce a veces problemas de inmiscibilidad con la consecuente precipitación del disolvente, según se detectó en la biomasa filtrada.
Resumen (Summary, in Spanish) 139 Conclusiones y trabajo futuro La investigación realizada en esta Tesis permite concluir que los disolventes de diseño, LIs y DEs, pueden sustituir a los disolventes orgánicos tradicionales dando lugar a procesos más sostenibles y/o de mayor eficacia. La desterpenación de aceites esenciales de cítrico puede llevarse a cabo mediante extracción líquido-líquido utilizando el LI [C4mim][OAc] como disolvente, puesto que las solubilidades y selectividades obtenidas son muy favorables y el LI sería relativamente fácil de recuperar. Se evitarían así los inconvenientes de la destilación fraccionada utilizada hoy en día (altos requerimientos energéticos y pérdida de calidad del producto). Ciertos LIs con carácter surfactante pueden ser empleados con el objetivo de mejorar las técnicas de drenaje micelar-polímero puesto que: - Permiten la formación de un sistema trifásico, asociado a una drástica reducción de la tensión interfacial agua-aceite, que liberaría el crudo atrapado en las rocas. Además, para la formación de este sistema, y a diferencia de los surfactantes tradicionales, no es necesaria la adición de co-surfactantes. - El sistema trifásico obtenido es estable ante cambios de temperatura y en presencia de sal. - La alta viscosidad de los LIs, al mezclarse con el agua, podrían evitar la formación de digitaciones permitiendo un avance homogéneo de la mezcla agua y crudo. De los dos LIs utilizados en este trabajo, los mejores resultados se obtuvieron con el [P6 6 6 14]Cl, reduciendo la tensión interfacial agua-n-dodecano un 98 % a 25 ºC y un 99% a 75 ºC. Estudios futuros deben basarse en la posibilidad de diseñar los LIs para conseguir una mayor solubilización del aceite y las extremadamente bajas tensiones interfaciales necesarias en esta aplicación, al mismo tiempo que se busca un surfactante medioambientalmente respetuoso. Otra investigación necesaria consiste en el estudio de la influencia de la presión, de la adsorción de los LIs en las rocas, comportamiento con el crudo real… En cuanto al pretratamiento de la biomasa con DEs, se concluye que es un proceso prometedor debido a la alta selectividad de los disolventes estudiados por la lignina en presencia de celulosa y hemicelulosa. Sin embargo, esta investigación está en sus comienzos y todavía se necesitan llevar a cabo numerosas optimizaciones del proceso para obtener resultados que sean la base de una implementación a mayor escala.