Traditional chemical processes show shortcomings caused by using volatile organic compounds as solvents during reactions and separations. Therefore, it is necessary to address this issue by moving toward more environmentally friendly processes. This is possible by using less toxic and hazardous solvents, such as ionic liquids and supercritical carbon dioxide (scCO2). Ionic liquids have attracted a lot of attention as potential "green" solvents to replace conventional organic solvents due to their unique properties. Ionic liquids are molten salts, which are liquid below 373 K. They consist of organic cations with organic or inorganic anions. Typically, ionic liquids possess an extremely low vapor pressure, a high thermal stability, and a wide liquid range. Ionic liquids show high solubility for both polar and apolar compounds. As a result of these properties, there are many reports where ionic liquids have been used as solvents for chemical reactions. Moreover, ionic liquids in combination with CO2 show numerous advantages. For instance, ionic liquids do not dissolve in CO2, but CO2 is highly soluble in ionic liquids. Moreover, it is shown before that CO2 is able to force two immiscible liquid phases to form one homogeneous phase as CO2 pressure increases. This interesting behavior, which is known as the miscibility switch phenomenon, is generally applicable for ternary systems containing ionic liquids, CO2 and organics. A new process set-up based on this phase behavior has been proposed, in which the reaction is carried out in a homogenous phase at high rate (no mass transfer limitations) by selection of a suitable pressure, temperature and CO2 concentration. After completion of the reaction, a phase separation is induced by changing the conditions. The product can be recovered with high purity from one of the two phases that is substantially free of ionic liquid. Two model reactions were studied to apply this new process concept. First, the epoxidation reaction of cinnamyl alcohol to 3-phenylglycidol was studied in the presence of ionic liquids as solvents. (2S,3S)-(-)-3-Phenylglycidol is an intermediate for a well-known potent active anti-inflammatory agent, Ibuprofen. In order to scale up the reaction in an ionic liquid/scCO2 miscibility switch system, a study was started to select the optimum ionic liquid in this case. The catalytic epoxidation reaction of cinnamyl alcohol in the presence of ionic liquids was optimized with respect to various parameters: (i) type and amount of oxidizing agent, (ii) type and amount of catalyst, (iii) type of ionic liquids (ranging from hydrophobic to hydrophilic), and (iv) temperature. Optimization of the conditions revealed that product stability is the key factor in determining the reaction conditions. Optimum conditions were obtained using the ionic liquid 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([bmim][Tf2N]) with 3 mol% vanadyl acetylacetonate as a catalyst and 1,5 equiv. of tert-butyl hydroperoxide as oxidant at 25 °C. In order to design the subsequent separation step using CO2 extraction, it is critical to have the knowledge on the phase behavior of the systems involved. Therefore, the high-pressure phase behavior of ternary systems containing epoxidation reaction compounds, ionic liquids and scCO2 were measured. Phase behavior experiments were carried out using a synthetic method in the Cailletet apparatus at five different concentrations of CO2, at temperatures and pressures up to 368 K and 12.1 MPa, respectively. Both ternary systems (cinnamyl alcohol + ionic liquid + CO2 and (2S,3S)-(-)-3-phenylglycidol + ionic liquid + CO2) only show one type of phase transition (liquid-vapor to liquid) at the highest pressures studied. A comparison between the two systems shows that both systems behaved similarly at low concentrations of CO2 (less than 40 mol%). However, by increasing the concentration of CO2 (more than 40 mol%) higher pressures are necessary to completely dissolve CO2 in the system containing cinnamyl alcohol compared to (2S,3S)-(-)-3-phenylglycidol. Moreover, a comparison between the binary system of [bmim][Tf2N] + CO2 and the studied ternary systems indicates that the addition of organic compounds decreases the solubility of CO2 in [bmim][Tf2N]. Finally, using data obtained during this work, the conditions for carrying out the epoxidation reaction in a homogeneous phase and extracting the product with scCO2 in the two-phase region are determined. The second model reaction was Friedel-Crafts acylation reaction of ferrocene to acetylferrocene. Ferrocene, which is solid up to 445 K, is a building block for wide range of applications from homogeneous catalysis and material science to biology and medicine. Acetylferrocene is the product of the acylation reaction of ferrocene and has applications as an intermediate in the production of functional groups, combustion catalysts for propellants, and in medicinal chemistry. Recently, a comprehensive investigation of the acylation of ferrocene into acetylferrocene has been conducted in the presence of ionic liquids as solvents instead of conventional organic solvents. The promising results showed that up to 100% conversion and yield could be reached using imidazolium-based ionic liquids, specially [bmim][Tf2N] as a solvent with scandium triflate (Sc(OTf)3) as a catalyst. In order to investigate the feasibility of the product extraction using scCO2, the solubilities of ferrocene and acetylferrocene in scCO2 were measured using an analytical method in a quasi-flow apparatus. High-performance liquid chromatography was applied through an online sampling procedure to determine the concentration of ferrocene and acetylferrocene in the scCO2 phase. The experiments were performed within a temperature range of 308 to 348 K and at pressures ranging from 7.7 to 24.4 MPa. The molar solubilities at the applied conditions range from 8.9 to 31.2 × 10-4 for ferrocene and 2.5 to 79.2 × 10-4 for acetylferrocene. The existence of a cross-over area for acetylferrocene is detected at a pressure of around 15 MPa and for ferrocene at a pressure of around 10 MPa. The comparison between the experimental solubility data shows that ferrocene is more soluble in scCO2 at lower pressures, while at higher pressures acetylferrocene was more soluble in scCO2. The reason for this behavior is a trade-off between the lower polarity of ferrocene (more dominant at lower pressures) and the molecular structure of acetylferrocene (more dominant at higher pressures). Results obtained in this work show that the solubility of the reaction product acetylferrocene in scCO2 was sufficiently high to use scCO2 extraction at high pressures to separate it from its reactant ferrocene in Friedel-Crafts acylation processes. Furthermore, to investigate the possibility of applying the miscibility switch phenomenon to perform Friedel-Crafts acylation reaction, the high-pressure phase behavior of the ternary system containing ferrocene or acetylferrocene, the ionic liquid [bmim][Tf2N] and CO2 were studied experimentally. The experiments were performed using a synthetic method in the Cailletet apparatus within a pressure range of 0.25 up to 10 MPa and in a temperature range of 278 up to 368 K. Five different concentrations of CO2 (10, 20, 31, 40 and 50 mol% of CO2) were investigated. While for the ternary systems containing ferrocene + ionic liquid + CO2 three phase transitions (SLV ¿ SL, SL ¿ L and LV ¿ L) were experimentally measured, for the ternary systems with acetylferrocene two different regions of liquid-vapor (LV) and liquid (L) were recognized. It is also shown that CO2 acts as a co-solvent in all measurements in the presence of ferrocene, even at high CO2 concentrations (50 mol%). Ferrocene is thus more soluble in the [bmim][Tf2N] + CO2 mixture than in pure [bmim][Tf2N]. Removing CO2 from the system by pressure release results in precipitation of the ferrocene, so that it can be recovered from the ionic liquid phase. This is contrary to many other ionic liquid + organic systems that show anti-solvency behavior in the presence of large amounts of CO2, where it is possible to recover the organic compound as precipitate by the addition of supercritical CO2. The solute effect on the phase behavior was studied by comparing the experimental results of the binary system [bmim][Tf2N] + CO2 with those of the ternary system acetylferrocene/ferrocene + [bmim][Tf2N] + CO2. It is shown that addition of an acetyl group to the ferrocene molecule dramatically changes the phase behavior of the binary system. Finally, the homogeneous liquid phase region was determined experimentally. This study indicates that performing the acylation reaction of ferrocene to acetylferrocene in the presence of [bmim][Tf2N] and CO2 in a homogeneous liquid phase is feasible. Finally, the benefits of applying new process set-up were investigated from economical and ecological point of view. For this purpose, the conventional production process is compared with the alternative process proposed in this study using ionic liquid/CO2 for both model reactions. From an ecological point of view, the ionic liquid/CO2 production process generates much less catalyst and solvent losses and consumes much less energy. From an economical point of view, there are capital expenditures associated with the purchase of new equipment for the ionic liquid/CO2 production process, but the savings in operational costs for both reactions are much higher, making the new process overall more attractive than the conventional one. For the market of 100 ton per year 3-phenylglycidol production, new process saving will be 3.10 million euros per year. For the production of 100 ton per year of acetylferrocene, it is estimated that using ionic liquid/CO2 process will save 0.94 million euros per year regarding variable and fixed costs for the production of acetylferrocene. In conclusion, even though the ionic liquid/CO2 production process needs high start-up investment but the amount of the money saved each year is higher than this investment, and therefore it is economically and ecologically feasible to replace the current production process.
|Qualification||Doctor of Philosophy|
|Award date||21 Jun 2013|
|Place of Publication||Delft|
|Publication status||Published - 2013|