Phase Equilibria in Ionic Liquid Facilitated Liquid-Liquid Extractions

Anand Bharti, Debashis Kundu, Dharamashi Rabari, Tamal Banerjee

• Publisher: CRC Press
• Year: 2017
• ISBN:
  9781315350066
• Edition: 1
• Topic: Technology & Engineering
• Subtopic: Environmental Management
• Index: Technology & Engineering

This book provides a comprehensive overview of ionic liquid based separation techniques. The glimpse of thermodynamic predictive models along with global optimization techniques will help readers understand the separation techniques at molecular and macroscopic levels. Experimental and characterization techniques are coupled with model based predictions so as to provide multicomponent data for the scientific community. The models will focus more on the a-priori based predictions which gives higher emphasis on hydrogen-bonded systems. Particle Swarm Optimization (PSO) technique will also eventually help the readers to apply optimization technique to an extraction process. The overriding goal of this work is to provide pathways for leading engineers and researchers toward a clear understanding and firm grasp of the phase equilibria of Ionic Liquid systems.

1

Introduction

Environment regulations and a strict legal framework motivate scientists and researchers to develop clean technologies for sustainable development. Researchers amplify their efforts to convert separation processes like extraction and distillation into clean technologies. The knowledge of thermodynamic properties like equilibrium data and activity coefficient at infinite dilution is required for adequate separation. Liquid–liquid equilibrium (LLE) data help in the selection of solvent for the selective separation of the solute component. In addition to this economic aspect, clean technology also considers the ecological aspect. Since the past few decades, eco-friendly solvents like ionic liquids (ILs) have been considered as a better replacement for conventional volatile solvents (Brennecke & Maginn, 2001; Welton, 1999). The rate of publication in the open literature has been rising at a greater rate for years. A total of 12,400 papers have been published since 2007 against 7,400 published between 1990 and March 2007. These data show the increase in new applications of ILs (Martins et al., 2014).

ILs are salts consisting of organic cations and inorganic or organic anions. Room temperature ionic liquids (RTILs) are liquids at room temperature. They have a bulky nonsymmetrical organic cation and an organic/inorganic anion. They are known as designer solvents due to the possibility of a large number of ILs (cation and anion combinations) which impart specific properties. These solvents are more attractive due to desirable special properties, such as wider liquid range, negligible vapor pressure at room temperature, tunable viscosity, low flammability, high heat capacity and favorable solvation properties for polar and nonpolar compounds (Anderson, Lin, Kuntzler, & Anseth, 2011; Mukherjee, Manna, Dinda, Ghosh, & Moulik, 2012; Neves, Carvalho, Freire, & Coutinho, 2011). These solvents are more stable at high temperatures and/or in the presence of chemicals and are suitable for the extraction of inorganic and organic compounds (Chapeaux, Simoni, Ronan, Stadtherr, & Brennecke, 2008; Garcia-Chavez, Garsia, Schuur, & de Haan, 2012; Simoni, Chapeaux, Brennecke, & Stadtherr, 2010; Wasserscheid & Keim, 2000; Welton, 1999). The physical properties of these solvents can be tuned by a suitable combination of cations and anions. Different types of RTILs are based on cations such as imidazolium, phosphonium, pyridinium, pyrrolidinium, ammonium and basionics. Salts are formed with anions possessing low nucleophilicity such as bis(trifluoromethylsulfonyl)imide, hexafluorophosphate, tetrafluoroborate, perfluoroalkylsulfonate, thiocynate, nitrate, acetate, ethyl sulphate and methane sulfonate. The applications of ILs cover many research domains like catalysis (Tao, Zhuang, Cui, & Xu, 2014; Xiao, Su, Yue, & Wu, 2014; Ying et al., 2015; Zhuo et al., 2015), nanotechnology (Beier, Andanson, Mallat, Krumeich, & Baiker, 2012), azeotropic separation (Cai, Zhao, Wang, Wang, & Xiao, 2015), liquid–liquid extraction (García, García, Larriba, Torrecilla, & Rodríguez, 2011; García, Larriba, García, Torrecilla, & Rodríguez, 2011), metal extraction (Papaiconomou, Vite, Goujon, Leveque, & Billard, 2012), aromatic–aliphatic separation (Manohar, Banerjee, & Mohanty, 2013; Shah, Anantharaj, Banerjee, & Yadav, 2013), membrane science (Santos, Albo, & Irabien, 2014) and gas absorption (Shiflett, Niehaus, & Yokozeki, 2010, 2011). Today ILs are not just confined to laboratories; they have some applications at industrial scale as well, for example, BASF (BASIL), IFP (Difasol), PetroChina (Ionikylation), Linde (hydraulic ionic liquid compressor), Pionics (batteries) (Berthod, Ruiz-Ángel, & Carda-Broch, 2008; Marsh, Boxall, & Lichtenthaler, 2004; Plechkova & Seddon, 2008).

The database of ILs is increasing day by day, but still it is not possible to explore all ILs. Therefore, prediction or even correlation of their thermophysical properties is essential and useful for engineering applications. Considering the limited database reported in the literature, prediction or even correlation of their thermophysical properties is essential and useful for engineering applications.

Extraction and recovery of valuable chemicals and products are the prime concern of any chemical industry. One such promising extraction process is biofuels such as ethanol, propanol and butanol from fermentation broth. Biobutanol is typically produced via the acetone–butanol–ethanol fermentation process using renewable feedstock. Butanol has been identified as a superior biofuel with excellent fuel properties. Compared to ethanol and other fermentation-derived fuels, butanol offers several advantages as a biofuel such as higher energy content, lower volatility, lower hygroscopy and better miscibility with gasoline. Apart from its use as a biofuel, butanol also makes a suitable platform chemical for further processing to advanced biofuels such as butyl levulinate. Ethanol and easily mixed with gasoline in any proportion. Butanol can be obtained from a petrochemical route as well as biochemical route. The butanol produced by fermentation is very dilute. Different ILs have been considered as solvents based on density and hydrophobicity. Various hydrophobic ILs (Garcia-Chavez et al., 2012; Simoni et al., 2010) have shown better selectivity for butanol separation from aqueous solution and are more economical when compared to hydrophilic ILs for extraction of water. The composition analysis can be carried out by nuclear magnetic resonance (NMR) spectroscopy (Anantharaj & Banerjee, 2011; Potdar, Anantharaj, & Banerjee, 2012; Shah et al., 2013).

The last decade also saw considerable research in the fast pyrolysis process for the production of liquid fuel and chemicals from biomass. Fast pyrolysis of biomass produces 60–75 wt% of liquid bio-oil depending on the feedstock used. Due to high concentration of the value-added chemical compounds, production of chemicals from bio-oil has received considerable interest. Fractionation of bio-oil with water is the easiest method which transforms bio-oil into two fractions: an aqueous top phase enriched in carbohydrate-derived chemicals and an organic bottom phase containing lignin-containing fractions (Mohan, Pittman, & Steele, 2006; Vitasari, Meindersma, & de Haan, 2011). The aqueous phase derived from bio-oil is a good feed for the extraction of acetic acid, levoglucosan and sugar compounds. In place of conventional organic solvents, ILs can be used for the extraction of valuable chemicals from the aqueous phase derived from the bio-oil.

The experimental equilibrium data are usually correlated with thermodynamic models. Among them the calculation of activity coefficients is an integrated part in phase equilibrium calculations involving liquid phases. Activity coefficients are calculated via excess Gibbs free energy models like nonrandom two-liquid (NRTL) (Renon & Prausnitz, 1968), UNIversal QUAsiChemical (UNIQUAC) (Abrams & Prausnitz, 1975). Each of these models requires proper binary interaction parameters that can represent LLE for highly nonideal liquid mixtures. These parameters are usually estimated from the known experimental LLE data via optimization of a suitable objective function. The optimization problem can be either the least-squares objective function minimization or likelihood function maximization. In both cases the objective function is nonlinear and nonconvex in terms of optimization variables; this possesses several local minima/maxima/saddle points within the specified bounds of the variables.

Nature-inspired metaheuristic algorithms are becoming increasingly popular to solve global optimization problems. They work remarkably efficiently and have many advantages over traditional, deterministic methods. These algorithms are broadly classified into Evolutionary Algorithms, Physical Algorithms, Swarm Intelligence, Bio-inspired Algorithms and others (Nanda & Panda, 2014). Genetic Algorithm (GA), developed by John Holland and his collaborators, is one of the most widely used optimization algorithms in modern nonlinear optimization (Yang, 2010). Cuckoo Search (CS) developed by Yang and Deb (2009) is a population-based method which mimics the breeding behavior of certain cuckoo species and is one of the latest nature-inspired metaheuristic algorithms. Recent studies have shown that CS is potentially far more efficient than other algorithms in many applications. In this book, binary interaction parameters were estimated using the GA and CS algorithm and a comparison has been made between these two algorithms.

COSMO (conductor-like screening model) (Klamt, 1995) based models such as COSMO-SAC (conductor-like screening model segment activity coefficients) also determine liquid phase nonideality using molecular interactions derived from quantum chemical solvation calculation. COSMO-RS parameters are optimized against a large data set of experimental data and can be used for other types of systems without sacrificing accuracy. By this way, the model can be termed as a predictive model. In COSMO-based models, a molecule is moved from a vacuum to a perfect conductor and then to a real solvent. The molecules are regarded as a collection of surface segments and chemical potential of each segment is self-consistently determined from statistical mechanical calculation. The difference in segment activity coefficient between mixture and pure liquid gives the segment activity coefficients and activity coefficient of a molecule is obtained from summation over the segment activity coefficients. Thermodynamic properties are necessary for more complex phase behaviors, such as vapor–liquid–liquid equilibrium (VLLE). If the presence of heterogeneous liquid mixtures is not correctly accounted for the systems that exhibit a miscibility gap, there may be multiple solutions to vapor–liquid phases, which in turn may be a possible reason for multiple steady states in heterogeneous distillation. The condition at which VLLE occurs can be obtained from experimental measurements; however, it is often economical and time saving to predict computationally. The objectives of VLLE computations are to determine which of the three phases are present and to determine the composition of phases. In the original COSMO-SAC modelling (Lin & Sandler, 2002), the hydrogen donors or acceptors were considered when their charge density exceeds a certain threshold value (e.g. |0.0084| e/Ų), which leads to the step function change in probability. However for aqueous systems, this approach did not converge with the modified Rachford–Rice algorithm, which led to the failure in predicting the mole fractions of extract and raffinate phases. We have thus adopted a continuous probability distribution function (Hsieh, Sandler, & Lin, 2010; Lin, Chang, Wang, Goddard, & Sandler, 2004; Wang, Sandler, & Chen, 2007) of charge density so that for the acceptor and donor segments, higher the charge density, the greater the possibility of forming a hydrogen bond. The hydrogen bonding portion was obtained from the combination of the electronegative atom and hydrogen atom only.

Equation of state (EoS) have also assumed an important place in the prediction of phase equilibria of fluids and fluid mixtures. Using Wertheim’s first-order thermodynamic perturbation theory, Chapman et al. (Chapman, Gubbins, Jackson, & Radosz, 1989, 1990) and Huang and Radosz (1990, 1991) developed the statistical associating fluid theory (SAFT) for pure fluids and mixtures containing associating fluids. Over the years, there have been numerous developments and modification of the SAFT EoS. In this contrast, the perturbed-chain statistical associating fluid theory (PC-SAFT) EoS developed by Gross and Sadowski (2001, 2002) has been successful for predicting the vapor–liquid and liquid–liquid equilibria of various systems based on only pure component parameters namely number of segments (m), segment diameter (σ), depth of potential (ε/k_B), association energy (ε^AiBj/k) and effective association volume (κ^AiBj). In Chapter 2, the PC-SAFT EoS has been used to predict the ternary LLE.

For transforming laboratory data into industrial application, the process optimization study is also necessary and useful. The binary interaction parameters generated by the NRTL model have been used for multi-stage extractor optimization. In the past, many popular stochastic algorithms such as GA (Goldberg, 1989), Simulated Annealing (Kirkpatrick, Gelatt, & Vecchi, 1983), particle swarm optimization (PSO) (Eberhart & Kennedy, 1995) and Differential Evolution (Storn & Price, 1997) have been investigated for optimization in science and engineering. PSO is an evolutionary algorithm based on social behavior of birds in swarm. The initial position and velocity of each particle are initiated randomly. During simulation, each particle in swarm (population) updates its position and velocity based on its experience as well as neighbors’ experience within the search space. The PSO technique was employed for optimizing the flow rate and number of stages in a multi-stage extractor. A multi-stage extractor containing more than two components requires detailed design like temperature, pressure, flow rate and composition in each stage. These are achieved by solving material balance equations (M), phase equilibrium relation (E), mole fraction summation for each stage (S) and energy balance equations (H), better known as MESH equations. In this book, the traditional Isothermal Sum Rate method (Tsuboka & Katayama, 1976) has been considered fo...