Measurement and modeling of CO₂ solubility in binary aqueous DMSO and MDEA and their ternary mixtures at different temperatures and compositions

Abstract

CO₂ capturing is an industrial tool in worldwide sustainable development to reduce the greenhouse effect. In this study, the new experimental vapor-liquid equilibrium (VLE) data for CO₂ solubility in the mixtures of physical and chemical solvents were investigated. The carbon dioxide solubility measurements were carried out using a quasi-static method through the DMSO (10 wt.%) +H₂O and sulfolane (20 wt.%) +H₂O systems at the temperatures of 328.15 and 343.15 K and pressure up to 3554 kPa. Moreover, the CO₂ partial pressure against its gas loading was obtained for the DMSO+MDEA+H₂O system at the compositions of (10-40-50), (20-40-40) wt. %, 328.15, 343.15 K, and pressure up to 3343 kPa. The influence of the physical solvents such as sulfolane and DMSO in the aqueous MDEA mixtures on CO₂ solubility was studied. We showed that DMSO was more effective than sulfolane in the absorption of CO₂. Finally, using the ion-pair concept, the eNRTL model was used to model the vapor-liquid equilibrium experimental data so that the molecule–molecule, ion pair-molecule binary interaction parameters of this model were obtained. An overall average relative deviation (AAD%) of 10.95 has resulted in the quaternary DMSO+MDEA+H₂O+CO₂ system.

Introduction

The alkanolamines through the chemical absorption as an industrial process have been widely used to reduce the CO₂ emission by post-combustion CO₂ capturing from oil refineries, power plants, and natural gas sweetening processes [1]. The aqueous solutions of alkanolamines such as DEA and MDEA react with carbon dioxide to remove this acid gas from natural gas using an absorber at high pressure and low temperature so that the amine is regenerated in the stripper by increasing temperature [2]. The various alkanolamines with different compositions and temperatures are often used in an acid gas removal unit to find an optimized solvent mixture to achieve maximum CO₂ absorption capacity, less corrosion, lower decomposing, and cost. Aqueous alkanolamine as a chemical solvent is protonated through the reactive solubility of CO₂ in the solution. However, this mechanism is limited by reactant stoichiometry. The high solubility of acid gases in alkanolamines at low partial pressure allows one to use this technique economically in a wide range of flow / acid gases, so that is a demanded and applicable method in different operating conditions of pressure and temperature. This chemical solvent has been widely used in the gas treating industry due to low vapor pressure and the several advantages of the aqueous MDEA solution [3]. However, the absorption of CO₂ in the solution is kinetically controlled. To overcome this problem at high pressure, one can use a combination of chemical and physical solvents to achieve both advantages of these solvents. The solubility of the acid gases in a physical solvent is taking place without any reaction, and the gas pressure is the primary driving force. Thus, using a physical solution leads to high loading at high partial pressure. Therefore, the chemical solvents are appropriate at low gas partial pressure, and the physical solvents are suitable for the absorption of acid gases at high partial pressures [4]. In other words, a combination of both physical and chemical solvents allows one to use these mixtures at the full range of pressures and temperatures conditions of gas capturing or a gas treatment unit.

Tetramethylene sulfone (sulfolane), which is abbreviated as TMS, is one of the well-known common physical solvents that are widely used in acid gas removal processes. Due to the presence of a sulfonyl group with a sulfur atom double-bonded to two oxygen atoms, it makes the possibility to interact with the sulfolane molecules with the sulfur components such as carbonyl sulfide (COS) and mercaptans. Thus, one can remove these contaminants from sour gases [5]. An alternative physical solvent such as dimethyl sulfoxide (DMSO) allows one to use in the absorption of carbon dioxide. The properties of this solvent are similar to tetramethylene sulfone (sulfolane). Moreover, DMSO has represented to be a versatile and powerful physical solvent with high potential to dissolve most aromatic and unsaturated hydrocarbons, organo-sulfur compounds, and many inorganic salts [6]. Thus, using a mixture of the physical and chemical solvents leads to reduce the concentration of acid gases in sour and industrial exhausted gases at low and high partial pressures and simultaneously remove the sulfur components such as mercaptans from stream gases [7].

Researchers have carried out several experimental works to measure the solubility of the acid gases in a mixture of chemical and physical solvents. However, the studies on the solubility of CO2 in a mixed-solvent of MDEA-DMSO in different compositions and temperatures are inadequate. Several researchers measured the CO₂ solubility in pure DMSO [8], [9], [10]. However, to our best knowledge, no experimental data has been reported for blends of mixed-solvent. Andreatta et al. measured the CO₂ solubility in two systems of pure and the aqueous DMSO at 280 and 370 K and pressure up to 15 MPa [11]. Besides, they modeled the experimental data with a group-contribution equation of state that takes into account the association effects (GCA-EOS). Wanren and Hua reported the solubility of SO₂ and CO₂ in DMSO within the temperature of 293.15 to 313.15 K, partial pressures of SO₂, and CO₂ from 0.15-2.62 kPa and 5-18 kPa, respectively, [12]. Dash and Bandyopadhyay measured the CO₂ solubility in the mixture of sulfolane+piperazine (Pz) with aqueous MDEA within the temperatures of 308-328 K and CO₂ partial pressures of 1-1400 kPa [13]. Their results showed that the CO₂ solubility decreases at a lower partial pressure range of 10–100 kPa, but enhanced at higher partial pressure ranges of 100–1000 kPa at 323.15 K for the sulfolane based composite solvents. They also figured out that adding Pz to aqueous MDEA presented a positive effect on the mixed-solvent CO₂ absorption. Rajasingam et al. produced the VLE experimental data for solubility of CO₂ in the mixed-solvent of DMSO and NMP at a pressure near the critical point of each binary system and temperature of 298, 308 and 318 K [14]. Moreover, they measured and correlated the volumetric expansion of the DMSO+NMP+CO₂ system against the solubility of CO₂ in the liquid phase. In another work, Shirazizadeh and Haghtalab reported the new experimental VLE data for simultaneous solubility of CO₂ and ethyl-mercaptan in the aqueous mixture of sulfolane and MDEA at the different solvent mass composition and the temperature range of 328.15 to 363.15 K and CO₂ partial pressure up to 1235 kPa [15]. To modeling of CO₂ capturing and adsorption of acid gases in pure and mixed-solvent, many models such as PC-SAFT, Electrolyte-NRTL, Kent-Eisenberg, and UNIQUAC have been developed so far. Zong and Chen applied the Electrolyte NRTL model for the aqueous sulfolane+DIPA and sulfolane+MDEA systems to calculate the activity coefficient of components in the liquid phase and the vapor phase nonideality was accounted for by PC-SAFT equation of state [16]. Moreover, Mondal et al. used Kent and Eisenberg's model for correlation of the experimental CO₂ solubility in the aqueous mixture of 2-amino-2-methyl-2-propanol and hexamethylenediamine with good accuracy [17]. Also, Ghalib et al. applied the Electrolyte-NRTL model for correlation of the CO₂+MDEA+PZ system [18].

Although researchers have performed extensive works for the solubility measurement of acid gasses in a mixed-solvent, to our best knowledge, no study has been reported for the mixed-sol DMSO's CO₂ solubility a physical solvent and MDEA as a chemical solvent. Thus, our objective is to investigate the solubility of CO₂ in the systems of MDEA + DMSO + water with the compositions of 40-10-50 and 40-20-40 wt. % and at temperatures of 328.15 and 343.15 K. Moreover, we measured the solubility of CO₂ in the aqueous sulfolane (20 wt.%) +water and DMSO (10 wt.%) +water at 328.15 and 343.15 K. Furthermore, the experimental VLE data were modeled by an eNRTL activity coefficient model for correlation of the experimental solubility of CO₂ in the present mixed-solvent systems. The eNRTL model consists of two terms of Pitzer-Debye-Hückel and the original NRTL for the long-range and short-range interactions, respectively.