Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach

For the design and development of new processes of gas sweetening using ionic liquids (ILs), as promising candidates for amine solutions, an amazing model to predict the solubility of acid gases is of great importance. In this direction, in the current study, the capability of artificial neural networks (ANNs) trained with back propagation (BP) and particle swarm optimization (PSO), to correlate the solubility of H₂S in 11different ILs have been investigated. Different structures of three-layer feed forward neural network using acentric factor (ω), critical temperature (T_c), critical pressure (P_c) of ILs accompanied by pressure (P) and temperature (T), as input parameters, were examined and an optimized architecture has been proposed as 5–9–1.Implementation of these models for 465 experimental data points collected from the literature shows coefficient of determination (R²) of 0.99218 and mean squared error (MSE) of 0.00025 from experimental values for PSO-ANN predicted solubilities while the values of R² = 0.95151 and MSE = 0.00335 were obtained for BP-ANN model. Therefore, through PSO training algorithm we are able to attain significantly better results than with BP training procedure based on the statistical criteria.     

Estimating solubility of supercritical H2S in ionic liquids through a hybrid LSSVM chemical structure model

Development of a predictive tool for H₂S solubility estimation can be very helpful in gas sweetening industry. Experimental databases on H₂S solubility were rarely available, so as reliable predictive models. Thus, in this study the H₂S solubility database was established, and then a Least-Squares Support Vector Machine (LSSVM) approach based on the established database is proposed. Group contribution method was also applied to eliminate the model's dependence on experimental data. Accordingly, our proposed LSSVM model can predict H₂S solubility as a function of temperature, pressure, and 15 different chemical structures of Ionic liquids (ILs). Root Mean Square Error (RMSE) and coefficient of determination (R²) are 0.0122 and 0.9941, respectively. Moreover, comparison of our model with other existing models showed its reliability for H₂S solubility in ILs. This can be very useful for engineers dealing with gas sweetening process in different applications of analysis, simulation, and designation.     

Efficient conversion of H2S into mercaptan alcohol by tertiary-amine functionalized ionic liquids

The resource utilization of hydrogen sulfide (H₂S) is of great significance in natural gas chemical industry. Described herein have developed a novel method mediated in tertiary amine-functionalized ionic liquids (ILs) to convert H₂S into mercaptan alcohols with enols. The effect of ILs, substrate scope, and regeneration experiments have been investigated. It is found that the conversion of 3-methyl-2-buten-1-ol by H₂S can reach 52% with a 50% (mol) catalyst loading of bis(2-dimethylaminoethyl) ether methoxyacetate within 12 h at 90 ℃. The reaction mechanism was speculated based on theoretical calculation. Besides, a plausible reaction-separation-integrated strategy was further proposed. This work reveals an effective insight into the capture and catalytic conversion of H₂S to high valuable mercaptan alcohol, which makes the utilization method of H₂S resource universal and has the potentiality for industrial application.     

Artificial Neural Network modeling of solubility of supercritical carbon dioxide in 24 commonly used ionic liquids

Application of supercritical CO₂ for separation of ionic liquids from their organic solvents or extraction of various solutes from ionic liquid solvents have found great interest during recent years. Knowledge of phase behaviors of the mixtures of supercritical CO₂+ionic liquids is therefore drastic in order to efficiently design such separation processes. In this communication, Artificial Neural Network procedure has been applied to represent the solubility of supercritical CO₂ in 24 mostly used ionic liquids. An optimized Three-Layer Feed Forward Neural Network using critical properties of ionic liquids and operational temperature and pressure has been developed. Application of this model for 1128 data points of 24 ionic liquids show squared correlation coefficients of 0.993 and average absolute deviation of 3.6% from experimental values for calculated/estimated solubilities. The aforementioned deviations show the prediction capability of the presented model.     

A new approach for correlating of H2S solubility in [emim][Lac], [bmim][ac] and [emim][pro] ionic liquids using two-parts combined models

Two bi-part combined models containing reaction equilibrium thermodynamic model and an equation of state (EoS) including cubic plus association (CPA) or modified Soave-Redlich-Kwong (SRK), were employed to correlate H₂S solubility in ionic liquids [emim][Lac], [bmim][Ac] and [emim][Pro]. RETM and CPA/mSRK are responsible for chemical and phase equilibria calculations, respectively. The RETM defines a complex formation reaction between an H₂S and two IL molecules into the liquid phase so that a complex in the form of H₂S(IL)₂ is probable to produce. On the other side, CPA/mSRK EoSs perform phase equilibria computations to find pressure and vapor phase concentrations. In the case of CPA, both of H₂S and the ionic liquids were considered as self-associating components which pursue 4C and 2B schemes, respectively. This model computes the pure IL parameters using experimental density data. In this work, critical properties were estimated through the modified Lydersen-Joback-Reid method, Eotvos and Guggenheim relations. Afterward, the binary systems were investigated by applying RETM. CPA presents average absolute deviations (AADs) equal to 2.41%, 13.42% and 3.52% for [emim][Lac], [bmim][Ac], and [emim][Pro], respectively. Moreover, the AADs obtained by mSRK are 3.75%, 5.07%, and 6.06%, respectively. As it is evident from the results, the combination of RETM with both CPA and mSRK EoSs will result in good correlation accuracy.     

Statistical mechanics and artificial intelligence to model the thermodynamic properties of pure and mixture of ionic liquids

In this paper, the volumetric properties of pure and mixture of ionic liquids are predicted using the developed statistical mechanical equation of state in different temperatures, pressures and mole fractions. The temperature dependent parameters of the equation of state have been calculated using corresponding state correlation based on only the density at 298.15 K as scaling constants. The obtained mean of deviations of modified equation of state for density of al pure ionic liquids for 1662 data points was 0.25%. In addition, the performance of the arti-ficial neural network (ANN) with principle component analysis (PCA) based on back propagation training with 28 neurons in hidden layer for predicting of behavior of binary mixtures of ionic liquids was investigated. The AADs of a col ection of 568 data points for al binary systems using the EOS and the ANN at various temperatures and mole fractions are 1.03%and 0.68%, respectively. Moreover, the excess molar volume of all binary mixtures is predicted using obtained densities of EOS and ANN, and the results show that these properties have good agree-ment with literature.     

Rigorous correlations for predicting the solubility of H2S in methylimidazolium-based ionic liquids

Two general and simple models, a group contribution correlation (model I) and an empirical relation (model II), were proposed to predict the solubility of H₂S in methylimidazolium based ionic liquids (ILs) over wide range of temperatures (303.15-363.15 K) and pressures (60.8-2016.8 kPa). The constants of the suggested functionality relations were found via the Nelder-Mead simplex algorithm. Both correlations were trained with 407 data points of H₂S solubility in 9 methylimidazolium based ILs and tested through 121 H₂S solubility data points of 3 different methylimidazolium based ILs to ensure generality. A comprehensive statistical evaluation showed that both suggested correlations are vigorous and have satisfactory error trends. The dataset was subjected to a statistical outlier diagnostic test and the validity of the database was confirmed. In addition, the sensitivity analysis revealed that the experimental data and both models have the same responses toward pressure and temperature, which indicates the reliability of the proposed correlations.     

High pressure solubility of CO2 in non-fluorinated phosphonium-based ionic liquids

High pressure solubility of carbon dioxide in three phosphonium-based ionic liquids has been measured experimentally. A synthetic method was used to measure vapor–liquid, vapor–liquid–liquid and liquid–liquid equilibria of carbon dioxide in the ionic liquids trihexyltetradecylphosphonium bromide [thtdp][Br], trihexyltetradecylphosphonium dicyanamide [thtdp][dca] and trihexyltetradecylphosphonium bis(2,4,4-trime-thylpentyl)phosphinate [thtdp][phos] for a temperature range of 271–363 K and up to 90 MPa. Furthermore, the densities and viscosities of these ILs have been measured in a temperature range of 293–363 K. The solubility of carbon dioxide in these ILs is (on mole fraction basis) significantly larger than most of the commonly used fluorinated imidazolium ionic liquids. The latter statement, however, does not hold for the [Br] and [dca] based IL if the solubility is compared on molality (mole/mass solvent) basis, where the solubility differences among physical ILs tends to vanish indicating a strong molecular weight effect. The solubility of carbon dioxide in [thtdp][phos], both on mole fraction and molality basis, is among the highest compared to all the other physical ILs reported so far in the literature. The Peng–Robinson equation of state in combination with Wong–Sandler mixing rules incorporating the NRTL Gibbs excess energy model was applied to represent the experimental data with acceptable accuracy.     

Separation of light hydrocarbons with ionic liquids: A review

Light hydrocarbons(C_1–C_4) are fundamental raw materials in the petroleum and chemical industry. Separation and purification of structurally similar paraffin/olefin/alkyne mixtures are important for the production of highpurity or even polymer-grade light hydrocarbons. However, traditional methods such as cryogenic distillation and solvent absorption are energy-intensive and environmentally unfriendly processes. Ionic liquids(ILs) as a new alternative to organic solvents have been proposed as promising green media for light hydrocarbon separation due to their unique tunable structures and physicochemical properties resulting from the variations of the cations and anions such as low volatility, high thermal stability, large liquidus range, good miscibility with light hydrocarbons, excellent molecular recognition ability and adjustable hydrophylicity/hydrophobicity. In this review, the recent progresses on the light hydrocarbon separation using ILs are summarized, and some parameters of ILs that influence the separation performance are discussed.     

Exceptionally effective H2S absorption and conversion into thiols in novel superbase protic ionic liquids

We demonstrate an effective strategy for capture and conversion of H₂S in novel superbase protic ionic liquids (SPILs). It is found that the synthesized SPILs with the multiple active sites exhibit the unprecedented H₂S uptake via chemical absorption (up to 1.81 mol mol⁻¹ at 298.2 K and 1 bar). More importantly, H₂S molecule is activated by these SPILs during the absorption process, so that the activated H₂S can be converted further into high value-added thiols in situ with excellent yields under mild conditions (303.2 K and 1 bar) without any solvents and metallic catalysts. Since H₂S-saturated SPILs can be regenerated by chemical conversion of absorbed H₂S into thiols, thereby eliminating the higher input of energy consumption during the process of H₂S stripping. This SPIL-mediated scheme provides an alternative approach for the capture and chemical conversion of H₂S.     

Prediction of thermodynamic properties of polymeric liquids using a new equation of state

In this paper, we have used a simple equation of state (EoS) to predict the density for polymeric liquid mixtures at different temperatures, pressures, and compositions. The excess molar volumes of these mixtures have been also calculated using this equation of state. Also, we have computed isothermal compressibility. A wide comparison with experimental data has been made for each thermodynamic property. The values of statistical parameters between experimental and calculated properties show the ability of this equation of state in reproducing and predicting different thermodynamic properties for studied polymeric mixtures.     

Extraction of dibenzothiophene from dodecane using ionic liquids

The effect of ionic liquid loading, extraction temperature, and extraction time in the removal of dibenzothiophene from dodecane were investigated. Eighteen (18) ionic liquids were screened for its dibenzothiophene extraction ability. Imidazolium based ionic liquids with thiocyanate, dicyanamide and octylsulfate anions exhibited the highest extraction capabilities with 66.1%, 66.1%, and 63.6% of extraction efficiency respectively. Tributylmethylammonium methylcarbonate ionic liquid gave 61.9% extraction efficiency, which showed that π-π interaction between aromatic rings of sulfur compound and ionic liquid (IL) was not be the main extraction mechanism. A trend between specific volume and desulfurization efficiency of ILs was put forward, enabling researchers to predict ILs' desulfurization efficiency from its specific volume. It was also found that [C₄mim][SCN] can be reused in extraction without regeneration with considerable extraction efficiency of 41.9%. Huge saving on energy can be achieved if we make use of this IL behavior in process design, instead of regenerating IL after every time of extraction.     

High pressure CO2 solubility in N-methyl-2-hydroxyethylammonium protic ionic liquids

The use of ionic liquids for CO₂ capture and natural gas sweetening is being object of intense research. Within the enormous group of existing ionic liquids, those based on conjugate bases of carboxylic acids seem to be particularly promising. This work addresses the study of the high pressure CO₂ solubility (up to 80 MPa) in two protic ionic liquids, N-methyl-2-hydroxyethylammonium formate and N-methyl-2-hydroxyethylammonium acetate, in a wide range of temperatures (293-353 K). A thermodynamic model based on the Peng-Robinson equation of state with the Wong-Sandler mixing rule, using the NRTL model for the activity coefficients, was here adopted to describe and evaluate the thermodynamic consistency of the experimental data. Furthermore, the study of a ternary mixture of CO₂ + CH₄ + N-methyl-2-hydroxyethylammonium acetate was investigated showing a high selectivity from the IL towards these solutes.     

Estimation of densities of ionic liquids using Patel–Teja equation of state and critical properties determined from group contribution method

In this paper, the Valderrama and Robles group contribution model for the critical properties is extended to the prediction of densities of ionic liquids (ILs) at varying temperatures and pressures, where the critical properties of ILs are represented by the modified Lydersen–Joback–Reid group contribution method, and the density is predicted by virtue of the Patel–Teja (PT) equation of state (EOS). The group increments for totally 47 groups with respect to the extended Group Contribution Patel–Teja (GC–PT) model were determined on the basis of experimental density data for 747 pure ILs at atmospheric pressure and ambient temperature. The group increments are suitable for both the GC–PT model and the original Valderrama and Robles model for the density prediction of ILs. The correlation accuracy in terms of overall average absolute relative deviation (AARD) is 4.4% for 918 data points of density at ambient temperature and atmospheric pressure. The applicability of the GC–PT model is justified by predicting the densities of imidazolium-, pyridinium-, and phosphonium-based ILs containing various anions over a wide range of temperatures and pressures and the vapor pressures of five alkylimidazolium-based ILs at varying temperatures, which implied the rationality of the group increments and the critical properties of ILs, as well as the potential uses of the GC–PT model for the thermodynamic modeling of IL-containing systems.     

Gas solubilities in ionic liquids using a generic van der Waals equation of state

A family of modified van der Waals equations of state (vdW EOS) is extremely useful for many industrial applications. For example, the generic Redlich-Kwong (RK) EOS or its modification by Soave (SRK EOS) and Peng-Robinson (PR EOS) are still of popular use in industry to the present day. These two most popular (“cubic”) EOSs are based on modifications [1/(V² + bV), or 1/(V² + 2bV − b²)] of the volume dependence on the attractive part of the original van der Waals EOS [1/V²] and also modifications of the temperature dependence of the attractive “a(T)” parameter of the original EOS (constant a). It is extremely rare in actual EOS applications to use the volume dependence of the original van der Waals EOS. In the present phase equilibrium calculations, we employ such a generic vdW EOS, P = RT/(V − b) − a(T)/V², with our well-tested mixing rule for multi-component mixtures. Using the same form of the “a(T)” parameter and the mixing rule, it has been found that all generic RK, PR, and vdW EOSs can present the phase behaviors (temperature-pressure-composition diagrams) equally well. It is shown that experimental gas solubility data (CO₂, CF₃-CFH₂, SO₂, and NH₃) in room-temperature ionic liquids are well correlated with the present EOS model, and also that the phase behaviors such as LLE (liquid-liquid separations) are satisfactorily predicted.     

Electrochemical synthesis of polypyrrole in ionic liquids

Electrochemical synthesis of inherently conducting polymers such as polypyrrole is traditionally performed in a molecular solvent/electrolyte system such acetonitrile/lithium perchlorate. We report the use of ionic liquids 1-butyl-3-methylimidazolium hexafluorophosphate, 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl) amide and N,N-butylmethylpyrrolidinium bis(trifluoromethanesulfonyl) amide, both as the growth medium and as an electrolyte for the electrochemical cycling of polypyrrole films. Use of the ionic liquid as the growth medium results in significantly altered film morphologies and improved electrochemical activities.     

Solution spinning of cellulose carbon nanotube composites using room temperature ionic liquids

Multiwall carbon nanotube (MWCNT)/cellulose composite fibers were processed from solutions in ethyl methylimidazolium acetate (EMIAc). Rheological percolation in MWCNT/Cellulose/EMIAc solution was observed above 0.01 mass fraction of MWCNT, while electrical percolation in oriented fibers was observed above 0.05 mass fraction of MWCNTs with respect to the weight of the cellulose. Cellulose orientation and crystal size were significantly higher in the composite than in the control cellulose fiber. In addition, in the composite fiber, carbon nanotube orientation was higher than cellulose orientation. At 0.05 mass fraction MWCNT, fiber tensile strength increased by about 25%, strain to failure increased by 100%, and modulus essentially remained unchanged. The composite fibers showed lower thermal shrinkage than the control cellulose fiber. The axial electrical conductivity at 0.1 mass fraction MWCNTs in these oriented fibers was more than 3000 S/m.     

Some trends and a prediction of the solubility of gases in liquids and the heat of dissolution

An equation is derived for an approximate calculation of the solubility of gases in liquids and of their heat of dissolution using the energy parameters (force cosntants) of the Lennard-Jones potential data of pure gases and solvents, but without using empirical coefficients determined from the properties of binary systems. The equation describes some trends including the linear dependence of the gas solubility logarithm and the dissolution heat on the square root of the force constant of the Lennard-Jones potential of the gas, the extremal dependence (via the maximum) of the gas solubility on the same constant of the solvent, and the possibility and condition of the appearance of a minimum in the gas solubility with variations in temperature. The theoretical results are compared with experimental data.     

Fixation and conversion of CO2 using ionic liquids

Ionic liquids (ILs), a kind of novel green media composed entirely of cations and anions, have recently attracted considerable attention due to their unique properties such as non-volatility, tunable polarity, high stability and so on. In this work, the latest progress on the fixation and conversion of carbon dioxide (CO₂) using ILs as absorbents, catalysts or promoters has been summarized. The absorption performance of conventional ILs and task-specific ILs was systematically investigated, the conversion of CO₂ with epoxides, propargyl alcohols and amines using ILs was critically evaluated, and the significant advantages in the fixation and conversion of CO₂ using the ILs were demonstrated compared to the conventional absorbents and the catalytic systems without ILs. This research progress may finally lead to building of an in situ fixation–conversion process of CO₂ with ILs. If so, we are near an epoch of the fixation and utilization of CO₂, although there is obviously a long way to go for us to achieve such a goal.