Zhonghao Li, Yuanyuan Yang, Huanong Cheng*, Yun Teng, Chao Li, Kangkang Li, Zhou Feng, Hongwei Jin,Xinshun Tan, Shiqing Zheng

College of Chemical Engineering, Qingdao University of Science and Technology, Qingdao 266042, China

Keywords:

ABSTRACT

1. Introduction

Hydrogen sulfide (H2S) is a colorless gas at room temperature with a special smell of rotten eggs.It is mainly produced in the processes of natural gas development, coal gasification, petroleum refining, ammonia production,sewage treatment, and paper manufacturing [1,2]. H2S has flammable, explosive properties and strong toxicity[3].H2S in gaseous materials can cause serious corrosion of metal equipment and poisoning of the catalyst.H2S in the atmosphere also causes acid rain,which is harmful to humans and the environment [4,5]. China has set strict standards for sulfur levels in environment and oil, which must not exceed 10 μg?g-1[6,7]. Therefore, to control the environmental pollution caused by industrial production, gas desulfurization must be carried out.

The removal of H2S by solvent absorption can be divided into physical absorption, chemical absorption, and wet oxidation. The physical absorption method (such as the Rectisol process [8]) has the advantage of low regeneration energy consumption. It is more suitable for high H2S concentration. But it needs to be operated under high pressure and low temperature [9]. The most popular chemical absorption method is amine scrubbing [10], but it has some disadvantages such as foaming of aqueous solution and high regeneration energy consumption.Wet oxidation methods include ADA method[11],tannin extraction method[12],and complex iron method[13].They have the advantages of high desulfurization efficiency, simple process, and operation at normal temperature and atmospheric pressure, so they have been widely in application[14]. However, there are still some problems with wet oxidation methods. For example, secondary salt (such as thiosulfate, sulfate,and sulfite) is produced during desulfurization [15]. To maintain the desulfurization efficiency, the waste liquid needs to be discharged from the absorption solution to keep the pH value of the solution stable between 9 and 10 [16].

Ionic liquids (ILs) are organic salt liquids at room temperature,composed of organic cations and inorganic or organic anions[17,18].Ionic liquids have stable physical properties and low vapor pressure, which can hardly cause solution loss in the desorption process [19,20]. The structure of the ionic liquids are adjustable,which means researchers can design the ionic liquids according to the research object. Some metal ions (such as iron, aluminum,etc.) can be combined in the ionic liquids to form metal base ionic liquids. The iron-based ionic liquids (Fe-ILs) have oxidation and acidity [21]. H2S is directly oxidized to elemental sulfur without the generation of secondary salt [22], which avoids the problem of waste liquid discharge of the existing wet oxidation methods.

He et al. [22] synthesized an imidazole-based Fe-IL named[Bmim]Fe0.9Cl4.7using [Bmim]Cl and FeCl3?6H2O. They found that the Fe-IL was hydrophobic and applied it to the removal of hydrogen sulfide gas. The desulfurization efficiency reached more than 99%.Guo et al.[23]used[Bmim]FeCl4ionic liquid to remove hydrogen sulfide gas. N,N-dimethylformamide was used as a solvent to improve the absorption performance.

Ma et al. [24] synthesized another kind of Fe-IL using Et3NHCl and anhydrous FeCl3, rFeCl3/[Et3NH]Cl (r, molar ratio of FeCl3to Et3NHCl), and used it to remove H2S in high-temperature gas.The results showed that 0.6FeCl3/[Et3NH]Cl had the best desulfurization efficiency, which was better than ordinary imidazolebased ionic liquid and Fe3+aqueous solution. In addition, the sulfur generated in the reaction process was easier to separate. Li et al. [25] synthesized rFeCl3/[Et3NH]Cl with r from 0.56 to 0.71. They studied the influence of r, temperature, and pressure on the absorption of H2S in ionic liquids. And they found that the maximal sulfur capacity of rFeCl3/[Et3NH]Cl was 2.178%(mass) at r of 0.67, the temperature of 303.15 K, and the pressure of 101.3 kPa.

In addition to the above two Fe-ILs (FeCl3/[Bmim]Cl and FeCl3/[Et3NH]Cl), Kogelnig et al. [26] first synthesized ionic liquid[A336][FeCl4]0.73[Cl]0.27using [A336]Cl (methyl trioctyl ammonium chloride) and FeCl3?6H2O. And magnetism and aggregation behavior of [A336][FeCl4]0.73[Cl]0.27were studied. Wang et al.[27] synthesized ionic liquid [A336]FeCl4using [A336]Cl and FeCl3?6H2O. And [A336]FeCl4was combined with [Bmim]OH to absorb hydrogen sulfide at room temperature and atmospheric pressure.The results showed that[A336]FeCl4was stable,and sulfur content was up to 2.13%.The raw material for the synthesis of r FeCl3/[A336]Cl ionic liquid is [A336]Cl. It is a quaternary ammonium salt that is widely used in the metal extraction [28,29] and has been industrially produced on a large scale. rFeCl3/[A336]Cl has the advantage of low cost compared to expensive imidazolebased Fe-IL. Moreover, compared with FeCl3/[Et3NH]Cl, rFeCl3/[A336]Cl has a larger molar volume, which means that the free space of the solution is larger and the solubility may be higher.Therefore, rFeCl3/[A336]Cl has great application potential in the industrial application of H2S removal from gas.

Density, viscosity, and solubility have important effects on the gas purification performance of Fe-ILs.Li et al.[30]measured density and viscosity of rFeCl3/[Bmim]Cl (r, molar ratio of FeCl3to[Bmim]Cl) in the range of r from 1/1.7 to 1.5/1, T from 293.15 to 343.15 K. Density and viscosity decreased with the temperature increasing.Density increased with the r increasing,while viscosity decreased with the r increasing. As 1/1.7 ≤r ≤1/1, the viscosity decreased extremely, and the viscosity turned smoothly as r ≥1.Cheng et al. [31] determined and modeled the solubility of hydrogen sulfide in rFeCl3/[Bmim]Cl ionic liquid (rFeCl3/[bmim]Cl,r = 0.6–1.4) at temperatures of 303.15 to 348.15 K and pressures of 100 to 1000 kPa.Henry’s constant and chemical reaction equilibrium constant were obtained. To date, significant parameters of rFeCl3/[A336]Cl,such as viscosity,density,and H2S solubility,have not been reported. The actual oxidative desulfurization processes are operated at room temperature and atmospheric pressure. In this work, r[A336]FeCl4(r, molar ratio of FeCl3to [A336]Cl,r = 0.1, 0.2, 0.4, 0.6, and 0.8) ionic liquid was synthesized. Its density and viscosity were measured in the temperatures of 313.15 to 358.15 K under atmospheric pressure. Its solubility of H2S was measured at temperatures of 318.15 to 348.15 K and pressures of 0 to 150 kPa. The experimental data of density, viscosity and H2S solubility were fitted by models. The effects of temperature,pressure and molar ratio on density, viscosity and H2S solubility were discussed.

2. Materials and Methods

2.1. Experimental reagents

H2S gas was purchased from Yuyan Special Gas Co., Ltd. with a purity of 0.99. Anhydrous FeCl3was purchased from Aladdin Reagent Company with a purity of 0.995. [A336]Cl was provided from Macklin Reagent Company with a purity of 0.97. The pure water used in the density verification experiment was prepared by the UPR-II-5/10T equipment produced by Sichuan Youpu Ultra Pure Technology Co., Ltd.

rFeCl3/[A336]Cl was synthesized according to the existing methods [26,27]. A certain amount of [A336]Cl was weighed into a beaker,and then slowly added anhydrous FeCl3with corresponding molar ratio while stirring. After full dissolution, the mixture was stirred continuously at 333.15 K for 24 h.Finally,the obtained rFeCl3/[A336]Cl was transferred to a vacuum drying oven and dried at 333.15 K and 10 kPa for 48 h. rFeCl3/[A336]Cl was further dried with 3A molecular sieve for several days before using. The water content was not detected due to the water content of rFeCl3/[A336]Cl was below the detection limit(100 μg?g-1)of the Karl Fischer titrator (DL31, Mettler Toledo). The purity of rFeCl3/[A336]Cl was determined by ultraviolet spectrophotometry (UV-1800,Shanghai Mepuda Instrument Co.,Ltd.).Hydroxylamine hydrochloride was used to reduce ferric iron to ferrous iron. Divalent iron and ophenanthroline can form an orange-red complex, and the complex has a good absorbance at 512 nm of UV light. Before the measurement of Fe-IL,the ultraviolet spectrophotometer was calibrated with 100 μg?ml-1standard-iron solution(Shanghai Aladdin Biochemical Technology Co., Ltd.). A sample description is presented in Table 1.

2.2. Density measurement

The density was measured by a Westphal balance (PZ-D-5,Shanghai Yueping Scientific Instrument Manufacturing Co., Ltd.).The measurement range was 0–2.0000 g?cm-3,and the uncertainty of the measured density was 0.0001 g?cm-3. The balance had a standard measuring hammer immersed in the liquid.Then various quantitative riding weights were putted in the V-shaped groove of the beam to maintain balance,and the specific gravity of the liquid was accurately measured. The calibration standard was that the specific gravity of pure water at 293.15 K was 1.0000.In this work,the temperature of the liquid was adjusted through a water recirculation bath (HH-Sc, Changzhou Langyue Co., Ltd.), and the temperature accuracy was ±0.03 K.

Table 1 Purities and sources of the samples used in the experiment

2.3. Viscosity measurement

The viscosity was measured using a rotary digital viscometer(NDJ-4S, Shanghai Hengping Instrument Factory) with a measurement range of (1–2) × 106mPa?s and measurement uncertainty of±1%. Different rotors and rotational speeds were selected according to the predicted viscosity of the tested liquid. The viscosity of the liquid was calculated by measuring the viscosity moment of the liquid acting on the rotor at a constant speed. In this work, the temperature of the liquid was adjusted through a water recirculation bath (HH-Sc, Changzhou Langyue Co., Ltd.), and the temperature accuracy was ±0.03 K.

2.4. Solubility measurement

2.4.1. Experimental equipment and procedures

The experimental equipment for measuring the hydrogen sulfide solubility of rFeCl3/[A336]Cl at temperatures of 318.15 to 348.15 K and pressures of 0 to 150 kPa is shown schematically in Fig. 1. The experimental equipment and procedures are similar to those reported in our previous work [31,32] for the measurement of H2S solubility. The experimental equipment consists of a nitrogen gas cylinder, a hydrogen sulfide gas cylinder, a vacuum pump,a buffer tank, a phase equilibrium kettle (SL-Q100, Sen Long Equipment Company) with stirring, and a lye absorption bottle. The volume of the buffer tank is 547.4 ml, and the material is 316 stainless steel. The volume of the phase equilibrium kettle is 122.8 ml, and the material is Hastelloy alloy. The temperature of the buffer tank and the phase equilibrium kettle is adjusted through two constant temperature water baths with stirring (HH-Sc, Changzhou Langyue Co., Ltd.). The temperature is monitored by TES 1320 TYPE-K thermocouple, and the uncertainty is ±0.03 K. The pressure of the buffer tank and phase equilibrium kettle is monitored and transmitted to the computer by an integrated digital pressure gauge (YK-100B,Yunyi Instrument Co., ltd), and the uncertainty is 0.4 kPa.

A certain mass of ionic liquid (wFe-IL) was added to the equilibrium kettle with an uncertainty of ±3.0 × 10-8kg. After checking the gas tightness of the experimental equipment, vacuumizing the system until the pressure was less than 0.1 kPa. At this stage,the system pressure was recorded as P1. A certain amount of H2S was filled into the phase equilibrium kettle through the buffer tank, and the pressure of the buffer tank before and after releasing H2S was recorded as P2and P3, respectively. Turned on the stirring device of the phase equilibrium kettle and started the absorption of H2S. When the absorption in the phase equilibrium kettle reached equilibrium, the pressure of the phase equilibrium kettle was recorded as P4. Finally, after the experiment had been completed,the residual H2S in the system was discharged into the lye absorption flask.

2.4.2. Calculation of solubility

At gas–liquid equilibrium,the total solubility of H2S in ionic liquid can be expressed by a molar fraction (xt,i) and molality based on unit ionic liquid mass (mt,i). They are given by Eq. (1) and Eq.(2).

where, mt,iis moles of H2S per unit mass of ionic liquid. nFe-ILand wFe-ILare the moles and mass of the ionic liquid, respectively. nt,iis the moles of H2S in the ionic liquid at the i-th absorption experiment. nt,ican be calculated from Eq. (3) and Eq. (4).

where,Δniis the incremental moles of H2S absorbed in the ionic liquid in the i-th experiment.nt,i-1is moles of H2S in the ionic liquid at the(i - 1)-th experiment. When i = 1, nt,i-1= 0. na,iis moles of H2S transferred from the buffer tank to the equilibrium kettle in the i-th experiment. ng,i-1and ng,iare moles of H2S in the gas phase of the equilibrium kettle when gas and liquid phases reach equilibrium in the (i - 1)-th and i-th experiment, respectively.

na,iand ng,iare calculated from Eq. (5) and Eq. (6) of state for a real gas.

Fig. 1. The experiment flowchart of absorption (1–6: valve; 7: buffer tank; 8: equilibrium kettle; 9,10: water bath; 11,12: thermocouple; 13,14: pressure gauge; 15:computer; 16: vacuum pump; 17: absorption bottle; 18: nitrogen cylinder; 19: hydrogen sulfide cylinder).

where T is the equilibrium temperature, R is the universal gas constant, R = 8.314. P2H2S,iand P3H2S,iare the partial pressure of H2S in the buffer tank before and after filling H2S from the buffer tank to the phase equilibrium kettle in the i-th experiment. PH2S,iis the equilibrium partial pressure of H2S in the equilibrium kettle in the i-th experiment. Z2,i, Z3,iand ZH2S,iare the compression factors corresponding to P2H2S,i, P3H2S,i, and PH2S,i, respectively. These compression factors can be obtained by the PR thermodynamic method in Aspen Plus software (AspenONE?for Universities, Aspen Technology, Inc., Bedford, MA, USA).

P2H2S,i, P3H2S,i, and PH2S,iare calculated by Eq. (7),Eq. (8), and Eq.(9).

where, P2,i, P3,iare the pressure of the buffer tank before and after injecting H2S to the phase equilibrium kettle in the i-th experiment.P4,iis the equilibrium pressure of the equilibrium kettle in the i-th experiment.P1is the pressure after the apparatus is vacuumized.PLis the partial pressure of the solvent.The vapor pressure of the ionic liquid in this work is low,so PL=0.For solvents with high saturated vapor pressure, PLcannot be ignored.

The solution volume VLcan be obtained from Eq. (10).

where wFe-ILis the mass of the ionic liquid.ρFe-ILis the density of the ionic liquid. The experimental absorption pressure is low, and the density of ionic liquid with dissolved H2S is approximately equal to that of pure ionic liquid.

3. Models

3.1. Density and viscosity models

Temperature variation of the density empirical correlation Eq.(11) [33] was used to fit the experimental data.

where ρ(g?cm-3)is density,T(K)is temperature.A and B are fitting parameters,which are obtained by fitting the equation with experimental data through least square regression.

VFT empirical Eq.(12)[34]was used to fit the experimental viscosity data.

where η (mPa?s) is viscosity, T(K) is temperature, A, B and T0are parameters of the equation,which are obtained by fitting the experimental data.

Coefficient of determination (R2), average relative error (ARD),and maximum relative error (MRD) are used to characterize the degree of fitting between the calculated value and the experimental data. These parameters are defined in the Supplementary Material.

3.2. Solubility model

Krichevsky-Kasarnovsky equation[35]and Henry’s law[36]can be used to model the solubility of hydrogen sulfide in ionic liquid by physical absorption. The redox reaction of rFeCl3/[A336]Cl occurred during the absorption of H2S. Thus, chemical absorption should also be taken into account. Huang et al. [37] modeled the solubility of H2S in 1-alkyl-3-methylimidazolium carboxylates ionic liquids using the reaction equilibrium thermodynamic model(RETM). Henry’s constant and chemical reaction equilibrium constant could be estimated by the RETM fitting. Henry’s constant characterized the physical solubility. The chemical reaction equilibrium constant characterized the chemical absorption. Wei et al. [38] applied the RETM to the mercapto carboxylic anions functionalized ionic liquid ([P4444][MSA]). Impressively, it was found that mercapto carboxylic anions captured H2S through 1:1 and 2:1 mole ratio manner. Zhang et al. [39] used the RETM to fit the gas–liquid equilibrium data of H2S in azole-based protic ionic liquid ([DBNH][1,2,4-triaz]). Furthermore, the enthalpy change of absorption was calculated according to the reaction equilibrium constant. Different ionic liquids and H2S were combined in different ways, which resulted in various expressions of the RETM. In this work, the corresponding RETM was established by analyzing the process of rFeCl3/[A336]Cl absorbing H2S.

rFeCl3/[A336]Cl has Bronsted acidic (Fig. S1 in the Supplementary Material), so the absorbed hydrogen sulfide mainly exists in the form of H2S molecules. The physical absorption is presented as the following Eq. (13), which means that H2S enters and dissolves in the ionic liquid from the gas phase.

where,subscripts g,and l denote gas and liquid phase,respectively.

According to the experimental results of H2S absorption by[bmim]FeCl4ionic liquid reported by Wang et al. [40], ferric iron still exists at the absorption equilibrium. Therefore, the chemical reaction between rFeCl3/[A336]Cl and H2S can be considered reversible. According to the literature [22,30], the chemical absorption of H2S in rFeCl3/[A336]Cl is expressed by Eq. (14).

In Eq. (14), subscript s denotes solid phases. rFeCl3/[A336]Cl is considered a virtual molecule in the form of Fe(III)-IL. Reduced rFeCl3/[A336]Cl is expressed as Fe(II)-IL. The Fe(II)-IL has similar properties to Fe(III)-IL, so the mixture of rFeCl3/[A336]Cl + H2S can be regarded as an approximate binary solution.

Based on the above assumption, Henry’s equation of H2S in r FeCl3/[A336]Cl is expressed by the Eq. (15). Since the equilibrium pressure does not exceed 200 kPa, the fugacity of H2S is replaced by partial pressure in Henry’s equation.

where, PH2S(kPa) is the equilibrium partial pressure of H2S. The vapor pressure of the ionic liquid is low. Thus, the gas phase in the equilibrium kettle can be considered as pure H2S without solvent.mH2S(mol H2S/kg IL,mol?kg-1)is the molality of the physically dissolved H2S in liquid phase.m0(1 mol H2S/kg IL, mol?kg-1) is the standard molality.γH2Sis the activity coefficient of H2S in the liquid phase. Happ(kPa) is an apparent Henry’s constant based on the assumption of a binary system.

The chemical absorption equilibrium of H2S in ionic liquid is expressed in Eq. (14). The equilibrium constant Kappis given by Eq. (16).

where mHFe(II)-ILand mFe(III)-ILare the molality of reacted and unreacted ionic liquids, respectively.γHFe(II)-ILand γFe(III)-ILare the activity coefficients corresponding to mHFe(II)-ILand mFe(III)-IL,respectively. It should be pointed out that rFeCl3/[A336]Cl is composed of FeCl-4, [A336]Cl-and Cl-ions. That is to say, the mixture of rFeCl3/[A336]Cl + H2S is a multicomponent solution. Therefore,Kappis also an apparent property.

The mass balance of H2S and ionic liquids are expressed by Eq.(17) and Eq. (18), respectively.

where, mt(mol H2S/kg IL, mol?kg-1) is the total solubility of H2S.mH2S,r(mol H2S/kg IL, mol?kg-1) is the molality of reacted H2S. mFe(II)-IL(mol/kg IL, mol?kg-1) and mHFe(II)-IL(mol/kg IL, mol?kg-1) are the molality of unreacted ionic liquid and reacted ionic liquid,respectively. According to Eq. (14), mH2S,r=1/2 mHFe(II)-IL. mIL0is the initial concentration of the ionic liquid based on mass, which can be calculated from Eq. (19).

where MIL0(kg?mol-1) is the molar mass of Fe(III)-IL. The molar mass of iron-based ionic liquids with different ratios r is shown in Table 2.

Under the experimental conditions,Fe(II)-IL has a similar structure and properties to Fe(III)-IL,so the liquid phase can be considered an ideal solution. The concentration of physically dissolved H2S molecules in the solution is low.Thus,the activity coefficients γH2S,γ(Fe(III)-IL,and γHFe(II)-ILare all 1.From Eqs.(15)-(19),the RETMfor the absorption process of hydrogen sulfide by rFeCl3/[A336]Cl is given by Eq. (20).

Table 2 The molecular weight of iron-based ionic liquids:

Taking PH2Sand mtas fitting data, Henry’s constant (Happ) and the chemical reaction equilibrium constant (Kapp) are obtained by ORIGIN software from the least squares fit of mtwith PH2S.

4. Results and Discussion

4.1. Validation of the experimental method

The liquid gravity balance for density measurement was validated with pure water at atmospheric pressure and temperatures of 303 to 353 K. The results of the validation experiments were compared with data in the literature [41]. The results are shown in Table 3.In Table 3,RD is the relative deviation,which is defined in the Supplementary Material.It can be seen that the average relative error (ARD) is 0.0880%, and the maximum relative error(MRD) is 0.1017%. The digital viscometer for viscosity measurement was verified with glycerol at atmospheric pressure and temperatures of 323 to 365 K.The results of the validation experiments were compared with data in the literature [42]. The verification results are shown in Table 3. It can be found that ARD is 0.996%and MRD is 1.16%.

To verify the reliability of the solubility measurement device,the absorption of H2S in pure water at a pressure range from 50 to 500 kPa was performed at 313.15 K, 323.15 K, and 333.15 K,respectively. To avoid solvent loss, the pressure after vacuumizing the device should be greater than the vapor pressure of water.The experimental steps are the same as those for determining the H2S solubility in the ionic liquid.The data of the validation experiment are listed in Table 4,where Ptotis the total pressure of H2S and H2O.The results of the validation experiments were compared with calculations (the thermodynamic method is SRK) by Aspen Plus software (AspenONE?for Universities, Aspen Technology, Inc.,Bedford, MA, USA) and data in the literature [43]. The results are shown in Fig. 2. The ARD% between the literature data and Aspen Plus data is 0.74%, and ARD% between the experimental data and Aspen Plus data is 0.62%. The results show that the experimental device is reliable.

Table 3 The density of water and viscosity of glycerol at (101.3±0.4) kPa

Table 4 H2S solubility in pure water at T = 313.15 K, 323.15 K, and 333.15 K

Fig. 2. Solubility of H2S in pure water (50–500 kPa).

4.2. Density

Under atmospheric pressure, the density data of rFeCl3/[A336]Cl obtained through experiments within the temperature range of 313.15 to 358.15 K are shown in Table 5. Considering the effect of purity on the uncertainty of density measurement[44], the standard relative uncertainty of density of rFeCl3/[A336]Cl is ±0.1%. The variation of experimental density with the temperature is shown in Fig. 3. It can be found that the density of rFeCl3/[A336]Cl decreases with the decrease of iron content. The density of FeCl3is greater than that of[A336]Cl.As the iron content increases,that is,the molar ratio of FeCl3increases,and the density of rFeCl3/[A336]Cl increases.Also it can be seen from Fig.3 that the density of rFeCl3/[A336]Cl decreases with the increase of temperature. With the increase of temperature, the molecular spacing increases, which leads to the decrease of the density of rFeCl3/[A336]Cl. The density variation trend is consistent with the density law of FeCl3/[Bmim]Cl ionic liquid studied by Li et al.[30]. The fitting results using Eq. (11) are shown in Table 6.

Fig. 3. Variation of density of rFeCl3/[A336]Cl with the temperature at 101.13 kPa ± 0.4 kPa.

Table 6 Parameter values of fitting Eq. (11) for density

It can be seen from Table 6 that the prediction value has a good correlation with the experimental data of rFeCl3/[A336]Cl. R2is greater than 0.99, ARD is not more than 0.07%, and MRD is not more than 0.12%. That means the empirical Eq. (11) can be used to predict the density of rFeCl3/[A336]Cl.

4.3. Viscosity

At atmospheric pressure, the viscosity data of rFeCl3/[A336]Cl measured within the temperature range of 313.15 to 358.15 K are shown in Table 7. The variation of viscosity with temperature and molar ratio(r)is shown in Fig.4.It can be seen that the viscosity of rFeCl3/[A336]Cl decreases with the increase of r, and the increase of temperature. The same conclusions were reported by Li et al. [30], who studied the effluence of iron content on the viscosity of FeCl3/[C4mim]Cl ionic liquid.

When the temperature is lower than 333.15 K, the viscosity changes significantly with temperature. For example, when the temperature increases from 313.15 K to 323.15 K, the viscosity of rFeCl3/[A336]Cl with r = 0.1, 0.2, and 0.4 decreases by 50.9%,45.2%, and 44.7%, respectively. When the temperature is higher than 333.15 K, the viscosity changes slowly with temperature.

Table 5 Density of rFeCl3/[A336]Cl at T = 313.15 to 358.15 K at (101.3 ± 0.4) kPa

Table 7 Viscosity of rFeCl3/[A336]Cl at (101.3 ± 0.4) kPa

Table 8 shows that the VFT empirical equation has a good correlation with the temperature variation of the viscosity for rFeCl3/[A336]Cl. R2is greater than 0.999. ARD is not more than 0.7%, and MRD is not more than 1.6%.

4.4. H2S solubility

4.4.1. Measurement data

The gas–liquid equilibrium data of H2S and rFeCl3/[A336]Cl(r = 0.1–0.8) was measured at temperatures of 318.15 to 348.15 K and pressures of 0 to 150 kPa. The results are shown in Table 9 and plotted in Fig.5.It can be seen that the total solubility of H2S increases with pressure increasing and temperature decreasing. This is consistent with the solubility of most gases.When the pressure is lower than 50 kPa, the relationship between solubility and pressure is nonlinear, indicating that chemical absorption has a great influence.When the pressure is greater than 50 kPa, the solubility and pressure show an approximately linear relationship, indicating that physical absorption has a great influence.

Fig. 4. Variation of viscosity of rFeCl3/[A336]Cl with the temperature at 101.13 kPa ± 0.4 kPa.

Table 8 Parameter values of fitting Eq. (12) for viscosity

Table 9 Experimental solubility data of H2S in rFeCl3/[A336]Cl: PH2S, partial pressure of H2S in the gas phase; mt, molality of H2S in liquid phase; u(mt) is standard uncertainty of mt; r,molar ratio of FeCl3 to [A336]Cl

Fig. 5. Variation of H2S partial pressure with molality of H2S dissolved in rFeCl3/[A336]Cl. (a) r = 0.1; (b) r = 0.2; (c) r = 0.4; (d) r = 0.6; (e) r = 0.8.

4.4.2. The RETM

The lines in Fig. 5 were calculated values using the RETM.Table 10 presents the fitting parameters of Henry’s constant(Happ)and chemical reaction equilibrium constant (Kapp). The ARD of the RETM for prediction is less than 1.3%, and the MRD is less than 3.4%. From Fig. 5 and Table 10, it can be seen that the RETM can predict the solubility of H2S in rFeCl3/[A336]Cl well.

Henry’s constant can be used to characterize the physical solubility of H2S in rFeCl3/[A336]Cl. Fig. 6 shows the variation of Henry’s constant(Happ)with the mole ratio(r)at different temperatures.It can be observed that Henry’s constant increases with the increase of r.From Fig.3 in Section 4.2,it can be observed that the larger r is,the greater the density is.That is to say,the free volume of the solution decreases with the increase of r,so fewer H2S molecules can be accommodated [45,46]. Also can be seen from Fig. 6 that Henry’s constant increases with the increase of temperature.Elevated temperature leads to the faster molecular motion of H2S, which makes H2S difficult to be immobilized in the liquid phase.

The chemical absorption of H2S in rFeCl3/[A336]Cl is characterized by the chemical reaction equilibrium constant. Fig. 7 shows the influence of the mole ratio (r) on the chemical reaction equilibrium constant (Kapp). It can be seen that Kappincreases with the increase of temperature, indicating that the forward chemical reaction that forms sulfur is endothermic. When r is less than 0.4, the concentration of FeCl-4in rFeCl3/[A336]Cl increases with the increase of r, and the oxidizing ability of ionic liquids becomes larger, and the chemical reaction equilibrium constant Kappincreases.

However, in addition to oxidizing properties, r also affects the acidity of rFeCl3/[A336]Cl.rFeCl3/[A336]Cl is hydrophobic,so water cannot be used as a solvent for pH determination.According to the literature [47,48], ethanol can be used as the solvent to prepare 0.01 mol?L-1ionic liquid-ethanol solution. pH values were measured directly by pH-meter (PHS-25, Shanghai Yueping Scientific Instrument Co.,Ltd.)with a pH composite electrode(65-1C,Shanghai Yi Electrical Scientific Instrument Co., Ltd.). The measurement accuracy is ±0.01 pH. Before reading the value, the pH standard buffer solution was used to calibrate.The pH values of ionic liquid are shown in Table 11.It can be seen that with the increase of r,the pH of ionic liquid decreases, which means the acidity of the ionic liquid increases. This is due to the strongly acidic anions of[21]. Therefore, with the increase of r, the acidity of rFeCl3/[A336]Cl increases.The strong acidity is not conducive to the combination of rFeCl3/[A336]Cl with acidic molecule H2S, which leads to the decrease of Kapp. This is consistent with conclusions on the effect of r on the acidity of imidazole-based Fe-ILs in the literature[49].

The joint action of the above two factors makes Kappincrease first and then decrease with the increase of r, and the maximum value is reached when r = 0.4, as shown in Fig. 8.

Table 10 Henry’s constant (Happ) and the chemical reaction equilibrium constant (Kapp) were obtained from the RETM

Fig. 6. The influence of r on Happ.

Fig. 7. The influence of r on Kapp.

Table 11 The pH values of rFeCl3/[A336]Cl ethanol solution at (298.15±0.03) K and(101.3±0.4) kPa

To understand the influence of the chemical reaction and the physical dissolution on the total solubility of H2S in rFeCl3/[A336]Cl, the amount of physical absorption and chemical absorption can be calculated according to Eq. (15) and Eq. (17),respectively. The chemical absorption isotherms and physical absorption isotherms of H2S in rFeCl3/[A336]Cl with r = 0.1 are plotted in Fig. 8. Chemical absorption isotherms and physical absorption isotherms with other r are given in the Supplementary Material.It can be seen from the figures that there is a linear relationship between the physical absorption of H2S and the partial pressure of H2S. When H2S partial pressure is low, the chemical absorption curve is higher than the physical absorption curve,and the chemical absorption is dominant. When the partial pressure of H2S reaches a certain value, the physical absorption curve is higher than the chemical absorption curve.Therefore,under high pressure, H2S is mainly absorbed physically.

4.4.3. Effect of r on total solubility of H2S

In addition to the influence of temperature and pressure on the solubility of H2S in ionic liquid, the molar ratio (r) of FeCl3to[A336]Cl is also an important factor. At a certain temperature,the variation of the total solubility of H2S with r is plotted in Fig. 9. It can be seen the total solubility of H2S decreases with the increase of r in the whole pressure range (from 0 to 150 kPa)when r ranges from 0.2 to 0.8.However,when r=0.1 it presents a different pattern.With the increase of pressure,the total solubility of H2S is first lower than that of r=0.2,and then higher than that of r = 0.2. This is because the absorption of H2S in 0.1FeCl3/[A336]Cl has a smaller Henry’s constant and chemical reaction equilibrium constant, which can be observed in Table 10.

Fig. 8. Physical and chemical absorption curves of H2S in 0.1FeCl3/[A336]Cl. (a)T = 318.15 K; (b)T = 328.15 K; (c)T = 338.15 K; (d)T = 348.15 K.

Fig. 9. The influence of r on the total solubility of H2S. (a)T = 318.15 K; (b)T = 328.15 K; (c)T = 338.15 K; (d)T = 348.15 K.

Shah et al.[1]sorted out H2S content in natural gas,biogas,syngas,and landfill gas.Natural gas had the highest H2S content with a molar fraction of about 16%. Chemical absorption units generally operate at atmospheric pressure,so the maximum partial pressure of H2S does not exceed 20 kPa. It can be seen from Fig. 9, when r = 0.2, rFeCl3/[A336]Cl has the maximum total solubility under the H2S partial pressure in the range of 0–20 kPa. This is because with the increase of r, for example, from 0.2 to 0.4, chemical reaction equilibrium constant increases. However, the corresponding Henry’s constant increases at the same time. The increase of Henry’s constant indicates that the amount of physically dissolved H2S is reduced.Therefore,the amount of chemical absorption may not increase, which results in a reduction in total solubility.

4.5. Compared to other iron-based ionic liquids

Fig. 10. Solubility of H2S in two Fe-ILs at r = 0.6. (a) T = 318.15 K; (b) T = 348.15 K.

Different anions and cations can form ionic liquids with different properties.Therefore,it will directly affect the solubility of H2S in ionic liquids, Cheng et al. [31] measured the solubility of H2S in iron-based ionic liquid rFeCl3/[bmim]Cl(r=0.6,0.8,1.0,1.2,1.4)at temperatures of 303.15 to 348.15 K and pressures of 100 to 1000 kPa. And corresponding Henry’s constant and chemical reaction equilibrium constant were calculated. Therefore, the RETM was used to calculate the solubility of H2S in rFeCl3/[A336]Cl and rFeCl3/[bmim]Cl at temperatures of 318.15 K and 348.15 K to compare the difference of H2S absorption.For r=0.6,the comparison of H2S solubility are plotted in Fig.10.The chemical reaction equilibrium constant and Henry’s constant of two iron-based ionic liquids are shown in Figs. 11 and 12, respectively. For r = 0.8, results are given in the Supplementary Material.

Fig.11. The chemical reaction equilibrium constant(Kapp)of two Fe-ILs with r=0.6.

Fig. 12. Henry’s constant (Happ) of two Fe-ILs with r = 0.6.

It can be seen from Fig. 10 that the solubility of H2S in rFeCl3/[A336]Cl is greater than that of rFeCl3/[bmim]Cl at the temperature of 318.15 K.This is due to that rFeCl3/[A336]Cl has greater Kappand smaller Happ, which means rFeCl3/[A336]Cl has greater reaction conversion and greater physical solubility.But at 348.15 K,Henry’s constant and equilibrium constant of rFeCl3/[A336]Cl are both larger than that of rFeCl3/[bmim]Cl.Less physical solubility results in lower concentrations of reactants,which limits the amount of product from a chemical reaction. The difference of H2S solubility of ionic liquids with temperature may be related to the different cationic structures of ionic liquids.

The raw material price of rFeCl3/[A336]Cl is much lower than that of rFeCl3/[bmim]Cl. In the actual wet desulphurization process, absorption is carried out under atmospheric pressure and normal temperature (lower than 318.15 K). Therefore, rFeCl3/[A336]Cl has a better prospect of industrial application than rFeCl3/[bmim]Cl.

5. Conclusions

A new absorbent FeCl3/[A336]Cl was proposed to absorb hydrogen sulfide from gas. The density and viscosity of rFeCl3/[A336]Cl(r = 0.1–0.8) were measured at temperatures of 313.15 to 358.15 K and atmospheric pressure. Temperature-dependent empirical models correlate well with measurement data. The gas–liquid equilibrium data of H2S in rFeCl3/[A336]Cl was measured at temperatures from 318.15 to 348.15 K and pressures from 0 to 150 kPa. Henry’s constant and chemical reaction equilibrium constant can be obtained by the RETM fitting,which can satisfactorily predict experimental results. For most gas desulphurization with H2S partial pressure not more than 20 kPa, the maximum total solubility of H2S in rFeCl3/[A336]Cl is obtained at r=0.2.Compared to rFeCl3/[bmim]Cl ionic liquid, rFeCl3/[A336]Cl has higher H2S solubility and lower price, which makes rFeCl3/[A336]Cl have a better potential for industrial application.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Financial support from the National Natural Science Foundation of China(21775081)and Shandong Province Natural Science Foundation (ZR2020MB145) is gratefully acknowledged.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2022.11.012.