An organic macromolecule, poly(1-vinylimidazole), with an appropriate polymerization degree was proposed and mixed with water to form a novel aqueous absorbent for SO capture. This aqueous solution absorbent has the advantages of simple preparation, good physicochemical properties, environment-friendliness, high ability in deep removal of SO , and excellent reusability. Moreover, pH-responsive behavior, pH buffering absorption mechanism, and their synergistic effect on absorption performance were revealed. The solubilities of SO in the absorbent were measured in detail, and the results demonstrated excellent absorption capacity and recyclability. Then, mathematic models that describe SO absorption equilibrium were established, and the corresponding parameters were estimated. More importantly, on the basis of model and experimental data, the absorption and desorption could maintain high efficiency within a wide operating region. In summary, this work pro - vided a low-cost, efficient, and unique absorbent for SO capture and verified its technical feasibility in industrial application. Keywords Poly(1-vinylimidazole) · pH buffer · pH responsibility · SO capture · Flue gas desulfurization Introduction application of dry desulfurization. Limestone-gypsum as a kind of absorbent in one of the wet desulfurization methods Combustion of fossil fuels leads to a large amount of sulfur has been widely used in practice [7–9]. However, limestone- dioxide (SO ) emission, which causes serious environmental gypsum has various inherent disadvantages, such as the problems and is harmful to human health . Therefore, a production of a large amount of wastewater, high operating crucial task is to reduce SO emission through flue gas des - cost, useless byproducts that cause secondary pollutants, and ulfurization (FGD) on account of the principles of sustain- intensive energy consumption. Another widely adopted wet able development and green chemistry . desulfurization method is the use of organic amine, such as Among all the FGD technologies, wet desulfurization [3, ethylenediamine , as absorbent in FGD. This method has 4] and dry desulfurization [5, 6] are widely used in prac- a high desulfurization efficiency, and SO can be desorbed tice. The efficiency of dry desulfurization is lower than that from the system by heating so that the absorbents can be of wet desulfurization, thereby preventing the large-scale recycled. N-methylimidazole and N-methylpyrrolidone also showed great performance in SO absorption . However, the large loss of the absorption agent is inevitable because the organic solvents will volatilize into the gas stream, espe- Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s1220 9-018-0168-0) contains cially in the desorption process. The development of renew- supplementary material, which is available to authorized users. able and ec ffi ient absorbents with the absence of byproducts is highly valued for industrial application. * Lühong Zhang Recently, ionic liquids (ILs) have attracted widespread email@example.com attention as promising absorbents for SO removal and cap- * Huawei Yang ture due to their unique properties [12–14], such as neg- firstname.lastname@example.org ligible vapor pressure, high thermal stability, and tunable School of Chemical Engineering and Technology, Tianjin structure. Wu et al.  first reported an IL—1,1,3,3-tetra - University, Tianjin 300350, China methylguanidinium lactate ([TMG]L) for SO absorption School of Chemistry and Materials Science, Ludong from a gas mixture of SO and N , and the result showed 2 2 University, Yantai 264025, China Vol.:(0123456789) 1 3 W. Feng et al. that 1 mol [TMG]L could selectively capture 0.978 mol SO (25.00 g) and cyclohexane (15.00 g) with a certain amount at 40 °C. Later, many other ILs, including hydroxyl ammo- of AIBN as initiator. The molecular weight of PVIM was nium ILs [16, 17], ether-functionalized ILs [18–20], imida- regulated by changing the dosage of AIBN, and the molar zolium-based ILs [21–23], guanidinium-based ILs [24–26], ratio of AIBN to VIM was selected as 3, 7, and 10%, respec- pyridinium-based ILs [27, 28], anion-functionalized ILs [29, tively. The reaction was placed in a preheated oil bath at 30], and, more recently, deep eutectic solvents [31–33], were 75 °C with magnetic stirring under nitrogen atmosphere investigated for SO absorption. for 24 h. After the reaction, the mixture was precipitated Absorption capacity and absorption rate are two impor- into acetone, and the solid was separated by filtrate. Then, tant aspects in assessing the absorption performance of the solid was purified by dissolution in methanol and pre- absorbents. Despite the high absorption capacity of ILs, cipitation with acetone. The steps were repeated two times. the high viscosities of most task-specific ILs induce the low Finally, the polymer was dried at 80 °C under vacuum for heat and mass transfer performance of the SO absorption 24 h before use. process [34, 35]. In addition, water and O are contained in 2− flue gas; thus, part of SO can be oxidized into SO in the Characterization of PVIM 2 4 2− FGD process. Moreover, SO will destroy the structure of the ILs and further affect the reusability of the absorbent [36 , A series of PVIM with different molecular weights was 37]. These limitations highlight the significance of develop- prepared in this work. Absolute polymer molecular weights ing new absorbents that have low viscosity and can adapt to were determined using aqueous size exclusion chroma- the complex compositions of flue gas. tography using Malvern’s OMNISEC and summarized in To take advantage of the non-volatility of ILs and the Table 1. The chemical structure of the as-prepared PVIM applicability of traditional organic amines, poly(1-vinylim- was confirmed by H nuclear magnetic resonance (NMR) idazole) (PVIM), an organic macromolecule with high ther- spectra and Fourier transform infrared (FTIR) spectra. H mal stability, was proposed and blended with water to form NMR spectra were measured on a Bruker DPX 500 MHz pH buffering aqueous solutions for SO capture. The aque- spectrometer using DMSO as a solvent, and the spectra ous absorbents can reversibly and efficiently capture SO are illustrated in Fig. S1. FTIR spectra were recorded on because of the suitable basicity of each repeated imidazole a NEXUS870 FTIR spectrometer. The thermal stability of group in PVIM . The absorptive/desorptive capability PVIM was investigated using TGA/DSC (STAR System, and the reusability of the absorbents were investigated in Switzerland) from 0 to 500 °C at a scan rate of 10 °C/min detail. Then, equilibrium models were built according to the under N atmosphere, and the decomposition temperature absorption mechanism. Parametric analysis was performed was determined according to the thermogravimetric analysis based on the models and the experimental data, and correla- (TGA) and differential thermal analysis (DTA) curves. tive thermodynamic parameters were obtained. The appli- cation feasibility of this absorbent was examined using the Preparation of SO Absorbents and Determination equilibrium models and the parameters. of Physical Properties A range of absorbents was prepared by dissolving PVIM Experimental Section (30 wt%) with different molecular weights in water (70 wt%). The viscosities were measured using a Brookfield Materials DV-II + Pro viscometer, which was supplied by Brookfield Engineering Laboratories. The temperature was controlled SO gas (≥ 99.9%) and N gas (≥ 99.99%) were supplied by a water bath accurate to ± 0.1 K. The viscosities of these 2 2 by Tianjin Shengtang Specialty Gases Co., Ltd. Ethanol (≥ 99.9%), acetone (≥ 99.9%), cyclohexane (≥ 99%), and azodiisobutyronitrile (AIBN, 99.9%) were purchased from Table 1 Viscosities of the absorbents with different molecular weights Aladdin Industrial Corporation. 1-vinylimidazole (VIM) was obtained from Shanghai Meryer Chemical Technology Co., Molar ratio of AIBN to Molecular weight of Viscosity (cP) VIM (%) PVIM Ltd. 3 5000 307.6 Synthesis of Poly(1‑vinylimidazole) 7 4000 194.7 10 3100 131.6 PVIM was synthesized via conventional free radical polym- erization . In a 100 mL round-bottomed flask, VIM The absorbents were aqueous solution of PVIM (30 wt%) (10.00 g) was dissolved in a binary mixture of ethanol The viscosities were measured at 298.2 K 1 3 pH‑Responsive and Buffering Macromolecule Aqueous Absorbent and Mathematic Model‑Based… absorbents before absorption at 298.2 K are listed in Table 1. for the equilibrium chamber. The SO uptake ( n P ) could 2 s The uncertainty of viscosity measurements was estimated be calculated using the following formula: to be ± 1%. n P = P , T V − P , T V − P , T V − V s g 1 1 g 1 g s 2 a (1) Absorption of SO where (P, T) , which is obtained from NIST standard refer- ence data , represents the density of SO in mol/cm at The measuring methods employed to determine the absorp- 2 P and T; V is the volume of the absorbent in cm at T; and tion of SO were similar to those used in other studies, a P , T is the density of SO gas remaining in the equilib- and the diagram of the experimental apparatus is shown g s 2 rium chamber when the absorption equilibrium was reached, in Fig. 1 . The whole test apparatus includes absorp- and it could be calculated by the following equations: tion and recording sections. The absorption section consists of two stainless steel chambers, the volumes of which are 3 3 P , T = − (2) 161.752 cm (V ) and 88.165 cm (V ), respectively. The g s 12 1 1 2 bigger chamber, which is called gas reserve, isolates SO before it contacts with the absorbent in the smaller chamber. y = P , T (3) 2 g s 12 The smaller chamber, which is called equilibrium chamber, is equipped with a magnetic stirrer. The temperature of the where ρ and ρ represent the density of the mixed gas and 12 1 chambers is controlled by a water bath with an uncertainty N in the equilibrium chamber, respectively; y is the mole 2 2 of ± 0.1 K. The recording section includes two pressure sen- fraction of SO in the mixed gas, and ρ is a function of 2 12 sors of ± 0.2% uncertainty (in relation to the full scale of y . The value of ρ could be calculated using a generalized 2 1 0–300 kPa), which are connected to a numeric instrument second virial coefficient ; the formula is as follows: to record the pressure changes of the two chambers online. BP P c c In a typical absorption procedure, a known mass (w) of Z = 1 + (4) RT T solvent was placed in the equilibrium chamber and nitro- c c gen was then purged into the chamber to exclude the air. BP After a sufficient amount of time, the initial pressure in the (0) (1) = B + B (5) equilibrium chamber was noted as P . The air in the gas RT 0 c reserve was exhausted (< 10 Pa), and then the gas reserve (0) 1.6 received a certain amount of SO from the gas cylinder. The B = 0.083 − 0.422 T (6) pressure was measured as P , which was much larger than (1) 4.2 P . The needle valve between the two chambers was turned B = 0.139 − 0.172 T (7) to introduce a certain amount of SO into the equilibrium = P Z RT chamber. The absorption equilibrium was considered to have (8) 1 0 1 been achieved when the pressure of the equilibrium chamber where Z represents the compressibility factor of N ; P is 1 2 0 stayed the same for at least 1 h. At this time, the pressure of the initial pressure of N before SO was introduced; and R 2 2 the two cells was recorded as P for the gas reserve and P Fig. 1 Diagram of the experimental apparatus of SO absorption: constant temperature water bath; (8) residual gas absorption bottle (1) N gas cylinder; (2) SO gas cylinder; (3) vacuum pump; (4) gas (sodium hydroxide solution); (9), (10) pressure relief valve; (11)–(18) 2 2 reserve chamber; (5) equilibrium chamber; (6) magnetic stirrers; (7) valve; (19), (20) pressure sensors 1 3 W. Feng et al. −1 −1 is the molar gas constant (8.314 J·mol ·K ). The value of introduced to the sodium hydroxide solution. The average ρ could be calculated using the generalized second virial uncertainty in absorption measurements was estimated to coefficient of gas mixture; the formula is as follows: be within ± 2%. 2 2 B = y B + 2y y B + y B (9) M 11 1 2 12 22 1 2 pKa Value and Reaction Enthalpy Determination of PVIM y = 1 − y (10) 1 2 First, approximately 1 g PVIM was dissolved in water to RT (0) (1) form a dilute solution. Then, the solution was titrated by HCl B = B + B (11) 12 12 aqueous solution (0.1 mol/L), and pH values were meas- ured using a PHS-3C-01 pH meter. The uncertainty of the √ pH value measurements was ± 0.01. The titration experi- � � T = k − 1 × T T (12) c12 12 c1 c2 ments were conducted under the temperature range from 25 to 70 °C. 1∕3 1∕3 In theory, every imidazole ring of the PVIM can bond V + V c1 c2 V = (13) with one H , thus resulting in pH buffering effects. With the c12 addition of HCl, H can spontaneously bond to the position that has the strongest protonation ability, leading to a change in the pKa value of PVIM. To calculate the pKa value evolu- Z RT c12 c12 P = 0 (14) c12 tion and the molar reaction enthalpy (Δ H ), a mathematic V m c12 model that describes the titration process was built using gPROMS software. The detailed mathematic model is listed Z + Z c1 c2 Z = (15) in the Supporting Information. c12 c12 Desorption of SO and Recycling of the Absorbents 1 2 (16) The diagram of the experimental apparatus for desorp- tion and recycling of SO is shown in Fig. 2. The whole B P apparatus consists of SO and N gas cylinders, a glass gas Z = 1 + (17) 2 2 RT absorption tube (inner diameter of 40 mm) with a reflux con- denser, a constant temperature oil bath with an uncertainty = P ∕ Z RT (18) 12 2 12 of ± 0.1 K, a pH sensor of ± 0.01 uncertainty (in relation to where Z represents the compressibility factor of the gas the full scale of 0–14), and a residual gas absorption bottle. mixture of SO and N in the equilibrium chamber, and The absorption tube is equipped with a magnetic stirrer, and 2 2 k (0.08) is the binary interaction parameter. Relevant the pH sensor is connected to a numeric instrument to record critical data are listed in Table S1. SO was continually the pH evolution of the tube online. introduced into the equilibrium chamber to reach new equi- The absorption process was operated at a temperature librium at different pressures. Unabsorbed SO was then and a pressure of 298.2 K and 101.3 kPa, respectively. In a Fig. 2 Diagram of the experimental apparatus for SO desorption and recycling: (1) N gas cylinder; (2) SO gas 2 2 cylinder; (3) numeric instru- ment; (4) pH sensor; (5) con- stant temperature water bath; (6) absorption tube (equipped with a reflux condenser); (7) residual gas absorption bottle (sodium hydroxide solution); (8), (9) pressure relief valve; (10)–(13) valve; (14), (15) rotor flow meter 1 3 pH‑Responsive and Buffering Macromolecule Aqueous Absorbent and Mathematic Model‑Based… typical procedure, a certain amount of absorbents (30–35 g) Table 2 Viscosities of the absorbents with various pH values was charged into the gas absorption tube. After the air tight- pH Viscosity (cP) ness of the experimental apparatus was checked, N gas was 298.2 K 303.2 K 318.2 K released at a rate of 30 mL/min for 30 min to drive away the air in the experimental apparatus. Then, N gas was cut 9.1 131.6 126.5 119.3 off by closing the valve, and SO gas was bubbled through 7.2 16.43 15.62 14.73 the gas absorption tube at a rate of 40 mL/min. The weight 5.0 2.42 2.03 1.85 changes of the tube combined with the absorbent during the 2.2 1.44 1.24 1.16 absorption process were monitored using an electron ana- lytical balance (Precision & Scientific FA2004, Shanghai, China) with an accuracy of ± 0.0001 g, and the molar ratio of absorbed SO to VIM could be measured. The absorption 100 0.0 equilibrium was considered to be reached when the pH value -0.4 no longer changed, and then SO gas was cut off. The desorption of SO from the absorbents was conducted 2 60 375 -0.8 by increasing the temperature of the oil bath to 373.2 K. The desorption equilibrium was considered to be reached when -1.2 the pH value did not change. The experiments of the absorp- tion (298.2 K and 101.3 kPa) and desorption (373.2 K and -1.6 101.3 kPa) cycles were repeated five times to test the reus- 0 0 100 200 300 400500 ability of the PVIM solution. Afterwards, the water in the Temperature ( absorbent was removed by rotary evaporation. Furthermore, to evaluate the absorption/desorption efficiency, another Fig. 3 TGA and DTG curves of PVIM mathematic model that describes the absorption/desorp- tion equilibrium is established and listed in the Supporting Information. to that of water (1.002 cP at 298.2 K) . The reason for this phenomenon was investigated by dynamic light Results and Discussion scattering. The size of micelles formed by PVIM could also respond to pH changes. As shown in Figure S2, the Physical Properties of the Absorbents average size of the micelles in the solution varied from 248.1 to 4.2 nm when the pH values reduced from 9.1 to The molecular weights of PVIM synthesized in different 2.2. This change is suspected to be caused by the protona- conditions and the viscosities of the corresponding absor- tion reaction of the PVIM. Before absorption, PVIM could bents are listed in Table 1. The viscosity of the absorbents associate with a little amount of H ionized by H O, thus under the same concentration decreased with the decrease in resulting in the original solution being alkaline. However, the molecular weights of the PVIM. As is well accepted, low the strong hydrogen bonding interactions between the par- viscosity is favorable to the absorption efficiency of SO . tial protonated polymers would cause a tight association. Thus, the absorbent with the lowest viscosity, which means As a result, the PVIM aqueous solution before absorption the PVIM whose molecular weight is 3100, was used in the exhibited a relatively high viscosity. When the concentra- subsequent experiments. tion of the H continued to increase during the absorption The pH-responsive properties of many polymers were of SO , more imidazole units of PVIM became protonated, frequently reported to have physicochemical properties that which resulted in the weakening of association between can spontaneously change with the variation of pH value in polymers and enhanced the solvation interaction in water. a narrow range [42, 43]. Hence, the pH values and the cor- Consequently, the absorbents possess notable pH-respon- responding viscosities of the absorbents during the absorp- sive properties. Moreover, lower viscosity can enhance tion process at different temperatures ranging from 298.2 to the phase transfer between the gas–liquid two-phase and 318.2 K were measured and listed in Table 2. is favorable to the efficient absorption of SO . Notably, the viscosities not only decreased with the To verify the thermal stability of PVIM, the TGA and increase in temperature but also dramatically decreased DTA curves of the PVIM are shown in Fig. 3. The PVIM when the absorbents changed from alkaline to acidic. For first showed an obvious loss of weight at 375 °C, which example, the viscosity at 298.2 K is 131.6 cP when pH means that the PVIM is stable enough under the operating is 9.1 and 2.42 cP when pH is 5.0, which is comparable conditions. 1 3 Weight (%) Derivative weight W. Feng et al. Absorption of SO 961.39 1544.84 Absorption Mechanism 2633.70 2744.39 1150.94 Saturated The proposed SO absorption mechanism is illustrated in with SO Scheme 1. Primarily, the gaseous SO was dissolved in water, and the dissolved SO could form H SO through 2 2 3 a hydration reaction. Then, H SO was ionized to HSO , 2 3 3 2− + Fresh SO , and H , while the imidazole groups in each repeated + 0 unit could effectively bond the H through a protonated reac- 3500 3000 2500 2000 1500 1000 tion, thus significantly promoting the solubility of SO in the -1 Wavenumber (cm ) absorbent. To confirm the proposed absorption mechanism, the FTIR spectra of the absorbents before and after SO Fig. 4 FTIR spectra of fresh absorbent and absorbent saturated with absorption were studied, and the results are shown in Fig. 4. SO As can be seen from Fig. 4, three new peaks appeared at −1 961, 1151, and 1545 cm after the absorption of SO . The −1 low pressure (0–10.0 kPa), whereas it increased gradually absorption bond at 961 cm corresponds to S–O stretching 2− − and almost linearly under high pressure (10.0–100.0 kPa). vibration in SO , HSO , or similar species, thereby indi- 3 3 For example, the SO absorption capacity at 298.2 K is cating the chemical interactions between SO and the mixed −1 0.431 mol SO per kg absorbents at 0.3 kPa, 1.719 mol absorbent . The other peak at 1151 cm can be assigned SO per kg absorbents at 1.7 kPa, and 2.215 mol SO per to the antisymmetric stretch of the dissolved SO , while the 2 2 −1 kg absorbents at 9.4 kPa. The large SO solubility at low peak at 1545 cm could be attributed to stretches of C–C pressures can be primarily attributed to chemical absorp- and C–N in the ring after protonation . Furthermore, a −1 tion according to the proposed absorption mechanism. The series of dispersed bonds in the region of 2800–2400 cm strong bonding of H with PVIM could enhance the ioni- after absorption could be assigned to NH . In sum, zation of H SO and significantly improve the content of the results of FTIR clearly prove the proposed absorption 2 3 − 2− HSO and SO in the system, while the linear increment mechanism. 3 3 of SO solubility at high pressures was due to the complete protonation of PVIM. Absorption Capacity In view of the relatively low SO partial pressure in flue gas, a strong chemical interaction is essential, while the reac- To investigate the absorption capability of the absorbent, tion equilibrium is always sensitive to temperature. As can the equilibrium solubility of SO in the absorbent was deter- be seen from Fig. 5, the increase in temperature has a nega- mined against SO partial pressure at temperatures ranging tive influence on the solubility of SO in the absorbent. For from 298.2 to 328.2 K, and the results are shown in Fig. 5 example, the solubility at 99.2 kPa is 3.409 and 2.468 mol (the solubility data are presented in Table S2). A detail that SO per kg absorbent at 298.2 and 328.2 K, respectively. should be noted first is that each absorption equilibrium could be reached within 5 min, thereby indicating the low transfer resistance of the absorbent, which was attributed to 4.0 the reduced viscosities in acidic conditions. The absorption isotherms shown in Fig. 5 demonstrated good absorption 3.5 capability. The solubility of SO in the absorbent increased 2 3.0 drastically with the increasing SO partial pressure under 2.5 2.0 1.5 298.2 K 1.0 313.2 K 328.2 K 0.5 0.0 020406080100 120 SO partial pressure (kPa) Fig. 5 Absorption isotherm of SO by the absorbent at different tem- Scheme 1 Absorption mechanism of SO in the absorbents peratures 1 3 / kg absorbent mol SO Transmittance (%) 2 pH‑Responsive and Buffering Macromolecule Aqueous Absorbent and Mathematic Model‑Based… This phenomenon was consistent with the absorption mecha- Given that the protonation reaction is an equilibrium process, nism we presented and the results obtained in most cases the H could spontaneously bond with the imidazole units [48, 49]. Low temperature is not only beneficial to physical that have the strongest bonding ability. Therefore, for a single absorption but also shifts the protonation reaction equilib- PVIM molecule with a certain protonation degree, its corre- rium toward a positive direction. The results also indicated sponding pKa value is constant (all the PVIM are assumed to that the absorbed SO could be released at high temperature. have the same molecular weight). Thus, the average protona- In this absorbent, PVIM, as the major absorption com- tion degree of the PVIM is defined as: ponent, could achieve the effective capture of SO through = c +∕ c + c + (19) VIMH VIM VIMH the pH buffering effect, whereas water, a green solvent, At a certain ω, the calculated pKa value can be approxi- was employed to realize the dissolution of PVIM and pro- mately considered as the pKa value of a single PVIM molecule vide the reaction environment for the absorption. Given its with the same protonation degree. easy preparation, low cost, and environment-friendliness, Through parametric analysis of the experimental results, the absorbent can be regarded as a promising candidate for the pKa values of the PVIM variation with the change in the application. Moreover, considering that SO capture is a protonation degree at different temperatures were calculated, continuous process of absorption–desorption, the suitable and the results are illustrated in Fig. 7. The pKa values of the alkalinity of PVIM, which is the key point to enable the PVIM decreased with the reduction of pH values of the aque- desorption at high temperature, was investigated in detail in ous solution and with the increase in the protonation degree of the following sections. the PVIM. As can be seen in Fig. 7, the pKa value decreased from 5.98 before the titration experiments to 4.46 when the Thermodynamic Analysis protonation degree was 100% at 298 K. Thus, the pKa value of the PVIM ranges from 4.46 to 5.98. The pKa value is known To calculate the reaction equilibrium constant (K), acidity to reflect the ability to dissociate H . The decrease in the pKa coefficient (pKa), and the thermodynamic parameters of the value along with the increase in the protonation degree indi- protonation reaction of PVIM, acid titration experiments at cates that the interaction between H and PVIM becomes different temperatures were conducted, and the experiment weaker. Subsequently, the molar reaction enthalpy, which is results are shown in Fig. 6. A corresponding reaction equi- crucial for the design of industrial processes, was estimated librium mathematic model to describe the acid titration pro- using the van der Hoff equation. Furthermore, the molar Gibbs cess was built using gPROMS software. energy of reaction and molar reaction entropy can be calcu- According to the absorption mechanism, PVIM could pro- lated using Eqs. (20) and (21). mote the capture of SO through the reversible protonation reaction, while, as a macromolecule, the already protonated 0 0 Δ G =−RT ln K (20) imidazole units would influence the bonding ability toward H 0 0 0 of other surrounding units via the inductive effect. Hence, we Δ S = Δ H −Δ G T (21) r r r m m m can reasonably consider that the pKa value of the macromol- With the pKa values integrated into these equations, the ecule would vary with the increment of the protonation degree. 0 0 0 average values of Δ H , Δ S and Δ G are presented in r r r m m m 6.5 298 K 6.0 313 K 328 K 5.5 5.0 4.5 4.0 3.5 0.00.2 0.40.6 0.81.0 1.2 Protonation degree Fig. 6 pH variation during HCl titration process at 298, 313, and Fig. 7 Relationship of pKa value versus protonation degree at differ - (aq) 328 K (solid line: experimental data; dash line: predicted results) ent temperatures 1 3 pKa value W. Feng et al. Table 3 Average molar reaction enthalpy, molar Gibbs energy of reaction, and molar reaction entropy at 298 K 0 −1 0 −1 0 −1 −1 Δ H (kJ mol ) Δ G (kJ mol ) Δ S (J mol K ) r r r m m m PVIMH − 40.30 ± 5 − 34.13 ± 1.14 − 20.69 ± 17.20 1.5 1.0 0.5 Fig. 9 Calculated and measured values of the molar ratio of SO to VIM 0.0 10 1.5 1234 5 Cycle times 1.2 Fig. 8 Reusability of the absorbents in SO absorption for five cycles 0.9 0.6 Table 3. Furthermore, in accordance with the calculated 0.3 parameters and the mathematic model, the pH variation with the addition amount of HCl was obtained and illustrated in 0.0 Fig. 6. The calculated results showed good agreement with 05 10 15 20 25 30 35 40 the experimental data, which demonstrates the accuracy of Time (min) calculated parameters. Fig. 10 Variations of pH and solubility of SO with time in the absorption process Desorption of SO and Reuse of the Absorbent On the basis of previous analysis, the pH value of the system which also proves that the absorption of SO in the mixed is a significant factor in evaluating the absorptive amount of absorbent is highly reversible and the absorbed SO can be SO . Hence, in this work, absorption/desorption capability easily stripped out by heating. and technological feasibility were studied by measuring the According to the absorption mechanism, the absorptive variation of the pH value and the weight change of the absor- amount of the SO in the absorbents is closely related to the bent in the repeated absorptive/desorptive process. Experi- pH value and temperature, and can be calculated using the ments on the absorption and desorption cycles of SO were absorption equilibrium model and the obtained thermody- conducted five times. The absorption or desorption equilib- namic parameters. First, to validate the model, the absorp- rium was considered to be achieved when the pH value of tive amount of SO that varied with the pH changes at 298 K the absorbents no longer changed. After five cycles, almost was calculated and compared with the experimental data. As no decline in the equilibrium absorption capacity of SO shown in Fig. 9, all the measured values are in good agree- occurred (see Fig. 8). ment with the calculated ones, which proves that the absorp- The solubility is 1.21 mol SO per mole VIM for the first tion equilibrium model is reliable. Taking the pH variation time and 1.17 mol SO per mole VIM in the subsequent data into the model, the corresponding absorptive amount cycles. The desorption efficiency could reach 96.6% after could be estimated. the desorption. The slightly incomplete desorption may be The absorption of SO in the absorbents at 298 K over due to the reflux water bringing back a small amount of SO . time was estimated. As shown in Fig. 10, the absorption The pH changes in the absorption and desorption process are capacity had an approximately linear increase with time in shown in the Supporting Information. The curves of the pH the initial 30 min with the corresponding pH value interval change with time are almost the same (see Figs. S3 and S4), of 9.02–1.51, which means that the absorbent could achieve 1 3 SO / VIM molar ratio pH value SO / VIM molar ratio 2 pH‑Responsive and Buffering Macromolecule Aqueous Absorbent and Mathematic Model‑Based… efficient capture of SO in this pH region. However, the 100 absorption efficiency decreased dramatically after 30 min, Regenerated which suggested that the reactive absorption had been com- 80 pleted when the pH value was lower than 1.51. Similarly, the optimum pH value interval in the desorption process could be obtained. The desorption of SO from the absorbents over time was conducted at 383 K under the condition of heating reflux. Fresh High temperature has an inhibiting effect on the protonation reaction of PVIM, while the decrease in the partial pressure of SO in the gas phase caused by steam stripping is the 4000 3500 3000 2500 20001500 1000 500 immediate cause of efficient desorption. With the proceed- -1 Wavenumber (cm ) ing of the desorption process, the desorption driving force decreased, and the pKa value of the PVIM increased gradu- Fig. 12 FTIR spectra of PVIM before use and after regenerating ally. Hence, the desorption rate gradually reduced as time proceeded. As shown in Fig. 11, in the initial 40 min of the desorption process and when the pH value of the system was PVIM, which further proves the excellent stability and reus- ability of the absorbent. lower than 4.80, the desorption efficiency could maintain a relatively high level and the residual amount was less than 10%. Therefore, the pH interval of 1.42–4.80 was considered a proper desorption operation range. Conclusion According to the absorption and desorption curves shown in Figs. 10 and 11, in a continuous operation process, both In this work, a novel macromolecule absorbent that was the absorption and desorption rates could maintain high prepared by blending PVIM and water is first proposed for when the pH value ranged from 1.51 to 4.80. In this region, SO absorption. This pH buffering absorbent combines the the ratio of the available absorptive amount can reach up advantages of the non-volatility of ILs and the applicabil- to 80%, the corresponding gas amount was 2.99 kg per ity of traditional organic amines. Furthermore, the unique mol absorbent, and, more importantly, the viscosity of the pH-responsive characteristic of the absorbent provides it absorbent was comparable to that of water. All these results with a suitable physicochemical property in the absorption proved the applicability of the absorbent in the SO capture process. The results of the absorption and desorption experi- process. ments showed that the absorbent can capture SO efficiently After five absorption–desorption cycles, the water in the at low partial pressure and release SO easily by heating absorbent was removed by rotary evaporation. The residual reflux. As verified by FTIR, the reversible absorption of PVIM was dried under vacuum and analyzed by FTIR. As SO was attributed to the reversible protonation reaction of shown in Fig. 12, no noticeable change can be detected PVIM due to its moderate alkalinity. Moreover, no obvi- after the cycles, as indicated by a comparison with the fresh ous reduction in absorption capacity and chemical structure change were detected after several cycles, which demon- strated the reusability and recyclability of the absorbent. According to the thermodynamic analysis, the pKa value 6 1.5 of the PVIM decreased with the increase in the protona- tion degree because of the inductive effect of the protonated 5 1.2 units, but the pKa value still remained within an applicable 4 0.9 range. With the use of the thermodynamic parameters, a technical feasibility study was conducted, and the absorbent 3 0.6 was proven to be an excellent candidate for SO capture in industrial application. In summary, this research provides an 2 0.3 absorbent that has advantages of simple preparation, good physicochemical properties, environment-friendliness, high 1 0.0 020406080100 120 ability in deep removal of SO , and excellent reusability. The Time (min) unique properties of this absorbent are also revealed. Acknowledgements This study was supported by the National Key Fig. 11 Variations of pH and solubility of SO with time in the des- R&D Program of China (No. 2016YFC0400406). orption process 1 3 pH value SO / VIM molar ratio Transmittance (%) W. Feng et al. Open Access This article is distributed under the terms of the Crea- 19. Wang J, Zeng SJ, Bai L et al (2014) Novel ether-functionalized tive Commons Attribution 4.0 International License (http://creat iveco pyridinium chloride ionic liquids for efficient SO capture. Ind mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- Eng Chem Res 53(43):16832–16839 tion, and reproduction in any medium, provided you give appropriate 20. Yu FL, Liu CY, Yuan B et al (2016) Energy-efficient extractive credit to the original author(s) and the source, provide a link to the desulfurization of gasoline by polyether-based ionic liquids. Fuel Creative Commons license, and indicate if changes were made. 177:39–45 21. Li N, Wang F, Zhang ZQ et al (2014) Synthesis and desulfuriza- tion performance of functional silica gel modified by polymeric 1-vinyl-3-ethylimidazolium tetrafluoroborate ionic liquids. Ind References Eng Chem Res 53(43):16664–16671 22. Firaha DS, Kavalchuk M, Kirchner B (2015) SO solvation in the 1. Kim KH, Jahan SA, Kabir E (2013) A review on human health 1-ethyl-3-methylimidazolium thiocyanate ionic liquid by incor- perspective of air pollution with respect to allergies and asthma. poration into the extended cation-anion network. J Solut Chem Environ Int 59:41–52 44(3–4):838–849 2. Hansen BB, Kiil S, Johnsson JE et al (2008) Foaming in wet flue 23. Jiang XC, Nie Y, Li CX et al (2008) Imidazolium-based alkylphos- gas desulfurization plants: the influence of particles, electrolytes, phate ionic liquids: a potential solvent for extractive desulfuriza- and buffers. Ind Eng Chem Res 47(9):3239–3246 tion of fuel. Fuel 87(1):79–84 3. Cordoba P (2015) Status of flue gas desulphurisation (FGD) sys- 24. Zhang ZM, Wu LB, Dong J et al (2009) Preparation and SO sorp- tems from coal-fired power plants: overview of the physic-chem- tion/desorption behavior of an ionic liquid supported on porous ical control processes of wet limestone FGDs. Fuel 144:274–286 silica particles. Ind Eng Chem Res 48(4):2142–2148 4. Kiil S, Michelsen ML, Dam-Johansen K (1998) Experimental 25. Lu X, Yu J, Wu J et al (2015) Novel guanidinium-based investigation and modeling of a wet flue gas desulfurization pilot ionic liquids for highly efficient SO capture. J Phys Chem B plant. Ind Eng Chem Res 37(7):2792–2806 119(25):8054–8062 5. Dahlan I, Mohamed AR, Kamaruddin AH et al (2007) Dry SO 26. Huang J, Riisager A, Wasserscheid P et al (2006) Reversible removal process using calcium/siliceous-based sorbents: deactiva- physical absorption of SO by ionic liquids. Chem Commun tion kinetics based on breakthrough curves. Chem Eng Technol 38:4027–4029 30(5):663–666 27. Zeng SJ, He HY, Gao HS et al (2015) Improving SO capture by 6. Renedo MJ, Gonzalez F, Pesquera C et al (2006) Study of sorbents tuning functional groups on the cation of pyridinium-based ionic prepared from clays and CaO or Ca(OH) for SO removal at low liquids. RSC Adv 5(4):2470–2478 2 2 temperature. Ind Eng Chem Res 45(10):3752–3757 28. Zeng SJ, Gao HS, Zhang XC et al (2014) Efficient and reversible 7. Kaplan V, Wachtel E, Lubomirsky I (2013) Carbonate melt regen- capture of SO by pyridinium-based ionic liquids. Chem Eng J eration for efficient capture of SO from coal combustion. RSC 251:248–256 Adv 3(36):15842–15849 29. Tian SD, Hou YC, Wu WZ et al (2013) Absorption of SO by 8. Zheng YJ, Kiil S, Johnsson JE (2003) Experimental investigation thermal-stable functional ionic liquids with lactate anion. RSC of a pilot-scale jet bubbling reactor for wet flue gas desulphurisa - Adv 3(11):3572–3577 tion. Chem Eng Sci 58(20):4695–4703 30. Chen KH, Lin WJ, Yu XN et al (2015) Designing of anion-func- 9. Kang Y, Lu J, Guo J (2017) Treatment of wet FGD wastewater by tionalized ionic liquids for efficient capture of SO from flue gas. a modified chemical precipitation method using a solid powder AIChE J 61(6):2028–2034 reagent. Trans Tianjin Univ 23(2):110–121 31. Zhao Y, Hu GX (2013) Removal of SO by a mixture of capro- 10. Tang ZG, Zhou CC, Chen C (2004) Studies on flue gas desul- lactam tetrabutyl ammonium bromide ionic liquid and sodium furization by chemical absorption using an ethylenediamine- humate solution. RSC Adv 3(7):2234–2240 phosphoric acid solution. Ind Eng Chem Res 43(21):6714–6722 32. Yang DZ, Hou MQ, Ning H et al (2013) Efficient SO absorption 11. Huang K, Xia S, Zhang XM et al (2014) Comparative study of by renewable choline chloride–glycerol deep eutectic solvents. the solubilities of SO in five low volatile organic solvents (sul- Green Chem 15(8):2261–2265 folane, ethylene glycol, propylene carbonate, N-methylimidazole, 33. Zhang Q, De Oliveira VK, Royer S et al (2012) Deep eutectic and N-methylpyrrolidone). J Chem Eng Data 59(4):1202–1212 solvents: syntheses, properties and applications. Chem Soc Rev 12. Rogers RD, Seddon KR (2003) Ionic liquids–solvents of the 41(21):7108–7146 future? Science 302(5646):792–793 34. Jin M, Hou Y, Wu W et al (2011) Solubilities and thermo- 13. Severa G, Bethune K, Rocheleau R et al (2015) SO sorption by dynamic properties of SO in ionic liquids. J Phys Chem B 2 2 activated carbon supported ionic liquids under simulated atmos- 115(20):6585–6591 pheric conditions. Chem Eng J 265:249–258 35. Wishart JF (2009) Energy applications of ionic liquids. Energy 14. Heldebrant DJ, Koech PK, Yonker CR (2010) A reversible Environ Sci 2(9):956–961 zwitterionic SO -binding organic liquid. Energy Environ Sci 36. Gao X, Guo RT, Ding HL et al (2009) Dissolution rate of lime- 3(1):111–113 stone for wet flue gas desulfurization in the presence of sulfite. J 15. Wu W, Han B, Gao H et al (2004) Desulfurization of flue Hazard Mater 168(2–3):1059–1064 gas: SO absorption by an ionic liquid. Angew Chem Int Ed 37. Ren SH, Hou YC, Tian SD et al (2012) Deactivation and regenera- 43(18):2415–2417 tion of an ionic liquid during desulfurization of simulated flue gas. 16. Huang K, Lu JF, Wu YT et al (2013) Absorption of SO in aque- Ind Eng Chem Res 51(8):3425–3429 ous solutions of mixed hydroxylammonium dicarboxylate ionic 38. Rallo M, Lopez-Anton MA, Perry R et al (2010) Mercury specia- liquids. Chem Eng J 215–216:36–44 tion in gypsums produced from flue gas desulfurization by tem- 17. Yuan XL, Zhang SJ, Lu XM (2007) Hydroxyl ammonium ionic perature programmed decomposition. Fuel 89(8):2157–2159 liquids: synthesis, properties, and solubility of SO . J Chem Eng 39. Huang K, Chen YL, Zhang XM et al (2014) SO absorption in 2 2 Data 52(2):596–599 acid salt ionic liquids/sulfolane binary mixtures: experimental 18. Hong SY, Im J, Palgunadi J et al (2011) Ether-functionalized ionic study and thermodynamic analysis. Chem Eng J 237:478–486 liquids as highly efficient SO absorbents. Energy Environ Sci 4(5):1802–1806 1 3 pH‑Responsive and Buffering Macromolecule Aqueous Absorbent and Mathematic Model‑Based… 40. National Institute of Standards and Technology. US Department 46. Pekel N, Guven O (2002) Synthesis and characterization of of Commerce (2016) Thermophysical properties of fluid systems. poly(N-vinyl imidazole) hydrogels crosslinked by gamma irra- http://webbo ok.nist.gov/chemi stry/fluid / diation. Polym Int 51(12):1404–1410 41. Smith JM, Van Ness HC, Abbott MM (1996) Introduction to 47. Kuba AG, Smolin YY, Soroush M et al (2016) Synthesis and chemical engineering thermodynamics. McGraw-Hill, New York integration of poly(1-vinylimidazole) polymer electrolyte in dye 42. Fang Y, Xue JX, Ke LY et al (2016) Polymeric lipid vesicles with sensitized solar cells by initiated chemical vapor deposition. Chem pH-responsive turning on-off membrane for programed delivery Eng Sci 154:136–142 of insulin in GI tract. Drug Deliv 23(9):3582–3593 48. Hong SY, Kim H, Kim YJ et al (2014) Nitrile-functionalized ter- 43. Nikouei NS, Ghasemi N, Lavasanifar A (2016) Temperature/pH tiary amines as highly efficient and reversible SO absorbents. J responsive hydrogels based on poly(ethylene glycol) and func- Hazard Mater 264:136–143 tionalized poly(e-caprolactone) block copolymers for controlled 49. Tian S, Hou Y, Wu W et al (2014) Hydrophobic task-specic fi ionic delivery of macromolecules. Pharmaceut Res 33(2):358–366 liquids: synthesis, properties and application for the capture of 44. Kestin J, Sokolov M, Wakeham WA (1978) Viscosity of liq- SO . J Hazard Mater 278:409–416 uid water in the range − 8°C to 150°C. J Phys Chem Ref Data 7(3):941–948 45. Tian SD, Hou YC, Wu WZ et al (2014) Absorption of SO at high temperatures by ionic liquids and the absorption mechanism. Bull Korean Chem Soc 35(9):2791–2796 1 3
Transactions of Tianjin University – Springer Journals
Published: May 28, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera