Effect of Sodium Benzenesulfonate on SO 42 − Removal from Water by Polypyrrole-Modified Activated Carbon

Sodium benzenesulfonate was doped into polypyrrole-modiﬁed granular activated carbon (pyrrole-FeCl 3 · (6H 2 O)-sodium benzene-sulfonate-granular activated carbon; PFB-GAC) with the goal of improving the modiﬁed GAC’s ability to adsorb sulfate from aqueous solutions. At a GAC dosage of 2.5g and a pyrrole concentration of 1mol L − 1 , the adsorption capacity of PFB-GAC prepared using a pyrrole:FeCl 3 · (6H 2 O):sodium benzenesulfonate ratio of 1000:1500:1 reached 23.05mg g − 1 , which was eight times higher than that for GAC and two times higher than that for polypyrrole-modiﬁed GAC without sodium benzenesulfonate. Adsorption was favored under acidic conditions and high initial sulfate concentrations. Doping with sodium benzenesulfonate facilitated polymerization to give polypyrrole. Sodium benzenesulfonate introduced more imino groups to the polypyrrole coating, and the N + sites improved ion exchange of Cl − and SO 42 − and increased the adsorption capacity of sulfate. Adsorption to the PFB-GAC followed pseudo-second-order kinetics. The adsorption isotherm conformed to the Langmuir model, and adsorption was exothermic. Regeneration using a weak alkali (NH 3 · H 2 O), which released OH − slowly, caused less damage to the polypyrrole than using a strong alkali (NaOH) as the regeneration reagent. NH 3 · H 2 O at a concentration of 12mol L − 1 (with the same OH − concentration as 2mol L − 1 NaOH) released 85% of the sorbed sulfate in the ﬁrst adsorption-desorption cycle, and the adsorption capacity remained > 6mg g − 1 after ﬁve adsorption-desorption cycles.


Introduction
e sulfate anion is ubiquitous in water. e main sources of SO 4 2− to water are dissolution of sulfur-containing minerals and atmospheric precipitation [1]. SO 4 2− is not toxic to humans, but drinking water with a high sulfate concentration can cause diarrhea [2], and SO 4 2− at a high concentration causes chronic toxic effects in freshwater organisms [3]. e highly concentrated SO 4 2− that enters a water environment through discharge will diffuse to the sedimentary layer at the bottom of the body of water. e anaerobic environment in the sediment is favorable for sulfate-reducing bacteria to convert SO 4 2− into S 2-or H 2 S gas [4]. e S 2produced in this reaction can combine with various metal ions to trigger metal sulfide precipitation, which reduces the concentrations of trace elements that are necessary for aquatic plants and disrupts the ecological balance of the water environment [5]. e H 2 S generated during this process is also highly toxic, and its presence can cause fish and other aquatic animals to die out in the location, thus ultimately leading the body of water to lose its original ecological function [6]. It is, therefore, extremely important to remove SO 4 2− from wastewater. Currently available methods for removing anions from water include precipitation, adsorption, ion exchange, electrode ionization, membrane treatments, and biological treatments [7,8]. Reverse osmosis membrane technology stands out among these strategies, because it can effectively separate SO 4 2− from water; however, the associated equipment investment is relatively high; the biological methods often require the addition of a carbon source to support the efficient performance of sulfate-reducing bacteria [9]. e adsorption method has attracted the attention of researchers because of its low cost and ease of use. Many novel adsorption materials, such as silicate materials [10], montmorillonite [11], and biosorbent [12], have been developed to separate pollutants from water. Based on the practical engineering applications, activated carbon is still the most common sorbent that is used. Modifying activated carbon with an ammonium salt, a surfactant, or quaternary ammonium increases the number of positive charges on the surfaces and strongly improves the efficiency of adsorption of negatively charged anions from water [10]. However, some (3%-10%) of the nitrogen-containing chemicals that are used to modify activated carbon can leach from the activated carbon during a water treatment process and cause secondary pollution of the aquatic system receiving the treated water [13]. Polypyrrole (PPy), which is a well-known conductive polymer, has been used to modify activated carbon in recent years, to improve the adsorption capacity of the carbon through its ion exchange properties and to prevent secondary pollution [14][15][16].
A conductive polymer is a kind of polymer material formed by chemically or electrochemically doping a polymer with conjugated π-bonds [17]. PPy, which has positively charged nitrogen atoms, can be formed by oxidizing and polymerizing pyrrole (Py) monomers using FeCl 3 ·(6H 2 O).
e PPy structure is conjugated and has alternating C-C and C�C bonds. Chloride ions (Cl − ) incorporated during the oxidation-polymerization process allow PPy to act as an ion exchange material [7,18,19]. PPy has been directly polymerized on activated carbon surfaces to effectively remove anions from water [8,20]. PPy has mainly been used to treat water to remove NO 3 − [21,22], F − [23,24], and SO 4 2− [9,25]. Polymerization of Py can equally occur at the α and β positions, so PPy grows in three dimensions during polymerization [26,27]. ree-dimensional growth of PPy will give relatively few planarization defects [24], meaning the PPy will readily be attacked by OH − (a strong nucleophile) once Cl − is incorporated, and deprotonation has occurred [26]. e presence of planar aromatic benzene sulfonate (e.g., sodium benzenesulfonate (BSNa)) anions during polymerization could limit three-dimensional growth of PPy and therefore promote an orderly arrangement of PPy chains, which would cause the PPy to grow in a laminar fashion and have a dense and ordered surface morphology [26,27]. Preparing PPy using this doping method should give PPy-modified activated carbon with a better anion exchange performance than PPy-modified activated carbon produced without doping. However, using doping anions that are too large (e.g., sodium dodecylbenzenesulfonate) will give PPy chains separated so much that interchain jumping by the carriers is hindered, and ion exchange cannot occur [28][29][30]. e ions in a PPy film will form galvanic cells on the surfaces because of the difference in the oxygen concentrations at the surface and in the bulk solution [31]. Hydroxide ions in the solution could attack the imino groups in PPy to form carbonyl groups, which will gradually decrease the ion exchange capacity of the PPy [27]. Doping with BSNa will increase the ion gap in the PPy and prevent galvanic cells forming.
In addition to efficient adsorption, efficient desorption and regeneration are extremely important when developing an adsorption material. OH − strongly attacks PPy and can remove the doped ions [26], so strong alkalis such as sodium hydroxide are often used to regenerate PPy. Bhaumik et al. [23] found that 97% of the F − ions adsorbed by a PPy/Fe 3 O 4 nanomagnetic material desorbed by increasing the pH to 12. Pahlavanzadeh et al. [32] used 0.1-1 M NaOH to desorb NO 3 − adsorbed to PPy/sawdust and found PPy/sawdust regeneration rates of between 10% and 35% at different NaOH concentrations. Hydroxyl anions and imino groups will react rapidly in a strong alkaline environment, meaning that PPy will be damaged [27], so ammonium hydroxide (NH 3 ·H 2 O), which is a weak alkaline, has also been used as a PPy regenerating agent.
In this study, a method involving doping with BSNa was used to produce granular activated carbon (GAC) modified with PPy (PFB-GAC). e effects of the PFB-GAC preparation conditions and reaction conditions on SO 4 2− adsorption were investigated. e PFB-GAC was characterized by scanning electron microscopy (SEM), Brunauer-Emmett-Teller (BET) surface area analysis, energy-dispersive X-ray spectroscopy (EDS), and Fouriertransform infrared spectroscopy to investigate the mechanism through which the PFB-GAC adsorbed SO 4 2− . e PFB-GAC performance was assessed using adsorption kinetic, adsorption isotherm, and adsorption thermodynamic data. Adsorption-desorption cycles were performed under alkaline conditions (using NH 3 ·H 2 O) to assess the recyclability of the PFB-GAC. e results were expected to indicate whether the PFB-GAC could be used to remove SO 4 2− from acidic wastewater.

Adsorbent Synthesis.
e PPy-modified GAC doped with BSNa was prepared from pristine GAC that had been ground and wet-sieved to Chinese mesh size 20 (0.9 mm). e GAC was repeatedly washed with ultrapure water and then dried at 328 K in a vacuum drying oven for 24 h to a constant weight. A 2.5 g aliquot of the dry GAC was added to a glass bottle, and then 50 mL of Py that had been distilled using a rotary evaporator and had a specified molar concentration was added to the bottle. e bottle was sealed and placed in a thermostat-controlled oscillator and then oscillated at 180 revolutions min −1 for 12 h at 298 K. e supernatant was discarded, then 50 mL of FeCl3·(6H2O) at a specified molar concentration was added, and the bottle was oscillated for 5 h. en, a specified molar concentration of BSNa was added, and the bottle was oscillated for 3 h. e PFB-GAC was then collected by passing the mixture through a funnel containing filter paper. e PFB-GAC was rinsed three times each with deionized water, anhydrous ethanol, and deionized water, in that order, and then dried in a vacuum drying oven. e PFB-GAC was dried for 24 h at 328 K. Unused PFB-GAC was stored in a nitrogen atmosphere at a low temperature in the dark. e improvement in SO 4 2− adsorption by PFB-GAC relative to other adsorbents was investigated by performing experiments using GAC modified separately with FeCl3·(6H2O), BSNa, and Py and with GAC modified with FeCl3·(6H2O) plus BSNa, GAC modified with Py plus FeCl3·(6H2O) (PF-GAC), and GAC modified with Py plus BSNa.

Batch Adsorption.
Adsorption experiments were performed to investigate the effects of different conditions on SO 4 2− adsorption and to identify the conditions that caused the largest amount of SO 4 2− to be removed. Each adsorption experiment was performed using 50 mL of a 500 mg L−1 SO 4 2− solution at neutral pH in a conical flask kept at 298 ± 1 K in a constant-temperature oscillator set to oscillate at 180 revolutions min−1. Different parameters were used in different experiments. e effects of using different initial SO 4 2− concentrations, pH values, and initial temperatures on the adsorption capacity at equilibrium were studied, and the adsorption kinetics and adsorption isotherms were investigated. e adsorbent dose m s was always 0.2 g except when tests were performed to investigate the effect of using different adsorbent doses. e pH was adjusted by adding 0.1 mol L −1 NaOH or HCl. e ion chromatography column was protected from organic matter by passing each supernatant sample through a C 18 cartridge and a 0.22 μm filter before the SO 4 2− concentration was determined. It was found in preliminary experiments that the sample pretreatment did not affect the SO 4 2− concentration. Each test was performed in duplicate, and the mean concentration is presented.
e adsorption capacity at equilibrium q e (mg g −1 ) was calculated using the following equation [33]: e adsorption capacity q t (mg g −1 ) at contact time t (min) was calculated using the following equation: e mass m ads (mg) of adsorbed SO 4 2− was calculated using the following equation: here, C 0 , C e , and C t are the initial adsorbate concentration (mg L −1 ), adsorbate concentration (mg L −1 ) at equilibrium, and adsorbate concentration (mg L −1 ) at time t (min), respectively, V is the volume of the solution (mL), and m s (g) is the dry weight of adsorbent.

Desorption Experiments.
A saturated PFB-GAC sample was transferred to a 10 mL bottle, and a specified concentration of the regeneration solution was added. e bottle was sealed and shaken for a specified time, and then the supernatant was analyzed to determine the concentration of SO 4 2− that had been desorbed from the PFB-GAC. e desorption and regeneration time was 40 min, the temperature was 298 K, and the regeneration solution volume was 1 mL. After desorption, the PFB-GAC was removed from the mixture by filtering the mixture, and then the PFB-GAC was washed with deionized water and dried in a vacuum oven. A fresh aliquot of SO 4 2− solution was then added before the desorption process was repeated. e mass of desorbed SO 4 2− m i (mg) was calculated using the following equation: where C i is the SO 4 2− concentration (mg L −1 ) in the desorption solution and V is the volume of the regeneration solution.
e desorption efficiency D (%) was calculated using the following equation:
e surface structure and element composition of the adsorbent were investigated by SEM and EDS using a JSM-7100F instrument (JEOL, Tokyo, Japan) at magnifications of 2000 and 7000.
e surface area and porosity of the adsorbent were investigated by performing BET analysis using MicroActive software for ASAP2460 system (Micromeritics, Norcross, GA, USA) at 418 K using N 2 . Fourier-transform infrared spectroscopy was performed using a TENSOR 27 system (Bruker, Billerica, MA, USA) to identify changes in functional groups during the PFB-GAC preparation procedure and through adsorption of SO 4 2− . 4 2− Adsorption. Modified GAC was prepared from 2.5 g of GAC and Py at a concentration of 1 mol L −1 . A 0.2 g aliquot of modified GAC was added to each of a series of conical flasks, and then 50 mL of 500 mg L −1 Na 2 SO 4 was added to each flask.

Effects of Using Different Py and FeCl 3 ·(6H 2 O) Ratios on SO
It can be seen from Figure S1 that the adsorption capacity of the PF-GAC reached 11 mg g −1 , which was about five times higher than the adsorption capacity of the GAC. However, the adsorption capacities were lower for the Py-BSNa-GAC and FeCl 3 ·(6H 2 O)-BSNa-GAC than for the Advances in Materials Science and Engineering GAC.
is indicated that oxidation of Py because of the FeCl 3 ·(6H 2 O) doping and polymerization to give PPy caused ion exchange to occur, as has been found in previous studies [7,18,19]. e effect of the Py : FeCl 3 ·(6H 2 O) (P : F) ratio on SO 4 2− adsorption by the PF-GAC was investigated, and the SO 4 2− adsorption capacity results are shown in Figure 1. It can be seen that the amount of SO 4 2− adsorbed gradually increased as the amount of Py used increased. At a P : F ratio of 1 : 1.5, the adsorption capacity was about seven times higher than the adsorption capacity of the GAC. e results indicated that the adsorption rate could still be increased when the P : F ratio was increased to more than 5 : 1. However, the cost of preparing the modified adsorbent led to subsequent PFB-GAC tests being performed using P : F ratios between 1 : 1 and 1 : 3. 4 2− Adsorption by the Modified GAC. Modified GAC was prepared using different BSNa concentrations (1, 0.1, 0.01, 0.001, and 0.0001 mol L −1 ) and P : F ratios (1 : 1.5 and 1 : 3). e adsorption capacity was determined after performing static 25 min adsorption tests at 298 K. e results are shown in Figure 2. SO 4 2− was adsorbed more effectively by the modified GAC prepared using a P : F ratio of 1 : 1.5 and a BSNa concentration of 10 −3 mol L −1 than by the other modified GAC samples. At the same BSNa concentration, SO 4 2− was adsorbed more effectively when the P : F ratio was 1 : 1.5 than when the P : F ratio was 1 : 3. e amount of SO 4 2− adsorbed by the modified GAC gradually increased as the BSNa dose increased. e adsorption capacity decreased when the BSNa dose increased above 10 −3 mol L −1 . For the same amount of PPy, increasing the BSNa dose may decrease ion exchange between SO 4 2− and Cl − . 4 2− by the PFB-GAC at Different Initial SO 4 2− Concentrations. Modified GAC was prepared using a Py : FeCl 3 ·(6H 2 O) : BSNa (P : F : B) molar ratio of 1000 : 1500 : 1. Adsorption of SO 4 2− by this modified GAC was assessed, and adsorption of SO 4 2− by PF-GAC and GAC was also assessed for comparison. e three types of GAC described above were used to adsorb SO 4 2− from solutions with different initial SO 4 2− concentrations (400, 500, 600, 700, and 800 mg L −1 ), and the results are shown in Figure 3. e SO 4 2− adsorption capacity of the PFB-GAC prepared using a P : F : B ratio of 1000 : 1500 : 1 was almost double the SO 4 2− adsorption capacity of the PF-GAC prepared using a P : F ratio of 1 : 1.5 and about eight times higher than the SO 4 2− adsorption capacity of the GAC. e amount of SO 4 2− adsorbed by the PFB-GAC gradually increased as the initial SO 4 2− concentration increased. When the initial SO 4 2− concentration was 800 mg L −1 , the adsorption capacity of the PFB-GAC was 20.23 mg g −1 , which was 1.3 times higher than the adsorption capacity of 15.49 mg g −1 found when the initial SO 4 2− concentration was 300 mg L −1 . Assuming that the number of adsorption sites was constant, the results indicate that increasing the SO 4 2− concentration difference at the adsorption interface may promote diffusion of SO 4 2− to the adsorption material surfaces. 4 2− Removal. e effect of the pH on SO 4 2− adsorption was studied using a PFB-GAC dose of 4 g L −1 , an initial SO 4 2− concentration of 500 mg L −1 , and pH values between 2.0 and 12.0. e results are shown in Figure 4. It can be seen that the PFB-GAC most effectively adsorbed SO 4 2− when the initial pH was 2.0. Between pH 3.0 and 5.0, the adsorption capacity decreased slightly as the pH increased. Increasing the pH further favored exchange  between OH − and SO 4 2− that had been adsorbed by the PFB-GAC, resulting in SO 4 2− desorption, because of the high OH − concentration especially under alkaline conditions. Münstedt [34] investigated the electrical conductivity of the PPy film treated with base or acid and found that its conductivity was markedly reduced following a strong alkali treatment, whereas it remained essentially unchanged under highly acidic conditions. Generally, PPy film is more stable in acidic environment [35], which might be the reason why it exhibited higher anion adsorption efficiency. e hydroxide ions can attack imino groups and damage the PPy, and deprotonation could also occur [24,26]. 4 2− Removal. e effects of the temperature on SO 4 2− adsorption by the PFB-GAC when the adsorbent dose was 4 g L −1 are shown in Figure 5. It can be seen that the SO 4 2− adsorption capacity of the PFB-GAC gradually increased as the temperature decreased whatever the initial SO 4 2− concentration is.

Effect of Temperature on SO
is indicated that the adsorption process was exothermic and that decreasing the temperature promoted SO 4 2− adsorption by the PFB-GAC.

Characterization of the PFB-GAC
3.3.1. SEM Analysis. e SEM results are shown in Figure 6. Images A 1 and A 2 are SEM images of the unmodified GAC. It can be seen that the GAC had abundant evenly distributed pores on the surfaces, making it a good carrier for PPy. Images B 1 and B 2 are SEM images of the PPy-modified GAC. It can be seen when comparing images A 1 and A 2 with images B 1 and B 2 that PF had been successfully loaded onto the PF-GAC. e pores on the GAC surfaces became fewer and smaller because of PPy being loaded onto the pore surfaces, partly blocking the pores [36].
Images C 1 and C 2 are SEM images of the oxidized polymerized PPy doped with BSNa. Some bulky substances were added to the surfaces. e BSNa played a role in the PPy polymerization process and caused the coating to be more continuous than the coating of the GAC coated only with PPy [37]. e BSNa connected PPy strands together, so it increased the degree of polymerization of the PPy and decreased PPy shedding during the adsorption process.

BET Analysis.
e BET analysis results indicated that the BET surface area of the PFB-GAC was 725.2964 m 2 g −1 , but the BET surface area of the GAC was 5.9400 m 2 g −1 . e adsorption and desorption isotherms and pore size distribution curves of the GAC and PFB-GAC are shown in Figure 7(a) (GAC) and Figure 7(b) (PFB-GAC). According to the International Union of Pure and Applied Chemistry classification, the PFB had a type I adsorption isotherm, but the macroporous GAC had a type II adsorption isotherm.
is and the pore size distribution suggest that the PFB coating occupied the large pores in the GAC and that more micropores formed. e PFB coating caused the macroporous GAC to become an adsorbent with well-developed mesopores (pore size <2 nm). e SEM and BET analysis results indicated that the PFB coating adhered well to the large pores and outer surfaces of the GAC and gave the PFB-GAC many micropores. Figure 8, EDS indicated that the N contents of the GAC, PF-GAC, and PFB-GAC were 3%, 10.2%, and 12.8%, respectively. Approximately 25% of the N in PPy would have been N + [22], and the nitrogen content of GAC loaded with PPy was markedly higher than the nitrogen content of the GAC. When the same number of moles Advances in Materials Science and Engineering of Py was used, the nitrogen content was higher when BSNa was present than when BSNa was not present.

EDS. As shown in
Loading the GAC with PPy increased the amount of N + on the surfaces, and the more positively charged sites would have been occupied by Cl − and would have provided active sites for ion exchange with SO 4 2− . Doping with BSNa controlled PPy growth in certain directions and would control and limit spatial growth of the PPy produced through polymerization of Py [37]. e PPy would therefore grow in an orderly manner in a specific direction or in a specific area rather than three-dimensionally. e PPy on the GAC surfaces therefore had a more regular morphology and was more stable when BSNa was present than when BSNa was not present. e BSNa also played a role in causing planes to form in the PPy, and therefore, more PPy to be loaded on the GAC surfaces, which would have increased the conductivity and ion exchange capacity of the PPy. e imino group in PPy can, to some extent, act as a reducing agent [38], but SO 4 2− is relatively stable. e ability of PPy to act as a reducing agent would therefore have decreased the SO 4 2− concentration in the water little.

Fourier-Transform Infrared Spectroscopy.
e Fouriertransform infrared spectra of the GAC, PFB-GAC, and PFB-GAC with adsorbed SO 4 2− are shown in Figures 9(a)-9(c), respectively. e PFB-GAC surfaces were found to have abundant functional groups.
ere was a strong wide absorption band at 3430 cm −1 , which was assigned to intermolecular hydrogen bonds and overlapping bands for O-H and N-H stretching vibrations. e main PPy bands were at 1543 cm −1 (C-C stretching vibrations in the pyrrole ring) and 1450 cm −1 (stretching vibrations of C-N in the ring), and there was a local absorption maximum at 1164 cm −1 (symmetric stretching vibration of the pyrrole ring). A band at 1039 cm −1 was assigned to deformation vibrations in the C-H and N-H planes of the pyrrole ring, and the lowest value of the spectrum corresponded to deformation vibrations outside the C-H plane of the ring [39]. PPy was successfully anchored to the GAC, and loading PPy onto the GAC clearly changed the functional groups attached to the GAC.
ere was a particularly large change at around 3430 cm −1 , which was assigned to N-H groups and clearly indicated that the amount of N + present was increased by the modification process.

Adsorption Kinetics.
e kinetics of SO 4 2− adsorption onto the PFB-GAC are shown in Figure 10.
e results indicated that the PFB-GAC strongly removed SO 4 2− in the first 25 min, and then the amount of SO 4 2− adsorbed remained stable. is indicated that the PFB-GAC could be used to quickly adsorb SO 4 2− . e pseudo-first-order and pseudo-second-order kinetics models were fitted to the experimental data. e data and lines are shown in Figures S2 and S3, and the fitting parameters are shown in Table 1. e pseudo-second-order kinetics model fitted the data well. e pseudo-first-order kinetics model is related to physical adsorption processes, and the pseudo-second-order kinetics model describes chemical adsorption processes based on electron sharing [39]. e fitting results indicated that the adsorption process included chemical adsorption. e linear forms of the pseudo-first-order and pseudosecond-order kinetics equations are shown in the following equations, respectively.
In equations (6) and (7), q e (mg g −1 ) and q t (mg g −1 ) are the amounts of SO 4 2− adsorbed onto the modified GAC at equilibrium and at time t, respectively, and k 1 (min −1 ) and k 2 (g (min mg) −1 ) are the pseudo-first-order and pseudosecond-order adsorption rate constants, respectively.

Adsorption
Isotherms. For the Langmuir model, it is assumed that adsorption occurs on the adsorbent surface uniformly and that no migration occurs. e Langmuir model is suitable for single-layer adsorption, adsorbents with homogeneous sites on the surfaces, and a uniform energy-level distribution.
e Freundlich model is not limited to single-layer adsorption and can be used as an empirical adsorption model for uneven surface conditions. e Langmuir equation can be rearranged to give a linear form to make it convenient to plot the model and determine the Langmuir constants (k L ), as shown below. e linear form of the Langmuir model is shown in equation (8). e q m and k L values can be determined from a linear plot of C e q e −1 against C e .  Figure 5: Effect of the temperature on SO 4 2− adsorption by the pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon. 6 Advances in Materials Science and Engineering In equation (8), Ce is the SO 4 2− concentration at equilibrium (mg L −1 ), q e is the amount of SO 4 2− adsorbed at equilibrium (mg g −1 ), q m is the maximum amount of SO 4 2− adsorbed (mg g −1 ), and k L is a constant (L mg −1 ). e Freundlich isotherm equation is based on adsorption onto nonuniform surfaces. e Freundlich isotherm equation is shown in the following equation: In equation (9), k F (mg g −1 ) is the Freundlich constant related to the adsorption capacity and adsorption strength and 1/n is the heterogeneity factor. e results of the adsorption isotherm experiments are shown in Figure 11, and the calculated parameters for the Langmuir and Freundlich models are shown in Table 2. e Langmuir and Freundlich isotherms are shown in Figures S4 and S5, respectively. e R 2 values for the Langmuir model were 0.9933-0.9977, and the R 2 values for the Freundlich model were 0.9482-0.9956, indicating that SO 4 2− adsorption onto the PFB-GAC was described better by the Langmuir model than the Freundlich model. is indicated that the PFB-GAC surfaces adsorbed a single layer of SO 4 2− , the adsorption sites were homogeneous, and the energy-level distribution was uniform. e SO 4 2− adsorption rate K L gradually decreased as the reaction temperature increased, indicating that the adsorption process was exothermic. is was consistent with the effect of the temperature on the adsorption capacity discussed in Section 3.2.3.

Adsorption
ermodynamics. ree thermodynamic parameters, standard enthalpy (ΔH 0 ), standard free energy (ΔG 0 ), and standard entropy (ΔS 0 ), were assessed. e ΔG 0 , ΔH 0 , and ΔS 0 values were calculated using the following equations: e relationship between ΔG 0 , ΔH 0 , and ΔS 0 is described by the following equation: e van 't Hoff equation is   Advances in Materials Science and Engineering where R (8.314 J mol −1 K) is the universal gas constant, T (K) is the absolute temperature of the solution, and Kc (L mol −1 ) is the Langmuir constant, which can be determined by identifying the intercept of the line fitted to a plot of C e q e −1 against C e . A plot of lnKc against 1/T is shown in Figure S6. A line was fitted to the plot to allow the van 't Hoff equation parameters to be determined. e enthalpy change was defined as the slope of the line (ΔH 0 R −1 ), and the entropy change was defined as the intercept of (ΔS 0 R −1 ). e calculated ΔG 0 , ΔH 0 , and ΔS 0 values for SO 4 2− adsorption onto the PFB-GAC are shown in Table 3. e enthalpy change (ΔH 0 ) was negative, indicating that lower temperatures favored adsorption, which was consistent with the conclusions drawn from the adsorption isotherms. e entropy change (ΔS 0 ) was negative, indicating that the PFB-GAC structure was not markedly changed by the adsorption process and that adsorption decreased randomness in the system. e negative ΔG 0 value indicated that adsorption was spontaneous.

Desorption and Reuse.
Batch desorption experiments were performed using deionized water, sodium hydroxide, and ammonium hydroxide (mass concentration 25%) as desorbents. When deionized water was used, <3% of the adsorbed SO 4 2− was desorbed. Tests were performed using different NaOH concentrations (0.01-2 mol L −1 ). e most SO 4 2− was desorbed (15 mg g −1 , 75% of the adsorbed SO 4 2− ) when 2 mol L −1 NaOH was used. However, the adsorption capacity after the first regeneration process had been performed was <10 mg g −1 , and the adsorption capacity was almost 0 mg g −1 after five adsorption-desorption cycles. Strong alkaline conditions (i.e., a NaOH solution) would cause OH − to readily attack the imino group and effectively desorb SO 4 2− [26] but excessively damage the PPy [27]. e OH − concentration will be roughly the same in 12 mol L −1 NH 3 ·H 2 O and 2 mol L −1 NaOH. SO 4 2− desorption over time using 12 mol L −1 NH 3 ·H 2 O is shown in Figure 12(a). e percentage of the adsorbed SO 4 2− that was desorbed gradually increased over time and was ∼30% at 5 min and >80% at 40 min. e desorption efficiency then remained stable. Regeneration tests were performed using 12 mol L −1 NH 3 ·H 2 O at 288, 298, 308, and 318 K for 40 min.
e desorption efficiency gradually decreased as the temperature increased (Figure 12(b)). e desorption efficiency was ∼75% at 288 K but only ∼30% at 318 K. is indicated that the desorption process was exothermic and that the desorption mechanism may have been spontaneous ion exchange involving OH − released by NH 3 ·H 2 O reacting with the imino groups in PPy to desorb SO 4 2− . Adsorption-desorption recycling experiments using the PFB-GAC were performed using 12 mol L −1 NH 3 ·H 2 O as the regenerant; the results are shown in Figure 13. e highest adsorption capacity and desorption efficiency   4 2− onto the pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon.

Langmuir model
Freundlich model  Table 3: ermodynamic parameters for SO 4 2− adsorption onto pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon. Advances in Materials Science and Engineering 9 were found in the first cycle. e adsorption capacity was ∼20 mg g −1 , and 12 mol L −1 NH 3 ·H 2 O gave a desorption efficiency of 85% (i.e., higher than the 75% desorption efficiency found using 2 mol L −1 NaOH). Four more adsorption-desorption cycles were performed. e adsorption capacity gradually decreased as the number of regeneration cycles increased, and the desorption efficiency decreased (from 71.6% in cycle 2 to 30.5% in cycle 5). e adsorption capacity in cycle 5 was still >6 mg g −1 (in contrast, the PFB-GAC did not adsorb SO 4 2− after five regeneration cycles using NaOH). NH 3 ·H 2 O is a weaker alkali than NaOH, so it damaged PPy less than NaOH. NH 3 ·H 2 O is also a weak reducing agent, which would have been conducive to SO 4 2− desorption. e surface morphology of the PFB-GAC was not markedly altered by five regeneration cycles using NH 3 ·H 2 O ( Figure S7). e modified GAC doped with BSNa adsorbed and desorbed SO 4 2− well, possibly because the BSNa improved the alkali-resistance and stability of the modified GAC. , so it was concluded that only a small amount of physical adsorption to the GAC pore surfaces would have occurred when the modified GAC was used. e PFB on the PFB-GAC therefore mainly adsorbed SO 4 2− . e GAC acted as a carrier of the PFB. GAC has advantages over powdered activated carbon as a carrier, because less physical shedding of PFB will occur from GAC than from powdered activated carbon after multiple regeneration cycles, and the PFB-GAC particles would have had more regular shapes than would PFB-powdered activated carbon particles. e PF-GAC and PFB-GAC clearly adsorbed SO 4 2− when the SO 4 2− concentration was low. is indicated that SO 4 2− was not removed from solution through a simple physical adsorption process. During the GAC modification process, Cl − would have temporarily occupied the positively charged sites, which would have prevented strong removal of SO 4 2− through electrostatic attraction [22]. It was found in previous studies [8,9] that SO 4 2− was removed by PPymodified GAC mainly through ion exchange between Cl − and SO 4 2− . e results of this experiment led to the same conclusion. During an adsorption test, the Cl − concentration in solution gradually increased as the SO 4 2− concentration decreased, and the lower the Cl − concentration, the poorer the adsorption effect. In summary, ion exchange played an important role in the adsorption of SO 4 2− by the PFB-GAC. e mechanism involved in SO 4 2− removal by the PFB-GAC is shown in Figure 14. During the GAC doping and modification process, Py would have been oxidized by   Advances in Materials Science and Engineering would have been involved in SO 4 2− removal [22]. e SEM images ( Figure 6) and EDS results ( Figure 8) indicated that doping with BSNa made the PPy more planar than it would otherwise have been and limited three-dimensional growth of the PPy.

Analysis of the
e PPy would therefore have been securely anchored to the surfaces and voids of the GAC. is would have caused the N + content to be higher than it would have been in the absence of BSNa. It can be seen from the results of the PFB and PF experiments described in Section 3.2.1 that SO 4 2− was removed more effectively by the PFB-GAC than the PF-GAC. is indicated that doping the PPy with BSNa increased the number of ion exchange sites and improved SO 4 2− adsorption. Strong acidic conditions will affect the surface charge of an adsorbent, the degree of ionization, and the adsorbent morphology [32]. Acidic conditions would have protonated the -NH groups in the PPy and increased the amount of SO 4 2− that became electrostatically adsorbed [38]. e positively charged sites in the PPy would have been attacked by OH − under alkaline conditions, causing the PPy to be damaged [26].
Regenerating the PF-GAC using NH 3 ·H 2 O allowed SO 4 2− adsorption and desorption to continue, because NH 3 ·H 2 O will slowly release OH − to desorb SO 4 2− . e SO 4 2− adsorption capacity after five adsorption-desorption cycles remained higher than the SO 4 2− adsorption capacity of the GAC.

Conclusions
Adsorption of SO 4 2− by the PFB-GAC was markedly improved by adding an appropriate amount of BSNa, and the PFB-GAC prepared using a P : F : B ratio of 1000 : 1500 : 1 had a higher SO 4 2− adsorption capacity than the PFB-GAC prepared using other P : F : B ratios. e SO 4 2− adsorption capacity of the PFB-GAC was 23.05 mg g −1 when the initial SO 4 2− concentration was 500 mg L −1 , the pH was 3, and the temperature was 298 K. Characterization and analysis of the PFB-GAC indicated that GAC doped with BSNa and loaded with PPy contained many N + sites and that ion exchange of Cl − and SO 4 2− could occur. SO 4  adsorption capacity after five adsorption-desorption cycles was still higher than the SO 4 2− adsorption capacity of the GAC. e PFB-GAC is a new type of adsorbent that could be used to remove SO 4 2− from acidic wastewater. Advances in Materials Science and Engineering activated carbon, PB-GAC � pyrrole-sodium benzenesulfonate-granular activated carbon). Fig. S2. Pseudo-first-order kinetics plot of SO 4 2− adsorption to the PFB-GAC at different initial concentrations of SO 4 2− (PFB-GAC � pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon). Fig. S3. Pseudo-second-order kinetics plot of SO 4 2− adsorption to the PFB-GAC at different initial concentrations of SO 4 2− (PFB-GAC � pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon). Fig. S4. Langmuir isotherms for SO 4 2− adsorption to the PFB-GAC at different reaction temperature (PFB-GAC � pyrrole-FeCl 3 ·(6H 2 O)sodium benzenesulfonate-granular activated carbon). Fig. S5. Freundlich isotherms for SO 4 2− adsorption to the PFB-GAC at different reaction temperature (PFB-GAC � pyrrole-FeCl 3 ·(6H 2 Osodium benzenesulfonate-granular activated carbon). Fig. S6. van 't Hoff plot used to determine thermodynamic parameters for SO 4 2− adsorption. Fig. S7. SEM image of PFB-GAC that had been regenerated five times using NH 3 ·H 2 O at a magnification of 7000 (PFB-GAC � pyrrole-FeCl 3 ·(6H 2 O)-sodium benzenesulfonate-granular activated carbon). (Supplementary Materials)