Preparation of composite adsorbents of activated carbon supported MgO/MnO₂ and adsorption of Rhodamine B

Dye wastewater from textiles and other industrial areas is difficult to remove due to its different chemical structure and large yield. Dyes in water have the characteristics of mutagenicity, toxicity, sensitization, carcinogenicity, non-degradation, good stability to light and oxidants, and resistance to aerobic digestion, and also have important effects on ecological systems and human health (Cheng et al. 2016; Gupta et al. 2016; Maeda et al. 2018). Rhodamine B (RB) is a basic dye molecule that is widely used in textile printing, dyeing, etc. (Ahmed et al. 2018; Qi et al. 2019; Song et al. 2019). Common methods for removing dyes in water include coagulation (Wang et al. 2016), electrolysis (Li et al. 2015), biodegradation (Liu et al. 2016a, 2016b), advanced oxidation (Xu et al. 2018) and adsorption (Qi et al. 2019), etc. As an adsorbent with a large specific surface area, activated carbon (AC) is often used to adsorb dye wastewater (Zhang et al. 2018). However, the adsorption capacity of AC to dye is still relatively small. In order to increase the adsorption capacity of activated carbon, many researchers have focused on research into activated carbon surface modification which can improve its surface characteristics to obtain better adsorption capacity.

Metal ion modification is a simple and effective method to improve the adsorption capacity of AC (Liu et al. 2017). The removal capacity for methylene blue of AC increased 17.12% by impregnating Fe(NO₃)₃ onto AC (Cheng et al. 2016). Granular activated carbon supported magnesium hydroxide had a better adsorption capacity to cadmium (II) than untreated granular activated carbon (Wang et al. 2016). Common modification methods include sol-gel (Ciesielczyk et al. 2017), impregnation (Rojasmayorga et al. 2015), calcinations (Jin et al. 2018), microwave (Liu et al. 2016a, 2016b) and co-precipitation (Wang et al. 2016). Calcining can effectively change the pore structure, dispersion and surface properties of the material (Yuan et al. 2013). Magnesium oxide and manganese dioxide are good catalysts and adsorbents for environmental protection (Shaabani et al. 2015; Ahmed et al. 2018; Zhang et al. 2019). A granular activated carbon microporous adsorbent coated with magnesium oxide was used to remove benzene from the air (Rashidi et al. 2019). Chemically modified activated carbon loaded with manganese dioxide nanoparticles enhanced the adsorption capacity of activated carbon for Li⁺ adsorption from aqueous media (Kamran et al. 2019). Although the use of manganese/magnesium modified adsorbents has been reported widely, there is little information about the effect of Mn/Mg on the adsorption properties of active carbons.

The main objectives of the presented work focus on loading both magnesium oxide and manganese dioxide onto AC by impregnation-coprecipitation-calcination. Batch adsorption tests were used to study the adsorption mechanism. The adsorption characteristics of RB on original and modified activated carbon were analyzed by scanning electron microscopy (SEM), Brunauer-Emmett-Teller (BET) and X-ray diffraction (XRD) analysis. Adsorption isotherms and adsorption kinetics were discussed.

Activated carbon (AC) was purchased from an activated carbon factory in Henan. It was screened into diameters of 1 to 3 mm. Rhodamine B was used as an adsorbate compound and was received from Tianjin Dongya Company. The molecular structure and molecular formula of RB are presented in Table 1. Stock solutions of Mn(NO₃)₂, MgCl₂ and Na₂CO₃ were prepared by dissolving reagent-grade metal salts, Mn(NO₃)₂ (Tianjin Guangfu), MgCl₂·6H₂O (Tianjin Chemical Reagent Co.) and Na₂CO₃ (Tianjin Wind Ship) in distilled water, respectively.

AC was washed several times with deionized water to remove surface dust and impurities, then dried at 90 °C for 3 h. A total of 10 g AC was mixed with 150 ml of a 0.1 mol·L⁻¹ Mn(NO₃)₂ and 0.5 mol·L⁻¹ MgCl₂ solution using a 250 ml conical flask, and the mixture was agitated at 25 °C and 180 rpm for 2 h. After filtering, washing and drying, MgCl₂-Mn(NO₃)₂-AC compounds are obtained. Then, the compounds were reacted with 100 ml of a 0.5 mol·L⁻¹ Na₂CO₃ and 0.5 mol·L⁻¹ NaOH solution for some time, then filtered, washed and dried to obtain a sample Mg(OH)₂-MnCO₃-AC compound. The sample was calcined in a furnace at 360 °C for 2.5 h, and named MgO-MnO₂-AC.

The surface micromorphology of MgO-MnO₂-AC composite materials was observed by SEM using a JSM-7800F scanning electron microscope (Quanta 200, FEI, Czech Republic). Element contents were observed by ICP-MAS (inductively coupled plasma mass spectrometry, Agilent 7000E, USA). The N₂ adsorption-desorption isotherms were recorded using a nitrogen gas adsorption analyzer (ASAP 2020, Micromeritics, USA). The specific surface area of the sample was determined by the BET method. The X-ray diffractometer (Ultima IV, Rigaku Corporation, Japan) was used on the crystal form of AC and MgO-MnO₂-AC.

Batch adsorption experiments were conducted in 250 ml conical flasks with different initial adsorbate concentrations, pH, temperatures, and addition of adsorbent. AC and MgO-MnO₂-AC modified materials were used as adsorbents for RB dye adsorption. The flasks were placed in a thermostatic oscillator at a speed of 200 rpm and reacted for 3 h without light. Then, the supernatant was filtered with a 0.45 μm filter membrane, the filtrates' concentration of RB was analyzed by a UV-Vis spectrophotometer (UV2550, Shimadzu, Japan).

The adsorption kinetics study of MgO-MnO₂-AC was performed by adding 0.3 g of sample and 100 ml of varying RB solutions (10, 50, 100 mg·L⁻¹) to 250 ml conical flasks with stoppers. Then, all conical flasks were shaken at a constant agitation speed (200 rpm) at 25 °C. Samples were obtained at 1–11 h, respectively. The supernatant solution was filtered with a 0.45 μm filter membrane and the absorbance was measured.

For the adsorption isothermal study, 0.3 g MgO-MnO₂-AC was mixed with 100 ml varying initial RB concentrations (10–60 mg·L⁻¹) using 250 ml conical flasks with stoppers. The flasks were shaken in a thermostatic oscillator at a constant agitation speed (220 rpm) for 10 h at pH 7.6. Reaction temperatures were set at 25 °C, 35 °C, and 45 °C. The supernatant solution was filtered with a 0.45 μm filter membrane and the absorbance measured.

The structures and properties of the original and modified activated carbon are presented in Table 2. The surface area and total pore volume of AC increased to some extent after modification by MgO and MnO₂, which can provide more active sites for RB. This indicates that the formation of MgO and MnO₂ increased the specific surface area and pore volume of the original activated carbon. The decrease in average pore diameter is due to the formation of load on the pore wall (Riverautrilla et al. 2017). Pore size distribution (PSD) is one of the important properties of AC, which determines the contact distribution of molecules in the whole pore volume. As can be seen from Figure 1(a), MgO-MnO₂-AC contains a large number of micropores and a part of mesoporous pores, and the number of micropores and mesoporous holes is more than the original AC. This may be due to the removal of impurities caused by calcining and the formation of new structures on the surface and inside the MgO-MnO₂-AC.

The nitrogen adsorption-desorption isotherms of the AC and MgO-MnO₂-AC are shown in Figure 1(b). According to the International Union of Pure and Applied Chemistry (IUPAC) classification, AC and MgO-MnO₂-AC materials conform to the characteristic of Type I (Sing et al. 1985). The Type I isotherm has significant adsorption properties under relatively low pressure and belongs to microporous adsorption. The N₂ isotherm slope of MgO-MnO₂-AC was larger than that of AC, indicating that MgO-MnO₂-AC had higher adsorption performance.

The SEM images of AC and MgO-MnO₂-AC are shown in Figure 2(a) and 2(b). The first image shows the smooth surface with regular pore structure of AC. The latter suggests the surface of AC loaded with MgO and MnO₂ is rough and irregular. Figures 1(a) and 2(b) show that regular holes were destroyed and micropores enlarged. This can be interpreted as the existence of pore expansion in the modification process. Rough surface and macropores improve the adsorption performance of the material. The manganese and magnesium contents of the materials under the best preparation condition are shown in Table 3. The contents of AC loaded with magnesium and manganese were 34.24 and 5.51 mg·g⁻¹ respectively. Magnesium and manganese also existed in AC with 16.21 and 0.42 mg·g⁻¹. This proved that the loading of manganese and magnesium on AC was successful.

X-ray diffraction patterns of AC and MgO-MnO₂-AC are shown in Figure 3. For the MgO-MnO₂-AC sample, MgO and MnO₂ crystals were found and accounted for the main crystalline part. Both indicated the presence of a crystalline structure evidenced by the appearance of peaks at 29.1°, 44.7° and 58.6° for MnO₂, and 36.9°, 42.7° and 62.3° for MgO. They are in good agreement with the reported Joint Committee on Powder Diffraction Standards (JCPDS) database. It is noted that the sharp peaks corresponding to 21°, 26° and 50° of the original AC disappear, showing that manganese and magnesium interact with AC during the roasting process (Shaabani et al. 2015). These characteristic peaks have a certain deviation from the standard diffraction peaks of manganese oxide and magnesium oxide. This shows that manganese oxide, magnesium oxide and AC exist in the form of solid dispersion.

The adsorption capacity of RB (50 mg·L⁻¹, pH = 7.6) is shown in Figure 4. The experiment was repeated three times. It exhibits that the adsorption rate of MgO-MnO₂-AC to RB was much higher than that of AC, and the adsorption capacity of RB was also significantly higher than that of AC. When adsorption time was 8 h, RB adsorption quantity almost remained stable at 16.0 mg·g⁻¹. The main reason for this phenomenon may be the greater availability of adsorption sites initially, and RB can contact the effective adsorption sites on the MgO-MnO₂-AC rapidly. After the adsorption was completely balanced at about 10 h, the adsorption capacity was 16.19 mg·g⁻¹, 1.68 times that of AC. It may be concluded that MgO-MnO₂-AC has a strong driving force for RB, and the removal rate of MgO-MnO₂-AC after modification is higher.

The dye concentration has a great impact on the adsorption of RB. The adsorption of RB by AC and MgO-MnO₂-AC increased with the initial concentration of RB (Figure 5). This may be due to active sites on the surface of activated carbon being fully utilized with the increase in RB concentration (Du et al. 2017). The adsorption capacity of MgO-MnO₂-AC was significantly higher than that of AC when the dye concentration remained unchanged. On the one hand, MgO-MnO₂-AC increases the specific surface area of activated carbon and the active sites. On the other hand, magnesium oxide and manganese dioxide have synergistic adsorption on dyes at the same time, so under the same conditions, their adsorption capacity increased. We found that when the initial concentration of solution gradually increased, the rate of adsorption increased slowly. This is because when RB concentration was low, there were many active sites of adsorbent and they were not saturated. As the dye concentration increased, more dye molecules reduce the number of free adsorption sites, and the adsorption eventually tends to be saturated.

Experiments investigated the effect of pH on adsorption of RB which is a weakly basic dye. With increasing pH, the adsorption of RB on MgO-MnO₂-AC first increased and then decreased for pH values between 3 and 11 (Figure 6). This is mainly due to the RB molecule containing the Cl⁻ ion. Under acidic conditions, the adsorbent adsorbs H⁺ on the surface and has a positive charge, which causes ion adsorption with the Cl⁻ ion. With increasing pH, the coexistence of a competitive adsorption with H⁺ decreased and RB adsorption increased (Liu et al. 2017). Under alkaline conditions, the adsorbent adsorbs hydrated hydroxide ions OH⁻ and MgO partially hydrolyzed to form OH⁻, making the surface of MgO-MnO₂-AC negatively charged, and generating electrostatic repulsion and competition with Cl⁻ ions in RB, thereby reducing the adsorption value of MgO-MnO₂-AC.

The adsorption kinetics is one of the most important characters in the whole adsorption process because it can demonstrate the solute uptake rate of dyes from aqueous solutions by the adsorbents. In order to investigate the influences of contact time and initial RB concentration on MgO-MnO₂-AC kinetic adsorption, RB initial concentration was in the range between 10 to 100 mg/L, which is presented in Figure 7. The result showed that the equilibration times for adsorption are approximately 8 h at 25 °C with pH of 7.6. From Figure 7, it was observable that the adsorbed amount increased with initial dye concentration. In the kinetic studies, determining the adsorption isotherms, 600 min was allowed to ensure the equilibrium was complete. When reaction time was 600 min, the equilibrium adsorption quantity reached 3.26 mg·g⁻¹, 16.19 mg·g⁻¹ and 30.30 mg·g⁻¹ for 10 mg·L⁻¹, 50 mg·g⁻¹ and 100 mg·g⁻¹ RB, respectively.

The maximum RB adsorption obtained is 30.30 mg·g⁻¹ in Figure 7 and, when compared with previously published reports in other studies (Table 5), it suggests that MgO-MnO₂-AC is a relatively better adsorbent in terms of performance than the adsorbent used before (Gupta et al. 2007; Wang. et al. 2013; Santhi et al. 2014; Azeez et al. 2018).

The adsorption isotherms of MgO-MnO₂-AC are presented in Figure 8. The adsorption of RB by MgO-MnO₂-AC is a nonlinear adsorption process. With the increase of temperature, the equilibrium adsorption capacity of the material at different dye concentrations also increased.

The Langmuir model and Freundlich model were employed to fit the adsorption equilibrium isotherm. The equations were expressed as follows (Yu et al. 2009).

Values of K_L, K_F, R² and 1/n were calculated and are summarized in Table 6. The correlation coefficients R² of the Langmuir isotherm model were all over 0.99. The fitting effect of the Langmuir isotherm model is better than that of the Freundlich isotherm model, and could better describe the adsorption of RB by MgO-MnO₂-AC. It suggested that RB exhibited monolayer adsorption behavior on the MgO-MnO₂-AC surface, which may be caused by electrostatic attraction, hydrogen bond and ion complexation interaction (Çalışkan & Göktürk 2010).

A new type of adsorption material, MgO-MnO₂-AC, was prepared by the impregnation-coprecipitation-calcination method, which made it possible to couple the hydrophilicity of magnesium oxide and manganese dioxide and the high adsorption capacity of activated carbon to greatly improve the pollutant adsorption performance. The content of manganese and magnesium in MgO-MnO₂-AC was 5.51 mg·g⁻¹ and 34.24 mg·g⁻¹ respectively. The surface area and total pore volume of MgO-MnO₂-AC were all increased to some extent compared to the original activated carbon. The pseudo-second-order adsorption kinetics modeling and Langmuir isotherm can better describe the adsorption of RB on MgO-MnO₂-AC. Internal diffusion process is the rate control step of RB dye adsorption by MgO-MnO₂-AC. The removal of RB on MgO-MnO₂-AC was significantly increased, and the maximum adsorption amount of RB on MgO-MnO₂-AC was 30.30 mg·g⁻¹ at 25 °C under the RB concentration of 100 mg·L⁻¹. The adsorption of RB by AC and MgO-MnO₂-AC increased with the initial concentration of RB. The adsorption of RB first increased and then decreased when pH was between 3 and 11. Based on a favorable removal performance, the modification process was confirmed to be useful and Mn-Mg-GAC can be used as a novel and promising adsorbent material to treat contaminated water with reactive dyes.

This work is supported by the Open Project of State Key Laboratory of Chemical Engineering (SKL-ChE-13C03) and the Technology Research and Development Program of Tianjin, China (16YFXTSF00390).