Removing Fluoride from Water by Nanostructured Magnesia-Impregnated Activated Carbon

Abstract

A facile method was employed to impregnate activated carbon, a commonly used water treatment medium, with nanostructured magnesium oxide for fluoride removal. Batch adsorption tests were conducted to evaluate the adsorption performance of the nanostructured magnesia-impregnated activated carbon (nMgO@AC) for fluoride removal. The results demonstrated that this composite material exhibited a good adsorption capacity, with a maximum equilibrium uptake of approximately 121.1 mg/g for fluoride. Kinetic studies revealed that the adsorption process followed the pseudo-second-order adsorption kinetic model, reaching equilibrium in about 100 min. Within the initial pH range of 3 to 11, the adsorption efficiency of nMgO@AC for fluoride remained above 95%, indicating that the initial solution pH had a minimal effect on the material’s fluoride removal capability. The adsorption mechanism was elucidated by characterizing the material properties before and after adsorption using SEM, TEM, XRD and XPS. Initially, magnesium oxide reacted with water and rapidly transformed into magnesium hydroxide. Subsequently, a ligand exchange occurred between the hydroxide groups in magnesium hydroxide and fluoride ions in the aqueous solution, resulting in the effective removal of fluoride. The findings of this study suggest that nanostructured magnesia-impregnated activated carbon holds significant potential for the treatment of fluoride-containing wastewater, particularly for highly alkaline wastewater.

1. Introduction

Water with elevated concentrations of fluoride is a major health concern worldwide. Over 200 million people from approximately 25 countries are facing the threat of fluorosis [1,2,3], particularly in regions such as India, China and East Africa, where groundwater fluoride concentrations can exceed 10 mg/L [4,5,6]. Although an adequate intake of fluoride helps protect our teeth from decay and keep our bones strong [7], excessive ingestion of fluoride will cause dental fluorosis and bone pain, among others [8]. Because of this, the U.S. EPA has set a maximum concentration level of 4 mg/L for drinking water [9]. An elevated fluoride concentration has been found in groundwater and areas where fluorine-containing minerals are abundant [10,11]. Major anthropogenic sources of fluoride discharges range from glass and cement production, semiconductor manufacturing, electroplating and aluminum smelting [12]. More recently, the water produced from coalbed methane production has drawn particular public attention. These process wastewaters generally possess very high salinity, and the fluoride concentration of these waters in some regions can reach 366 mg/L [13], posing severe threats to the safety of drinking water if not managed adequately.

Commonly used technologies for fluoride removal include adsorption [14,15], precipitation [16,17], ion exchange [18,19], membrane separation [20,21] and electrodialysis [22,23]. Adsorption has been widely used for fluoride removal because of its cost-effectiveness and operational simplicity [11]. In comparison, other methods have some limitations [24,25,26]. For instance, precipitation would generate substantial amounts of sludge while membrane processes have fouling problems and high energy demands. Numerous adsorbents have been investigated, such as activated alumina [27,28,29], magnesium oxide [30,31], rare earth metals [32,33,34] and other metal oxides/hydroxide, including Fe, Ca, Zr and Ti [35,36,37,38]. Activated alumina is the most commonly used commercial adsorbent for fluoride removal. Generally used around pH 5–7, this sorbent has an adsorption capacity about 1–20 mg/L [39]. One disadvantage associated with this material is the residual dissolved aluminum ion in the processed water, which is considered to have negative impact to human health [28]. Compared with alumina, rare earth metal oxides exhibit a fairly high affinity to fluoride. Despite this, their exceedingly high cost prevents them from being regarded as a competitive sorbent. Magnesium oxide is a material with unique physicochemical properties. Having an extremely high PZC (12.1–12.7) [40], this material is expected to carry net positive charges on its surface when the solution pH is less than 12. Earlier investigations have found that active magnesia has a strong affinity for fluoride in water [41,42]. More recently, many magnesium oxide-based materials with nano/porous structures have been reported, exhibiting a fairly good performance in fluoride removal [31,43,44,45]. For instance, Jin et al. [45] synthesized porous MgO nanoplates with an adsorption capacity of over 185.5 mg/g. Gao et al. [31] reported hierarchical meso/macroporous nanostructures of MgO, which showed a high adsorption capacity of 290.67 mg/g. However, direct use of these materials in water treatment may encounter potential difficulties in solid–liquid separation due to their relatively small particle size. Wan et al. [46] synthesized a hybrid MgO–biochar through one-step pyrolysis of peanut shells and MgCl₂. At 25 °C, the synthesized MgO-BC material demonstrated an optimal fluoride adsorption capacity of approximately 50 mg/L, achieving adsorption equilibrium within 400 min. The material exhibited its maximum fluoride removal efficiency under neutral pH conditions, with both acidic and alkaline conditions resulting in a decreased adsorption performance. Therefore, it is necessary to develop an adsorbent that can efficiently remove fluoride while maintaining particle stability, making it convenient for use in water treatment.

In the current study, a nanostructured sorbent nMgO@AC was synthesized by impregnating activated carbon with magnesium oxide. Activated carbon is a commonly used reagent in water treatment due to its abundant pores and high surface area (500–2000 m²g⁻¹) [47]. As an ideal carrier, activated carbon can also enhance the mechanical strength of MgO particles, improve the fluoride adsorption capacity of the material and facilitate solid-liquid separation without changing the water treatment infrastructure. This work aims to develop a material that can effectively remove fluoride in highly alkaline wastewater. The prepared binary material was expected to have a good fluoride-binding property and fast kinetics in a wide range of pH. Numerous tests, including the kinetic, adsorption isotherm and the effect of pH on fluoride sorption by nMgO@AC, were conducted to identify its removal performance. An adsorption mechanism was proposed by analyzing the data from batch experiments and characterization of the material by SEM, TEM, XPS and XRD, prior and post fluoride sorption.

2. Materials and Methods

2.1. Materials

All the chemicals used in the experiment were of analytical grade. Magnesium methoxide solution (6–10 wt.%) was purchased from Sigma-Aldrich(Shanghai)Trading Co., Ltd. (Shanghai, China). Methylbenzene, nitric acid, sodium chloride, sodium bicarbonate, sodium perchlorate and hydrochloric acid were obtained from China Pharmaceutical Group Chemical Reagents Co., Ltd. (Beijing, China). A stock solution of 1 g/L fluoride was prepared from sodium fluoride dissolved in deionized water, and stored in a 4 °C refrigerator before use. Activated carbon was purchased from Calgon Carbon(Suzhou) Co., Ltd. (Suzhou, China).

2.2. Synthesis

Pretreatment of activated carbon [48]: Activated carbon was first ground and then passed through a 200-mesh sieve. The fraction less than 200 mesh was then soaked in concentrated nitric acid and refluxed at 93 °C for 5 h to enhance its hydrophilicity and surface reactivity. After it was cooled to room temperature, the solid fraction was rinsed with deionized water to a neutral pH and dried at 60 °C in a vacuum oven overnight.

Preparation of nMgO@AC [49]: We added 0.36 g of the acid-treated activated carbon prepared above and 20 mL of magnesium methoxide to 100 mL methylbenzene contained in a 250 mL volumetric flask under constant magnetic stirring (resulting in an approximate Mg:C mass ratio of 1:1 based on theoretical calculations). Then, 1.8 mL of deionized water was added dropwise to the mixture, and the content was continuously stirred overnight to ensure complete hydrolysis. The product was then autoclaved at 265 °C and 100 psi for 10 min using a temperature ramp of 1 °C/min. The solid material was then collected and cooled to room temperature. The resulting material was then transferred to a tubular furnace and heated at 450 °C for 300 min under vacuum condition. The magnesia-coated activated carbon thus prepared was named nMgO@AC.

2.3. Characterization

The synthesized material was characterized by a number of instrumentational methods. The specific surface area of the material was evaluated by Quadrasorb SI-MP (Autosorb-iQ2, Quantachrome Instruments, Boynton Beach, FL, USA). The surface morphology of the material was observed by scanning electron microscopy (SEM, Hitachi S-4800, Tokyo, Japan). X-ray diffraction (XRD) data were collected on a SmartLab diffractometer (Rigaku, Tokyo, Japan). The surface chemical composition of the material was characterized by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Fisher, Waltham, MA, USA).

2.4. Batch Adsorption Experiments

Batch experiments were conducted to evaluate the adsorption kinetics and capacity of nMgO@AC for fluoride and examine the impact of solution pH on fluoride removal. The adsorption kinetic test was carried in a polyethylene bottle containing 100 mL of 40 mg/L fluoride solution and 0.1 g of nMgO@AC, with solution pH unadjusted. The bottle was then transferred to a thermostatic shaker working at 180 rpm and at 25 °C. At the desired time of adsorption, a sample was collected and filtered with a 0.22 μm membrane filter, and the fluoride concentration in the filtrate was determined by ion chromatography (ICS1100, Thermo Fisher, Waltham, MA, USA). The equilibrium tests were conducted in 15 mL centrifuge tubes containing 0.01 g nMgO@AC and 10 mL of fluoride solution with varying concentration from 2.5 to 300 mg/L, without adjusting the solution pH. The content of the test bottle was shaken for 24 h at 25 °C to reach an equilibrium before testing the residual fluoride concentration in solution. In the test of the influence of pH on fluoride removal, the initial fluoride concentration was fixed at 40 mg/L, and the initial pH (3–12) of the solution was adjusted using NaOH or HCl solution. For all the experiments, the ionic strength of the solution was fixed at 0.1 M NaClO₄.

3. Results

3.1. Sorbent Characterization

The SEM and TEM images of the MgO-impregnated activated carbon nMgO@AC are shown in Figure 1. Figure 1a showed the low-magnification SEM images of the obtained material, which suggested the magnesium oxide particles were successfully loaded on the surface of activated carbon. The MgO composites were made of multiple small particles, and each small particle exhibited an irregular shape. Under high magnification (Figure 1b), it could be seen the particles had a polyhedral appearance with a secondary particle size of about 200–500 nm, which was consistent with the TEM image shown in Figure 1c. From the selected area diffraction pattern, the bright diffraction rings (111), (200), (220) and (311) of MgO could be observed, suggesting that the product was polycrystalline [50]. Figure 1d demonstrates N₂ adsorption/desorption isotherms and the pore size distributions of the material. The physisorption isotherm could be ascribed to type IV with H3 hysteresis loops, suggesting the product to be a mesoporous adsorbent with slit-shaped pores [51]. The pore diameter and the BET specific surface area were determined to be 12.33 nm and 329.70 m²/g, respectively.

3.2. Adsorption Isotherms

Figure 2 illustrates the fluoride adsorption isotherm of magnesia-coated activated carbon. Fitting of the equilibrium data from batch experiments revealed that the adsorption process followed the Langmuir model, indicating a maximum fluoride-binding capacity of approximately 121.1 mg/g under the given operating conditions, which was higher than that of most MgO-based composites [46,52]. Sundaram et al. [52] synthesized a magnesia/chitosan composite by mixing commercial MgO with chitosan, which has a fluoride adsorption capacity of 4.44 mg/g. Wan et al. [46] using peanut shell and MgCl₂ fabricated biochar-supported magnesium oxide, and the material showed a maximum fluoride-binding capacity of about 50 mg/g based on the Langmuir model at 25 °C. Compared with other common fluoride removal adsorbents such as aluminum [28,29,53] and lanthanum [32,54]-based materials, the adsorbent reported by the current work also exhibits a good performance in fluoride removal. Activated alumina, as a frequently used commercial material in water treatment, enable the uptake of fluoride with capacity of about 1–20 mg/g [28,29]. Lanthanum hydroxide is considered a promising adsorbent for fluoride removal, and it is reported that its adsorption capacity for fluoride was 242.2 mg/g [32]. The adsorption capacities for fluoride of the commercial activated carbon (F400, Calgon) treated with aluminum and lanthanum are 1.07 and 9.96 mg/g, respectively [53,54]. Al³⁺- and La³⁺-exchanged zeolite particles were investigated for their capacity to remove fluoride from water, exhibiting maximum adsorption capacities of 37.27 and 51.50 mg/g, respectively [55]. Considering safety, cost and the defluorination performance, it seems MgO is a promising adsorbent for fluoride removal.

3.3. Adsorption Kinetics

Figure 3 shows the sorption efficiency of nMgO@AC for fluoride from the solution as a function of time. Over 80% of fluoride was removed by the material within an hour. After this, the rate of fluoride binding gradually decreased, with the equilibrium reached at about 100 min. Fitting of the experimental data indicate that the sorption process followed the pseudo-second-order kinetic model, with a calculated rate constant of 0.0244 min⁻¹ (R² = 0.99), suggesting that fluoride adsorption by the material might be a chemisorption process. Compared with magnesium-based materials reported in other literature, nMgO@AC exhibited a fast adsorption kinetic [46,50,56]. For the commercial MgO reported by Kameda et al. [56], it took more than 8 h to remove 70% of the fluoride at the initial concentration of 100 mg/L. The specific surface area of the synthesized material was larger than that of commercial magnesium oxide (4.8 m²/g), which might lead to a faster kinetic process. Jin et al. [50] reported that hierarchical MgO microspheres had a similar kinetic process to this material, with 88% of the fluoride removed in the first 30 min at an initial concentration of 20 mg/L. Wan et al. [46] found that the nanoporous biochar-supported magnesium oxide required 400 min to reach adsorption equilibrium for fluoride removal, potentially due to the biochar’s small pores and electrostatic repulsion caused by its negatively charged surface.

3.4. Influence of Solution pH

Figure 4 illustrates the effect of initial solution pH on fluoride removal by nMgO@AC. The initial pH of the solution had only a minor impact on fluoride binding by nMgO@AC, with more than 95% of fluoride removed in the initial pH domain investigated. Interestingly, despite the great difference in the initial pH of the solutions for the pH domain of 3–11, their final pH at the end of the tests fell into a narrow pH range of 10.2–10.5. The final solution pH was close to the original one, however, when the initial pH was further increased to 12. Similar trends were observed by Sundaram et al. [52] and Yu et al. [57] in their study of fluoride adsorption by a magnesia-based composite. It is speculated that, although designed as an active fluoride-binding reagent, nanostructured MgO impregnated on nMgO@AC happens to work as an alkaline buffer material, causing the final solution pH to be maintained at an elevated level of 10. This performance stood in contrast to most adsorbents, which preferentially adsorbed fluoride under acidic or neutral conditions. For instance, commonly used fluoride adsorbents, such as activated alumina and granular ferric hydroxide, operated optimally at pH 5–7 [28] and pH 3–8 [35], respectively, while lanthanum hydroxide performed best below pH 8 [32]. This exceptional performance was attributed to the high point of zero charge (PZC) of magnesium oxide, which ranged from 12.1 to 12.7. As a result, the surface of magnesium oxide-based materials remained positively charged below pH 12, facilitating fluoride adsorption. In comparison, activated alumina, ferric hydroxide and lanthanum hydroxide, with PZC values of approximately 8.5, 8.0 and 8.8, respectively [27,32,35], became negatively charged under alkaline conditions, leading to electrostatic repulsion with fluoride ions and limiting their effectiveness to acidic and neutral environments. The material demonstrates unique advantages for treating alkaline wastewater that is difficult to handle with conventional adsorbents, as it can be directly applied without requiring any pH pretreatment.

3.5. Adsorption Mechanism

To further clarify the mechanism of fluoride removal by nMgO@AC, the material before and after adsorption was characterized by SEM, XRD and XPS. Figure 5a–c illustrate the material’s surface morphology post adsorption. It is clear that the morphology and structure after adsorption are different from those before reaction. After adsorption, the morphology of the sample turned sharp (Figure 5a). The higher-magnification SEM image (Figure 5b) shows the presence of needle- or flake-like nanostructures. The flaky magnesium oxide particles were arranged vertically, and the needle-like structure formed the edge of the nanosheet. TEM investigation was employed to further confirm the microstructure of the sample (Figure 5c). The particles were assembled with nanoplates with a size of 100–300 nm, which was similar to the reported images of magnesium hydroxide [58].

The structure of the synthesized powder was characterized via XRD (Figure 5d). The peak at a 2θ diffraction angle of 26.60° (JCPDS: 26-1080) could be assigned to the existence of activated carbon. Other peaks at 36.83°, 42.76°, 62.12°, 74.54° and 78.47° were in agreement with the (111), (200), (220), (311) and (222) planes of magnesium oxide (JCPDS: 45-0946), which was in accordance with the result of selecting electron diffraction described above (Figure 1c). The crystal structure changed prior to and post adsorption according to the XRD patterns. As mentioned above, the angles of the diffraction peaks were proven to be that of MgO before adsorption. After adsorption, the peak positions had changed and were consistent with Mg(OH)₂ (JCPDS: 75-1527). It was demonstrated that magnesium oxide hydrolyzed in the adsorption process, resulting in the formation of magnesium hydroxide, which explained the appearance of the lamellar structure in the SEM image. From the XRD patterns, we found no obvious evidence to support the formation of MgF₂ during the adsorption process, and the same conclusion has also been reported [59,60].

Figure 6 shows the XPS survey scan of nMgO@AC before and after adsorption. Comparison of survey-scan XPS spectra (Figure 6a) before and after reaction revealed that there were two new peaks at 830.98 eV and 685.12 eV, which could be readily indexed to F KLL and F 1s [61]. Figure 6c,d show the O 1s spectra from non-absorbed and absorbed magnesia-impregnated activated carbon. Before adsorption, three peaks at 532.59, 531.15 and 529.39 eV were observed on the surface of the sample. The peaks at 532.59 and 531.15 eV could be attributed to absorbed water and O 1s in hydroxyl groups of magnesium hydroxide, respectively. The peak at 529.39 eV could be assigned to O 1s-Mg [62,63]. Post adsorption, the peaks at 532.23 and 531.15 eV ascribed to absorbed water and hydroxyl groups of magnesium hydroxide were observed [62,63,64]. The peak at 529.39 eV disappeared, while the peak area ratio of OH⁻ increased from 34.09% to 79.63%, suggesting the formation of magnesium hydroxide.

On the basis of the analysis above, a schematic adsorption mechanism for fluoride removal by nMgO@AC was proposed, as illustrated in Figure 7. Magnesium oxide particles on nMgO@AC played a major role in the process of fluoride removal. Instead of directly reacting with fluoride, magnesium oxide interacted with water first and led to a fast hydrolysis reaction with the formation of magnesium hydroxide (Mg(OH)₂). Subsequently, fluoride removal occurred through ligand exchange, where fluoride ions (F⁻) in the solution replaced hydroxyl groups (OH⁻) on the surface of magnesium hydroxide. This process is driven by the stronger affinity of fluoride ions for magnesium ions (Mg²⁺) compared to hydroxyl groups, causing fluoride removal from the system.

4. Conclusions

This study successfully developed a novel magnesium oxide-impregnated activated carbon composite (nMgO@AC) with outstanding fluoride removal performance. Systematic batch adsorption experiments demonstrated the material’s superior defluorination characteristics, including rapid adsorption kinetics and an impressive maximum capacity of 121.1 mg/g. Comprehensive characterization through XRD and XPS analyses before and after fluoride adsorption revealed that the exceptional removal efficiency stems primarily from ligand exchange between fluoride ions and hydroxyl groups in magnesium hydroxide formed during hydrolysis. Remarkably, nMgO@AC maintained a consistently high fluoride removal efficiency (>95%) across a broad pH range (3–11). This feature distinguished it from conventional adsorbents, such as aluminum-based materials, which typically perform better under acidic or neutral conditions. The developed composite showed great potential as an efficient fluoride removal material, particularly for treating high-alkalinity wastewater.