Application of Mesopore-Activated Red Mud for Phosphorus Adsorption

In this paper, the mesopore-activated red mud (M-ARM) was prepared by treating red mud (RM) with acid under an ultrasonic batch and through heat treatment at 750 C. The surface area and adsorption average pore width of M-ARM were calculated and obtained values of 13.408 m2 g-1 and 25.160 nm, respectively. Therefore, the maximum adsorption capacity of M-ARM for phosphorus was 200.85 ± 1.66   mg   g − 1 at 318 K and pH = 5 . At a low initial concentration (75 mg L-1), the phosphorus removal capacity by M-ARM material was up to 97.09 ± 0.15 % at 313 K. With the temperature scales varying from 298 to 313 K, the values of Gibbs free energy change ( Δ G ° ) were negative and also vary from -37.47 to -36.68. The phosphorus adsorption process in an aqueous solution is spontaneous, and this adsorption process was exothermic with enthalpy change ∆ H o = − 14.36 . From the results of investigations and calculations of thermodynamic values, kinetics, and adsorption capacity of materials, we can confirm that the materials in this study had a low-cost and potential material for applications to treat phosphorus-contaminated water. In addition, the adsorption kinetics of this material for phosphorus were also studied and discussed.


Introduction
In agricultural production, phosphorus (P) is a necessary nutrient for plant growth. However, if chemical fertilizers are not used properly, phosphorus will leach into the water, leading to harmful eutrophication of the aquatic [1]. When phosphorus levels are elevated in the water system, it stimulates the growth of cyanobacteria and degrades water quality. The phosphate concentrations in domestic wastewater are usually in the range of about 10 to 15 mg L -1 [2]. In natural water bodies, phosphate sources exist significantly in agricultural production, industrial production, household production, and several other fields. To reduce emissions of this nutrient to the aquatic environment, various techniques have been investigated to remove phosphate composition from wastewater before releasing them into the environment. There are several methods of phosphorus treatment such as biological removal, precipitation, adsorption, and ion exchange [2,3]. Biotechnology is widely applied to remove phosphorus in water due to low operating costs, but the phosphorus removal efficiency is not satisfied with the requirements [4]. In recent years, many modern techniques and technologies have also been applied such as electrolysis and reverse osmosis membrane [5][6][7]. A few studies have also reported on the adsorption technology applied to remove organic phosphorus [6,7]. In 2020, two projects at the Tan Rai Alumina Factory in Lam Dong province, Vietnam, exploited more than 1.4 million tons of alumina. Large quantities of red mud (RM) as a by-product of the plant are discharged daily. Therefore, the study of the application of red mud in industrial production such as brick making, cement, and construction concrete, and their application in environmental treatment has also been of great interest to many people [5][6][7][8][9].
Nowadays, considerable attention has been paid to the application of low-cost, effective adsorbents from natural materials, or by-products from a variety of industries. RM is a sludge formed after the alkaline treatment of bauxite ore to produce alumina. RM often has a brick-red color because of the shape of iron compounds [8]. The main compositions of RM include fine particles that are metals in the form of oxides or hydroxides such as Al, Fe, Si, Ti, Ca, and Mg [9]. In recent times, there are many scientific publications on the applications of RM as an adsorbent to treat water contaminated with arsenic, heavy metals [10][11][12][13], dyes, antibiotics, and phenols [14][15][16]. The RM must be activated before being used as an adsorbent; the treatment of RM is to neutralize and remove some unnecessary compounds such as sand, gravel, sodium, and potassium. Normally, M-ARM is by methods such as acid treatment, heat treatment, and combined heat. The purpose of RM activation is to increase the surface area or porosity of the material [17]. The RM usually has a pH value of 10 to 13, so it needs to be treated before use. Several studies have reported the application of activated or inactivated RM to remove phosphorus from aqueous solutions. Some other studies have reported the RM activation method and the maximum adsorption capacity of RM for phosphorus. There are very few studies reported on the kinetics and thermodynamics of adsorption [18][19][20].
In this study, we focus on investigating and reporting the results of RM activated by HCl combined with ultrasonic batch and thermal activation. M-ARM is used as a material to remove phosphorus from an aqueous solution. Remarkably, we investigated their morphology and composition after activation, the influences of pH, contact times, adsorption isotherms, adsorption kinetics, and thermodynamics related to the adsorption process.

Experimental Details
2.1. Materials. The RM used in this study was provided by Tan Rai Alumina Factory in Lam Dong province, Vietnam. In the first step, the RM was dissolved in distilled water, the supernatant (emulsion mixture) was recovered, and the undissolved rock and gravel residue were removed. In the second step, the above emulsion mixture was centrifuged, and the residue (raw RM) was taken to the next step. The raw RM sample was first treated with 4 N hydrochloric acid, at a liquid/solid ratio of 20 mL/gram, for 24 h under stirring at 300 rpm and ultrasonic batch (Elma S300H). After the second step of treatment, we used distilled water several times to wash the residue until the pH reached neutral and continued drying at 100°C for 24 hours. In the third step, heat treatment was performed by placing the treated sample in a Nabertherm LT 15/12 furnace (Germany), calcined at 750°C for 5 h. The samples after heating, cooling, and grinding with agate mortar, M-ARM, were used for further study. The orthophosphate stock solution (1 g L -1 ) was modulated by dissolving the KH 2 PO 4 (Merk) salt into double distilled water. Phosphorus solutions in this study were prepared from the above stock solutions with different concentrations. From the P mass concentration can be calculated the orthophosphate content in the solutions.

2.2.
Characterization of M-ARM Adsorbent. The elemental compositions of M-ARM were performed by the total reflection X-ray fluorescence spectrometry (TXRF) analysis on a TXRF S2 PICOFOX Bruker spectrometer using the gallium internal standard. To examine the surface area and the pore characteristics of M-ARM, the M-ARM adsorbent was analyzed by the Micromeritics-TriStar II 3020 3.02 (USA), Gas (Nitrogen) adsorption/desorption isotherms at 77 K. Scanning Electron Microscopy (SEM) images of M-ARM adsorbent was taken by JEOL JSM-6510LV (Japan) scanning electron microscopy.

Adsorption of Phosphorus by M-ARM.
To study the batch adsorption process of phosphorus by M-ARM, the 500 mL Erlenmeyer flask contained 0.2 g M-ARM and 200 mL of the phosphorus solution, phosphorus in the solution has a concentration in the range of 75-250 mg L -1 . The reaction vessel was shaken at 300 rpm for 24 h. The adsorption process was studied in the temperature range from 298 to 313 Kelvin (K). HNO 3 0.01 N or NaOH 0.01 N was used to adjust pH. After adsorption, the mixture was centrifuged at 10000 rpm for 5 minutes. The supernatant was filtered through a 0.22 μm membrane using for phosphorus analysis. This procedure was adopted for all adsorption experiments, including pH effect, contact time, and evaluations of the isotherms. The phosphorus adsorption capacity by the M-ARM materials was calculated by the [21]: where q e is the phosphorus adsorption capacity by the M-ARM (in mg g -1 of M-ARM); V is the volume of solution (L) used; M is the mass (g) of M-ARM used for the phosphorus adsorption process; C i and C e are the phosphorus concentrations (mg L -1 ) before and after the phosphorus adsorption processes, respectively.

Adsorption Isotherm Study.
Based on the study results in Section 2.3.2, the absorbance values of q e corresponding to each initial concentration of phosphorus solutions were the basis for the study of the following isothermal adsorption processes from 298 to 313 K. The Langmuir adsorption and Freundlich adsorption were calculated as follows [21]: where q e and C e have been described in Formula (1); q m is the maximum phosphorus adsorption capacity (in mg g -1 ); K L is the Langmuir constant (L mg -1 ); n is the Freundlich constant; and Kf = mg ð1−1/nÞ L 1/n /g.

Thermodynamic
Studies. The adsorption of phosphorus by M-ARM has been performed in Section 2.3. With temperatures ranging from 298 to 313 K. The ΔG 0 , ΔH 0 , and ΔS 0 thermodynamic parameters were calculated by using the Van't Hoff equation.

Characteristics of M-ARM
3.1.1. Surface Area (BET) and Pore Volume of M-ARM. Figure 1 shows the SEM image of RM treated only with HCl and without heat treatment. Figure 2 shows the SEM image of RM heated and treated at 750°C. When comparing the results in Figures 1 and 2, we can see that the RM materials after heat treatment have a smaller particle size and improved porosity.
To examine the surface area and the pore characteristics of M-ARM, N 2 adsorption/desorption measurements were performed and depicted in Figure 3. The isothermal plots of N 2 adsorption/desorption for the M-ARM show type IV isotherms. At low pressure, the adsorption process was monolayer, meanwhile at high-pressure adsorption occurred in multilayers.
The pore size distributions of M-ARM were calculated from adsorption and desorption data using the Barrett-Joyner-Halenda (BJH) model. The results in Figures 4(a) and 4(b) showed that most of the peaks in the range of 2 to 50 nm and the pore size of M-ARM were mostly mesopores with a pore width of 2 to 50 nm. Besides, the results of the physical properties analysis of M-ARM materials by the BET method were described in detail in Table 1.    Figure 5 and Table 2. When compared with the elemental compositions RM of different regions around the world shows that the main components are relatively similar, but the content of the elements is different. The results of this comparison are presented in Table 2.

Adsorption of Phosphorus by M-ARM
3.2.1. Effect of Solution pH. The results in Figure 6(a) showed that the phosphorus adsorbed by M-ARM materials was pH dependent. When the pH value changed from 1 to 2, this anion almost was not adsorbed because now, in solution, they exist in the form of H 3 PO 4 (aq) [27]. The maximum adsorption capacity of M-ARM materials for the phosphate anion was maximized when the pH reached a value between 4 and 5 due to at this pH value; the phosphate anion existed in solution in the form of H 2 PO 4 -. When the pH value increased from 7 to 10, the maximum adsorption capacity decreased significantly. At this time, the OHradicals in the solution dominate, and phosphate anions exist in the solution in the form of HPO 4  [19]. From the above statements, there are some differences which, in our opinion, may be due to different RM activation methods. To confirm, we need to survey a point of zero charges (PZCs) of M-ARM materials.
Therefore, we conducted the investigation and established the PZC of M-ARM in the aqueous adsorption. The results were shown in Figure 6(b). The PZC defined the M-ARM surface density of positive charge as equal to that of negative charge in the solution, which means ΔpH = pH f − pH i = 0. Here, pH i was the initial pH value, ranging from 1 to 10; pH f was the pH value after the phosphate ions were absorbed by M-ARM. When ΔpH > 0, the surface of M-ARM has a positive charge due to the surface of this adsorbent becoming protonated. When ΔpH < 0, the surface of the adsorbent has a negative charge due to the surface of M-ARM becoming deprotonated [28].
When the pH i value was from 1 to 2, ΔpH = 0, adsorption did not occur; moreover, phosphorus in the solution exists in the form H 3 PO 4 (aq). When the value of pH i changed from 4 to 10, ΔpH > 0, the surface of the adsorbent has a positively charged due to the surface of M-ARM becoming protonated. When the value of pH i increases from 7 to 10, ΔpH decreases rapidly, approaching zero. At pH i = 5, ΔpH has the most positive value of +0.38; at this pH value, the maximum adsorption capacity of M-ARM materials for the phosphorus was maximized. In solution, phosphorus exists as negative anions; therefore, if the surface of the adsorbent has a positive charge, the phosphate anion adsorption process takes place more favorably (the opposite poles will attract one another). Interestingly, in this study, we determined a pH value of 5 was the best phosphorus

Effect of Contact Time.
We know that adsorption kinetics is an important parameter when investigating the adsorption process. Therefore, in this study, we also investigated the adsorption kinetics. Figure 7 described the adsorption kinetics of phosphorus by M-ARM in the form of pseudo-first-order (PFO) and pseudo-second-order (PSO) models. The results in Figure 6 showed that time affects the adsorption process. During the period from 30 to 90 min, the adsorption process took place rapidly, but when the time was 120 min, the adsorption reached equilibrium. When compared with the results published by Huang et al., there is not much difference [21]. The adsorption kinetics of the process was examined through two adsorption kinetic equations, PFO and PSO adsorptions. The PFO and PSO adsorption models were described by Trung et al. and Abd El-Rahman et al. according to the following formula [28,29]: where q e and q t are the phosphorus amounts uptake per mass of M-ARM at equilibrium and at any time t (min), respectively; K 1 and K 2 (min -1 ) are the constant rates of the PFO and PSO kinetic model, respectively. With the boundary conditions (t = 0, q t = 0 and t = t, q e = q t ), integrating Equations (4) and (5) lead to the following PFO and PSO nonlinearized forms: q t = q 2 e k 2 t q e k 2 t + 1 : ð7Þ Table 3 lists the parameters through the adsorption kinetic investigation. The value coefficient R 2 of PFO was 0.98, and the adsorption capacities calculated by this model were higher than the experimental value. The R 2 value of PSO was extremely high at 0.99 (see Figure 6), and the experimental data fitted the PSO model. Therefore, we can confirm that the PSO model was suitable for the description of phosphorus adsorption kinetics by M-ARM. The adsorption rate of the process depends on the phosphorus concentration gradient (△C) for PFO, and (△C 2 ) for PSO. Therefore, the adsorption rate of phosphorus by M-ARM was fast for 90 min; then, the adsorption rate of phosphorus by M-ARM slows down and the reaction reached adsorption equilibrium at 120 min.

Effect of the Initial Phosphorus Concentration at
Temperatures from 298 to 318 K. The adsorption capacity of M-ARM materials for phosphorus corresponds to separate initial concentration ranges, different temperatures from 298-313 K. Experimental results were shown in Table 4; from the results show that when the phosphorus concentration

Adsorption Science & Technology
increases from 75-200 mg L -1 , the adsorption capacity also increases. However, when the initial concentration of phosphorus was higher than 230 mg L -1 , the adsorption capacity of the material did not increase any more for all 3 temperature ranges of 298, 303, and 313 K. The adsorption centers were filled, or the adsorption process of the material was saturated. When at a low initial concentration (75 mg L -1 ), the phosphorus removal capacity by M-ARM material was up to 97:09 ± 0:15ð%Þ at 313 K. Table 4, Langmuir and Freundlich's isotherm adsorption models were applied to calculate the     Figure 8 showed the relationship between C e and q e of phosphorus adsorption at pH = 5, and temperature ranges of 298, 303, and 313 K. The calculation results obtained in Table 5 showed that when the temperature increases, the phosphorus adsorption capacity by M-ARM also increases. From the above analysis results, it was shown that the adsorption process of phosphorus by the M-ARM material was not only the process of monolayer adsorption on the surface but also by the pores of the material. Based on the Langmuir isotherm adsorption model, we can determine that the maximum adsorption capacity of M-ARM materials for phosphorus at 313 K was 200:85 ± 1:66 mg g −1 . The correlation coefficient (R 2 ) values of the Langmuir adsorption model under different temperature scales vary from 0.95 to 0.98, and the correlation coefficient (R 2 ) values of the Freundlich model were relatively low. Therefore, Langmuir isotherm can be used to describe the phosphorus adsorption process by M-ARM.

Thermodynamic Study.
The adsorption thermodynamic parameters were determined through these parameters to confirm the phosphate ion adsorption feature by M-ARM materials, corresponding to conditions such as temperature and pH of the adsorption process. The Gibbs free energy change (ΔG 0 ), enthalpy change (ΔH 0 ), and entropy change (ΔS 0 ) are the basic thermodynamic parameters of the adsorption process that need investigating following equations: where R = 8:314 J mol -1 K -1 is the gas constant, T is the temperature (K), and K C is the equilibrium constant. K C From the Van't Hoff equation, ΔG 0 , ΔH 0 , and ΔS 0 can be determined through the equation [31]: Clausius-Clapeyron equation describes the relationship between the equilibrium constant K C and two parameters ΔH 0 and ΔS 0 : The values of ΔH 0 and ΔS 0 were determined from the slope and intercept of the plot of LnK C versus 1/T (see Figure 9). The results are presented in detail in Table 6.
From the calculation results in Table 6, the ΔG 0 value was negative for all temperature scales. So phosphorus was adsorbed by M-ARM material in a spontaneous process [32]. Positive entropy values (ΔS 0 = 67:06) showed that phosphate ions were randomly concentrated on the surface of M-ARM materials. When the ΔH 0 = −14:36 value was negative, the adsorption process was exothermic [33]; however, when the temperature of the process increased, the adsorption capacity increased (see Table 5). This phenomenon can be explained that as the temperature increased, the Brownian motion of the particles in the solution increased. Then, the phosphate ions easily got into the pores of the material. Yuan et al. [34] reported that phosphorus adsorption by dolomite mineral was the exothermic process, and Kalaitzidou et al. [35] reported phosphorus adsorption by iron oxyhydroxide material also gave the same result.

Comparing the Results in This Study with Other
Studies. A comparison of the results of this study with the results of other studies is presented in Table 7; the M-ARM material in this study has a relatively high adsorption capacity for phosphorus. For RM, to be able to use them as phosphorus adsorbents, it is necessary to be activated. Depending on the activation method, they have different high or low maximum adsorption capacities, which were consistent with results reported by Yin et al. [36], Li et al. [19], and Liu et al. [37]. When compared with the results of Li et al. [19], the maximum adsorption capacity of M-ARM for phosphate was much higher. According to the report of Liu et al., when activating red mud by heat alone, the phosphate adsorption capacity was not high [37]. When   combined with acid and heat treatment, the phosphorus adsorption capacity of the material increased significantly, which was also true of the results we reported in this manuscript. So, red mud used for phosphorus adsorption must be activated, depending on the activation method, it has a high or low adsorption capacity.

3.3.
Mechanism of the Adsorption. The mechanism of phosphorus adsorption by M-ARM was complex and depends on both the composition and physical properties of the adsorbent. When adsorption occurs, the mechanism of phosphorus adsorption on M-ARM was not simple adsorption in surface complexation mode. There had been some reports that phosphates are adsorbed by gibbsite [16], goethite [39,40], and calcite [41] groups, which were also the main con-stituents of M-ARM. In solution, when the pH = 5:5, the metal oxides Al 2 O 3 , Fe 2 O 3 , and CaO interacted with water molecules and became new hydrous metal oxide sites [40,42]. These hydrous metal oxides are favorably complex with phosphorus on the surface or inside the pores of M-ARM material [17]. Figure 10(a) shows the FTIR spectrum of M-ARM material before phosphorus adsorption. From Figure 10(a), a narrow δ(HOH) band centered at 1636 cm -1 was observed and there was a strong peak at 3445 cm -1 , which was the symmetrical stretching oscillation of ν(H-O-H) [43]. So the composition of this material contains water molecules, perhaps moisture in the air was adsorbed by this material. FTIR spectrum also has an obvious band centered at 469, 618, 736, and 1452 cm -1 assigned to the ν(Fe-O), ν(Al-O), ν(Si-O), and   [42,44]; in addition, a band centered at 1008 cm -1 could be identified as quartz and hematite phases which were typical of a red mud spectrum [45]. Figure 10(b) shows the FTIR spectrum of M-ARM material after phosphorus adsorption. From Figure 10(b) also appeared a narrow δ(HOH) band centered at 1636 cm -1 , and the peak of the symmetrical stretching oscillation of ν(H-O-H) has been shifted to 3550 cm -1 but with a stronger intensity than in Figure 10(b). There were some peaks compared to Figure 10(a) that had been shifted some units, such as ν(Ca-O) (1452 cm -1 shifted to 1461 cm -1 ), ν(Fe-O) (469 cm -1 shifted to 470 cm-1), and ν(Al-O) (618 cm -1 shifted to 619 cm -1 ). In Figure 10(b), some overlapping peaks appear at a strong intensity from 889 to 1134 cm -1 . Evidence obtained from ν(Fe-O-P) and ν(Al-O-P) showed bands centered at 889 and 1143 cm -1 , respectively [42], and a new δ(P-O-P) band was observed at 424 cm -1 [42]. Because in the composition of M-ARM, the calcium content was low, in Figure 10

Conclusions
We have successfully prepared activated red mud, the mesopore-activated red mud (M-ARM) was prepared by treating red mud (RM) with acid HCl under ultrasonic batch and through heat treatment at 750°C. M-ARM has a surface area of 13.408 m 2 /g and an adsorption average pore width of 25.160 nm. M-ARM has a maximum adsorption capacity for phosphorus of 200:85 ± 1:66 mg g −1 at pH = 5 and an adsorption temperature of 313 K. When at a low initial concentration (75 mg L -1 ), the phosphorus removal capacity by M-ARM material was up to 97:09 ± 0:15% at 313 K. The PSO model was suitable to describe the adsorption kinetics of phosphorus by M-ARM; the reaction reached adsorption at 120 min. Langmuir model can be used to describe the phosphorus adsorption process by M-ARM. The ΔG°has negative values from -37.47 to -36.68, corresponding to the temperature scale from 298 to 313 K. From these results, it can be seen that the phosphorus adsorption process in an aqueous solution was spontaneous. ΔH 0 = −14:36 was a negative value indicating that the adsorption process was exothermic. Therefore, M-ARM is a low-cost, easy-to-be-applied, readily available material, and a potential material for treating phosphorus-polluted water by adsorption.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors declare that they have no conflicts of interest.