Functionalization of SBA-15 with EDTA and its application in removing Ca 2 + and Mg 2 + ions from hard water

Recebido: 20/01/2020 Aprovado: 26/06/2020 Poor management, changes in drinking parameters and drought subject many parts of the world to survival in the face of shortages of drinking water and the quality of drinking water needs to be assessed for adequate consumption. Hard water can lead to the buildup of mineral deposits in pipes and appliances that use water on a regular basis, affecting the performance and life cycle of these items. In addition, in the presence of soap, fatty acids form an insoluble precipitate with calcium ions, making foam and cleaning difficult, as well as other problems generated by the high content of Ca and Mg ions present in water. The goal of this paper was to study the removal of Ca and Mg ions using SBA-15, a mesoporous material, functionalized with ethylenediaminetetraacetic acid and compared to its pure matrix. This adsorbent was tested in Ca and Mg ions solution with concentration of 250 mg.L and pH = 9, varying the temperature in 25°C and 50°C. Mesoporous materials were synthesized by hydrothermal method and characterized by TG/DTG (Thermogravimetry and Derivative Thermogravimetry), XRD (X-Ray Diffraction) and BET (Brunauer, Emmett and Teller method). Adsorption results showed removal of up to 50% of the Ca and Mg ions in a short time, approximately 5 min.


INTRODUCTION
The quality and quantity of available water resources, changes and management of consumption parameters, as well as the drought itself make many parts of the world survive in the face of scarcity of drinking water (MARINOSKI et al., 2018). Water quality is related to various dissolved and suspended particles and impurities, as well as physical, chemical and biological characteristics (MARCAL-SILVA et al., 2017). Among the types of drinking water, it has those with hard characteristics (high Ca 2+ and Mg 2+ ions content) that represent serious environmental, industrial and domestic problems (OLIVEIRA JUNIOR et al., 2016). Furthermore, in the presence of soap, fatty acids form an insoluble precipitate with leading to the formation of soap foam.
For the treatment of contaminated waters there are different techniques, such as solvent extraction, micro/ultrafiltration, sedimentation and gravity separation, flotation, precipitation, coagulation, oxidation, evaporation, distillation, reverse osmosis, adsorption, ion exchange and electrodialysis (SANTHOSH et. al., 2016). Among these techniques adsorption stands out for being highly versatile and ecologically effective employed in wastewater treatment due to its easy operation, low cost and availability of a wide range of adsorbents. In addition, adsorption can also be applied for the removal of soluble and insoluble organic, inorganic and biological pollutants (SANTHOSH et al., 2016;FIORILLI et al., 2017).
New organic/inorganic mesoporous ordered structures have been widely investigated as adsorbents for waste or contaminated water treatment due to their high surface area, large and uniform pore size. However, the surface area can be easily functionalized with heteroatoms or organic groups to confer specific hydrophobicity/hydrophilicity or to introduce specific binding sites in order to achieve greater interaction (MAJDA et al., 2016;SOLTANI et al., 2017). Among the classes of mesoporous materials stand out the SBA-15 (Santa Bárbara Amorphous) (ZHAO et al., 1998), materials considered promising for application in adsorption because they have a hexagonal array of mesoporous and uniform size with a highly ordered channel matrix. Its surface is easily functionalized with numerous organic functions without compromising the mesostructure (LUZ et al., 2010;MORITZ and GESZKE-MORITZ, 2016;FONSECA-CORREA et al., 2016;THUNYARATCHATANON et al., 2017).
In the adsorption phenomenon, ethylenediaminetetraacetic acid (EDTA) has also attracted considerable attention as adsorbent due to the formation of a strong metal-binder complex, cost effectiveness and availability. Divalent ion capture has been studied by immobilizing EDTA on some surfaces, but there is little about the applications of EDTA-SBA-15 in the removal of ions such as Ca 2+ and Mg 2+ (IQBAL and YUN, 2017), as well as the capture of ions. Calcium ions in EDTA functionalized mesoporous silica matrix have not been found to date. Therefore, the goal of this paper is to synthesize SBA-15 and its functionalization with EDTA to evaluate its potential as adsorbent material in the removal of Ca 2+ and Mg 2+ ions in aqueous solutions in finite bath. It is also made an analysis of the structural, textural and zero charge point. Finally, kinetic and thermodynamic studies were also evaluated.

MATERIAL AND METHODS
SBA-15 mesoporous support was synthesized by the hydrothermal method, according to the methodology proposed by Zhao et al. (1998). The reagents used in the syntheses are showed in Table 1. Post graft modification was applied to modify the structure of the synthesized mesoporous support and the EDTA was impregnated in the SBA-15 support by the wet spot impregnation method. In this technique a minimum volume of solution is used to cover all the dust to the point of leaving it pasty (POPA et al., 2011;DONG et al., 2013;SHENG et al., 2014). Initially, 2 mL of the EDTA solution (0.2 M) is added to 1g of SBA-15 mesoporous support then dried at 30°C for 6h. Figure 1 shows a schematic of the synthesis and impregnation process. Table 2 summarizes a nomenclature to distinguish the synthesized materials: the SBA-15 mesoporous matrix and functionalized by wet spot impregnation.

B)
The thermogravimetric curves (TG-DTG) of SBA-15 mesoporous support were obtained on a SHIMADZU DSC-60 and DTG-60H thermocouple with coupled DTA. For the analysis, approximately 5 mg of sample were used in alumina crucibles that were heated to 750°C, with heating rate of 10°C/min, in an inert nitrogen atmosphere. Through the thermogravimetric curves generated from the samples of noncalcined materials, it was possible to determine the amount of water and organic template present, as well as the temperature ranges in which these molecules are removed. These data are of fundamental importance in determining the lowest calcination temperature required to remove the pore director from the mesoporous support.
For the X-ray analysis of SBA-15 mesoporous materials, the powder method proposed by SETTLE (1997) was employed, which basically consists in uniformizing the sample to make it a fine and homogeneous powder. Depending on the type of material synthesized, analyzes were performed in two steps, according to the low and high angle range. In the first stage, the low angle diffractograms in the range of 0.5 to 3.0º were obtained by a Rigaku MineFlexTM II diffractometer, operating with Cukα radiation at 40kv, 30mA, velocity of 0.5ºg/min and step of 0.01deg. In a second step, the high angle diffractograms in the range of 10 to 60º were obtained by a Shimadzu model XRD -6000 diffractometer, operating with Cukα radiation at 40 kv, 30mA, velocity of 0.5deg/min and step of 0.02deg (GÓMEZ et al., 2016). Structural analyzes as a function of diffractograms were characterized based on reference standards calculated from the COD database for standard chart (00-035-1779) and standard chart (00-035-1779) and using the MAUD® program. Phase Identification from Powder Diffraction Version 2.064.
The zero charge point (ZCP) is defined as the pH at which the surface of a material has a neutral charge. The methodology used for its determination is called "11-point experiment" (BARBOSA et al., 2014). 0.10 mol.L -1 NaCl solutions of different pH values (ranging from 0 to 12) were prepared using distilled water at room temperature, 25ºC. The pH of each solution was adjusted to the required value with HCl or NaOH solution. SBA-15 ZCP and EDTA-SBA-15 were estimated by direct pH measurements before and after contact with the solid in standard solutions (HCl/NaCl and NaOH/NaCl). Approximately 0.1g of the matrix was weighed and added in 50 mL pH adjusted solution starting from 0 to 12. The solutions were constantly stirred for 24h, and then centrifuged and the pH of the final solution was then measured. The obtained data were represented in a graph of the pH variation (ΔpH = final pH -initial pH) as a function of the initial pH of the solutions with the solid. The value of the zero charge point was determined, considering ΔpH equal to 0.
Nitrogen adsorption/desorption spectra were performed at 77K using the BET method. Each analysis contained about 0.1g of sample previously calcined which underwent a degassing process at 300°C for 10 hours. This treatment aims to remove moisture from the surface of the solid. Nitrogen adsorption isotherms for the samples were obtained in the P/P0 range (approximately from 0 to 0.95), providing important information about the materials, such as: surface area, average pore diameter and pores volume. These parameters (SBET, Dp and Vp) were calculated using the autosorb Surface Area & Pore Size Analyzer Program® version 1.55.
To verify the efficiency of removal of Ca 2+ and Mg 2+ ions from drinking water using the matrix SBA-15 B and SBA-15 C the adsorption experiments were performed in a finite bath. To investigate the removal rate in the pure (SBA-15 B) and impregnated (SBA-15 C) silica matrix, aqueous solutions were used at a concentration of 250 mg.L -1 of CaCO3 and pH=9. For each 1.0 g of adsorbent, 100 mL of the solution was used, varying the temperature and constant stirring (200 rpm). At predetermined intervals aliquots of solution were removed and filtered. The supernatant liquid was diluted and the metal concentration determined by titration.
The collected supernatants were analyzed by EDTA-Na titration method according to standard method (19th edition, 1995). To study the removal of Mg 2+ ions at initial concentrations of 250 mg.L -1 , a 0.01 M EDTA solution was used as titrator, a pH 10 NH4Cl/NH4OH buffer solutions and the turning point Eriochrome Black T indicator from violet to blue. Whereas for the study of Ca 2+ ions removal in solutions with initial concentrations 250 mg.L -1 a 0.01 M EDTA solution was used as titrant, pH 12 NaOH solution and the murexide indicator with pink to purple turning point. For both Ca 2+ and Mg 2+ analyzes were performed in triplicate.  template output of the mesoporous material structure relative to the peak in the range 150°C to 200°C on the DTG curve, with 5% and 45% of mass loss, respectively. The stability of the mesoporous silica structure begins from 550°C. Analyzing Figure 2 (B) the material SBA-15 B shows a mass loss below 100°C which is attributed to water desorption. The second mass loss occurs between 200°C and 600°C due to dehydroxylation of silicate networks. According to EZZEDDINE et al., 2015, at high temperatures condensation of the silanol groups of the silica matrix is observed. For EDTA functionalized mesoporous silica (SBA-15 C), mass losses in the same temperature ranges are also observed, as shown in Figure 2 (C). However, in this case EDTA has significant mass losses in this same temperature range between 200°C and 600°C, which represents the breakdown of CN (amine) bonds corresponding to the decomposition of EDTA (SENNA et al., 2015). This can be verified by the decay in the DTG curves of Figure 2 (C), proving the impregnation of the silica matrix with EDTA (EZZEDDINE et al., 2015).

RESULTS AND DISCUSSION
It is important to note that, as the SBA-15 support, the concept of crystallinity cannot be used, as with zeolites, because its walls are made of amorphous silica. The absence of peaks at greater angles indicates that the material is not crystalline. However, it is known that there is an orderly hexagonal network, where one pore is surrounded by six others, generating the characteristic reflection of this material. Thus, phase identification occurs when three to five peaks are observed corresponding to the planes (100), (110), (200), (210) and (300) Figure 3 show three peaks, which confirm the crystal structure of the synthesized materials. According to the powder X-ray diffraction data of the adsorbent, as illustrated in Figure 3 presents three peaks that appear in the range of 2θ between 0.5 to 2°. These peaks are indexed to the reflections (100), (110) and (200), showing that the mesoporous structure of the SBA-15 is maintained after both calcination and functionalization with EDTA. However, the reduction in diffraction intensity reveals some degree of disorder, especially for the less intense peak attributed to organic monolayer installations within the pore structure when functionalized with EDTA as can be seen in the diffractogram of Figure   The high angle X-ray diffraction pattern is typical of SBA-15 amorphous walls with 2θ between 20 and 25° (IQBAL and YUN, 2017) and can be observed in the three diffractograms presented. For the diffractogram for SBA-15 C (with impregnation), was attributed to peaks concerning the molecular structure of EDTA are observed, indicated by * in Figure 4 (C). These peaks are characterized based on reference patterns calculated from the COD database for the standard chart (00-035-1779) and the standard chart (00-035-1779) and using the Phase Identification from Powder (MAUD®) program. The result presented from the molecular crystalline characterization confirms once again the presence of EDTA in the SBA-15 mesoporous silica matrix, demonstrating that the adopted wet spot impregnation technique is efficient in modifying and functionalizing the surface of the SBA-15.
The zero charge point (ZCP) is a parameter that indicates the pH value at which a given solid has a zero charge on the surface. This parameter is relevant because it allows predicting the adsorbent surface charge as a function of pH. The results found for calcined and functionalized SBS-15 are presented in Figure 5. values, it is possible to predict the adsorbent behavior: for solution pH values lower than the ZCP value the adsorbent surface is positive (adsorbs anions) and for solution pH values greater than the ZCP, the surface charge adsorbent is negative (adsorb cations). In this case, below these values, the hydroxyls are protonated and therefore positively charged and for values above those obtained by ZCP analysis, the hydroxyls are deprotonated and negatively charged (FERREIRA and PARK, 2012). It can be seen that SBA-15 C (functionalized) has a higher negative surface charge than SBA-15 B, showing that the impregnation with EDTA was successful, since the EDTA has coordinating oxygen and nitrogen atoms in its structure, promoting a greater number of negative active sites in the mesoporous silica matrix. In this context, the complexation reaction of calcium and magnesium atoms by EDTA is facilitated at a more alkaline pH than in most metals (eg Zn 2+ , Cd 2+ and Pb 2+ ), since at more acidic pH their complexes are less stable, being protonated EDTA instead of complexing calcium and/or magnesium (LEE et al., 2016). Thus, in order to ensure that the adsorption process occurs without the phenomenon of precipitation of Ca 2+ and Mg 2+ ions and the formation of stable complexes, the study was carried out at pH 9, knowing that the adsorption of cations occurs above the point. The zero-charge tests obtained were performed at this pH based on studies already carried out in the capture of calcium ions in zeolitic materials (HERRMANN and KLEIN, 1987;QIN et al., 2010aQIN et al., , 2010b. Figures 6 and 7 show the 77 K nitrogen adsorption/desorption isotherms for the specific surface area obtained by the BET equation and the pore size and pore volume distribution of synthesized mesoporous materials (SBA-15 A, SBA-15 B and SBA-15 C). According to the IUPAC classification, type IV isotherms are observed, as shown in Figure 6. The SBA-15 A, SBA-15 B and SBA-15 C also exhibit type H1 hysteresis, characteristic of mesoporous solids (BIRTH et al., 2014). In addition, adsorption isotherms show a sharp inflection at high relative pressure (P/P0 = 0.6 -0.8), which according to Dong et al., (2013) indicates typical capillary condensation within uniform pores.  Table 3 summarizes the surface area (SBET), pore diameter (Dp) and pore volume (Vp) corresponding to pure SBA-15 with calcination and functionalized with EDTA (SBA-15 A, SBA-15 B and SBA-15 C, respectively). SBA-15 C 435.5 6.5 1.0 As shown in Table 3, the calcined and EDTAimpregnated mesoporous material has a surface area of 197.1 m 2 g -1 , 744.9 m 2 g -1 and 435.5 m 2 g -1 , respectively. As expected, SBA-15 without calcination presents low surface area (197.12 m 2 g -1 ) when compared to areas of calcined SBA-15 and EDTA-functionalized SBA-15. This is due to the blockage of the pores by the organic director used in the synthesis of the silica matrix. Meanwhile, SBA-15 B (calcined) exhibits the largest surface area (744.9 m 2 g -1 ) even when compared to EDTA functionalized SBA-15 C (435.5 m 2 g -1 ). This reduction is a phenomenon also expected after mesoporous silica functionalization, which occurs due to the implantation or suspension of functional group portions inside and outside the walls of the silica structure. Functional group loading tends to decrease specific surface area and pore volume after functionalization. These functional group units block adsorption of nitrogen molecules and suppress the dew point of functionalized mesoporous silica toward lower relative pressure. Thus, it was observed a decrease in pore size and surface area after functionalization (EZZEDDINE et al., 2015). Chong et al. (2019) synthesized SBA-15 with surface area, pore volume and pore diameter equal to 856 m 2 /g, 0.99 cm 3 /g and 7.45 nm, respectively. Although it is the same synthesis method used, such variations can be attributed to the shorter drying and/or calcination time of SBA-15, 12 hours and 3 hours, respectively.
Pore and diameter volume distributions also decreased with functionalization from 1.9 cm 3 g -1 and 6.6 nm (SBA-15 B) to 1.0 cm 3 g -1 and 6.5nm SBA-15 C. This behavior is also related to implantation or suspension of functional group portions within the pores of the silica thereby reducing pore volume. However, no significant changes in the shape of the mesoporous matrix are observed after functionalization and the Revista Verde, v.15, n.3, p.300-308, 2020 structure remains intact, which is in agreement with those presented by structural analysis (XRD).
Adsorption kinetics is one of the most relevant features as it represents the efficiency of the system. The effect of different contact time on the adsorption of Ca 2+ and Mg 2+ ions by SBA-15 B and SBA-15 C is shown in Figures 8 to 11, respectively. The efficiency of SBA-15 as a function of temperature (25°C and 50°C) and pH (without pH adjustment and equal to pH 9) is presented. As shown in Figure 8, initially there is a significant increase in Ca +2 adsorption at both 25°C and 50°C. However, SBA-15 B shows a 5% increase in adsorption with increasing temperature over the same period of time. After this time, at both temperatures, there is a small reduction in adsorption, which remains constant over time, but more significant at 25°C, ie, a reduction of 5% at 50°C and 8% at 25°C. It can be further observed that Ca +2 adsorption at 50°C is almost 2 times higher than at 25°C.
When SBA-15 was functionalized with EDTA (SBA-15 C) (Figure 8b), a similar adsorption profile was observed, that is, an increase in adsorption in the first five minutes and then remained constant over time. However, with SBA-15 C the adsorption capacity is much higher at both temperatures (25°C and 50°C), being more pronounced at 50°C, about 3 times higher when compared to SBA-15 B.
Finally, it can be concluded that in the system without pH adjustment, the temperature increase favors the adsorption of calcium ions from the medium, being a process that demands a higher energy to promote the reaction. The profile of the Ca +2 ion removal curve is similar with and without the pH adjustment of the medium at 9, as shown in Figure 10. However, there is an inversion in temperature behavior with pH adjustment, ie adsorption is greater at 25°C than 50°C for SBA-15 B.
Analyzing SBA-15 C 25°C pH 9 SBA-15 C 50°C pH 9 12 and 20, while at pH 9 a removal of 50 and 40% at the same temperatures, 25°C and 50°C, respectively. In this context, the pH variation favors the adsorption process of calcium ions (Ca 2+ ) at room temperature, which proves to be a spontaneous reaction without the need to increase the medium temperature, just adjusting the pH to charge above the zero charge potential. For the EDTA-impregnated mesoporous silica SBA-15 (SBA-15 C), a similar behavior to the previous one was observed, with an increase in the adsorption capacity, which may be justified by the presence of amine groups present in the matrix surface that contributed with the largest number of active sites for calcium ion adsorption. Removal rates were 38 and 60% for medium without pH adjustment and 50% with pH adjustment at 9 to 25°C and 50°C, respectively. When SBA-15 was functionalized with EDTA, pH adjustment did not influence the removal rate, showing that only the active sites present in the silica matrix after impregnation were sufficient to reach a high removal rate. Analyzing the adsorption of Mg +2 ions (Figure 11), the adsorption profile is similar to the Ca +2 ion adsorption profile with pH adjustment at 9, that is, the adsorption is higher at lower temperature: 40% and 26% for SBA-15 B and 65% and 34% for SBA-15 C at 25°C and 50°C, respectively. As well, the functionalization with EDTA promotes higher adsorption when compared to SBA-15 B. However, the adsorption capacity of magnesium ions (Mg 2+ ) is more effective than calcium ions (Ca 2+ ) adsorption. This may be justified by the fact that the size of magnesium atoms is smaller than those of calcium and thus favor their migration into the pores of mesoporous silica (SBA-15), in addition to the surface bonds that are formed by the presence of active sites after EDTA impregnation. For SBA-15 B, where the medium was not modified at pH the removal rates of 40 and 26% and at pH 9 of 65 and 34% at temperatures of 25°C and 50°C, respectively. Given these data, it is noted that the pH adjustment favors a higher adsorption rate of magnesium ions. It can also be observed that adsorption is fast for both Ca 2+ and Mg 2+ ions, occurring within five minutes and reaching the adsorption equilibrium within ten minutes for both SBA-15 B and SBA-15 C. Short-time adsorption rate shows that the modified SBA-15 surface has a high density of active sites. According to Ferreira and Park (2012) this is related to wet spot impregnation that does not break the mesoporous structure and functional groups being distributed around the matrix and Ca 2+ and Mg 2+ adsorption on the SBA-15 C support behaves as a Ca 2+ /Mg 2+ /SBA-15 chelate.

CONCLUSION
The heat treated or EDTA functionalized lead the SBA-15 a promising adsorbent for reducing the hardness of waste or drinking water. However EDTA functionalized SBA-15 has much higher adsorption capacity, about three times higher than heat treated SBA-15. It should also be noted that EDTA functionalized SBA-15 has higher adsorption potential regardless of pH or temperature, and the adsorption capacity of magnesium ions is about 20% higher than calcium ions adsorption under the same conditions.