Application of Factorial Design In The Optimization of A Procedure For Antimony (Sb) Remediation From Wastewater Employing Mesoporous Array

This study describes the sustainable and eco-friendly synthesis of the silica-based mesoporous structure from the use of alternative amorphous silica extracted from rice husk ash (RHA). The mesoporous material was called MCM-48 (RHA) and its application as adsorbent to the antimony (Sb) remediation in wastewater was tested. The adsorbent was prepared by an ecient and sustainable hydrothermal method, which exhibited an amorphous framework with type IV isotherms and type H1 hysteresis, high surface area (820.94 m 2 g −1 ) and total pore volume (0.55 cm 3 g −1 ) with a narrow mesopores distribution, uniform spherical particles, and well-dened architecture. Multivariate optimization using a factorial design (2 4 ) was employed in the adsorption tests of Sb. The variables evaluated and the conditions selected were: adsorbent mass (45 mg); adsorption time (60 min); pH (ranged from 2 to 10); and concentration of the Sb standard (8 mol L −1 ). The adsorbent material proposed in this study proved to be ecient for Sb remediation in aqueous media, mainly because it is a material with easy access, low-cost, and eco-friendly.


Introduction
Industrial waste, commonly thrown into aquatic environments, contains both inorganic and organic contaminants, representing a global concern for many countries (Akhil et al. 2021). Among the various inorganic contaminants, antimony (Sb) is considered pollutant non-biodegradable and that can cause many environmental damages. Furthermore, carcinogenic contaminants cause health problems, especially for human beings (Viczek et al. 2020). The main industrial activities that can contribute to the discharge Sb in aquatic environments come from the industries of re retardants, pigments, mining, and ceramics (Aquino et  is still necessary to nd an absorbent material that has the following characteristics: (i) low-cost, eco-friendly, and sustainable; (ii) good physical and chemical stability; (iii) excellent textural and structural features; and (iv) high selectivity, so on (Costa and Paranhos 2020; Costa et al. 2020b). In this context, the synthesis mesoporous materials has attracted a great deal of interest for the adsorption process and has already been used with success for removal of the organic compounds (Santos et  The mesoporous structures have attractive features, such as good thermal and mechanical stability, and high surface area, which allows the diffusion and/or adsorption process of the organic and inorganic compounds through their uniform pores and high mesoporous arrangement, as well as ease in the synthesis and functionalization process of these ordered structures (Costa et al. 2014(Costa et al. , 2015(Costa et al. , 2017a. The silica-based mesoporous arrays are synthesized via the hydrothermal method from the use of a surfactant (directing agent), a catalyst (acid or basic), and mainly from a silica source, which is responsible for forming the framework of the amorphous material (Costa et al. 2015(Costa et al. , 2017b.  (Norsuraya et al. 2016), so on. In the present approach, we use the rice husk ash (RHA) as an alternative, inexpensive, eco-friendly, low-cost, abundant, and accessible source of amorphous silica for the synthesis of the mesoporous material with a cubic phase (MCM-48 (RHA)), which was later used as an adsorbent material in the Sb removal in aqueous media.
The most of the approaches found in the literature, which are dedicated to the adsorption studies, are carried out from the univariate optimization of the adsorption tests, which aim at understanding the adsorption mechanism between the adsorbent material and the adsorbate, especially from the correlation of experimental adsorption data with kinetic and isothermal theoretical models (Costa et al. 2014; Costa and Paranhos 2019). However, these approaches are laborious and requires expertise analyst.
Recently, a demand has emerged for the optimization step of the procedures are fast and with reduced number of experiments (Ferreira et al. 2018). In this sense, the multivariate optimization techniques have been shown powerful to evaluation the variables that affect the analytical response in order to obtain the best conditions of optimization to ensure the procedure reliability. Among the multivariate optimization tools factorial design is more employed and allows a preliminary evaluation of the variables for development of linear models Gamela et al. 2020). These tools have numerous advantages, such as: (i) possibility of evaluating synergistic and antagonistic interactions between variables; (ii) possibility of forecasting the system under study in a condition that has not been tested in practice; and (iii) reduces the generation of chemical waste which contributes to the principles of green chemistry (Ferreira et al. 2017;). Factorial designs have been used in several areas, but its use in absorption procedures has not been explored su ciently.
In this context, the factorial design was employed to optimize a procedure for Sb remediation in wastewater. In addition, the adsorbent material used was obtained from a cleaner, low-cost, and ecofriendly approach from the use of alternative amorphous silica extracted from RHA.

Preparation of MCM-48 (RHA) mesoporous array
The mesoporous array (named MCM-48 (RHA)) was synthesized from RHA of the agulhinha variety. Thus, amorphous silica was extracted from the RHA by leaching with sodium hydroxide solution and the MCM-48 (RHA) was synthesized by a hydrothermal route. The extraction of the sodium silicate solution was performed according to our methodology developed recently (Costa and Paranhos 2018), and the MCM-48 (RHA) was synthesized as follows: (i) 10 g of CTAB was dissolved in 70 mL of NaOH (0.75 mol L − 1 ) under constant stirring at room temperature for 1 h; (ii) after this time, 50 mL of the sodium silicate solution from RHA were added slowly into the solution; (iii) therefore, this mixture was stirred at room temperature for 2 h and then transferred to a Te on-lined stainless steel autoclave, which was placed in a vacuum oven and heated at 100 ºC for 48 h; (iv) after this time, the solution pH was adjusted to ~ 10 with HCl (1 mol L − 1 ) and the reactor was left in the oven for another 24 h at 100 ºC; (v) the solid product obtained was ltered, washed with deionized water, and dried in a vacuum oven at 100 ºC for 12 h; (vi) nally, CTAB removal was performed by calcination at 550 ºC for 6 h at a rate of 1 ºC min − 1 .

Characterization of MCM-48 (RHA) adsorbent
The characterization of the RH and RHA was complemented using Scanning Electron Microscopy/Energy Dispersive X-Ray Spectroscopy (SEM/EDS). The prepared MCM-48 (RHA) was characterized using Fourier transform infrared spectroscopy (FTIR) spectra for powder samples in the form of KBr pastilles achieved in the region of 4000 to 400cm − 1 using a Varian 3100 equipment (at room temperature, 32 scans, and a resolution of 4 cm − 1 ). Powder X-ray diffractometry (XRD) analysis was required on a LabX XDR-6000 (Shimadzu) equipment using Cu Kα radiation source (λ = 1.5406 Å) at a voltage/current display of 30 kV/30 mA. The data were collected with a diffraction angle (2θ) ranging from 5 to 80º and scanning rate of 2º min 1 . SEM analysis was achieved in a FEG-XL30 (Philips) equipment with an EDS accessory, operating with the help of a secondary electron (SE) detector and an accelerator power of at 3 kV.
Nitrogen adsorption and desorption isotherms of MCM-48 (RHA) (evacuated for 2 h at 150°C) was acquired using a NOVA 1200 apparatus in the at liquid nitrogen temperature (− 196.15°C). Additionally, the surface area (S BET ) and the pore size distribution (D BJH ) values were calculated from the adsorption data, using the Brunauer-Emmentt-Teller and the Barrett-Joyner-Halenda methods, respectively. Although the MCM-48 (RHA) mesoporous matrix presents a high degree of ordering from the small angle XRD analysis, it is typical that mesoporous materials also have an amorphous diffraction pattern, as seen from the high angle XRD analysis ( Fig. 1(b), which can be attributed to the amorphous condensed silica framework of MCM-48 (RHA) array from the silica source used in its hydrothermal synthesis (Costa et al. 2020d).  S BET : BET surface area; V: pore volume; V T : total pore volume; D BJH : pore diameter.
The SEM images obtained for untreated and treated RH and RHA, as well as for the MCM-48 (RHA) are shown in Fig. 2a(i-iii). The untreated RH showed an external epidermis, which is well-organized and has a rippled surface with an elongated and contorted shape, as well as the appearance of a corn cob. However, after the acid treatment carried out therein ( Fig. 2a(ii), it is possible to observe that the surface of the RH has become more rough, due to the dilution or destruction of the amorphous region of the bers present in the rice husk (Johara et al. 2012; Costa and Paranhos 2018). Thus, the external epidermis of the RHA, presents the same characteristic of the raw RH, however, in the external epidermis it concentrates the greater percentage of silica (Della et al. 2002(Della et al. , 2005. Therefore, Fig. 2a(iii) shows the internal epidermis of the RHA, which shows the porous structure known as the silica skeleton, from the burning of the organic matter of the RH bers, and this region also contains a considerable amount of silica (Liou 2004;Ahmed and Adam 2007).
The morphology of the MCM-48 (RHA) mesoporous material was evaluated by SEM, as shown in the image presented in Fig. 2a(iv). Thus, it is possible to observe that the MCM-48 (RHA) presented an agglomerate of uniform spherical particles from the amorphous silica matrix, which is characteristic of the nanostructured mesoporous architecture of the M41S family (Costa et al. 2015). From the SEM-EDS analysis it was possible to determine the chemical composition of the MCM-48 (RHA) (Fig. 2b). The MCM-48 (RHA) presented a high content of Si and O, as these are the main constituents of the framework of the amorphous material responsible for the formation of the MCM-48 (RHA) mesoporous material.

Multivariate optimization of adsorption procedure
Usually, the adsorption procedures are carried using univariate methodology, which requires a high number of experiments, and consequently greater waste generation and higher cost. To get around these problems, a factorial design was used to optimize the variables involved in adsorption procedures. Table   2 shows a matrix experimental for the full factorial design (2 4 ) containing 19 experiments with real and coded values and the response in function of the % removal of Sb. Before of evaluating the best conditions for the % removal of Sb, it was necessary to evaluate the quality of the linear model obtained from the factorial design. To evaluate check the quality of the linear model, an analysis of variance (ANOVA) was performed, and the results are available in Table 3. Table 2 Matrix of the full factorial design (2 4 ) was performed. In this case, the F_calculated (51.80) was 17-fold higher than the F_tabulated (3.111) at the 95% con dence level. These data demonstrate that the regression of the model is highly signi cant, which gives credibility to the linear model. The second step for the evaluation of model and check if there is lack of t. In the case, a comparison between the ratio of Mean square of lack of t (MS lof ) and Mean Square of pure error (MS pe ) was performed. It was observed that the model does not presented lack of t because the F_calculated (6.156) was lower than the value of F_tabulated (19.00) at the 95% con dence level. In addition, the quality of the linear model also was evaluated by analyzing the graphic of predicted values versus observed, as shown in Fig. 3. Thus, it is possible to observe that the model is well adjusted with 99% regression percentage. This observation con rms the good t of the model that was veri ed in the Table 3.
After data processing, the evaluation of best conditions for adsorption procedure for the % Sb removal was performed by analysis of Pareto graphic (Fig. 4), evaluating the signi cance of the variables and their interactions, at a con dence level of 95% (Ferreira et al. 2018). According to the results presented in  Table 2, that the adsorption times evaluated showed excellent results, especially when combined with the condition of greater mass of adsorbent was used. This behavior can be seen between experiments 9 to 19 as shown in Table 2. Figure 4 also shows that the interaction between adsorbent mass and adsorption time is signi cant with a negative effect, that is, one of the variables must be tested at the maximum level and the other at the lower level. As the adsorbent mass has been xed at the maximum level (45 mg), the adsorption time can be xed using the condition of the central point (60 min).
The pH solution is an important factor affecting the removal of the metal species in aqueous solution.
The dependence on metal adsorption in function of the pH solution is related to the type of metal that is in the solution, and to the state of ionization of the adsorbent functional group, which affects the availability of the binding sites. Under the established conditions, the experiments were carried at pH solution ranging from 2 to 10, as shown in Table 2. Under the established conditions, the experiments were carried at pH ranging from 2 to 10. It is possible to observe in Table 2 that the removal percentage was adequate in the pH range evaluated, except in experiments 1, 2, 3, 5, 6 and 7, which showed removal e ciency below 80%. However, it is possible to verify that in the mentioned experiments, the adsorbent mass tested was 5 mg, which seems to interfere in the Sb removal in function of the tested pH range. In the experiments that were used mass of 25 and 45 mg, the removal percentage stayed above 85%. These observations show that the studied material has a good adsorption capacity in the Sb remediation in solutions with pH (2 to 10) using the mass greater than 25 mg. As the purpose of the applying MCM-48based mesoporous array is to remove Sb from different water samples, we understand that there is no need to establish an optimal pH condition. In this case, we can establish a pH range between 2 and 10 using an adsorbent mass of 45 mg. The no need of pH adjustment is interesting and increases the frequency of the analytical method proposed. In this study, the variable concentration of the Sb standard showed a positive effect. This behavior is very interesting because it shows that adsorbent has the capacity to remove concentrations of Sb considered high.

Application using real samples
The proposed method of adsorption was applied for the Sb removal in samples of environmental interest, in the case of this study in water samples. Commonly, the concentration of Sb in water samples is at a trace level (ppb), in this sense the samples analyzed were enriched with known concentrations of Sb. A total of ve samples were analyzed and Sb concentrations on the order of 8 ppm were added. These samples were submitted to the adsorption procedure with the optimized conditions, and subsequently the nal aqueous solution was subjected to analysis by ICP OES. From Eq. 1 it was possible to calculate the % removal of Sb which varied from 88 to 96%. The adsorbent proposed had not matrix effect in the adsorption process, and thus con rming that can be used in the remediation of Sb in water samples.

Conclusions
In this present approach, we carry out the synthesis of MCM-48-based mesoporous array via an inexpensive, sustainable, and eco-friendly hydrothermal method using an alternative silica source extracted from the rice husk ash. The prepared mesoporous material was subsequently tested as an adsorbent material in the Sb remediation, which was completely useful in the adsorption of this metal. The prepared MCM-48 (RHA) array exhibited an amorphous framework with the N 2 adsorption/desorption isotherms of type IV and type H1 hysteresis, due to the high nitrogen adsorption, surface area, and pore volume, and average pore diameter between the range of mesoporous materials. In addition, the mesoporous matrix presents a narrow mesopores distribution and uniform spherical particles, typical the architectures with well-de ned regular channels. In addition, the variables of adsorption procedure were optimized using a full factorial design (2 4 ). In this case, the factorial design was useful to nd the optimized conditions using a smaller number of experiments. We also highlight the effectiveness of the MCM-48, which does not need to adjust the pH of aqueous solutions, this increases the analytical frequency method proposed.

Declarations
Compliance with ethical standards Ethics approval Not applicable.

Consent to participate
Not applicable.

Consent for publication
The Authors transfer to Springer the publication rights.