Use of red mud activated at different temperatures as a low cost adsorbent of reactive dye

Red mud, a waste product generated during alumina extraction from bauxite, could be used as a low-cost adsorbent. Here, the effect of thermal treatment on the adsorption of Reactive Blue 19 (RB19) dye by red mud was compared with the adsorption capacity of untreated red mud. Thermal treatment of red mud at 500°C results in an increase in adsorption capacity from 357 mg g -1 (untreated red mud) to 416 mg g -1 , under acidic conditions. Red mud samples thermally treated at 600°C and 800°C show a reduction in adsorption capacity, however, falling to 337 mg g -1 , in acid medium. The change in the maximum adsorption capacity of red mud to RB19 following thermal treatment is associated with specific surface area. Red mud subjected to 500°C can be used for the treatment of water and wastewaters with a higher efficiency than untreated red mud, thus finding possible application in the textile industry.


INTRODUCTION
Brazilian mining activities contribute significantly to global mineral production, including the largest global production of niobium, the second largest production of iron and manganese, and the third largest production of bauxite (BRASIL, 2018).
Mineral production, however, carries with it the generation of a huge amount of residue.In 2017, Brazil benefited more than 35 megatons of bauxite via the Bayer Process, which involves the digestion of crushed bauxite in concentrated caustic soda (NaOH) under elevated temperatures and pressure, generating a residue denominated red mud.According to Fortes et al. (2016), about 10 to 25 million tons/year of red mud are generated in Brazil.Red mud is generally characterized as a complex slurry (45% liquor and 55% solid) of mixed oxides and trace metals in a highly alkaline matrix.The disposal of red mud usually occurs in tailing dams, producing high financial and environmental costs (JONES; HAYNES, 2011; POWER; GRÄFE; KLAUBER, 2011;GOMES et al., 2016).
The textile industry is of great importance in the economies of several countries, generating a large volume of industrial effluents with high concentrations of dyes and chemical products, which can cause serious environmental problems.
In particular, it should be noted that reactive dyes, due to their advantages of having bright and stable colors and being easy to apply, have been identified as the dyes most commonly consumed by the textile industry worldwide (PAPIC et al., https://doi.org/10.1590/S1413-415220180180Antunes, M.L.P. et al. 2004).The processing techniques applied to these effluents are combined biological treatment with physical or chemical processes, which have high efficiency in the removal of organic matter (FIGUEIREDO et al., 2011).However, color removal from effluents is a major problem.Many dyes are stable to light, heat, and oxidizing agents, and are resistant to conventional treatments (OLIVEIRA et al., 2016).
The removal of reactive dyes by a single treatment process is difficult due to their molecular arrangement and excellent solubility in water.The best solution for color removal is a combination of conventional treatments, followed by additional processes, such as chemical oxidation (MOHAN; BALASUBRAMANIAN; SUBRAMANIAN, 2001;LING;WANG;PENG, 2010), membrane filtration (SOJKA-LEDAKOWICZ et al., 2010), different low-cost adsorbents for dye removal (GUPTA; SUHAS, 2009) and coagulation (STEPHENSON;DUFF, 1996).
Adsorption shows the best results and is more frequently applied in industry (CRINI, 2006).Most commercial systems for the removal of dyes from industrial wastewater use activated carbon as adsorbents.Good results have been obtained with different alternative materials and at low cost (GUPTA; SUHAS, 2009).
In this study, the effects of thermal treatment at different temperatures, i.e., 400, 500, 600, and 800°C, of bauxite residue on the adsorption capacity of an anionic reactive dye used in the textile industry, reactive blue 19 (RB19) (Figure 1), was compared to the adsorption capacity of untreated red mud.Untreated and thermally treated red muds used in this work were characterized by pH, electrical conductivity, major oxide composition, mineralogy, and specific surface area.The influence of pH on the adsorption tests in untreated and thermally treated red mud was also evaluated.Adsorption isotherms were studied using Langmuir and Freundlich models to better understand the reactive dye adsorption characteristics of these low-cost adsorbents.The results provide new insight into the relative importance of the use of low-cost adsorbents on cleaner production, especially the reduction of the toxic substances generated by the textile industry.
The pH and electrical conductivity (EC) values for RM, RM400, RM500, RM600, and RM800 were determined using the ratio 1:25 (g of red mud per mL of distilled and deionized water).The system was stirred continuously for 10 minutes at 250 rpm and, then, remained at rest for 60 minutes.After this, the pH and the electrical conductivity in the supernatant liquid were measured using a direct reading equipment by YSI, model YSI 556.High purity standards were used for calibration of pH 4.00 (4.00 ± 0.01 at 25 ± 0.2°C) and 7.00 (7.00 ± 0.01 at 25 ± 0.2°C).The electrical conductivity meter was calibrated using a standard solution of KCl (1.0 mmol L -1 ) of known conductivity, namely 147 μS cm -1 at 25°C.X-Ray diffraction (XRD) was used to determine mineral composition using a Philip X-Pert wide angle X-ray diffractometer, operating at 40 kV and 40 mA, with CuKα radiation.The morphologies of untreated and thermallytreated red mud samples were observed using a JEOL JSM-6010LA Scanning Electron Microscope with an Energy Dispersive X-ray Spectrometer (SEM-EDS).For each sample, the specific surface area (SSA) was measured by BET/ N 2 adsorption, using a Micromeritics ASAP 2010 instrument.

Adsorption studies
Adsorption experiments were undertaken in triplicate using RM, RM400, RM500, RM600, and RM800 and the Reactive Blue 19 (RB19) dye as adsorbate, whose characteristics are shown in Souza et al. (2013).Maximum absorbance at 590 nm was observed for the aqueous solution, regardless of the pH.Adsorption Isothermal studies were carried out using 0.20 g (± 0.01 g) of different samples of red mud and 50 mL of RB19 solution in concentrations of 100 to 5,000 mg L -1 , at pH 4 and 10.The system was maintained under continuous stirring at 250 rpm, for 1 hour and a temperature of 25°C.An aliquot was subsequently collected from the supernatant, which was centrifuged at 3,500 rpm for 5 minutes, and the RB19 concentration was determined.RB19 concentrations were determined by using a Hach DR-2800 spectrophotometer, measuring at a wavelength of 590 nm.Deionized water was used as a blank sample.The amount of RB19 adsorbed onto the different samples of red mud, q e (mg g -1 ) was calculated using the mass balance relationship presented in Equation 1.The percent adsorption (%A) of RB19 was calculated using Equation 2. 5 0.01 at 25 ± 0.2°C).The electrical conductivity meter was calibrated using a standard solution of KCl (1.0 mmol L -1 ) of known conductivity, namely 147 μS cm -1 at 25°C.X-Ray diffraction (XRD) was used to determine mineral composition using a Philip X-Pert wide angle X-ray diffractometer, operating at 40 kV and 40 mA, with CuKα radiation.
The morphologies of untreated and thermally-treated red mud samples were observed using a JEOL JSM-6010LA Scanning Electron Microscope with an Energy Dispersive X-ray Spectrometer (SEM-EDS).For each sample, the specific surface area (SSA) was measured by BET/N2 adsorption, using a Micromeritics ASAP 2010 instrument.

Adsorption studies
Adsorption experiments were undertaken in triplicate using RM, RM400, RM500, RM600, and RM800 and the Reactive Blue 19 (RB19) dye as adsorbate, whose characteristics are shown in Souza et al. (2013).Maximum absorbance at 590 nm was observed for the Isothermal studies were carried out using 0.20 g (± 0.01 g) of different samples of red mud and 50 mL of RB19 solution in concentrations of 100 to 5,000 mg L -1 , at pH 4 and 10.
The system was maintained under continuous stirring at 250 rpm, for 1 hour and a temperature of 25°C.An aliquot was subsequently collected from the supernatant, which was centrifuged at 3,500 rpm for 5 minutes, and the RB19 concentration was determined.Red mud as a low cost adsorbent of reactive dye

Adsorption isotherms
ysis of the adsorption isotherms was achieved by adjusting the Langmuir and dsorption models.An assumption of the Langmuir Model (Equation 3) is that dsorption corresponds to a saturated monolayer of adsorbate molecules on the rface at a constant energy.The Freundlich Model is an empirical isotherm that for non-ideal sorption involving heterogeneous surface energy systems (Equation uir constant related to the energy of adsorption (mg -1 ); um amount of adsorption corresponding to complete monolayer coverage of the g -1 ); lich constant; (1/n is the adsorption intensity).

Adsorption isotherms
Analysis of the adsorption isotherms was achieved by adjusting the Langmuir and Freundlich adsorption models.An assumption of the Langmuir Model (Equation 3) is that maximum adsorption corresponds to a saturated monolayer of adsorbate molecules on the adsorbent surface at a constant energy.
The Freundlich Model is an empirical isotherm that can be used for non-ideal sorption involving heterogeneous surface energy systems (Equation 4).

Adsorption isotherms
of the adsorption isotherms was achieved by adjusting the Langmuir and ption models.An assumption of the Langmuir Model (Equation 3) is that tion corresponds to a saturated monolayer of adsorbate molecules on the e at a constant energy.The Freundlich Model is an empirical isotherm that on-ideal sorption involving heterogeneous surface energy systems (Equation onstant related to the energy of adsorption (mg -1 ); mount of adsorption corresponding to complete monolayer coverage of the constant; is the adsorption intensity).

Adsorption isotherms
of the adsorption isotherms was achieved by adjusting the Langmuir and ption models.An assumption of the Langmuir Model (Equation 3) is that tion corresponds to a saturated monolayer of adsorbate molecules on the at a constant energy.The Freundlich Model is an empirical isotherm that on-ideal sorption involving heterogeneous surface energy systems (Equation nstant related to the energy of adsorption (mg -1 ); mount of adsorption corresponding to complete monolayer coverage of the onstant; is the adsorption intensity).

RESULTS AND DISCUSSION
(4) Where: K L = Langmuir constant related to the energy of adsorption (mg -1 ); q m = maximum amount of adsorption corresponding to complete monolayer coverage of the surface (mg g -1 );

Characterization and influence of pH on adsorption experiments
Table 1 presents the characteristics of RM, RM400, RM500, RM600, and RM800.The pH ranges between 9.9 (RM600) and 10.8 (RM500), confirming the high alkalinity of this material, which is caused by the addition of NaOH during the Bayer process, and is retained even after heat treatment.The values of EC were almost constant, ranging from 3,700 to 3,600 μS cm -1 .Furthermore, thermal activation causes an increase in SSA for RM400 (43.5 m 2 g -1 ), RM500 (44.3 m 2 g -1 ), and RM600 (30.2 m 2 g -1 ) in relation to RM (18.7 m 2 g -1 ).However, the SSA value for RM800 (15.9 m 2 g -1 ) is lower than that of RM.
The morphology of RM, RM400, RM500, RM600, and RM800 particles was observed by scanning electron micrographs, with Figure 2 illustrating the morphology of RM and RM500.Particles of different sizes and shapes were observed.Heterogeneous materials with particle diameters ranging from lower than 1 μm to greater than 10 μm are seen.In addition, the thermal treatment did not change the mineral morphology.Using the SEM-EDS, it is possible to discern that the smallest particles correspond to iron oxides and the largest ones to silicon.
The mineralogical characterizations of RM, RM400, RM500, RM600, and RM800 are shown in Source: elaborated by the authors.Antunes, M.L.P. et al.
i.e.Al(OH) 3 → Al(OH) (S) + H 2 O → Al 2 O 3 + H 2 O.This fact explains the increase in specific surface area in the RM400 and RM500, since the transition aluminas have a higher specific surface area than the original hydroxide (WEFERS; MISRA , 1987).At 800°C, the mineral phase was the same found for RM400, RM500, and RM600, with the exception of calcite, which was decomposed at 681°C (CaCO 3 → CaO (S) + CO 2 ), according to Mercury et al. (2011).In addition, at 800°C the amorphous content decreases and two crystalline phases (Ca 2 Al 2 SiO 7 -gehlenite and Ca 2 SiO 4 ) may be formed by thermal reaction between kaolinite and calcite (LIAO; ZENG; SHIH, 2015).Finally, sodalite is unaffected by the thermal treatment used in this study, since it decomposes only at 1,198°C (MERCURY et al., 2011).
Figure 3 shows the removal of RB19 by RM, RM400, RM500, RM600, and RM800 at pH 4.0 and 10.0.The analysis of RB19 removal demonstrated small differences under each condition.The relative standard deviation for all samples was lower than 4%.For the acidic condition (pH 4.0), the maximum percent adsorption (%A) of RB19 was 96% for lower concentrations (500 mg L -1 ), decreasing to 8% in high RB19 concentration (5,000 mg L -1 ).For all concentrations tested at pH 10.0, the percentage removal of RB19 was between 4 and 27%.This pH value is near the point of zero charge, with a value of around 9.0 for untreated red mud collected in the same aluminium plant located at Alumínio City and previously described by Antunes et al. (2012), indicating that the surface charge of red mud is neutral and, therefore, suggesting a lower adsorption capacity.

Adsorption isotherms
The equilibrium relationship between the concentration of an adsorbate in the solid phase and its concentration in the liquid phase at constant temperature and pressure is described by isotherms.Figure 4 shows the adsorption of RB19 dye onto RM, RM400, RM500, RM600, and RM800, at pH 4 (a) and pH 10.0 (b).The parameters of Freundlich and Langmuir isotherms are shown in Table 3.The relative standard deviation for all samples was lower than 4%.
RB19 adsorption in untreated and thermally treated red mud was higher when the concentration of RB19 in the contact solution was lower.The highest values of R 2 were obtained with the Freundlich model for untreated or thermally treated red mud, regardless of pH.Thus, for this application, the Freundlich model represents heterogeneous adsorption of RB19 in untreated or thermally treated red mud.
At pH 10, the maximum adsorption capacities were lower than those measured at pH 4, as explained above.At pH 4.0, the maximum adsorption capacity (q m ) was observed at 400°C (403 mg g -1 ) and 500°C (416 mg g -1 ) in relation to RM (357 mg g -1 ).For sample RM600, the q m value (370 mg g -1 ) was lower than the values found for RM400 and RM500.Condition RM800 presents the lowest value of adsorption capacity, i.e., 337 mg g -1 .
%A: percent adsorption of RB19.Source: elaborated by the authors.Sodalite is a calcium and sodium tectosilicate of open porous structure.
It can be considered a material with zeolite-type properties and exhibits permanently negatively-charged surfaces (SMICIKLAS et al., 2014).As the RB19 is a cationic dye, the presence of sodalite can be the main mineral phase related to the adsorption capacity of untreated and thermally treated red mud.
The influence of the thermal treatment on the q m values, however, is not associated with the presence of sodalite, since this was found in all samples studied, regardless of the temperature range.Thus, another factor must be responsible for the change in the q m values during thermal treatment.Figure 5 illustrates the linear relationship between q m and SSA and clearly indicates that equilibrium of the RB19 was reached within one hour.The influences of the pH on the experiments were studied using two different values: (a) natural pH of Brazilian red mud (10.0); and (b) in the best pH condition for adsorption of RB19 by red mud (4.0), as described by Souza, Antunes e Conceição (2013).
aqueous solution, regardless of the pH.Adsorption equilibrium of the RB19 was reached within one hour.The influences of the pH on the experiments were studied using two different values: (a) natural pH of Brazilian red mud (10.0); and (b) in the best pH condition for adsorption of RB19 by red mud (4.0), as described by Souza, Antunes e Conceição (2013).
RB19 concentrations were determined by using a Hach DR-2800 spectrophotometer, measuring at a wavelength of 590 nm.Deionized water was used as a blank sample.The amount of RB19 adsorbed onto the different samples of red mud, qe (mg g -1 ) was calculated using the mass balance relationship presented in Equation1.The percent adsorption (%A) of RB19 was calculated using Equation2.Source: elaborated by the authors.
C e = initial and equilibrium concentration of dye, respectively (mg L -1 ); V = volume of solution (L); W = weight of adsorbent sample (g).

Figure 3 -
Figure 3 -Dependence of adsorption of RB19 on C 0 under different pH conditions: (A) pH 4 and (B) pH10.

Figure 4 -
Figure4-Relationship between the quantity of RB19 adsorbed onto the different red mud samples (q e ) versus the equilibrium dye concentration (C e ).RM, RM400, RM500, RM600, and RM800 at (A) pH 4 and (B) pH 10.
Source: elaborated by the authors.

Table 3 -
Parameters of adsorption using Langmuir and Freundlich models for the RM, RM400, RM500, RM600, and RM800 at pH 4.0 and pH 10.