Optimization of Methylene Blue Ultrasound-Assisted Adsorption onto Magnetic Sugarcane Bagasse Activated Carbon Using Response Surface Methodology

This study focused on the optimization of methylene blue removal on a magnetic activated carbon from the carbon-rich agro-industrial residue, sugarcane bagasse, synthesized by microwave method. The adsorption process was assisted by ultrasound. The magnetic sugarcane bagasse activated carbon (MSB) was characterized by FTIR and SEM-EDX. Based on FTIR results, the functional groups found in magnetic sugarcane bagasse activated carbon are O-H, C=C, C-O, and Fe-O. The SEM results show that MSB is porous with a rough surface. In addition, EDX data found the presence of three main elements, namely C, O, and Fe. Response Surface Methodology (RSM) Box-Behnken Design was applied to analyze the effects of three parameters, including adsorbent dosage (50-100 mg/L), ultrasonic power (100-200 W), and contact time (30-60 min). The obtained optimum conditions of the adsorption process were the sonication power of 155.65 W, the adsorbent dosage of 89.77 mg/L, and the sonication time of 57,81 minutes. The results indicated that the parameters of adsorbent dosage, ultrasonic power, and contact time influenced the response (qe and methylene blue removal).


Introduction
Organic dyes are one of the most significant pollutants in wastewater released by the textile and other manufacturing industries. Methylene blue (MB), a cationic dye, has been used widely in textile industries. However, it is challenging to remove MB from wastewater due to its complex structure. Untreated organic dyes create not only aesthetic problems but also have a negative impact on health, such as carcinogenesis, teratogenesis, and mutagenesis [1]. It is mandatory to use effective treatment methods to remove water contamination before releasing them into water streams. Different wastewater treatment methods are adopted for removing textile dyes, such as biological, chemical, physical, and hybrid treatments. Among several wastewater treatment methods, the adsorption process is a proper mechanism to remove dyes from wastewater. Purification or contaminants removal utilized adsorbent, small porous materials, which can attract target contaminants [2].
Activated carbon is a part of carbon-based adsorbent. Biomass, biomass waste, or industrial waste as carbon sources of activated carbon were selected due to the availability of raw materials and simple fabrication methods. The surface area, the pore structure, the particle size of carbon, the surface acidity, and the functionality of activated carbon influenced the adsorption capacity and interaction between MB and activated carbon. Fabrication of activated carbon is divided into carbonization and activation, which generally utilize thermal activation or chemical activation to promote surface functionality [3]. As mentioned later, the fabrication of activated carbon is generally conducted at relatively high temperatures in conventional convective or conductive heating systems. More extended time to achieve the activation process is the disadvantage of conventional heating systems. There is also a significant risk of overheating, which leads to the complete combustion of activated carbon [4]. Alternatively, microwave heating has the advantages of short heating time, low energy consumption, accurate temperature control, and small equipment size. Microwave-assisted activation can enhance the impregnation process [5].
The spent activated carbon has to be separated from the solutions. Fe(NO3)3·9H2O was used as a precursor to producing magnetic activated carbon. Combining iron oxide and activated carbon will generate magnetic separation by an external applied magnetic field [6]. Many researchers have tried mechanical aid such as ultrasound to improve adsorption capacity. The advantages of ultrasoundassisted adsorption are cavitation formation. The growth and collapse of cavities release energy. This energy is higher than conventional processes like heat or agitation. This cavitation is responsible for increasing the adsorbate mass transfer to the adsorbent's surface [7]. Ultrasound-assisted adsorption has been conducted to remove fenoterol [8], pyridine [9], dyes [10], heavy metal [11] [12], dinitrophenol [13], levofloxacin [7], and other contaminants. Many studies reported that ultrasound could improve adsorption efficiency and gained interest due to its advantages over batch adsorption [14]. Adsorption efficiency could be measured and analyzed by qe and % removal value. qe (mg/g) is related to how many milligrams of adsorbate are adsorbed by the adsorbents. % Removal is proportional to concentration reduction of final and initial concentration of methylene blue solution after the adsorption process.
This study aimed to evaluate the adsorption ability of magnetic sugarcane bagasse activated carbon (MSB) using the ultrasound-assisted adsorption method. This research focuses on how the process parameters, namely adsorbent dosage, contact time, and ultrasound power, affect the efficiency and adsorption capacity. The effects of the process conditions and the optimization of adsorption process were measured with Design Expert software under RSM. Response surface methodology (RSM) is a combination of mathematical and statistical methods to analyze various independent variables effects on a dependent variable. RSM can minimize the number of experiments including chemicals usage, time, cost, and expensive characterization [15].

Experimental Section
Materials. The source of MSB was sugarcane bagasse that provided from local micro enterprises (Jember, Indonesian). Methylene blue was purchased from Merck (Germany). MB stock solution was prepared by dissolving appropriate amount of MB in deionized water. Fe(NO3)3·9H2O (Loba Chemie) was used for the precursor of magnetic activated carbon. All chemicals were reagent grade. Preparation of MSB. Fabrication of MSB refers to previous studies from Jiang et al. [4] Sugarcane bagasse is washed with deionized water to remove dirt particles from the surface. After drying (6.77% moisture content) it was screened to 80 mesh. MSB precursors were prepared by mixing SB and Fe(NO3)3·9H2O with a mass ratio of 1:1.5 in a mortar for 20 minutes. Then, it was calcined for 150 seconds in a microwave oven with a microwave input power of 800 W, a frequency of 2450 MHz. When the calcination is complete, the calcined material is rinsed repeatedly with distilled water until it reaches a neutral pH and then dried. The composite obtained by this microwave heating method is referred to as Magnetic Sugarcane Bagasse Activated Carbon (MSB). Characterization of MSB. The morphology and elemental composition of MSB were investigated using Scanning Electron Microscope-Energy Dispersive X-ray spectroscopy (SEM-EDX). The functional groups present in the adsorbents were analyzed using Fourier Transform Infrared (FTIR) spectroscopy within the range of 4000-600 cm -1 . Design expert software. In this study, three parameters were adsorbent dosage (A), ultrasonic power (B), and contact time (C). All the parameters were studied with three levels. Run data has been processed in Design Expert software using the BBD (Box-Behnken Design) method. BBD (Box-Behnken Design) was chosen to determine the unknown optimal point of the MB adsorption process by MSB. Table 1 shows the ranges of the experimental values of independent variables. Table 2 shows the adsorption results of MB in the 17 runs done. Equilibrium concentration and percent removal was set as the system response (dependent variable).

44
Engineering Chemistry Vol. 2 Ultrasound-assisted adsorption. The ultrasound-assisted adsorption was conducted in the erlenmeyer flask (250 ml). The first experiment determined the optimum conditions (adsorbent dosage, contact time, and power). The operating range is: Adsorbent dosage (250-750 mg/L), contact/sonication time (30-60 min), power (100-200 Watt). The sample solution was centrifuged, and the effluent solutions was record with a UV-Vis (Ultra Violet-Visible) Spectrophotometer at a wavelength of 665 nm. The quantity of adsorbed dyes in each unit mass of the adsorbent (qe) was calculated by equation 1.
C0 represents the initial concentration of the dye in mg/l, and Ce represents the equilibrium concentration of the dye in mg/l. V stands for the volume of the solutions (l), and W represents the mass in grams of the adsorbent. Percent removal was calculated by equation 2.
C0 represents the initial concentration of the dye in mg/l and Ct stands for concentration after time (mg/l).

Results and Discussion
Characterization of MSB. Scanning Electron Microscope-Energy Dispersive X-ray Spectroscopy (SEM-EDX). Magnetic Sugarcane Bagasse Activated Carbon (MSB) samples were characterized using SEM-EDX at the Integrated Laboratory of Diponegoro University. Fig. 1 shows the morphology of MSB at different magnifications. The MSB exhibits rough surface with small number of pores. Pore structure can affect the surface area of activated carbon. The smaller the resulting pore size, the larger the surface area obtained, so that activated carbon can optimally adsorb methylene blue. Characterization was performed using scanning electron microscopy coupled with X-ray spectroscopy via energy dispersive (EDX), which allows qualitative determination of chemical composition. The results show that the elements contained in the adsorbent are 41.32% C, 15  field. However, using microwave-assisted activation and Fe(NO3)3·9H2O as precursor, activated carbon has magnetic response. Fourier Transform Infrared Spectroscopy (FTIR). Fig. 3 shows the results of the FTIR spectrum of MSB. The broadband appears in the absorption area of 3216.62 cm -1 , indicating the presence of an O-H (Hydroxyl) functional group [16]. The absorption peaks observed at 1588.39 cm -1 correspond to the presence of the C=C functional group of the aromatic ring. The band located in the 1104.48 cm -1 is assigned to the presence of the C-O stretching vibration [4]. The sharp peak appears in the absorption area of 542.79 cm -1 , indicating the presence of Fe-O bending vibration of Fe3O4 or γ-Fe2O3 [17]. The functional groups found in magnetic sugarcane bagasse activated carbon are O-H, C=C, C-O, and Fe-O. RSM Approach and Statistical Analysis. Table 2 shows the 17 data which contain the process parameters and the response (qe and % removal). The responses varied between 1-3 mg/g for qe and 6.5-30.5% for % removal. In this study, obtained R square value is equal to 0.9605 for qe and 0.9701 for % removal, near the value of 1. This shows the dye removal variability could be explained by the model [18]. The quadratic polynomial model was obtained for the responses: which, A represents adsorbent dosage (mg/l); B stands for ultrasound power (Watt), and C represents contact time (minute).   The significance of the BBD model design was determined by variance analysis. Table 3 and Table  4 show the ANOVA results for MB adsorption. Based on the results of ANOVA, p-value < 0,05 ensures that the second order polynomial model is significant. A non-significant lack-of-fit test (low F-value and P-value > 0.05) indicates that the model is well validated for MB adsorption [19]. Three process parameters (A, B, C) were found to be significant. The parameters with p-value < 0.05 were assumed to be significant. The coefficient of variation (CV) has an acceptable limit between 0.5 and 13.5% [14]. According to the results, C.V. % on qe response = 8.4 and C.V. % on % removal response = 10.25. The lower the value, are more reproducible. The optimum conditions for MB on MSB were determined using this model, and the results are adsorbent dosage 89.77 mg/l, ultrasound power 155.65 W, and contact time 57.81 min. The supposed adsorption capacity is 2.68 mg/g and % removal = 24.37%. Fig. 4 shows the impacts of two variables: two dimensional contour. Fig. 4 (a, b, c) indicated the combined effects of adsorbent dosage-ultrasound power, dosage-time, and power-time. It can be concluded that both responses increase as the adsorbent dosage and contact time rises. The plot showed that the higher the adsorbent dosage, the greater the gain of Qe and % removal. A higher adsorbent dosage increases the availability of the adsorbent surface area, thereby increasing the removal of methylene blue.

Conclusions
The morphology of MSB exhibits rough surface with small number of pores. The EDX results show that the elements contained in the adsorbent are 41.32% C, 15.76% O, and 36.91% Fe elements. The functional groups found in magnetic sugarcane bagasse activated carbon are O-H, C=C, C-O, and Fe-O. The element Fe in SEM and FTIR results, at once magnetic response of MSB to an external magnetic field have proved that iron oxide compounds are effectively immobilized on activated carbon. The purpose of this study was to evaluate various process parameters on the adsorption capacity MB by MSB. Design expert software was used in this research to find out the optimum condition and analyze the effects of three parameters. The obtained optimum conditions of the adsorption process were the sonication power of 155.65 W, the adsorbent dosage of 89.77 mg/L, and the sonication time of 57,81 minutes.