Nascent Rice Husk as an Adsorbent for Removing Cationic Dyes from Textile Wastewater

We assessed the applicability of rice husk (RH) to remove cationic dyes, i.e., methylene blue (MB) and crystal violet (CV), from water. RH thermally treated at 75 °C showed a higher adsorption capacity than that at high temperatures (300–700 °C). For a suitable CV-adsorption model, a pseudo-first-order model for MB adsorption was followed by the kinetics adsorption process; however, a pseudo-second-order model was then suggested. In the qt versus t1/2 plot, the MB line passed through the origin, but that of CV did not. The Langmuir isotherm model was better than the Freundlich model for both dye adsorptions; furthermore, the adsorption capacity for MB and CV was 24.48 mg/g and 25.46 mg/g, respectively. Thermodynamically, the adsorption of both MB and CV onto the RH was found to be spontaneous and endothermic. This adsorption increased insignificantly on increasing the solution pH from 4 to 10. With an increasing dosage of the RH, there was an increase in the removal percentages of MB and CV; however, adsorption capacity per unit mass of the RH was observed to decrease. Therefore, we conclude that utilizing RH as an available and affordable adsorbent is feasible to remove MB and CV from wastewater.


Introduction
Effluents from textile, leather, cosmetics, paper, pulp and dye manufacturing industries are highly colored because they contain organic chemicals called dyes [1]. The various types of dyes include methylene blue (MB) and crystal violet (CV). MB is an odorless, dark green powder that turns blue in an aqueous solution; it is used for dying cotton or silk, surgeries, and as a hair-colorant [2][3][4]. However, on direct contact, MB causes eye-burns and nausea, and, on ingestion, vomiting and profuse sweating [5]. CV, a protein-based dye also known as gentian violet, is used to enhance bloody-fingerprints and on adhesive tapes to prevent adhesion of powders on their sticky sides [6]. Conversely, CV is carcinogenic, recalcitrant, and non-biodegradable in nature [7]. Therefore, disposal of colored wastewater into an aquatic environment causes reduced photosynthesis due to low light penetration [8], and mutagenesis in both aquatic organisms and humans [9].
Several dye-removal methods have been investigated of late, which include biological [10], and physio-chemical technologies [11,12]. Photochemical and advanced oxidation process [13][14][15], electro kinetic coagulation [16], flocculation [17], nano filtration [18], adsorption [19], and reverse osmosis [20] have also been found to be impressive in removing dye contaminants from aqueous solutions. On the other hand, the majority of these methods are relatively expensive and energy-demanding [21,22]. Biological treatment processes are inefficient due to the non-biodegradability of dyes [11]. Therefore, dye-removal from effluents using adsorption has been found to be more desirable and effective in developing countries [23,24]. This is because adsorption exhibits simplicity, design flexibility, high contaminant selectivity, low operational cost, and ensures excellent treated-effluent quality [25,26].
In recent years, many researchers have focused on developing cheaper adsorbents for dye-removal such as peanut husk [27], wheat straw [28], seed husk [29], walnut shells [30], straw dust [31,32], orange peel [33], date stones [34], cashew nut shell [35], and plum kernel [36]. These agricultural by-products are being increasingly used due to their low ash content, good mechanical strength, few processing needs, superior adsorption potential, availability, and regenerative ability [37][38][39] Rice husk (RH), a major by-product of the rice milling industry is abundant and readily available [40], accounting for 20 % of global gross rice-paddy production, which is estimated to be at 190 million tons as of 2016 [41]. Vietnam and Myanmar together account for 69 million tons of rice production (>7.20 % of global production), thereby, producing 13.8 million tons of RH as by-product [42]. Thus, possible use of RH will yield high economic and environmental benefits [43,44]. RH has been found to consist mainly of 32 % cellulose, 20 % silica, 3 % crude proteins, and 21 % of both hemicelluloses and lignin [45]. Furthermore, chemical stability, good intrinsic strength, inherent resistance, and a silica-cellulose structural arrangement characterize RH [45,46].
Here, RH was an adsorbent to remove MB and CV from water under batch experimental conditions. MB and CV were selected for being cationic, strongly colored, non-biodegradable, mutagenic, and carcinogenic even in very small doses and at short exposure periods [47][48][49]. The impact of thermal treatment on the RH adsorption potential was assessed under different temperatures. The sorption characteristics of MB and CV on the RH were further explored by performing kinetic, equilibrium, thermodynamic, pH, and competitive anion adsorption experiments. In addition, model studies for kinetic, equilibrium, and thermodynamic adsorption experiments were also conducted to elucidate the adsorption characteristics and their mechanisms.

Preparation of Adsorbate Solution
MB (≥ 82.0 % dye content) and CV (≥ 90 % anhydrous basis) of analytical grades purchased from Sigma-Aldrich, Korea, were used here. Amounts of 1000 mg/L MB and CV stock solutions were made by mixing 1 g of the dye (MB and CV) with 1 L deionized water in a volumetric flask. The solutions were conserved in a dark place to impede depolarization.

Thermal Treatment and Characterization of Rice Husk
The RH was acquired from Oryza sativa L. cultivated in paddy fields, Anseong City, Republic of Korea. It was rinsed with distilled water to remove dirt and other contaminants, followed by drying in an oven for 24 h at 80 °C to reduce moisture content after which it was ground and strained using a 30mesh standard sieve (0.595 mm). The impact of thermic operations on the adsorption capacity of the RH for MB was evaluated. A conventional tube furnace with a horizontal stainless-steel tube (diameter: 5.5 cm; length: 55 cm) at various temperatures (75 °C, 100 °C, 300 °C, 500 °C, and 700 °C) was used to activate the RH under anoxic conditions. For the comparison, RH dried at room temperature (RT) was also used. The samples were labeled as untreated RH (RH), RH-75, RH-100, RH-300, RH-500, and RH-700, respectively. A field emission-scanning electron microscope (FE-SEM, S4700, Hitachi, Japan) was employed to inspect the surface morphologies of the RH. N2 adsorption-desorption experiments were performed to measure the specific surface area of RH using a surface area analyzer (Quadrasorb SI, Quantachrome Instrument, USA). From the N2 adsorption-desorption isotherms, the specific surface area, pore volume, and pore diameter were determined by Brunauer-Emmett-Teller (BET) analysis using a surface area analyzer.

Batch Adsorption Experiments
The RH was employed to elucidate the removal of MB and CV under batch conditions. The concentrations for MB and CV were set to be 300 mg/L and 400 mg/L, respectively, and they were made from MB and CV stock solutions (1000 mg/L). Batch adsorption tests were carried out by reacting 0.5 g of the RH with 30 mL dye solutions (MB and CV) in a 50 mL polypropylene tube using a quivering incubator with a stirring rate of 100 rpm and temperature of 25 °C. Kinetic experiments were conducted by differing the reaction time (0.5-24 h) under the concentrations of 300 mg/L and 400 mg/L for MB and CV, respectively. Equilibrium adsorption analyses were conducted using diverse initial concentrations (10-500 mg/L) for MB and CV, respectively; the samples were tested after 24 h. Thermodynamic analyses were carried out by mixing the RH (0. Kinetics adsorption results were analyzed using pseudo-first-order (PFO), pseudo-second-order (PSO) and intra-particle diffusion (IPD) models [50][51][52]. Equilibrium isothermal adsorption data were modeled with the Langmuir and Freundlich models [50,51]. Thermodynamic parameters were explored using the equations published in literatures [50,51]. Figure 1 depicts the adsorption capacity of RH under different temperatures from RT-700 °C with two initial MB concentrations (300 and 500 mg/L). At 300 mg/L, the MB adsorption by untreated rice was similar to that of RH treated at 75 °C and 100 °C, respectively, but was larger than that obtained under high temperatures (300-700 °C). The low adsorption of RH treated under the high temperatures was also observed when the initial MB concentration was 500 mg/L. These outcomes revealed that pyrolysis of the RH under high temperatures did not improve the adsorption of MB. The RH treated at 75 ºC showed higher MB adsorption than untreated RH and that treated at 100 ºC. MB adsorption capacity slightly increased from 25.64 mg/g at RT to 27.58 mg/g at 75 °C. Lower temperatures resulted in higher adsorption efficiency. The decrease in adsorption capacity was ascribed to the certainty that, at higher temperatures, the RH surface melts; and due to high silica content, it is easily transformed to crystabolites, which cause partial blockage of pores [53,54]. The specific surface area and pore and pore diameter were determined by Brunauer-Emmett-Teller analysis and those showed that the RH without thermal treatment had the highest surface-area (35.14 m 2 /g) and smallest pore diameter (5.11 nm). However, with increasing temperature, a smaller specific surface-area and bigger pore-diameter were obtained-as for RH-100 (24.63 m 2 /g and 6.34 nm) and RH-300 (18.38 m 2 /g and 7.56 nm)-due to high silica content, which contributes to agglomeration, reducing the porosity of RH and is consistent with other literature [55,56]. Surface morphology of the RH treated at different temperatures was studied with FE-SEM; the results are depicted in Figure 2. All thermally treated RH was observed to have nodule-like surfaces structures; increased temperature caused their destruction, resulting in the reduction of specific surface-area.

Kinetic Studies
The experimental data for MB and CV adsorption onto RH as a function of reaction-time were represented in Figure 3, with PFO and PSO kinetic model fits. The rates of adsorption for both dyes (MB and CV) were fast during the first 15 min; 16.56 mg/g (MB) and 18.20 mg/g (CV) were adsorbed, respectively. Equilibrium was attained within 3 h for MB but the adsorbed amounts for CV increased continuously during 24 h. The fast increase in the adsorbed dye amount was ascribed to enough vacant adsorption sites at the beginning, leading to easy access of dyes to adsorption sites [57,58]. The slow adsorption-rate at equilibrium was owed to overloading of favorable active sites [8].
To investigate the mechanism involved in the adsorption process, the PFO and PSO models demonstrated a relatively high correlation coefficient (R 2 ) for MB and CV as depicted in Table 1. The PFO model best represents MB adsorption, divulging that the sorption rate involves either intra-particle or film-diffusion [59,60]. The PSO kinetic model best fits CV adsorption, indicating a chemisorption process involving ionic exchange and valence forces via the splitting or swapping of electrons between the RH and CV [61].
The kinetic data was also explored using an IPD model. In order to explore the relevance of the IPD model, the amount adsorbed, qt, vs. the square root of time, t 1/2 , adsorption of MB and CV onto RH at different conditions are mapped according to the theory proposed by Weber and Morris [52]. Figure  4a presents that the rate limiting process is not only influenced by intra-particle diffusion during the adsorption of MB onto RH. The adsorption process of MB involved three stages of diffusion mechanisms. The first being the diffusion of MB through the solution to the external surface of the RH, forming a boundary layer. The passage of the MB adsorption plot through the origin indicated that intra-particle diffusion controlled the adsorption-rate [62]; this was consistent with the PFO observed for the MB adsorption process. Secondly, the slow adsorption of MB through internal pores, which limited the adsorption-rate; and thirdly, the retarded rate of intra-particle diffusion caused by low MB concentration, which helps attain final equilibrium. [63,64]. MB adsorption at the final stage involved the lowest slope of 0.2062 and the highest intercept of 16.401, indicating the largest boundary layer diffusion effect [65]. Conversely, the straight-line CV adsorption plot was observed not to pass through the origin, indicating that the adsorption-rate was not only controlled by intra-particle diffusion [66]. The one-directional linearity of CV adsorption (Figure 4b), suggests a one-stage adsorption process, involving chemisorption, which is consistent with the PSO model [67]. The difference in the adsorption mechanisms exhibited by both, MB and CV, could be due to their molecular structure and differing molecular weight (MB: 319.86 g/mol; CV: 407.99 g/mol).  Table 1. Pseudo-first-order and pseudo-second-order model parameters for methylene blue (MB) and crystal violet (CV) adsorption onto rice husk. qe: the amount of adsorbed dyes at equilibrium (mg/g); k1: PFO rate constant (1/h); k2: PSO rate constant (g/mg/h).

Pseudo-first-order kinetics
Pseudo-second-order kinetics

Adsorption Isotherm
The equilibrium adsorption isotherms are indispensable in understanding the interactive nature between adsorbents and adsorbates, as they are useful in designing adsorption systems. The influence of different MB and CV concentrations (0-120 mg/L) was tested and presented in Figure 5, where equilibrium of the different concentrations in an aqueous phase, Ce (mg/L), and the amount of MB and CV adsorbed on the RH, qe (mg/g), were plotted as a function of the equilibrium concentration of MB and CV in aqueous solution. With increasing concentrations of MB and CV in the aqueous phase ( Figure  3a,b), the adsorbed amounts of MB and CV, respectively, increased linearly; each isotherm represented the observed data fairly well. The selection of best fit model was based on the R 2 value provided in Table 2. Thus, the adsorption data fit well with the Langmuir isotherm, with R 2 values of 0.972 and 0.979, rather than the Freundlich isotherm, for both MB and CV adsorption by the RH. The Langmuir isotherm is an equation assuming monolayer adsorption onto a uniform surface with no interaction between the adsorbed molecules, suggesting the adsorption of MB and CV on the RH surface as a homogenous monolayer [68]. The isotherm implies a monolayer of MB and CV over a homogenous RH surface, reaching saturation after attaining equilibrium, and making it impossible for further adsorption [69]. The CV adsorption onto almond skin waste followed a similar Langmuir isotherm [70]. Furthermore, the values of 1/n for MB and CV were determined to be 0.32 and 0.28, respectively. The 1/n values between 0 and 1, indicate a favorable adsorption process [71].

Langmuir isotherm
Freundlich Isotherm The adsorption capacities of RH for MB and CV were compared with other adsorbents. The maximum amounts of MB and CV adsorbed were found to be 25.46 mg/g and 24.48 mg/g, respectively. The RH showed a favorable adsorption capacity, even though it was low in comparison to that of various adsorbents, as shown in Table 3 for MB, and in studies by Chakraborty, S., Chowdhury, S., Saha, P.D. [44] and Sharma, P., Kaur, H., Sharma, M., Sahore, V. [72] for CV. The maximum reported capacities of the adsorbent materials ranges from 20.29 mg/g to 555.56 mg/g; that of RH (25.46 mg/g) lies at the tail end of the table. It is worth mentioning that most adsorbents ranking higher also had a smaller size (nano or micron). Nano/micro-sized particles cannot be used in filtration systems without post separation process, to separate micro particles from water after treatment. The low adsorption capacity of RH could be attributed to its chemical composition, which includes lignin and silica, that cause a reduction in the bonding strength between functional groups available on its surface and an adsorbate; thus, resulting in low adsorption of MB and CV. Additionally, the surface of RH is smooth, containing wax and fats that shelter the grain, chemically and physically affecting its adsorption properties [73]. However, RH is readily available in Myanmar and Vietnam at an affordable price of ~USD 0.025 per kg [38]. Thus, the RH treated at 75 ○ C (RH-75) for 24 h, with particle size of 0.075-1.16 mm, is environmentally friendly and more cost effective as compared to other adsorbents. Firstly, the granular nature of the adsorbent is advantageous as it is easily separated after filtration and sedimentation without forming a sludge. Secondly, with regard to recycling, by utilizing an agricultural by-product, it helps to valorize waste as a resource to minimize environmental hazards.

Thermodynamics
To better comprehend the thermal impact, the adsorption process was conducted at temperatures of 15 °C, 25 °C and 35 °C. Table 4 shows the magnitude and extent of the adsorption processes. The positive values of ΔH° were 151.53 kJ/mol and 28.50 kJ/mol for MB and CV, respectively, affirming that the adsorption of MB and CV onto RH was endothermic [87]. The type of adsorption may be derived from the magnitude of ΔH°: physisorption, when ΔH° < 20 kJ mol −1 ; ionic exchange, if ΔH° falls between 20 and 40 kJ mol −1 ; and chemisorption, when ΔH° > 40 kJ mol −1 [88,89]. Therefore, the adsorption of MB and CV onto RH demonstrates ionic exchange and chemisorption, respectively. In addition, the positive values of ΔS° suggested an increase in haphazardness at the solid/solution interface during the adsorption of MB and CV onto RH [87,90]. The negative values of Gibb's free energy change (∆G°) with increasing temperature revealed the spontaneous adsorption of MB and CV onto RH. More so, the decrease in ∆G° with increasing temperature indicated that the adsorption process of both dyes on the RH was more favorable at higher temperatures, probably due to the rise in dye molecules movement in solution.

Influence of solution pH
The adsorption of MB and CV onto RH (Figure 6), is presented concerning solution pH as the pH increased from 4 to 10. The removal percentages of MB and CV all exceeded 95 % at the pH range of 4-10. The adsorption capacity of molecules of both, MB and CV, increased as pH increased from 4 to 10, even though the increase was insignificant. The adsorption efficiency is dependent on the solution pH, as variation in pH affects the degree of ionization of adsorptive molecules and the adsorbent surface properties [91]. MB and CV are cationic dyes that exist in an aqueous solution as positively charged ions. Consequently, for charged molecules, the degree of adsorption of these cations onto the surface of the RH could be primarily impacted by the surface charge of the adsorbent, which is dependent on the solution pH. The point of zero charge (pHpzc) of unmodified RH has a value of 2.7 [92]; when the solution pH > pHpzc, the surface of the adsorbent is negatively charged, with more functional groups available for bonding (attributed to deprotonation). In this experimental condition (pH > 4), there was an electrostatic interaction between positively charged dyes and the negatively charged surface of RH particles, prompting favorable adsorption but the electrostatic repulsion between anionic dyes and RH surfaces. This can alternatively be explained by other adsorption methods, e.g., ion exchange, which is consistent with other studies [93].

Effect of Adsorbent Dosage
The efficiency of any adsorbent in the treatment of polluted wastewaters depends mainly on the number of available active adsorption sites [49]. To enhance the removal efficiency of RH and find the optimum dosage amounts, the adsorption capacity of RH and removal percentage were analyzed by varying the dosage in the range of 0.05-1.0 g at initial dye concentrations of 400 mg/L (MB) and 300 mg/L (CV). The results, as plotted in Figure 7, indicate that the MB removal percentage marked up from 53.74% to 97.74% while that of CV increased from 57.40% to 98.19%. The increase may be attributed to the increase in the number of adsorption sites [6,19]. In contrast, with an increasing dose of RH, the adsorption capacity of MB and CV onto RH decreased from 160.41 mg/g to 25.6 mg/g and 68.81 mg/g to 11.78 mg/g, respectively. This decrease could be attributed to the split in the flux or concentration gradient between the solute in solution and solute at the surface of the adsorbent, or to the overlaying of adsorption sites following a reduction in the total adsorption surface-area available to the dye, and a rise in the diffusion pathway. Therefore, the amount of MB and CV adsorbed onto unit weight of RH is reduced, leading to a decrease in the adsorption capacity.

Conclusions
RH, an agricultural byproduct, was utilized as an affordable resource to remove MB and CV from an aqueous solution. Pyrolysis of the RH above 300 C was ineffective in enhancing its adsorption capacity for MB and it was opposite to the removal of dyes by the adsorbents derived from biomass. RH treated at 75 C, had a superior removal capacity than RH at other temperatures (100 C, 300 C, 500 C, and 700 C). The kinetic adsorption for MB and CV was well represented by the PFO model and PSO model, respectively. The adsorption rate of MB by RH was mainly dominated by film-diffusion or intra-particle diffusion, but the adsorption for CV was by chemisorption. The equilibrium adsorption of both MB and CV by RH were well-fitted to the Langmuir model, revealing a homogenous monolayer coverage of MB and CV on the outer surface of the RH. The adsorption of MB and CV onto RH was endothermic and spontaneous. Solution pH barely influenced the adsorption of both, MB and CV onto RH. An increasing RH dosage led to an increase in the removal percentage and a decrease in the adsorbed dyes per unit mass of RH. The adsorption capacity of RH for both, MB and CV was lower than that of other adsorbents in the literature; however, it can be a good candidate for the removal of MB and CV from textile wastewater due to its low price and availability.

Conflicts of Interest:
The authors declare no conflict of interest.