Removal of Acid Brown 354 from wastewater by aminized cellulose acetate nanofibers: experimental and theoretical study of the effect of different parameters on adsorption efficiency

Wastewater effluents usually involve dyes that are dangerous for aquatic life and other environments. Many of these dyes are toxic, carcinogenic, and can cause skin and eye irritation. In this study, firstly aminized cellulose acetate was prepared from cellulose acetate and applied for the adsorption of Acid Brown 354 from aqueous solutions. The effects of different parameters including adsorbent dosage, pH, temperature, and initial concentration of dye on adsorption capacity were examined. Results showed that removal efficiency of dye declined by increasing values of all parameters. Finally, maximum removal of dye was achieved in the presence of 0.1 g adsorbent, pH of 2, and 10 mg/L of initial dye concentration at a temperature of 25 C. Also, different adsorption isotherms were investigated including Langmuir, Temkin, and Freundlich models and results demonstrated that the adsorption isotherm of dye followed the Freundlich model with a correlation coefficient of 0.988 revealing that the bond between the dye and the adsorbent is strong. Finally, kinetic study indicated that the adsorption of dye is exactly governed by pseudo-second-order kinetics explaining that the adsorption process is chemical and the adsorbent can not be reused.


INTRODUCTION
Industrial wastes, which usually include different kinds of dyes and other pollutants, are one of the major causes of contamination of water resources. Dyes are colorful substances that are widely used in printing, textile, plastics, leather tanning, food, pharmaceuticals, cosmetics, pigment, and many other industries (Ahmad et al. ; Seow & Lim ; Katheresan et al. ; Ruan et al. ). Dye pollutants can endanger aquatic life as they are naturally toxic. Therefore, their entrance into water resources is unacceptable and should be prevented by different removal methods (Katheresan et al. ).
Due to the environmental and health concerns associated with wastewater effluents, a wide range of techniques has been developed and utilized for removing dyes, which can be classified into three main categories of chemical, physical, and biological techniques (Seow & Lim ). Some of these conventional techniques include membrane filtration, coagulation and flocculation, sedimentation, chemical precipitation, ion exchange, and adsorption (Wu et  . Adsorption, as a combination of physical-chemical methods, is found to be the most favorable and economic method to eliminate dye pollutants from wastewater (Nasar & Shakoor ; Saber-Samandari et al. ; Guo et al. ). This method has been widely accepted as one of the most effective effluent treatment techniques for removing hazardous inorganic and organic pollutants because it is fast, inexpensive, efficient, and easy to design (Raval et al. ; Nasar & Mashkoor ). An ideal adsorbent for dye elimination should have some favorable features such as high adsorption capacity, large surface area, high porosity, easy availability, mechanical stability, environmental friendliness, compatibility, ease of recovery, and high selectivity to remove a wide range of dyes (Yagub et al. ). Finding a highly efficient sorbent substance is the most important issue for practical utilization of the adsorption method. Bhowmik et al. () investigated the performance of Fe 2 O 3 /Mn 3 O 4 nanocomposite for the removal of methyl orange (MO) from aqueous solutions. They reported that the maximum removal efficiency of MO was achieved at pH of 2.0 and ranged between 70.35 to 98.78% for different sorbent doses of 0.25 to 1.0 g/L, with an initial dye concentration of 100 mg/L. Konicki et al. () assessed the efficiency of ZnO/ZnMn 2 O 4 composite for the removal of basic yellow 28 (BY 28) from aqueous solutions. They reported that the amount of dye adsorbed was improved by increasing the solution pH, where an increase in pH from 3.4 to 11 resulted in the increase of dye adsorption capacity from 36.4 to 55.4 mg/g. Also, they investigated the effect of initial dye concentration on adsorption efficiency and demonstrated that its increase from 10 to 50 mg/L leads to the increase of adsorption capacity from 25.9 to 44 mg/g at pH of 7.0 and temperature of 30 C. Fazlzadeh et al. () investigated removal of acid black dye from aqueous solutions using magnetite nanoparticles. Their results showed that by decreasing the initial pH of the solution and the adsorbent concentration, the adsorption capacity of nanoparticles was increased, while increasing temperature and/or concentration decreased removal efficiency. The highest removal efficiency was achieved for the initial adsorbent concentration of 25 mg/L at pH of 2, reaction time of 60 min, and nanoparticle concentration of 1.2 g/L which was 98%. Song et al. () investigated removal of Acid Brown 348 from aqueous solutions using ultrasound-assisted adsorption on exfoliated graphite. They examined the effect of different parameters such as temperature, contact time, pH, etc. The yield of their proposed method was 90% within 120 min at 40 C and pH of 1 using 2 g/L of the adsorbent. Pandima Devi & Muthukumaran () focused on adsorption kinetics of Acid Brown 48 from aqueous solutions using chicken feathers as biowaste. Equilibrium conditions were established after 24 h of contact time at optimum pH of 2 and confirmed that both Freundlich and Langmuir isotherms can predict adsorption isotherms with good accuracy. Kałużna-Czaplińska et al.
() utilized an advanced oxidization method for the degradation of 1 × 10 À4 M aqueous solution of C.I. Acid Brown 349. They used hydrogen peroxide and sodium hypochlorite together with different concentrations of Fe 2þ ions and obtained 100% removal efficiency under optimum conditions; for example, at pH of 2.5. Pansuk & Vinitnantharat () studied adsorption of Acid Brown 75 and Direct Yellow 162 from aqueous solutions using unmodified fly ash and surfactant-modified granules. They investigated the adsorption capacities of the adsorbents at different contact times, temperatures, and adsorbent concentrations, and found that by increasing temperature, the adsorption capacity increased, which revealed that the adsorption process of anionic dyes is endothermic. Dehghan Abkenar () studied removal of Acid Brown 214 from water solution using fluorene functionalized nanoporous SBA-15 adsorbent. She synthesized the adsorbent and performed different analyses to identify the structural characteristics of the adsorbent. Then she used the synthesized adsorbent for the removal of Acid Brown 214 with a concentration of 50 mg/L and found that the maximum adsorption capacity was achieved at pH value of 4 after 5 min of contact time. Also, kinetic study showed that the adsorption process follows pseudo-second order kinetics. Hassan et al. () investigated removal of Acid Brown 354 using Haloxylon recurvum stem biomass as the adsorbent. They found that the high removal was obtained at acidic pH of 2 with minimum contact time of 20 min. Also, increasing temperature decreased removal efficiency. Their kinetic study showed that the removal process follows pseudosecond order kinetics and isotherms of Langmuir and Temkin best fitted the equilibrium results.
Recently, 3-Chloro-2-Hydroxypropyl Trimethyl Ammonium Chloride (CHPTAC) has attracted much attention in the cationization of cellulose because of commercial availability, good reactivity, and low toxicity (Hashem ; Pal et al. ). For example, Li et al. () prepared a cationic cellulose derivative, cellulose-CHPTAC, by homogeneous etherification of cellulose with CHPTAC in a NaOH/urea aqueous solution and used it as filler/modifier. Also, Morantes et al. () used CHPTAC agent as a cationic graft on cellulose nanocrystals, to form a novel water treatment flocculant. In another study, Golizadeh et al. () reported modification of electrospun cellulose nanofibers using CHPTAC monochloroacetic acid. One of the modified nanofibers that they fabricated was Aminized (cationic) Cellulose Acetate (ACA) which is also used in this study as an adsorbent.
In this study, ACA was used to adsorb and remove Acid Brown 354 (AB354) from aqueous solutions. Then, the effects of different parameters including pH, initial concentration of dye, amount of adsorbent, and temperature on dye removal efficiency were investigated. Also, equilibrium studies on adsorption isotherms were performed to find the best model for describing the adsorption mechanism. Furthermore, kinetic modeling of the adsorption process was carried out to determine the rate of adsorption at different conditions, and the adsorption mechanism.
Based on the above literature review and according to our knowledge, no study has been conducted until now for the removal of AB354 dye from aqueous solutions using ACA adsorbent. In this work, it is shown that the utilized adsorbent can adsorb high amounts of the AB354 dye. Table 1 shows all material together with their chemical formula, molar mass, and suppliers used in this study for the removal of AB354.

Synthesis of aminized cellulose acetate
Firstly, 0.5 g of 1 M NaOH, 8 g of distilled water, 30 g of ethanol (96%), 28 g of CHPTAC, and 5 g of cellulose acetate (CA) were simultaneously poured into a 250 mL Erlenmeyer flask and then the flask was placed on a thermal-magnetic stirrer (Farashafaq, Iran) at a temperature of 70 C for 30 minutes. The desired temperature was observed and controlled manually by a thermometer. Then, the insoluble contents in the flask were filtered through a filter paper (Whatman Ashless, UK). Finally, filtered contents were placed in an oven (BF120E, Iran) for 2 hours to dry the sample at 45 C.
The chemical structure of both CA and ACA nanofibers was analyzed by FTIR method in order to identify functional groups and changes in the chemical structure of CA nanofibers and new functional groups generated due to modification. As FTIR spectra give functional groups and elements of those groups, no other analysis method is needed for chemical structure analysis of ACA nanofibers. Also, average diameters of nanofibers were reported in that work. Because the synthesis procedure was from the mentioned work, therefore no more analysis was needed for investigation of the physical structure of the synthesized nanofibers. Only FTIR analysis was done in this work in order to make sure that the procedure of ACA fabrication was correct.

Acid Brown 354 azo dye
AB354 dye with the chemical formula of C 30 H 20 N 8 Na 2 O 12-S 2 is an acidic azo dye with a red light brown color and is used for wool dyeing and printing, and leather shading (Hassan et al. ). The structural formula of the dye is shown in Figure 2. Due to the acidic nature of the dye, an adsorbent with alkaline nature can remove the dye more effectively. ACA nanofibers are insoluble in water and due to the alkaline nature of the amine group it can be a good candidate for the removal of AB354. The preparation method of the adsorbent used in this study is easy and cheap, which are its advantages over other conventional adsorbents.

Dye adsorption experiments
Aqueous supply solutions (1,000 ppm) of AB354 dye were prepared by dissolving 1 g of solid dye in 1,000 mL of deionized water. Then, working solutions with different concentrations of dye were prepared by consecutive dilution of the stock solution with distilled water. For example, to prepare a working solution containing 2 ppm (mg/L) of dye, 2 mL of the supply solution containing 1,000 ppm of dye was taken and poured into a container and the volume of the liquid within it was augmented to 1 L by adding deionized water. Calibration curves were plotted using the absorbance values of the standard solutions at concentrations of 0.5, 1.5, 2, 4, 6, 8, 10 mg/L at maximum wavelength of 440 nm using a UV-Vis spectrophotometer (Photonix, Denmark) (El-Ashtoukhy et al. ) (Figure 3).
To study the performance of aminized cellulose acetate, adsorption of dye from contaminated waters were performed and concentrations of remaining acid brown 354 were measured at maximum wavelength of 440 nm based on the calibration curve. Batch adsorption experiments were performed by contacting 0.1 g of the adsorbent with desired concentrations (10-150 ppm) of dye in 100 mL of aqueous solution at a temperature of 25 C and pH of 7 for 60 min as base values of the parameters. After 10 min, the suspension was centrifuged and the concentration of AB354 adsorbed on ACA nanofibers was determined by the method described above. The equilibrium adsorption capacity q e (mg/g) and removal rate (R %) were calculated   according to the following equations: where C 0 and C e (mg/L) are initial and equilibrium concentrations of dye, and m (g) and V (L) are the mass of the adsorbent and volume of the dye solution, respectively (Wang et al. ; Hamedi et al. ).
In order to identify changes in the structure of Aminized cellulose acetate nanofibers, FTIR analysis was performed using Fourier-transform infrared spectroscopy apparatus (TENSOR 27, Brucker, Germany) and FTIR spectra were obtained and analyzed.

Adsorption isotherms
The adsorption isotherm is an expression that indicates the equilibrium concentration of adsorbate at the surface of an adsorbent at a constant temperature (Gautam & Chattopadhyaya ). Three isotherm models including Freundlich, Langmuir, and Temkin isotherms were applied and equilibrium adsorption data were fitted. The following equations describe these isotherm models (Asnaoui et al. ; Biftu & Ravindhranath ).
Freundlich adsorption isotherm model Langmuir adsorption isotherm model Temkin adsorption isotherm model where C e is the equilibrium dye concentration (mg/L) on an adsorbent surface, q e is the amount of dye adsorbed at equilibrium (mg/g), k f is the Freundlich constant (mg/g), n is considered as the heterogeneity of the adsorbent surface and its affinity for the adsorbate, q max is the maximum adsorption capacity (mg/g), k is the Langmuir constant (L/mg), K t reveals the equilibrium binding constant (L/ mol), R is the universal gas constant ( Equation (3) can be expressed in linear form as below for use in linear regression: B T is related to the heat of sorption and a positive value for it implies that the sorption process is exothermic (Togue Kamga ).
In order to calculate constants of isotherms, experiments were repeated at optimum values of the parameters and with different values for the initial concentration of dye in aqueous solutions. The volume of the solution was 100 mL in all experiments. Then, constants of isotherms were calculated using linear regression on experimental data.

Adsortion kinetics
Kinetic modeling of any reactive process, including adsorption, gives valuable data about the rate of consumption of a component in a process such as a dye, which is essential for designing a continuous or batch processing system for the process. Also, it helps to identify the limiting step, wherein adsorption can be controlled by different mechanisms including complexation, film diffusion, pore diffusion, and ion-exchange processes (Ofomaja ; Jarrah & Farhadi ). In this study, pseudo-first-order and pseudo-secondorder kinetic models were applied to model the dye removal rate. The pseudo-first-order kinetic model states that the adsorption rate is proportional to the number of free adsorption sites. It is represented by the following equation (Mahmoodi et al. ): where, q t is the amount of dye adsorbed (mg/g) at time t (min), and k 1 is the pseudo-first-order rate constant (min À1 ) (Ho & McKay ). Integration of the above equation in the range of [0, t] results: Reaction curve method (Huang & Jeng ) was used for calculation of kinetic parameters of this model. In this method, the slope of the tangent line on the curve at t ¼ 0 is related to k 1 .
The other kinetic model that was applied on experimental data was the pseudo-second order rate equation. It assumes that chemisorption is the rate-limiting step and is described as: where k 2 (g/mg.min) is second order rate constant (Doulia et al. ). Integration of the above equation with respect to time and re-arranging it will give (Wang et al. ): According to the above equation, the diagram of t/q t with respect to time should give a straight line if pseudo-second order kinetics hold, and q e and k 2 can be determined from the slope and intercept of the diagram, respectively (Wang et al. ).

FTIR analysis
FTIR spectra were taken for both CA and ACA nanofibers to identify their functional groups and changes in chemical structure of CA during fabrication of ACA nanofibers and to investigate the presence of functional groups on the chemically-modified surface by CA ( Figure 4). As indicated in Figure 4(a), vibrations of the acetate group are observed at 1,734 cm À1 (C ¼ O, carbonyl stretch), 1,380 cm À1 (methyl C-H symmetric/asymmetric bend), and 1,131 cm À1 (C-O-C, alkoxyl ester stretch). Also, a broad absorption band at 3,300-3,500 cm À1 corresponding to stretching of OH group was observed (Saleh , ). The absorption band at 2,700-2,900 cm À1 is related to -CH 2 groups (Pielesz & Biniaś ; Kamal et al. ). Figure 4(b) depicts the spectrum of ACA, where new bands are observed at 1,021-1,161 cm À1 , which belong to the stretching vibrations of ammonium groups. In addition, the peaks at 1,453 and 2,921 cm À1 are associated with the stretching and bending vibrational states of methylammonium groups, respectively.

Effects of operating parameters
In this study, the method of one-factor-at-a-time was used for better assessment of each parameter (Frey et al. ). In this method, all variables except the desired one were kept constant during each experiment and only the desired factor was changed to find its optimum value with the objective of maximum dye removal. The starting conditions were selected as solution volume of 100 mL, pH of 7, dye concentration of 10 ppm, stirring speed of 120 rpm, experiment time of 240 min, and temperature of 25 C, based on the literature (El-Ashtoukhy et al. ).

Effect of adsorbent dosage
The adsorbent dose is an essential parameter influencing sorption processes by determining the sorption capacity of an adsorbent for a given initial concentration of the adsorbate.
For investigating the effect of the amount of adsorbent on dye removal performance, five experiments with different values including 0.1, 0.3, 0.5, 0.7, and 1 g of ACA nanofibers were performed while other parameters were kept constant at their base values. Based on the results shown in Figure 5, the removal percentage of dye decreased by increasing the amount of adsorbent. This observation would be explained considering the fact that the accumulation of adsorbent resulted in decreasing space and concentration of dye molecules within the solution and hence at active sites, and diminishing the removal percentage. The optimum adsorbent dose was observed to be its base value, i.e. 0.1 g, and was applied in further experiments. The maximum removal percentage was obtained as 82.2% at optimum amount of the adsorbent.

Effect of pH
The initial pH is also a very significant parameter in the adsorption performance which can influence charges of the adsorbent, ionization/dissociation values of organic pollutants, dissociation of functional groups on the adsorbent active sites, the absorbent surface load, and the dye structure This optimum value has also been reported by Hassan et al. (), which establishes the validity of the experiments at least to this point. As can be seen from Figure 6, removal percentage of the dye at this pH value is 84.53%.

Effect of initial dye concentration
The initial dye concentration plays an important role in the amount of dye adsorbed by the adsorbent and dye removal efficiency. Usually, increasing initial dye concentration leads to a diminishing percentage of dye removal, which is due to the impregnation of adsorption sites on the adsorbent surface (Salleh et al. ; Dawood & Sen ). To study the effect of this parameter, five dye concentrations of 10, 30, 50, 100, and 150 mg/L were prepared at the optimum pH value of 2 and then mixed with the optimum amount of adsorbent, found earlier. The other parameters were kept constant at their base values. Figure 7 shows the effect of dye concentration on its removal using ACA nanofibers as adsorbent. It is clear that by increasing dye initial concentration, its removal percentage decreases which can be attributed to the decreasing and saturation of active sites with the increase of dye concentration. Hassan et al.
() tested dye concentrations as low as 10 À4 mg/L, which is not comparable with the amounts in this work. As reported in the previous works (Yaseen & Scholz ), minimum dye concentration in wastewater is 10 mg/L, which has been tested in this work. Furthermore, dye concentrations as high as 150 mg/L were also tested to show the effectiveness of ACA nanofibers for the removal of this dye. Based on the results shown in Figure 7, an initial dye concentration of 10 mg/L was reported as the optimum value for subsequent experiments. The removal percentage at this value was obtained as 85.24%.

Effect of temperature
Adsorption temperature also plays a significant role in the removal efficiency of dye. Based on the literature, if adsorption capacity improves with decreasing temperature, the adsorption process is exothermic (Dawood & Sen ; Hassan et al. ). In this study, the effect of temperature was investigated at four values of 25, 30, 40, and 50 C while other factors were kept constant at their optimum values. As can be seen from Figure 8, the removal percentage has been decreased with temperature increase, which indicates dye adsorption is an exothermic process (Dawood & Sen ; Hassan et al. ). This result indicates that the adsorption process is spontaneous, because in this case there is a decrease in enthalpy and Gibbs free energy of the system. According to the following equation: where K d is the equilibrium constant. According to the above equation, for removal percentages greater than 50%    K d >1 which is the case in all of the experiments and therefore ΔG is always negative for the adsorption of AB354 dye by the adsorbent, which implies that the adsorption process is spontaneous. Based on the results, the best reaction temperature was considered as 25 C, which is equal to the optimum value reported by Hassan et al. ().
Based on the experimental results and optimum values of different parameters reported here, the maximum removal percentage was obtained as 85.24% with the adsorbent amount of 0.1 g, and initial dye concentration of 10 mg/L at pH of 2 and solution temperature of 25 C.

Modeling of adsorption isotherms
The experimental adsorption isotherm data of AB354 on the surface of aminized cellulose acetate are shown in Figure 9. As was mentioned before, different experiments were conducted at optimum values of the parameters, with different values for the initial concentration of dye while the volume of the solution was 100 mL. Constants of isotherms were calculated using linear regression on experimental data for the case of different initial dye concentration for each model. Trend lines and fitting equations are shown in Figure 9. Also, correlation coefficients are calculated for three models, which are shown in Table 2 and Figure 9. Considering R 2 values of three models, the Freundlich adsorption isotherm fitted experimental data better than the Langmuir and Temkin models, which indicates that the adsorbent surface is heterogeneous and the dye does not form a monolayer on the adsorbent and rather follows multilayer adsorption (Piccin et al. ; Al-Ghouti & Da'ana ). The value of n is larger than unity, which implies that the bond between the dye and absorbent is strong. Also, a positive value is obtained for B T for the Temkin isotherm model, which means that the process is exothermic.

Adsorption kinetics
As mentioned in the Materials and Methods section, kinetic modeling was performed on experimental data in order to reveal the kinetics and mechanism of adsorption on the adsorbent. In this study, pseudo-first-order and pseudo-second-order kinetic models were applied to model dye removal rate.
The plot of the pseudo-first order kinetic model for an initial concentration of 10 mg/L of dye and optimum values of other parameters is demonstrated in Figure 10. Kinetic parameters of k 1 and q e can be calculated from the diagram of q t  versus t (Table 3). As mentioned before, the reaction curve method (Huang & Jeng ) was used for calculation of kinetic parameters. As can be seen, the curve fitting reveals a low correlation coefficient of 0.441 and therefore it can be concluded that the adsorption of AB354 by ACA nanofibers does not follow pseudo-first order kinetics. Also, curve fitting using the pseudo-second order model on experimental data was performed and the plot of t/q t with respect to time was depicted, which is shown in Figure 11. According to Figure 11, exact linearity has been achieved by utilizing the pseudo-second order model with R 2 ¼ 1.000, denoting chemisorption (Doulia et al. ) which means that the adsorbent cannot be regenerated for further uses. Calculated kinetic parameters for this model are also shown in Table 3.

CONCLUSIONS
Removal of dyes from wastewater effluents in textile industries is a challenging problem due to the wide range of dyes and their properties. Many investigations are devoted to this area and different methods are applied for the removal of dye. Adsorption is the most widely used and more reliable method compared to other methods. In this study, removal of acid brown 354 dye from aqueous solutions using a modified adsorbent was performed. For this purpose, the surface of cellulose acetate was modified by CHPTAC for fabrication of aminized cellulose acetate nanofibers as an efficient adsorbent for dye removal. Different adsorption experiments were performed to find optimum conditions of effective parameters. Effective parameters on the adsorption capacity including pH, initial dye concentration, adsorbent dosage, and temperature of the solution were studied. The results demonstrated that the maximum removal of acid brown 354 was achieved at 10 mg/L of initial dye concentration with 0.1 g of the adsorbent while optimum pH and temperature were 2 and 25 C, respectively. Also, equilibrium analysis with different isotherm models was performed and it was revealed that Freundlich isotherm fitted dye removal experimental data better than other models with a correlation coefficient of 0.988, proposing that acid brown 354 does not form a monolayer on the adsorbent and rather follows multilayer adsorption. Also, the obtained value for the exponent of the model was greater than one (1.602), revealing strong bond formation between dye and adsorbent (Al-Ghouti & Da'ana ). Furthermore, it was shown by equilibrium studies that the process is spontaneous. Finally, kinetic modeling of the process was performed and it was shown that adsorption kinetics of the dye followed the pseudo-second order kinetic model with exact fitting on experimental data and the model coefficient was found to be 0.118. Because the pseudo-second order kinetic model governs experimental data, it indicates a chemisorption process and thus the adsorbent could not be used again (Doulia et al. ).

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.