Removal of Remazol Black B dye using bacterial cellulose as an adsorbent Remoção do corante Remazol Black B usando celulose bacteriana como adsorvente

Effluents from textile processes, when discarded of inappropriately, have been shown to be a major environmental concern. In this way, different methods can be used, among them adsorption is an economical and efficient technique in the removal of dyes. Therefore, we propose to analyze the adsorptive capacity of bacterial cellulose (CB) against effluents containing the dye Remazol Black B (RBB). CB was produced by the bacterium Gluconacetobacter hansenii and characterized by the techniques of FTIR, DRX, TGA / DTG and pH (PCZ). The RBB removal tests were initially performed at different pHs. From the best experimental condition, new tests were performed at temperatures of 30, 40, 60 and 100 ± 2 °C, 150 rpm, pH 3.5, using 0.5 g of adsorbent in the concentrations of dye from 25 to 65 mg·L. The kinetic study showed that the system balance was achieved in 80 minutes. The experimental data were better described by the pseudosecond order model. The equilibrium results showed that the experimental data fit the Langmuir model (qmax 17.513 mg·g). The thermodynamic parameters of adsorption showed that the process is exothermic, not spontaneous and also presented low system randomness. The activation energy (Ea) was 23.8 kJ·mol characterizing physical adsorption. The residual water was not toxic to animal or microbial cells. Bacterial cellulose proved to be a good low-cost adsorbent, easy to acquire and which can be used in the adsorption process.


INTRODUCTION
The global textile industry is the fastest growing sector economically. World trade in textile fibers -natural and chemical (synthetic / artificial). The sector had revenues of around US $ 1.9 trillion in 2019 and in 2030 this revenue is expected to be US $ 3.3 trillion .Brazil is the 5th largest world producer of textiles, its industries generate approximately 1.5 million direct employees [1]. This industrial segment is one of the largest consumers of water [1,2]. And it is estimated that during the dyeing and finishing stages, around 50 to 100 L of effluent are produced per kilogram of fabric produced [3,4].
Among these compounds present in the textile effluent there are dyes, complex molecules whose purpose is to give coloring to synthetic or natural fibers [5,6]. When discarded inappropriately in aquatic environments, it prevents the penetration of sunlight into the deeper layers, altering the photosynthetic activity of the environment, decreasing the oxygen solubility, and resulting in toxic effects on aquatic fauna and flora [7]. In addition, azo-type dyes (Remazol Black B, for example) can decompose into aromatic amines with possible carcinogenic potential under anaerobic conditions [7,8].
Due to this problem, different treatments have been used for the remediation of environments contaminated by dyes, among them oxidation or ozonation, electrochemical treatment, reverse osmosis, precipitation, coagulation and flocculation, membrane separation and photochemical degradation can be mentioned [9,10]. However, these methods are not widely used due to their high cost [10]. An alternative method for the treatment of textile effluents is adsorption, an efficient and economical process that involves the adhesion of molecules of the fluid (pollutant) on the solid surface of the adsorbent (biological matrix) [11,12]. Different adsorbents of synthetic or natural origin can be used to remove contaminants [13]. Among these, bacterial cellulose (CB), obtained by different bacteria has shown to be very promising, as it has functional groups on its surface capable of capturing these contaminants [14]. In addition to being a renewable and easily obtainable material [15,16].
Studies have shown how effective the adsorbents from bacterial cellulose are in the treatment of textile effluents. CB was able to adsorb 654 mg·g -1 of the Congo red dye in solution [17]. Chen et al. (2020) [18], also studying the adsorption of the Congo red dye obtained an adsorptive capacity of 230 mg·g -1 . Ali et al. (2012) [19] obtained 25 mg g -1 against the direct blue dye 15. The activated carbon obtained from bacterial cellulose was able to adsorb 507.5 mg·g -1 from the methylene blue dye [20]. When functionalized, CB also shows good removal results. Studies by Huang et al. (2020) [21] with PEI-Pt modified CB showed that this material was able to capture 1157.9 mg·g -1 of the acid-black dye and 13.5 mg g -1 of methylene blue. These results reveal how potential bacterial celluloses are for the treatment of textile effluents.
This work aimed to obtain and characterize the bacterial cellulose produced by Gluconacetobacter hansenii and use it as an adsorbent for the removal of the Remazol Black B dye in synthetic solutions. The influence of pH on the solution and temperature were investigated. In addition to kinetic, equilibrium and thermodynamic studies of the adsorption process.

Inoculum production
The inoculum was produced according to the methodology proposed by Coimbra (2016) [22]. The culture of G. hansenii was initially activated in HS broth (without the addition of bacteriological agar). After broth incubation for 48 h at 30 °C, 100 µL of this culture was sown in 90 mm diameter Petri dishes containing solid HS (obtained by adding bacteriological agar), the volume being distributed over the entire surface of the medium with the aid of a Drigalski handle. The plates were incubated for seven days at 30 °C for growth of cell mass on the surface of the medium. The grown cell mass was transferred to Erlenmeyer flasks containing HS broth in a volume equivalent to 1/5 of its capacity and maintaining the biomass proportion, that is, two plates for each 50 mL of broth. The medium was subjected to stirring with a magnetic bar until the lumps of cells and cellulose were broken up and incubated at 30 ± 1 °C and 150 rpm of agitation for 48 h, before being used as inoculum.

Determination of moisture and ash content
The moisture content was determined by drying the samples in an oven (Tecnal, TE-393/1) at 105 ± 2 °C for 8 hours. This analysis helped to determine the wet and dry mass. After drying, the cellulose was crushed and sieved in a range of 100 µm. The total ash content was determined by measuring the residue obtained after the material was incinerated in a muffle furnace (Thermo scientific) at 650 ± 5 ºC for 4 hours. All analyzes were performed in triplicate.

Infrared spectroscopy with fourier transform (ATR-FTIR)
ATR-FTIR analysis was used to identify the main functional groups of the adsorbent. For this, a Bruker Tensor 27 spectrometer with the attenuated total reflectance accessory (Platinum ATR) was used. The spectra were recorded in the spectral range from 4000 to 500 cm -1 , with a resolution of 2 cm -1 and 20 scans.

X-ray diffraction-(DRX)
In order to evaluate the behavior of the cellulose structure, that is, crystalline and amorphous regions, an X-ray diffractometer (XRD-6000 / Shimadzu) was used. The applied conditions were: 40 kV, angular range 4º to 60º (Bragg -2θ angle), angular variation 0.05º and counting time of 1s. Through Equations 1 and 2 it was possible to determine the values of the average width of the crystallites in the lattice plane of greater intensity ( ℎ ) and the crystallinity index (CI).

Determination of the zero-charge point (pHpcz)
The zero-charge point (PCZ) is the pH value at which the adsorption of potential-determining ions (H + and OH -) are equal. To determine the PCZ, the methodology proposed by Silva et al.
The solutions were prepared with distilled water. The concentrations of the solutions were diluted from a stock solution (100 mg·L -1 ). To adjust the pH of the solutions, solutions of hydrochloric acid (Merck Millipore) and sodium hydroxide (Merck Millipore, CAS 1310-73-2), both 3 mol·L -1 , were used. The initial and final concentrations of the initial dye solution were determined by a spectrophotometer (Hewlett-Packard, model 8453), measuring the absorbance at a length of 597 nm, with water as white.

Influence of pH on the adsorption process
The influence of pH was assessed according to studies by   [25] and Nascimento et al. (2017) [26] with few modifications. The adsorbent was subjected to tests conducted in a batch system, varying only the initial pH value of the solution (3.5 to 9). The fixed experimental conditions were: mass of the adsorbent (0.5 g) and the concentration of the dye solution (25 mg·L -1 ). The system was conducted at 30 ºC, 150 rpm and a contact time of 12 hours. At the end of each test, the samples were centrifuged at 3000 rpm for 10 minutes and then the final concentrations of the solutions were read on a UV-Vis spectrophotometer at 597 nm with water as white. Through the results it was possible to determine the highest percentage of removal, this value was determined using Equation 3.
Where: Co is the concentration of the solute in the initial solution (mg L -1 ); Ce, refers to the residual concentration in the equilibrium (mg L -1 ).

Adsorption kinetics
The study of adsorption kinetics was carried out with the objective of establishing the ideal contact time between the adsorbent and the adsorbate [27]. For this, the experiments were carried out in a batch system under conditions previously determined by   [25] and Nascimento et al. (2017) [26], with few modifications. The adsorbent (0.5 g) was subjected to an aqueous solution 25 mg L -1 of RBB dye (100 mL). The system was kept under agitation at 150 rpm on a shaking table (New Brunswick Scientific, model C25KC), with pH 3.5 and the aliquots were analyzed, in the time interval between 0 and 120 minutes.
The study was conducted at three different temperatures (30, 40, 60 and 100 ± 2 ºC) and 5.0 mL samples were taken at 2-minute intervals to be analyzed, and subsequently centrifuged at 5000 rpm for 3 minutes. The supernatant was analyzed by spectrophotometry to determine the final concentration. The determination of adsorptive capacity was determined by Equation 4.
Where: q is the adsorption capacity (mg·g -1 ), Co is the concentration of the solute in the initial solution (mg·L -1 ); Ce, refers to the residual concentration in the equilibrium (mg·L -1 ); V corresponds to the volume of the solution (L); m is the amount of mass of the adsorbent (g).
In order to know some characteristics about the adsorption kinetics of RBB in cellulose, the experimental data obtained in the kinetic study were evaluated by three kinetic models, namely: the pseudo-first order, pseudo-second order and intraparticle diffusion models (Morris-Weber) using Linear Equations 5, 6 and 7 [25,28,29].
Where qe is the adsorptive capacity at the saturation point in mg·g -1 , qt the adsorptive capacity varying in time in mg·g -1 , K1 and K2 are the speed constants of the pseudo-first and pseudo-second order models in min-1 e g mg-1 min-1, respectively, Kid is the intraparticle diffusion constant in mg·g -1 ·min -0.5 , D the constant related to the thickness of the diffusion layer in mg·g -1 and the contact time between species in minutes.

Balance study and determination of thermodynamic parameters
The adsorption isotherms represent a balance between the concentration of dye in the liquid phase and the amount of dye present in the adsorbent [28,29]. All equilibrium tests were performed in triplicate, the conditions used were: 0.5 g of adsorbent, 150 rpm for 12 h and pH 3.5, in different concentrations of the dye solution (25,35,45,55 and 65 mg L -1 ) and different temperatures (30, 40, 60 and 100 ± 2 ºC). The models used were Langmuir and Freundlich according to linearized Equations 8,9 and 10 [25,[28][29][30][31].
Where: qe is the adsorptive capacity at the saturation point in mg·g -1 , Ce is the residual concentration in equilibrium (mg·L -1 ), Co is the concentration at the beginning (mg·L -1 ), b (L·mg -1 ) and qm represent the Langmuir constants and the adsorption capacity and energy (mg·g -1 ), Kf the Freundlich constant. and RL is the separation factor (dimensionless). The value of RL indicates whether the adsorption is irreversible (RL = 0), favorable equilibrium (0 <RL<1).
The thermodynamic characteristics of the process were expressed at different temperatures and concentrations. Through linearized Equations 11, 12 and 13 it is possible to determine the variations of entropy, enthalpy, Gibbs free energy and activation energy of the adsorptive process [25,28,29,32].
Where: K is the thermodynamic equilibrium constant (L·g -1 ), Kc is the kinetic constant of the thermodynamic equilibrium ΔH°, ΔS° and ΔG° refer to enthalpy, entropy and Gibbs free energy of the system, respectively. R is the universal gas constant (8.314 J mol -1 ·K -1 ), T is the temperature (K), Ea corresponds to activation energy and A is the Arrhenius frequency factor.

Cytotoxicity assays in animals and microbials cells
The water obtained after treatment, to be discarded in the environment, cannot cause damage to microbiota or animals [25]. The residual water for cytotoxicity assays for both cells, animal and microbial, was used in the best experimental condition (dye concentration 25 mg·L -1 , 150 rpm at 30 ºC, 0.5 g of adsorbent and pH 3.5). This water was neutralized in order to decrease toxicity to cells and the environment when this effluent is discarded.
The cytotoxicity assay in J774 macrophage cells was performed according to Souza et al. (2014) [33] with few modifications. The cells were grown in 96 well plates at a concentration of 6x10 5 cells / well and incubated in a 5% CO2 oven at 37 °C for 24h (to adapt the cells to the wells surface). After this incubation period, the supernatant was removed and 100 µL of the neutralized residual water was added to the cells in RPMI medium and again the system was incubated in a 5% CO2 oven at 37 °C for 48 hours. After this step, the supernatant was removed and to the wells RPMI medium without phenol red plus 10μL of 3-(4,5-dimethyltriazol-2-yl) -2,5-diphenyltetrazolium (MTT) 0.5 mg mL -1 .
The plates were incubated for an additional 3 hours to allow MTT to react with the cells and form the insoluble crystals of Formazan. After 3 hours, Formazan crystals were dissolved in DMSO and immediately after the absorbance of the system, it was determined in an ELISA Benchmark Plus plate reader (Bio-Rad) at 490 nm.
In the supernatants of macrophage cultures, the nitric oxide content was determined by the colorimetric method of Griess [34]. The nitric oxide concentration was estimated using a standard curve (3.12-100 μmol·mL -1 ). The reading was performed on a spectrophotometer (Bio-Rad 3550, Hercules, CA) at 595 nm. This experiment was carried out in order to confirm whether the water used had a toxic effect.
For the microbial growth viability assays, the following strains Escherichia coli, Pseudomonas aeruginosa, Enterococcus faecalis and Staphylococcus aureus were used, maintained in Mueller-Hinton (MH) agar culture medium. These belong to the collection of microorganisms of the Industrial Microbiology Laboratory located in the Department of Chemical Engineering of the Federal University of Pernambuco. These cultures were inoculated in MH broth at 35 °C for 18 hours. Then, a 0.1 mL aliquot was transferred to another tube containing Mueller-Hinton broth and incubated for 3 hours at 35 °C. Thus, obtaining a suspension calibrated with the tube 0.5 of the MacFarland scale that corresponds to 10 8 UFC·mL -1 . Subsequently, with the aid of a swab, sowing was carried out in plates containing MH agar. Once this was done, 10 µL of the neutralized residual water was inoculated to the plates containing the culture medium, and then incubated at 35 °C for 24 hours, in order to check whether the residual water would cause cell death. All tests were performed under aseptic conditions.

Physical and chemical characterization
The bacterial cellulose presented dry weight 0.34 ± 0.01 g, wet mass 5.43 ± 0.01 g, ashes of 0.005 ± 0,0% and humidity of 92%. These results are close to those obtained by Mikkelsen et al.  [37] and Gayathri and Srinikethan (2019) [38] characterizing bacterial cellulose. Figure 3 shows the results of the FTIR, DRX, TGA/DTG and PCZ analyzes.
The Figure 3A shows the FTIR spectrum for bacterial cellulose. This was analyzed according to the data obtained by Cheng  In the X-ray diffractogram ( Figure 3B), the bacterial cellulose produced is a semicrystalline material with crystalline and amorphous regions. The two predominant peaks found in the analyzes make it possible to assess the presence of type I and II cellulose. In X-ray analysis, cellulose type I is characterized by peaks of 18 ° ≤ 2θ ≤ 19 ° in the amorphous region and 22 ° ≤ 2θ ≤ 23 ° in the crystalline region. Type II cellulose shows peaks of 13 ° ≤ 2θ ≤ 15 ° in the amorphous region and 18 ° ≤ 2θ ≤ 22 ° in the crystalline region. The presence of cellulose type I and II is characteristic of bacterial cellulose, while the vegetable cellulose presents only type I cellulose [42,43].
The analyzes by thermogravimetry (TGA) and Derivative Thermogravimetry (DTG) ( Figure  3C) show the variation in mass of a sample as a function of temperature. The cellulose obtained had three different stages of mass loss. During an initial phase in the ambient temperature ranges up to approximately 100 °C -there is a small loss of mass related to the evaporation of the residual water molecules (from the drying process) physically adsorbed by hydrogen bonds to the sample surface. A second phase, from 220 °C to 420 °C, where the samples suffer a marked loss of mass, which can be explained by combustion with consequent degradation of cellulose, this step extends to the region of 420 °C up to 700 °C, finally, the third and last phase shows the thermo-oxidative degradation of cellulose. The DTG curve shows the maximum cellulose degradation around 350 ºC. This degradation profile is similar to those obtained by Kiziltas  The Figure 3D shows the relationship between the initial pH and the end of the solutions. From it, it is possible to determine the zero-load point by making an arithmetic average of the points where the final pH remains constant. Through the analysis of Figure 3D, it is observed that the PCZ is 6.8. Thus, based on the graph, it can be predicted that below these values of the pH of the PCZ the adsorbent has a positive surface charge favoring the adsorption of negatively charged adsorbates and above the pH of the PCZ the surface is negatively charged, favoring the adsorption of loaded adsorbates positively.

Influence of ph on the adsorption capacity of bacterial cellulose
In the adsorption process, factors such as the properties of the adsorbent and adsorbate, temperature and pH of the solution can directly influence the efficiency of the process [47][48][49]. The pH determines the intensity of the electrostatic interactions between adsorbent and adsorbate [47,49].
The adsorption mechanisms involving materials of biological origin are not yet fully understood, as they involve several processes of capture of the adsorptive species. In general, it is accepted that the surface of an adsorbent has an important role in the mechanism of capture of chemical species [47,49]. These mechanisms, in general, involve surface groups that in an aqueous medium can act in the capture and retention of ionic or molecular species. Such groups can be modified in an aqueous medium by varying the pH, which affects the adsorption process [48]. Figure 4 shows the profile of the bacterial cellulose adsorption capacity as a function of pH. The highest values of dye removal are observed at low pH values, this is explained by the fact that the molecules of the RBB anionic dye are negatively charged in aqueous solution [47,48], while cellulose adsorption sites at a lower pH have increased the number of positive charges [30,31,47,48]. Cellulose has hydroxyl functional groups which, under acidic conditions, are protonated, enabling the interaction between the different adsorbent and adsorbent charges. Cellulose acquires a positive charge that, by electrostatic attraction, binds to negatively charged dyes [30,31,47]. In order to prove this behavior on the surface of cellulose as a function of pH, a preliminary study of the acid-base behavior of the surface of these adsorbents in an aqueous medium was carried out by determining the PCZ [23,24] as shown in Figure 3D.

Kinetic modeling of the adsorptive process
The kinetic study of adsorption of dyes on adsorbent materials is a prerequisite for choosing the best removal conditions [27]. This evaluates the adsorbent adsorption rate and controls the residence time of the adsorbate at the solution interface [50,51]. The Figure 5 shows the kinetic evolution of the Remazol Black B dye removal process in contact with the adsorbent. This experiment was carried out in batch tests at temperatures of 30, 40, 60 and 100 ± 2 ºC for a period of 120 minutes. And the adsorptive capacities in the respective times were calculated using Equation 4.  It can be seen through Figure 5, that the process reaches equilibrium around 80 minutes for all temperatures studied. This fact is probably related to the high number of adsorption sites available on the surface of the material [49]. Thus, as time increases, saturation of the available surface area occurs and thus the equilibrium situation is achieved and no significant change in the amount of adsorbed dye can then be perceived [29]. The rapid removal of adsorbate and the achievement of equilibrium in a short period is one of the indications of a good adsorbent material [50,51].
The kinetics experiments, in addition to finding the time in which the adsorption equilibrium occurs ( Figure 5), aimed to assess the stage that controls the adsorption process [50,51]. Thus, three linearized kinetic models were tested to fit the experimental data, namely, the pseudo-first order model ( Figure 6A), pseudo-second order model ( Figure 6B) and intraparticle diffusion ( Figure 6C).  Tables 1, 2 and 3 show the results of the constants for each model and the respective correlation coefficients, that is, how the theoretical model fits the experimental data ( Figure 6D). A good correlation of kinetic data proves the possible kinetic mechanism of adsorption in the solid phase [50,51].  Through the correlation coefficient, it was found that the pseudo-second order model obtained a good adjustment to the experimental data obtained for bacterial cellulose, since they approach the theoretical qe values ( Figure 5B). This model consists of the adsorption capacity of the solid phase and reports the behavior of the process over the entire contact time range [49,51].
The models of pseudo-first order (applied in times above the initial 20-30 minutes of the adsorption process) and intraparticle diffusion (mechanism as a limiting step) did not fit satisfactorily with the experimental data because they have a low correlation coefficient [52,53].
Other authors carried out the evaluation of the removal of the RBB dye and observed that the capture of this dye also adjusted to the pseudo-second model, as observed by Alves et al. (2020) [29] using soil samples, Cruz Filho et al. (2016) [48] using mixed biomass and Ziapour et al. (2016) [54] using bagasse.

Balance study and thermodynamic parameters
The adsorption isotherm is fundamental for understanding the adsorption mechanism [55]. An adsorption isotherm shows the amount of a particular solute adsorbed to an adsorbent surface, depending on the equilibrium concentration of the solute [56]. The adsorption isotherms indicate how the adsorbent will remove the contaminating molecules and whether the desired purification can be obtained, as well as presenting a maximum estimate of the adsorption capacity [57]. The most used models for balance study are those of Langmuir and Freundlich The Langmuir model assumes a monolayer and considers the adsorbent surface with identical sites in terms of energy, that is, it considers the adsorbent surface to be homogeneous [55,56]. The Freundlich model describes multilayer adsorption systems and is applied to the heterogeneous surface, where the adsorption sites are considered to have different adsorption energies, which vary according to the surface coverage [55][56][57]. The Figure 7 shows the linearization of data for Langmuir ( Figures 7A e 7C) and Freundlich ( Figure 7B e 7D) models. All experiments were conducted in batches until the equilibrium time obtained by the kinetic tests and the equilibrium tests of the bacterial cellulose adsorbent with the RBB adsorbate were carried out at different temperatures. The values of the linearization parameters of the Langmuir and Freundlich models are shown in Table 4. Where qm corresponds to the maximum adsorption capacity and KL refers to the Langmuir  The results of the adjustments showed that the Langmuir model presented better correlation coefficients when compared to the Freundlich model. According to the Langmuir model, maximum adsorption occurs in a monolayer of solute molecules on the surface of the adsorbent, without any lateral interaction between the adsorbed molecules [52,53,58]. That is, the adsorbent has a limited number of positions available on the surface and that the molecules can be adsorbed until the available surface sites are vacated [49,51]. Thus, it is considered that the molecules were adsorbed only on free sites [55,56]. The effect of temperature is a significant physical-chemical parameter in the adsorptive process [59,60]. In general, adsorption increases with increasing temperature [61]. This fact is related to the increase in the molecular solubility of the dye and also to its chemical potential [59]. The increase in temperature promotes an increase in the molecular diffusivity of the dye and, consequently, by an increase in the adsorption speed. In addition to the increased mobility of the dye molecules, there is an increase in the number of active adsorption sites, due to the swelling of the material's pores [59][60][61]. However, at high temperatures (100 ºC) there is a decrease in the adsorption capacity, and this is because the increase in the process temperature can damage its surface of natural adsorbents and indicates that at this temperature the process is not favorable [48,62].
This behavior is also confirmed through the data obtained in Figure 8, which show that at temperatures below 100 ºC the process is favorable, because the value of the constant RL is between 0 and 1 ( Figure 8A) [30]. Figure 8B shows that at temperatures above 100 ºC the adsorptive process is unfavorable. Through the results obtained in the kinetic and equilibrium studies it was possible to make a comparison with different natural adsorbents. Table 5 shows the type of adsorbent, the maximum adsorption capacity (Qo), equilibrium time and models that were adjusted, both in the kinetic study and in the equilibrium.
According to Table 5, the adsorbent capacity of the adsorbent from bacterial cellulose is lower when compared to activated carbon from fungus residue, chitosan, fiber composed of polyethyleneimine-polyvinyl chloride and activated carbon. And it has greater adsorption capacity when compared to mixed biomass of Aspergillus niger/elephant grass, green coconut mesocarp, macadamia seed husks, Eichhornia crassipes/chitosan composite. These differences in the adsorption efficiency are directly related to the nature of the adsorbent and the process conditions [31]. The determined thermodynamic parameters were Gibbs free energy (∆°), enthalpy variation (∆ º) and entropy (∆°), the same being of great importance for the adsorption process [25,70].The estimates of the thermodynamic parameters ∆S° and ∆H° were possible to determine by linear regression Ln (K) as a function of T -1 as shown in Figure 9 [32,70]. The temperature of 100 ºC was not used to obtain the thermodynamic parameters because it showed a decrease in the adsorption capacity. Table 6 shows the obtained thermodynamic parameters.  From the data in Table 6, it can be seen that Gibbs free energy has positive values, indicating that the reverse reaction is negative, it also demonstrates that the process is non-spontaneous and can be reversible, therefore, the cellulose can be recovered. for reuse [71,72].
The negative enthalpy value suggests an exothermic nature of adsorption and this is accompanied by an energy release [25,70]. The value of ∆° gives an indication of the type of adsorption that occurs, the two main types being physical adsorption (physisorption) and chemical adsorption (chemisorption). In physisorption, the adsorption heat has values less than 20.9 kJ·mol -1 and in chemisorption, the adsorption heat has energy values greater than 20.9 kJ mol -1 [73]. Therefore, the value of ∆° for the adsorption of the RBB dye indicates that the adsorption process is of a physical nature.
Negative entropy values indicate an increase in the degree of organization of the system, associated with the accommodation of the dye on the cellulose surface, also indicating that there is no increase in mobility and there is no dissociation of the dye particles on the cellulose surface [25,71,72].
The activation energy was extrapolated using the Arrhenius equation and Figure 10 shows the Arrhenius graph of Ln K versus 1/T where K was obtained from the pseudo second order model. The calculated Ea value was 23.8 kJ·mol -1 and the exponential factor of the Arrhenius constant 1. 65. This means that for the adsorbate interaction to occur with the cellulose surface, it is necessary to overcome an energy barrier of 23.8 kJ·mol -1 .

Cytotoxicity assays in animal and myocrobia cells
The residual water from the process with the best experimental condition was neutralized and only after that was the cytotoxicity assay performed against macrophage cells of the J774 strain ( Figure 11A). The cells did not suffer significant cell death (p = 0.0022). These values were confirmed due to the low nitric oxide production ( Figure 11B). Yang et al. (2020) [74] removing methylene blue noted that the residual water was not cytotoxic against Chinese hamster ovary cells. In addition to not being toxic to animal cells, it also did not cause death to the bacterial cells of this study Pseudomonas aeruginosa, Escherichia coli, Enterococcus faecalis and Staphylococcus aureus ( Figure 12). Therefore, this water can be reused and has no harmful effects.

CONCLUSION
In this work, bacterial cellulose was used as an adsorbent for the removal of the Remazol Black B dye in aqueous solutions at different pHs. The results showed that at pH 3.5 the best percentage of removal was 92% of the RBB dye. The kinetic study showed that the balance between the liquid and solid phases was reached in about 80 minutes and that the experimental data adjusted to the pseudo-second order model. The equilibrium tests showed that the experimental data fit the Langmuir model and that the increase in temperature causes a decrease in the adsorption capacity. The thermodynamic behavior of the adsorption was evaluated according to Gibbs' enthalpy, entropy and free energy. The value of Gibbs' free energy showed that the process is not favorable with increasing temperature. The negative enthalpy value confirms the exothermic nature of the adsorption process. And the negative entropy value suggests a reduction in the disorder of the adsorbate particles during the adsorption process. Thus, bacterial cellulose proved to be an economical and efficient alternative for the treatment of textile effluents containing the dye Remazol Black B.

ACKNOWLEDGEMENTS
This study was supported by Brazilian agencie Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)