Circular economy approach for rice husk modification: equilibrium, kinetic, thermodynamic aspects and mechanism of Congo red adsorption

Equilibrium, kinetic, thermodynamic aspects, and mechanism of congo red (CR) adsorption onto rice husk treated with effluent from cotton pre-treatment (ERH) are examined by altering the initial dye concentration (0.1, 0.2, 0.3, 0.4, and 0.5 g/L), contact time (0 to 1440 min), temperature (298, 323, and 343 K), and adsorbent dosage (10, 20 and 40 g/L). When 10 and 20 g/L adsorbent dosages are applied, the treated rice husk adsorbs the CR following the Langmuir model, while at 40 g/L ERH, the adsorption follows the Freundlich model. A maximum of 149 mg CR per gram ERH is adsorbed with 10 g/L ERH at 70 °C. This adsorption capacity is one of the better ones found in the literature. The calculated Dubinin–Radushkevich activation energy lower than 8 kJ/mol, indicates the physical nature of CR adsorption. The adsorption kinetic follows a pseudo-second-order kinetic model because R2 for all tested samples is higher than 0.999. The activation energy (Ea) varies from 0.045 to 40.1 kJ/mol, while the isosteric heat of adsorption (ΔHX) varies from 1.06 to 36.0 kJ/mol. Ea and ΔHX lower than 40 kJ/mol and 80 kJ/mol, respectively, confirm the physical adsorption of CR onto ERH. The other thermodynamic analysis indicates spontaneous and endothermic adsorption. These results showed the applicability of the circular economy concept in the effort to obtain an efficient adsorbent without wasting additional chemicals and energy that could be used to create a continuous column-mode process of rice husk modification and purification of colored effluent from the textile industry.


Introduction
Direct (substantive) dyes are water-soluble colorants for dyeing cellulosic fabrics that impart good light fastness and medium washing fastness of the dyed textile goods (Broadbent 2001). As sulfonated azo chemicals, they are serious pollutants (Blackburn 2004) because their degradation products are highly carcinogenic (Gičević et al. 2020). Effluent from dyeing of cellulosic fibers, with 5-30% residual direct dye (Blackburn 2004), has a negative environmental impact and a toxic influence on aquatic organisms because of its high biochemical oxygen demand that Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/ s10570-021- 04312-9. results in the suffocation of aquatic flora and fauna. Several studies for cleaning the colored effluent have been examined. Biological treatment using activated sludge (Abu Elella et al. 2019), chemical treatment with oxidation (Alaton and Teksoy 2007) and ozonation (Ciardelli et al. 2001), and physical methods by implementing membrane technology (Gorgieva et al. 2019;Males et al. 2020), ion-exchange materials (Shao et al. 2021), polyelectrolytes (Qiao et al. 2021), biomasses (Halysh et al. 2020), biosorbents (Mohebali et al. 2019), natural low-cost pumice (Shayesteh et al. 2016), zeolites (Nodehi et al. 2020), nanostructures (Aboamera et al. 2018;Batmaz et al. 2014;Nodehi et al. 2021), nanocomposites (Koohi et al. 2021), cellulose-based hydrogels (Ching et al. 2018;Mohamed et al. 2018), cellulose-based aerogels (Chong et al. 2015), and biopolymers (Perez-Ameneiro et al. 2014) have been explored. Many of these treatments are expensive and have limited effectiveness that requires additional extensive examination. Adsorption of dye molecules is one of the cheapest methods for purifying colored effluent from the textile industry based on employing cheap and highly effective adsorbents. Many adsorbents show effective dye adsorbency, but most of them are either expensive or have limited adsorptivity. One of the commercially available adsorbents, which possess excellent adsorption ability, is activated carbon (Foo and Hameed 2010). Despite its high efficiency, its expensive fabrication limits its use as a dyed effluent purifier (Abdelwahab et al. 2005).
The latest development in colored effluent cleaning is related to using an agricultural byproduct, or industrial waste, as an adsorbent, that is inexpensive and abundant. Grass waste, leaf, fly ash, sugarcane bagasse, sludge ash, wheat straw, natural fibers, rice husk, etc., have been tested as potential adsorbents (Chakraborty et al. 2011;Gupta et al. 2009;Sanghi and Verma 2013;Sharma et al. 2011;Tarbuk et al. 2020;Toshikj et al. 2019). Some are highly efficient, but many require additional modification steps to make them more amenable for dye sorption. The chemical composition of the rice husk comprising 32% cellulose, 21% hemicellulose, 21% lignin, 20% silica, and 3% crude proteins  is similar to the composition of natural cellulose fibers (Ivanovska et al. 2020;Lazic et al. 2018), emphasizing its use as a purifier for colored effluent (Chakraborty et al. 2011;Mladenovic et al. 2020;Shamsollahi and Partovinia 2019). Despite the theoretically good ability for dye adsorption, the content of 21% lignin in its outer protective layer i.e. the hydrophobic ligninsilicone-cellulose shield, makes the rice husk less adsorptive (Ndazi et al. 2007). In order to improve its adsorption capacity, the rice husk was physically grounded (Vadivelan and Kumar 2005) or incinerated (Alam et al. 2020), or chemically treated with acids (Abdelwahab et al. 2005), ethylenediamine (Ong et al. 2007), and bases (Chowdhury et al. , 2012. These modifications are effective but require an investment in expensive equipment that operates at high temperatures under pressure. Alkaline scouring of cotton is a process that uses dilute sodium hydroxide solution and surfactants to remove non-cellulosic hydrophobic components from cotton's cuticle and to make it hydrophilic (Toshikj et al. 2016(Toshikj et al. , 2017Lazic et al. 2018). This process creates an enormous amount of effluent with high chemical oxygen demand (COD) and pH higher than 12 (Jordanov et al. 2010), that is usually discharged in the environment. This process belongs to the traditional linear economy concept. On the other side, the circular economy is a new concept, mainly focused on using the products and materials over and over in a continuous loop. In this system, products and wastes are either reused or recycled. The applicability of the circular economy concept for rice husk modification has been proved in previous research (Mladenovic et al. 2020). Because of the similar chemical composition of the cotton fibers and the rice husk, the alkaline scouring or effluent from this process could be used for the rice husk modification. Rice husk was treated with cotton scouring effluent as an inexpensive way of successful modification by avoiding the need for additional energy and chemicals. Attenuated total reflectance-infrared spectroscopy (ATR-IR) and scanning electron microscopy-energy dispersive X-ray (SEM-EDX) was used to observe the occurred changes after this modification. The results of these analyzes showed that the treatment with effluent from alkaline scouring effectively removed the ligninsilicone shield from the rice husk surface. This treatment, which follows the circular economy concept, increased the percentage of cellulose in the husk, and the structure became cellulose-dominant and more amenable for purifying the colored effluent. Thus, treating rice husk as agricultural waste with effluent from the alkaline scouring of cotton fibers (other waste) is an elegant concept for producing material able to clean colored effluent from cotton dyeing. Moreover, this cheap and effective process of rice husk modification avoids the use of chemicals, energy for carbonization or grounding of the rice husk, and investment in a new machinery system.
In the present study, rice husk, modified by the concept of circular economy (with the effluent from the alkaline scouring of cotton), is used to explore equilibrium, kinetics, thermodynamics, and the adsorption mechanism of congo red (CR) by altering the initial dye concentration (0.1, 0.2, 0.3, 0.4, and 0.5 g/L), contact time (0 to 1440 min), temperature (298, 323, and 343 K), and adsorbent dosage (10, 20 and 40 g/L). The results show that this approach is effective in obtaining an efficient adsorbent for direct dye removal. Moreover, 10 g/L ERH modified rice husk at 70°C has an exceptional absorption capacity of 149 mg CR per gram ERH, which is higher than other capacities of chemically modified rice husks found in the literature.

Materials
The rice husk was collected from the local rice milling factory in Kochani, Republic of North Macedonia. The collected rice husk was sifted through a sieve to eliminate tiny particles with a size less than 1.25 mm and dried at 105°C for 2 h. CR dye-C.I. 20,120 (Sigma-Aldrich, Milwaukee, WI). All experimental dye solutions with defined concentrations were prepared by diluting the stock solution which has 1 g/L concentration. Sodium hydroxide and sodium chloride were supplied by Sigma-Aldrich (Milwaukee, WI). Cotoblanc HTD-N (CHT Switzerland AG), anionic washing surfactant and Kemonecer NI (Kemo Croatia), nonionic wetting surfactant. All chemicals are per analysis (p.a.) and were used as received.

Rice husk modification
Rice husk was treated with effluent from cotton scouring. The effluent was obtained after cotton yarn scouring, done with a solution containing 25 g/L of sodium hydroxide, 2 mL/L Cotoblanc HTD-N, and 1 mL/L Kemonecer NI in a bath with 50:1 liquor ratio at 100°C for 60 min in Ahiba Turbomat TM-6 apparatus. Then, the collected effluent was mixed with the raw rice husk (with a ratio of effluent and husk 20:1) and treated for 30 min in the Linitest apparatus. During this treatment, in a period of 30 min, the temperature of the effluent decreased from 100 to 40°C. Then, the rice husk was rinsed with effluents from cotton yarn rinsing, performed at 70°C for 15 min (warm rinsing) and 25°C temperature for 15 min (cold rinsing).

Adsorption studies
The dye adsorption ability of treated rice husk is examined by employing a batch adsorption process where 2, 4, and 8 g rice husk (i.e. 10, 20, and 40 g/L adsorbent dosage) are added in a 200 mL colored solution (in a glass stoppered Erlenmeyer flask) with an initial dye concentration of 0.1, 0.2, 0.3, 0.4, and 0.5 g/L CR, 10 g/L NaCl at pH 7, at 298, 323, and 343 K with constant agitation of 60 min -1 in a GLFshaking water bath 1083 (EURO lux GmbH & Co. KG Karlstadt-Karlberg, Germany). A sampling of the dye solution was done at 10,20,30,40,60,90,120,360,720, and 1440 min. The sampled solution was filtered (to remove some parts of the rice husk). Then, the concentration of CR (max. wavelength at 507 nm) in the supernatant solution was measured with a UV/VIS spectrophotometer (Model Hitachi-2800, United Kingdom). The calculated values are mean values of three measurements. The dye removal efficiency, expressed as a percentage, was determined using Eq. 1.
C 0 (g/L) is the initial concentration of the dye and C t (g/L) is the concentration of the dye after sorption at any time. The mass balance equation (Eq. 2) was used to calculate the amount of adsorbed dye onto the rice husk Q e (g/g).
C 0 (g/L) and Ce (g/L) are the initial concentration and the concentration of dye at equilibrium, respectively, V (L) is the volume of the solution and W (g) is the weight of the sorbent used.

Results and discussion
The influence of adsorbent dosage, temperature, and initial dye concentration on the adsorption ability of the sorbent has to be inter-correlated to obtain a clear picture of the sorption phenomena for each sorbatesorbent system. Hence, the adsorption ability of CR employing three masses of ERH, five initial dye concentrations, and three temperatures is examined in this study. The capability of the ERH to remove CR from solution with 0.3 g/L initial dye concentration at 298, 323, and 343 K, tested with 10, 20, and 40 g/L ERH are shown in Fig. 1a-c, respectively. The increase in the ERH dosage and temperature of the dye solution increases the percent of removed dye. Increased ERH adsorbent dosage increases the specific area and available adsorption sites, while higher temperatures make the dye molecules more energyactive with a higher potential for adsorption onto the ERH surface. Opposite, a higher initial dye concentration, at a constant adsorbent dosage and temperature of the solution, results in a lower percentage of removed dye (Fig. 2). A constant adsorbent dosage has a defined adsorption ability, so the percentage of removed dye decreases as the initial dye concentration increases. The complete data of the dye removal versus contact time, conducted under all combined parameters, are given as a supplementary file.

Adsorption isotherms
In order to examine, explain and interpret the used adsorbent-adsorbate sorption system, the adsorbent capacity must first be assessed. Langmuir and Freundlich's isotherms are the most used equilibrium relationships between the adsorbate and adsorbent at a given sorption condition. As such, they are essential for the evaluation of the physicochemical interaction between CR and ERH, and the dye adsorption equilibrium analysis. According to the Langmuir theory, dye adsorption is achieved through certain functional groups in the adsorbent. During the reaction, these groups are saturated forming a monolayer, which prevents further absorption of the dye. Also, in the Langmuir equation (Eq. 3), the dye concentration in the adsorbent increases with the dye concentration in the bath but to some extent, i.e. until all active centers of the adsorbent are saturated with dye. This isotherm has a linear form, as expressed in Eq. 4 (Nodehi et al. 2021).
Ce (g/L) is the equilibrium concentration of CR, Qe (g/g) is the amount of CR adsorbed per unit mass of adsorbent, Qm (g/g) is constant related to adsorption capacity, and K L (L/g) is constant related to the rate of adsorption. Qm is calculated from the slope of the straight line when Ce/Qe is plotted against Ce.
The Freundlich equation, presented in Eq. 5, explains sorption on a heterogeneous surface with sites of varied affinities .
The logarithmic form of Freundlich is given by Eq. 6.
Qe (g/g) is the equilibrium dye concentration, Ce (g/L) is the equilibrium dye concentration in solution, K F (g/g) (L/g) 1/n is the constant related to adsorption capacity, and n F is the adsorption isotherm constant. K F and 1/n F are calculated from the intercept and slope of the straight line of the plot log(Qe) versus log(Ce), respectively.
Langmuir and Freundlich isothermal curves for 10 g/L ERH at 298 K are shown in Fig. 3. The other curves (10 g/L ERH at 323 and 343 K, 20 and 40 g/L ERH at all temperatures) are given in Figs. 2, 3 and 4, in the Supplementary file.
The Dubinin-Radushkevich (D-R) isotherm, which rules out the homogenous surface or constant sorption potential, is an empirical model for the adsorption of the adsorbate to the volume filling of micropores (Dabrowski 2001). This isotherm is very important for determining the adsorption characteristics of most of the industrial adsorbents with a complex and welldeveloped porous structure, including pores of different shapes and sizes, but micropores play the most significant role. The non-linear equations of Dubinin-Radushkevich isotherm can be illustrated as Eqs. 7 and 8 (Chen 2015).
Q s (g/g) is a constant related to adsorption capacity, K DR (mol 2 /KJ 2 ) is a constant related to the mean free energy of adsorption, e is Polanyi potential, Ce (g/ L) is the concentration at equilibrium, R (J/mol K) is the gas constant, and T (K) is the absolute temperature.
The linear expression of the Dubinin-Radushkevich isotherm model is shown in Eq. 9.
Q m (g/g) is the maximum adsorption capacity and b (mmol 2 /J 2 ) is a coefficient related to the mean free energy of adsorption. Qm and b are calculated from the intercept and slope of the plot between ln Qe and e 2 , respectively.
b is used to calculate E (kJ/mol), the mean free energy of adsorption per mole of the adsorbate, when the adsorbate is transferred from the infinity of the solution to the surface of the solid. The relation between them is shown in Eq. 10 (El Haddad 2016).
E ranged from 8 to 16 kJ/mol indicates proceeds of the sorption process via chemisorption. Values of E \ 8 kJ/mol indicate a physical sorption process (El Haddad 2016).
The experimental adsorption capacity of ERH, the calculated constants of Langmuir, Freundlich, and D-R, and the coefficients of correlation between calculated and experimental data are presented in Table 1. The highest adsorption capacity of 149 mg/g is achieved when the lowest adsorbent dosage (10 g/L ERH) and the highest temperature of 343 K are applied. The dye removal is directly proportional, while the adsorption capacity of the ERH, at all temperatures, is inversely proportional to the adsorbent dosage. Higher dosage reduces the amount of the adsorbed dye per unit weight of the adsorbents. The experimental adsorption capacity of the ERH, tested with 10 and 20 g/L adsorbent dosage, increases by about 63% as the temperature of the adsorption testing rises from 298 to 343 K. The adsorption capacities tested with 40 g/L ERH are not that temperaturedependent and increase for about 5% when the temperature of adsorption increases from 298 to 343 K. The calculated coefficients of correlation presented in Table 1 show that the equilibrium data for 10 g/L ERH fits the Langmuir better than the Freundlich isotherm at all tested temperatures, confirming a monolayer coverage of the CR onto ERH when a lower adsorption dosage is used. The adsorption capacity for 20 g/L ERH at 298 and 323 K fit the Langmuir isotherm, while the one at 343 K fits the Freundlich isotherm model better. The Freundlich isotherm proved to be a better fit for the equilibrium data for 40 g/L ERH at all tested temperatures.
The Langmuir isotherm could be also used to determine the dimensionless equilibrium separation parameter R L calculated by Eq. 11 (Omidi and Kakanejadifard 2018).
K L (L/g) is the Langmuir constant related to adsorption capacity, while Co (g/L) is the initial concentration of the dye solution. When R L \ 1, the adsorption is favorable, while for values of R L [ 1, it is unfavorable.
The D-R isotherm is employed to calculate the magnitude of E which indicates the type of adsorption. As mentioned above, the value of E higher than 8 kJ/mol is an indicator of chemical ion exchange, while E below 8 kJ/mol indicates physical sorption (Chakraborty et al. 2011). The E values presented in Table 1 are lower than 8 kJ/mol in most cases, confirming the physical sorption of the CR onto ERH.
A list of published data for adsorption capacities of chemically modified rice husks (Table 2) shows that ERH, modified by the circular economy concept using effluent from the alkaline scouring of cotton, has the highest capacity. The adsorption capacity of some renewable adsorbents to adsorb CR (Table 3) indicates that ERH capacity is one of the better, found in the literature. Some of the capacities presented in Table 3 are higher than the adsorption capacity of ERH. Ballmilled clinoptilolite, modified with a cationic surfactant; powdered, acid-modified celery; and powdered, cetyltrimethylammonium bromide-acid modified celery have higher adsorption capacity than ERH. Even though these adsorbents have higher adsorption capacity, they are ball-milled or powdered (mechanical processes that are high-energy consumptive) and additionally modified with chemicals. On the other side, ERH is modified with the effluent from alkaline scouring without using energy and chemicals for modification, following the circular economy concept.

Adsorption kinetics
The amount of adsorbed CR versus the contact time serves to evaluate the adsorption kinetics. It is essential for the examination of the mechanism by which the adsorption occurs. Pseudo-first-order (Eq. 12), pseudo-second-order (Eq. 13), Elovich (Eq. 14), intraparticle diffusion (Eq. 15) as well as liquid film diffusion (Eq. 16) models are employed in this study to evaluate the kinetics of adsorption. Qt k 1 (min -1 ) and k 2 (g/gmin) are rate constants for pseudo-first and pseudo-second-order, respectively, Qe (g/g) is the amount of adsorbate at equilibrium, Qt (g/g) is the amount of adsorbate at time t, a (g/gmin) is the initial sorption rate constant, b (g/g) is desorption constant, k i (g/gmin 0.5 ) is intraparticle diffusion rate constant, F is fractional attainment of equilibrium, equal to Qt/Qe, and k fd (min -1 ) is liquid film diffusion rate constant.
The calculated kinetic parameters for linear kinetic models are presented in Table 4. The coefficients of correlation of the pseudo-second-order vary from 0.9998 to 1, while those of the pseudo-first-order range between 0.1513 and 0.9128. The substantially higher coefficients of correlation of the pseudo-second-order and the small differences between Q e, exp , and Q e, cal of this kinetic model, reinforce its applicability. The plots t/Q t versus t of 0.1 to 0.5 g/L initial CR concentration, for 10 g/L adsorbent dosage tested at 298 K show excellent linearity (Fig. 1 in Supplementary file).
Pseudo-second-order rate constant k 2 is calculated from the intercept of the slope t/Qt versus t. Since k 2 has the highest R 2 , it is then used for the calculation of the initial adsorption rate h according to Eq. (17). The k 2 values increase along with the adsorbent dosage and temperature of adsorption testing. Increased k 2 at a higher temperature (Table 4) indicates the endothermic process of the adsorption of CR onto ERH when the temperature of adsorption increases . The h value, which is the initial adsorption rate, is useful for creating a continuous adsorption system in which a higher initial adsorption rate is crucial for the proper selection of the adsorbents.
Higher h values at elevated temperature and higher ERH dosage favor building a continuous adsorption system filled with more ERH that will work at a higher temperature.
The Elovich model is another kinetic model used to examine adsorption/desorption kinetic on the solid adsorbent. This equation assumes the heterogeneous nature of the solid surface active sites that exhibit different activation energies for chemisorption ). The correlation coefficients R 2 determined from the Qt versus ln (t) plots are in the range between 0.458 and 0.906 (Table 4). In this model, the initial adsorption rate is presented as the constant a, while b is the surface coverage or desorption constant. These constants are temperature-dependent. An increase in temperature is accompanied by an increase in a and b, indicating that both the initial adsorption and the available adsorption area would increase. Moreover, the adsorbent dosage also influences the values of a and b. More rice husk, at the same initial dye concentration, increases both a and b constants.
During the process of dye adsorption onto some adsorbent, the diffusion of the dye from the bulk dye solution to the surface of the rice husk, and the transport from the surface into the solid rice husk may have occurred. Since these mechanisms follow liquid film diffusion and intraparticle diffusion models, these models have been explored to rate CR adsorption onto ERH. The liquid film diffusion parameters calculated from the plots of ln (1 -F) versus t have R 2 between 0.124 and 0.894 with intercepts ranging from -1.09 to -4.86, and k fd ranging from -0.0019 to -0.0071 min -1 (Table 4). On the other hand, the intraparticle diffusion parameters calculated from the plots Qt versus t 0.5 have R 2 between 0.124 and 0.534, intercepts between 0.0097 and 0.0837, and k i ranging from 0.0001 to 0.0073 g/gmin 0.5 . The plots of Qt versus t 0.5 for 10 g/L ERH at 298 K for 0.1 to 0.5 g/L CR are shown in Fig. 4. The other Qt versus t 0.5 graphs for 10 g/L ERH at 323 and 343 K, and 20 and 40 g/L ERH at all temperatures are given in the supplementary file (Fig. 5, 6, and 7). These plots have mainly two linear parts that influenced the adsorption process and imply that the intraparticle diffusion is not the sole rate-limiting step and that the adsorption is controlled  Table 4 Pseudo-first-order and pseudo-second-order kinetic, Elovich, liquid film diffusion and intraparticle diffusion kinetic parameters for adsorption of CR onto ERH   Table 4 continued by two processes (Wu et al. 2009). The first linear part determines the instantaneous adsorption on the surface, while the second one is the gradual adsorption stage that approaches an equilibrium step when the adsorption rate is the slowest. Both models (intraparticle diffusion and liquid film diffusion models) show that the intercepts of linear plots did not pass through the origin. Theoretically, if one of these models intends to be the sole ratedetermining step, the plots must have zero intercepts. Since the linear plots of both models do not meet this requirement, the applicability of these models in the present adsorption system is limited. In that case, during the interaction of CR with ERH, surface adsorption and intraparticle diffusion could both happen.

Activation parameters
The activation energy Ea is a crucial parameter for determining the origin of the sorption forces that bind the sorbate and sorbent. Since physical adsorption contains weak forces, Ea is lower than 40 kJ/mol. As chemical adsorption involves much stronger forces than physical adsorption, Ea is higher than 40 kJ/mol. The activation energy for adsorption of CR onto ERH is determined by the Arrhenius equation (Eq. 17) (Zhu et al 2009) using coefficient k of the pseudo-kinetic orders that fit better. Since k 2 , determined by the pseudo-second-order kinetic model, has the highest R 2 (Table 4), this coefficient is used for the calculation of the activation energy (Ea) according to k is the rate constant, A is the Arrhenius constant, Ea (kJ/mol) is the activation energy, R (8.314 J/mol K) is the gas constant and T (K) is the temperature. In this study, Ea was obtained from the slope of the linear plot of ln k 2 (calculated from the pseudo-second-order kinetic model) versus 1/T. The Eyring equation (Eq. 19) was also used to determine the standard enthalpy of activation (DH # ), the entropy of activation (DS # ), and free energy of activation (DG # ) in the sorption process Chowdhury et al. 2010): where k is the rate constant, T is the temperature (K), k B is the Boltzman constant (1.3807 9 10 -23 J/K), h is the Planck constant (6.6261 9 10 -34 J s) and R is the gas constant. By plotting ln (k 2 /T) vs 1/T, the values of DH # and DS # were determined from the slope and intercept of the plot, respectively. The free energy of activation (DG # ) was calculated by using these values in the following equation: The results of Eyring's thermodynamic parameters and energy of activation shown in Table 5 indicate  (Chowdhury et al. , 2012, which means that ERH can be reused. The obtained Ea values for the adsorption of CR onto ERH, for all initial concentrations (g/L) and adsorbent mass (g), suggest that the adsorption process is, in fact, physisorption.

Adsorption thermodynamics
Van Hoff's classical equation (Eq. 21) is used to calculate the change in Gibbs free energy which is crucial for determining the spontaneity of a process : where R (J/mol K) is the universal gas constant, T (K) is the absolute temperature and K C is the distribution coefficient for adsorption defined by the Eq. 22: in which Ca (g/L) is the equilibrium dye concentration on the adsorbent and Ce (g/L) is the equilibrium dye concentration in solution. The enthalpy (DH 0 ) and entropy (DS 0 ) were calculated using Eq. 23.
By plotting DG 0 versus T, the values of DH 0 and DS 0 were determined from the intercept and slope of the plot, respectively. The calculated values of the thermodynamic parameters for the adsorption of CR onto ERH at all temperatures are listed in Table 6.
The negative DG 0 values for most of the tested samples indicate a spontaneous adsorption process of CR onto ERH. The positive DG 0 for ERH tested at 10 g/L adsorption dosage and 0.3, 0.4, and 0.5 g/L initial CR concentration show non-spontaneous adsorption of these samples. The inverse correlation between DG 0 and the temperature implies more favorable adsorption at high temperatures. The positive DH 0 indicates an endothermic reaction of adsorption. The positive values of DS 0 show the modified rice husk's affinity for CR, an increase in the degree of freedom of the adsorbed dye, and increased randomness at the solid/solution interface (Gedam et al. 2019).

Isosteric heat of adsorption
Isosteric heat of adsorption (DH X , kJ/mol), presented as the heat of adsorption at a constant amount of adsorbed adsorbate, is of great importance for obtaining an optimized adsorption process. The Clausius-Clapeyron equation is used to estimate this thermodynamic quantity (Simsek and Beker 2014).
In the present study, the calculations for DH X are made at constant surface coverage of Qe = 0.04, 0.06, 0.08 and 0.1 g/g. The ln of equilibrium concentration (Ce) at a constant amount of adsorbed dye plotted versus 1/T is used to determine DH X from the slope of the plots. In the case of physical adsorption, DH X should be below 80 kJ/mol while for chemisorptions, the DH X values range between 80 and 400 kJ/mol . The calculated values of isosteric heat of adsorption are presented in Table 7. The values between 2.71 and 72.60 kJ/mol mean physical adsorption of CR onto ERH.

Mechanism of adsorption of CR onto ERH
The results presented above showed the physical adsorption of CR onto the ERH. Lower values of D-R free energy (E) than 8 kJ/mol, lower energy of activation (Ea) than 40 kJ/mol, and the isosteric heat of adsorption (DH X ) below 80 kJ/mol confirm the physical adsorption of CR onto ERH. CR is an Na-salt of diazo dye that contains two amino and two sulfonic groups. The amino groups are responsible for the creation of hydrogen bonds, while sulphonic groups make CR water-soluble. Theoretically, the adsorption of the direct dye onto cellulose occurs with physical hydrogen bonds between amino and hydroxyl groups (Broadbent, 2001). The physical hydrogen bonding between CR and ERH is confirmed by performing two tests. In the first (Mladenovic et al. 2020), the adsorption of CR on ERH is tested in the presence of sodium chloride (salt). In the second one, the adsorption of CR is tested in different pH of CR-solution. The earlier paper (Mladenovic et al. 2020) has shown a strong influence of sodium chloride on CR removal by ERH. The unmodified rice husk showed higher adsorption ability relative to ERH. By non-cellulosic components removal, the ERH became more hydrophilic with increased negative f-potential. The sulfonated direct dyes easily dissociate in the water into negatively charged ions that act as an electrostatic barrier between the dye and ERH, making the dye adsorption onto the ERH less feasible. To overcome this, the removal of the CR with ERH was improved by adding 10 g/L sodium chloride. Moreover, the test of CR removal relative to pH (Fig. 8 in Supplementary file) indicates no changes in CR removal in the first 120 min of contact time as pH changes. The abovementioned information confirmed that the adsorption of CR on ERH is influenced by the presence of salt and is not pH-dependent. All of these, combined with the calculated parameters for E, Ea, and DH X confirmed the physical bonding between CR and ERH under the proposed schematic of adsorption presented in Fig. 5.

Table 6
Thermodynamic parameters for adsorption of CR onto ERH Model Initial con.

Conclusion
Equilibrium, kinetic, thermodynamic aspects, and mechanism of CR adsorption onto rice husk treated with effluent from the cotton scouring (ERH) are explored. The initial dye concentration, contact time, the adsorption dosage of ERH, and temperature are crucial in determining the adsorption efficiency of modified rice husk. Adsorption of CR follows the Langmuir model, when the adsorption dosage of ERH is lower, and the Freundlich model when the dosage is high. The dye removal increases with an increase in adsorbent dosage of ERH, but the adsorption capacity of the ERH, performed at all temperatures, decreases with an increase in ERH dosage. The adsorption capacity of the ERH tested with 10 g/L ERH at 298 K and 323 K is 91.7 mg and 115 mg CR per g ERH, respectively. The maximum adsorption capacity of 149 mg CR per gram ERH was obtained with 10 g/L ERH at 70°C. It is one of the more remarkable results mentioned in the literature. Almost all of the tested samples have Dubinin-Radushkevich activation energy lower than 8 kJ/mol. Kinetic of adsorption, tested with all kinetic models, shows the highest values of coefficient of correlation for pseudo-secondorder kinetic model (R 2 [ 0.999). These results indicate the physical nature of CR adsorption onto ERH followed by the pseudo-second-order kinetic model. The Arrhenius's energy of activation, which varies from 0.045 to 40.1 kJ/mol, increases as the adsorbent dosage of ERH increases. Ea lower than Positive DH # symbolizes an endothermic reaction, while negative DS # indicates a lack of significant change in the adsorbent's internal structures during the adsorption, implying that the ERH can be reused. The isosteric heats of adsorption, estimated by Clausius-Clapeyron equation, with values below 80 kJ/mol, also confirm the physical bonding of the adsorbed CR onto ERH. This modification, performed with effluent from the alkaline scouring of the cotton, completely agrees with the circular economy concept. This is a straightforward process for obtaining a cheap adsorbent capable of removing CR from an aqueous solution and presents a solid base for creating a continuous column-mode process of rice husk modification and purifying the colored effluent from the textile industry.

Declarations
Conflict of interest The authors declare no competing financial interest.