Chromium adsorption studies using brewer’s spent grain biochar: kinetics, isotherm and thermodynamics

ABSTRACT: Brewer’s spent grain (BSG) is a residue from brewery production, that can be reused as adsorbent of heavy metals like Chromium (Cr). In this study, BSG was used as a biochar for Cr adsorption in batch adsorption experiments. The biochar pyrolysis temperature (500, 600 and 700ºC), pyrolysis time (30, 105 and 180 minutes) and zinc chloride concentration (5, 12.5 and 20%) were evaluated and optimized from a fractional factorial design. An equilibrium adsorption capacity of 78.13 mg.g-1 and a yield of 26.42% were achieved using 700ºC, 30 min, 12.5%. This biochar was applied to the adsorption of Cr in aqueous solution, under different stirring speeds (100, 150 and 200 rpm). The higher agitation speed reduced the adsorption capacity of the coal from 90 to 72 mg.L-1. In addition, the Cr adsorption equilibrium was reached before 100 min. The pseudo-first order model best described the Cr adsorption kinetics. The Redlich-Peterson isothermal model best fitted the experimental data, with the parameter g (close to 1) suggesting Langmuir’s assumptions as the most appropriate to describe the adsorption of Cr on the optimized activated carbon.


INTRODUCTION
The implementation of new brewing industries has increased significantly in recent years (CORDEIRO et al., 2012;SARAIVA et al., 2019).Among the types of waste generated by beer production, Brewer's spent grain (BSG) represents 85% of the total waste produced by the industry (MATHIAS et al., 2014).In general, after the beer fermentation process, approximately 20 kg of BSG is obtained for every 100 L of beer produced, the same being mainly composed of barley grain hulls, which contain cellulose, polysaccharides, lignin, some lipids and polyphenolic components, in addition to proteins, fibers, vitamins, minerals and amino acids (IVANOVA et al., 2017;MUSSATTO et al., 2006).
The large volume of BSG generated by the production process associated with its low cost and its inherent physicochemical characteristics make this residue an attractive biotechnological raw material to be used for the most varied purposes (SARAIVA et al., 2019).Currently within the scope of the research, many alternatives have been proposed for the use of BSG in the most varied segments, such as in the feeding of animals and ruminants, production of breads and hamburgers, production of biodegradable Ciência Rural, v.53, n.10, 2023.Pelissari et al. foams, among others (ALIYU & BALA, 2011;MELLO & MALI 2014;KTENIOUDAKI et al., 2012;SARAIVA et al., 2019).
In addition to these applications, the production of activated carbon (AC), currently known as biochar from BSG has emerged as a promising alternative for the use of this waste (VANDERHEYDEN et al., 2018;GONÇALVES et al., 2017;MACHADO et al., 2020).In a recent study, LOPES et al. (2021) optimized the parameters involved in the preparation of biochar from BSG through physical activation and showed that BSG is a potential precursor highly available and ecologically correct to be used with a biochar.In addition, biochar from BSG has shown high rates of adsorption of organic matter and organic compounds such as paracetamol (NADOLNY et al., 2020).
However, the process of preparing the biochar is essential for the adsorption performance (WONG et al., 2018).The preparation processes can be of chemical and physical order.In the chemical activation process, the activator, which may be acid, alkali and some salts, are mixed with the precursor material of the biochar and pyrolyzed at temperatures varying between 500 to 900 ºC.Meanwhile, during physical activation, oxidizing gases (CO 2 , air), steam or a mixture of them are used together with the precursor material that is heated at temperatures between 800 and 1000 ºC (MARSH, 2006).Both activations have advantages and disadvantages (WONG et al., 2018).Therefore, chemical activation exclusively for the case of BSG seems to offer advantages over physical activation.
Factors such as time and temperature of pyrolysis are essential factors for obtaining a biochar with a high adsorptive capacity (SCHETTINO et al., 2007).The effect of these variables on the surface area, the adsorption capacity and the pore volume have been studied (LOPES et al., 2021).However;although, currently BSG is commonly used as a precursor to biochar, when it comes to chemical activation, it is not known what are the optimized conditions that result in good adsorption performance for this material.
Significant Cr effluents are generated from electroplating industry, leather processing, oil refineries, textile industries, ceramic plants and other industrial processes, whereas CrIII and CrVI are the main species found in aqueous environment (KARRI et al., 2020;RODRIGUES et al., 2020).Cr III causes risks to human health, as well as negative environmental impacts.For human health, Cr is considered a toxic and carcinogenic metal, which can cause chemical pneumonia, perforation in the nasal septum, lung cancer and dermatitis (WANG et al., 2011).While for the environment, Cr III causes pollution of surface water bodies, being lethal for aquatic organisms and causing death for terrestrial biota (ADRIANO, 1986).At the same time, Cr III, an essential trace element in mammalian metabolism, exhibits lower toxicity and mobility (MOHAN, 2006).In addition, it is important to note that in Brazil, the current legislation regarding the standards for effluent disposal into surface water bodies presents restrictive values in relation to the concentrations of Cr (0.05 mg.L -1 ) (BRASIL, 2019).
Due to the fate and the need of Cr removal from aqueous environment, the development of alternative and cheap Cr removal techniques such as adsorption have been developed.BSG has attracted attention as a potential matrix for biochar production to be further used as adsorbent.In this sense, this research proposed to evaluate the conditions of biochar production from BSG and the efficiency of Cr adsorption through kinetics, isotherm, and thermodynamic studies.

BSG preparation
The wet BSG was supplied by a beer producer located in the city of Videira, in the state of Santa Catarina, located at 27º00'30" S and 51º09'06" W, Brazil.First, the BSG was washed and then dried in an oven for 24 h, at a temperature of 105 ºC, crushed into particles with diameters ranging between 250 and 425 µm, and stored in amber flasks until the moment of use.The moisture content, the volatile matter, ash and fixed carbon content were estimated via centesimal analysis, following recommendations of the American Society for Tests and Materials (ASTM -D1762-84).

Biochar production optimization
To produce the biochar, the BSG was initially impregnated with the activating agent (ZnCl 2 ) and submitted to constant agitation at 25ºC, for 24 hours.Then, after maintaining the pH of the samples close to neutrality, they were dried at 105 ºC for 6 h and stored, prior to the pyrolysis process.Pyrolysis was carried out in a muffle furnace.
A full factorial design was applied to evaluate the combination of the factor influencing the biochar preparation.The Box-Behnken matrix of 23 factorial design was performed for 15 runs.The pyrolysis temperature (500, 600 and 700ºC), pyrolysis time (30 105 and 180 minutes) and ZnCl 2 concentration (5, 12.5 and 20%) were the evaluated factors, and their levels and central values are detailed table 1.The combined effects for the tested parameters for the biochar preparation were evaluated according to the response parameters named: surface area (m².g -1 ), yield (%) and adsorption capacity (mg.g -1 ), with the same experimental conditions used for the preliminary adsorption experiment above described.
Statistica (StatSoft, USA) software was used for the experimental design, with a randomized experimental order to avoid systematic mistakes.A 2-way interaction between factors was used to calculate the model, at 95% of confidence level.Factors or their combinations resulting in p-values higher than 0.05 were discarded.Significant effects, those considered for the model and that most influenced the response, were evaluated by analysis of variance (ANOVA) at 95% confidence from Fisher test (F-value) and p-values.The parameter is significant if P < 0.05.

Biochar characterization
The biochar was characterized according to its surface area, pore volume, and apparent density using N 2 adsorption isotherms at 77 K using a Quantachrome ® analyzer (Nova 1200e).The specific area was calculated according to the Brunauer-Emmet-Teller (BET) method (BRUNAUER et al. 1938), and the volume of the micropores was calculated by the t-plot method and the distribution of the mesopores was determined by the BJH method (BARRETT, et al. 1951).In addition, the biochar was also characterized by scanning electron microscopy (SEM) (JEOL JSM 6360-LV).The samples were fixed on supports, using a copper tape and metallized with a thin layer of gold in the metallization device (SCD 050 from the Bal-Tec brand), using 40 milliamps for 2 minutes.

Biosorption studies Preliminary adsorption experiment
The preliminary adsorption experiments to select the best biochar preparation conditions were performed in Erlemeyer flasks filled with 300 mL of a 100 mg.L -1 Cr(NO3) 3 solution at pH 5.0 and 0.25g of biochar.The flasks were maintained at 25ºC and stirred at 150 rpm during 24 hours.The Cr concentration was measured using an atomic absorption spectrophotometer (AA 500 PG instruments).The solution pH and adsorbent dosage were parameters previously defined (data not published).

Adsorption kinetics
The adsorption kinetics experiments were performed as follows: 30 mg.L -1 Cr at pH 5.0; 0.25g of biochar and 25ºC in a reciprocal shaking.Samples were collected each 5 minutes until 180 minutes of testing, in order to ensure sufficient time for the system to reach equilibrium.Homogeneous samples were collected to maintain the biochar/solution ratio, followed by centrifugation to remove biochar and to  measure Cr concentration as immediately identified by means of the atomic absorption spectrophotometer.
(1) Where: qt: represents the amount of metal adsorbed at time t (mg.g -1 ); v: the volume of the solution (L); Co and Ct: absorbate concentrations (mg.L -1 ) in the initial time and time t, respectively; m: mass of the adsorbent (g).
Nonlinear kinetic models were tested to describe the adsorption of the metal (LAKSACI et al., 2017).The nonlinear models of pseudo-first order (PFO), pseudo-second order (PSO) and Elovich used are expressed in table 2.

Isotherm equilibrium and themodynamics
The adsorption equilibrium isotherms were performed at different temperatures (25, 35 and 45 ºC) with different Cr concentrations (5, 10, 15, 20, 30 mg.L -1 ), using 0.25 g of biochar which resulted in the highest adsorption capacity of 78.13 mg g -1 (700ºC pyrolysis temperature, 30 min pyrolysis time, 12.5% ZnCl concentration).The solutions were stirred at 120 rpm for 180 min.After this, the samples were centrifuged and the Cr concentration was assessed by reading on an atomic absorption spectrophotometer (AA 500 PG instruments).The adsorption capacity at equilibrium time was determined using Equation 2(DALL 'AGNOL et al., 2022).
(2) Where, q e : represents the amount of metal adsorbed at equilibrium (mg.g -1 ); Ce: final concentration of the adsorbate in solution at equilibrium (mg.L -1 ); Co: the adsorbate concentration (mg.L -1 ) in the initial time, V: the volume of the solution (L); m: mass of the adsorbent (g).
The models that showed the highest values for determining the correlation coefficient (R 2 ) and the lowest values for the normalized standard deviation (Δq) were considered the best to describe the kinetic and isothermal adsorption curves (Equation 5).
(5) Where: N is the total number of experiments, qcal and qexp are the calculated and experimental adsorption capacities in equilibrium (qe) (isotherms) or at each time t (qt) (kinetic curves), respectively.
The nonlinear regressions and plots for the kinetic and isotherm models were performed using the software Origin ® 2017 (ORIGINLAB CORPORATION, USA).

BSG and Biochar characterization
The percentage of humidity, volatile matter, ash and fixed carbon of BSG were 9.0%, 82.8%, 2.3% and 12.4%, respectively.Similar values were reported by Lopes et al. (2021) using a BSG obtained from a local brewery located in south Brazil, for humidity (7.03%), volatile matter (79.42%), ash (1.68%) and fixed carbon (11.87%).This behavior showed that BSG does not show great variation in relation to the local production characteristics of each brewery.
The biochar which resulted in the highest adsorption capacity, as presented in the next section, have an average pore diameter and pore volume of 1.44 nm and 0.27 cm³.g -1 , and a surface area of 412 m 2 g -1 .Sing et al (1985) indicated a classification according to pore size.Pores with widths exceeding about 50 nm (0.05m) are called macropores.Pores with widths not exceeding about 2 nm are called micropores and pores of intermediate size are called mesopores.In this way, results showed the predominance of micropores.
However, in the SEM images presented figure 1, it is possible to observe micropores ranging from 10 to 150 µm.This heterogeneous pore sizes are also observed for activated carbon when adsorbing 2-phenylethanol (CARPINÉ et al., 2013).It was possible to identify in the BSG an irregular and rough structure, characteristic of a structure arrangement of vegetable raw material (YOKOYAMA et al., 2019).After the process of preparing and obtaining the biochar, the surface of the charcoal showed significant changes with greater heterogeneity of the particles, as well as channels which can be associated with internal pores.

Optimization of Biochar from BSG
In general, the highest adsorption capacity was obtained with higher pyrolysis temperature (700 ºC), lower pyrolysis time (30 min) and middle ZnCl 2 concentration (12.5 %) (Table 1).Higher pyrolysis temperatures may cause the decomposition of organic compounds in the precursor material, resulting in greater pore formation and, consequently, a greater surface area of the biochar (MOHANTY et al., 2005;LENG et al., 2021).Previous studies conducted in biochar originating from tomato processing residues, also activated with ZnCl 2 , showed that the increase in the temperature of pyrolysis (from 400 to 600 o C) generated an increase in its surface area from 648 to 1093 m 2 .g - (SAYGILI & GUZEL, 2016).
Moreover, the highest adsorption capacity was identified together with the highest values of surface area (412 m².g -1 ).According to Weber & QUICKER (2018), larger and accessible surface area are key features for biochar reactivity.MACHADO et al. (2020) showed that a larger surface area of the biochar from BSG caused a greater adsorption performance.The authors showed specific surface area values close to 545 m² g -1 and 161 m² g -1 for ZnCl 2 -activated biochar and CO 2 -activated biochar, respectively.These values were extremely higher in relation to the biochar without activation (6.5 m² g -1 ), and confirmed that the activation efficiently performed.
The Pareto diagram showed that almost all the tested parameters were statistically significant, even with linear or quadratic interactions, at a confidence level of 95%, for the adsorption capacity response (Figure 2).The pyrolysis time and the ZnCl 2 mostly influenced the surface area of the biochar.Moreover, the isolated pyrolysis temperature has no significant influence over the biochar surface area but, its combined effect with pyrolysis time and ZnCl 2 was significant.At the temperatures around Ciência Rural, v.53, n.10, 2023.
7 500 o C amorphous carbons is transformed into crystalline carbons via condensation, more volatiles are removed, creating sparse regions, which leads to cracks in the material, and thereby generating mesopores (KEILUWEIT et al., 2010).Micropores exert the greatest effect on the surface area, whereas the value of total pore volume is positively correlated to the pore size (LENG et al., 2021).Moreover, activating agent concentration is directly linked to the dilation of the pores, i.e., the higher the concentration of ZnCl 2 the greater the porous area.However, pore widening will result in a larger total pore volume but a lower surface area (LENG et al., 2021;MIAO et al., 2013).In this way, the surface response showed that higher pyrolysis time combined with lower ZnCl 2 concentration led to higher biochar surface area.

Adsorption kinetics
The experimental data were evaluated using the non-linear adjustments of PFO, PSO, and the Elovich models (Table 3).According to the figure 3, the equilibrium time was reached after approximately 100 minutes for those experiments with agitation rates of 150 and 200 rpm.In the case of the stirring rate of 100 rpm, equilibrium was reached around 130 minutes of testing.The agitation rate clearly influenced the adsorption kinetics since different adsorption capacities were achieved for each agitation rates.
For the experimental data, the Elovich model fitted best with higher correlation coefficients (R 2 > 0.98) and the lowest ∆q (< 4.85), for all the tested agitation speed.The Elovich model suggested that the adsorption is regarded to chemical bonds between the Cr and the adsorbent.Also, the increasing in agitation rate increased the initial adsorption rate (α) from 10.82 ± 0.52 to 349.7 ± 142.8 g.mg -1 .min - , showing how this parameter influenced the adsorption process.
The PSO model for the 100 rpm experiment presented the closest adjusted qe values (93.3 ± 1.8 mg.g -1 ) to the experimental qe (87.6 mg.g -1 ), with lower qe standard deviation and higher R 2 when compared to the PFO model; although, this model also represented well the experimental data.The same result was found by Ramos et al. (2016), using carboxylate-functionalized sugarcane bagasse to adsorb copper (II), with k 2 value 0.154 g.mmol.min - .
Other studies carried out with different materials (soil, red mud) used for adsorption of Cr were better adjusted with the PFO model (GUPTA et al., 2001;RENGARAJ, et al., 2002).The k 1 PFO rate constant refers to how fast the Cr adsorption occurred, and their value ranged from 0.01 to 0.03 min -1 .

Equilibrium isotherms and thermodynamics
The adsorption isotherms parameters together with the thermodynamic analysis allow the understanding of the surface properties, interaction mechanisms and the affinity degree of the biochar and the Cr.The equilibrium isotherms are also used to calculate the maximum adsorption capacity in the tested conditions (AL-GHOUTI & DA'ANA, 2020; BENI & ESMAEILI 2020).Equilibrium is achieved when the adsorbent and adsorbate are in contact for long enough to establish a dynamic balance between the concentration of the solution  The adjustment of the Langmuir model to the experimental data obtained under the different temperatures showed an R 2 ranging from 0.92 to 0.98 and a Δq with values ranging from 3.70 to 4.39.This model reflects the monolayer adsorption on a surface with energetically equivalent sites, where there is no lateral interaction or steric impediment between the adsorbed molecules (GUPTA et al., 2001;GUPTA & BHATTACHARYYA, 2011).
According to the table 4, the maximum Cr adsorption capacity (Qm) increased considerably (from 31.8 to 64.6 mg.g -1 ) with the increase in temperature, indicating that the increase in temperature favored the adsorption process.The maximum adsorption capacity is influenced by several factors such as, for example, the characteristics of the activated charcoal (AC), the type of adsorbate and the conditions of the solution (FOO & HAMEED, 2010).Previous studies showed that the Qm can vary between 22.31 to 199.7 mg.g -1 (Table 5).Based on the results obtained, the biochar used in this study has good adsorption capacity.
The adjustments of the Freundlich model provided a value of R 2 (0.79 to 0.88) and ∆q (5.65 to 9.15%) lower than the comparison with the Langmuir model (Table 6).The empiric Freundlich model does not represent the maximum adsorption capacity and does not present the equilibrium plateau, not fitting well the final experimental points of the equilibrium curve (SOUZA et al., 2021).The Temkin model showed R 2 values from 0.87 to 0.99 and ∆q from 2.12 to 6.78 (Table 6).This model considers that the heat of adsorption of all molecules in the layer decreases linearly with the coverage of the adsorbent surface (Foo & Hameed, 2010).The Redlich-Peterson model showed the best fit for the experimental data, with the highest R² (0.97 to 0.99) and the lowest Δq value (1.66 to 3.48) (Table 6).This model presents similar premises presented by the Langmuir and Freundlich models.The exponent g can assume interval values between 0 to 1, where at the limit of g is 0, the most appropriate model is the Freundlich model.On the other hand, if the value of g is close to 1, the most suitable model is Langmuir's (BRDAR et al., 2012).From the values of g (0.87 to 0.89) identified in this study, the evidence fits from the Redlich-Peterson model that the adsorption of Cr is best described by Langmuir's assumptions.
The thermodynamic parameters were calculated considering the Langmuir model, as presented in table 7 and figure 5.It can be seen in table 7 that change in the Gibbs free energy (∆G) are negative for all the tested temperatures, indicating that the adsorption process is spontaneous and favorable.According to the Langmuir model, higher temperature increased the adsorption capacity.The positive enthalpy is considered for endothermic reactions.The slightly positive value obtained in this study (3.07 kJ.mol -1 ) suggested a physisorption reaction (SOUZA et al., 2021).The change in the entropy (∆S) indicated the randomness of the reaction at the solid/liquid interface.The positive entropy reveals the system disorder and randomness.

Environmental application of BSG biochar for Cr removal
The state of Santa Catarina has the highest number of breweries per inhabitants in Brazil, according to the ministry of agriculture (BRASIL, 2019).Most of micro and craft Ciência Rural, v.53, n.10, 2023.breweries are in south and southeast of Brazil (80%), whereas Santa Catarina state accounts for 13%.In this way, the region is a potential BSG producer, which requires correct management and disposal, frequently associated to costs for the breweries.Alternative use of this residue may result in environmental benefits and reduction for BSG disposal.

°C 35 °C 45 °C
Another study presented Cr concentrations up to 16 mg/kg of dry sediment in the same study area (BONATTI et al., 2004).
Cr contamination due to anthropogenic activities like industrial processes may be controlled using end-of-pipe solutions including wastewater treatment technologies, since the pollution source can be identified and controlled.Adsorption has been regarded as a promising technology due to its cheap and efficient nature, when using agro-industrial residues such as BSG.This enables the applicability of this residue, instead of its disposal.In this way, many breweries can value their by-product of their production process, thus reducing costs with final disposal and management.

CONCLUSION
According to this study, higher pyrolysis temperature (700 ºC), lower pyrolysis time (30 min) and higher ZnCl 2 concentration (12.5 %) resulted in the biochar with better adsorption capacity (78.13± 0.87 mg.g -1 ) and higher surface area (412 m².g -1 ).Also, 130 minutes were required to reach the equilibrium in the kinetic experiments, which revealed that PSO model best described the Cr adsorption kinetics.Isothermal studies showed that the Langmuir model best fitted to the experimental data, whereas the monolayer adsorption drives the process mechanism.The temperature exerts some effect over the adsorption process.According to this study, BSG biochar has the potential to be applied for adsorption of heavy metals, contributing to the reuse of this residue and the removal of a toxic pollutant from liquid streams.

Figure 1 -
Figure1-SEM images at 450x (A) and 1.500x (B) of the biochar under the conditions defined as ideal for the adsorption process.

Figure 2 -
Figure 2 -A) Pareto diagram for the factors affecting the adsorption capacity with statistically significance; B, C, D) Response surface graphs for the combined factors affecting the biochar surface area.
Pelissari et al.    volume and the interface concentration(GUPTA  & BHATTACHARYYA, 2011).The figure4shows the adsorption isotherms of Cr obtained at different temperatures (25, 35 and 45 °C), adjusted according to the non-linear models of Langmuir, Freundlich, Redlich-Peterson and Tempkin.

Table 1 -
Experimental design with real and coded levels and the experimental responses values.

Table 2 -
Kinetic and isothermal models and non-linear equations.

Table 3 -
Parameters calculated from the nonlinear fits of kinetic models to the experimental data.

Table 4 -
Parameters calculated from the nonlinear fits of theoretical isotherm models to the experimental data of Cr adsorption onto activated carbon.

Table 7 -
Thermodynamic parameters of Cr adsorption onto activated carbon.

Table 6 -
Analysis of variance (ANOVA) of quadratic regression model for surface area as experimental response.SQ is the sum of squares; DF is the degrees of freedom; MS is the mean for squares.