Modeling the removal of Reactive Red 120 dye from aqueous effluents by activated carbon

Environmental pollution has recently become a severe problem worldwide. The synthesis of industrial products such as cosmetics, food and textiles has participated in the formation of wastewater contaminated with dyes. In fact, dyeing a textile loses about 10–60% of reactive dyes, and produces large quantities of colored wastewater (Hessel et al. 2007; Cardoso et al. 2011).

Dyes are extensively used for industrial, chemical, printing, cosmetic use, and clinical purposes as well as textile dyeing because of their chemical stability; however, they cause serious environmental and health problems and are not biodegradable (Srinivasan & Viraraghavan 2010; Eren 2012). It has been demonstrated that most of these dyes can cause several allergies, skin irritation (Brookstein 2009) cancer proliferation (Carneiro et al. 2010), and mutation in humans (de Lima et al. 2007; Carneiro et al. 2010). Their presence at even low concentrations can be an obstacle to the use and reuse of water.

Reactive dyes are typically azo-based chromophores combined with different types of reactive groups (Dias et al. 2003). These complex molecular structures make them very difficult to be removed from wastewaters (Allege et al. 2016). Among reactive dyes in the textile industry, Reactive Red 120 (RR-120) is one of the frequently used ones. Recently, great scientific attention has been paid to remove dyes from wastewaters not only due to their toxicity but also because of their visibility. Adsorption, coagulation, advanced oxidation, and photodegradation (Naushad et al. 2019a) are some commonly used methods. With low capital investment cost, the adsorption process is one of the most commonly employed techniques for the extraction of synthetic dyes from aqueous effluents (Ozcan & Ozcan 2005; Machado et al. 2011; Mosallanejad & Arami 2012; Somasekhara et al. 2012; Vargas et al. 2012).

Several low-cost materials are used for the removal of different dyes, such as arginine-modified activated carbon (Naushad et al. 2019b), nano-sized chitosan blended polyvinyl alcohol (Anitha et al. 2016), eggshell (Rápó et al. 2020) and composite sorbent coated with siderite nanoparticles (Alshammari et al. 2020). However, activated carbon is found in large quantities and it is easy to use, in addition to being low cost.

The carbon-based materials are the most used adsorbents to remove water pollutants like dyes. The adsorption capacity of this type of adsorbents could change by many orders of magnitude, depending on its preparation conditions (Tan et al. 2008; Umpierres et al. 2018; Lima et al. 2019). In general terms, activated carbons must be thought of as being the most effective adsorbents. Their performance in removing from water contaminants such as metals, rare earth elements, phenolic and aromatic derivatives (including dyes and pesticides), pharmaceuticals, and drugs has been reported (Haghseresht et al. 2002; Dąbrowski et al. 2005).

Calvete et al. (2009) have achieved the RR-120 adsorption on the Brazilian pine-fruit shell activated carbon. The equilibrium adsorbed quantities as a function of RR-120 concentration were determined and modeled empirically at 298 K. The experiment was achieved in order to evaluate the adsorption capacity of activated carbon as a function of the environment change (at 303, 308, 313, 318 and 323 K).

The objective of this work is to investigate the physical nature of the adsorption of RR-120 on particles using models based on the formalism of statistical physics. The five models were performed and proposed to fit the RR-120 isotherms curves. The fitting process was achieved in the ideal gas law approximation and the approach of the law of real gas interactions. The possible adsorption mechanisms associated with this dye (RR-120) and tested adsorbents were theoretically analyzed via the interpretation of the parameters of the statistical physics model. The novelty of this investigation is to provide detailed information about the stereographic mechanism of the adsorption process through the fitted parameter values of the used models. The energetic characteristics of the stages of the process are given at a microscopic level as well as global thermodynamic potentials.

The Brazilian pine-fruit (piñon) shell was dried and milled, as previously reported (Calvete et al. 2009). The carbonization of the Brazilian pine-fruit shells was achieved by adding 20.0 g of milled shell and 50.0 mL of concentrated sulfuric acid (98% weight, 1.98 g·mL⁻¹) in a 1 L glass beaker to produce a black carbonaceous residue. This solid material was magnetically stirred for 15 minutes and combined with 350 mL of water. The system was then heated to 373 K and kept at this temperature for 2 hours under magnetic agitation. Afterward, the non-activated carbonized Brazilian pine-fruit shell was filtered, washed with water and then with 1.0 mol L⁻¹ of K₂CO₃ solution, followed by extensive washing with water until the washing liquids reached pH 6.0. The carbonized pine waste was then dried at 423 K for 2 hours and kept in a desiccator. The yield of this carbonization step was 70%.

A chemical and physical activation method was followed. An amount of 10.0 g of previously non-activated carbonized material was placed in a quartz reactor provided with a gas inlet and outlet, which was then placed in a vertical cylindrical furnace to activate the carbon material. The system was heated up to 1,123 K, at a heating rate of 7 K·min⁻¹ under N₂ atmosphere (flow rate: 100 mL·min⁻¹). Secondly, the temperature was kept isothermal for 1.5 h, and the gas flow was switched to CO₂ (flow rate: 150 mL·min⁻¹). Afterward, the system was cooled down to room temperature for 2 h, and the gas was again switched to N₂. The chemically and physically activated carbon obtained was further ground in a disk mill, followed by sieving, obtaining an adsorbent with a particle size lower than 53 μm. The chemically and physically activated carbon obtained was assigned as CPAC. The yield of this activation process was 45%. Considering the earlier carbonization step carried out with sulfuric acid, the total yield for CPAC adsorbent was 32%.

Considering that this paper focuses on the statistical physics models, the characterization of the activated carbon is presented in the Supplementary Material.

The RR-120 dye belongs to the large family of azo dyes, which are characterized by the presence of the azo group (-N=N-). It also belongs to the large family of reactive dyes which owe their name to the presence of the reactive chemical function of triazine or vinylsulfone type. Figure 1 shows the chemical structure of RR-120 (Calvete et al. 2009). The solubility in water is ensured by the multiple sulfonate groups which account for the anionic molecules in aqueous media. The deionized water was used to prepare the solutions of RR-120, which was purchased from Sigma Aldrich as a commercially textile dye with 80% dye content and was used without further purification. RR-120 has six sulfonate groups in highly acidic solution, so the main characteristic structural features of a typical reactive dye molecule are (Calvete et al. 2009):

• the reactive system, enabling the dye to form covalent bonds between the dye and the cotton fiber;

• the chromophoric group;

• a bridging group that links the reactive system to the chromophore;

• solubilising groups that make the dye soluble in water (Calvete et al. 2009).

Some adsorption experiments have been carried out to evaluate the removal of RR-120 by using the activated carbon as adsorbent (Calvete et al. 2009):

• 50.0 mg of adsorbent was put in 50 mL cylindrical polypropylene flasks containing 20.0 mL of the dye solutions (50.00–1,200.0 mg·L⁻¹).

• The mixture was agitated for 3 h using an acclimatized shaker at temperatures ranging from 298 to 323 K to attain the equilibrium.

• After the batch contact, the slurries were centrifuged at 3,600 rpm for 5 minutes to separate the solid adsorbent from the remaining dye solution.

• After aliquots of dye solution were properly diluted from 1 to 10 mL into volumetric flasks of 50–100 mL, concentrations of this dye solution were determined by visible spectrophotometry at 534 nm, using a Fenton 600 S spectrophotometer.

This is the final analytical expression of this model that will be used to describe and interpret the adsorption process in terms of the parameters n, N_m, and C_1/2. The corresponding partition function and expression from other models proposed to fit the adsorption isotherm are present in Table 1.

When we focus on the previous work (Calvete et al. 2009) we notice that the model used for the adsorption isotherm does not have, in its expression, parameters which give information about the adsorption phenomenon on the microscopic scale, namely the distribution of molecules on the surface as well as its adsorption energies. In our work, Equation (8) clearly shows the importance of the statistical physics model since it includes steric and energetic parameters making it possible to clarify the adsorption process at the microscopic level. So, the importance of this paper is to give a physical sense to the adsorption process from the statistical model that the empirical models do not allow.

Now for modeling of our experimental adsorption isotherms, we have to use five statistical physics model expressions. The following five adsorption models were considered: monolayer adsorption at one energy, monolayer adsorption with two horizontal energies, double layer adsorption with one energy, double layer adsorption with two vertical energies and multilayer adsorption with saturation. The best fitting model will be the one that shows a better correlation with the experimental adsorption isotherms. This is expressed by the best values of the R² adjustment coefficient (Ayachi et al. 2018).

In Table 2, we present the values of the adjustment coefficients for each model given by a computer-assisted modeling analysis. These values of the adjustment coefficients show that the double layer model with one energy is the best fitting model.

We can clearly understand from Equation (2), which is the adsorption reaction, that n is the stoichiometric coefficient of this equation. Nevertheless, this number of adsorbed molecules per site, n, can also be considered as a steric parameter. The role of the previously mentioned number is to give information about the geometrical properties of the adsorbed molecules on the adsorption surface: that is to say, on the topography of the adsorbed molecule (Khalfaoui et al. 2012).

The linking of the molecule at the receptor site can be achieved through two ways, but that depends on its geometry as well as the angle of incidence on the adsorption surface: perpendicular and parallel anchoring. Parallel anchoring occurs if the number n of adsorbed molecules per site is greater than or equal to unity. In this case, the site is occupied by a portion of the molecules. In this respect, Figure 3 shows the number of adsorbed molecules of the RR-120 dye per receptor site of the adsorbent and for each temperature.

It is worth noting that whatever the value of the temperature, n is greater than 1. This number varies between 1.06 and 1.67. It is essential to mention that n is a decimal number. This can be justified by the fact that these values are averages between plenty of anchorages on the sites such as occupied by one dye molecule and the sites occupied by two dye molecules.

We consider the presence of only anchorages by one or two molecules, to be more understandable.

If we symbolize by x, the percentage of sites occupied by one dye molecule of RR-120 and by (1 − x) those occupied by two dye molecules of RR-120, for n = 1.29, for instance, we can write 1.29 = x.1 + (1 − x).2, which gives a rate of 71% of sites occupied by one molecule of RR-120 dye and a rate of 29% of the occupied sites by two molecules of RR-120 dye.

The values of n for each temperature are presented in Table 3. Along with that, Figure 3 illustrates that a rise in temperature leads to a rise in n. Equation (2) better explains this. It can, indeed, be split into two stages, the aggregation process: nD → D_n where n dye molecules are gathered; then, this aggregate is anchored on the receptor site: nD + S → D_nS. Thus the increase of n with temperature is due to the aggregation process, which is endothermic.

We have found out that the density of the receptor sites gets reduced with temperature: this is a plain effect that concerns thermal agitation causing, in turn, an inhibition of dye molecules anchoring on the receptor sites. Figure 4 represents the reduction of N_m with temperature, while Table 4 provides us with N_m values for every temperature.

A clear presentation of the value of the half-saturation concentration for each temperature is embodied in Table 5. This shows that an increase in temperature has led to an increase in the half-saturation concentration. Figure 5 better illustrates these outcomes. However, the next paragraph states that the growth of C_1/2 has paved the way for a variation in the adsorption energy.

Looking at Figure 6, we can notice clearly the variation of the adsorption energy with the temperature. By realizing that the values of the adsorption energy are positive in modulus, this can signal the exothermicity of the adsorption reaction, which will be verified by the study of the internal energy in the next section. One other thing worth noting is that the adsorption energy values are unable to exceed 40 kJ·mol⁻¹. That leads us to deduce that the adsorption is physisorption. Thus, the structure of the RR-120 cannot be subject to modification. It is clear that the boost in temperature goes hand in hand with a boost in the adsorption energy proving the influence of temperature on the environment of the adsorbate.

To understand any adsorption phenomenon, we need to introduce the energy concept. Looking at Figure 7, we can notice clearly the variation of the internal energy with the temperature.

It has been proved that internal energy keeps the same pace and is always negative. Hence there is an energy release, and this is how the spontaneity as well as exothermicity of the phenomenon is proved. As long as the adsorption of the RR-120 dye on the activated carbon is physisorption, the previously mentioned result remains predictable and expected. In fact, what explains well the increase of internal energy in absolute value with the concentration is the increase of adsorbed quantity. Added to that, we should mention that the increase of the adsorption energy leads systematically to the increase in absolute values.

Looking at Figure 8, we can notice undoubtedly the variation of the free enthalpy with the temperature. Either from zero or all other temperature degrees, the free enthalpy gets reduced algebraically. Nevertheless, we should bear in our minds that the absolute value of the free enthalpy increases with the concentration. As soon as the adsorption process is set in motion, we become able to notice that the receptor sites are vacant, proving the way to an easy molecule retention.

Getting close to a saturation status and through the high concentration rate, RR-120 dye molecules are by no means able to reach the receptor sites of activated carbon since they are all occupied. In this respect, we can conclude from the figure that the increase of temperature obviously results in a rise in the free enthalpy. At this point, the system becomes clearly more stable. More than this, the spontaneity of the RR-120 dye adsorption on activated carbon is proved by the negative value of the free enthalpy.

Figure 9 illustrates the variation of entropy as a function of temperature. When concentration reaches a low degree (i.e., for concentration levels which are below the half-saturation concentration), we may notice that the receptor sites remain empty; the RR-120 dye molecules have a diversity of receptor sites to choose to give a more significant room for the disorder to increase. Hence, we can understand why the entropy, along with the concentration, has risen at first. When the surface of activated carbon tends to be saturated, the disorder lessens. This occurs after the half-saturation. By doing so, the entropy lowers, with a high possibility of becoming zero.

This work allowed us to present the analytical models developed by the formalism of statistical physics. An attempt has been made to model the adsorption isotherms of the dye RR-120 on activated carbon. The double layer model with one energy is the best fitting model. This model has allowed us to achieve the following results. Steric and energetic interpretations were performed. The adsorption of RR-120 on activated carbon is of the physisorption type, as confirmed by the adsorbed energy values and it is exothermic as verified by the internal energy. The free enthalpy has negative values, whatever the temperature, which demonstrated the spontaneity of the phenomenon. The monitoring of the disorder of the system was realized through the study of entropy.

The authors thank the Foundation for Research Support of the State of Rio Grande do Sul (FAPERGS), National Council for Scientific and Technological Development (CNPq, Brazil), and Coordination of Improvement of Higher Education Personnel (CAPES, Brazil) for financial support and sponsorship. Authors are grateful to Nanoscience and Nanotechnology Center (CNANO-UFRGS), and Microscopy and Microanalysis Center (CME-UFRGS) of Federal University of Rio Grande do Sul (UFRGS). We are also grateful to ChemAxon for giving us an academic research license for the Marvin Sketch software, Version 20.3 (http://www.chemaxon.com), used for molecule physical–chemical properties.

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