Thermodynamics, kinetics, and isotherms for phenol removal from wastewater using red mud

a₁

Elovich constant

b

Langmuir constant (L/mg)

B₁

Temkin constant connected with adsorption heat (J/mol)

b₁

Elovich constants

C_a

Amount of phenol adsorbed on the adsorbent per liter of the solution at equilibrium (mg/L)

C_e

Phenol concentration in solution at equilibrium (mg/L)

C₀

Initial phenol concentration in solution (mg/L)

C_t

Phenol concentration in solution at time t (mg/L)

D_e

Effective diffusion coefficient of adsorbate in the adsorbent phase (m²/sec)

E

Mean adsorption energy (kJ/mol)

F(t)

Ratio of the amount of Phenol adsorbed per gram of adsorbent at any time to that at equilibrium time

ΔG⁰

Gibbs free energy (kJ/mol)

ΔH⁰

Enthalpy (kJ/mole)

K_ad

Pseudo-first-order rate constant (/min)

K'

Pseudo-second-order rate constant (mg/g/min)

K_bq

The constant obtained by multiplying q_max and b

Apparent equilibrium constant

Thermodynamic equilibrium constant

K_f

Freundlich constant

K_T

Equilibrium binding constant (L/g)

M

Mass of the adsorbent per unit volume (g/L)

m_s

Amount of adsorbent added (g)

n

Freundlich constants intensity of adsorption (mg/g)/(mg/L)

q_e

Amount of adsorbate (mg) adsorbed per gram of the adsorbent at equilibrium

q_max

Maximum adsorption capacity (mg/g)

q_t

Amount of adsorbate (mg) adsorbed per gram of adsorbent at time t

q_∞

Amount of adsorbate (mg) adsorbed per gram of adsorbent at infinity

r²

Correlation coefficient

R

Ideal gas constant (kJ/mol/K)

R_a

Radius of the adsorbent particle (m)

R_L

Separation factor

S_S

External surface area of the adsorbent per unit volume (/m)

ΔS₀

Entropy (kJ/mol K)

t

Time (min)

T

Temperature (K)

t_o

Elovich constant equals to 1/(a₁b₁)

V

Volume of the solution (L)

X_m

Maximum adsorption capacity of adsorbent (mill mol/g)

β

Mass transfer coefficient (cm/sec)

λ

Constant related to adsorption energy (in square mol/square kJ)

ε

Polanyi potential (in square kJ/square mol)

Phenol (C₆H₅OH) has been considered as a significant contaminant present in several domestic, agricultural, and industrial wastewaters (Mu'azu et al. 2017). It is treated as a severe human pollutant because of its high toxicity, since the toxicity leads to several chronic issues like anorexia, vomiting, fainting, kidney and liver damage, etc. (Busca et al. 2008). Phenol is treated as a severe environmental pollutant because it reduces soil porosity, affects the natural germination of seeds, and increases soil seepage, changing groundwater quality (Lakshmi et al. 2016). The impact owing to these contacts, however, varies with the quantum of phenol and its duration. Hence, phenol removal from its wastewater to an acceptable limit is necessary.

The industrial wastewater that contains phenol is usually discharged from industries like steel factories, oil refineries, coal-tar units, industrial resins, rubber plants, paint/disinfectant/pharmaceutical plants (Dargahi et al. 2017). Phenol can also be present in agricultural and domestic wastewater. The phenol content in industrial wastewater may group to the extent of 6,800 mg/L, whereas the permissible limit set by the EPA, USA, is 1.0 mg/L. Therefore the phenol quantum from wastewater must be reduced to this permissible limit before the wastewater is discharged into the surface water (Almasi et al. 2018). Although it is a significant pollutant in the wastewater, there are several other pollutants like metal ions, dye, etc. also prevailing in the wastewater, which can affect the water purification process.

Several methods are used to remove phenol from wastewater. They are solvent extraction, separation, flocculation, electro-Fenton, photo-degradation, oxidation, adsorption, etc. These methods are classified into physical, chemical, and biological types. Although there are many advantages associated with these methods, there are also certain disadvantages in many of them (Uddin et al. 2007). Many of these methods are not suitable because of the possibility of discharge of additional toxic materials or the high cost of the process (Naiya et al. 2009). The adsorption technique is usually considered as effective and eco-friendly (Divate & Hinge 2014). Although the activated carbons are generally considered the most efficient adsorbent, they are not favoured due to their high cost (Shokoohi et al. 2017). Therefore the applications of various natural adsorbents like algae, greensand, pumice, etc. are being investigated to replace activated carbon (Safari et al. 2015; Naghipour et al. 2016, 2018; Jaafari & Yaghmaeian 2019a, 2019b). The vital factor for choosing an adsorbent is, of course, a high surface area, microporous structure, and a high degree of surface reactivity (Sarkar & Das 2015, 2016). The applications of industrial and agricultural waste as adsorbents are studied through many researchers (Mandal et al. 2018, 2019a, 2019b, 2020; Mandal & Das 2019a, 2019b). Like the agricultural wastes, industrial wastes are also available in large quantities and almost free of cost. In this perspective, the red mud, industrial waste, is used as an adsorbent in this research. The porous structure of the red mud, having a high surface area, is usually supportive of adsorption. The presence of compounds like iron oxide, aluminum oxide, titanium oxide, calcium carbonate, sodium aluminum silicate, dicalcium silicate, and silicon oxide in red mud provides an enhanced binding spot for adsorption. The disposal of the used red mud is usually possible by incineration at a temperature above 800 °C, and its regeneration is not required since the red mud is generated in large quantities and is thus available almost free of cost. There is no cost involved in incineration as red mud, when used for brick manufacturing purposes, is required to be heated above 1,000 °C. The gradual increase of red mud applications for brick manufacturing purposes will thus reduce the disposal problem of the used adsorbents.

The main apparatuses used were pH meter, Hach Germany; UV-Spectrophotometer, Hach Germany; scanning electron microscope (SEM), Hitachi Japan; X-ray diffractometer (XRD), Bruker Germany; Fourier transform infrared (FTIR) spectroscope, Thermo Fisher USA; BET, Quantachrome USA. A B.O.D. incubator shaker, heating furnace, and digital balance were also used for the experimentation. The chemicals, like analytical grade phenol (purity 99.99%), NaOH, and HCl from M/s Merck, India, were arranged for the experiments.

The red mud was collected from NALCO, Damonjodi, Odisha, India. It is a by-product during alumina (aluminum oxide) production from bauxite. In general, 1.2–1.6 tonne of red mud is produced from 1 tonne of alumina production. The red mud was appropriately washed in distilled water and was kept for 8 hours at 110 °C in a furnace. It was then cooled at ambient condition, powdered in a grinder, and screened using −44 and +52 mesh sieves. The sieved quantity was then placed in a desiccator for batch experiments.

The surface topography assessed through the SEM apparatus is placed as Figure 1. The Figure 1 suggests that the red mud particle has numerous uneven pores. The SEM image further implied that the surface area of the red mud was high due to high porosity, with a relatively loose microstructure. Thus it was supportive of favourable adsorption.

The XRD spectra in Figure 2 suggest the presence of Fe₂O₃, Al(OH)₃, TiO₂, CaCO₃, NaAlSiO₄ Ca₂SiO₄, and SiO₂.

The FTIR spectra in Figure 3 suggest that the 468 cm⁻¹ and 521 cm⁻¹ peaks were associated with Fe-O of Fe₂O₃. The peak at 560 cm⁻¹ represented O-Si-O. The 730 cm⁻¹ peak represented the Al–O bond. The 998 cm⁻¹ peak corresponded to Si-O-Si in the silicate group. The band at 1,002 cm⁻¹ was for Si–O–Si in the silicate group. The band at 1,410 cm⁻¹ was described as the C = O stretching vibration and indicated the carbonate group. The peaks at around 3,415 cm⁻¹, 1,640 cm⁻¹ and 2,850 cm⁻¹ were assigned to the hydroxyl group vibration. The identified absorption band at 3,600–3,400 cm⁻¹ corresponds with the vibration. These functional groups in red mud helped in chemical interactions through hydrogen bonding with the phenolate ions. The BET surface area was 300 m²/g. The bulk density of the adsorbent was found to be 3 g/cm³. The point of zero charge (pH_pzc) by the salt addition process was pH 7.65. The physical features and the chemical compositions of red mud are placed in Tables 1 and 2.

The stock solution of aqueous phenol was prepared to take 1gm phenol (C₆H₅OH) in 1,000 ml double-distilled water. The concentration of this solution was 1,000 mg/L. This solution was diluted to prepare the desired concentration of phenol (40–200 mg/L). This procedure of dilution of the stock solution was considered to reduce the error in the weight measurement of a small quantity of crystal phenol required for desired low concentrations. The calibration chart was prepared using the absorbance values from the spectrophotometer corresponding to a known concentration of phenol. The same was used for the determination of the resultant phenol concentrations after adsorption.

The phenol adsorption by red mud was tested in batch procedures with the variation of phenol content, pH, contact period, dose, and temperature. In every batch, 100 ml desired phenol solution (40–200 mg/L) was taken in 250 ml conical flask. The phenol concentration range of 40–200 mg/L was selected in this research work since the contaminant phenol available in the wastewater of many medium scale industries is in this range. The pH was varied from 2 to 12 using HCl and NaOH. The red mud dose was varied from 2 to 7 g/L. The temperature and contact time were maintained at 30–50 °C and 120–600 min using a thermostatic shaker at 120 strokes/minute. At the end of the experiment, the absorbance of the filtrate was measured in the UV-spectrophotometer, and the residual phenol content was checked using the calibration curve. The experiments were repeated thrice, and the average was accounted for more precision. Reproducibility and relative variance were found to be ±0.5% and ±0.25%, respectively. The standard deviation of the variance was shown in the relevant graphical presentation.

The batch tests were conducted with the variation of time (120–600 min) and phenol concentration (40–200 mg/L). The pH, dose, and temperature were unchanged at 8, 6 g/L, and 50 °C. The phenol removal versus time was as shown in Figure 4. The graph described that the removal increased in the beginning and then became constant after 480 min. Similar results were found for all phenol concentrations (40–200 mg/L). The removal was faster initially as the adsorbent had empty sites. Meanwhile, the oxides and silica present in red mud took part in the adsorption, and the oxygenated surface of the adsorbent facilitated the adsorption by hydrogen bonding with the hydroxyl surface of the phenol. However, when the equilibrium arrived at 480 min, regardless of phenol concentration the adsorption sites became saturated and therefore, after that, no more adsorption was possible. The decrease of phenol removal with the increase of the initial phenol concentration can be explained as the red mud has a limited number of active sites; and can be quickly saturated (Tor et al. 2006).

The experiments were conducted with the variation of dose (2–7 g/L) and phenol contents (40–200 mg/L). The pH, time, and temperature were unchanged at 8, 480 min, and 50 °C. The phenol removal versus dose is shown in Figure 5. The graph describes that the removal increased in the beginning and then became constant after the dose of 6 g/L. Similar results were found for all phenol concentrations (40–200 mg/L). The removal was faster initially, as the adsorbent had empty sites. The higher rate of adsorption continued because of the addition of surface area through the continued addition of adsorbent. This phenomenon was observed up to the adsorbent dose of 6 g/L. However, the addition of more quantity of adsorbent beyond 6 g/L was not required because the active adsorbent sites got saturated, and the adsorbent sites were overlapped due to overcrowding of red mud particles.

The experiments were conducted with the variation of pH (2–12) and phenol contents (40–200 mg/L). The dose, time, and temperature were unchanged at 6 g/L, 480 min, and 50 °C. The phenol removal versus pH is shown in Figure 6. Similar results were found for all phenol concentrations (40–200 mg/L). The graph showed an increase in phenol removal with pH 2–8. After pH 8, the removal, however, decreased considerably. Hence, the optimum pH was 8. The reason for this observation was the amphoteric properties of red mud at the point of zero charge (pH_pzc) 7.65. The amphoteric behaviour was due to the zero surface charge at pH 7.65 when the adsorbent came in contact with water. The functional groups present in the adsorbent played a significant role in such amphoteric characteristics. The pK_a value of phenol is 9.9, which indicates that the phenol is a weak acid and exists as anions at higher pH. As the pH of the solution surpasses the pK_a, the phenol exists as phenolate ions, and at below the pK_a value, it exists as a neutral molecule (Kumar et al. 2014). The pH_PZC values of Fe₂O₃, Al₂O₃, CaO, and SiO₂ are 8.5, 8.3, 11.0, and 2.2, respectively. The optimum pH is 8; hence, for pH ≤ pH_PZC an electrostatic attraction exists between the positively charged surface of the red mud and the phenolate ion. At pH ≥ 9, the red mud surface and phenolate anions will repel each other because of the negative surface charge of Fe₂O₃, Al₂O₃, the significant component in the red mud. At low pH, that is, the acidic solution that causes extra protons to exist in the solution, there will be a competition between the protons and phenol, hence the significant drop in adsorption (Halouli & Drawish 1995; Nagda et al. 2007).

The tests were conducted with the variation of temperature 30–50 °C) and phenol contents (40–200 mg/L). The pH, time, and dose were unchanged at 8, 480 min, and 6 g/L. The phenol removal versus temperature is shown in Figure 7. Similar results were found for all phenol concentrations (40–200 mg/L). The intra and intermolecular hydrogen bond becomes weak at a higher temperature; hence, phenol molecules are freely available for adsorption (Gupta et al. 2004). The graph describes that the removal increased with temperature because of better chemical interaction between red mud and phenol due to the increase of electrostatic forces. At higher temperatures, the mass transfer rate from the bulk to the boundary layer around the red mud surface increases due to a decrease in the boundary layer thickness. This enhances the rate of chemical reaction, hence adsorption capacity increases, and the adsorption process is chemical. The reduction of Gibbs free energy evaluated by thermodynamic analysis also supported better chemical interaction at the higher temperature. Hence, the adsorption was endothermic (Ekpete et al. 2010).

The phenol removal rate with time was studied using different phenol contents (40–200 mg/L) at constant pH, dose, and temperature. The results showed maximum phenol removal at 480 min. These results were tested using seven kinetic models.

This model suggests physisorption with the homogeneous surface of the adsorbent involving Van Der Waal forces of attraction. The model parameters were calculated from the graph of versus t. The higher the value of rate constant K_ad, the faster the rate of adsorption, and the higher the value of q_e, the higher the amount of phenol that is adsorbed by red mud in equilibrium.

This model suggests chemisorption with the heterogeneous surface of the adsorbent involving chemical bonding forces of attraction (Chaudhary et al. 2014). The parameters were calculated from the graphs against t. The higher the value of rate constant K’, the faster will be the rate of chemical adsorption and the higher the value of q_e, the higher will be the amount of phenol uptake capacity by red mud in equilibrium.

The results of pseudo kinetic models reported in Table 3 indicated that the pseudo-second-order model was better statistically. Table 3 also shows that the derived quantities of q_e for the pseudo-second-order model were more proximate than the experimental q_e. Therefore, the adsorption method would be chemical and heterogeneous (Singha & Das 2012).

The conformity of this model could be established by studying their correlation coefficients after the graphs are plotted using the values of Bt against t. The support of this model suggests film diffusion, and the higher the diffusion rate constant Bt, the faster the rate of film diffusion will be through mass transfer. The plot is a straight line but not passing through the origin, indicating the film diffusion-controlled process (Gupta 1998).

This model can be analyzed using graphs of q_t/q_e versus t. If the linear plot of the figure shows a single line, the process will have intra-particle diffusion. The first straight line will signify film diffusion if the figure shows two or more multi-linear lines after linearization. The second will signify pore diffusion, and the third will mean intraparticle diffusion. The corresponding time for each type of diffusion can be evaluated from the graph. The involvement of maximum time will decide the rate-determining step occurring in the adsorption process. The ratio of film diffusion time to intra-particle diffusion time was found to be 1:1; hence, the process was regulated jointly by film and intra-particle diffusion.

Here, the K_bq was evaluated by the product of b and q_max evaluated from the Langmuir isotherm. From the graph of versus t, the parameters were estimated. This model used for mass transfer mechanism is developed for external film diffusion during the initial adsorption process. The coefficient of mass transfer, β, can be calculated from the graphs for estimating whether phenol movement from solution to the adsorbent surface is faster. The higher the value of mass transfer rate constant β, the quicker will be the rate of phenol movement through external film diffusion.

This model is tested using graph q_t versus lnt. The higher the value of rate constant a₁, the faster will be the rate of adsorption, and the higher the value of b₁, the higher will be the rate of desorption.

This model can be tested using graph and t. If the external mass transfer is greater than the internal one, film diffusion controls, and if the internal mass transfer is greater than the external one, particle diffusion controls. This can be verified from the linear plot of the graph. In the case of a linear plot passing through the origin, particle diffusion controls or else film diffusion controls. The diffusion coefficient, D_e, can then be derived from the plot. The process will be chemical if D_e falls between 10⁻⁹ and 10⁻¹⁷ m²/sec.

The results of kinetic testing are reported in Table 4. The Reichenberg model suggested film diffusion, whereas the Fick model indicated that the film and intra-particle diffusion were both equally important. The Furusawa and Smith model suggested faster phenol movement. The Elovich model and the Boyd model suggested the chemisorption process.

The phenol removal was tested at different temperatures and phenol contents (40–200 mg/L) with constant pH, dose, and time. The results suggested that the phenol removal was increased with temperature and indicated the endothermic chemical reaction. The isotherm study was performed using four models.

The Langmuir model is favourable if R_L lies between 0 and 1.

From the graph of log q_e against log C_e, K_f and are derived. The higher the value of K_f, the higher will be the multilayer adsorption capacity and the higher the value of n, the higher will be the adsorption's free energy. The Freundlich model supports that the adsorbent surface is heterogeneous, suggesting multilayer adsorption over a large number of available sites acting simultaneously. The values of between 0 and 1 indicate heterogeneous adsorption.

From the graph of q_e against lnC_e, the parameters B₁ and K_T can be evaluated. The Temkin constants B₁ and K_T signify adsorption heat and equilibrium binding energy, respectively.

The plot of against determines the values of ΔG^o, ΔS^o, and ΔH^o, which are reported in Table 6. The values of ΔG^o being negative, the adsorption process was spontaneous. The positive ΔH^o supported an endothermic reaction. The positive ΔS^o suggested a higher degree of randomness in the adsorption process.

The adsorption mechanism is summarized as follows:

• i.

  Pore capture: the SEM image shows that the red mud is very porous and consists of many uneven pores. The phenol adsorption takes place on the pores of the red mud.

• ii.

  Hydrogen bonding: oxygen-containing functional groups, as identified by FTIR spectra, promote the phenol adsorption by hydrogen bond formation by the interactions between the red mud surface and phenol. The peak in the FTIR (Figure 3) at 2,850 cm⁻¹ corresponds to the –OH group present on the red mud surface. The intermolecular hydrogen bonding between phenol and red mud increases the adsorption.

• iii.

  Electrostatic interaction: The constitution of red mud is of different metal oxides. There is the formation of the metal-phenol complex onto the hydrous metal oxide. It is due to the ligand exchange process between the phenol and hydroxyl group on the red mud surface. The pH_PZC values of Fe₂O₃, Al₂O₃, CaO, and SiO₂ are 8.5, 8.3, 11.0, and 2.2, respectively. The optimum pH is 8; hence, for pH ≤ pH_PZC, electrostatic attraction exists between the positively charged surface of the red mud and the phenolate ion.

The comparison of the maximum adsorption capacity for phenol removal by different adsorbents and the present study is shown in Table 7. Red mud has comparable adsorption capacity and therefore is suitable for practical use. The low cost and the vast availability of red mud are the added advantages for its selection by many industries.

Equation (23) is applied for scale-up design, and the weight of red mud (g) required for phenol removal is shown in Table 8.

The red mud used as adsorbent was generated at the bauxite refinery plant as a waste by-product. It is available at little cost, its regeneration is also not significant, and the researchers do not insist on any desorption technique for reuse of such adsorbents. The used adsorbents were destroyed by incineration at 800 °C. The incinerated ash is suitable for rood making or land-filling. The used adsorbents are also suitable as one ingredient for the production of bricks, which requires above 1,000 °C.

The phenol removal rate on red mud was investigated in this research. The surface characterizations of red mud were carried out. The batch tests were conducted with phenol content (40–200 mg/L), pH (2–12), time (120–600 min), dose (2–7 g/L), and temperature (30–50 °C). The maximum removal (87.5%) was obtained at pH 8, time 480 min, and dose 6 g/L. The adsorption kinetic was the pseudo-second-order. The Reichenberg model supported film diffusion, and the Fick model supported the equal importance of film and intra-particle diffusions. The Furusawa and Smith model suggested rapid phenol movement to the adsorbent surface. Both Boyd and Elovich models suggested chemisorption. The Freundlich model was the best-fitted isotherm model among Langmuir, Temkin, and Dubinin–Radushkevich models. The process was endothermic, random, and spontaneous. The mechanism of adsorption involved in the phenol removal process was deliberated with conclusive explanations. The safe dumping of used red mud was tested by incineration. The scale-up design was calculated to find the quantity of adsorbent required for phenol removal. The study concludes that red mud is suitable for phenol removal from wastewater.

The authors have no conflict of interest to disclose.

Data cannot be made publicly available; readers should contact the corresponding author for details.