Coconut husk-raw clay-Fe composite: preparation, characteristics and mechanisms of Congo red adsorption

The release of unspent dyes from industries constitutes hazard and environmental challenges. For rapid and efficient removal of Congo red from aqueous solutions, a composite was prepared from coconut husk, raw clay, Fe(II) and Fe(II) compounds. Adsorption variables (initial pH of the solution, contact time, temperature and initial concentration of Congo red) were varied to understand the characteristics and mechanisms of the adsorption process. The composite was characterised using Fourier Transform Infrared (FTIR) spectroscopy, Scanning Electron Microscopy (SEM)–Energy dispersive X-ray (EDX) spectroscopy, nitrogen adsorption–desorption isotherm, X-ray Diffraction (XRD) spectroscopy and pH of the point zero charge (pHpzc). The optimal values of the pH, equilibrium time and temperature for adsorption of Congo red by the composite are 2, 40 min and 50 °C, respectively. The kinetic and equilibrium data followed Avrami fractional order and Langmuir models, respectively. A 1.0 g of the composite could maximally take up 1649.3 mg of Congo red at 50 °C. The values of ΔG° are in the range of − 27.901 to − 24.492 kJ mol–1 while the value of ΔH° is − 72.239 kJ mol–1. Hence, the removal of the Congo red by the composite was spontaneous, feasible and exothermic. The adsorption process was biphasic and followed physisorption process. Electrostatic interaction played a significant role in the removal of Congo red by the composite. The combine data in this study have proven that the clay composite, a cheap adsorbent, can be used for remediation of water contaminated with Congo red.


Materials and methods
Preparation of the composite. Analytical grades of reagents were used without further treatment.
Coconut husks were obtained from a dumping site of a local market in Akure, Nigeria while raw clay was furnished by the Department of Industrial Design, The Federal University of Technology, Akure, Nigeria. The two precursory materials were taken to the Department of Crop, Soil and Pest Management of The Federal University of Technology, Akure, Nigeria for proper identification and authentication. The coconut husks were washed with tap water and subsequently with distilled water, sun-dried for 7 days to drive off the moisture, pulverised and sieved with a 2 mm mesh sieve to get a sample of smooth and small particles. The sample was properly kept until usage. The raw clay sample was sun-dried for 5 days, oven-dried at 90 °C for 24 h, crushed with mortar and pestle, sieved with a 2 mm mesh to get clay of fine particles, and kept until usage.
A 100 g of the husk, 100 g of the raw clay, and 300 mL of distilled water were adequately mixed and this mixture was labelled as mixture A. In another preparation, a 40 g ferric sulphate (Fe 2 (SO 4 ) 3 www.nature.com/scientificreports/ and dissolved in 50 mL of water while 40 g of ferrous sulphate (FeSO 4 ·7H 2 O) was also weighed and dissolved in 50 mL of water. The two solutions were mixed together, and this formed mixture B. Lime (20 g) was prepared by weighing and mixing CaO, Ca(OH) 2 and CaCO 3 , and dissolving in 50 ml of water and this solution was named as mixture C. The three mixtures were mixed and stirred on the magnetic stirrer for 1 h at 70 °C. To the resulting mixture still on the magnetic stirrer, a few drops of 10 mol L -1 NaOH was added to the solution while stirring to maintain a pH of 10 24 . The final mixture was allowed to stand for 24 h for proper equilibration. The  clear solution was removed and the residue (adsorbent) was washed with distilled water, oven-dried for 48 h at  80 °C, sieved with a 106 µm mesh sieve, labelled as CHCFe (coconut husk-clay-ferrate composite), and kept in an airtight plastic container in a desiccator until usage. The lime in the composition was to allow proper mixing of the components 25 .
Techniques for characterisation of the composite. The functional groups of the composite were obtained using Fourier Transform Infrared Spectroscopy (FTIR; Shimadzu Spectrometer, IR Prestige 21, Japan). The CHCFe and KBr were separately oven-dried at 393 K for 7 h, and stored in a desiccator before analysis. The spectrum of the CHCFe was recorded between 4000 and 400 cm -1 . The surface feature of CHCFe was visualised using Scanning Electron Microscopy (SEM; Zeiss DSM 982 Gemini). The specific surface area, pore size distribution and total pore volumes of CHCFe were obtained by using BET (Brunauer, Emmett and Teller) multipoint technique, and BJH (Barret, Joyner and Halenda) from N 2 adsorption-desorption isotherm at 77 K using a surface analyser (Micrometrics TriStar II 3020). The composite was also subjected to X-ray diffraction (XRD) spectroscopy to obtain its diffraction patterns using Philips X'pert MPD diffractometer (Netherlands), which was operated at 40 kV and 40 mA with CuKa radiation (λ = 1.5406 Å). By using pH drift method 2,5,25 , the value of pH of the point zero charge (pHpzc) of CHCFe was obtained.
Adsorption experiments, modelling and statistical analysis. A 5.00 g L -1 stock solution of Congo red was prepared and this solution was diluted serially for preparation of various working solutions. Removal of Congo red, using the composite, was achieved by using batch adsorption method. In various adsorption experiments to probe the effects of pH, contact time, initial concentration of Congo red and temperature on adsorption characteristics of CHCFe, 20 mL Congo red solutions (0.02-3.00 g L −1 ) was added to 0.05 ± 0.003 g of CHCFe in 25 mL Falcon tubes at varying pH values (2-10), contact times (0-480 min), and temperatures (25, 30, 35, 40 and 50 °C). Different dye solution-composite mixtures were agitated thermostatically on a shaker at 160 rpm (revolution per minute) and centrifuged at the end of the adsorption experiments for separation of adsorbent from un-adsorbed Congo red in solution. The concentration of the dye left (un-adsorbed) in the filtrate solution was measured at 497 nm on UV/Visible spectrophotometer (Shimadzu, Japan). The absorbance readings from the spectrophotometer were converted to concentrations using a standard calibration curve. All adsorption experiments were carried out three times, and the mean values were used for further analysis. The amount of Congo red adsorbed (Q e , mg g −1 ) from solution and the percentage of adsorption (%Ads) were evaluated with the aids of Eqs. (1) and (2), respectively.
where C o and C e stand for respective initial and final (equilibrium) concentrations of Congo red in mg L −1 ; V denotes the volume, in litre (L), of the aqueous phase; and m is the weight (g) of the CHCFe used for each experiment.
The data obtained from time-dependent experiment were subjected to non-linear forms of pseudo-first order, pseudo-second order, Modified Ritchie second order, Avrami fractional order and intraparticle diffusion kinetic models, which are shown respectively in Eqs. (3)- (7). The concentration and temperature-dependent (isothermal) data were further analysed using non-linear forms of Langmuir, Freundlich, Dubinin-Radushkevich (D-R), and Toth equilibrium models, which are presented in Eqs. (8)-(11), respectively. www.nature.com/scientificreports/ In the above equations, Q t is the amount (mg) of Congo red removed by 1.0 g of CHCFe at time, t; k f denotes the pseudo-first order rate constant (min -1 ); k s represents the pseudo-second order rate constant (g mg -1 min -1 ); k mR stands for the modified Ritchie second order rate constant (min -1 ); β stands for the Ritchie's constant, which is related to initial loading of the pollutant on adsorbent (min -1 ); k Av represents the Avrami fractional order rate constant (min -1 ); n AV denotes the Avrami's fractional order, which is related to the mechanism of adsorption; k ipd is the intraparticle diffusion rate constant (mg g -1 min -0.5 ); C represents the intraparticle diffusion constant, which is related to the thickness of the boundary layer (mg g -1 ); Q max is the maximum adsorption capacity of the adsorbent (mg g -1 ); C e represents the equilibrium concentration of the dye (mg L -1 ); K L is the Langmuir equilibrium constant (L mg -1 ); K F is the Freundlich equilibrium constant (mg/g)(mg/L) − 1 n F ; n F denotes the Freundlich's exponent; K DR is the activity coefficient, which is related to mean free energy of adsorption (mol 2 kJ −2 ); K T denotes the Toth equilibrium constant (L mg −1 ); and n T stands for the Toth exponent (dimensionless).
where R is the universal gas constant (8.314 J mol -1 K -1 ) and T denotes the temperature in K.
Non-linear curve fittings and calculations of parameters were done on OriginPro 9. The R 2 adj (adjusted determination coefficient) and SD (standard deviation) values of each model were used to determine the best model of the adsorption data. The statistical representations of R 2 adj and SD are presented in Eqs. (14) and (15), respectively.
where q i,exp represents an individual, i, data obtained from the batch experiment, q i,model represents the estimation of the corresponding q i,exp generated by each model; n is the number of data points; and p is the number of parameters in the model. A small value of SD indicates a good curve fit by the model while a value of R 2 adj close to unity also signifies a good curve fit by a model 2,5 .

Results and discussion
Preparation and characterisation of CHCFe. In a quest for an effective adsorbent material for wastewater treatment, a composite (CHCFe) was prepared from coconut husk (an agricultural by-product), raw clay (abundant and cheap), magnetic particles (have a good tendency for capturing environmental pollutants 26 and lime that ensured proper mixing of all the components 3,25 . The composite was successfully used for removal of Congo red from water. The characteristics of the adsorbent were elucidated by using SEM-EDX, FTIR, XRD, nitrogen adsorption-desorption, and pH pzc . Scanning electron microscopy is a versatile method for visualisation of the microstructure morphologies of materials to obtain information such as the shape and size of the particles of the materials 27 . The surface of the CHCFe (Fig. 1a) is coarse and rough. There are a few non-uniform aggregates on the surface of the adsorbent. The surface is characterised with heterogeneous non-uniform particles, which dye molecules can adhere to. There are also holes on the surface that can encapsulate Congo red molecules 28 . The EDX spectrum (Fig. 1b) confirmed the presence of Fe, Si Al and Ca apart from the major elements, carbon and oxygen. The revelation of the elements is in agreement with the composition of the CHCFe.
Fourier transforms infrared spectroscopy is a technique that provides information on available active sites, which can bind or interact with pollutant molecules, on the surface of the adsorbent 29 and this technique was used The nitrogen adsorption-desorption isotherm of CHCFe (Fig. 1e) was used for evaluation of surface area and other parameters. The nitrogen adsorption-desorption data was subjected to Brunauer, Emmett and Teller (BET) and Barret, Joyner and Halenda (BJH) models. The parameters obtained from BET and BJH are BET surface area of 100.30 m 2 g -1 , pore volume of 0.099 cm 3 g -1 and half pore width of 23.392 Å. The good surface area of the composite substantially enabled excellent adsorption of Congo red.
Preliminary study and variation of pH on adsorption behaviour of CHCFe. The initial pH of the solution mostly has profound effects on adsorption capacities of adsorbents 5 . The effects of variation of pH on percentage adsorption of the dye by CHCFe (Fig. 1f) was investigated between pH 2 and 10, initial Congo red concentration of 2 g L -1 , contact time of 4 h and 25 °C. The adsorbent adsorbed Congo red best at lower pH values. The percentage removal at pH 2 is 99.9 while that of pH 4 is 98.22. Hence, pH 2 is the optimal pH for the removal of Congo red by CHCFe. The result is in agreement with the previous report by Lafi et al. 10 , who reported pH 3 as optimum for removal of Congo red using activated coffee waste. Popoola et al. 5 also reported optimum pH 3 for removal of Congo red using synthesised coal graphene. Removal of Congo red dye onto the adsorbents is usually influenced by pH because pH influences polarity of the surface of adsorbents as well as ionisation and ionic mobility of Congo red 13 . Congo red has a bipolar characteristic and can behave as anionic and cationic compound in solution depending on the pH of the medium 5 . This behaviour makes Congo red to change colour as pH is increased from acidic pH to alkaline pH. Removal of Congo red with anionic character is expected to increase as the initial pH of the solution decreases 10 . The value of pH pzc of CHCFe is 6.97 (Supplementary Fig. S3) and this value signifies that the surface of the CHCFe is positively charged relative to the optimum pH, that is pH < pH pzc 5,7,25 .
In the preliminary experiments conducted at pH 2 ( Supplementary Fig. S4), coconut husk, clay and coconut husk-clay (untreated) and coconut husk-clay-Fe composite were evaluated for their potentials to adsorb Congo red from aqueous solutions. The composite exhibited the best adsorption capacity than other adsorbents shown in the figure. Inclusion of iron compounds in the preparation of the composite could be responsible for surface activation, and hence increase the binding of Congo red molecules onto the surface of CHCFe.
Kinetics of the adsorption process. Amount of Congo red adsorbed by CHCFe is greatly time-dependent (Fig. 2a) at the beginning of the experiments. Amount of Congo red adsorbed increased progressively as the contact time increased. It means that there was a rapid removal of the dye molecules at the start of the adsorption process and this is because there were enough vacant binding sites on the surface of CHCFe at that moment. Adsorption of Congo red onto CHCFe was not time dependent after the equilibrium time of 40 min. Figure 2a also displays the modelling of time-dependent data to kinetic models. Pseudo-first order, pseudo-second order, modified Ritchie second order and Avrami fractional models were employed to elucidate the kinetic behaviour of the adsorption process, and the kinetic parameters (Table 1) were obtained from the models. To adjudge the best kinetic model, the values of SD and R 2 adj were used. Avrami fractional model exhibited the lowest SD value and highest R 2 adj value among the four kinetic models. This model presented the lowest difference between the values of experimental q (Q e(expt) ) and calculated/theoretical q (Q e(calc) ). Avrami model, which is an empirical model, is useful for explaining the behaviour of solid-solution interactions in adsorption systems 5,34,35 . The equilibrium adsorption value predicted by the model is 1383.3 mg g -1 at 25 °C while the value of Avrami rate constant (k AV ) is 0.23737 min -1 .
The mechanism of adsorption process can be obtained from the analysis of kinetic data 36 . Therefore, the timedependent data were equally subjected to intraparticle diffusion model (Fig. 2b) to ascertain the mechanism of the adsorption process. The intraparticle diffusion parameters (Table 1) were evaluated from the figure. If the kinetic data are subjected to intraparticle model and multilinear plots are obtained, it means that there are two or more adsorption stages that govern the adsorption process 37 . There are two adsorption stages for the removal of Congo red by CHCFe because there exist two linear sections in the intraparticle curve. The first linear adsorption stage (external mass transfer) is regarded as a fast-kinetic process and in this stage, there was a rapid migration of Congo red molecules to the surface of CHCFe. The second adsorption linear stage (intraparticle diffusion) is a stage that was attained after equilibrium. At this stage, there was a diffusion of the dye molecules through pores of CHCFe 38 . The intraparticle diffusion plot did not pass through the origin. This observation means that there were some levels of boundary layer control and the intraparticle diffusion was not the sole rate-limiting step that controlled the rate of adsorption 37 .
Equilibrium/isotherm study. Adsorption isotherms describe the relationship between the concentration of substrate in solution and the amount of pollutant concentration (mg) adsorbed by a gram of adsorbent at a given temperature 2,39 . Hence, isotherm provides information on the interaction between the molecules of adsorbate in solution and the adsorbent surface 39 . Congo red removal by CHCFe at 25 °C (Fig. 2c) was concentration dependent, that is, Congo red uptake increased as the equilibrium concentration of dye increased. When the concentration of dye is low, the numbers of dye molecules available in the solution will be relatively low compared to the active sites on the surface of the adsorbent. At a high concentration of dye, more dye molecules will be present and these molecules will be more than what the active sites of adsorbent can take.  Thermodynamic evaluation. Pieces of information on feasibility, spontaneity, nature (either endothermic or exothermic) and disorderliness of adsorption process are always obtained from temperature-dependent data. From the van't Hoff plot ( Supplementary Fig. S5), some of the thermodynamic parameters in Table 3  Negative values mean that the adsorption of Congo red onto CHCFe was generally spontaneous and feasible. The value of standard enthalpy change (ΔH°) is negative and this observation signifies that adsorption system was exothermic. In the same manner, the value of standard entropy change (ΔS°) is negative and this implies that there was a decrease in the randomness at Congo red solution-CHCFe interface during adsorption process 10,40 . including CHCFe, utilised for removal of Congo red were compared. Table 4 presents the comparison data. Among the twelve (12) adsorbents listed in the table, CHCFe outperformed all other adsorbents. Modification of the surface with Fe(II)/Fe(III) and good adsorbing feature of clay are some of the factors that are responsible for fantastic adsorption behaviour of CHCFe. The composite, CHCFe, could be utilised as an alternative and efficient adsorbent for Congo red removal from contaminated water.
Proposed mechanisms of adsorption. Adsorption could be described as a surface-based process in which molecules of an adsorbate (a substance in gaseous or liquid phase) are accumulated on the surface of an adsorbent's material 41 . Mechanism of adsorption implies details of events that take place at each stage of adsorption or during entire adsorption process. There are different schools of thought on adsorption mechanisms, and adsorption mechanisms can be propositionally explained in several ways using different operating parameters. The mechanisms of Congo red adsorption onto CHCFe will be discussed by using (1) type of functional groups present in CHCFe, (2) nature of the isotherm curve, (3) initial pH of the Congo red solution, (4) the linearity of intraparticle diffusion curve, and (5) thermodynamic data. As discussed earlier, removal of Congo by CHCFe followed Langmuir equilibrium model. This model assumes that (1) adsorbate molecules are adsorbed on the binding or active sites located on the adsorbent's surface, (2) all the adsorption sites are energetically identical, (3) the binding of adsorbate on the adsorbent's surface is monolayer adsorption, (4) each active site can only bind a molecule of an adsorbate at a time, and (5) there is no interaction among the adsorbed molecules on adsorbent surface [42][43][44] . On the basis of IUPAC classification of adsorption isotherm, the isotherm of CHCFe (Fig. 2c) followed Type L isotherm with a long well-defined plateau 41,45 . In this isotherm, it is assumed that adsorption of adsorbate occurred on a microporous adsorbent in which the adsorbent's surface is covered with a monolayer of adsorbate 43 .
Congo red, an anionic/acidic dye, has sulphonic group (-SO 3 -) in its structure 10 . At a low pH, Congo red dye will dissociate to a polar group (R-SO 3 -), hence, Congo red will be favourably adsorbed at low pH since the surface of CHCFe is positively charged relative to the dye molecules 10,13 . The opposing charges on the surface of CHCFe and dye molecules enhanced electrostatic interactions between CHCFe surface and R-SO 3 -of dye     Table 3. Thermodynamic characteristics of the adsorption process. Conditions: initial pH of the solution = 2; mass of adsorbent = 0.05 mg; temperature = 25-50 °C; agitation time = 1 h; initial concentration of Congo red = 0.02-3 g L -1 ; volume of Congo red = 20 mL; and agitation speed = 160 rpm.  Interactions of adsorbent with adsorbate, which will eventually lead to removal of adsorbate from aqueous solutions, include hydrogen bonding, π-π interaction, n-π interaction, anion-π interaction, electrostatic interaction, dispersion interaction, and hydrophobic interaction 8,30,46 . Some of the existing interactions between Congo red and CHCFe include hydrogen bonding, n-π, π-π, anion-π, and electrostatic ( Fig. 3).
Weber and Morris 47 proposed intraparticle diffusion model for explanation of the adsorption process if the diffusion of adsorbate to the porous adsorbent is the rate-limiting step. In the present study, adsorption of Congo red onto CHCFe is biphasic, that is, bilinear curve was obtained. This scenario confirmed that adsorption process occurred through a fast external mass transfer of adsorbate followed by a slow intraparticle diffusion of adsorbate molecules into the inner pores of CHCFe. There exist variations in mass transfer of Congo red at the initial and final stages, hence, the linear segments of the plot did not pass through the origin 25 . In addition to this, diffusion was not the only mechanism responsible for Congo red adsorption onto CHCFe and that the adsorption process was governed by either pore or film diffusion 37,48 . The value of C (617.42 mg g -1 ) obtained for CHCFe-Congo red interaction is high. This value is higher than the values reported for a few studies for adsorption of Congo red 3,5,10,11 . When the value of C is high, it means that the effect of the boundary layer is pronounced 43 .
Molecules of adsorbate can be adsorbed onto the surface of adsorbent via physisorption (physical adsorption) or chemisorption (chemical adsorption) depending of the type of interaction between the adsorbent material and adsorbate. Physisorption is always exothermic and occurs at high relative pressures, physisorption process mainly takes place as a multilayer. A physisorption process will possess adsorption energy < 40 kJ mol -1 but a chemisorption process possess will exhibit energy > 40 kJ mol -146, 49,50 . The value of adsorption energy (enthalpy change) for removal of Congo red by CHCFe is − 72.239 kJ mol -1 , and this signifies that physisorption process predominates during adsorption process.

Conclusion
This research prepared an efficient adsorbent material from raw clay, coconut husk, and inorganic compounds and the adsorbent (CHCFe) was utilised for removal of Congo red from aqueous solution in a process that is feasible, spontaneous and exothermic. The data from FTIR, SEM-EDX, N 2 adsorption-desorption, XRD, and pHpzc of CHCFe correspondingly revealed that the adsorbent possessed suitable adsorptive functional groups, heterogeneous surface, good surface area, four major compounds, and positive surface. The rate of removal of Congo red by CHCFe was a fast process and followed Avrami fractional model. The adsorption process occurred in two phases: mass migration of Congo red molecules onto the CHCFe and intraparticle diffusion of the dye molecules through the inner pores. The CHCFe-Congo red system was spontaneous, feasible and exothermic. The value ΔS° for the adsorption process is negative-an indication that a decrease in the randomness at Congo red www.nature.com/scientificreports/ solution-CHCFe interface manifested during adsorption process. The maximum Congo red uptake by CHCFe as suggested by Langmuir model is 1649.3 mg g -1 and this occurred at 50 °C. The adsorbent (CHCFe) could be deployed for treatment of wastewater contaminated with dyes, especially Congo red.

Data availability
All data generated or analysed during this research are included in this published article and its supplementary information files.