Experimentally and theoretically approaches for Congo red dye adsorption on novel kaolinite-alga nano-composite

ABSTRACT A comprehensive study combining experimental, computational, and field experiments was conducted to find out the most suitable catalysis method to assist industries using Congo red dye to get rid of this waste from industrial wastewater in Beni-Suef area. The adsorption potential of kaolinite, Liagora farinose (Egyptian marine macroalgae) and kaolinite modified by Liagora farinose macroalgae assessed for the removal of Congo red dye from aqueous solutions. The kaolinite/alga nano-composite with a crystal size of 40 nm was fabricated using a wet impregnation technique. Our results indicate that surface modification of kaolin with Liagora farinose results in an obvious increase in adsorption of toxic dye for nano-composite than individuals. Batch experiments were applied and both kinetics and isotherms of Congo red dye adsorption were also explored in order to find out the influence of different experimental factors. Congo red removal percentage is highly affected by adsorbent dose, working temperature, and pH value. The best temperature for Congo red adsorption onto kaolinite/alga nano-composite is 40°C at pH > 7. The maximum adsorption capacities were found to be 5.0, 7.0 and 10 mg/g for kaolinite, alga and kaolinite/alga nano-composite, respectively. Computational simulations studies have shown that the adsorption of the Congo red molecule on Kaolinite surfaces is exothermic, energetically favourable and spontaneous. Congo red adsorption on kaolinite/alga nano-composite is well handled with the first-order diffusion model, while kaolin and Liagora farinose follow two different kinetic adsorption models depending on the Congo red dye concentration. Finally, the field tests showed optimistic results with nearly 94% efficiency for kaolinite/alga nano-composite in removing mixed dyes from industrial wastewater, which in turn verified the foundation of new eco-friendly nano-adsorbents to help reuse industrial wastewater.


Introduction
Water is considered an essential source for the subsistence of life on earth.Industrialisation and innovation have been participated in a bad way to contaminate clean water resources [1][2][3].Many industrial sectors focused on the utilisation of synthetic dyes in their industrial processes which resulted in huge effluents of wastewater to the environment daily [4][5][6].These toxic effluents containing dyes that are reported as harmful organic materials of low biodegradability and it's a major reason in some environmental problems such as aesthetic pollution, eutrophication and perturbations of the aquatic system [7][8][9].The first contaminant to be recognised in wastewater is colour even if it is found in very small amounts as it inhibits re-oxygenation in water.Moreover, it also inhibits sunlight penetration and hence disrupts the biological activity of aquatic organisms.Besides, highly hazardous effects on the living systems due to the toxic, carcinogenic, mutagenic and allergic nature of dyes, the discharge of dyes containing effluents in natural waters [10].For example, serious eye and skin irritation in a few minutes if a person is exposed to Congo red (CR) dye.In addition to stomach irritation, nausea, vomiting, and diarrhoea caused by the ingestion of Congo red [2,11].Benzidine, a carcinogenic product obtained during metabolisation of CR that can cause definite allergic reactions.Although CR, a human carcinogen, has been banned in many countries due to health hazards, but it is still widely consumed in several countries [10,12].Many technologies have been developed for the removal of dyes from wastewaters in a wide range to decrease their environmental effect [13][14][15].Physical, chemical or biological techniques are widely used.Adsorption, advanced oxidation, filtration, coagulation, flocculation, and microbial degradation have been applied to remove dyes from wastewater [16][17][18][19][20][21][22].The most effective and convenient method is the adsorption process, for some reasons it's a very simple process to use and it can remove the pollutants at very low concentrations [8,17,[23][24][25].Naturally occurring clay materials such as bentonite, fly ash, kaolinite and also algae are preferred as adsorbents they are considered a very good adsorbent because of their high cation exchange capacity, very abundant, available with low cost and high surface area [25][26][27][28].Also the presence of active functional groups that inspire contaminant attached to the wall of the biomaterial these functional groups may be carboxylic, hydroxyl, amino, carbonyl, phosphate or sulphonic [29,30].kaolinite is an important industrial raw material that has great applications such as paper coating and filling, ceramics, paints, cracking catalysts, cements, wastewater treatment, and pharmaceutical industries [31,32].Kaolinite, [Al 2 Si 2 O 5 (OH) 4 ], with stacking layer structure (1:1) at which each separate layer contains a sheet from Si 2 O 5 2− tetrahedrons and another from the alumina [Al 2 (OH) 4 ] 2+ octahedrons, whereas these sheets bond together through common oxygen atoms and keep their lamellar structures through the hydrogen bonds [33,34].The existence of OH function groups in kaolinite structure is the most important because they are involved in a wide variety of chemical reactions [35,36].In our work, a comprehensive study including computational, experimental, and field experiments is conducted to find out the most appropriate adsorbent system that effectively removes waste dyes especially Congo red dye from industrial wastewater.The adsorption performance of kaolinite (K), alga (LF) and kaolinite/alga nano-composite (KLF) was studied for Congo red dye removal from wastewater under different experimental conditions to explore the effects of addition of LF on K's adsorption capacity.However, such adsorbents are not novel at all, it has been firstly reported long time ago and many times later.The innovation of this paper focused on the effect of the introduction of natural algae on Kaolinite adsorbent performance.LF and K have several factors such as their natural abundance, low cost, cheap, reusability, and recyclability, which qualify them for dye removal.These factors make them more economical in the application and popularisation of this low-cost technology.Batch experiments were performed including the study of the effect of starting CR concentrations, reaction times, nano adsorbent doses, reaction temperatures, and pH values on CR dye removal %.Adsorption isotherms and kinetics are also studied.

Raw materials, dyes, and reagents
Kaolinite ore was supplied from El-Nassr company for mining and used without any further modification.LF macroalga was collected from the inter-tidal area of Egyptian Red Sea shores at the distance between Quoseir and Marsa-Alam cities. Congo Red dye was purchased from Sigma Aldritch and dissolved in distilled water.Sodium hydroxide granules with 99.99% purity and 36% hydrochloric acid were supported by Sigma Aldrich and used for pH adjustment.

Preparation of kaolinite/alga (KLF) nano-composite
The wet impregnation technique was selected to be technique of choice for the fabrication of kaolinite/alga nano-composite (KLF) [37,38].Kaolinite/algae nano-composite was prepared in several steps, starting by mixing, 1 gm of kaolinite, 1 gm of algae and 20 ml of deionised (DI) water followed by magnetically stirred at 500 rpm for 60 min then ultrasonication for 60 min and repeated for 3 times after that, the resultant kaolinite/algae nano-composite was subjected to filtration, washing by DI water for several times, and finally dried using a vacuum oven at 60°C during 24 h.The K, LF, and KLF nano-composites were characterised using an X-ray diffractometer (XRD), Scanning electron microscope (SEM) and Fourier transformer-infrared (FTI-R) spectrometer.The pH at zero point charge (pHzpc) and the effect of pH values on the zeta potential were followed up using Zetasizer, Malvern Panalytical, Nanoseries, zs90, UK.

Preparation of adsorbate
The regular and well-known anionic dye, Congo red (CR), was chosen as the adsorbate in this study.CR dye is a sodium-based salt of  S1 (Supplementary data) illustrates the structure of the CR dye.A 1000 mg/l solution in stock was prepared by dissolving an adequate amount (1000 mg) of CR in 1000 ml of DI water.The freshly prepared stock was diluted by DI water to obtain the required concentrations of the working solutions.The pH of all prepared solutions was tuned to 3, 5, 7 and10 by using either 0.1 M solution of HCl or NaOH.

Samples characterisations
The XRD characterisation was carried out by PANalytical diffractometer (Empyrean) using CuKα of wavelength λα = 0.154045 nm and operating at 40 kV,35 mA with scan step of 0.02°within the range 20-70°.The average crystallite size, D s , of the prepared nanoparticles was acquired by Scherer formula, D s = 0.94 λα/β w cosɸ; where β w and ɸ are the corrected full width at half maximum and the diffraction angle [39].SEM micrographs were measured using Quanta FEG 250 microscope (Switzerland).FT-IR spectra were measured using a Bruker VERTEX 70 FT-IR spectrophotometer using the dry KBr pellet technique.

Adsorption studies
Batch mode experiments were conducted for all CR adsorption experiments in various conditions including initial CR concentrations (5-25 mg /l), contact time (up to480 min), adsorbent dosage (10-50 mg), pH (3)(4)(5)(6)(7)(8)(9)(10), and temperature (25-90°C) with continuous shaking.Four adsorption experiments series were implemented on K, LF, and KLF adsorbents at diverse adsorption circumstances including, initial dye concentration, the temperature of adsorption, adsorbents dose, and initial pH of the solution as displayed in Table S1 (Supplementary data).The experiment time was set at 480 minutes with 25 ml solution volumes in all experiments.The variation in the CR concentration was elucidated from the absorption peak measured by UV/Vis spectrophotometer.The limits of detection and quantification (LOD/LOQ) for the used instrument were found to be 0.0066 and 0.02 mg /L, respectively.The reusability tests of both K, LF, and KLF adsorbents were examined 5 times using 0.02 g of all adsorbents, 25 ml of 10 mg/L initial concentration of CR for 480 minutes' contact time at 25 °C and pH 7. The three adsorbents K, LF, and KLF were collected from the solution after each run, then cleaned from dye residues by distilled water and set for the next run.The quantity of CR uptake by the synthesised nanocomposite at equilibrium (q e (mg/g) and time t (q t ) and CR dye removal% have been obtained utilising equations 1 and 2, respectively [40,41]: At which C o , C t , and C e are the concentrations of CR in mg/l at the start, after time t, and at equilibrium, respectively.V is the CR volume in mL and m is the K, AS, and KAS mass in mg.

Adsorption isotherms
Langmuir, Freundlich, and Tempkin isotherms have been used to explain the reaction isotherm of the fabricated K, LF, and KLF nanocomposites for the tested CR [42][43][44][45].All linear isotherms equations and their parameters are explained in supplementary data.
Tendency and favourability to Langmuir isotherm for the equilibrium data could be predicted from the value of the dimensionless separation factor(R L ) based on equation 3 [46].
Where C max represents the maximum initial CR concentration.

Adsorption kinetics and mechanism
Different adsorption mechanisms and kinetics models such as intra-particle diffusion, pseudo-first-order, pseudo-second-order and simple Elovich kinetic model are used for identifying the adsorption mechanisms and kinetics models that best match with the adsorption of CR onto K, LF, and KLF adsorbents [30,[47][48][49].All linear kinetics equations and their parameters are explained in supplementary data.

Statistical analyses
All adsorption results were measured in triplicates and the average values are presented.The values of regression coefficients (R 2 ) for the different kinetic and isotherm models were obtained using the statistical functions of Origin Pro 2016.

Computational calculations
The kaolinite and Congo red structures were optimised by density functional theory (DFT) using the GGA-PBE (Generalised Gradient Approximation-Perdew Burke Ernzerhof) functional.The double numerical polarised (DNP) basis set was assigned.No spin-polarisation effects were included in the exchange-correlation functional.the core electrons of kaolinite and Congo red structures were treated with the effective core potential and all electrons, respectively.The calculations were performed by DMol3 module [50,51] in the Biovia Materials Studio.The energy of the bulk structure of the kaolinite unit cell was minimised then the kaolinite unit cell was cleaved with (001) and (002) planes.We constructed three supercells (4x4x1), (6x6x1), and (8x8x1) for each plan with the vacuum thickness of 20 A° and optimised them at all the previous conditions.To identify the effects of different sizes of kaolinite surfaces on the adsorption energy and find the desorption sites of Congo red on the kaolinite surface, a Monte Carlo (MC) simulation was performed.MC simulation was carried out by the Adsorption Locator module in the Materials Studio using The COMPASS force field (Condensed-phase Optimised Molecular Potentials for Atomistic Simulation Studies) as a force field and use current in the charges section.The basic principles of MC simulation have been described by Frenkel and Smit [52].

Field experiments
The newly synthesised adsorbents system was tested as an effective eco-friendly adsorbent that could be applied on a large scale to remove industrial waste dye from industrial wastewater.For this purpose, wastewater containing waste dye was supported by clothes dyeing plant in Beni-Suef city, and the wastewater containing waste dye was used as it is without any further treatment or dilution.The appropriate adsorbent system was selected depending on our adjusted computational and experimental results.

FT-IR analyses
FT-IR characteristic peaks for K, LF, and KLF adsorbents are displayed in Figure 2(a).The FT-IR spectrum of K, Figure 2(a) (red colour) illustrated a wide mode of OH groups.The modes at 3691and 3621 cm −1 refer to the inner OH stretching [53][54][55].The peaks at 1109 and 1023 cm −1 are related to Si-O vibration modes [56].The peaks at 469, 543, and 919 cm −1 are related to the Si-O-Si bending, Si-O-Al, and octahedral aluminium (Al-OH) [57].All peaks in the region from 400 to 800 cm −1 are related to the metal oxides [58].
Peaks related to LF algae, Figure 2(a) (blue colour), exhibit well known characteristic band at 3410 cm −1 for hydroxyl function (-OH) of phenolic groups.The peak at 2930 cm −1 was allocated to the stretching mode of alkyl groups -CH, whereas the mode at 1618 cm −1 was corresponding to -C = O.The band at 1477 cm −1 was attributed to the C-H vibration [59,60].The bands located around 1110, 1120, and 1140 cm −1 are attributed to the C-O bond or may refer to the sulphate group [61].The bands at 3300-3500 cm −1 and 2500-3000 cm −1 are referred to as amine N -H stretching and carboxylic acid O -H stretching, respectively [62].Finally, FT-IR peaks of the newly synthesised adsorbent KLF are shown in Figure 2(a) (green colour).The presence of the characteristic peaks representing the two phases (kaolin and algae) confirms the presence of a new compound.The disappearance of some characteristic peaks especially those representing the amino group in the algae, which confirms the interlocking that happened between algae and the pores and surface molecules of kaolin.Not only peaks disappearance was noticed, but also the peak shift happened for characteristic peaks for both kaolin and algae.Both the peak shift and peak disappearance come in line with data obtained from other characterisation techniques which confirms the formation of a new compound.Table S2, supplementary data, lists the positions of the characteristic FT-IR bands for K, AS, and KAS adsorbents.

XRD characterisation
The XRD charts of K, LF and KLF are presented in Figure 2(b).The main XRD peaks of kaolinite minerals were observed at 2 theta 12.44° and 24.9°due to the crystallographic growth alongside the (001)and (002) planes [63,64] [59,65].The XRD chart of KLF displays a small shifting in the position of the main XRD peaks of kaolinite to shift to 24.98° at 26.2°, 26.9° and 45.8°.The average calculated crystallite size using the Scherer equation was 40.3 nm, which confirms the nanoscale nature of the newly synthesised composite.

Effect of initial dye concentration
The variations in the removal % and the amount of CR adsorbed with time using K, LF, and KLF nano-adsorbents at different initial CR concentrations are shown in Figure 3(a-c, d-f), respectively.It can be observed from these Figures that; during the first stage of the adsorption process, the adsorption capacity and the dye removal % were elevated at the beginning of the process and then, their increasing rates are reduced to reach the equilibrium state at the end.It was observed also that; contact time has no marked effect on the adsorption process using new sorbents after reaching the equilibrium.The prompt removal rates at the early stage of the reaction are allocated to the existence of a huge surface density of uncovered active spots on the nano adsorbent's surfaces.By increasing the period of contact between adsorbent and adsorbate, the hot spots converted to fully occupied sites by CR molecules.As a result, repulsion forces are established between the adsorbed CR molecules on the surface of adsorbents and CR molecules in the bulk liquid phase [41].
The clay nano-composite, KLF, revealed higher efficiency for Congo red adsorption at lower concentrations in comparison with the other adsorbents K and LF.At 5 and 10 ppm initial dye concentration; the CR removal% reached 98% and 90% for KLF, 49% and 43% for LF; and 37.7% and 28.7% for K, respectively.This behaviour matches with previously reported composites but with lower performance than ours does [32,66,67].With increasing concentration to 15ppm, the dye removal % reaches 77% and 36% for both KLF and LF,and 23.3% for K adsorbent.At relatively high concentrations, 20 and 25 ppm, the dye removal % was in the order KLF>LF>K.
The quantities of adsorbed CR are increased by raising the starting CR concentration as shown in Figure 3(d-f).This could be accredited to the growth of the concentration gradient with raising the initial CR concentration.Hence, appropriate growth in the draft forces occurs to overawed the mass transfer resistance between CR adsorbate and K, LF, and KLF adsorbents [68,69].The maximum adsorption capacities of K were found to be 1.88, 2.8, 3.5, 4.4, and 5.0 (mg/g), while the adsorption capacities of LF were found to be 2.45, 4.3, 5.4, 6.0, and 7.0 (mg/g) for CR with initial concentrations of 5, 10, 15, 20 and 25 mg/l respectively, at pH 7 and 25°C.Whereas, the maximum adsorption capacities of KLF adsorbent were found to be 4.9, 9.0, 11.55, 11.0 and 10 (mg/g) at these starting concentrations.The results showed that the modification of K with LF is a feasible approach to enhance the CR removal performance by KLF.

Influence of nano-adsorbent dosage
To determine the adsorption cost, the influence of the nano-adsorbent dose on the CR removal% was assessed for determining the optimised nano-adsorbent dosage that offers the maximum performance.This is graphically depicted in Figure 4(a).The adsorbent doses were varied from 0.01 to 0.05 g.It was found that 0.02 g nano-adsorbent per 20 ml of CR solution of an initial concentration of 10 mg/l was the best nano-adsorbent dosage.
From Figure 4(a), The CR removal% for all adsorbents rises as the nano-adsorbent dosage increased from 0.01 to 0.02 g.The removal% increases from 26% to 28.7% in the case of K adsorbent, from 27% to 43% in case of LF adsorbent and from 79% to 91% in case of KLF adsorbent.Which refers to the increase in the number of active dots with increasing the nano-adsorbent dosage [41].For nano-adsorbent dosage over 0.02 g, the removal% decreases again.For K adsorbent, the CR removal% decreases to 25, 22, and 15% with increasing the adsorbent dose to 0.03, 0.04, and 0.05 g, respectively.For LF adsorbent, the CR removal% decreases to 35, 18, and 10% with increasing the adsorbent dose to 0.03, 0.04, and 0.05 g, respectively.Also, the CR removal% decreases to 70, 37, and 36% with increasing the adsorbent dose to 0.03, 0.04, and 0.05 g, respectively, in the case of KLF adsorbent.This phenomenon may be attributed to the screening effect that occurs at elevated nano-adsorbent dosage due to the accumulation of the nano-adsorbent particles and decreasing the distance between the nano-adsorbent particles.The condensed layer at the surface of the adsorbent blocks the binding sites from CR molecules.Also; K, LF, and KLF overlapping resulted in a competition between CR molecules for restricted available binding sites.Aggregation or agglomeration at greater, K, LF, and KLF doses increases the diffusion path length for CR adsorption causing a decrease in adsorption % [30,[70][71][72].

Effect of pH
Due to its influence on the dissociation/ionisation of the K, LF, and KLF nano-adsorbents and their impact on the absorbent surfaces, the starting pH value of the CR solution can be a crucial factor in controlling the nano-adsorbent performance [72].So, the electrostatic charges on the K, LF, and KLF adsorbents and the CR sorbate are greatly affected by the pH of the solution.The effect of pH on the CR removal efficiency of the adsorbent was studied between pH 2 and pH 10 as shown in Figure 4(b) at initial CR concentration of 10 mg/l, sorbent dosage of 0.02 g.The K adsorbent shows removal percentages of 16.6%, 17.0%, 28.7%, and 33% for CR solutions of pH 2, 5, 7, and 10, in that order.The LF adsorbent shows removal percentages 62%, 34.5%, 43% and 36% while KLF adsorbent shows removal % of 59%, 37%, 90% and 90% at pH values of 2, 5, 7, and 10, respectively, at the same previously mentioned conditions.As such, a significant role for pH in controlling the surface charge of adsorbents was apparent.To investigate the effects on K, LF and KLF, we determined the zeta potential of composites in the solution.The effect of pH on the zeta potential of K, LF and KLF in an aqueous solution shown in Figure 4(e).The surface charge of K, LF and KLF shifted from higher to lower values with increasing the pH value from 2 to 5, which resulted in a gradual decrease in electrostatic attraction between the composites and the negatively charged CR.Such a decrease should lower adsorption capacity.A larger fluctuation of adsorption capacity was observed at a pH ranging from 5 to 7 to reflect a large change in zeta potential.Generally, the more the shift in the zeta potential value to the positive values the more the removal %.The lower zeta potential indicates that the adsorbents surface were partially negatively charged at a pH of 2 to 5 and that the electrostatic force between K, LF and KLF and CR through the sulphonic acid group (SO 3

−
) was mainly repulsive during the experiment [73,74].The pH at which the adsorbent has zero-point charge (pHzpc) was 5.8 in case of KLF adsorbent and above this pH the adsorbent surface became positively charged and consequently a large increase in the adsorption capacity take place on KLF surface.The pHzpc values were not detected in case of K and KLF under the investigated pH ranges.The gradual increase the adsorption capacity at pH 7 and 10 in case of K adsorbent could be related to the shift on zeta potential to less negative values.

Effect of temperature
The influence of reaction temperature on the uptake % of CR onto K, LF, and KLF was done at different adsorption temperatures degrees.The adsorption tests were done at 25, 40, 50, 60, 80, and 90°C and the results were presented in Figure 4(c).For both LF and KLF adsorbents, the CR removal % increase from 73 to 100% and from 92 to 100%, respectively, with increasing temperature from 25 to 40°C.This performance could be owed to the growth of the CR diffusion rates with rising temperature as a result of the decrease in viscosity of the solution [75].With more temperature rise, the CR removal% remains constant at 100% till 60°C for LF and KLF adsorbents.This could be attributed to the fact that the maximum limit of CR adsorption is reached by the nano-adsorbent.With more and more increase in temperature, the dye removal% drops down again, and this is owed to the collapse of adsorption force responsible for CR dye molecule adsorption on the LF and KLF surface.This may ascribe to active site damage and adsorptive force reduction between nano-adsorbent's active site and CR molecule [8,76].Therefore, the optimum temperature for adsorption of CR is from 40 to 60°C for LF and KLF adsorbents.For K adsorbent, CR elimination % increase with rising adsorption temperature from 25 to 40°C, where it increased from 28.7 to 53.3% with changing temperature from 25 to 40°C.The CR elimination % reduced from 53.8 to 33% by increasing temperature from 40 to 60°C.By increasing the temperatures from 25 to 40°C in the initial stage, the CR removal% increases due to the growth in the CR diffusion rate.By raising temperatures from 40 to 70°C, a decrease in the CR elimination% occurs and this could be attributed to the desorption of CR molecules caused by the destruction of adsorptive forces responsible for the CR dye adsorption on the K nano-adsorbent surfaces [77].With further increase in temperature from 60 to 90°C, the removal% increase from 33% to 36%.The best temperature for CR adsorption onto K is 40°C.

Reusability of adsorbents
The K, LF, and KLF reusability for the elimination of CR was performed four 4 times with the same adsorbent and the same adsorbent dosage (Figure 4(d)).The results showed that; the removal strength of all used adsorbent greatly varied throughout the four adsorption cycles.For K adsorbent, the recorded dye removal % was 28.7%@1 st cycle, 12.5%@2 nd cycle, 10%@3 rd cycle, and ~10%@4 th cycle.For LF adsorbent, the dye removal % was decreased from 73%@1 st cycle to 15%@4 th cycle.For KLF adsorbent, a decrease in the calculated CR removal% occurs where it changes from 92%@1 st cycleto 20%@4 th cycle.
The reduction on the CR removal% could be ascribed to the agglomeration of the CR molecules onto the surface of K, LF, and KLF adsorbents, which consequently blocks adsorbent surface and pores from the dissolved CR molecules and so, a reduction in adsorption capacity take place [78].

Linear regression analysis
The statistical significance of R 2 (the correlation coefficient) for the linear fitting of Ce/ qe versus Ce, log(q e )versus log(C e ), and q e versus Ln(C e )was the criteria by which the data fitted to Langmuir isotherm, Freundlich isotherm, and Tempkin isotherm, correspondingly.From the linear plots, the values of Q o , K L , K F , 1/n, K T , B, and R 2 were determined from Figure 5(a-c) and recorded in Table 1.The results in Table 1 demonstrate that CR adsorption on K and KLF adsorbents track the Langmuir isotherm models where the R 2 value is the highest; the adsorption process almost tracks the Langmuir isothermal model [79][80][81].Therefore, the elimination of the dye occurs at the active sites of the nanoadsorbents on a single surface layer, and the adsorbed CR molecules do not interact with each other.At 25°C, the obtained R 2 values by the Langmuir isotherms were 0.9911and 0.9893 for K and KLF adsorbents, respectively.The value of R L is < 1, which means that the CR adsorption is preferred in the study case [82].CR adsorption on LF adsorbents tracks the Temkin isotherm models where the R 2 value is the highest.

Nonlinear regression analysis
Redlich and Peterson proposed a nonlinear empirical isotherm model [83].Adsorption that does not follow perfect monolayer adsorption can be described by this hybrid mechanism.The model combines elements of the Freundlich and Langmuir models and can be used to describe sorption equilibrium over a wide range of adsorbate concentrations.The nonlinear form this empirical model is where C e is the adsorbate concentration (mg/L) in solution at equilibrium, P 1 (L/g), P 2 = 1 and P 3 (mg/L) −g are the Redlich-Peterson constants and g is an exponent with a value between 0 and 1.This model becomes a linear isotherm when g = 0, reduces to the Langmuir isotherm when g = 1, and converts into the Freundlich isotherm when P 1 , P 3 >>> 1 and g ≤ 1.The ratio of P 1 /P 3 indicates the adsorption capacity [84,85].
The parameters of the sorption isotherm model may be easily determined using the nonlinear regression approach and OriginPro 2018 by minimising the sum of the square differences between experimental data and model outputs.The theoretical q e versus C e values are then calculated using initial estimations of the unknown parameters in the model equation, and the residual sum of squares (RSS) between the experimental data and the theoretical model output is obtained.Following that, iterations are done in which the initial estimated parameter values are changed by a tiny amount and the RSS is recalculated numerous times until the parameter values result in the lowest feasible RSS value.As shown in Table 1 and Figure 5(d), nonlinear regression provides a more appropriate and precise determination of model parameters than linear regression [86].For K nanoadsorbent, P 1 and P 3 >>> 1, g < 1, and P 1 /P 3 = 0.99.For LF nanoadsorbent, P 1 > 1 and P 3 < 1, g < 1, and P 1 /P 3 = 6.46.For KLF, the P 1 /P 3 ratio is enhanced to 9.55, indicating an increase in adsorption capability.Also, because the g value for KLF is so close to unity, the nonlinear Redlich-Peterson model approaches the Langmuir model, which correlates well with the linear regression data in Table 1.

Error function analysis
An error function assessment is frequently necessary to evaluate the applicability of a model equation to experimental results.Error functions are statistical equations that are used to calculate the difference between theoretically expected data and actual experimental data values.The best fitting model was validated by three different statistical error functions namely coefficient of determination (R 2 ), reduced Chi-square(χ 2 ) test, and the reduced sum of square error (SSE).The best-fitting model is the one with the lowest value of SSE and χ 2 (close to zero) and the highest value of R 2 (close to unity).The values of these error functions are presented in Table 1.The obtained R 2 values by the Redlich-Peterson model were 0.9928, 0.9902, and 0.9981 for K, LF, and KLF adsorbents, respectively.Also, χ 2 values are very close to zero (0.02, 0.05, 0.03), this implies that our experimental data are more fit with the nonlinear regression.

Adsorption kinetic models
To investigate the most appropriate adsorption kinetics model, the adsorption of CR on K, LF, and KLF under various starting CR concentrations was addressed.The first-order, second-order, and Elovich kinetics linear graphs were represented in Figure 6 by plotting ln(q e -q t ) versus t, t q t versus t, and q t versus ln(t), respectively.The adsorption kinetics parameters k 1 , k 2 , q e , β, and α of the evaluation model in addition to R 2 were obtained using the linear plots and depicted in Table 2.The linear fit and regression coefficient values in Table 2 for all the studied kinetic models confirmed that CR adsorption onto K is well handled with the second-order diffusion model till 20 ppm dye concentration and over this concentration, the adsorption process follows the first-order rate law and this was also confirmed from the good approximation between the calculated qe and experimental qe [87,88].The CR adsorption onto KLF is well handled with the Elovich diffusion model and this appears in the higher R 2 values [87].The CR adsorption onto LF follows two different kinetics adsorption models depending on the CR concentration where it follows the Second order kinetics up to 15 ppm concentration and above this concentration, it follows the Elovich diffusion model.

Sorption mechanisms
To comprehend the adsorption kinetics process and rate-controlling steps, the practical data are fitted for Weber's Intra-particle diffusion.A straight line in the chart of q t versus t 1/ 2 , Figure S2 (Supplementary data), proposes the applicability of the Intra-particle diffusion model.The values of K 3 and I, Table 3, are obtained using the slope and intercept of the linear fitting, respectively.The intercept I ≠ zero, demonstrating that the Intra-particle diffusion model may not be the sole rate-controlling route in estimating the adsorption reaction kinetics [89].The intercept in Figure S2 refers to the boundary layer effect.The larger intercept, the greater the contribution of surface adsorption in the rate control stage [89].

MC simulation
The lowest configurations obtained due to the adsorption of Congo red on (001) and (002) kaolinite for different three supercells are summarised in Figure 7. MC simulation aims to elucidate the influence of diverse planes and sizes of kaolinite on the adsorption of CR.The adsorption energies of each kaolinite-Congo red system are titled in Table 4. Figure 7(a-c) demonstrates the CR adsorption on the kaolinite (001) plane in a dry system without any solvent.The CR molecule holds various hydrogen bond (HB) donor/acceptor spots, and hence, it forms a number of HBs with the nitrogen and oxygen atoms of the kaolinite superficial.The oxygen and the nitrogen atoms of CR form HBs and intra-molecular HBs through the hydroxyl hydrogen atoms of the kaolinite superficial.Figure 7(d-f) displayed the formation of HBs and intra-molecular HBs between CR and the hydroxyl hydrogen atoms of kaolinite alongside the (002) plane.The adsorption energy, ΔE ads , interaction energy, E int , deformation energy (E def ), and the substrate/adsorbate configurations (dE ads /dN i ), wherein one of the adsorbate constituents is missing, are reported in Table 4. ΔE ads for all configurations in this study are negative.I.e. the adsorption reactions of the CR molecules on Kaolinite surfaces are exothermic, energetically preferred and spontaneous, due to the existence of the intermolecular interactions.Also, increasing the Kaolinite surface size does not affect significantly the adsorption energies for all configurations.But the HBs and intramolecular HBs with the hydroxyl hydrogen atoms of the kaolinite (001) surface are weaker than those in (002).So, this may cause a decrease in ΔE ads values, whereas the adsorption energies for all configurations in the state of kaolinite (001) are larger than those in kaolinite (002) as shown in Table 4.The complete supercells for adsorption configurations of the adsorbed CR on kaolinite (001) and (002) facets are displayed in Figure S3 for clarity purpose.

Field experiments
The optimised parameters for the newly synthesised KLF adsorbent were 0.02 gm adsorbent mass, near room temperature, and contact time of 480 minutes, while the pH of the wastewater containing the waste dye remained unchanged.The presence of different wavelengths corresponding to different dyes was detected by the optical scanning of the as-received industrial wastewater.Absorbance at various wavelengths recorded at the end of the contact period to measure the removal efficiency of dyes from industrial wastewater.At various wavelengths, the data revealed removal efficiency reaches about 94%.

Comparison of adsorption capability of K, LF, and KLF with other adsorbents
The relation between the qm (adsorption capacity) values of the various adsorbents mentioned in the literature and that of K, LF, and KLF for CR dye adsorption is shown in Table 5.A comparison of qm values also showed that a fair adsorption potential of CR dye from aqueous solutions was demonstrated by K, LF, and KLF [10,14,32,[90][91][92][93].Our optimised composite showed qm grater then Kaolin-based and Liagora farinose adsorbents [10,14,32,[90][91][92][93].

Conclusion
The wet impregnation approach was used to produce a unique kaolinite/alga (KLF) nanocomposite.The performance of the KLF nano-composite as a nanoadsorbent for CR dye from aqueous solutions was investigated and compared to that of Kaolin (K) and Liagora Farinose (LF) alga.Various morphology and structure characterisation techniques for K, LF, and KLF indicated aggregation of LF nanoparticles with kaolinite nanopores to produce 40.3 nm nanocomposite crystallites.The adsorption tests revealed ~98%, 49%, and 37.7% removal efficiency of 5 ppm CR dye employing KLF, LF, and K, respectively.The highest adsorption capacities of K, LF, and KLF nanocomposite were 5.0, 7.0, and 10 mg/g, respectively, implying that the order of performance was KLF > LF > K for all CR concentrations.Furthermore; adsorbent dose, working temperature, and pH value all have a significant impact on CR removal percentage, with 40°C at pH > 7 being the best working temperature for CR adsorption onto KLF.The reusability tests for K, LF, and KLF revealed that none of the adsorbents were favoured for CR removal reuse, but the novel composite KLF demonstrated improved stability.The CR adsorption isotherms and kinetics on K, LF, and KLF indicate that the Langmuir isothermal models are monitored by K and KLF adsorbents, while LF is better suited to Tempkin isotherms.With the Elovich model, CR adsorption on KLF is well controlled, while K and LF adopt two separate models of kinetic adsorption depending on the concentration of CR.Increasing the surface size of Kaolinite also does not greatly affect the energy of adsorption for all configurations.Finally, field tests showed surprising results of 94% dye removal efficiency from industrial wastewater, which in turn confirms the foundation of a modern eco-friendly adsorbent device that could assist in the reuse of industrial wastewater.Future research should focus on improving the stability of the developed nanocomposite by incorporating plasmonic or metal oxide nanostructures.

Figure 1
Figure 1 illustrated SEM images of K, LF, and KLF adsorbents.For kaolinite, Figure 1(a)shows that agglomerated rounded regular shape particles, rough surface, different particle sizes, and porous cavities on the surface.The SEM image of LF, Figure1(b), revealed that LF exhibits a less porous surface, which consequently affects the surface area for LF which in turn affects its adsorption capacity.When kaolinite is treated with algae LF the SEM image of the nano-composite, Figure1(c), showed covered pores in the kaolinite surface with the LF particles and converted into agglomerated particles.The formation of KLF nano-composite could be established from the noted changes in the morphology of the nanocomposite relative to that observed for K and LF.

Figure 3 .
Figure 3.Effect of CR dye concentration and contact time on the removal% and the amount of CR dye adsorbed at 25°C and pH 7 by 20 mg of (a, d) K, (b, e) LF, and (c, f) KLF.

Figure 4 .
Figure 4. Effect of (a) adsorbent weight, (b) Initial pH of the solution, (c) adsorption temperature, and (d) reusability test on the removal% of 20 ml CR solution of 10 mg/l by K, LF and KLF (e) effect of pH on zeta potential.

Figure 5 .
Figure 5. Plots of (a) Langmuir, (b) Freundlich and (c) Temkin adsorption isotherms for the adsorption of CR dye by 50 mg of K, LF and KLF at 25°C and pH 7; and (d) nonlinear regression fitting for Redlich-Peterson isotherm.

Figure 7 .
Figure 7. Snapshots for the adsorption configurations of Congo red adsorbed on (a-c) kaolinite (001) facet and (d-f) k.

Table 1 .
Isotherm parameters for CR adsorption on K,LF and KLF.

Table 2 .
Parameters of the kinetic models for CR dye adsorption on K, LF and KLF.

Table 3 .
Intra-particle diffusion constant's at different initial CR concentrations at 25 O C.

Table 4 .
Adsorption energies for the adsorption configurations of Congo red adsorbed on kaolinite (001) and (002) facets.

Table 5 .
Comparison of the optimised conditions, removal%, and adsorption capacity of different CR adsorbents relative to our K, UL, and KUL nanoadsorbents.