Dauda Mohammed *,Muhammad H.Al-Malack ,Basheer Chanbasha

1 Department of Civil and Environmental Engineering,King Fahd University of Petroleum &Minerals,Dhahran 31261,Saudi Arabia

2 Department of Chemistry,King Fahd University of Petroleum &Minerals,Dhahran 31262,Saudi Arabia

Keywords:Functional groups Optimization Zwitterion Adsorption Waste treatment Industrial waste

ABSTRACT Surface functionalization of blast furnace slag with sulfamic acid (a zwitterion) was performed for the removal of Cr3+ and methylene blue dye (MB) from water samples.The slag functionalization process was optimized using Response Surface Methodology Design.Statistical analysis of the parameters that include the sulfamic acid amount (A),reaction time (B),and temperature (C) revealed that (A) increase had a negative effect on the adsorption of both pollutants by the zwitterion slag,whereas (B) and (C)increase presented a positive impact.At the optimum condition of 2 g sulfamic acid amount,50 min reaction time and 37 °C temperature,the prepared slag showed a removal efficiency of more than 90% for both Cr3+ and MB.Surface characterization by SEM/EDS,FTIR,XPS and surface area analyser,showed an improvement in surface properties and the incorporation of zwitterionic NH2+ and S=O groups of sulfamic acid.Adsorption isotherm and kinetic studies conducted with the zwitterion slag showed the adsorption process was suited to Freundlich isotherm model and pseudo-second-order kinetic model.The thermodynamic study conducted revealed the spontaneity of the process based on the calculated negative ΔG (Gibb’s free energy) values.The prepared zwitterion slag offered easy regeneration with dilute HCl solution and showed a considerable removal(Cr3+:65%and MB:80%)for both pollutants even after 3 cycles of usage.

1.Introduction

Water contamination with toxic and recalcitrant compounds generated from industrial processes are a potential source of human health and environmental hazards.Effluents from industrial activities are often contain a high level of toxic heavy metals and organics such as dyes,surfactants,and phenol generated as a by-product of the production process[1,2].For example,in the tannery industry,chromium(Cr)and organic azo dyes are the primary contaminants of interest generated in the effluent due to their extensive use during leather tanning operations[3,4].Reports have shown that these contaminants are both toxic and capable of inducing mutagenic and carcinogenic responses in humans [5].

Considering the impact of these toxic contaminants on human health and their effect on receiving water bodies,several environmental organizations have imposed stringent regulatory limits and guidelines for them in water and wastewater treatments [6,7].Based on these regulatory constraints,several treatment approaches are currently in use for the removal of aqueous organic dyes and Cr compounds,namely,coagulation/flocculation,adsorption,chemical precipitation,oxidation/reduction processes,membrane processes,and ion exchange.

Recently,adsorption treatment using low-cost or waste materials for the removal of varieties of organic and inorganic contaminants have attracted significant research interest.This interest stem from the numerous benefits it offers towards both waste management and environmental pollutants remediation [8–10].The use of blast furnace slag(BFS),an industrial steel waste material for Cr removal from water sample has been reported [11,12].Raw slag application shows it could offer reduction,precipitation,and adsorption of Cr compounds from aqueous solutions.Studies attributed these slag properties to be due to its mineral compositions made up of iron oxide,calcium oxide,aluminum oxide,and magnesium oxide [13–15].

Despite its exceptional results for the precipitation of metals from aqueous phase due to its alkali constituents (CaO,Al2O3,and MgO),raw BSF tends to offer limited adsorption capacity for organic contaminants owing to its poor surface properties and lack of appropriate surface functional groups.This limitation can be overcome by introducing reactive groups on its surface via a proper functionalization procedure.

In this study,for the first time,we functionalized BSF with sulfamic acid(NH2SO3H,relatively stable inorganic acid),and utilized it as an adsorbent for aqueous removal of Cr and methylene blue dye(MB).Generally,sulfamic acid possesses a unique zwitterionic property that results from its oppositely charged aminoand sulfonicfunctional groups.With the aid of the zwitterionic group,sulfamic acid can interact with several ionic components either through the sulfonic group or amino group [16].Hence,in this study,it is speculated that the loaded zwitterionic group on the BFS surface would aid in the uptake of the pollutants via electrostatic interaction or complexation reaction.

Previous studies have loaded this acid on adsorbents to enhance their surface properties [17,18].For example,El-Hakam et al.[17]were able to synthesis chromium metalorganic frameworks(MOFs) modified with sulfamic acid via the single-step impregnation method.The modified MOFs with 55%(mass).sulfamic acid offered significant removal of methyl orange dye from the aqueous samples.However,no study has reported the functionalization of BFS with sulfamic acid,and it uses for metals or organic contaminants removal from water samples.

In this study,chromium and MB were selected as a model inorganic and organic contaminant to evaluate the removal efficiency of the zwitterion functionalized BFS (Z-BFS).Also,the influence of the surface functionalization parameters on the performance of the Z-BFS was evaluated and optimized using the Box-Behnken design of response surface methodology (RSM).

2.Materials and Methods

2.1.Raw materials and chemicals

The blast furnace slag utilized in this study was obtained from an iron processing plant in the Eastern Province of Saudi Arabia.High purity analytical grade (>99%) chemical reagents were used throughout the study and were purchased from Merck(Darmstadt,Germany).Cr3+stock solution (1 g∙L-1) was prepared from chromium (III) nitrate nonahydrate (Cr (NO3)3∙9H2O),and methylene blue stock solution(1 g∙L-1)was prepared from its organic chloride salt(C16H24ClN3O3S).Ultrapure water was obtained from a Milli-Qsystem(Millipore,Milford,MA,USA).The ultrapure water used for the preparation of stock and working solutions.0.1 mol∙L-1solutions of NaOH and HCl were prepared and used for sample pH adjustment.Cr3+measurements were performed using Atomic Absorption Spectroscopy (AAS) (Elmer Perkin,USA),while MB measurements were conducted using UV–Vis Spectrophotometer(Shimadzu,Japan)

2.2.Surface modification of slag

A facile wet impregnation method was adopted for the preparation of the surface-modified slag.The process involves the following steps;the collected slag was thoroughly washed with deionized water to remove any impurities on it and,subsequently,it was dried in an oven at 110 °C for 24 hrs.Afterwards,the dried slag was grounded and sieved through a sieve of pore size less than 1 mm.Furthermore,5 g of the sieved slag was weighed and premixed with a specific amount of the solid sulfamic acid (2 g,6 g,10 g).Then,a 50 mL volume of deionized water was added to give a mixture of the desired reactant ratios.The resulting mixture was sonicated for 15 minutes and then vigorously stirred on a magnetic stirrer with a temperature regulator to react for a specified temperature and contact time under the experimental design.Subsequently,the product was filtered and washed with deionized water until it showed a pH value in the range of 5–6 (deionized water pH value).Finally,the resulting product designated as ZBFS after the washing step was dried in an oven at 110 °C for 12 hours.

A statistical experimental design approach based on response surface methodology(RSM)was employed for optimizing the conditions of the surface modification process.Accordingly,a Box-Behnken design (BBD),which is a three-level factorial design,was selected in designing the sets of experiments for the modification process.The three independent factors,namely,reaction temperature,contact time,and sulfamic acid amount,were varied at three equally spaced levels.The three levels selected for each factor were coded as low(-1),medium(0),and high(+1),as given in Table T1(Supplementary Material).The effect of the selected factors and their interactions on the process was evaluated based on the two studied responses,namely,MB(Y1)and Cr3+(Y2)removal efficiencies by the modified slags.Based on the selected standard BBD(33)with three replicated centre points,a total number of 15 experimental runs were conducted for both responses.The creation of the experimental layout(design matrix),as well as the statistical analyses of the two responses obtained,were performed using a statistical software package(Design-Expert version 11).

2.3.Characterization of the raw and modified slag

Surface property characterization,as well as elemental composition of the raw and modified slag samples,were determined using different characterization techniques.The surface area and pore volume of the samples were determined by nitrogen adsorption/desorption technique using Micromeritics ASAP 2020,USA.The surface morphology was acquired using scanning electron microscopy coupled with energy dispersive spectroscopy (JEOL SEM-EDS,Japan).Functional groups of the samples were identified using Fourier Transform Infrared Spectroscopy (FTIR) analyser(NICOLET Thermo Scientific,USA).X-ray photoelectron spectroscopy analysis (Thermo scientific Escalab 250Xi spectrometer)was employed in characterizing the elemental composition of the raw slag and the loaded zwitterionic groups.The point of zero charges pH of the Z-BFS was acquired using Malvern Zetasizer(United Kingdom).

2.4.Adsorption experiments

Batch adsorption experiments were conducted for both Cr3+and MB uptake from aqueous phase to evaluate the performance of the raw and modified slag samples.For these experiments,test solutions of 20 mg∙L-1of both Cr3+and MB were prepared based on the experimental design presented in Table T2 (Supplementary Material).Also,the dosage of the modified slag,corresponding pH variation range,and the proposed contact time exploited in the experiments are shown in the Table.Moreover,for the thermodynamic study,the temperature was varied according to the values reported in the Table,while the other process parameters were kept constant.During the batch adsorption experiments,50 ml or 100 ml volume of known concentration of the contaminant was contacted with a certain amount of the slag in a 125 ml Erlenmeyer flask.The resulting mixture was placed on a shaker and agitated at a speed of 200 r∙min-1at room temperature (except for the thermodynamic study)for a specified time.After the agitation process,samples were collected,centrifuged,and filtrated.These experiments were conducted in duplicate and with a sample blank,which has no added adsorbent.Residual concentrations of the constituents were analyzed accordingly,where Cr3+levels were measured using AAS,and MB concentrations were determined using a spectrophotometer at a wavelength of 665 nm.The amount adsorbed per gram (qe) of the adsorbent and the removal percentage for each of the pollutants were calculated using the following expression.

Fig.1.SEM micrograph and EDS Spectrum of (a) Raw Slag and (b) Z-BFS (Modified Slag).

Fig.2.Comparison of FTIR spectra of the raw slag and Z-BFS.

where Co(mg∙L-1) is the initial concentration of the contaminants,and Ct(mg∙L-1) is the residual concentration at a corresponding time,m is the amount of adsorbent used in g and V is the volume of the solution in litres.

3.Results and Discussion

3.1.Material characterization

The surface structure and elemental composition of raw and modified slag samples were investigated using scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy(SEM-EDS) technique.The obtained SEM micrographs presented in Fig.1(a) and (b) shows the raw slag surface with no clear pore structures,while the Z-BFS shows a roughened surface that implied its interaction with the acid.

The EDS spectra presented in the same figure shows the presence of oxygen (32.7%),iron (20.1%),calcium (18.6%),carbon(15.4%)as the major elements,while silica,aluminium and magnesium were detected as the minor elements.The appearance of sulphur on the Z-BFS sample [Fig.1(b)] implied the loading of sulfamic acid on its surface.

FTIR analysis performed for both the raw and modified slag samples was used to confirm the successful functionalization of the slag with sulfamic acid.Results of the investigations,as presented in Fig.2,shows a clear difference in the spectra bands between the raw slag and Z-BFS sample.

The spectra of the Z-BFS shows broadband with a peak at 3432 cm-1that was assigned to the presence O—H functional group.The presence of this peak could be due to the SO3H group of sulfamic acid or to the presence of adsorbed water [19].The same peak can be observed in the raw slag sample but with a much lower intensity.Peaks identified between 3000–2800 cm-1bands in both samples were assigned to the presence of slag carbonate mixture [20].

Similarly,peaks in the band range of 2360–2100 cm-1for both samples were associated with the occurrence of silica group in the form Si—H silane.Besides,the peaks at 1000 cm-1and 900 cm-1for the raw slag was ascribed to the stretching vibration of Si—O—Si silica group.In the case of Z-BFS,the small peak at band 1383 cm-1was assigned to S=O stretching vibration of sulfamic acid,thus,affirming the occurrence of the sulfonic group [17,18].Moreover,the absorption peak at 1063 cm-1band is also a characteristic of —SO3group vibration of sulfamic acid [21].Other investigators reported similar peaks for sulfamic acid functionalized materials [22,23].The absorption band at 1627 cm-1for the Z-BFS was assigned to N—H stretching vibration of sulfamic acid.

To further confirm the successful functionalization of the slag,the Z-BFS sample surface elemental composition was identified through the XPS analysis.The XPS survey spectrum of the sample as illustrated in Fig.3(a) showed numerous peaks at 105.08 eV(Si 2p),170.6 eV (S 2p),280.4 eV (C 1s),403.08 eV (N 1s),533.6 eV (O 1s),715.5 eV (Fe 2p) and 1037.2 eV (Mg 1s) binding energies that were attributed to the mineral constituents of the slag and loaded zwitterion of sulfamic acid.The high resolution S 2p and N 1s spectra in Fig.3(a)and(b)with broad symmetric peaks were attributed to the sulfonic(S=O)and protonated aminogroups of sulfamic acid [24]. The identification of these functional moieties agrees with the EDS and FTIR analysis findings.

The surface textural properties of the raw slag and the Z-BFS sample acquired using the Micro metrics surface area analyser are provided in Table 1 as shown. As expected, the raw slag presented a small BET specific surface area value of 2.32 m2∙g-1and pore volume of 0.004165 cm3∙g-1, while the Z-BFS sample showed an improved surface area and pore volume of 16.57 m2∙g-1, 0.019 cm3∙g-1, respectively. The finding suggests the surface modification process enhance the slag surface properties, in addition to its functionalization with the zwitterionic group. The reduction in the average pore diameter from 71.85 to 45.68 Å (1Å = 0.1 nm) of the modified slag also justified the surface properties improvement after the functionalization process. It worth noting that the porestructures of both samples are still within the mesoporous range(2–50 nm).

Table 1Surface properties of the raw and modified slag

Table 2Experimental design matrix and responses

Table 3Key features of the models

Fig.3.X-ray Photoelectron Spectroscopy of Z-BFS (a) Survey scan,(b) high resolution S2p,and (c) high resolution N1s,spectra.

3.2.Surface modification of slag

3.2.1.Response surface modelling

The result of the experimental design matrix for the surface modification process showing the factor combinations,as well as the corresponding responses for each of the combined factors,are presented in Table 2,as shown.The two experimental responses,percent removal of MB (Y1) and Cr3+(Y2),were fitted to a reduced cubic model in terms of the coded values as presented in Eqs.(3) and (4).

Fitted Response Models in Terms of Coded Factors

Fig.4.3D plot showing the effect of sulfamic acid,contact time and temperature on (a) MB removal and (b) Cr3+ removal.

The above-fitted models were obtained after the elimination of insignificant terms while at the same time satisfying the model adequacy evaluations.Results of the model adequacy evaluations,which included the lack of fit test,correlation coefficient R2,adjusted R2,predicted R2,and adequate precision values,are shown in Table 3.The analysis of lack of fit for the two models was found to be insignificant,as indicated by their P-values of more than 0.05(5% confidence level).High R2value close to 1,and agreement between adjusted R2and predicted R2values (occurs when their difference is less than 0.20) are established criteria for a highquality model with good prediction capability [25].Obtained R2values for the two response models (Y1-0.9999 and Y2-0.9989)were close to unity,and the calculated differences between their adjusted and predicted R2values were far less than 0.2.The adequate precision value is a measure of the signal-to-noise ratio of the experimental responses.The values presented in Table 3 for both models,imply adequate signal(a value of more than 4 is often desirable) and signify the capability of the models to predict the experimental data with high accuracy within the design space[26].

3.2.2.Effect of the factors on the surface modification process

To understand and visualize in detail the impact of each of the factors and their interaction on the two modelled responses,a 3D Response surface plots depicted in Fig.4 was developed.Fig.4(a)(1–3) show the effect of sulfamic acid amount,contact time,and temperature on MB removal.A reduction in MB removal is observed with an increase in the sulfamic acid amount.A possible explanation to this finding is that at a high sulfamic acid amount,a significant dissolution of the metal oxides on the surface of Z-BFS occurred [Fig.1(b):EDS Spectra],which led to poor functionalization and thus,impacting MB uptake by the Z-BFS.

On the contrary,an increase in contact time from 30 min to 150 min led to a significant rise in MB removal as noticed.The effect of temperature is presented in plots a (2) and b (3),and it shows an increase in MB removal with temperature.An increase in temperature from 25 C to 75 C led to the rise in MB uptake to the optimal amount of 94%.

Fig.4(b)(1–3)show the effect of the process parameters on Cr3+removal.The impact of the sulfamic acid amount and contact time on Cr3+removal follows the same trend as that of the MB removal.An increase in contact time resulted in an increased Cr3+removal,while the sulfamic acid amount increase led to a decrease in Cr3+removal.Furthermore,temperature increase led to a rise in Cr3+removal,as depicted in the plots.

3.2.3.Models validation and responses optimization

The fitted models presented in Eqs.(3) and (4) were validated through an additional set of four experimental runs conducted with factor level combinations that were within the domain of the experimental design.The results provided in Table T3(Supplementary Material) show the predicted values based on the equations,actual values from the additional experimental run,and percent error difference.The models showed a reasonable agreement between the predicted and the actual experimental values with absolute percent error ranging from 1%to 14%.Hence,the fitted models show a high possibility of providing satisfactory predictions of the responses within the experimental design space.

The three surface modification parameters influencing MB and Cr3+removal responses were optimized using the desirability function.In simple terms,this approach entails setting a constraint and the desired goal (minimize or maximize) and identifying the optimum conditions,where the constraints and objectives are satisfied[27].The selected constraints,desired goals,and the result for the optimization of MB and Cr3+responses are provided in Tables T4 and T5 (Supplementary Material).Table T5 case 5-factors level combination was selected as the optimum condition because it provided desirability of 0.920 and high removal efficiency of more than 90 percent for both Cr3+and MB.Hence,the optimum conditions of the surface functionalization process,which can provide a removal efficiency of more than 90% for both MB and Cr3+,are 2 g sulfamic acid,50 min contact time,and a temperature of 37 C.The relative lower optimum conditions attained in this study suggests that the scale-up of the slag functionalization process would present an attractive means for producing a low-cost functionalized adsorbent for real application.

Fig.5.Effect of initial pH on single solute adsorption of(A)MB and(B)Cr3+(Cr3+reduction curve shows the effect of pH variation on its concentration without adsorption).

3.3.Adsorption experiment

Initial adsorption experiments were performed to evaluate the removal of Cr3+and MB from aqueous phase using the raw slag,and Z-BFS prepared at the optimum condition.The result is presented in Table 1,and it shows the raw slag(unmodified)exhibiting poor adsorption for MB (13% removal).Whereas,Cr3+was significantly removed from the solution via a combined adsorption and precipitation mechanism with the resulting solution having a pH value greater than 11.In the case of the Z-BFS,a significant improvement in MB removal (94%) was noticed,with a removal efficiency of seven times higher than that of raw slag.While for Cr3+,a significant removal efficiency of 99%was noted.These findings were able to prove the capability of Z-BFS to provide a substantial removal for both Cr3+and MB from aqueous samples.Further,the effect of the various adsorption parameters(pH,adsorbent dosage,and contact time)on the performance of the Z-BFS for the removal of Cr3+and MB was investigated.

3.3.1.Effect of initial pH

The importance of pH on adsorption processes is highly significant since it determines,in most cases,the interaction of the adsorbate with the functional groups of the adsorbent in aqueous phases [28].Hence,in this study,the influence of initial pH variation on the adsorption of both Cr3+and MB was studied,and the results are presented in Fig.5(a) and (b).

Fig.5(a)shows the removal of MB with the modified slag at different pH values (2 to 10).A notable rise in MB removal from 53%to 95%was noted as pH was increased from 2 to 4,and then a constant removal efficiency of around 96%was attained for higher pH values.A similar trend was also reported by Etim et al.[29]in their study on MB adsorption using coconut coir dust.Methylene blue(C16H18ClN3OS) being a basic cationic dye possess a pH-dependent dissociation in the aqueous phase forming cations(MB+),which is the chromophore of the dye and chloride (Cl-)anions [30].Hence,it is proposed that MB removal by the Z-BFS is due to the electrostatic interaction between the positively charged MB+cations and the negatively charged sulfonic groups of the Z-BFS.Moreover,the point of zero charge pH of 4.6(Fig.S1 (Supplementary Material)) determined for the Z-BFS further confirms this interaction,as its net surface charge was negative at pH value higher than 4.6.

The result on the effect of initial pH on Cr3+removal is presented in Fig.5(b) as shown.An increasing trend in Cr3+removal can be observed for the pH range investigated.Also included in the figure is the result of pH variation on Cr3+concentration(reference solution).As depicted,a reduction in Cr3+concentration by 50% is observed at pH values between 6 and 7,which indicates its precipitation as chromium hydroxide Cr (OH)3.Based on this,it was deduced that the high Cr3+removal efficiencies noticed at pH 6 and 7 were due to the combined effect of adsorption and precipitation.To further understand the possible role played by Cr3+speciation in the removal process,a chemical speciation calculation was performed using visual MINTEQ 3.1,and the speciation diagram is presented in (Fig.S2) (Supplementary Material).The system studied for the calculation were Cr (NO3)3,HCl,NaOH,H2O,where HCl and NaOH were the acid and base used for pH adjustment.From the diagram,the dominant species of chromium with high potential for adsorption onto the slag surface at the pH range investigated (3–5) were the two Cr3+hydroxo-complexes of CrOH2+and Cr2(OH)+42 .Gomez-Gonzalez et al.[19]reported that at acidic pH range (3–6),sulfonic acid ligand gets deprotonated,implying that it will be available for interaction with the chromium hydroxides species to form complexes via electrostatic interaction.In the study,FTIR and XPS analysis has confirmed the presence of sulfonic groups on the modified slag.Based on this,it is proposed that chromium removal by the Z-BFS was aided through the electrostatic interaction between the Cr3+species (CrOH2+and Cr2(-OH)2+4) and the sulfonic group of the Z-BFS [19].

Also,the influence of the Z-BFS composition on Cr3+removal cannot be ruled out,as the ionic exchange of Ca2+and Mg2+for Cr3+species on the exchangeable surface sites of the Z-BFS might have occurred at acidic pH values (3–5).It is worth noting that similar ionic exchange mechanism was reported by Kyzioł-Komosinska et al.[31]for Cr3+removal by biomass material.Based on the high removal efficiency noted for both contaminants at pH 5,and to prevent Cr3+precipitation at higher pH value,pH 5 was selected as the optimum in this study and was recognized for subsequent experiments.

3.3.2.Effect adsorbent dosage

The effect of Z-BFS dosage on the removal efficiency and adsorption capacity for both contaminants were investigated as presented in Fig.6(a) and (b).

Fig.6.Effect of modified slag dosage on adsorption of (a) MB and (b) Cr3+.

Fig.7.Adsorption rate for (a) MB and (b) Cr3+ at different time.

Batch adsorption experiments conducted involved contacting different dosage of the modified slag(25–200 mg) with 20 mg∙L-1concentration of Cr3+and MB (sample volume of 50 ml).For MB adsorption depicted in Fig.6(a),a steep rise in adsorption was observed as Z-BFS dosage was increased from 25 to 100 mg.While at higher dosages,a constant removal efficiency was noticed.This trend indicates the optimum dosage required to achieve more than 90% MB removal using the Z-BFS should be greater than 100 mg(equivalent to 4 g∙L-1dosage).In the case of Cr3+adsorption shown in Fig.6(b),increased adsorption was observed as Z-BFS dosage was increased.However,unlike MB removal,the optimum dosage for Cr3+could not be ascertained based on the dosage range investigated.Still,it was identified that a dosage of 100 mg resulted in more than 50% removal.

The observed increase in removal efficiency noticed for both MB and Cr3+at higher adsorbent dosages can be attributed to increased surface area and surface functional group availability [32].On the contrary,adsorption capacity (adsorbed amount qe) was observed to decrease with increase in adsorbent dosage.A maximum experimental adsorption capacity of 24 mg∙g-1and 3 mg∙g-1were noted for MB and Cr3+,respectively,at a dosage of 0.5 g∙L-1,which reduce to 2.33 mg∙g-1and 1.66 mg∙g-1at the dosage of 8 g∙L-1.The decrease in the adsorbed amount noted at high dosages could be attributed to the condition of unsaturation of adsorption sites at higher dosage for a fixed amount of pollutants,as compared to complete saturation taking place at lower dosages of the amount of the same pollutant [33].The above findings were consistent with those reported by [34,35] for MB and Cr3+removal,respectively,while using different adsorbent material.

3.3.3.Effect of adsorption time

The effect of adsorption rate on the removal of 20 mg∙L-1initial concentrations of Cr3+and MB was investigated,and the results are shown in Fig.7(a) and (b).

For this experiment,500 mg Z-BFS dosage and 100 ml volume of both contaminants were used.Result for MB presented in Fig.7(a)shows a rapid removal efficiency of more than 60 percent achieved at the initial stage (5 minutes) and equilibrium condition appearing after 720 minutes of contact time.Results on Cr3+adsorption presented in Fig.7(b) also depicts a rapid removal efficiency of more than 40%at 5 minutes of contact time,but unlike MB equilibrium condition was not attained even after 24 hrs.The fast adsorption rate observed at the initial stage and the slow rate noticed at later stages were attributed to the presence of numerous vacant active sites on the adsorbent surface that gets filled with time[36].A similar trend was also reported by [37] in their study on Cr adsorption on a waste tire,where the slow rate of adsorption at a later stage was attributed to the intra-particle diffusion process that requires longer time as compared to surface adsorption.

3.4.Adsorption isotherm models

The experimental adsorption isotherm data based on adsorbent dosage variation for both Cr3+and MB were modelled using some selected adsorption isotherm models;namely,Langmuir,Freundlich,Temkin,and Redlich-Peterson [16,38,39] as depicted inEqs.(5)–(8),respectively.The modelling approach was based on a non-linear regression method,and the obtained model parameters are presented in Table 4.

Table 4Isotherm model parameters of Cr3+ and MB adsorption onto the modified slag

Average relative error (ARE) and the coefficient of determination (R2) [38] presented in Eqs.(9) and (10) were employed as the error function terms for evaluating the fitness of the selected models.Non-linear regression model fittings were carried out with the aid of the SOLVER add-in tool in Microsoft Excel,and the procedure adopted involved minimizing(ARE)or maximizing(R2)the error functions between the experimental and the theoretical values that were predicted by the models.

Langmuir Isotherm

where qm(mg∙g-1) and KL(L∙g-1) is the maximum adsorption capacity for the adsorbates,and the Langmuir isotherm constant,respectively.

Freundlich Isotherm

where Ceis the equilibrium aqueous concentration of adsorbate(mg∙L-1),and qeis the adsorption capacity (mg∙g-1).K (mg∙g-1)and n are the Freundlich constants related to the sorption capacity and sorption affinity of the adsorbent.

Temkin Isotherm (TEM)

where B=RT/bT,A constant that depicts the heat of adsorption(J∙mol-1),AT(L∙g-1) is the Temkin equilibrium binding constant,and b is the Temkin isotherm constant.R is the universal gas constant (8.314 J∙mol-1∙K-1),and T denote Temperature in Kelvin.

Redlich-Peterson Isotherm (R-P)

A (mg-1),and B (L∙g-1) denotes the isotherm constants,while β is the Redlich-Peterson isotherm exponent.At a high concentration of adsorbates and when β=1,the model reduces to Freundlich and Langmuir isotherms,respectively [38].

Average Relative Error (ARE)

where qe,actualis the experimental adsorption capacity that was measured and qe,pred.is the predicted adsorption capacity calculated from the models.

Results of the non-linear regression fitting provided in Table 4 show the parameters of the selected models together with the respective ARE and R2values.It is highly desirable to have low ARE value and high R2value (closer to unity) because they signify how close the predicted values are to the actual experimental ones.In this study,ARE value of less than 5% (0.05) was chosen as the acceptable error limit because it suffices for a satisfactory model fitting[40,41].Redlich-Peterson model and Freundlich model with higher R2values (0.9577),and ARE values of 3.688 and 3.690,respectively,offered the best fit to MB adsorption by Z-BFS.Redlich-Peterson’s model incorporates elements from both Langmuir and Freundlich models.Therefore,it is expected to provide a better description of heterogeneous or homogenous surface than the two models [42].The model reduces to Langmuir equation for β=1 and the Freundlich equation for high adsorbate concentration.Hence,the obtained β value of 0.223(less than 1)in this study indicates the isotherm behaviour is most likely suited to the Freundlich model than the Langmuir model.Besides,the Freundlich model presented a much lower ARE value than the Langmuir model(ARE-13.74%).Therefore,it is concluded that MB adsorption by the Z-BFS proceeded via the multilayer approach to cover the various surface binding sites.Meili et al.[42]and Tong et al.[43]have also reported the Redlich-Peterson model to offer the best fit to MB adsorption based on its low ARE and R2values.However,unlike in this study,the adsorption behaviour was most suited to the Langmuir isotherm because its β value was close to 1.The isotherm plot(qevs.Ce) provided in Fig.8(a) and (b) show the degree of fitness between actual experimental and predicted models values based on the non-linear regression fitting.

Table 5Kinetic parameters for single solute adsorption of Cr3+ and MB

Fig.8.Isotherm plot for actual experimental values and the model predicted value,(a) MB and (b) Cr3+.

In the case of Cr3+adsorption,the relatively high error function value(20%≥ARE>17%)for all selected models,indicate that none of them was able to offer a satisfactory description for the Cr3+adsorption data.This observation can be confirmed from the isotherm plot presented in Fig.8(b),where all fitted model’s curves deviated significantly from the actual experimental data curve.Similar observations were reported by [44] in their study on Cr3+adsorption using plant residues.Since Freundlich and Redlich-Peterson models showed the lowest ARE values,18.15 and 17.59,respectively,the two models were regarded as the best of the four models.Thus,it was suggested that Cr3+adsorption might have also occurred in multilayer over the heterogeneous surface binding sites of the slag.

3.5.Adsorption kinetic models

In order to identify the rate-determining step of Cr3+and MB removal by the Z-BFS,the experimental adsorption data on adsorption rate study was fitted to both pseudo-first-order and pseudosecond-order kinetics models as presented in Eqs.(11) and (12)[45].The kinetic parameters obtained for the two models are provided in Table 5.Also,the fitted kinetic model plots are presented in Figs.S4 and S5 (Supplementary Material)

Table 6Adsorption thermodynamic parameters for MB and Cr3+

Linear pseudo-first-order model

Linear pseudo-second-order model

Here,(qe) represents the theoretical equilibrium adsorption capacity (mg∙g-1) of the pollutants derived from the model’s plot,(qt) denotes the adsorption capacity (mg∙g-1) at any time (t)derived from the experimental data,k1is the pseudo-first-order rate constant (min-1) which value was computed using the intercept of a plot of ln(qe-qt) against t (time) and k2denotes the pseudo-second-order rate constant (g∙mg-1∙min-1) and its value calculated from the slope and intercept of t/qtagainst t.

Results of Cr3+adsorption at the concentrations investigated revealed that the second-order model well describes it with R2values of 0.9581.A second-order rate constant (k2) of 0.0033 L∙g-1-∙min-1was obtained for Cr3+removal by the Z-BFS.Also,its theoretical adsorption capacity (qe) was found to deviate fairly from the experimental adsorption capacity value.Similarly,MB adsorption data was observed to fit well to the second-order model with R2values of 0.9999.However,in this case,its second-order rate constant value of 0.035 L∙g-1∙min-1was found to be higher than that of Cr3+,thus,suggesting its fast uptake rate by the ZBFS.In addition,the theoretical adsorption capacity obtained for MB was found to agree with the experimental adsorption capacity value.

In general,the second-order model gave the best fit for the adsorption data than the first-order model.Thus,suggesting that electrostatic interaction earlier proposed might have been the dominant step in the removal of both contaminants from the aqueous phase [46–48].

3.6.Effect of temperature and thermodynamic study

The influence of temperature on adsorption of both MB and Cr3+by the Z-BFS was studied at four different temperatures(25,40,55,and 70°C)that included the room temperature(25°C)in which all the previous adsorption experiments were conducted.Results from the study displayed in Fig.9(a)shows a slight increase in both MB and Cr3+removal efficiency as the temperature was increased.

It is proposed that the temperature increase might have provided more energy to the molecular species of both contaminants such that they were able to overcome the activation barrier quickly and,thus resulting in more adsorption.The increase in removal efficiency with temperature as observed indicates an endothermic driven adsorption process [49].A similar rise in adsorption trend with temperature has also been reported by other investigators[50–52].

Fig.9.(a) Effect of temperature variation on adsorption,and (b) Van’t Hoff thermodynamic parameter plot.

The nature of the adsorption process,its feasibility and the spontaneity of the adsorbent-adsorbate system interaction was elucidated through the thermodynamic parameters [51].While employing the Van’t Hoff relationship provided in Eqs.(13)–(15),the change in Gibbs free energy (ΔG),enthalpy (ΔH),and entropy(ΔS) of the adsorption system were calculated.

where R is the universal gas constant,T the temperature in Kelvin and Kdis an equilibrium constant that can be calculated from Eq.(9).

Cae(mg∙L-1)is the amount of the contaminant sorbed on the adsorbent surface at equilibrium,while Ce(mg∙L-1) is the residual amount of the contaminant left in the solution at equilibrium after adsorption.

To obtain the thermodynamic parameters[ΔH(J∙mol-1)and ΔS(J∙mol-1∙K-1)],a linear plot of ln Kdagainst 1/T was made as depicted in Fig.9(b).Values of ΔH and ΔS for both MB and Cr3+adsorption reported in Table 6 were derived from the slope and intercept of the plot,respectively.The change in Gibbs free energy provided in Table 6 for each of the temperature variation investigated was calculated from Eq.(16).

Table 7Comparison of adsorption capacities of different industrial waste adsorbents for Cr3+and MB

The calculated thermodynamic parameters clearly show that MB and Cr3+adsorption by the modified slag were highly feasible and spontaneous based on the negative Gibbs free energy values obtained [52,53].

Fig.10.Reusability of the Z-BFS for the removal of MB and Cr3+.

Moreover,positive values of change in enthalpy (ΔH) for both contaminant removal further confirm the endothermic nature of the adsorption process.It was reported that an enthalpy change(ΔH) in the range of 8–60 kJ∙mol-1often indicate that complexation reaction is the dominant mechanism of the adsorption process[50,54].In this current study,the change in enthalpy (ΔH) of 11.58 kJ∙mol-1calculated for Cr3+adsorption system justifies the earlier proposed mechanism of electrostatic complexation reaction between its hydroxo-complexes and the sulfonic group of the ZBFS.Furthermore,the positive values of change in entropy ΔS calculated indicate an increased random interaction between the contaminants and the adsorbent during the adsorption process.

3.7.Regeneration and reuse of the Z-BFS

Adsorbent reusability is of immense importance in its use for industrial application purposes.Hence,a regeneration study of the spent Z-BFS was investigated in this study while employing a dilute HCl solution (0.1 mol∙L-1) as the desorbing eluent.It was observed that the spent Z-BFS can be effectively regenerated by first washing with dilute HCl for 2 hr,rinsed with deionized water and then dried at 60°C for a period of 3 h.A significant desorption of the contaminants was confirmed based on the amount of Cr3+and MB measured in the eluent after the regeneration process.The results of the reusability study shown in Fig.10,depicts the regenerated Z-BFS exhibited a considerable removal for the contaminants even after 3 cycles of regeneration.

A maximum removal efficiency of more than 90%was achieved for both Cr3+and MB with the fresh Z-BFS and after the 3rd cycle of regeneration,it reduces to 65% and 80%,respectively.Hence,the above result affirms the excellent regeneration performance of the Z-BFS for Cr3+and MB remediation.

3.8.Proposed adsorption mechanism for Cr3+ and MB

To understand the active role played by the zwitterion functional groups of the sulfamic acid in the removal of both Cr3+and MB,additional FTIR spectra of the contaminants loaded Z-BFS(after adsorption) were acquired as shown in Fig.S6 (Supplementary Material).As highlighted earlier,the peak at band 1627 cm-1,1383 cm-1and 1063 cm-1previously assigned to N—H and S=O stretching vibrations of sulfamic acid were found to have diminished in the Cr3+loaded Z-BFS.This finding suggests the interaction of these groups with Cr3+through the electrostatic complexation reaction previously reported.

As for MB loaded Z-BFS (Z-BFS-MB),its FTIR spectrum shows the elimination of sulfamic acid functional groups bands and the appearance of new bands that confirm the successful loading of MB on Z-BFS.For this spectrum,the peak at 2979 cm-1was assigned to C—H stretching vibration of the aromatic group of MB [55].Similarly,the peaks at bands 1384–1070 cm-1were assigned to C—H and C=C aromatic structure of MB [55].Hence,the earlier proposed MB removal mechanism that involves the interaction between its cationic species (MB+) and the negatively charged sulfonic group of the Z-BFS can be affirmed from this finding.

A comparison of Langmuir maximum adsorption maximum capacities (qe) for different industrial waste materials,as reported in the literature are provided in Table 7.It was noticed that the ZBFS (Cr3+:6.117 mg∙g-1;MB:58.58 mg∙g-1) exhibited better adsorption capacity for both pollutants than some reported waste materials and thus,proving it could serve as a promising low-cost adsorbent for these pollutants remediation.

4.Conclusions

The sulfamic acid-functionalized slag exhibited significant removal for MB and Cr3+than the raw slag,as observed in this study.The influence of slag surface modification parameters investigated using the Box-Behnken experimental design revealed that sulfamic acid amount (B) had a major influence on the amounts of contaminants removed by functionalized slag.Also,the optimum conditions determined for the modified slag were 2 g of sulfamic amount,30 min of reaction time,and room temperature of 27 °C,which can result in more than 90% removal for both Cr3+and MB.Surface characterizations performed for the modified slag confirmed the incorporation of the sulfamic acid groups on its surface.MB adsorption was highly favoured at pH value greater than 5,while the occurrence of Cr3+precipitation as chromium hydroxide was significant at pH values greater than 5.The proposed adsorption mechanism for both pollutants was based on electrostatic interaction between their respective cationic species (Cr3+and MB+)and the sulfonic group of the slag.The adsorption process in both cases was found to be spontaneous (negative ΔG),endothermic driven (positive ΔH),and were suited to the Freundlich isotherm model and the pseudo-second-order adsorption kinetic model.The prepared functionalized slag can serve as a low-cost alternative adsorbent for the removal of aqueous phase organic and inorganic contaminants from tannery industries wastewater.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to express their gratitude to King Fahd University of Petroleum &Minerals (Dhahran,Saudi Arabia) for the technical and financial support provided.CB thank the Deanship of Scientific Research for the financial support through the project No:DF181034.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2020.08.047.