Mesoporous activated carbon-zeolite composite prepared from waste macadamia nut shell and synthetic faujasite

Surachai Wongcharee*,Vasantha Aravinthan,Laszlo Erdei

School of Civil Engineering and Surveying,University of Southern Queensland,Toowoomba 4350,QLD,Australia

Keywords:Activated carbon Adsorption Synthetic faujasite Functional composite Methylene blue

ABSTRACT Novel activated carbon-zeolite composite adsorbent was prepared from macadamia shell bio-waste and synthetic zeolite X using hydrothermal treatment.Characterisation studies revealed mainly mesoporous structure with 418 m2·g-1 BET surface area with faujasite clusters on the carbon carrier.Sorption capacity for methylene blue model pollutant increased from 85 to 97 mg·g-1 with the temperature increase from 25 to 45 °C,and improved with increasing pH.Nonlinear regression analyses found accurate fit to the pseudo- first-order kinetics model and intra-particle diffusion rate controlling mechanism.Excellent fits to the Jovanovic isotherm model indicated monolayer coverage on chie fly homotattic surfaces with variable potential.The thermodynamic analysis confirmed spontaneous and endothermic physisorption process.The spent adsorbent was regenerated with 20%capacity loss over five reuse cycles.Although the adsorbentwas developed for ammonia,heavy metaland organic matter removal from water sources,the results also indicate good performance in cationic dye removal from wastewaters.

1.Introduction

Activated carbons(ACs)are the most widely used adsorbents for many pollutants,owing to their low cost,sustainable production,and high performance.Activated carbons can be produced from coals,wood,and many other materials having high carbon content.Macadamia is a plant native to Australia but also cultivated in Hawaii,Brazil,California,South Africa,Israel,Bolivia,Guatemala,Colombia,Kenya,Costa Rica,Malawi,Mexico,New Zealand and China.Macadamia nut shell(MNS)is one of the leading biomass wastes in Australia.According to Australian the Macadamia Society,macadamia nut production was around 50000 tons in-shell in 2016[1]representing about 39800 tons of nutshell waste.Only around 5000 tons of nutshell waste was utilised as solid fuel and garden mulch,and the disposal of the rest is a concern.MNS is a suitable precursor for the production of activated carbons due to its hardness and durability.Activated carbons produced from macadamia nut shell have higher surface area,surface reactivity,and percent yields compared to carbons produced from pistachio,hazelnut,pecan,almond,black walnut and English walnut[2]Macadamia nutshell-based activated carbons(MACs)were successfully used as adsorbents for removing aurocyanide[3],phenol[4],and methylene blue[5].

Zeolites(ZEOs)are crystalline microporous aluminosilicates thatare built up of a 3-dimensional framework of SiO4and AlO4tetrahedra,linked by sharing oxygen atoms,weakly bonded cations and water molecules within the pores and voids of the structure[6].Isomorphous replacement of Si(IV)by Al(III)produces a negative charge in the lattice,and the balancing Na,K or Ca cations are readily exchangeable with cations present in solutions.Zeolites are widely used as molecular sieves,catalysts,ion-exchangers,and adsorbents.Various zeolite types/frameworks show high af finity toward specific pollutants,and this has been exploited in water in wastewater treatment over decades.Zeolites were recommended for the removal of ammonia[7],heavy metal ions[8],dyes[9],arsenic[10],radionuclides[11],and emerging pollutants[12].

Zeolites and activated carbons may be used in combination to take advantage of their different sorption characteristics.To achieve this aim,the simplest solution is to use ordinary mixtures of activated carbon-zeolite granules in packed columns as macrocomposites to enhance removal efficiencies for targeted contaminants[13].Unfortunately,this method is impractical for powdered materials due to the substantially different material densities and the resulting dosing and separation problems.Employing hybrid and composite materials can resolve the problems of particle separation and different settling rates in water.Suitable hybrids can be produced by combining activated carbon and natural zeolite particles with an inorganic binder,such as Portland cement[14].This approach allows the manufacture of desirable granularshapes and sizes by extrusion;however,italso diminishes the specific surface area.

The development of zeolite-carbon composites mainly stems from the utilisation of fly ash wastes,rich in aluminium and silica with significant unburnt coal content[15,16].Composites prepared from coal ash are suitable for various environmental protection applications[17,18].Such materials are produced with simple methods to ensure low production cost,hence their sorption performance can be compromised by the composition(including impurities)of the precursor,coupled with limited coalcontent[19].Improved composites with higher carbon content can be produced by separating and purifying the precursors.In a recentpublication,Purnomo[20]demonstrated the advantages ofthis approach by preparing a composite from bagasse fly ash that showed good performance in phenol removal.

Similar outcomes can be achieved using suitable activated carbons and synthetic zeolites prepared from readily available raw materials.Using verified recipes and methods from the International Zeolite Association Synthesis Commission and other sources,appropriate hydrophilic or hydrophobic(high Si to Al ratio,organophilic)zeolites can be selected to create functional composites.This line of research has been a major direction in related material research[21]leading to the introduction of many advanced materials[22]but so farwith no reports on the development of functional composite adsorbents engineered for water purification.

In this study,therefore,we investigate a functionalmesoporous composite adsorbent made from MAC and hydrophilic synthetic faujasite.The material characteristics of the novel composite were determined using established laboratory methods.Methylene blue dye model adsorbate was employed in batch experiments at natural pH for activity characterisation,also covering desorption and adsorbent reuse.Experimental data were analysed with nonlinear fitting to kinetic and equilibrium models and thermodynamic study to identify the controlling mechanism and parameters.

2.Materials and Methods

2.1.Activated carbon-zeolite composite(ACZ)preparation

Waste macadamia nut shell of the Integrifolia species(Fig.1)was obtained from a factory in Southeast Queensland,Australia.The crushed shells were washed with distilled water and dried at105°C for 24 h.In a typical batch,100 g of the dried shell was charred at 400°C for 2 h at a heating rate of 5 °C·min-1in an iron reactor furnace(F62730,Thermolyne)with the exclusion of air.About 50 g of char was placed in a crucible boat and loaded into a quartz tube reactor having the dimensions of 60 mm diameter,10 mm thickness,and 100 mm length.The tube reactor was installed inside a vertical tube furnace(STF55433C-1 model,Lindberg/Blue MTM).The furnace temperature was increased from 30 °C to 900 °C at a heating rate of 30 °C·min-1under a nitrogen gas flow rate of 0.30 L·min-1,followed by activation for 1 h using carbon dioxide gas at a flow rate of 0.3 L·min-1.After being cooled to room temperature,the activated carbon was milled in a Vibrating Cup Mill(Pulverisette 9,Fritsch)to obtain powdered MAC that was used as a parent component to prepare ACZ composite.

Zeolites were synthesised using the established sol–gel method combined with hydrothermaltreatment,both with and without natural zeolite(10%dry weight,Raw Supply Company,US)added as seeding material.In a typical batch,synthesis hydrophilic zeolite,sodium aluminate anhydrous(NaAlO2,Sigma-Aldrich)and NaOH pellets(Analytical grade,Ajax Finechem)amounting to 10.76 g and 11.25 g were mixed with 50.18 g of Milli-Q and gently stirred for 20 min on a hot plate at 90°C.Once a clear solution was obtained,the seed material and 127.81 g of sodium silicate solution(Na2SiO3,QP,Panreac)was added to the solution and vigorously mixed on a magnetic stirrer at room temperature to achieve a soft creamy consistency.For carbon pretreatment,10.0 g MAC was added to 100 ml of NaOH 0.4 mol·L-1solution and stirred on a hot plate for 12 h at 90°C temperature.This preparatory step is essential to ensure the penetration of ZEO into the carbon carrier and achieve effective impregnation with strong bondage between the two materials.Instead of NaOH,hydrogen peroxide can also be used for the purpose[23].Surface-treated MAC was added to pre-determined amounts ofzeolite precursor gel,and vigorously stirred for 24 h on a magnetic stirrer at room temperature.Subsequently,the mixture was loaded into a PTFE vessel and inserted into a Te flon lined stainless steel reactor for hydrothermal treatment in an autoclave at 120°C for 24 h or more.The synthesised raw ACZ material was set aside to cooldown to room temperature,washed with Milli-Q water several times until pH was around 8,and dried in an oven at 105°C for 24 h.The dry material was milled by Ball Miller(Pulverisette 5,Fritsch)to powdered ACZ for material characterisation and adsorption studies.

2.2.Characterisation of ACZ composite

Specific surface area,pore volume,and pore size distribution were determined with N2at 77 K using an AUTOSORB-1(Quanta Chrome)instrument and AS1Win v1.50 software.The samples were degassed at 80°C for more than 10 h prior to analyses.The specific surface area was estimated using the multipoint Brunauer–Emmet–Teller(BET)method,and the pore volume and pore size distribution were determined with the Density Functional Theory(DFT).Total pore volume and pore volume distribution were estimated with the t-plot method.Adsorption and desorption isotherms were investigated between 0 and 1 relative pressures(P/P0).

The surface morphology and features of the composite material were studied with a SEM-JCM6000 scanning electron microscope(SEM).A FEI TECNAI G2 20 S-Twin instrument was used for Transmission Electron Microscopy(TEM)imaging,coupled with Energy Dispersive X-ray(EDX)for spot analysis.Elemental compositions were determined with a CHONS auto analyser(Perkin-Elmer 2400,series II),while functional groups of samples were identified with Fourier-Transformed Infrared(FT-IR)spectroscopy in the 400–500 cm-1range using a Perkin Elmer Two ATR-FTIR equipment and Spectrum Standard v10.4.2 software.

Fig.1.Photographic images of raw material:a)nuts,b)dehusked nuts in shell and c)MNS.

Powder X-ray diffraction(XRD)patterns were investigated with a Philips X'Pert instrument with radiation CuKα=0.1542 nm,operating at 40 kV and 20 mA in the 10°–80°2θ scanning range.Powder diffraction files(PDFs)from the International Center for Diffraction Data(ICDD)were used as references with X'Pert HighScore Plus software for data analysis.

2.3.ACZ composite adsorption studies

2.3.1.Dye solution preparation

Stock methylene blue(MB)dye solution(1 g·L-1)was prepared by dissolving MB(C16H18ClN3S·2H2O,Ajax Finechem,Australia)in pure(Milli-Q)water,with a resulting pH of 6.5.A calibration standard formula(absorption versus concentration)was established using successive dilutions of a stock sample employing a UV–vis spectrophotometer(Jenway-6715)at peak absorbance of MB at 668 nm.

2.3.2.Batch sorption procedures

Adsorption kinetic studies were performed in batch experiments at natural pH,using 100 ml wide-mouth sample containers containing 100 ml of 150 mg·L-1MB solution with 0.1 g powered ACZ added.The samples were shaken at 140 r·min-1in a water bath with temperatures maintained at 25 °C,35 °C,and 45 °C in separate runs.Aliquots were withdrawn during the 0–240 min time period and filtered using Whatman GF/C filters before measuring absorbance in disposable cuvettes.Adsorption isotherms were obtained with 0.1 g adsorbent in a similar manner except that the MB concentrations varied from 10 to 160 mg·L-1and the samples were shaken for 24 h to reach equilibrium.Triplicate measurements and blank samples were used to establish the reliability and accuracy of the experimental data.The relative standard deviations of all measurements(including pH effect studies described later)were average 1.5%and maximum 5%.The mass ofMB adsorbed qe(mg·g-1)atequilibrium was calculated from the equation

2.3.3.Investigation of pH and surface charge effects

The effect of pH on MB adsorption at equilibrium was studied at 25 °C using samples with 150 mg·L-1MB concentration.The pH of samples was adjusted in the 2 to 11 unit range using laboratory grade(Sigma-Aldrich)HCl and NaOH solutions.Samples containing 0.1 g of powdered ACZ in 100 ml MB solution were shaken in a water bath at140 r·min-1for24 h to reach equilibrium.Aliquots werefiltered and measured for absorbance as described in batch sorption procedures.The point of zero charge(determined as the pH value with no change)was determined by the pH drift method[24].

2.3.4.Regression analyses and calculations

Numerical analyses,other calculations,and charting were carried out using the Origin Pro v7.5(OriginLab)software package.

2.3.5.Desorption studies

MB-loaded ACZ was gently rinsed with Milli-Q water to remove any unadsorbed MB,mildly sonicated in 100 ml of dilute 0.1 mol·L-1HCl solution for 5 min,and the supernatant was analysed for MB concentration by UV–vis spectrometry.The procedure was repeated until negligible desorbed MB was found in the supernatant.The recovered powder was washed in pure water and used again after drying for consecutive adsorption tests as described earlier.The complete adsorption–desorption cycle was repeated five times to assess reusability.

3.Results and Discussion

3.1.Material characteristics of ACZ composite

Table 1 presents the results of proximate analyses of MAC and ACZ obtained with a CHONS analyser.ACZ has 47% fixed carbon with 33.4%ash content and 7%volatiles,in good agreement with the 55%to 45%carbon to zeolite mass proportion used in the preparation.Compared with the MAC,the carbon content in the ACZ decreased by 6.95%,while the oxygen and hydrogen contents increased by 6.21%and 0.75%due to the presence of alumino silicates.

Table 1 Proximate and elemental compositions of MAC and ACZ material obtained

Semi-quantitative(surface spot)chemical compositions were obtained for the two parent materials and the ACZ using an EDX analyser coupled to TEM as shown in Fig.2(ACZ)and Table 2.The ACZ peaks shown in Fig.2 are in agreement with MAC and ZEO presented in Table 2.The major elements are C,O,Na,Aland Siwith smaller amounts of K,Ca,and Fe attributable to the seeding natural zeolite.Fig.2 also shows significant Cu content,which occurs as an artefact of the examination method.

Fig.2.Elemental composition of ACZ by EDX analysis.

Table 2 Elemental compositions of parent materials and ACZ

Table 3 lists the physical properties of the parent materials(MAC and ZEO)and ACZ composite.The ACZ sample has 418 m2·g-1specific surface area with an intermediate value between ofthe parentactivated carbon(830 m2·g-1)and the zeolite(100 m2·g-1).The other parameters such as pore volumes and widths follow a similar pattern.Such results are expected since the zeolite particles partly occupy the pores and surface ofthe activated carbon.The key finding is thatthe ACZ composite inherited the desired,predominantly mesoporous character of the zeolite and activated carbon.

Table 3 Physical properties of parent materials and ACZ composite

The sorption/desorption isotherms and pore size distribution of the ACZ adsorbent are illustrated in Fig.3.

Fig.3a shows the adsorption–desorption isotherm plots of ACZ.The volume of N2absorbed(Vads)at constant pressure(760 mm Hg)and temperature(273.15 K)per adsorbent weight unit are presented as a function of the relative pressure(P/P0)at 77 K.According to the classification of the International Union of Pure and Applied Chemistry(IUPAC),the isotherm plot of ACZ exhibits a combination of type I(partly)and type II(dominantly)curves.The rate is steep below 0.025 relative pressure,followed by a milder monotonous increase over the rest of the domain.Therefore,microporefilling is completed at 0.025 P/P0,followed by monolayer formation until about 0.2 P/P0,and then by multilayer absorption.ACZ is characterised by the simultaneous presence of micropores and mainly mesopores as shown by the presence of hysteresis loop above 0.6 relative pressure.

IUPAC classifies the pore sizes as micropores below 2 nm,and as mesopores from 2 to 50 nm.The pore volume is presented in Fig.3b as a function of the pore size.The ACZ composite displays broad peaks in the mesopore region,with the main peak at 3.50 nm that confirms a mesoporous character.The pore size distribution in the micropore region is bimodal,with a primary(0–0.7 nm)and a secondary(1–2 nm)range.The computed results indicate that ACZ comprises about 75.87%mesopores and 24.13%micropores.

Surface morphologies of MAC,ZEO and ACZ composites as observed with a SEM are presented in Fig.4.

The SEM micrographs reveal that MAC has a glassy but porous surface with a variety of pore sizes owing to the activation with CO2gas at high temperature.The SEM micrograph of the ZEO shows the alignment of the surface structure comprising small spherical particles and their disrupted clusters.Ozaki et al.[25]concluded that carbon supportwith large mesopore surface area and less acidic surface functionality are the most suitable to achieve nanocrystalline zeolite formation,and our results confirm their recommendation.The presence of spherical particles as observed in SEM shows that the chosen temperature of 120°C and the reaction time of 24 h the hydrothermal process ensured good crystallisation.The ACZ micrograph confirms the effective distribution of the small ZEO particles on the activated carbon career.

Fig.3.N2 adsorption and desorption isotherms at 77 K(a)and pore size distribution of ACZ by the DFT method(b).

Fig.4.SEM image of MAC,ZEO,and ACZ.

Fourier Transformed Infrared(FTIR)spectroscopy was used to analyse the surface functional groups of ACZ(Fig.5).

Fig.5.FTIR spectrum of ACZ composite.

The peaks in Fig.5 show the bands at 2093 and 1640 cm-1that are attributed to the O--H stretching vibrations owing to proton vibration occurring in water molecules.The bands at 979,735,and 684 cm-1represent the asymmetric and symmetric stretching vibrations related to the bridging phenomenon with SiO4or AlO4structure.The band at 593 cm-1indicates the deformation vibration of the Al--O--Si groups.The broad peak around 3300–3700 cm-1is attributed to the stretching vibration of the O--H bond hydroxyl groups of the zeolitic structure,involving bridging with Al and Si(Si--OH--Al)[26].In summary,the results are in agreement with the characteristic numbers ascribed cardinally to Na-X type zeolites.

X-ray diffraction was used for the analysis of the crystal structure of the ACZ composite by measuring the intensity of radiation re flected at various angles.Powder diffraction files(PDF)from the International Center for Diffraction Data(ICDD)were used as references using X'Pert HighScore Plus software forthe analysis ofdata peaks as shown in Fig.6.

Fig.6.Typical XRD pattern(background subtracted)for ACZ composite and Powder Diffraction Files(PDF).

The XRD pattern shows clear broad peaks around 11.80°,15.56°,18.53°,20.20°,21.64°,22.63°,23.48°,26.84°,28.00°,29.40°,30.51°,31.16°,32.22°,33.81°,34.40°,37.60°,41.02°,42.83°,50.08°,52.00°,53.52°,57.79°,63.56°,68.26°and 72.37°2θ,which is matched with PDF 00–038-0232(Zeo-X)and PDF 01–080-2463(Zeo-Y).As expected,the amorphous activated carbon does not show strong peaks.The presented resultreveals the successfulpreparation offaujasite-type zeolite.Faujasites are attractive materials for technological and environmental applications due to their wide micropores,high exchange capacity,and hydrophilic surface with an af finity toward polar molecules[6].We note that it is relatively simple to produce either hydrophilic or hydrophobic(mordenite type,organophilic)zeolites to tailor the composite for the enhanced removal of certain pollutants,by changing the preparation recipe and/or duration of hydrothermal treatment.In addition,surface modifications using acid/base treatmentand surfactantimpregnation also can improve the sorption efficiency oftargeted ions and organics[27].

3.2.Effect of solution pH

The effect of pH is an important process parameter in adsorbate uptake by determining the surface charge and the degree of ionisation of the adsorbentand adsorbate.To examine the effectofpHon the adsorption capacity at equilibrium,solution pH was investigated over a pH range of 2–11(Fig.7a).

It can be seen that the adsorbed mass of MB onto ACZ steeply increased from pH 2 to 7,and then reached a plateau with no further uptake over pH 8.The adsorption capacity increased significantly from 41.08 mg·g-1to 87.58 mg·g-1when the initial pH was increased from 2 to 8,showing the strong in fluence of pH on the equilibrium adsorption capacity.The pH of zero surface charge(PHZ)was determined in the presence of pH-neutral(NaCl)electrolyte with results shown in Fig.7(b).The relatively high value of PHZ(about 10)indicates the dominantrole ofthe synthetic zeolite component.The faujasite was prepared using alkaline hydrothermal method and shows a strong screening ion effect[28].If desired,the PHZ of raw synthetic zeolites can be reduced by acidic treatment,which is frequently used in the preparation of zeolite-based catalysts to increase their durability.From these findings,it can be deduced that the sorption of the cationic dye is hindered by the competing,like-charged hydronium ions at low pH values,and the full MB sorption capacity can be utilised at basic pH values.

3.3.Effect of contact time and temperature

The in fluence of contact time and temperature on MB uptake was studied at three different temperatures at natural pH over four hours duration and the results are illustrated in Fig.8.

It can be observed that the mass of MB dye adsorbed on ACZ increases with time at each temperature to saturation limits,and the uptake is relatively rapid during the first 60 min of contact.Thereafter,the rate of adsorption declines and equilibriumis achieved at 150–180 min,depending on the temperature.The adsorbed masses of MB at equilibrium were 85.86 mg·g-1,91.58 mg·g-1,and 97.01 mg·g-1at 25 °C,35 °C,and 45 °C,respectively for 150 mg·L-1MB initial concentration.The increase of initial rates with temperature immediately indicates that the process is endothermic and spontaneous,like that often found in dye adsorption studies.The increase of temperature reduces the viscosity of the solution,and the resulting increase of mobility of dye molecules is manifested in higher rates of diffusion through both the liquid–solid interface and within the adsorbent.The increased kinetic energy also increases the adsorption capacity by counteracting the surface charge barrier created by the competing hydronium ions[29].The proposed major mechanisms of MB removal for the carbon component is an electrostatic attraction between the nitrogen groups of the dye and carboxylic surface groups[24],and molecular(π–π,n–π,and p–π)interactions[30].For the zeolite components,the removal is attributed to cationic ion exchange.The possible mechanisms that may happen in the MB adsorption process onto ACZ material are shown in Fig.9.

Fig.7.EffectofpHon adsorption ofMBonto ACZ(a)at150 mg·L-1 MB initialconcentration,25°C temperature,120 r·min-1 shakerspeed and 24 equilibriumtime,and(b)determination of the PHZ.

Fig.8.Effect of contact time on MB adsorption efficiency(ACZ dose=0.1 g,pH?8,MB volume=100 ml;initial MB concentration=150 mg·L-1;shaker speed=120 r·min-1;temperature=25 °C,35 °C and 45 °C).

3.4.Adsorption kinetic studies

Adsorption kinetic studies provide valuable information about the mechanism of adsorption and essential design information for applications.In this study,the adsorption kinetic of MB dye solution onto ACZ composites was probed by the pseudo first order(PFO),pseudo second order(PSO),general(rational)order,Elovich,diffusion chemisorption,and Avrami fractional-order models.Noting the various problems created by data linearization[31],in this study only nonlinear regression was used.The best fitting kinetic model was selected by comparing the individual and mean numerical values of the adjusted coefficient of determination,chi-square[32],and the root of mean square error[33].Table 4 shows the results at three different temperatures for the six models chosen.

The numerical results reveal excellent fits to three models.The general order model has the best statistics considering the averages of X2and RMSE values obtained at the three temperatures,while it is ranked second after the PSO model regarding R2adj.The PFO model predicts the equilibrium capacities best,followed by the so-called Avrami fractionary-order(a misnamed but widely used)empirical model.The PFO and PSO models are special cases of the general order model with n=1 and n=2,respectively[39].In this case,the average of the n exponents nav=1.0364 is very close to the unit value of the PFO model.Similarly,the Avrami fractionary-order model also reduces to the PFO model for nav=1 exponent,which is nearly matched by the 0.98783 calculated average value.These results show that the sorption process follows first order kinetics.The best model among these candidates is clearly the classical PFO model,as it is the simplest one with only two parameters.

The PSO kinetic model has appreciable statistics,however the predicted values of qesignificantly(minimum 15%)deviate from the measured values.The Elovich and diffusion-chemisorption kinetic models show poorer fits,suggesting that the sorption process does not involve a chemisorption mechanism.

Fig.10 illustrates and confirms the very good predictive performance of the PFO model.

3.5.Internal diffusional model

The rate controlling step of adsorption processes is commonly investigated by the application of the Crank(short time),Boyd,Weber–Morris and Bangham diffusional models[40].Mass transfer resistance across the liquid film on the adsorbent surfaces can be neglected as the rate-determining step in well-mixed reactors[41].In this study,experimental data was fitted to the Bangham model,which is an application of the Freundlich and Helle model[42],using the equation

The kinetic plots in Fig.10 show that the adsorbent is half-saturated in 30 min or less,hence it is appropriate to consider the 0–30 min initial period in modelling.Table 5 and Fig.11 provide very good fits to the Bangham model at each temperature,indicating that the controlling mechanism of the adsorption process is intra-particle diffusion.

3.6.Equilibrium studies

Fig.9.Possible MB adsorption mechanism onto ACZ obtained.

Sorption equilibrium is achieved when the adsorbent is in contact with the adsorbate for a suf ficiently long time,resulting in a steady mass balance between the bulk solution and deposited material.There are many empirical and semi-empirical isotherm models available to predict equilibrial conditions.In this study,various models were examined to quantify the adsorption characteristics of the system.Table 6 presents the calculated parameters for the Freundlich,Langmuir,Temkin,Harkin–Jura and Jovanovic isotherm models.

Table 4 Kinetic modelling results of MB adsorption on ACZ

Fig.10.PFO kinetic model plots of MB adsorption onto ACZ composite at 25 °C,35 °C,and 45°C.

Table 5 Bangham model results of MB adsorption onto ACZ in the 0–30 min period

The best results are obtained by fitting data to the Jovanovic model.The predicted values of qm(86.14,89.61 and 95.51 mg·g-1)are very close to experimental values of qm,with small 0.6%–2.1%deviations thatare within the range ofexperimentalerrors.This modelemphasises a dynamic balance by assuming homotattic surface with variable potential rather than the homogenous surface.Monolayer coverage is assumed,whereby the collision of impinging molecules with desorbed molecules also affects the equilibrium.Lateral movement(hopping)of adsorbed molecules to vacant adjacent sites is allowed but lateral interactions are excluded.Fig.12 visually confirms very good fits to the Jovanovich model.

Fig.11.Bangham modelplots ofMB adsorption onto ACZ obtained at25 °C,35 °C and 45 °C temperatures.

Table 6 Isotherm models of MB adsorption on the ACZ material at 25,35,and 45°C

The Langmuir model provides a relatively good fit but overestimates the saturation capacities.This model assumes monolayer(twodimensional)coverage on a homogenous surface with equivalent active sites.There are no interactions among the molecules of the adsorbate,and the number of desorbed molecules is proportional to the covered portion of the surface.The model allows the calculation of the separation factor,and the obtained small values(0.058–0.014 at C0=150 mg·L-1)indicate favourable adsorption of MB onto the ACZ composite at each temperature.

Fig.12.Jovanovic isotherm plots of MB adsorption onto ACZ composite at 25 °C,35 °C and 45C°C.

Lastly,the relatively poor fits for the Freundlich,Temkin,and Harkin–Jura models do not provide numerical support to consider multilayer adsorption on heterogeneous surfaces.Therefore,the likely adsorption mechanism of MB onto ACZ involves monolayer coverage over homotattic surfaces.

Table 7 shows reported maximum adsorption capacities of MB onto various related adsorbents,such as natural zeolites,modified zeolites,fly ash based zeolites and composites.

Table 7 Comparison of reported MB adsorption capacity of MAC,zeolite and carbon-zeolite composites

The measurementof the MB number is an established method to estimate the specific mesoporous surface area,and also allows obtaining information about the total and micropore volumes[54].The ACZ composite has a significantly smaller specific surface area(418 m2·g-1)than MAC(830 m2·g-1)and larger than the zeolite(100 m2·g-1)parent compounds,and a resulting capacity nearly in the proportion of the ratio of the constituents.

These results show that while the carbon-zeolite composite was specifically developed for the specific pollutant(ammonia,heavy metals and organic matter)from water sources,it is also suitable for MB removalfrom aqueous solutions.Arecently completed investigation found a good performance by ACZ in simultaneously removing natural organic matter(NOM)and ammonium from poor quality drinking water source.Furthermore,both activated carbons and zeolites have been used successfully in wastewater treatment to enhance nitrification,remove recalcitrant compounds,and improve sludge separation and dewatering.Therefore,the ACZ hybrid may find additional applications in biological wastewater treatment.

3.7.Thermodynamic studies

The change ofGibbs free energy,enthalpy,and entropy can be determined from Eqs.(3)and(4)using the Kevalues obtained at different temperatures[55].

The negative values of ΔG°from Table 8 indicate the feasibility and spontaneous nature of adsorption,confirming the validity of the small separation factor values calculated from the Langmuir model.The positive values of ΔS°suggest increased randomness during adsorption at the solid–liquid interface.The positive value of ΔH°confirms that the adsorption reaction is endothermic,and the enthalpy change of 55.61 kJ·mol-1being much smaller than the lower limit of chemisorption(80–450 kJ·mol-1)range.This finding confirms physisorption phenomena,in agreement with earlier research concerning MB adsorption on carbons and zeolites.

3.8.Desorption studies

The changes of efficiencies over consecutive adsorption–desorption cycles are shown in Fig.13.The normalised adsorption capacity decreased by 20%after five uses in the order of 100%,95%,92%,87%and 80%.The desorption efficiency(relative desorbed mass)showed a greater decrease from 100%to the final 69%value.No significantchange in the UV–vis spectrum of the dye was observed before and after desorption,underlying again that the process involved physisorption.These results show that the ACZ composite has good reuse potential in MB dye removal.

Fig.13.Reusability of ACZ composite.

4.Conclusions

A novel mesoporous activated carbon-zeolite composite was produced from macadamia nut shell-based activated carbon and synthetic zeolite by the hydrothermal treatment process.The extensive qualitative and quantitative characterisation studies conducted on the composite confirmed the impregnation of the MAC carrier by mesoporous faujasite nanocrystalline clusters.The product contained mainly mesoand macropores with a specific surface area of 418 m2·g-1,nearly the average of the parent materials.Batch sorption experiments with MB dye at natural pH found sorption capacities from 85 to 97 mg·g-1, which increased with the temperature in the 25 °C to 45 °C range.The solution pH significantly in fluenced the uptake capacity,and alkaline conditions provided better performance.

Table 8 Determined thermodynamic parameters

Kinetic studies performed with nonlinear regression analyses found that the classical PFO model accurately described the sorption process.The general order and Avrami-fractionary models also provided excellent fits,and their exponent values confirmed first order kinetics.The Jovanovic model provided the best fit for the isotherms to indicate monolayer coverage on the chie fly homotattic surface with variable potential.The calculated thermodynamic parameter values confirmed that the adsorption process was spontaneous,endothermic and involved physisorption.The physisorption nature of adsorption enabled the effective regeneration of the composite for subsequent reuse,with 20%capacity reduction over five cycles.

The results confirmed the successful preparation of a composite adsorbent using a simple hydrothermal method.Although the ACZ is primarily intended for specific pollutant(organic matter in the presence of ammonia and heavy metals)removal,it also showed good performance in sorption of the cationic MB dye,and thus may find wider application in the areas of water and wastewater treatment.Point to be noted is that the ACZ composite is an affordable functional adsorbent,whose characteristics can be tailored to achieve enhanced af finity and selectivity toward specific contaminants that resist removal by currently available adsorbents.It offers desirable material characteristics,durability,and reusability.

Nomenclature

aHHarkins–Jura isotherm parameter

bHHarkins-Jura isotherm constant

btTemkin isotherm constant,J·mol-1

CeConcentration of MB at equilibrium,mg·L-1

C0Initial concentration of MB,mg·L-1

KeDistribution coefficient

kavAvrami kinetic constant,min-1

kbBangham rate constant,g·mg-1·min-?

kDCDiffusion-chemisorption constant,mg·g-1·min-n

kfFreundlich isotherm constant,(mg·g-1)(L·g-1)n

kjJovanovic constant

klLangmuir isotherm constant,L·MB-1·mg-1

krRate constant of general order model,h-1·(g·mg-1)n-1

ktTemkin equilibrium binding constant,L·mol-1

k1PFO rate constant,min-1

k2PSO rate constant,g·mg-1·min-1

m Mass of ACZ adsorbent,g

n Order of reaction

navAvrami model exponent

nfHeterogeneity factor

qePredicted mass of adsorbed MB at equilibrium,mg·g-1

qexMass of adsorbed MB at equilibrium(experimental),mg·g-1

qmMaximum adsorption capacity,mg·g-1

qtMass of adsorbed MB at t time,mg·g-1

R Universal gas constant(8.314 J·mol-1·K-1)

R2adjAdjusted nonlinear coefficient of determination

RlSeparation factor

RMSE Root of mean square error

T Absolute temperature,K

t Adsorption contact time,min

V Volume of solution,ml

αeInitial adsorption rate of Elovich model,mg·g-1·min-1

βeElovich desorption constant,g·mg-1

ΔG° Gibbs free energy change

ΔH° Enthalpy change

ΔS° Entropy change

? Bangham constant

χ2Chi-square test

Acknowledgement

The first author thanks the University of Southern Queensland and Australian GovernmentResearch Training Program Scholarship for providing a scholarship to pursue this work.