Zeolite P synthesis based on fly ash and its removal of Cu(II)and Ni(II)ions☆

Yong Liu*,Guodong WangLu WangXianlong LiQiong LuoPing Na

1School of Environmental Science and Engineering,Tianjin University,Tianjin 300350,China

2School of Chemical Engineering and Technology,Tianjin University,Tianjin 300350,China

Keywords:Class C fly ash Zeolite P Kinetics Thermodynamics

A B S T R A C T Zeolite P wassynthesizedthroughhydrothermalmethod based on a kindof ClassC fly ash(FA).X-ray diffraction(XRD),scanningelectronmicroscopy(SEM),andBrunauer-Emmett-Teller(BET)wereusedtoanalyzeandcharacterizethesyntheticsample.Thekineticsandthermodynamicsofcopperandnickelionsremovedbythezeolite samples were experimentally explored in detail.The results of kinetic treatment showed the second-order exchange second-order saturation model(SESSM)can well describe the removal process of copper ions,while the first-order empirical kinetic model(FEKM)isthe best kinetic model for nickelions.Langmuir and Freundlich isothermswereusedtofit theequilibriumconcentrationofCu(II)orNi(II)undercertainconditions.Whetherfor copperornickelion,theLangmuirmodelisingoodagreementwiththeexperimentalequilibriumconcentration.The apparent theoretical removal capacities for Cu(II)and Ni(II)can reach to 138.1 mg·g?1and 77.0 mg·g?1,respectively.

1.Introduction

Because of its strong toxicity,environmental persistence and bioaccumulation,heavy metalhas greatharm to human health,biological resources and environment system[1].So it must be effectively treated before discharge.At present,precipitation,ion exchange,electrochemical method,solvent extraction and membrane separation technology were the main methods to remove heavy metal ions from wastewater[2].Nowadays,adsorption is considered as an effective method,and the key is to find more cheap adsorbents[3].

Zeolites are a kind of porous aluminum silicate crystals,it has good adsorption performance and ion exchange capability,which has been widely used in sewage and gas control[4-6],chemical separation[7]catalytic reaction[8]and materials industries[9].In the early stage of zeolite synthesis,pure chemical reagents were often used as raw materials,which led to the uneconomical preparation and environmental pollution.For saving the cost of zeolite synthesis,many researchers have tried to use some waste or low value materials as a raw material instead of pure chemical reagent.

Coal combustion produces two kinds of solid waste,that is,bottom ash and fly ash.Fly ash has a huge annual output in China[10],but its comprehensive utilization rate is low.The utilization rate of FA is estimated about 16%globally[11].A large amount of FA is idle or abandoned,resulting in severe environmental and ecological problems[12-14].Therefore,the comprehensive utilization of FA is very necessary.

FA contains a lot of silicon and aluminum elements,which is similar withthecompositionofzeolites.ConvertingFAintozeolitescanachieve theharmlessofsolidwasteandtherealizationofresourceutilization,so this topic has aroused great interest for the researchers and various zeolites have been synthesized from FA.Volli et al.[15]synthesized several zeolites from coal fly ash(CFA)by alkali fusion followed by hydrothermal reaction.And a single phase and high quality zeolite X can form at 110°C for 12 h.Wang et al.[16]prepared an initial gel using Si source dissolved from fly ash by NaOH solution,then pureform zeolite A was synthesized with the extra Al source under 100°C hydrothermal treatment for 340 min.Zhu et al.[17]converted colloidal silica and CFA into ZSM-5 zeolite by conventional and microwaveassisted hydrothermal reaction,respectively.It was shown that microwave-assisted synthesis effectively reduced time from 6 h to 1 h.By adding sodium hydroxide and sodium aluminate into CFA,Behin et al.[18]adjusted the molar composition of the final aluminosilicate gel to 4.45Na2O:1Al2O3:1.79SiO2:193.78H2O,then synthesized zeolite LTA using a single-mode microwave oven equipped with a reflux condenser under atmospheric pressure.Franus et al.[19]synthesized Na-X,Na-P1 and sodalite from Class F fly ash under different conditions,and three types of synthetic zeolites all have relatively high ion exchange capacities.

In recent years,there are lots of reports about the zeolite 4A,zeolite X and zeolite Y synthesis[20-25],but the research for zeolite P is relatively few[26].Also many reports mainly focused on the factors influencing zeolite synthesis and the removal effect for pollutants[27-29],and the thermodynamics and kinetics of heavy metal ions on zeolite P were rarely reported[30].In this paper,zeolite P was successfully synthesized from FA by hydrothermal method.Meanwhile,the removal characteristics for Cu(II)and Ni(II)on the samples were explored,and the kinetics and thermodynamics of the heavy metal ion removal process were studied in detail.This paper will provide theoretical guidance and technical support for the application of this technology.

2.Materials and Methods

2.1.Materials and chemical reagents

The FA was provided by a heating company in Tianjin,China.Hydrochloric acid was obtained from Tianjin City Yuan Li Chemical Co.,Ltd.,China.Sodium hydroxide,copper sulfate,and nickel chloride were purchased from Tianjin Kwangfu Institute of Fine Chemicals Co.,Ltd.,China.Hach reagents were obtained from Chemart(Tianjin)Chemical Technology Co.,Ltd.,China.All reagents are analytically pure.

2.2.FA pretreatment

Firstly,therawFAwasgroundbyacrusherandsievedthrougha100 μmaperturesieve,andtheundersizecomponentswereannealedat750°C for 1.5 h in a muffle furnace to remove unburned carbon in FA.Then in accordance with the solid-liquid ratio of 1:5(g·ml?1),the calcinate was soaked in 4 mol·L?1hydrochloric acid solution and stirred at 90°Cfor2h,thenfiltered,washed,andputintoanoventodry.Thesetreatments aimed to remove theiron,calcium,sulfur and other impurities in the FA.

2.3.Preparation of samples

10 g of the pretreated FA was used as raw material,then 2 mol·L?1sodium hydroxide solution was added into FA according to the solidliquid ratio of 1:5(g·ml?1),the mixed solutions were placed at room temperature for 12 h,then they were transferred into a 200 ml hydrothermal reactor,and crystallized for 24 h at the preset temperature.Afterthereactantswerecooledtoroom temperature,theywerefiltered andfullywashedwithdistilledwater,thendriedat105°Cintheovento obtain products.

2.4.Removal experiments

Artificial solutions of Cu(II)and Ni(II)were respectively prepared with CuSO4and NiCl2in advance,and diluted with distilled water to the required concentrations.Firstly,25 ml ion solution with a certain concentration was put into a 250 ml conical bottle,and then 0.05 g of synthetic zeolite was added into the conical bottle,finally the conical bottle was immediately put into a thermostat oscillator(150 r·min?1).Supernatants were withdrawn at different time intervals through a 0.45 μm membrane filter,then the filtrate was diluted with distilled water to a suitable level and the ion concentration in the diluent was measured by a Hach automatic analyzer.The solution pH in all experiments was not adjusted.The initial concentrations of Cu(II)or Ni(II)for kinetic experiments which were carried out at 28 °C were both 100 mg·L?1.All equilibrium concentration experiments for Cu(II)or Ni(II)on the synthetic zeolite P were run for 24 h and the initial concentrations of Cu(II)and Ni(II)are 50 mg·L?1,100 mg·L?1,150 mg·L?1,200 mg·L?1,250 mg·L?1and 350 mg·L?1,respectively.All experiments were repeated,and the average value of three experiments was taken as thefinal result.

2.5.Characterization and analysis

In this study,the compositions and contents of FA were obtained by an X-ray fluorescence spectrometer(XRF)of S4 Pioneer(Bruker,Germany).The powder X-ray diffractometer(XRD)of D/MAX-2500(Rigaku,Japan)was used to obtain the diffraction data of the synthetic products,and samples were scanned with a range of 5°-60°and a scanning speed of 6(°)·min?1.With the help of JCPDS files,crystal phase of inorganic and zeolite compounds could be identified.A scanning electron microscope(SEM)of S-4800(Hitachi Limited,Japan)was used to observe the morphology of samples.The specific surface and pore volume of zeolite samples were determined through a Brunauer-Emmett-Teller(BET)analyzer(Quantachrome Ins.,United States).The metal ion content in aqueous solutions was determined using a DR3900 type Hach apparatus(Hach,United States).And the amount of ions per unit mass of zeolite at different time(qt)and the apparent equilibrium capacity(qe)were evaluated using the following expressions.

3.Results and Discussion

3.1.FA analysis

The components,morphology and X-ray diffraction data for FA are shown in Table 1,Figs.1 and 2,respectively.Afterpretreatment,XRFresults indicate that the impurities such as iron(Fe2O3),calcium(CaO),sulfur(SO3)and magnesium(MgO)were removed by about 64%,87%,80%and 78%,respectively,compared to raw FA.Besides,the content of Al2O3decreases from 14.19%to 10.62%,while the content of SiO2increasesfrom33.50%to73.10%,sothemolarratioofSitoAlisgreatlyimproved from 4 to about 11.

Table 1 Compositions of raw FA and treated FA with hydrochloric acid(%)

XRD patterns show that quartz and mullite are the main crystal phases whether it is raw FA or the treated FA,indicating pretreatment has little effect on quartz and mullite in raw FA.Comparing with raw FA,the few stray peaks in the XRD of the treated FA are possibly due to fewer impurities.

3.2.Characterization of zeolite samples

3.2.1.XRD analysis

Fig.1.SEM diagrams for raw and treated FA:a—raw,b—treated.

The XRD results of hydrothermal synthesis products under different conditions are shown in Fig.3.By contrast with the standard JCPDS cards,the crystal phases are mainly zeolite P[31,32]and unreacted quartz under the condition of 95 °C and 105 °C,but the diffracted intensity is weak.Under 120°C,there is not only zeolite P with higher diffraction intensity,but also sodalite.From Fig.3,the diffraction intensityforquartzdecreaseswiththeincreaseofreactiontemperature,indicating quartz is likely to involve in the synthetic reaction of zeolites when temperature is higher.Thus,higher temperature is beneficial to the formation of zeolite P and the reaction of quartz,but easily leads to the formation of sodalite.

3.2.2.SEM analysis

Fig.4 is the microscopy of prepared samples under various temperatures.The synthetic products contain spherical particles,which are obviously different from those of raw FA(Fig.1).The spherical particles should belong to zeolite P.The size of the spherical particles respectively synthesized at 95 °C,105 °C and 120 °C is very uniform,which is basically 7 μm.In addition,there are some cracks on the particle's surface.

3.2.3.BET-DFT analysis

A nitrogenabsorption/desorption isothermal curve forzeolite P synthesizedat120°CispresentedinFig.5.Itshowsthattheabsorptionisothermal curve is similar to a typical IV-type shape,indicating that the synthetic zeolites contain mesoporous structures.And a H3 hysteresis loop appears in the N2adsorption-desorption isotherm,which may be due to the condensation of nitrogen between the zeolite particles.The sample has an equivalent BET specific surface area of 42.01 m2·g?1and a pore volume of 0.111 cm3·g?1.

Fig.2.XRD patterns for raw and treated FA:1—raw,2—treated(Q—quartz;M—mullite).

3.3.Removal of Cu(II)and Ni(II)

Thissectionmainlyaimsattheremovalbehaviorsofcopperionsand nickel ions from water solution by the zeolite samples synthesized at 120°C.The kinetics and thermodynamics of ion removal by the synthesized sample will be investigated.

3.3.1.Ion removal kinetic models

For Cu(II)and Ni(II),the change of ion concentration with time is shown in Fig.6.The initial concentrations for Cu(II)or Ni(II)are both 100 mg·L?1,and the ratios of solid(zeolite)to liquid(solution)are both 2 g·L?1.But the copper ion concentration reduces faster than that of Ni(II).For Cu(II),after about 3 h the concentration will be kept at constant.While for Ni(II),it needs about 10 h to reach a stable concentration.Thus,the removal efficiency of copper ion by the synthetic zeolite is higher than that of nickel ion under these studied conditions.

Fig.4.SEM images for the synthesized products:a.95 °C,b.105 °C,c.120 °C.

In this work,the synthetic sample belongs to sodium type zeolites,which has a large specific surface area.Thus,the zeolite has not only the ion exchange ability,but also the adsorption property.In addition,the ion of Cu(II)or Ni(II)easily produces hydroxide precipitation.Therefore,the process of ion removal in this study is more complex.Therefore,based on the driving force between the solution concentration and the equilibrium concentration,a simple empirical kinetic model can be written as:

When thevalues of minEq.(3)are1 and 2,respectively,twokinetic models can be obtained as follows:

and

where

and

In this paper,Eqs.(4)and(5)are known as the first-order empirical kinetic model(FEKM)and second-order empirical kinetic model(SEKM),respectively.For ion-exchange,nowadays the kinetic models include many mechanism models[33]and some empirical models[34].According to the bottle point method in the literature[34],we deduced four kinds of ion exchange kinetic models,which include the first-order exchange first-order saturation model(FEFSM),first-order exchange second-order saturation model(FESSM),second-order exchange first-order saturation model(SEFSM)and second-order exchange second-order saturation model(SESSM).These ion exchange kinetic models are based on the hypothesis of two processes,that is,the ion exchange process and the active spot saturation process.In our work,the four kinds of ion-exchange kinetic models mentioned above can be linearly expressed as follows:

Fig.5.Isothermal absorption/desorption curve of nitrogen.

Fig.7.Fitting effects of kinetic models:a.FEKM,b.SEKM,c.FEFSM,d.FESSM,e.SEFSM,f.SESSM.

Using Excel software,the model fitting effects for Eqs.(4),(5),(8)-(11)and the optimum parameters are shown in Fig.7 and Table 2,respectively.The effect of the kinetic model cannot be simply evaluated fromtheR2ofitslinearizationequation.Itmustbejudgedbytheconsistency between the model concentration and the experimental concentration.Therefore,the concentration calculated by various kineticmodelswith theoptimumparameters and theexperimentalconcentration are shown in Fig.8.The agreements between the model value and the experimental data are shown in Table 3.

Table 2 Kinetic model parameters for Cu(II)and Ni(II)

According to the slope,regression index(R2)and mean relative error(MRE)of the linear equation between the model value and the experimental concentrations(Fig.8 and Table 3),it is clear that the ion exchange kinetic models have good fitting effects,the MRE between the experimental data and the model values of the four ion exchange kinetic models ranges from 2.29%to 3.47%.But the other two empirical kinetic models(FEKM and SEKM)also have good fitting effects to the experimental data,and the corresponding MREs are in the range of about 2.14%to 4.00%.For Cu(II),SESSM is more suitable for the removal process of copper ions,and the slope and R2for the linear in Fig.8a are 0.990 and 0.992,respectively,and the MRE between the experimental data and the model values is 2.85%.But for the removal process of nickel ions,the empirical kinetic model of FEKM is more applicable,and the slope and R2for the linear in Fig.8b are 0.999 and 0.993,respectively,and the MRE between the experimental data and the model values is about 2.14%.

According to the best kinetics for Cu(II),the calculated initial concentration(C0),k and kaare 46.18 mg·L?1,5.54 × 10?4L·mg?1·min?1and 1.22× 10?3,respectively.Obviously,the calculated initial concentration(46.18 mg·L?1)is less than the initial experimentalconcentration (100 mg·L?1),which indicatesthatthe precipitation phenomena of copper ions exist in theremovalprocess.According to the solubility product of cupric hydroxide[35]and the calculated initial concentration,thesolution pH in the Cu(II)removal process can be determined as about 5.9.For Ni(II),the calculated initial concentration(C0)and k1for FEKM are 82.48 mg·L?1and 4.42× 10?3min?1,respectively.Similarly,there are precipitation phenomena in the removal process of nickel ion.And the solution pH is about 7.8 based on the solubility product of Ni(OH)2[36].The values of solution pH are basically consistent with the experimental conditions.

Fig.8.Models'and experimental concentrations:a.Cu(II),b.Ni(II).

Table 3 Fitting effect of calculated and experimental concentrations

3.3.2.Equilibrium concentration

3.3.2.1.Influencesofconcentration and temperature.Theeffectsof theinitial concentration of metal ions and temperature on the equilibrium concentration are shown in Fig.9.Obviously,the higher the initial concentration is,the larger the equilibrium concentration is,at the same time,the greater the apparent removal ability of the synthetic samples is.Under the same initial concentrations,the removal amount of Cu(II)or Ni(II)on the zeolite increases as the temperatures rises,proving that the higher temperature is favorable for the removal of Cu(II)and Ni(II).In general,for pure physical adsorption,the increase of temperature will reduce the adsorption capacity.Hence an ion exchange reaction must exist during the removal process.This phenomenon further proves the rationality of the former ion exchange kinetic models.

The isotherm data also indicates the apparent capacity of Cu(II)on zeolitePishigherthanthatofNi(II).This ispossiblyattributedtotheaffi nityofzeolitesfordifferentmetalions.Theresultisconsistentwiththefindings by the literature[37],which discovered that the adsorption capacityfor differentcationsonzeolite P follows theorder of Ba2+＞Cu2+＞Cd2+＞Co2+＞Ni2+.

3.3.2.2.Thermodynamic models.According to the kinetic models,the ion equilibrium process not only contains adsorption,but also precipitation and ion exchange.Thus,the ion saturation process is very complex.In order to simplify the treatment process,the classical adsorption isotherm models including Langmuir and Freundlich were used to deal with the saturated concentration.Langmuir and Freundlich isotherms can be expressed as follows[38]:

Fig.9.The saturated capacity vs equilibrium concentration:a.Cu(II),b.Ni(II).

Fig.10.Isotherm fitting of Cu(II)on zeolite P:a.Langmuir,b.Freundlich.

Fig.11.Isotherm fitting of Ni(II)on zeolite P:a.Langmuir,b.Freundlich.

The effects of different isothermal models for Cu(II)and Ni(II)are shown in Figs.10 and 11,respectively.Corresponding model parametersarelistedinTable4.ForCu(II)andNi(II),theLangmuirmodelfitted well with experimental data,the average correlation indexes of R2are 0.979 and 0.990,respectively.

Table 4 Parameters of isothermal adsorption model

For the Langmuir model,qmof Cu(II)is much higher than that of Ni(II),indicating that the zeolite P has better effect on the removal of Cu(II)than Ni(II).At 65°C,the maximum theoretical removal amounts for the two kinds of ions are 138.1 mg·g?1and 77.0 mg·g?1,respectively.Thesephenomenaare in accordance with thekinetic experimental observation.For the Freundlich model,the values of n are within the range of 2-10,indicating the synthetic zeolite can well remove copper and nickel ions.

4.Conclusions

The zeolite P was successfully synthetized from FA through the hydrothermal method.Characterization indicates it has good quality and appearance.The main conclusions are as follows.

(1)Higher temperature hydrothermal reaction is beneficial to the synthesis of zeolite P and the participation of quartz in zeolite synthesis,but it easily forms sodalite.

(2)The second-order exchange second-order saturation model(SESSM)can well fit the removal processes of Cu(II)from watersolutionbythesyntheticsample,whilethefirst-orderempirical kinetic model(FEKM)is the best kinetic model for Ni(II).

(3)Thesyntheticproducts caneffectivelyremoveheavy metals such as Cu(II)and Ni(II),and higher temperature is favorable for the removal.TheLangmuirmodelisinbetteragreementwiththeexperimental equilibrium data than the Freundlich isotherm,and the apparent theoretical removal capacities for Cu(II)and Ni(II)by the synthetic samples can reach to 138.1 mg·g?1and 77.0 mg·g?1,respectively.

Nomenclature

C ion concentration,mg·L?1

C0initial ion concentration,mg·L?1

Ceequilibrium concentration,mg·L?1

KFFreundlich isotherm constant,mg·g?1

KLLangmuir isotherm constant,L·mg?1

k exchange rate constant,min?1or L·mg?1·min?1

kasaturation rate constant

k1empirical kinetic rate constant,min?1or L·mg?1·min?1

n Freundlich constant

qeequilibrium capacity,mg·g?1

qmmaximum capacity,mg·g?1

qtcapacity at the t moment,mg·g?1

V initial solution volume,L

w dosage of zeolite,g