Nano-zeolite-graphene oxide composite for calcium hardness removal: isotherm and kinetic study

Water quality plays a vital role for its use in industrial, domestic as well as agricultural purposes. Zhi & Zhang (2014) reported interior water quality with respect to chemical contaminants; that is, hardness, nitrate, heavy metals and soluble ions, in urban as well as a rural area of the world (Amarasooriya & Kawakami 2019; Ahamad et al. 2020). Hardness is a major challenge for groundwater, which is ensured to be the source of water for various purposes. It creates the problem of scaling of boilers, washing machines, and pipes, and difficult lathering for industries. Hard water also creates health problems in human beings. Disorders like urolithiasis, cardiovascular disorders, kidney problems, and even cancer were reported by Rolence (2016). Currently, several techniques have been used to solve the problems such as the deionization process, adsorption, lime soda, nano-filtration (Madarasz et al. 2014), microbial and ion exchange processes. The natural zeolites have shown good adsorption properties and are used for purification of water (Gougazeh & Buhl 2013; Madarasz et al. 2014).

Graphene is a mono-atomic layer of graphite which is an emerging star on the horizon of material science in the 21^st century. Geim & Novoselov (2014) synthesized two-dimensional graphene material for the application of ground-breaking experiments. The unique physicochemical characteristics of graphene oxide; that is, its large surface area (2,600 m² g⁻¹), electrical mobility (em²v⁻¹s⁻¹), conductivity (500 W m⁻¹ K⁻¹) and mechanical properties have made graphene oxide an ideal adsorbent (Maab et al. 2016).

Being a carbon nanomaterial, graphene oxide contains a number of active functional groups like carboxyl, carbonyl, epoxy and hydroxyl which makes graphene a strong candidate for target-specific adsorption (MeiJiao et al. 2013). The study by Soltani & Lee (2016) reveals that, due to their high surface area, negative surface charge and highly mobile π-electrons, graphene, graphene oxides, and its composites are efficient adsorbents and are widely used for environmental applications (Kar et al. 2016).

Although fly ash has wide applications (Koshy & Singh 2016), mainly about 20% of total fly ash generated from the thermal power plant is used in the cement production (Zhang et al. 2011; Izidoro et al. 2013). A large quantity of fly ash is dumped in land with resulting harmful effects on the environment (Molina & Poole 2004). Fly ash mostly consists of Al and Si in the form of aluminosilicate, with traces of Fe, Na, K, Ca, Ti and S. It also contains crystalline minerals in minor quantities such as mullites, lime anhydrite and quartz (Soni & Shukla 2018). However, the adsorption efficiency of fly ash is lower in comparison to the commercial adsorbents, consequently (Hong et al. 2017), the experiments are being performed with various chemical reactions for the improvement of the fly ash adsorption capacity (He et al. 2016).

Zeolites are a significant group of crystalline aluminosilicate crystalline materials and they have generally used in adsorption of contaminants and ion exchange process (Ferrarini et al. 2016). It is mainly composed of crystalline solid structures made of silicon, aluminium, and oxygen. Zeolites naturally occur in small quantities only, hence synthetic zeolites are very popular for their catalytic characteristics and other properties like ion exchange and adsorption (Liu et al. 2011).

Literature on the removal of hardness from water by zeolite is scarcely available. The present work demonstrates a novel synthesis method for nano-zeolite-graphene oxide (nZGO) composite from fly ash to remove calcium hardness from water (Liu et al. 2016; Hailu et al. 2019). The natural zeolites have shown good adsorption properties and are used for purification of water, but the limited reserves of zeolite do not permit its application at the commercial level (Gao et al. 2011). The fly ash derived nano-zeolite-graphene oxide composite is a better adsorbent for hardness removal from water. The synthesized nZGO composite may also be investigated as a catalytic converter for auto exhausts.

Graphite powder, sodium nitrate (NaNO₃), sodium hydroxide (NaOH), potassium permanganate (KMnO₄), and hydrochloric acid (HCl) were purchased from Fisher Scientific, India. Sulphuric acid (H₂SO₄) and hydrogen peroxide (H₂O₂) were supplied by Merck. All reagents were used without any further purification.

GO was synthesized using the modified Hummer's method. For this, concentrated sulphuric acid (H₂SO₄) 69 ml, 3 gram of graphite powder and NaNO₃ (1.5 gram) were added in the round bottom flask at less than 5 °C for 20 minutes with continuous stirring. Then subsequently 9 gram of potassium permanganate (KMnO₄) was added slowly for 1 hour. The reaction was stirred for 7 hours at 35 °C. After 7 hours, 9 gram of potassium permanganate was added slowly for 15 minutes and the reaction was stirred for 12 hours at room temperature. The gradually brown color was developed which indicates the complete oxidation of graphite (Marcano et al. 2010; Zamani & Tabrizi 2016). The reaction mixture was cooled at room temperature and transferred into an ice bath with the addition of 300 ml distilled water. Finally, the brown suspension was further treated with 10 ml hydrogen peroxide (H₂O₂) to reduce residual oxidant, the brown suspension turns to pale yellow. This suspension was highly acidic (pH <4). The mixture was repeatedly washed with dilute ammonium hydroxide solution followed by distilled water until the mixture attained pH 7. The final yield was 5.1 g dried solids after drying in the oven at 60 °C. The mother liquor was analyzed on the atomic absorption spectrophotometer, to ensure that the graphite oxide was free of sodium, potassium and manganese; the solution was free of the trio.

The fly ash <120 μm particle size was obtained by sieving for the synthesis of nano-zeolite. The dissolved iron oxides were removed by washing the fly ash with 10% HCl for 12 hours. Pre-treated 15 gram of dry fly ash was mixed in 50 ml 8 N NaOH in a Teflon-lined autoclave reactor. The hydrothermal reaction was carried out at 130 °C for 3 hours. The autoclave reactor was cooled until it reached room temperature, and the mixture was washed with distilled water up to pH 7 and dried in the oven at 80 °C for 3 hours (Chen & Lu et al. 2018; Zhao et al. 2018).

250 mg of graphene oxide was mixed in 50 ml of distilled water for ultra-sonic exfoliation for 3 hours. Subsequently, 5 gram of nano-zeolite was added to the exfoliated graphene oxide solution and stirred for 6 hours. The solution of the reaction was filtered and washed several times with distilled water. The reaction mixture of the reaction was dried in the oven at 60 °C for 3 hours. To ensure composite formation, the process was undertaken three times.

Investigate the functional groups, Fourier Transformation Infrared spectroscopy (FTIR) (Shimadzu IR Affinity-1) with intensity from 400 to 4,000 cm⁻¹ was applied. The crystallographic study was done on an X-ray diffractometer (LabX XRD-6100) with monochromatized Cu-Ka (λ = 0.154056 nm) radiation at 36 kV, 20 mA samples were scanned in the range of 5–80° with continuous scanning mode and scanning rate was set at 2°/min. The morphological characteristics of the sample were investigated using Field Emission Scanning Electron Microscopy (Hitachi FESEM S-4800) with voltage accelerated voltage at 20 keV. The composition of the elements of the samples were studied using Energy dispersive analysis of X-rays (EDX, Bruker Nano). The chemical composition of fly ash was carried out with the help of an X-ray fluorescence spectroscope with glass tablets FLX-SP1, Voltage/kV: 38.0, Current/μA: 15.0 for MCA peak calibration (ED-XRF SPECTRO XEPOS III).

In order to find the batch adsorption behaviour of calcium hardness onto nZGO, the experiments were performed at room temperature. The standard stock solution of calcium hardness was made by dissolving 3.6 gm of calcium chloride in 1,000 ml DI water at neutral pH. The working solution of the required concentration was prepared by serial dilution of the stock calcium hardness solution. For the evaluation of the hardness removal, different amounts of nZGO (250–1,000 mg/L) were mixed into 100 ml calcium hardness solution of known concentration (250–1,000 mg/L) and the mixture was stirred using an orbital shaker at 110 rpm for different time intervals. Subsequently, the supernatant was separated from the mixture by filtration, and the residual calcium hardness was determined by titration with the EDTA method.

The pH optimization was performed with the nZGO (250 mg/L) at different pH (1–14). The pH of the solution was attuned using 0.1 N HCl/NaOH before the addition of nZGO. For dose on optimization, different amounts of adsorbents (250–1,000 mg/L) were used and an optimum adsorbent dose of 1,000 mg/L at 110 rpm stirring speed was observed. All the experiments were performed three times and the mean was considered as a final value.

The XRF results reveal the chemical composition of the fly ash used for the preparation of nZ. The major constituents were silica (76.68%), aluminium (17.33%), calcium (1.6%) and titanium dioxides (1.42%) with other elements in trace quantity. The loss on ignition was found to be 47.52%. Table 1 also shows the chemical composition of major and minor elements obtained by XRF analysis. In general, results are similar for most of the elements.

Figure 1 shows the FTIR spectra of GO synthesised by the modified Hummer's method. The FTIR spectrum of GO shows strong and broad O-H stretching vibration band at 3,630 cm⁻¹, 3,439 cm⁻¹, -OH from (H₂O) at 3,190 cm⁻¹, at 2,824 cm⁻¹ for –OH of carboxylic group, carboxyl C = O stretching band at 1,737 cm⁻¹, aromatic C = C at 1,574 cm⁻¹ C-O-H deformation band at 1,386 cm⁻¹, 1,196 –OH (Phenolic), C-C skeletal stretching vibration at 1,053 cm⁻¹, and C-O skeletal stretching band can be seen at 968 cm⁻¹, 828 cm⁻¹ and 698 cm⁻¹.

The FTIR spectra in Figure 2 of nano zeolite are helpful for the identification of bands related to the construction of the aluminosilicate framework of zeolites. The asymmetric stretching vibrations bridge bonds of Si-O-Al are assigned to the band at 1,006 cm⁻¹ and around 670 cm⁻¹ is symmetric stretching vibrations of bridge bonds (Mozgawa 2001). The characteristic single band of zeolite around 1,006 cm⁻¹ has resulted in a hydrothermal reaction of fly ash with NaOH. The bands at the wave number 489 cm⁻¹ are the characteristic bending vibration of oxygen hexagonal rings present in zeolites.

The FTIR spectra in Figure 3 depict the bands of the nZGO. In this spectrum, the nano zeolite crystal framework dominates, whereas the O-H of the graphene oxide is reduced to a considerable extent and the band for the hydroxyl stretching vibration at 3,439 cm⁻¹ is hardly visible. The other bands of GO are observed as at 1,737 cm⁻¹ (C = O), 1,386 cm⁻¹ (-C-OH), and (-C-C-) 1,053 cm⁻¹ (Xia & Gerson 2018). The asymmetric stretching vibration of bridge bonds in nano zeolite for Si-O-Al is also present around 1,016 cm⁻¹ with symmetric stretching vibration at 670 cm⁻¹ (Agarwal & Rani 2018). All these observations validate the formation of the nano-zeolite-graphene oxide composite. The reduction of the hydroxyl group in the nZGO indicates the reducing nature of the nano zeolites; for further clarity, a detailed experimental study is necessary (Jha & Singh 2018).

The XRD patterns of the solid products of GO, nZ, and nZGO composite are shown in Figure 4. The XRD pattern of graphene oxide shows diffraction peaks at 12.5° and 43.2°. The values of the diffraction peaks are in good agreement with the previously reported literature (Gebhardt et al. 2014). The peak intensity at 25.97° in the nZGO composite is reduced considerably in comparison to pristine GO. After the Reitvield refinement in PANalytical High ScorePlus V. 30, search and match of peaks revealed the structure of the nano zeolite as almost similar to wonesite. The crystal structure was formed with the help of the data processing software VESTA (Figure 4(b). The lattice type was C with the symmetry C2/m, the unit cell dimensions were found to be a = 6.00, b = 10.60, c = 9.6 Å, α = 90, β = 104.100, γ = 90° and the unit cell volume was 592.16 Å3. From Figure 4(b) it is clear that the hexagonal oxygen ring formation is present in the structure. The synthesized nano zeolite shows a triclinic structure in which the two tetrahedral sheets are set apart through the interlayer region. The interlayer region is partially occupied by sodium ions (Kogure et al. 2005).

The surface characteristics of graphene oxide were studied using FESEM. The graphene oxide shows a layered and three-dimensional nature (Figure 5(a)), thick edges of graphite have exfoliated after oxidation. The image shows the multilayer graphene oxide sheet as well as folding at the edges. Energy dispersion X-ray analysis was carried out for estimation of the elemental composition (carbon and oxygen) in the sample. The increased oxygen content in GO can be seen in the inset table of Figure 5(b).

FESEM images of fly ash show the irregular spheroidal crystal morphology, it has spherical particles with smooth surfaces and edges (Figure 5(c)). The spheroidal fly ash particles were transformed into smooth edged-irregular elliptical and spherical shaped nano-zeolites (Figure 5(e)). This shape and size transformation provides a large surface area with highly active sites (Figure 5(j) and 5(k)).

The rough surface of the nZ adsorbent resulted in a comparatively increased surface area compared to fly ash. The average diameter of the nZ was found to be approximately 20 nm. The size of the fly ash particles size reduces significantly after hydrothermal reaction with sodium hydroxide and forms nano zeolites. The EDX results (Figure 3(f)) indicates the replacement of aluminium by sodium in the fly ash. Figure 3(f) also indicates the elemental presence of Na, Al, Si and O in nZ.

FESEM observation of nZGO shows that the nano zeolites are uniformly coated with GO sheets as a result of pressure and temperature during the hydrothermal reaction. The bright crystals of the nZ turned black in color after the hydrothermal treatment. The composite formation is evidenced by the color change as well as the EDX analysis results (Figure 5(h)). The GO sheets with hydrophilic nature have favourable interaction with nano zeolite crystals. Figure 6(a) is the histogram of the nZ particles size, it reveals the particles ranged between 10 and 130 nm; however, the particle size 10–40 nm is of maximum frequency. In Figure 6(b), the histogram of the nZGO particle can be seen with a size distribution between 10–140 nm; the particle size of 10–30 is dominated by high frequency.

The pH of the calcium hardness solution is a significant influence on the adsorption process, which is directly affected by the functional groups present on the adsorbent surface by changing the charge of the adsorbent surface.

The effect of contact time was studied at room temperature and with intervals of 15–300 minutes. Figure 9 shows the relationship between adsorbent adsorption time and calcium absorption efficiency. It was revealed that calcium removal is rapid for 15–120 minutes and thereafter removal capacity is at a slower rate. The percentage of calcium hardness removal comes to equilibrium within 120 minutes. Further increase in contact time did not show significant change in hardness removal. This may explore the adsorption on the outer part of the adsorbent surface (Geethamani et al. 2013).

The effect of initial calcium hardness concentration in the water solution was studied to find the adsorption capacity of nZGO. The experiments were conducted at pH 7 for 120 minutes (Figure 10) duration with four different adsorbent doses (250, 500, 750 and 1,000 mg/L) for four different calcium hardness concentrations, viz. 250, 500, 750, and 1,000 mg/L being used to examine efficiency. It can be observed in Figure 7 that the adsorption capacity declines in both acidic and basic conditions, and the decrease in adsorption capacity may be a result of saturation of the active binding sites on nZGO. The increased dose increases the adsorption up to 1,000 mg/L. With increasing adsorbent dose from 250 to 1,000 mg/L, the adsorption of calcium ions increased from 69 to 98%, respectively. An increase of adsorption of calcium ions with increasing adsorbent dose may be due to the availability of more binding sites for adsorption.

The various parameters of adsorption kinetics are necessary for determining the adsorption rate, with substantial information for modelling and designing the adsorption mechanism (Jayarathna et al. 2015). The kinetics of calcium hardness by nZGO absorption were studied using the pseudo-first-order, pseudo-second-order (Figure 11) and intraparticle diffusion models (Figure 12). Table 3 illustrates the various parameters and correlation coefficients (R² values near 1) values. For the conformity, the predicted values of model and experimental data were expressed by the correlation coefficient. The kinetics equations are given below:

The kinetics study shows that the nZGO obeys the pseudo-second order kinetics and also satisfies the Elovich equation.

The adsorption isotherm study of the adsorption process for water purification is an important step. The adsorption isotherm models were used to investigate the interaction between calcium ions and adsorbent; that is, nZGO. As per Langmuir (1918), is the amount of adsorbate; that is, Ca⁺⁺ ions adsorbed per unit weight of adsorbent (nZGO) and is the equilibrium concentration of adsorbate solution (mg/L), at constant temperature; this is the adsorption isotherm (Jethave et al. 2017). The experimental values of adsorption are tested with Langmuir, Freundlich and Temkin isotherm models. The linearized forms of the equations predict adsorbent-adsorbate behaviour.

The Temkin isotherm model anticipates that the adsorption of the heat of all molecules should decrease linearly with increasing adsorbate overlapping on the adsorbent (i.e. nZGO) surface, and adsorption of adsorbate (calcium ions) is characterized by uniform distribution with maximum binding energy (Kadir & Ismail 2017; Petriciolet et al. 2017). The Temkin isotherm equation is given as Equation (10). Where KT is the equilibrium binding constant (mg/L), T is the absolute temperature in K, and B₁ is for the heat of adsorption. The constants calculated from the Temkin isotherm are shown in Table 3. The heat of adsorption (B₁) value should decrease with increase in adsorbent dosage but the increasing values of (B₁) indicates the Temkin isotherm model is not in agreement with the experimental data.

Where is the calcium ion initial concentration (mg/L), represents the condition for adsorption and is the Langmuir constant⁠. The calculated values in the present study for were found to be between 0.186 and 0.012, which indicates the favourable condition for calcium ion adsorption onto nZGO surface (Figure 14) (Geethamani et al. 2013).

In order to check the interference of other cations and anions in the adsorption process, an open well water sample was collected from the university campus and treated with the nZGO composite. The physiochemical parameters of well water sample are shown in Table 4. Physiochemical parameter such as pH, electric conductivity, calcium and Magnesium hardness, sulphate and nitrate, were determined with the standard methods. The synthesized nZGO (250 mg/L) was introduced in a 100 ml well water sample and stirred for 120 minutes. The results of adsorption can be seen in Table 4; the calcium ion concentration of the well water sample is reduced by 73.33%; this value is obviously less than the adsorption of calcium from the standard solution used in the study. The interference of the Mg⁺⁺ on the adsorption can be seen clearly.

This is also a favourable one, as the magnesium ions are also a hardness contributor. The 73% of calcium hardness, along with 67% of the magnesium from the open well water sample, are removed. Apart from both cations; that is, Ca⁺⁺ and Mg⁺⁺, the anions; that is, nitrate, is reduced by 45% while sulphate is reduced by 54%; on a cumulative scale, the adsorption capacity of the composite is more than the q_max; that is, 217 mg/g. The adsorption process for the real water sample is processed at neutral pH. Though the cationic adsorption on the graphene is favourable at lower pH conditions, in this composite the presence of zeolite molecular sieves governs the adsorption at neutral conditions.

The nZGO composite was synthesized and explored for a possible application for hardness removal. The increased surface area and structural diversity of the synthesized material have demonstrated good adsorption capacity. The possible ion exchange process of the nZGO with the Ca⁺⁺ ions is illustrated in the following Figure 15.

Zeolites normally attract cations by interfacing with the adjacent medium due to their ion-exchange properties. The heavy metal cations like rubidium, caesium, silver, cadmium, lead, zinc, barium, strontium, copper, mercury, cobalt and chromium have a great affinity for ion exchange on the zeolite surface. However, their selectivity by the zeolite framework strongly depends on the hydrated molecular size of the cations, zero point charge of the adsorbent surface and pH of the medium associated with the process. The physicochemical properties; that is, the porosity and oxygen-containing functional groups, have considerable influence on the adsorption of cations. The porous structures of the composite with a large surface area provides favourable diffusion energy levels. The adsorption performance of the native components of the composite; that is, GO and nZ, is not very satisfactory in comparison with the composite (nZGO). This is due to escalated surface interaction provided by the functional groups of GO in the composite with the molecular sieve of the nanozeolite. The study strongly advocates the application of nZGO composite as a promising alternative to the adsorbent materials currently available in the field of water softening treatment.

Table 5 compares various natural and commercial adsorbents' capacity for calcium hardness removal. Among all of the absorbents, the nZGO synthesized in the present study has shown maximum calcium ions removal capacity and the Q_max is 980 mg/L.

The nZGO nanocomposite was synthesized and successfully demonstrated its efficient adsorption potential for calcium hardness. More than 98% of calcium ions were adsorbed from 250 mg/L synthetic solution, at 120 min equilibrium time and 250 mg/L adsorbent dose. The maximum adsorption was observed at neutral pH. The removal efficiency increased with increasing adsorbent concentration. The adsorption study indicated the adsorption of calcium hardness was by the Langmuir and Temkin models, which indicates monolayer adsorption. The kinetic equilibrium data showed that the adsorption process follows pseudo-second-order kinetics and shows the adsorption process is through chemisorption. The intraparticle diffusion model indicated that it is a multi-step adsorption process. The removal of hardness from a real water sample by nZGO is also greater, it is also a good adsorbent for removing other contaminants from water. It is shown that nZGO has the best performance as an adsorbent to remove calcium hardness compared to other composites. In the view of the above, nZGO composite could be used as an effective and economical adsorbent with good practical value in the future for industrial applications.

One of the authors, Mr Ram A. Konale, sincerely acknowledges the financial support provided under DST INSPIRE Fellowship for Research Students, Scheme by Department of Science and Technology, Government of India. Grant number: DST/INSPIRE Fellowship/[IF170674].

All relevant data are available from an online repository or repositories at https://drive.google.com/drive/folders/1fIJLjhk2KfQ8QSwDzlx06o1xY9KvAwiL?usp=sharing.