Zu Rong Ang, Ing Kong*, Rachel Shin Yie Lee, Cin Kong,Akesh Babu Kakarla Ai Bao Chai, Wei Kong

（1. Department of Mechanical, Materials and Manufacturing Engineering, University of Nottingham Malaysia,Jalan Broga, 43500 Semenyih, Selangor, Malaysia;2. School of Computing, Engineering and Mathematical Sciences, La Trobe University, Bendigo, Victoria 3552, Australia;3. School of Pharmacy, University of Nottingham Malaysia, Jalan Broga, 43500 Semenyih, Selangor, Malaysia;4. Faculty of Engineering, Science and Technology, Infrastructure University Kuala Lumpur, Corporate Block,Unipark Suria, Jalan Ikram-Uniten, 43000 Kajang, Selangor, Malaysia）

Abstract： Novel hybrid aerogels were prepared by adding ZnCl2, NiCl2·6H2O, FeCl2·4H2O and FeCl3·6H2O to a suspension of equal weights of graphene oxide and oxidized carbon nanotubes, followed by co-precipitation under basic conditions. The aerogels were then crosslinked with polyvinyl alcohol in water and freeze-dried. They consisted of magnetic Ni0.5Zn0.5Fe2O4 nanoparticles,graphene oxide, carbon nanotubes and polyvinyl alcohol, which have active sites that attract dye molecules and could be extracted from water by applying a magnetic field. Using optimum mass ratios of ZnCl2/NiCl2·6H2O/FeCl2·4H2O/FeCl3·6H2O/(graphene oxide+oxidized carbon nanotube) at 6∶6∶12∶12∶1, the hybrid aerogel has a high adsorption capacity of 71.03 mg g?1 for methylene blue and a moderate magnetic strength of MS = 3.519 emu g?1. Its removal efficiencies for methylene blue, methyl orange, crystal violet and a mixture of equal masses of the three were 70.1%, 4.2%, 8.9% and 11.1%, respectively for the same dye concentration of 0.025 mg mL?1. It could be used for 3 regeneration cycles with a regeneration efficiency of over 82%. It was also not toxic to the living organisms, suggesting that it is a promising adsorbent for treating industrial wastewater.

Key words: Magnetic aerogel；Dye removal；Graphene；Carbon nanotubes；Caenorhabditis elegans (C. elegans)

1 Introduction

Conventional industrial dyes used in textiles,leather, printing, plastic, food, pharmaceutical and cosmetics, can cause serious environmental issues[1,2],which are toxic and carcinogenic[2]. Several methods have been developed to remove dyes, including biodegradation, filtration and adsorption[3–5]. Adsorption is popular owing to its high flexibility, efficiency and compatibility to various chemical compounds[6].

The most commonly used adsorbents are silica[7]and activated carbon[8,9]. More recently, graphenebased adsorbents have been discovered, which have large surface and abundant active sites, and are promising for dye adsorption[10,11]. Graphene oxide (GO)-magnetite-chitosan nanocomposite[12]and GO-oxidized carbon nanotube (O-CNT) aerogel[10,13]are developed for this purpose recently. Their adsorption capacities, rates and regeneration efficiencies depend on their chemical components and microstructures.Modification of the adsorbent magnetic by incorporating magnetic nanoparticles such as Fe3O4[14]and ZnFe2O4[15]makes their separation easy from liquid phase by using external magnetic field.

Using polyvinyl alcohol (PVA) as a cross-linking agent strengthens the structure of the 3D porous graphene-based aerogel. In addition, its hydroxyl groups interact with polar functional groups of graphene-based materials, providing new adsorption sites for dye removal[16]. Previous studies have shown that the properties and dye adsorption capability of PVA hydrogel are improved by incorporating graphene-based materials[5,17–19].

A graphene-carbon nanotube hybrid magnetic polyvinyl alcohol aerogel (3DmGT-PVA) that contains GO, O-CNTs, PVA and Ni0.5Zn0.5Fe2O4nanoparticles was developed by a combined in-situ deposition and low temperature co-precipitation method. No one has previously incorporated nickel zinc ferrite into a graphene-based aerogel for dye extraction.

2 Experimental

2.1 Materials

Graphite fine powder was obtained from Synergy Scientific Sdn Bhd, Malaysia. Multiwalled carbon nanotubes (CNTs) (Nanocyl NC7000 series; average diameter 9.5 nm; average length of 1.5 μm; purity 90%; and surface area 250-300 m2g?1) were provided by Nanocyl SA, Belgium. Potassium permanganate(KMnO4, 99%), 30% hydrogen peroxide solution(H2O2), iron (II) chloride tetrahydrate (FeCl2·4H2O),zinc chloride (ZnCl2), polyvinyl alcohol (PVA, Mn80 000), sodium hydroxide pellets (NaOH, reagent grade), methylene blue (MB), crystal violet (CV) and methyl orange (MO) were obtained from R&M Chemicals, Malaysia. 96% sulphuric acid (H2SO4),37% hydrochloric acid (HCl), 70% nitric acid (HNO3)and nickel (II) chloride hexahydrate (NiCl2·6H2O)were obtained from Fisher Scientific (M) Sdn Bhd,Malaysia. Iron (III) chloride hexahydrate (FeCl3·6H2O) was provided by Bendosen Laboratory Chemicals, Malaysia.

2.2 Preparation of GO

Graphene oxide (GO) was synthesized via the modified Hummers’ method[20]. 2.5 g graphite and 2.5 g NaNO3were mixed with 120 mL 96% H2SO4. The mixture was stirred for 1 h in an ice bath. 15 g KMnO4was added to the mixture by keeping the temperature below 15 °C and stirring for 2 h. After adding 100 mL deionized water, the mixture was stirred for 1 h at 35 °C. After that, another 150 mL deionized water was added and the mixture temperature was raised to 90 °C. The mixture was stirred for 1 h before 50 mL 30% H2O2solution, 100 mL deionized water and 100 mL 10% HCl were added to the suspension.The suspension was left for 10 min. The precipitate was washed with deionized water and centrifuged repeatedly until pH=7, and finally dried at 80 °C for 24 h.

2.3 Preparation of O-CNT

2.5 g multiwalled CNTs were dispersed in 150 mL 70% HNO3solution in a round-bottom flask in an ultrasonic water bath for 1 h. The suspension was refluxed at 80 °C under vigorous stirring for 24 h,washed with deionized water and centrifuged repeatedly until pH=7. Then it was dried at 80 °C for 24 h to obtain O-CNTs.

2.4 Synthesis of 3DmGT hybrids

3DmGT hybrids were synthesized using the combined in-situ deposition and low temperature co-precipitation method[12]. 20 mg GO and 20 mg O-CNTs were mixed into 20 mL deionized water and sonicated, forming homogeneous suspension of GO and OCNTs (GT). Various mass of ZnCl2, NiCl2·6H2O, Fe-Cl2·4H2O and FeCl3·6H2O with a constant molar ratio of 1∶1∶2∶2 were dissolved into the suspension under a constant flow of nitrogen. The mixture was stirred for 30 min. After 1.0 mol L?1NaOH solution was added dropwise to the mixture until pH 11, it was stirred for 30 min. The precipitate was washed with deionized water until pH 7 and dried at 60 ℃ for 24 h.3DmGT hybrids were assigned from F1 to F6, with mass ratios of FeCl3·6H2O to GT at 4∶1, 8∶1,12∶1, 16∶1, 20∶1 and 24∶1, respectively.

2.5 Synthesis of 3DmGT-PVA aerogels

The aerogels were fabricated using the ice-templated method[21]. PVA was dissolved in water and heated at 70 ℃ for 2 h to form a homogeneous 10 mg mL?1PVA solution. 1 g 3DmGT powder was mixed with 2 g PVA solution. The mixture was sonicated for 1 h, forming a homogeneous suspension. It was then poured into a 50 mL mould and freeze-dried for 48 h to form 3DmGT-PVA aerogel. Six 3DmGTPVA aerogels were prepared from six 3DmGT hybrids. For comparison, pure PVA aerogel, PVA aerogel with GT and PVA aerogel with Ni0.5Zn0.5Fe2O4nanoparticles were also prepared.

2.6 Adsorption experiments

MB, MO and CV dye powders were dissolved in deionized water to produce 0.025 mg mL?1dye solution, respectively. The 3DmGT-PVA aerogels with specific mass were added into a specific dye solution.The mixture was manually shaken for 1 min and left for 48 h. By using a magnetic separation method, the aerogels were extracted. The UV-Vis was used to determine the equilibrium concentrations of dye solutions by measuring the peak intensity at 589 nm for CV and peak intensities at various wavelengths for MO. 0.025 mg mL?1pure dye solutions were taken as a reference concentration for calibrating the relationship between absorbance and wavelength.

Eq. 1 and 2 were used to calculate the removal efficiencyηand the amount of dye adsorbed in equilibrium by the aerogelsqe.

Where,Ce(mg mL?1) is the dye concentration at equilibrium,C0(mg mL?1) is the initial dye concentration,V(L) is a volume of the dye solution andmi(g)is the mass of an aerogel.

The dye adsorption experiments were performed on F1 to F6. The best adsorbent was identified for balanced dyeqeand magnetic strength. The different aerogel mass (0.5 to 3.0 mg), different dye pH values(4 to 10) and various dyes (pure and a mixture of 0.025 mg mL?1MB, MO and CV) were then tested.The kinetic adsorption experiment used 50 mL 0.025 mg mL?1MB solution and 67 mg of the selected aerogel at 23 °C and pH 7. At 30 min interval,10 mL of MB solution was collected from the bulk solution. This step was repeated 6 times. The concentration of each MB solution collected at the specified time was measured by UV-Vis. Pseudo kinetic models was then used to study the adsorption rate of the best aerogel.

2.7 Regeneration capability

21.9 mg of the selected aerogel was added to 0.025 mg mL?1MB solution and left for 48 h. It was extracted magnetically and repeatedly washed with 0.3 mol L?1NaOH solution. It was vacuum freezedried for 24 h, then reused by adding it into a new 0.025 mg mL?1MB solution. The concentration was then measured by UV-Vis to investigate regeneration capability.

2.8 Characterization

X-ray diffraction analysis (XRD, Bruker D8 Advance, Bruker AXS, Germany) was used to study the crystal structure of 3DmGT-PVA aerogels, with Cu-Ka radiation (λ=0.154 06 nm). Field emission scanning electron microscopy (FE-SEM, FEI Quanta 400F, accelerating voltage=20 kV) was used to study its morphology. The composition of the aerogel was analyzed by energy-dispersive X-ray spectroscopy(EDX, Oxford Instruments INCA 400 with X-Max detector). To study its thermal stability, 2 mg aerogel was used for thermogravimetric analysis, (TGA,PerkinElmer STA6000) with a constant heating rate of 20 °C min?1from 30 to 955 °C. To determine dye solution concentration, 1.5 mL was extracted to a quartz cuvette for UV-Vis spectroscopy (UV-Vis, Perkin Elmer Lambda 365, dual beam, optical path length=1 cm). The magnetic properties of the aerogels were studied by a vibrating sample magnetometer (VSM, Lake Shore 7404) under an applied magnetic field of 12 000 G at room temperature. The magnetic parameters such as saturation magnetization (MS),remanence (MR), and coercivity (HC) were then determined.

2.9 In vivo toxicity assessment in a Caenorhabditis elegans (C. elegans) model

The toxicity of F3 aerogel to living organism was tested using a live animalC. elegansmodel. AC. eleganslifespan assay was carried out as previously described[22]with minor changes. Briefly, thirty age-synchronized adult worms were manually transferred to three agar plates (n=90) supplemented with 100,200 and 500 μg mL?1F3. A negative control without the addition of the F3 aerogel was run in parallel. The plates were then incubated at 25 °C and the number of surviving and dead worms were scored under a Nikon SMZ745 stereomicroscope every alternate day until the entire population of worms were found dead. The experiment was repeated independently for three times. Comparison was then made between the control and treatment groups using Log-rank (Mantel-Cox) statistical test.

3 Results and discussion

3.1 Characterization of 3DmGT-PVA aerogels

Fig. 1 shows how to make 3DmGT-PVA aerogels. Graphite was oxidized into GO via the modified Hummers’ method. GO and O-CNTs are dispersible in water due to oxygen-containing functional groups such as hydroxyl and carboxyl[10]. They were co-precipitated with Fe3+, Ni2+and Zn2+ions. These functional groups serve as active sites to attach Ni0.5Zn0.5Fe2O4nanoparticles. Once PVA was added,the mixture was vacuum freeze-dried, and the aerogels formed. Fig. 2(a) shows the optical photographs of 3DmGT-PVA aerogels. Fig. 2(b) is a SEM image of F1 aerogel with 200× magnification. The 3D microstructure of the F1 aerogel has an average pore size of ～21.9 μm, lower than that of previous reports of pure GO aerogel (pore size (86 μm))[16]. Connecting graphene layers and PVA chains increases density[16].Fig. 2(c) shows the Ni0.5Zn0.5Fe2O4nanoparticles,CNTs, graphene sheets and clusters of PVA in the F3 aerogel with a magnification of 1500×. A large number of the Ni0.5Zn0.5Fe2O4nanoparticles distributed homogeneously on the graphene sheets and polymer layers can be observed in the low magnification. In Fig. 2(d) at a magnification of 60 000×, the graphene sheets were covered with PVA chains. Wrinkles and folds are seen on the graphene sheets, linked by PVA chains, following the “bricks” (graphene) and “mortar” (PVA) microstructure model[23]. Fig. 2(e) shows the SEM image of F6 aerogel at a magnification of 30 000×. From F1 to F6, the amount Ni0.5Zn0.5Fe2O4nanoparticles in the microstructure increases, as expected. Ni0.5Zn0.5Fe2O4nanoparticles are still embedded in the graphene sheets and PVA chains, the less stacked structure reduced its homogeneity and structural stability, causing slight, but increasing decomposition of the aerogel into powder[22].

As shown in Fig. 3, XRD was used to investigate the crystalline structure of the PVA, GT,Ni0.5Zn0.5Fe2O4nanoparticles and F3 aerogel. In Fig. 3(a), the strong peak at 2θ= 19.8°, which corresponds to (101) of PVA, shows the presence of PVA in semi-crystalline form in the aerogel[24]. The presence of graphene in GT is evidenced by the small peak at 2θ=26.15° (002) for GO and 2θ=13.30°, 42.70°,53.80° and 73.00°, corresponding to (001), (100),(004) and (110) for O-CNTs, respectively[25,26]. The moderate narrow and small peaks are identified as the formation of two-dimensional graphene sheets. Based on Bragg’s law and peak (002), the interlayer spacing of GO is estimated to be a 0.341 nm. Ni0.5Zn0.5Fe2O4nanoparticles show the peaks at 2θ=30.07°, 34.42°,43.09°, 56.94° and 62.56°, corresponding to (220),(311), (400), (511) and (440), respectively[27].

TGA analysis was performed to measure the thermal stability of aerogels. Fig. 4 shows the onset decomposition temperature of PVA aerogel at 230 ℃,it is fully burnt out at 800 ℃. GT aerogel decomposes from 130 to 500 ℃ with 29.4% of graphene left at 995 ℃, which could be ascribed to moisture loss and decomposition of oxygen-containing functional groups on the surface of GO and O-CNTs[28].Ni0.5Zn0.5Fe2O4in the aerogel decomposes apparently from 250 to 310 °C and from 550 to 650 °C with 39.2% of Ni0.5Zn0.5Fe2O4left at 995 °C. 3DmGT-PVA aerogels are thermally stable before 130 °C. The weight loss rate is higher above 230 °C due to the decomposition of PVA and Ni0.5Zn0.5Fe2O4hybrid.Above 650 ℃, graphene and Ni0.5Zn0.5Fe2O4are the main componets in 3DmGT-PVA aerogels. At any temperatures, the weight percentages remained increase from F1 to F5 and are closer to that of Ni0.5Zn0.5Fe2O4due to the increase in Ni0.5Zn0.5Fe2O4content from F1 to F5. Taking the temperature at 995 °C as an example, there is an increase in the Ni0.5Zn0.5Fe2O4content from F1 to F3 and F5.

VSM analysis was conducted to study the magnetic properties of the 3DmGT-PVA aerogels. Fig. 5 shows the magnetic hysteresis loops for the aerogels.The saturation magnetization (MS), coercivity (HC)and retentivity (MR) at room temperature are listed in Table 1. The negligible values ofHCandMRindicate their superparamagnetic nature, which means that the Ni0.5Zn0.5Fe2O4nanoparticles can overcome magnetic anisotropy when thermally activated due to their fine crystalline sizes[29]. From F1 to F6,MSincreases as the amount of Ni0.5Zn0.5Fe2O4nanoparticles increases.TheseMSvalues are high enough for the aerogels to be magnetically attracted against gravity (Fig. 5),demonstrating the ability of 3DmGT-PVA aerogels to be magnetically extracted from treated solution.

Table 1 Saturation magnetization (MS), coercivity (HC)and retentivity (MR) at room temperature for the aerogels.

3.2 Adsorption properties

3.2.1 Mechanism

For wastewater treatment application, the main property of 3DmGT-PVA aerogels is their dye adsorption capability, which can be achieved in multiple ways. First, negatively charged oxygen-containing functional groups on GO and O-CNTs have electrostatic attraction for and form hydrogen bonding with cationic and hydrophilic dye molecules[10,16].Second, graphene layers have conjugated π domains that interact with aliphatic chains and benzene rings in dye molecules. Third, the hydroxyl groups in PVA form hydrogen bonding and electrostatic attraction with dye particles[16]. Finally, Ni0.5Zn0.5Fe2O4nanoparticles have dye adsorption capability that depends on the type of charges on their surfaces[30]. The combination of graphene and Ni0.5Zn0.5Fe2O4nanoparticles allows the aerogels to adsorb dye molecules and can be attracted by an ordinary magnet. F3 aerogel was selected as the optimum aerogel for the rest of the experiments due to its highqe(up to 71.03 mg g?1with MB) with a moderateMsvalue (3.519 emu g?1).

3.2.2 Adsorption properties of the F3 aerogel

The effects of amount of GT and magnetic nanoparticles in 3DmGT-PVA aerogel on itsqeandηwere studied. The photos of MB solutions treated by F3 aerogel and GT are shown in Fig. 6(a). Fig. 6(b)shows the UV-Vis spectra of MB solutions after the adsorption by 3DmGT-PVA aerogels. As shown in Fig. 6(c), F1 aerogel has the highestη=73.2% andqe=135.65 mg g?1while F6 has the lowestη=3.6% andqe=6.42 mg g?1. The obtainedqevalues are lower than that reported by Lee et al.[13], where theqeon MB is up to 626 mg g?1. One of the reasons for the lowerqevalues for the prepared aerogels is due to the higher percentages of PVA and magnetic nanoparticles and the lower percentage of GT in the aerogels, as proven by the TGA results. A lower GT content in the aerogel means a lowerqeas the dye adsorption is caused by the π-π interaction and van der Waals forces between the nano carbon materials and the organic dyes. From F1 to F6,qeandηdecrease because the amounts of graphene layers and CNTs decrease, more adsorption sites are occupied by Ni0.5Zn0.5Fe2O4nanoparticles, hence the total specific surface area is reduced, resulting in lower absorption capability and removal efficiency. The larger the surface area of the structure, the more dye molecules can be adsorbed by the structure. With the highestηandqeof GT aerogel as a control sample, it is found that the dye adsorption capacity is dominated by of graphene other than Ni0.5Zn0.5Fe2O4nanoparticles in the aerogel.

3.2.3 Effect of the mass loading of F3 aerogel on adsorption capacity and removal efficiency of MB

The effect of mass of F3 added onqeandηis presented Fig. 7. The experiment is performed with 6 different mass loadings of F3 hydrogel in 12 mL MB solution with a pH value of 7 and a concentration of 0.025 mg mL?1. As the F3 mass added increases,ηincreases because the amount of graphene increases.However, the trend ofqevalue is unsteady. At the mass loadings of 0.5 and 1.0 mg,qevalues are low with a uncertainty of ±0.20 mg g?1. At the mass loadings above 1.5 mg,qevalues are high with uncertainty of ±0.68 mg g?1. 1.5 mg loading is selected to be optimal for F3 hydrogel as it has the highestqe(71.03 mg g?1) with a moderate magnetic strength.

3.2.4 Effect of the pH value of dye solutions

The effect of pH value of MB solutions onηandqeof F3 aerogel is shown in Fig. 8. The pH values of MB solutions is adjusted by 0.3 mol L?1NaOH or 0.3 mol L?1HCl solutions.ηincreases from 42.6% to 57.06% andqeincreases from 65.20 to 88.23 mg g?1as the pH value increases from 4 to 10. This trend is due to charge behavior of MB molecules as pH value increases. MB has cationic quaternary ammonium groups. There is a higher concentration of H+ion at a lower pH value of MB solutions, which will compete with the cationic MB molecules for the adsorption sites on the graphene layers of F3 aerogel. Also, the protonated functional groups in the aerogel will become positively charged and repel MB molecules. As the solution pH value increases, the H+ion concentration decreases, the functional groups are deprotonated,which reduces the competition between H+ions and MB molecules and enhances the electrostatic attraction[2]. Therefore, the higher the solution pH value, the higher the adsorption efficiency of F3 aerogel.

3.2.5 Adsorption kinetics

An adsorption kinetics experiment was conducted to study the adsorption behavior of F3 aerogel.Fig. 9(a) shows a plot of the adsorption capacity at the specified time,qtagainst timet, with a logarithmic line of best fitting applied to the obtained trend. The adsorption rate is initially high but decreases as the adsorption time increases. The adsorption reaches its saturation point at the time near 180 min, with a saturationqeof 63.05 mg g?1. Two kinetic models are purposed. The Lagergren pseudo-first-order kinetic model is suitable for the case with a high initial concentration of the MB solution[2]. It is expressed as Eq. 3[3].

Wherek1(L min?1) is the pseudo-first-order rate constant. Pseudo-second-order model assumes that the adsorption is chemically dominated, which involves electron exchange and sharing between aerogel and dye molecules[31]. It is expressed as Eq. (4)[2].

Wherek2(g mg?1min?1) is the pseudo-second-order adsorption rate constant. Fig. 9(b) and 9(c) show the fitting results based on these models. The values ofqe,k1andk2are determined from the intercepts and gradients of these graphs and are listed in Table 2.The percentage error ofqeis evaluated withqe=71.03 mg g?1(F3, from Fig. 6(c)) as reference. The initial adsorption rate,V0(mg g?1min?1) is determined by Eq. 5[32].

Table 2 Kinetic parameters for F3 aerogel.

Theqefor the pseudo-second-order kinetic model has a percentage error of 19.3%. This is due to uneven distribution of nanofillers in F3 aerogel, which varies the composition of active sites. The data are considered to obey this kinetic model when the fitting has a higher correlation coefficient (R2>0.95), which means that chemisorption is involved and diffusion of MB molecules through the pores of F3 aerogel limits its adsorption rate[33].

3.2.6 Regeneration of 3DmGT-PVA aerogel

The regenerative capability of F3 aerogel was studied by performing a series of MB adsorption experiments. NaOH solution was used as an eluent during dye desorption[23]. Four adsorption-desorption cycles are performed with results presented in Fig. 10.ηris relative toηfor the first cycle.ηof F3 aerogel is restored with a slight decrease inηrat the first threecycles (ηr=82.0% at 3rdcycle). Such an outcome is similar to other graphene-based magnetic aerogels reported[12,34]. At the 4thcycle,ηris significantly reduced to 52.3%. These prove a high reusability of F3 aerogel in the first three cycles.

3.2.7 Adsorption for other dyes and their mixture

The ability of F3 aerogel to adsorb MB from water mixed with single dye and with a combination of dyes is investigated. The F3 aerogel is left in three single dye solutions with the concentration of 0.025 mg mL?1for 48 h. The resultingηvalues in MB(22.6%) and in CV (8.9%) are more significant than that in MO (4.2%), as shown in Fig. 11. MO molecules are anionic, forming repulsive electrostatic forces with negatively charged functional groups on the graphene surface[35]. Although CV molecules are cationic[36], F3 aerogel has a higher adsorption capability in a highly cationic dye. A dye combination is prepared by mixing an equal mass of MB, MO and CV dyes to form a 0.025 mg mL?1cocktail (Ct) solution.ηvalue of F3 aerogel in theCtsolution is 11.1%.This shows the ability for F3 aerogel to adsorb dye combinations[37].

3.3 Toxicity of F3 in a living organism

As F3 displays a good adsorption behavior (alone and in combination with other dyes) and demonstrates great potential to be used as an adsorbent to treat industrial wastewater, it is crucial to assess the possible toxic effect of F3 towards living organism and the environment. Therefore, the potential toxic effect of F3 was investigated by using a live animal model, the nematodeC. elegans. No significant changes are observed in theC. eleganslifespan when the worms are exposed to 100 , 200 and 500 μg mL?1of F3 as compared with the control group (p>0.05) as shown in Fig. 12, suggesting that F3 is safe to a biological organism. This further supports the use of F3 as an effective adsorbent to remove dyes from industrial wastewater as it may not harm the living organism in water and environment.

4 Conclusion

Graphene-based aerogels, which incorporate nickel zinc ferrite to introduce magnetic properties to improve dye adsorption performance, were prepared.The SEM images show the 3D porous microstructure of the aerogels with PVA chains that strengthen the structure. Its highqeis contributed by electrostatic attraction, π-π interactions and hydrogen bonding between dye molecules and active sites in aerogels.As the mass ratio of FeCl3·6H2O:GT increases, its magnetic strength increases, but itsqeand structural strength decrease. F3 aerogel with a mass ratio 12∶1 is the best due to its highqe(up to 3.77 mg g?1with MB) and moderateMs=3.519 emu g?1. It has a high MB removal efficiency (η) of 70.1% with a high aerogel mass loading of 3.0 mg mL?1. In a basic dye solution at a pH value of 10, theηis 17.1%. Its adsorption follows the pseudo-second-order kinetic model. It can be easily regenerated by washing with a NaOH solution while maintainingηfor the first three cycles(ηr=82.0% at the third cycle). It can be also used to adsorb MO (η=4.2%), CV (η=8.9%) and a mixture solution of MB, MO and CV with a relatively high efficiency (η=11.1%). Additionally, the aerogel prepared in this study is found to be non-toxic to a living organism. These results suggest that the aerogel may have application potential in dye adsorption for industrial wastewater treatment. Future work will be conducted to test the aerogel for adsorption of polluted water produced by industrial dyes, to develop processes for magnetic separation of the aerogel, which comply with existing industrial infrastructure, and potentially even to develop processes for dye reclamation.

Acknowledgements

The authors gratefully acknowledge the facilities,the scientific and technical assistance of Engineering Research Department, University of Nottingham Malaysia and School of Applied Physics, National University of Malaysia.