Novel Fe₃O₄/HNT@rGO composite via a facile co-precipitation method for the removal of contaminants from aqueous system

Abstract

In this study, a novel series of ternary composites of halloysite nanotubes (HNT), Fe₃O₄ nanoparticles (NP), and graphene oxide (GO) was prepared via a facile co-precipitation process. After the simple heat treatment process, the Fe₃O₄/HNT@GO composite was converted to the Fe₃O₄/HNT@rGO composite (FHGC), which was investigated as a potential adsorbent for the removal of contaminants from water. This series of designed ternary systems FHGC-n (n = 1, 2, 3) was identified by SEM, TEM, HRTEM, FT-IR, powder XRD, Raman and X-ray photoelectron spectroscopy (XPS) analyses. As anticipated, for both the simulated organic contaminant species (RhB) and the inorganic contaminant species (As(V) ions) in water, the multifunctional FHGC exhibited good adsorption performance and easy separation. In each case of the FHGC series, the adsorption process for As(V) ions was less than 30 min, which is close to the best example in literature.

---

1. Introduction

Sustainable development has been the “common future” that is aspired to worldwide, since its modern concept was defined in the 1987 Brundtland Report.¹ One aim is to achieve new, clean and powerful energy resources, while scientists still try to improve the utilization of the existing energy sources.^2,3 The treatment of conventional pollution is also one of the most important aspects, prompting researchers to tirelessly design new material and develop the treatment procedures.^4,5 For instance, the pollution of surface water and groundwater has become a tough problem of this century.^6–8 Undoubtedly, water treatment has been a source of challenge and opportunity, especially adsorption techniques for the removal of contaminants from water because of their high efficiency, low cost and operational simplicity.^9,10 In recent years, due to their unique structure and abundant action sites, both graphene oxide (GO) and reduced graphene oxide (rGO) exhibit excellent performance for the adsorption of contaminants in water.^11–14 However, such potential application is hindered by the “too good” dispersibility of GO or the “too bad” dispersibility of rGO in water.^15,16

Herein, we report that one compelling choice for obtaining suitable dispersion properties in aqueous system is to construct organic–inorganic hybridized graphene-based composites.¹⁷ Currently, inorganic Fe₃O₄ nanoparticles (NP) are promising candidates because of several unique advantages, such as low cost, environmental friendliness and fine recyclability.¹⁰ Apart from being a material with good affinity for inorganic ions, Fe₃O₄ NP has also been explored for waste water treatment, such as the removal of As(V). In the recent literature, examples of graphene-based composites incorporated with Fe₃O₄ NP show favorable adsorption behaviors toward organic dyes and inorganic ions. In these cases, the presence of Fe₃O₄ NP hinders the re-stacking of graphene layers to modulate the dispersion properties of composites in aqueous systems, as well as the available area of graphene sheets.¹⁸ On the other hand, the graphene sheet as the substrate material modifies the Fe₃O₄ NP aggregation resulting from its magnetic nature. In comparison to the relatively tiny Fe₃O₄ NP, the size of natural clay halloysite nanotubes (HNT) is on the meso/macroscopic level, and they present negatively charged outer surfaces, and positively charged inner surfaces.^19,20 After surface modification, the HNT@graphene composite is rendered a relatively large specific surface area and big pore volume to improve the total adsorption capacity,⁹ and incorporating HNT species into the Fe₃O₄@graphene binary system may produce larger tunnels to enhance the adsorption rate.^22–24 It is also most likely beneficial for suppressing the aggregation of Fe₃O₄ NP. Hence, the ternary composite of graphene, HNT and Fe₃O₄ NP obtained by rational design can be an excellent candidate for the multifunctional adsorbent material for surface water and groundwater treatment. Such ternary systems may solve the problem of several pollutants coexisting in aqueous solution, and simplify the practical application process to reduce the costs.

Depth consideration is necessary before designing a facile hybridization process for these three species.^25,26 Firstly, the GO sheet with the negatively charged outer surface²⁷ and the nearly neutral rGO sheet,²⁸ would hardly form a composite with HNT without suitable surface modification, because HNT also displays a negatively charged outer surface.²⁹ Secondly, because of the magnetic feature of Fe₃O₄ NP, it easily aggregates and is dispersed with difficulty under mechanical mixing conditions.³⁰ In our previous work, both HNT@rGO⁹ and Fe₃O₄@rGO¹⁰ binary composites were prepared by a convenient method triggered by the electrostatic self-assembly process, and displayed favorable adsorption performance in aqueous systems.

Herein, we develop a more convenient co-precipitation procedure to obtain a series of new Fe₃O₄/HNT@rGO composite materials (FHGC-1, FHGC-2 and FHGC-3) without any surface modification. In this work, iron ions can be easily adsorbed onto GO and HNT to form nucleation centers because of the electrostatic interaction. Furthermore, Fe₃O₄ NP generated on the surface of both GO and HNT will eliminate the electrostatic repulsive forces between the two negatively charged components. In the case of FHGC-n (n = 1, 2, 3), the experiments for the removal of the organic dye rhodamine B (RhB) and the inorganic arsenic(V) ion were respectively investigated in aqueous systems. As we expected, because of the excellent adsorption behaviors and the simple operations of separation and recycling, multifunctional FHGC-n (n = 1, 2, 3) displayed good application prospects in wastewater treatment. In particular, FHGC-2 showed better adsorption performance toward simulated organic contaminant species than FHGC-1 or FHGC-3, while for the inorganic arsenic(V) ion removal, FHGC-1 was best.

2. Experimental section

Chemicals

HNT originated from Henan Province, China. Natural graphite flake (∼200 mesh) was purchased from Sigma-Aldrich. Potassium permanganate (KMnO₄), ferric chloride hexahydrate (FeCl₃·6H₂O), ferrous sulfate heptahydrate (FeSO₄·7H₂O), ammonium hydroxide (NH₄OH, 25% ammonia), hydrogen peroxide (H₂O₂, 30%), concentrated hydrochloric acid (HCl, 36.5%), and concentrated sulphuric acid (H₂SO₄, 98%) were obtained from Tianjin Chemical Co. (Tianjin, China). They were all analytical grade, and were used without further purification. Doubly-distilled water was used throughout.

Synthesis of FHGC

GO (0.8 g) and HNT (0.8 g) were dispersed into 800 mL water to obtain an aqueous suspension, where the GO precursor was prepared by a modified Hummers' method.²⁷ After 2 g FeCl₃·6H₂O was added into this aqueous suspension, it was stirred for 12 h. Subsequently, 1.22 g FeSO₄·7H₂O (mole ratio of Fe²⁺/Fe³⁺ = 1 : 2) was added into the mixture, and then 24 mL of NH₄OH was added to the suspension at 90 °C. A black precipitate was obtained after 1 hour at 90 °C,^32,33 and was separated by centrifugation. The raw product was calcined at 500 °C for 5 h to obtain a new magnetic black powder composite that was named FHGC-1. All the processes mentioned above were conducted under N₂ atmosphere. Similarly, the new composites FHGC-2 and FHGC-3 were prepared through the same co-precipitation procedure, where the different Fe₃O₄ content was controlled by the added weight of FeCl₃·6H₂O and FeSO₄·7H₂O (Table S1†). The yields of the three FHGC samples were 79.7% (FHGC-1, 1.82 g), 73.3% (FHGC-2, 1.51 g), and 67.0% (FHGC-3, 1.23 g).

Materials characterization

Fourier transform infrared (FT-IR) spectra were recorded by a Bruker IFS 66v/S infrared spectrometer, while the Raman spectra were detected by a Renishaw RM-1000 instrument (excited at 514 nm, Ar ion laser ∼5 mW). The phase structures of the samples were identified by powder X-ray diffraction analysis (PXRD, Bruker D8 Advance, λ(Cu_Kα) = 1.5418 Å). X-ray photoelectron spectra (XPS) of samples were obtained by a PHI-5702 multi-functional X-ray photoelectron spectrometer (Physical Electronics Inc., Chanhassen, MN, USA) with pass energy of 29.35 eV and the Mg Kα line excitation source as well as the C 1s binding energy of 284.6 eV. The surface morphologies of composites were measured by a JEOL JSM-6301F instrument. The transmitting electron microscopy (TEM) pattern was performed on a JEOL JEM-2010F electron microscope (Japan) operating at 200 kV. The powder samples for the TEM experiments were firstly dispersed in water in an ultrasonic bath for 5 min, and then deposited on a copper grid covered with a perforated carbon film. M–H measurements were performed to investigate the magnetic behavior of FHGC-2 at room temperature (298 K), with the magnetic field sweeping back and forth between 10 and −10 kOe (1 Oe = 103/4π A m⁻¹ = 79.59 A m⁻¹; M_s is the specific saturation magnetization and H_c is the coercive field). The gas sorption isotherm of FHGC-2 was collected on a Micromeritics 3Flex surface area and pore size analyzer under ultrahigh vacuum in a clean system, with a diaphragm and turbo pumping system. Ultrahigh-purity-grade (>99.999%) N₂ gas was applied in all adsorption measurements. The experimental temperature was maintained by liquid nitrogen (77 K). Prior to measurement, the bulk sample was dried in a vacuum oven at 60 °C for 12 h.

Adsorption measurement of organic contaminant

Rhodamine B (RhB) was used as the model organic contaminant in water. When FHGC-1 (5 mg) was added into the RhB aqueous solution (1.0 × 10⁻⁵ M, 20 mL) at room temperature as the adsorbent, the change in the concentration of RhB over time was monitored by a UV-Vis spectrophotometer (UNIC Corp. UV-2102PC). After the adsorption experiments, the adsorbed sample was easily separated by a small magnet. The RhB adsorption experiments with FHGC-2 and FHGC-3 were executed under the same conditions.

Adsorption measurement of inorganic contaminant

The experiments were performed at ambient temperature at pH = 7.0 ± 0.2 (adjusted by 1 M NaOH or HCl). FHGC-1 (5 mg) was added to the 10 mL As(V) aqueous solution, the concentration of which was in the range of 1–10 mg L⁻¹. After the equilibrium experiments were carried out in a constant temperature oscillation box for 24 h, the sample was separated using an external magnet. The initial and residual concentrations of As(V) were measured using the 8220 atomic fluorescence spectrophotometer. The equilibrium adsorption capacity (q_e) was calculated by using the following equation:

q_e = V(C₀ − C_e)/m (1)

where C₀ and C_e (mg L⁻¹) are the initial and equilibrium concentrations of the adsorbate in solution, respectively; V is the volume (mL) of the arsenic(V) aqueous solution and m is the mass (mg) of adsorbent used in the experiments. Furthermore, the data were analyzed using the Langmuir (eqn (2)) and Freundlich (eqn (3)) isotherms.

C_e/q_e = 1/q_mK_L + C_e/q_m (2)

log q_e = log C_e/n + log K_F (3)

where maximum adsorption capacity of adsorbents q_m (mg g⁻¹) and the free energy of adsorption K_L are Langmuir constants. The Freundlich constant K_F (mg^1−(1/n) L^1/n g⁻¹) is indicative of the relative adsorption capacity of the adsorbent, and n is related to adsorption intensity. In the adsorption kinetics experiment, 5 mg FHGC-1 were thoroughly mixed with 10 mL As(V) solution (initial concentration of 5 mg L⁻¹), which was allowed to be adsorb for a given period (1–150 min). The As(V) adsorption experiments for FHGC-2 and FHGC-3 were executed under the same conditions.

3. Results and discussion

Commonly, because the outer surface of the HNT^10,28 or GO sheet^27,28,34 is negatively charged, the direct composite process between HNT and GO sheets is hardly accomplished without additional assisting techniques.¹⁰ However, HNT and GO were easily dispersed in water to generate a stable suspension, which resulted from the existing electrostatic repulsion force. As shown in Scheme 1, the external iron(III) could be easily attached to either the outer surface of HNT or that of GO by the strong electrostatic attraction in such a suspension.^21,31 Similarly, the attachment process for iron(II) ions was also probably accomplished by the strong electrostatic attraction. In the following co-precipitation procedure, Fe₃O₄ NP would quickly appear, being triggered by heating and the added base (NH₄OH). In this process, the Fe₃O₄ NP could play the role of the adhesive to combine HNT and GO sheets (Scheme 1). As we expected, the total combination of the three components was accomplished in 1 hour by the co-precipitation procedure. After centrifugation and evaporation, the raw product Fe₃O₄/HNT/GO composite was treated by heating under N₂ atmosphere to prepare the target Fe₃O₄/HNT/rGO composite. Normally, the HNT and Fe₃O₄ NP are easily encapsulated by the sheet-like rGO generated during the last heating treatment. When the final composite was washed three times with the mixture of ethanol and H₂O (v/v = 1 : 1) under ultrasonication, no magnetic material was observed in the aqueous phase, indicating the excellent stability of the new composites. Herein, the new Fe₃O₄/HNT@rGO composites were denoted as FHGC-n (n = 1, 2, 3).

Scheme 1 The construction process of FHGC.

In this study, the microstructure of FHGC-n (n = 1, 2, or 3) was determined by SEM, TEM and HRTEM analyses (Fig. 1). SEM results show that the Fe₃O₄ NP and the tubular HNT are distinctly dispersed on the rGO sheet (Fig. 1a–c). Further detailed evidence for the formation of the ternary composite was traced by TEM (Fig. 1d–g and S1†) and HRTEM (Fig. 1h) analyses. Although the morphologies of these three FHGC samples are very similar, the amount of Fe₃O₄ NP is visibly different in the three samples (Fig. 1d–g). The HNT and Fe₃O₄ NP are loaded onto the surface or intercalated between the rGO sheets; independent Fe₃O₄ NP and HNT are absent in these images of FHGC-n (n = 1, 2, or 3) composites. The morphology of the HNT is consistent with that of raw HNT (Fig. S1i†), while the morphology of Fe₃O₄ NP produced by using the facile co-precipitation process is irregular spheres with diameters of about several to 20 nm. Additionally, it is obvious that the dispersion of Fe₃O₄ NP onto graphene sheet is not quite homogenous. This may have originated from the distribution of the oxygen groups on the GO sheets, which could act as the nucleation centers to capture the Fe³⁺. On the other hand, the tendency of Fe₃O₄ NP toward aggregation is driven by their magnetic nature, which may also result in the non-homogeneous distribution of Fe₃O₄ NP on the graphene sheets. Moreover, in Fig. 1g, the distinctly resolved lattice (around 0.295 nm corresponding to the (220) plane of Fe₃O₄,³⁵) and fine diffraction spots (inset, Fig. 1h),³⁶ suggest that the irregular spherical-Fe₃O₄ NP were highly ordered crystals.

Fig. 1 SEM images of (a) FHGC-1; (b) FHGC-2; (c) FHGC-3. TEM images of (d) FHGC-1; (e) FHGC-2; (f) FHGC-3; (g) Fe₃O₄ on HNT, the inset is the corresponding SAED pattern. (h) HRTEM image of Fe₃O₄.

In this study, the FT-IR spectra of samples FHGC-n (n = 1, 2, 3) are very similar (Fig. S2a† and 2a); thus, only FHGC-2 is discussed in detail. As shown in Fig. 2a, the presence of rGO, Fe₃O₄ NP and HNT is undoubtedly demonstrated, as well as their combination in the FHGC. In comparison with the spectra of GO and rGO, the typical stretching vibrations of carboxyl groups (1715 cm⁻¹) or epoxide groups (1090 cm⁻¹) were absent in the FHGC-2 spectrum, suggesting the total conversion from GO to rGO.^9,10 The peaks at 1571 cm⁻¹ (C=C) and 1086 cm⁻¹ (C–O) in the spectrum of FHGC-2 were assigned to the stretching vibrations of rGO,³⁷ while the typical peaks of HNT were present at 1640, 1040, 750 and 550 cm⁻¹.^10,20 The peaks at 567 and 416 cm⁻¹ correspond to the Fe²⁺–O²⁻ and Fe³⁺–O²⁻ in the FHGC-2 spectrum, which also confirmed the presence of Fe₃O₄.^9,21,31

Fig. 2 (a) FT-IR spectra of GO, rGO, Fe₃O₄, and FHGC-2; (b) XRD patterns of GO, rGO, HNTs, and FHGC-2; (c) Raman spectra of GO, rGO and FHGC-2; (d) XPS spectra of FHGC-2; (e) and (f) XPS spectra of FHGC showing Fe and C 1s peaks, respectively.

The phase structure of FHGC-2 is described in detail and shown in Fig. 2b (the XRD patterns of FHGC-1 and FHGC-3 are shown in Fig. S2b†). As expected, the disappearance of the GO peak at 9.7° and the presentation of two broad rGO peaks centered at 24.3 and 43.6° indicate the complete reduction of GO to rGO,³⁸ while the peaks at 24.8° and 35.0° originate from HNT.^10,20 The positions and relative intensities of the diffraction peaks match well with the standard Fe₃O₄ (JPPDS no. 19-0629, 30.1° (220), 35.4° (311), 43.0° (400), 57.2° (511), and 62.6° (440)).³⁹ As shown in Fig. 2c and S3,† the D (1340 cm⁻¹) band of the disordered carbons and the G (1580 cm⁻¹) band of the sp² hybridized carbon are present in the Raman spectra of FHGC-n (n = 1, 2, or 3), and are very close to those in the spectra of GO and rGO.⁴⁰ In addition, the increased D band intensity of the FHGC-n (n = 1, 2, 3) results from the apparent structural interactions of the ternary system (Fe₃O₄ NP, HNT and rGO sheet).^41,42

Importantly, XPS analysis is a quite useful tool for identifying the details of elemental composition in material science. Herein, the XPS spectra of FHGC-n (n = 1, 2, or 3; Fig. 2d and S4†) distinctly show the carbon (C 1s), oxygen (O 1s), iron (Fe 2p), silicon (Si 2p), and aluminum (Al 2p). The observation of Si 2p (75.42 eV) and Al 2p (103.42 eV) is evidence for proving the existence of HNT in the ternary system FHGC.^10,43 As shown in Fig. 2e, S4b and e,† Fe 2p_3/2 (711.27 eV) and Fe 2p_1/2 (724.92 eV) are also visible in the high resolution Fe 2p scan. It is clear that the fitting peaks broaden with higher binding energy in Fe₃O₄ as a result of the existence of Fe²⁺ (2p_3/2) and Fe²⁺ (2p_1/2).^9,21 Moreover, the O 1s peak is observed at 531.42 eV, belonging to the lattice oxygen of Fe₃O₄, while the weak shoulder peak at 710 eV also confirms the presence of Fe₃O₄.^9,21 In addition, Fig. 2f, S4c and f† show that the typical C 1s peak occurs at 284.6 eV, while the peak located at 286.02 eV is commonly attributed to surface-adsorbed hydrocarbons and their oxidative forms (–COOH, epoxide). In the C 1s spectrum of FHGC-n (n = 1, 2, or 3), the peak centered at 288.72 eV originates from the carbon element in association with oxygen in the carbonate ions.^9,21,44 The remaining smaller peaks at higher binding energies (785.6, 788.2, 802.1, and 805.2 eV) are satellite shake-ups of the assigned components.⁴⁵

As we know, Fe₃O₄ NP are paramagnetic. In the control experiments, it is clear that each of the samples FHGC-n (n = 1, 2, and 3) designed is easily removed by one small magnet and can be re-dispersed into aqueous solution without aggregation. Hence, M–H measurements were performed to further prove the paramagnetic behavior of FHGC-n (n = 1, 2, or 3) at room temperature. In each magnetization curve of FHGC-n (n = 1, 2, or 3; Fig. S5a†), the magnetic saturation value is lower than that of pure Fe₃O₄ NP (FHGC-1: 17.07 emu g⁻¹, FHGC-2: 14.07 emu g⁻¹, and FHGC-3: 6.88 emu g⁻¹).⁴⁶ These results may be caused by the relatively low content of Fe₃O₄ NP in FHGC-n (n = 1, 2, or 3), where the existence of non-magnetic rGO and HNT decreases the density of the magnetic component. The H_c (Fig. S5b†) indicates that the FHGC-n (n = 1, 2, or 3) is a kind of soft magnetic material. FHGC-n (n = 1, 2 and 3) samples were selected to investigate the microstructure of the ternary system by the N₂ adsorption–desorption test. As shown in Fig. S6a,† the FHGC-n samples are mesoporous, showing the specific surface areas of 107.3 (FHGC-1), 152.4 (FHGC-2), and 103.8 m² g⁻¹ (FHGC-3), with the most probable pore widths of 5.7, 4.2, and 6.0 nm, respectively (Fig. S6b†); the total pore volumes of FHGC-n are 0.240, 0.282, and 0.289 cm³ g⁻¹, respectively, for pore widths ranging from 1.7 to 300 nm.

According to the reported graphene-based composite,^9,10 both HNT and intercalated Fe₃O₄ NP prevented the rGO from restacking during the heat reduction process to effectively maintain the large surface areas and pore volumes. In addition, both rGO and HNT are good candidates for adsorbing organic material, and the separation process is definitely facile, as a result of the presence of magnetic Fe₃O₄ NP in the FHGC-n (n = 1, 2, or 3) ternary system. In this paper, the typical organic dye RhB was selected for the study of the absorption properties in water.

Absorption was monitored using the UV-Vis absorption spectrometer. The typically strong absorption band of RhB was immediately decreased when FHGC-n (n = 1, 2, or 3) was added into the test solution (Fig. 3a and S7a†). For the samples FHGC-n (n = 1, 2, and 3), more than 70% RhB in water disappeared after 3 min, and then the adsorption process gradually slowed down (Fig. S7c and d†). These results show that the concentration of remaining RhB is substantially unchanged after 15 min (C₀ = 1.0 × 10⁻⁵ M is the initial concentration of RhB; C represents the concentration of RhB at a given moment), indicating the adsorption rate is relatively fast, compared to that in recent literature.^9,10,47 Because of the magnetic Fe₃O₄ NP in this ternary system, FHGC is easily separated by using a small magnet (Fig. S7b†), which is very important to avoid secondary pollution in the waste water treatment. Interestingly, FHGC-2 exhibits the highest adsorption rate and lowest residual concentration of RhB.

Fig. 3 (a) UV-Vis spectra of the original RhB solution (1.0 × 10⁻⁵ M, 20 mL) and those after treatment with FHGC-2 (50 mg) at different times; (b) the ln(C/C₀) versus time plots with FHGC samples 1–3; (c) C/C₀ versus time plots with Fe₃O₄, HNTs, GO, and rGO at various times; (d) FHGC-2 in various cycles: (i) 1st, (ii) 2nd, (iii) 3rd, (iv) 4th, (v) 5th, (vi) 6th, (vii) 7th, (viii) 8th, (ix) 9th cycle.

Moreover, the maximum adsorption amount of FHGC-2 (19.39 mg g⁻¹) is higher than that of FHGC-1 (15.22 mg g⁻¹) and FHGC-3 (16.56 mg g⁻¹). As shown in Fig. 3b, each adsorption curve of the three FHGC cases is well fitted by the first-order reaction rate kinetic model, which is similar to the binary system FGNC or HGC that we reported previously.^9,10 Control experiments with Fe₃O₄ NP, HNT, GO, and rGO as adsorbents were carried out to determine the advantages of the FGNC ternary system (Fig. 3c). In comparison to the performances of GO and rGO, HNT showed bad adsorption behavior for RhB, while RhB in water was scarcely adsorbed by the pure Fe₃O₄ NP. It is evident that the significant adsorption ability exhibited by the FHGC-n ternary system mainly resulted from the physical adsorption onto the rGO surface.

Normally, the ratio of rGO in the sample is the key factor for the maximum adsorption amount (A_max). In this work, although the weight content of rGO calculated is gradually increased (FHGC-1: 20%, FHGC-2: 25%, and FHGC-3: 32%), the presentation of FHGC-2 is the most dazzling of the FHGC samples. We found that FHGC-2 was well dispersed in water under ultrasonic conditions in 30 min, while the other two samples were still slurries and precipitated easily under the same conditions; this may be the reason for the special behavior in the case of FHGC-2. To further investigate the capability of FHGC-n ternary systems, the absorption experiment was repeated for FHGC-2 under the same conditions. After the first 20 min (Fig. 3d), FHGC-2 was separated from the system by the magnet and directly reimmersed into a new RhB aqueous solution (20 mL, 1.0 × 10⁻⁵ M); the second line is almost the same as the first one, illustrating a similar absorption, still efficiently executed. It is noteworthy that the adsorption rate was retained with increased separation–absorption cycles, even in the 9th cycle.

Notoriously, arsenic is one of the top 20 hazardous substances in the world, especially the inorganic arsenic species, which are more dangerous than the organic forms. Arsenate (As(V)) is easily enriched in the surface water to aggravate the pollution of drinking water⁴⁸ and consequently, it has attracted worldwide attention to develop facile and cheap techniques for the remediation of arsenic contamination. In a recent study by our group, the Fe₃O₄@rGO dyad composite (FGNC) prepared by the colloid electrostatic self-assembly method indicated favorable behavior in the removal of the simulated inorganic contaminant As(V) (Na₃AsO₄, Freundlich parameter n = 3.40).⁹ Herein, Na₃AsO₄ aqueous solution was again selected as the simulated As(V) water pollutant for the adsorption experiment of FHGC-n (n = 1, 2, or 3) samples. The adsorption results under neutral conditions are displayed (Fig. 4, S8 and Table S2,† pH ≈ 7). As illustrated in Fig. 4a, the equilibrium adsorption capacity (q_e, mg g⁻¹) increased visibly with increasing original concentration of As(V). For the FHGC-n (n = 1, 2 and 3) samples, both the Langmuir and Freundlich isotherm models were applied to analyze the adsorption equilibrium data of simulated inorganic contaminant As(V). Interestingly, sample FHGC-1, with the highest content of Fe₃O₄ NP (Fig. 4b and S8†), has a similar fine value of R² in the Langmuir and Freundlich models, and the adsorption process on a homogeneous surface is possibly either a monolayer adsorption or a multilayer adsorption. In contrast, in sample FHGC-2 and FHGC-3, the monolayer adsorption process is only reasonable (Fig. 4b).

Fig. 4 (a) As(V) adsorption isotherms of the FHGC-n (n = 1, 2, 3) samples; (b) linearized Langmuir isotherm for As(V) adsorption of the FHGC-n (n = 1, 2, 3) samples; (c) As(V) kinetic absorption data plot of the FHGC-n (n = 1, 2, 3) samples (removal rate q_t vs. time t) and (d) pseudo-second-order kinetic plot for the adsorption of As(V) onto the FHGC-n (n = 1, 2, 3) samples.

As we know, under neutral environmental conditions the major form of As(V) is the anionic species (including H₂AsO₄⁻ and HAsO₄²⁻); therefore, there is a relatively strong electrostatic repulsive interaction between the adsorbed As(V) species and unabsorbed As(V) species in solution. When the content of Fe₃O₄ NP is comparatively low in the sample composite, such electrostatic repulsion is the key effect promoting a dominant monolayer adsorption. In contrast, when the electrostatic repulsive effect is weakened by the increased content of Fe₃O₄ NP in the sample composite, there is competition between the monolayer adsorption process and the multilayer adsorption process. Furthermore, in the case of FHGC-1, the maximum adsorption capacity (q_m, Table S2†) of 4.42 mg g⁻¹ was calculated using the Langmuir model, while the Freundlich parameter (n, Table S2†) value of 7.06 indicated that the designed FHGC-1 is a potentially fair arsenate adsorbent. Normally, when the content of Fe₃O₄ is decreased in the sample composite, FHGC-2 or FHGC-3 shows the lowest q_m value (Table S2†). This result is still remarkably higher than that of commercial Fe₃O₄ (1.35 mg g⁻¹),⁴⁹ which therefore implies that because of the strong affinities of the FHGC series developed in our work for arsenate, especially sample FHGC-1, they might serve as efficient adsorbents for the removal of arsenic ions in water treatment.^49,50 In order to detect the stability of FHGC-n (n = 1, 2, or 3) in the adsorption process, the concentrations of Fe, Al, and Si in simulated As(V) polluted water were also investigated by ICP MS. As shown in Fig. S9,† it is clear that the concentrations of such ions were almost unchanged during the adsorption process, indicating that no secondary pollution is produced by FHGC-n (n = 1, 2, or 3) adsorbent.

Additionally, the adsorption data for arsenate ions by the FHGC series is illustrated in the pattern of Fig. 4c at different time intervals, which indicates the importance of adsorption kinetics. As shown in Fig. 4c, in each of the FHGC series, the adsorption process of As(V) is rapid at first, which is ascribed to the process of arsenic adsorption on the exterior surface of the Fe₃O₄ particles without any repulsion. When the diffusion resistance and the electrostatic repulsive interactions gradually played their roles, the adsorption process slowed down and at last reached equilibrium. In each case of the FHGC series, equilibrium was obtained in less than 30 min, which is almost the best example in literature.^9,48,50–53 As shown in Fig. 4d, the above adsorption kinetic experimental data were best fitted by the pseudo-second-order rate kinetic model in each of the three FHGC cases, which is presented as follows:

t/q_t = 1/(k₂q_e²) + 1/q_t

V₀ = k₂q_e²

where q_t and q_e (mg g⁻¹) are the adsorption amounts of As(V) at a given time t (min) and at equilibrium, k₂ (g mg⁻¹ min⁻¹) is the rate constant of the pseudo-second-order model and V₀ (mg g⁻¹ min⁻¹) is the initial pseudo-second-order rate.

4. Conclusions

Three components, HNTs, Fe₃O₄ and GO, were successfully combined together to form a new FHGC-n (n = 1, 2, or 3) material via a facile co-precipitation method, followed by a heat treatment process. The resulting FHGC exhibits superior performance for the adsorption of both RhB and As(V). The maximum adsorption capacity of 19.39 mg g⁻¹ for RhB and 4.42 mg g⁻¹ for As(V) was achieved. These results show the promising applications of FHGC toward environmental issues. The facile co-precipitation method is a versatile platform for synthesizing composites of several different components.

Acknowledgements

Financial was supported from the National Science Foundation of China (No. 21373189) and the China Postdoctoral Science Foundation (No. 2013M531681).

Notes and references

1. M. W. Holdgate, Environ. Conserv., 1987, 14, 282.
2. X. Huang, X. Y. Qi, F. Boey and H. Zhang, Chem. Soc. Rev., 2012, 41(2), 666–686.
3. S. Xin, Y. G. Guo and L. J. Wan, Acc. Chem. Res., 2012, 45(10), 1759–1769.
4. F. Perreault, A. F. de Faria and M. Elimelech, Chem. Soc. Rev., 2015, 44(16), 5861–5896.
5. M. M. Khin, A. S. Nair, V. J. Babu, R. Murugan and S. Ramakrishna, Energy Environ. Sci., 2012, 5(8), 8075–8109.
6. S. C. Smith and D. F. Rodrigues, Carbon, 2015, 91, 122–143.
7. R. K. Upadhyay, N. Soin and S. S. Roy, RSC Adv., 2014, 4(8), 3823–3851.
8. K. C. Kemp, H. Seema, M. Saleh, N. H. Le, K. Mahesh, V. Chandra and K. S. Kim, Nanoscale, 2013, 5(8), 3149–3171.
9. J. Ding, B. J. Li, Y. S. Liu, X. S. Yan, S. Zeng, X. D. Zhang, L. F. Hou, Q. Cai and J. M. Zhang, J. Mater. Chem. A, 2015, 3, 832–839.
10. Y. S. Liu, X. Q. Jiang, B. J. Li, X. D. Zhang, T. Z. Liu, X. S. Yan, J. Ding, Q. Cai and J. M. Zhang, J. Mater. Chem. A, 2014, 2, 4264–4269.
11. A. Y. Romanchuk, A. S. Slesarev, S. N. Kalmykov, D. V. Kosynkin and J. M. Tour, Phys. Chem. Chem. Phys., 2013, 15(7), 2321–2327.
12. F. Fang, L. T. Kong, J. R. Huang, S. B. Wu, K. S. Zhang, X. L. Wang, B. Sun, Z. Jin, J. Wang and X. J. Huang, J. Hazard. Mater., 2014, 270, 1–10.
13. C. Wang, H. J. Luo, Z. L. Zhang, Y. Wu, J. Zhang and S. W. Chen, J. Hazard. Mater., 2014, 268, 124–131.
14. L. M. Cui, Y. G. Wang, L. Gao, L. H. Hu, L. G. Yan, Q. Wei and B. Du, Chem. Eng. J., 2015, 281, 1–10.
15. H. R. Thomas, S. P. Day, W. E. Woodruff, C. Valles, R. J. Young, I. A. Kinloch, G. W. Morley, J. V. Hanna, N. R. Wilson and J. P. Rourke, Chem. Mater., 2013, 25(18), 3580–3588.
16. B. J. Li and H. Q. Hua, J. Mater. Chem., 2011, 21(10), 3346–3349.
17. G. Gnana kumar, K. J. Babu, K. S. Nahm and Y. J. Hwang, RSC Adv., 2014, 4(16), 7944–7951.
18. J. Salamon, Y. Sathishkumar, Y. S. Lee, D. J. Yoo, A. R. Kim and G. Gnana kumar, Biosens. Bioelectron., 2015, 64, 269–276.
19. W. O. Yah, A. Takahara and Y. M. Lvov, J. Am. Chem. Soc., 2012, 134(3), 1853–1859.
20. W. O. Yah, H. Xu, H. Soejima, W. Ma, Y. M. Lvov and Y. Lvov, J. Am. Chem. Soc., 2012, 134(29), 12134–12137.
21. B. J. Li, H. Q. Cao, J. Shao, M. Z. Qu and J. H. Warner, J. Mater. Chem., 2011, 21(13), 5069–5075.
22. C. Nethravathi, B. Viswanath, C. Shivakumara, N. Mahadevaiah and M. Rajamathi, Carbon, 2008, 46(13), 1773–1781.
23. C. Zhang, W. W. Tjiu, W. Fan, Z. Yang, S. Huang and T. X. Liu, J. Mater. Chem., 2011, 21(44), 18011–18017.
24. F. R. Zhang, Y. W. Song, S. Song, R. J. Zhang and W. G. Hou, ACS Appl. Mater. Interfaces, 2015, 7(13), 7251–7263.
25. S. L. Li, H. Li, J. Liu, H. L. Zhang, Y. M. Yang, Z. Y. Yang, L. Y. Wang and B. D. Wang, Dalton Trans., 2015, 44(19), 9193–9199.
26. L. Yu, Y. T. Zhang, B. Zhang and J. D. Liu, Sci. Rep., 2014, 4, 4551.
27. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
28. S. B. Yang, X. L. Feng, S. Ivanovici and K. Mullen, Angew. Chem., Int. Ed., 2010, 49, 8408–8411.
29. J. Liang, Z. Y. Fan, S. Chen, S. J. Ding and G. Yang, Chem. Mater., 2014, 26(15), 4354–4360.
30. D. A. Robinson, J. J. Yoo, A. D. Castañeda, B. Gu, R. Dasari, R. M. Crooks and K. J. Stevenson, ACS Nano, 2015, 9(7), 7583–7595.
31. Y. S. Liu, P. Liu, Z. X. Su, F. S. Li and F. S. Wen, Appl. Surf. Sci., 2008, 255, 2020–2025.
32. G. N. Li, L. Liang, B. H. Wu, J. X. Li, Y. Yuan and J. L. Shi, Nanoscale, 2015, 7(42), 17855–17860.
33. Z. T. Hu, J. C. Liu, X. L. Yan, W. D. Oh and T. T. Lim, Chem. Eng. J., 2015, 262, 1022–1032.
34. W. Wei, S. Yang, H. Zhou, I. Lieberwirth, X. Feng and K. Müllen, Adv. Mater., 2013, 25(21), 2909–2914.
35. W. L. Gu, X. Deng, X. F. Jia, J. Li and W. K. Wang, J. Mater. Chem. A, 2015, 3, 8793–8799.
36. C. L. Liang, S. C. Huang, W. X. Zhao, W. Y. Liu, J. Chen, H. Liu and Y. X. Tong, New J. Chem., 2015, 39, 2651–2656.
37. K. Zhang, L. Mao, L. L. Zhang, H. S. O. Chan, X. S. Zhao and J. S. Wu, J. Mater. Chem. A, 2011, 21, 7302–7307.
38. Y. F. Fan, Y. S. Liu, Q. Cai, Y. Z. Liu and J. M. Zhang, Synth. Met., 2012, 162(21–22), 1815–1821.
39. Q. Guo, P. R. Ye, J. Wang, F. F. Fu and Z. J. Wu, J. Hazard. Mater., 2015, 298, 28–35.
40. G. X. Wang, J. Yang, J. S. Park, X. L. Gou, B. Wang, H. Liu and J. Yao, J. Phys. Chem. C, 2008, 112, 8192–8195.
41. L. H. Ai, C. Y. Zhang and Z. L. Chen, J. Hazard. Mater., 2011, 192, 1515–1524.
42. Z. H. Ni, H. M. Wang, J. Kasim, H. M. Fan, T. Yu, Y. H. Wu, Y. P. Feng and Z. X. Shen, Nano Lett., 2007, 7, 2758–2763.
43. Z. Long, W. Tingmei and L. Peng, Appl. Surf. Sci., 2008, 255, 2091–2097.
44. L. Cui, Y. Wang, L. Gao, L. Hu, L. Yan, Q. Wei and B. Du, Chem. Eng. J., 2015, 281, 1–10.
45. D. K. Padhi and K. Parida, J. Mater. Chem. A, 2014, 2, 10300–10312.
46. V. S. Zaitsev, D. S. Filimonov, I. A. Presnyakov, R. J. Gambino and B. Chu, J. Colloid Interface Sci., 1999, 212, 49–57.
47. B. J. Li, H. Q. Cao and G. Yin, J. Mater. Chem. A, 2011, 21, 13765–13768.
48. K. Zhang, V. Dwivedi, C. Y. Chi and J. S. Wu, J. Hazard. Mater., 2010, 182, 162–168.
49. T. Wang, L. Y. Zhang, H. Y. Wang, W. C. Yang, Y. C. Fu, W. L. Zhou, W. T. Yu, K. S. Xiang, Z. Su, S. Dai and L. Y. Chai, ACS Appl. Mater. Interfaces, 2013, 5(23), 12449–12459.
50. V. Chandra, J. Park, Y. Chun, J. W. Lee, I. C. Hwang and K. S. Kim, ACS Nano, 2010, 4(7), 3979–3986.
51. L. Q. Guo, P. R. Ye, J. Wang, F. F. Fu and Z. J. Wu, J. Hazard. Mater., 2015, 298, 28–35.
52. X. B. Luo, C. C. Wang, S. L. Luo, R. Z. Dong, X. M. Tu and G. S. Zeng, Chem. Eng. J., 2012, 187, 45–52.
53. G. D. Sheng, Y. M. Li, X. Yang, X. M. Ren, S. T. Yang, J. Hu and X. K. Wang, RSC Adv., 2012, 2(32), 12400–12407.

---

Footnote

† Electronic supplementary information (ESI) available: Synthesis of graphene oxide, TEM images, FTIR and XRD results, VSM isotherms, nitrogen adsorption desorption isotherms, and tables. See DOI: 10.1039/c6ra01279e

---