Dyes Adsorption Behavior of Fe₃O₄ Nanoparticles Functionalized Polyoxometalate Hybrid

Abstract

The magnetic adsorbent, Fe₃O₄@[Ni(HL)₂]₂H₂[P₂Mo₅O₂₃]·2H₂O (Fe₃O₄@1), is synthesized by employing the nanoparticles Fe₃O₄ and polyoxometalate hybrid 1. Zero-field-cooled (ZFC) and field-cooled (FC) curves show that the blocking temperature of Fe₃O₄@1 was at 120 K. Studies of Fe₃O₄@1 removing cationic and anionic dyes from water have been explored. The characterization of Fe₃O₄@1, effects of critical factors such as dosage, the concentration of methylene blue (MB), pH, adsorption kinetics, isotherm, the removal selectivity of substrate and the reusability of Fe₃O₄@1 were assessed. The magnetic adsorbent displayed an outstanding removal activity for the cationic dye at a broad range of pH. The adsorption kinetics and isotherm models revealed that the adsorption process of Fe₃O₄@1 was mainly governed via chemisorption. The maximum capacity of Fe₃O₄@1 adsorbing substance was 41.91 mg g⁻¹. Furthermore, Fe₃O₄@1 showed its high stability by remaining for seven runs of the adsorption-desorption process with an effective MB removal rate, and could also be developed as a valuable adsorbent for dyes elimination from aqueous system.

1. Introduction

With the development of world’s population and socioeconomics, fresh and clean water has become one of the major concerns of 21^st century. The pollutants in aquatic ecosystems have caused serious environmental problems and threatened public health. Developing sustainable and economical technologies for treating wastewater process faces true challenges [1,2,3]. The pollution caused by organic dyes has been a great menace to the water ecosystem and human health [4,5]. Many researches have been implemented for the organic dyes removal, such as electrochemical oxidation, coagulation, adsorption, membrane filtration, biological schemes and photocatalytic degradation [6,7,8,9]. Among these strategies, adsorption is recognized as a promising approach due to its operating environment, simple procedures and high efficiency [10]. Many adsorbents have been used for the removal of dyes from aqueous solutions such as activated carbon, chitosan particles, aerogels, etc. [11,12,13]. However, it should be noted that the traditional adsorption materials have low removal efficiency and limited recycling ability. Therefore, developing effective adsorbents to remove the pollutants from wastewater is of great concern to researchers.

Phosphomolybdates are key to this issue due to their special structure and potential application in water pollution treatment [14,15,16]. One of the most prominent structures of phosphomolybdate is [P₂Mo₅O₂₃]⁶⁻. Transition metal ions also have possible application in the dyes removal by coordinating with dyes containing –N=N–, –C=C– and heterocyclic compounds [17,18]. In particular, nickel is a vital element to contribute to the removal of dyes [19,20]. Additionally, thiosemicarbazones are also beneficial for the dyes removal by forming hydrogen bonds and π–π stacking interactions [21]. Thus, the combination of [P₂Mo₅O₂₃]⁶⁻, nickel ion and 2-acetylpyrazine-thiosemicarbazone at a molecular will contribute to the dye removal by synergic effect. However, it is difficult to separate the hybrid phase from the liquid phase. Moreover, the non-recyclability may limit the application of the hybrid on a large scale. Recently, magnetic separation has gained increasing attention due to the fast and noncontact magnetic response by external magnetic field [22,23,24]. Fe₃O₄ nanoparticles are ferrite magnetic materials, which have strong magnetic properties. They have been taken as a carrier to synthesize nanocomposites, which can be easily removed and regained under a magnetic field [25].

Taking the above considerations, a Fe₃O₄-based magnetic adsorbent Fe₃O₄@1 was synthesized and characterized. The adsorption activity was investigated by adsorbing MB, rhodamine B (RhB) and methyl orange (MO) dissolving in water. Fe₃O₄@1 has demonstrated an effective selectivity for adsorbing cationic dyes. Moreover, the effects of some key factors such as the adsorbent dosage, the concentration of MB and pH on adsorption performance were studied. The adsorption behavior of Fe₃O₄@1 was explored by adsorption kinetics and isotherm models, along with its recyclability and stability for multiple cycles of adsorption-desorption process.

2. Results and Discussion

2.1. Structural Descriptions

1 comprises a strandberg-type [P₂Mo₅O₂₃]⁶⁻ anion, two [Ni(HL)₂]²⁺, two protons and two crystal water molecules (Figure 1a). [P₂Mo₅O₂₃]⁶⁻ and [Ni(HL)₂]²⁺ are bound together by hydrogen bonds, as shown by dotted orange lines in Figure 1a and electrostatic interaction. The anion [P₂Mo₅O₂₃]⁶⁻ possesses two central PO₄ tetrahedrons which combine the edge-sharing MoO₆ octahedrons into a single cluster. In cationic [Ni(HL)₂]²⁺, two Ni²⁺ show the similar coordination geometry by coordinating four N atoms and two S atoms from the two ligands HL, which shows the same coordination patterns of distorted octahedral geometries. The amino and pyrazine groups in ligand HL will conduce to form the hydrogen bonds and π–π stacking interactions with dyes. Figure 1b shows the three-dimensional network of 1. The anionic [P₂Mo₅O₂₃]⁶⁻ and amino-included ligands coexisting in a molecule are conducive for the interaction with cationic dyes by electrostatic interaction and hydrogen bonds.

2.2. FTIR Spectra

In the spectrum of Fe₃O₄@1, the characteristic peaks of 1 and Fe₃O₄ are included (Figure 2). Both Fe₃O₄@1 and Fe₃O₄ exhibit the same peak at 588 cm⁻¹, matching with the Fe–O band. The broad bands around 3492 cm⁻¹ are attached to the O–H stretching vibration from water molecules [26]. Two characteristic peaks at 3280 and 3154 cm⁻¹ are assigned to N–H stretching vibrations. The peak at 1614 cm⁻¹ is associated with ν(C=N) bonds. The peak at 1219 cm⁻¹ corresponds the ν(P–O) bands. The peaks between 950 and 853 cm⁻¹ are appointed to the terminal Mo=O vibration. The range of 859–664 cm⁻¹ is attributed to Mo–O–Mo vibration [27]. Therefore, 1 and Fe₃O₄ are included in Fe₃O₄@1.

2.3. UV-Vis Spectra

The spectra of 1 and Fe₃O₄@1 show four absorption peaks in water (Figure 3). The peaks at 212 and 233 nm correspond to the charge-transfer of O_t→Mo, and the peaks at 312 and 410 nm are assigned to the charge-transfer of O_b→Mo [28]. There are no absorption peaks obtained for Fe₃O₄ nanoparticles. These results show that 1 exists in Fe₃O₄@1.

2.4. XPS Characterization

The XPS analysis was recorded to further identify the successful preparation of Fe₃O₄@1. Results indicate that the presence of C 1s, N 1s, S 2p, P 2p, Mo 3d, Fe 2p and Ni 2p are at the sample surface. The corresponding detailed peaks of Fe and Mo are shown in Figure 4. The binding energy (BE) was evaluated by C 1s peak (284.8 eV). The BE of Fe 2p_1/2 and Fe 2p_3/2 are around 725.0 and 711.1 eV, respectively. The BE at 708.8 and 710.3 eV conforms to Fe²⁺ 2p_3/2 and Fe³⁺ 2p_3/2, respectively (Figure 4a) [29,30]. The XPS spectrum of Fe₃O₄@1 for Mo atoms gave two peaks with the BE of 235.2 and 232.2 eV respectively referring to Mo 3d_3/2 and Mo 3d_5/2 (Figure 4b) [31]. It indicated that Fe₃O₄ and 1 did coexist in Fe₃O₄@1.

2.5. Magnetic Properties of Fe₃O₄@1

Figure 5a,b demonstrates the hysteresis curves of Fe₃O₄@1. Under the 10,000 Oe magnetic field, Fe₃O₄@1 exhibits ~35 Oe coercivity and ~43 emu g⁻¹ magnetization at 300 K, while the coercivity of ~280 Oe and the magnetization of ~47 emu g⁻¹ are at 5 K. The hysteresis curves of Fe₃O₄@1 indicates that Fe₃O₄@1 change from soft ferromagnetic at 300 K to ferromagnetic at 5 K. Meanwhile, the magnetization increases when the temperature decreases. Under 500 Oe, the curves of ZFC and FC between 5 and 300 K were also studied (Figure 5c). It turned out that the blocking temperature of Fe₃O₄@1 was circa 120 K [32]. Under the temperature, the FC curve is almost flat, whereas the ZFC curve falls sharply. Figure 5d represents the segregation and diffusion process of Fe₃O₄@1 in water. The solution changed from brown uniform diffusion to achromous transparence under the influence of a magnetic field. After removing the magnetic field, the collected Fe₃O₄@1 could be easily dispersed with agitation. The segregation and diffusion process could be repeated for many times, which further indicated that the Fe₃O₄@1 were magnetic. In addition, the process also indicated that hybrids 1 and Fe₃O₄ were successfully combined together as 1 was nonmagnetic. It should be noted that the magnetic responsiveness and redispersibility of Fe₃O₄@1 play important roles in the repeatability of adsorption application.

2.6. Morphology and Particle Size Analyses

Figure 6a represents the TEM image of pure Fe₃O₄@1. It is clear that pure Fe₃O₄@1 has uniform particle morphology of roundness shape with an average size of 20.8 nm. The distribution was properly illustrated by the Gaussian function (Figure 6b). Figure 6c represents the HRTEM images of Fe₃O₄@1 alone, with the 2.53 Å spacing matching the (311) reflection of Fe₃O₄ nanoparticle. Even though 1 covered on the Fe₃O₄ exterior was hardly classified, the coarseness and expanded thickness of the periphery indicated that 1 has been profitably incorporated on Fe₃O₄. Furthermore, elemental mappings (Figure 6d–l) illustrate the distribution of the elements C, O, Ni, Mo, N, Fe, P and S in the Fe₃O₄@1 sample, indicating that 1 and Fe₃O₄ coexist in Fe₃O₄@1. These further confirm the successful formation of Fe₃O₄@1 nanocomposites.

2.7. The XRD Patterns and BET Analysis

The pattern of Fe₃O₄@1 indicates that the peaks of 1 and Fe₃O₄ are included (Figure 7a). The diffraction peaks of Fe₃O₄ correspond to the stand card of JCPDS No. 75-0449. The average particle size of Fe₃O₄@1 is 19.7 nm. The value is calculated by the Debye-Scherrer equation. The numerical value is approximate to the TEM analysis. The result of PXRD analysis demonstrates that 1 and Fe₃O₄ are well combined.

Figure 7b exhibits the N₂ adsorption-desorption and the aperture distribution curve of Fe₃O₄@1. According to the classification of IUPAC, Fe₃O₄@1 shows a symbolic type IV isotherm curve with H1 hysteresis loop [33]. Besides, the adsorption curve is closer but leveled below the desorption curve. Therefore, the Fe₃O₄@1 is typical mesoporous structure [34]. The generated pores are conducive to the removal of MB [35]. For the Fe₃O₄@1, the BET surface area was 15.05 m² g⁻¹, the pore volume was 0.07 cm³ g⁻¹ and the average diameter of BJH pore was 3.83 nm, respectively.

2.8. Dye Adsorption Experiments

In recent decades, nanocomposites have attracted increasing attention in adsorbing organic dyes from wastewater. To determine whether Fe₃O₄@1 is a suitable adsorbent for organic dyes, MB was used as an exemplary role because of its extensive usage in academic studies and numerous industries. Fe₃O₄@1 was put into 10 mL water of MB (15 mg L⁻¹) at pH 6.98 for 3 h. Fe₃O₄@1 displayed a removal rate of 97.84% and adsorption capacity of circa 41.91 mg g⁻¹ on MB. After adsorption, the MB solution almost turned colorless. Some major aspects which might influence the adsorption activity, such as the adsorbent dosage, MB concentration, pH, adsorption kinetics, isotherm models and substrate selectivity were studied.

The optimum usage amount of Fe₃O₄@1 was studied by injecting multifarious amounts (2, 4, 6, 8, 10, 12 and 14 mg) to aqueous MB of 15 mg L⁻¹ (experimental condition: pH, 7; temperature, 25 °C; volume, 10 mL; time, 3 h). As shown in Figure 8a, the removal rate reached 97.84% when 12 mg of the adsorbent was used. Thus, 12 mg/10 mL was concluded to be the optimum usage/volume proportion for the adsorption process, which was applied for the following experiments. The influence of MB concentration was explored as well (Figure 8b). While increasing MB concentration from 5 to 30 mg L⁻¹ (5, 10, 15, 20, 25 and 30 mg L⁻¹), the removal rate changed to 95.02%, 97.53%, 97.84%, 97.76%, 95.81% and 94.70%, accordingly. Under low concentrations, saturation was achieved after 30 minutes. Indeed, MB was almost completely removed within 3 h with different concentrations. Therefore, 15 mg L⁻¹ is considered to be the optimum concentration for the adsorption process.

The influence of solution pH was investigated as it was a crucial aspect that dictated the adsorption property due to its relationship with the extent of ionization with substrate and apparent charge of adsorbent. As demonstrated in Figure 9, Fe₃O₄@1 represents varying degrees of affinity towards MB at the range of 3–13. The MB adsorption efficiency retained a relatively high situation, while the initial values were between 5 and 11. The activity of Fe₃O₄@1 at the initial pH value of 7 was the best. At a high pH value of 13, the removal rate was only 36.57%, which might be the reason that the structure of 1 of Fe₃O₄@1 was damaged by the excess OH⁻. Therefore, further studies were explored under neutral pH condition (pH circa 7.0).

The adsorption kinetics was analyzed using the pseudo-first-order and pseudo-second-order model. Figure 10a,b shows the correlation coefficient (R²) of the above two models, respectively. It can be observed that the pseudo-second-order model showed a larger R² than that of the pseudo-first order, demonstrating that the adsorption process was mainly governed via chemisorption [36].

The adsorption isotherms of MB at 298 K were investigated with the adsorbents of 12 mg for 3 h. Langmuir and Freundlich isotherm models were utilized to further study the adsorption capacities of adsorbents and the interaction between adsorbate and adsorbent. The two isotherms models are described in Equations S5 and S6, respectively [37]. As shown in Figure 10c,d, it can be found that the Langmuir isotherm model was more reasonable for the description of the adsorption behavior of the Fe₃O₄@1 toward MB, which indicated that the recognition tendency was mainly because of the charge interactions referring to negative charge of 1 and the cationic of MB, hydrogen bonds and π–π stacking interactions [10,38,39]. The comparison experiments of Fe₃O₄, 1 and Fe₃O₄@1 also revealed that 1 of Fe₃O₄@1 might be the active site in adsorption process. The Fe₃O₄ nanoparticles of Fe₃O₄@1 were beneficial for the recovery of adsorbents (Figure S2). In order to affirm this assumption, cationic dye rhodamine B (RhB) and anion dye methyl orange (MO) were adopted to examine the adsorption behavior of Fe₃O₄@1 at the same condition. It turned out that the removal rate of RhB was 91.22%, while Fe₃O₄@1 has little effect on MO removal (Figure 10e,f). These results can be regarded as a solid support for the removal of cationic dyes in the presence of Fe₃O₄@1.

The reusability of nanocomposites plays a crucial role in developing reliable, economic and sustainable applications [40,41,42]. The reusability of Fe₃O₄@1 was investigated by means of retired adsorption cycles on the same sample (MB solution, 15 mg L⁻¹). In the process, the adsorbent was isolated by a magnet after every adsorption reaction and was washed by ethanol for six times and dried at 60 °C for 2 h. The adsorbent almost preserved the same adsorption capacity, even after seven runs with a small decline in the yield to 97.84%, 96%, 95.4%, 94.75%, 93.36%, 91.34% and 90.44%, respectively (Figure 11a). The variation of adsorption capacities between the first and seventh adsorption cycle was found to be less than 8%. The slight decrease might be caused by the loss of the used adsorbents in recovery process. It should be mentioned that further optimization of the elution might improve the recycling. Interestingly, the FTIR and XPS spectra were nearly identical before and after seven cycles of adsorption reaction, which demonstrated the high stability of Fe₃O₄@1 (Figure 11b,c). Thus, it is concluded that Fe₃O₄@1 has a good recyclability and remarkable stability in the removal of MB at present experiments. This aspect is very important in practical applications of economic and stringent ecological demands for sustainability.

3. Experiment

3.1. Materials and Methods

All reagents and chemicals were acquired from commercial sources. Nickel(II) perchlorate hexahydrate (Ni(ClO₄)₂·6H₂O, reagent grade), sodium molybdate dihydrate (Na₂MoO₄·2H₂O, 99%), phosphoric acid (H₃PO₄, 85%), 1,2-hexadecanediol (C₁₄H₂₉CH(OH)CH₂(OH), 98%), dioctyl ether (C₈H₁₇OC₈H₁₇, 96%), iron(III) acetylacetonate (Fe(acac)₃, 98%), Poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethyleneglycol) (PEO-PPO-PEO, Mr = 5800) were purchased from J&K Scientific Ltd. (Beijing, China). Methanol (CH₃OH, 99.5%), ethanol (CH₃CH₂OH, 99.7%), 2-acetylpyrazine (C₄N₂H₃COCH₃, 99%) and thiosemicarbazide (CH₅N₃S, 99%) were purchased from Aldrich (Shanghai, China). The Fe₃O₄ nanoparticles were prepared according to the article scheme [43]. X-ray powder diffraction patterns (PXRD) were tested with a Cu Kα radiation. A Flash 2000 analyzer was employed to test C, H, and N. Ni and Mo were confirmed by an inductively-coupled plasma spectrometer (Leaman, city, state abbrev if USA, Country). A Nicolet FTIR 360 spectrometer (company, city, state abbrev if USA, Country) was used to obtain the infrared (FTIR) spectrum. Ultraviolet-visible (UV-Vis) spectrum was captured by a TU–1900 spectrometer (company, city, state abbrev if USA, Country). The physical property measurement system (PPMS) and vibrating sample magnetometer (VSM, company, city, state abbrev if USA, Country) were used to explore the magnetic property. Transmission electron microscopy (TEM, JEOL2010F) and high resolution (HRTEM) were collected on a JEOL2010F (company, city, state abbrev if USA, Country). X-ray photoelectron spectrum (XPS) was studied by Al Kα X-ray as the excitation resource by a Thermo ESCALAB 250XI photoelectron spectrometer (company, city, state abbrev if USA, Country).

3.2. Preparations of [Ni(HL)₂]₂H₂[P₂Mo₅O₂₃]·2H₂O (1)

A mixture of Ni(ClO₄)₂·6H₂O (0.091 g, 0.25 mmol), 2-acetylpyazine-thiosemicarbazone (0.096 g, 0.5 mmol), methanol and water solution (25 mL, 3:2) was stirred for 30 minutes at 60 °C. After cooling, the above solution was mixed with 10 mL water with Na₂MoO₄·2H₂O (0.24 g, 1.0 mmol). The pH of the solution was maintained around 3.0 with concentrated H₃PO₄. Such obtained solution was stirred at 60 °C for another 30 minutes, then cooled and filtrated. The filtrate evaporated slowly at room temperature in an open beaker and brown rod crystals were obtained after five days. Yield: circa 65.80% (based on Ni). Elemental analysis for C₂₈H₄₂Mo₅N₂₀Ni₂O₂₅P₂S₄. Calculation (%): C 18.22, H 2.29, N 15.17, Ni 6.36, Mo 25.99; Found (%): C 18.36, H 2.60, N 15.41, Ni 6.45, Mo 26.14.

3.3. Preparation of Nanocomposites Fe₃O₄@1

Fe₃O₄ (7.5 mg) and 1 (50 mg) were added to a 10 mL solution (V_water:V_ethanol = 1:1) which was then dispersed with ultrasonic for 10 h. During the ultrasound process, the solution is thermostatic (30 °C). Then, the materials (Fe₃O₄@1) were segregated by a magnet and washed with water. The weight ratio of 1 in Fe₃O₄@1 is 89.57% (Figure S1).

3.4. X-Ray Crystallographic Study

Data of 1 were carried out on a Bruker D8 SMART APEX-II CCD X-ray diffraction with graphite-monochromated Mo Kα radiation (λ = 0.71073 Å) (company, city, state abbrev if USA, Country). Lorentz and polarization rectification were utilized, and an applied multi-scan assimilation rectification was achieved with the SADABS scheme. The structure was determined by direct methods with the SHELXTL program package [44,45]. All atoms were anisotropically refined—except for hydrogen. Hydrogen atoms were fixed geometrically. The detailed crystallographic data are given in Table S1. The CCDC number of 1 is 1918602.

3.5. Adsorption Experiments

The adsorption experiments were conducted in the dark at room temperature. Typically, 12 mg of the adsorbents were added to 10 mL MB solution (15 mg L⁻¹). The reaction suspension (4 mL) was taken out from the beaker periodically by a pipette. The isolated clear solution was explored by a UV-vis spectrophotometer. The removal efficiency and adsorption capacity (q_e) of adsorbents was calculated by Equations S1 and S2 [46]. The experimental info was plotted using pseudo-first order and pseudo-second order kinetic model (Equations S3 and S4) [47].

4. Conclusions

In conclusion, a magnetic adsorbent Fe₃O₄@1 was successfully prepared and systematically characterized. The adsorption behavior of Fe₃O₄@1 for organic dyes from aqueous solution was explored with MB as a model substrate. The ZFC-FC measurements showed that the blocking temperature of Fe₃O₄@1 was at 190 K. Although the low BET surface area is 15.05 m² g⁻¹, Fe₃O₄@1 displayed an effective adsorption efficiency of 97.84% for MB. The adsorption of MB depends on adsorbent usage, initial MB concentration and solution pH. Moreover, the potential reason of adsorption was discussed as well according to results from adsorption experiments, which revealed that the adsorption of MB through Fe₃O₄@1 was mainly caused by chemisorption. The maximal adsorption capacity of Fe₃O₄@1 was 41.91 mg g⁻¹. Furthermore, Fe₃O₄@1 demonstrated a great performance for selectively adsorbing cationic dyes. Lastly, the adsorbent can last at least for seven cycles of adsorption-desorption process with high stability, indicating that Fe₃O₄@1 has potential application in cationic dyes removal.