Magnetic Carbon Nanofibers as Potent Adsorbents for Phosphate Removal and Regeneration

Abstract

Researches on the removal and regeneration of phosphate from rivers and industrial wastewater have been accelerated gradually due to eutrophication and global phosphorus scarcity. Herein, magnetically recyclable carbon nanofiber (CNF) adsorbents were rationally designed by the coprecipitation method. Based upon preliminary filtering of magnetic CNFs with different Fe-to-CNF mass ratios in the aspect of phosphate sorption capacity, magnetic CNFs with an Fe-to-CNF mass ratio of 2:1 were selected to explore characterization and phosphorous adsorption performance further. Numerous adsorption experiments exhibited that magnetic CNF (Fe₃O₄-CNF₂) adsorbent showed a preferable phosphate uptake capacity of 7.26 mg P/g, a prompt sorption kinetic, strong interference immunity in the presence of low concentrations of anions, and an optimum range of pH (3–6). Furthermore, magnetic CNF (Fe₃O₄-CNF₂) adsorbent exhibited an excellent sorption performance in a secondary wastewater effluent. A 0.5 g/L of this adsorbent could reduce phosphate consistence from 0.95 to 0.05 mg P/L. In addition, exceeding 80% of initial phosphate uptake capacity could be preserved after five continuous regeneration cycles. This consequence showed the outstanding regenerative performance of Fe₃O₄-CNFs₂. Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and X-ray powder diffraction spectroscopy were used to reveal the phosphate adsorption mechanism, including surface precipitation and electrostatic attraction. Those developments are expected to have a far-reaching meaning for urban sewage treatment and eutrophic river.

Introduction

Phosphorous is a fundamental factor affecting the environmental quality of water bodies and an essential element in cellular composition (Hao et al., 2018). Industrial and agricultural runoff is the main source of excess phosphorus in water. Phosphorus pollution needs to be solved urgently because of limited advanced wastewater treatment technologies. There are plenty of conventional methods, such as biological processes, chemical precipitation, and ion-exchange. The biological processes are the most commonly applied because ammonia, phosphorus, and biological oxygen demand can be removed simultaneously by these processes (Kajjumba et al., 2019). Biological methods needed special care and rigid control (e.g., organic extent, nitrate and ammonium, water temperature) (Zheng et al., 2014).

Adsorption has been utilized for phosphate removal in the industrial wastewater according to its advantages of effectiveness and reliability (Loganathan et al., 2014). However, rapid and effective separation was considered a restrictive factor in practical application, such as activated carbon and industrial by-products. It was worth mentioning that magnetic separation might address this issue excellently ascribed to the properties of separating adsorbents effectively and conveniently in wastewater compared with sophisticated membrane filtration. Composites of iron and other carriers were necessarily studied.

Cellulose is one of the best candidates because there is a large amount of cellulose on the Earth, which is safe and a wide range of sources. Some researches indicated that it can function as the adsorbent due to adsorptive properties (Ma et al., 2011; Liu et al., 2015). In recent years, researchers have been working on the preparation of nanosorbents on account of the high surface area of nanomaterials (Abdelrahman and Hegazey, 2019; Abdelrahman et al., 2019). Therefore, nanofiber adsorbent is worthy to be explored in the actual application because of few studies in these aspects.

Consequently, the main objectives of this work are twofold: (1) utilize magnetic carbon nanofibers (Fe₃O₄-CNFs₂) to handle phosphorus-containing wastewater efficiently; and (2) elaborate underlying phosphorus adsorption mechanisms according to the consequences of batch experiments (e.g., influences of pH and coexisting anions), spectroscopic investigations (field emission scanning electron microscopy, energy-dispersive spectroscopy [EDS], energy dispersive X-ray photoelectron spectroscopy [XPS], Fourier transform infrared spectroscopy [FTIR], X-ray powder diffraction spectrometer [XRD]), and vibrating sample magnetometer (VSM). Corresponding results would have a better understanding of CNF materials (such as magnetic CNFs) in the aspect of removing phosphate. These magnetic CNFs would increase the effectiveness of the wastewater treatment.

Materials and Methods

Materials

The purity of Na₂CO₃, Na₂SO₄, NaNO₃, KH₂PO₄, FeCl₃▪6H₂O, FeSO₄▪7H₂O, and NaOH used in this study was analytical reagent (AR). CNFs were purchased from Chengdu Jiacai Technology Co., Ltd. Distilled water (DW) was used to configure all solutions.

Preparation of the adsorbents of magnetic CNFs (Fe₃O₄-CNFs)

CNFs were washed by DW and cut into 4 × 2 mm pieces. Then, they were washed with DW several times to remove the impurities. Since Fe²⁺ is easily oxidized, the molar ratio of Fe³⁺ to Fe²⁺ is 1:1. After that, ∼1.1674 g of FeCl₃▪6H₂O and 1.2007 g of FeSO₄▪7H₂O (a molar ratio of 1:1) were well dissolved into DW (100 mL). CNFs, 0.5 g, were added to the mixed solution and stirred at 60°C for 30 min. One molar NaOH was used to adjust the mixed solution to pH 11.0 and then 0.25 g trisodium citrate was added to the mixed solution. All samples were incubated at 80°C for 1 h, followed by cooling them to room temperature. Magnetic CNFs were washed several times with DW and ethanol until the filtrate became nearly neutral. Finally, samples were dried at 60°C overnight. The Fe₃O₄-CNFs with various mass ratios between Fe and CNFs of 0.5:1, 1:1, 2:1, and 5:1 were synthesized and named as Fe₃O₄-CNFs_0.5, Fe₃O₄-CNFs₁, Fe₃O₄-CNFs₂, and Fe₃O₄-CNFs₅, respectively.

Characterization of adsorbents

Scanning electron microscopy (SEM) was taken on an instrument equipped with EDS accessory. The composition of adsorbents was examined by using energy dispersive XPS (Thermo Scientific K-Alpha), all spectra were calibrated using the C1s signal located at 284.80 eV. A VSM (Lake Shore 7404) with external magnetic field intensity of 10,000 Oe was used to inspect the magnetic properties of adsorbents. The functional groups of sorbents were revealed by using an FTIR spectrometer (Thermo Scientific Nicolet iS5) in the range of 4,000–450 cm⁻¹. The crystalline structures of adsorbents were discussed using a XRD (Rigaku Ultima IV) equipped with Cu Kα radiation for 2θ ranging from 5° to 80°. The actual contents of Fe in each sample were determined using an inductively coupled plasma optical emission spectrometer (ICP-OES).

Batch adsorption experiments by magnetic CNFs (Fe₃O₄-CNFs₂)

According to the Chinese National Standard Water Quality Determination of total phosphorus-ammonium molybdate spectrophotometry method (GB 11893-89), 0.2197 ± 0.001 g of anhydrous KH₂PO₄ was dissolved in DW to obtain 50 mg/L phosphate solution, and then, we could prepare different concentrations of solution by diluting the mother solution. Throughout the test, the concentration of phosphate was determined with the UV-754N spectrophotometer (INESA). All phosphates determined in this work were orthophosphate unless otherwise specified.

Adsorption kinetics

To evaluate the adsorption kinetics, 0.05 g of Fe₃O₄-CNFs₂ was added to 100 mL phosphate solution (2 mg P/L) in a flask. The obtained mixture was shaken in a constant temperature shaker for 8 h at a shaking rate of 230 rpm. Finally, the adsorption capacity at equilibrium time (q_e, mg/g) during the phosphate adsorption process was calculated by Equation (1) (Liao et al., 2018): $q_{e} = \frac{V (C_{0} - C_{e})}{m}$ (1)

where C₀ and C_e (mg P/L) were the phosphate concentrations at the initial and equilibrium status, respectively. V (L) represented the bulk of the phosphate solution. m (g) was the quantity of adsorbent.

The phosphorus uptake efficiency was taken by the following Equation (2): $R = \frac{C_{0} - C_{e}}{C_{0}} \times 100 %$ (2)

The pseudofirst-order [Eq. (3)] and pseudosecond-order [Eq. (4)] kinetic models were applied to fit the curve according to the adsorption experimental data (He et al., 2017; Feng et al., 2020): $l g q_{e} - lg q_{e} - q_{t} = k_{1} t$ (3) $\frac{t}{q_{t}} = \frac{1}{k_{2} q_{e}^{2}} + \frac{t}{q_{e}}$ (4)

where q_e and q_t (mg/g) were the phosphorus uptake capacity at equilibrium and a specific time t, respectively. The k₁ (min⁻¹) and k₂ (g/[mg·min]) were the adsorption rate constants of the pseudofirst-order and pseudosecond-order kinetic models, respectively.

Adsorption isotherms

The adsorption isotherm about phosphorus capacity by Fe₃O₄-CNFs₂ can be analyzed by releasing 0.05 g of Fe₃O₄-CNFs₂ in 100 mL phosphate solutions with different primary concentrations (1, 2, 4, 6, and 8 mg P/L) in a flask. Those solutions were shaken at three different temperatures (25°C, 35°C, and 45°C) for 24 h. In addition, the phosphorus adsorption equilibrium data can be indicated by Langmuir [Eq. (5)] and Freundlich [Eq. (6)] isotherm models (Ning et al., 2008; Ozcan et al., 2009), as shown below: $L a n g m u i r m o d e l : \frac{1}{q_{e}} = \frac{1}{q_{m} K_{L}} \cdot \frac{1}{C_{e}} + \frac{1}{q_{m}}$ (5) $F r e u n d l i c h m o d e l : l g q_{e} = l g K_{f} + \frac{1}{n} l g C_{e}$ (6)

where q_m (mg P/g) represented the maximum Langmuir capacity, and K_L (L/mg) presented the Langmuir adsorption constant. C_e (mg P/L) was the equilibrium concentration in phosphate solution. K_f (mg/g) was the Freundlich constants associated with uptake capacity, and n was the adsorption intensity in the Freundlich model.

Effects of initial solution pH

Considering the influence of solution pH in the process of phosphate adsorption, 0.05 g of Fe₃O₄-CNFs₂ was added to 100 mL solution within pH ranging from 3.0 to 11.0 at 25°C ± 1°C when the solution concentration was 2 mg P/L. The pH value was altered by dosing 1 mol/L HCl or NaOH. After adsorption, 0.45 μm membrane filters were used in all samples before measurement.

Effects of coexisting anions

The effect of coexisting anions in the course of phosphate adsorption was performed by adding 0.05 g Fe₃O₄-CNFs₂ in 100 mL of 2 mg P/L solution, which contained 0.01 or 0.10 mol/L coexisting anions (CO₃²⁻, NO₃⁻, and SO₄²⁻).

Adsorbent regeneration

To test phosphate adsorption/desorption cycles of Fe₃O₄-CNFs₂, 0.050 g of Fe₃O₄-CNFs₂ was added into 100 mL of 2 mg P/L solution under the condition of 24-h shake at 25°C ± 1°C. After that, Fe₃O₄-CNFs₂ were collected and then dried at 80°C. In the desorption experiment, 100 mL of 1 mol/L NaOH solution was applied to treat this substance at 85°C ± 1°C for 18 h. We could calculate the desorption efficiency of phosphate through the ratio of the desorbed phosphate to the adsorbed phosphate. After desorption, this adsorbent was washed to neutrality by DW and dried in the oven. After completing the above operations, the adsorbent was applied to the next adsorption. The reusable performance of the regenerated adsorbent was evaluated by performing five continuous adsorption/desorption recycle experiments.

Results and Discussion

Effects of Fe₃O₄-to-CNF mass ratio on phosphate uptake

The phosphate removal efficiency of CNFs, Fe₃O₄-CNFs_0.5, Fe₃O₄-CNFs₁, Fe₃O₄-CNFs₂, and Fe₃O₄-CNFs₅ is compared in Fig. 1. As shown in Fig. 1, the sorption of phosphate by Fe₃O₄-CNFs increased distinctly as Fe₃O₄- to-CNF mass ratio increased from 1:2 to 2:1, then the phosphate removal efficiency decreased. This is likely due to the gradually increasing Fe amounts in the Fe₃O₄-CNF sorbents and the adsorbent load is not complete when the iron content is too high. The adsorbent phosphate removal efficiency of Fe₃O₄-CNFs_x was all higher than CNFs (6.3%). As can be seen in Fig. 1, the phosphate sorption capacity accelerated continuously when Fe contents in nanocomposites increased gradually, suggesting that the Fe compound functioned as the main active site for phosphate binding. Among the Fe₃O₄-CNFs_x, Fe₃O₄-CNFs₂ had the best phosphate adsorption capacity. Taking adsorption capacity into consideration, Fe₃O₄-CNFs₂ were applied for detailed study in terms of phosphorus removal.

FIG. 1.

Phosphate removal efficiency for CNFs, Fe₃O₄-CNFs_0.5, Fe₃O₄-CNFs₁, Fe₃O₄-CNFs₂, and Fe₃O₄-CNFs₅ in the solution with an initial phosphate concentration of 2 mg P/L. CNFs, carbon nanofibers.

Characterization of magnetic CNFs (Fe₃O₄-CNFs)

SEM micrographs in Fig. 2 depicted an extremely regular fibrous channel appearing in CNFs and its surface was smooth. As demonstrated, surfaces of all magnetic CNFs were entirely different from that of original CNFs. Fe₃O₄ particles were deposed on CNFs homogeneously. The iron oxide loaded on the surface of the carbon fiber can adsorb more phosphorus. Elemental distribution mapping (EDS) (Fig. 3) further confirmed that composites were modified successfully. The representative elements (Fe) of Fe₃O₄ were clearly observed and were concentrated all over the CNF structure. The corresponding EDS spectra apparently revealed that magnetic CNFs had high peaks for C and O, while additional peaks for Fe were detected for magnetic CNFs.

FIG. 2.

SEM images of (a) CNFs; (b) Fe₃O₄-CNFs₂; (c) Fe₃O₄-CNFs₂ after adsorption. SEM, scanning electron microscopy.

FIG. 3.

EDS spectra of samples (a) Fe₃O₄-CNFs₂; (b) Fe₃O₄-CNFs₂ after adsorption. EDS, energy-dispersive spectroscopy.

VSM was accustomed to measure the magnetic properties of Fe₃O₄-CNFs₂ at room temperature. Figure 4 shows the magnetic hysteresis loop of Fe₃O₄-CNFs₂. Figure 4a indicates that this substance was a type of super magnetism because the coercivity (Hc) of Fe₃O₄-CNFs₂ was close to zero (Zeng et al., 2009). The magnetization gradually increased with the increase of the external magnetic field and saturation magnetization (Hs) was 5.36 emu/g. Figure 4b shows that the magnetic material gathered around the magnet and could be quickly separated from the solution under the action of an external magnetic field.

FIG. 4.

(a) Magnetization curves of Fe₃O₄-CNFs₂; (b) photogram showing the effective separation of Fe₃O₄-CNFs₂ under an external magnetic field.

Further ICP-OES measurements indicated that the Fe content of Fe₃O₄-CNFs_0.5, Fe₃O₄-CNFs₁, Fe₃O₄-CNFs₂ and Fe₃O₄-CNFs₅ was 20.61%, 30.77%, 35.26%, and 31.25%, respectively. When the initial iron ion concentration was too high during the preparation process, the iron ion could not be completely loaded on the CNFs. This assumption could explain the phosphate sorption capacity of various adsorbents.

Phosphate sorption by magnetic CNFs (Fe₃O₄-CNFs₂)

Adsorption kinetics

The sorption experimental study of time influence was taken to evaluate rates of phosphate sorption of magnetic CNFs (Fe₃O₄-CNFs₂). As shown in Fig. 5, magnetic CNF (Fe₃O₄-CNF₂) adsorbent performed a fast sorption kinetic, the sorption equilibrium could be reached within 300 min, and over 80% of phosphate was removed. Phosphate was difficult to be captured from an aqueous solution to sorbent owing to inadequate active sites with the increase of preliminary concentration. However, contact time was only 5 h when the adsorption reached equilibrium in 2 mg P/L solution by 0.5 g/L sorbent, which is shorter than plenty of previously reported Fe-based sorbents such as Fe(OH)₃@CNFs (Luo et al., 2019) (nearly 11 h).

FIG. 5.

Phosphate adsorption efficiency. Experimental conditions for 2 and 4 mg P/L, 0.5 g/L of sorbent dosage, pH 7.0 ± 0.1, contact time 0–480 min.

The pseudofirst-order and the pseudosecond-order kinetic models were used to fit the adsorption kinetic data, and the major fitting parameters are shown in Table 1. The pseudosecond-order kinetic model (Fig. 6) generated a higher correlation coefficient (R²) of 0.9866 than the pseudofirst-order kinetic model (R² = 0.9433), suggesting that chemical adsorption could play a significant role in the phosphorus uptake process.

FIG. 6.

Phosphate adsorption kinetics: (a) pseudofirst-order model; (b) pseudosecond-order model.

Table 1.

Parameters for the Adsorption Kinetics Model of Magnetic Carbon Nanofibers (Fe₃O₄-CNFs₂)

T (°C)	Pseudofirst-order model			Pseudosecond-order model
T (°C)	q_e (mg/g)	k₁ (min⁻¹)	R²	q_e (mg/g)	k₂ (g/[mg·min])	R²
25	17.69	0.0177	0.9433	9.2081	0.0011	0.9866

Adsorption isotherms

In consideration of the possible sorption mechanism and evaluating the maximum phosphate sorption capacity of magnetic CNF (Fe₃O₄-CNF₂) sorbents, adsorption isothermal experiments were performed according to the initial phosphate concentration. To minimize the respective error functions, Langmuir and Freundlich models (Tran et al., 2017) were used to fit the experimental isotherm data.

As displayed obviously in Fig. 7, phosphate adsorption by magnetic CNF (Fe₃O₄-CNF₂) sorbents increased constantly when the initial phosphate concentration was added little by little and then gradually arrived at the maximum adsorption capacity of 7.26 mg P/g, outshining CNFs (Yao et al., 2011). The Freundlich and Langmuir models were used to make the adsorption isotherms data curve with the related parameters, illustrated in Table 2. With regard to R² values, the Langmuir model (R² = 0.9496, 0.8911, 0.9970) is more adaptive for describing the phosphorus adsorption mechanism onto magnetic CNF (Fe₃O₄-CNF₂) sorbents than that of the Freundlich model (R² = 0.9873, 0.7394, 0.9930), implying that the distribution of active sites on magnetic CNF (Fe₃O₄-CNF₂) sorbents was uniform and the phosphate sorption mode was monolayer. Table 3 lists the adsorption capacity of different magnetic adsorbents and it could conclude that Fe₃O₄-CNFs₂ had better adsorption capacity toward phosphate compared with the other modified sorbents when adsorbents tested under the same condition.

FIG. 7.

Phosphate adsorption isotherms: (a) Langmuir models; (b) Freundlich models. Experimental conditions for 2 mg P/L, 0.5 g/L, pH 7.0 ± 0.1, 25–45°C, 24 h.

Table 2.

Parameters for the Adsorption Kinetics Model of Magnetic Carbon Nanofibers (Fe₃O₄-CNFs₂)

Temperature	Langmuir			Freundlich
Temperature	q_e (mg/g)	K_L (L/mg)	R²	K_F (mg^{1 −} ⁿ ·L ⁿ /g)	1/n	R²
25°C	7.70	0.12	0.9496	4.30	0.25	0.9873
35°C	6.00	0.13	0.8911	3.26	0.36	0.7394
45°C	4.45	0.46	0.9970	2.30	0.41	0.9930

Table 3.

The Comparison of Phosphate Adsorption Capacity of Different Adsorbents

Adsorbent	Qmax (mg/g)	References
Magnetic iron oxide nanoparticles	5.03	Yoon et al. (2014)
Amorphous magnetic ZrO₂/Fe₃O₄	29.5	Wang et al. (2020)
Magnetic lignin-based nanoparticles	0.012	Li et al. (2021)
Zirconium- and iminodiacetic acid-modified magnetic peanut husk	13.2	Aryee et al. (2021)
Zerovalent iron supported activated carbon	1.75	Khalil et al. (2017)
Magnetic sugarcane biochar	2.47	Li et al. (2016)
Magnetic carbon nanofibers	7.26	This study

Effects of initial solution pH

Taking account of factors affected by pH (surface charges of Fe₃O₄-CNFs₂ and existing species of phosphate in different pH solutions), sorption experiments were taken to explore the influence of initial solution pH on phosphorus uptake when the pH ranged from 3.0 to 11.0. In Fig. 8, initial solution pH made evident influence on phosphate sorption by magnetic CNF (Fe₃O₄-CNF₂) sorbents. Specifically, in the acid condition (3.0–5.0), phosphate uptake efficiency by this magnetic nanocomposite enhanced step by step. With a further pH increasing from 5.0 to 11.0, phosphate adsorption by magnetic CNF (Fe₃O₄-CNF₂) sorbents dropped dramatically. In general, pH has an effect on phosphorus adsorption.

FIG. 8.

Effects of initial solution pH. Experimental conditions for 2 mg P/L, 0.5 g/L, pH 3.0–11.0, 25°C, 24 h.

This could have been explained by two reasons. One was that pH influenced the surface charge of magnetic CNFs (Fe₃O₄-CNFs₂). The other was that pH had the impact on the ionization state of phosphate in the aqueous solution. The reaction of phosphate dissociation in the aqueous medium could be stated as the following equation (Ajmal et al., 2018): $H_{3} P O_{4} + H_{2} O ⇌ H_{2} P O_{4}^{-} + H_{3} O^{+} p K_{1} = 2.15$ (7) $H_{2} P O_{4}^{-} + H_{2} O ⇌ H P O_{4}^{2 -} + H_{3} O^{+} p K_{2} = 7.20$ (8)

H P O_{4}^{2 -} + H_{2} O ⇌ P O_{4}^{3 -} + H_{3} O^{+} p K_{3} = 12.33

(9)

Under highly acidic condition (pH 3.0), phosphate existed as H₃PO₄ molecule that was electrically neutral, which rejected the uptake of phosphate through electronic interaction. With the pH ranging from 3.0 to 5.0, monovalent H₂PO₄⁻ played a dominant role. Magnetic CNF (Fe₃O₄-CNF₂) sorbent surface was positively charged because active groups indicated protonation and facilitated the phosphate uptake quickly through electrostatic attraction. When the solution pH was above 5, we could witness an increasing portion of HPO₄²⁻, which was lower sorption affinity. This result illustrated that protonation of sorbent trended to alleviate, resulting that phosphate sorption capacity was a slight decrease.

The experimental statistics in Fig. 9 illustrated that the pH_pzc of magnetic CNFs (Fe₃O₄-CNFs₂) was calculated as 7.07 through the pH drift method (Nasiruddin Khan and Sarwar, 2007). When pH is below pH_pzc, the surface charge on magnetic CNFs (Fe₃O₄-CNFs₂) would appear positive and H₂PO₄⁻ (dominant species) was likely to be attracted by Fe₃O₄-CNFs₂ (Harijan and Chandra, 2017). However, the adsorbent would become a negative charge at pH > pH_pzc and the electrostatic repulsion between Fe₃O₄-CNFs₂ and phosphate would accelerate (Paltrinieri et al., 2017). Moreover, the presence of excess amount of OH⁻ would occupy the adsorption sites, thereby resulting in a lower phosphate removal efficiency under alkaline conditions.

FIG. 9.

Point of zero charge of magnetic CNF (Fe₃O₄-CNF₂) adsorbents by pH drift method.

Effects of coexisting anions

Considering that natural water bodies and industrial wastewaters exist interfering substances, where plenty of coexisting anions may reduce phosphate adsorption capacity via contending active sites, it is urgent to inspect the selectivity of magnetic CNFs (Fe₃O₄-CNFs₂) for phosphate uptake (Awual, 2019). Hence, three kinds of common anions (i.e., SO₄²⁻, NO₃⁻, and CO₃²⁻) and two different concentrations of anion were researched in the sorption experiments. As presented in Fig. 10, introducing exogenous anions had no or minimal effect on phosphate uptake by magnetic CNF (Fe₃O₄-CNF₂) sorbents. It may be the main reason for specific selectivity that there is specific binding between Fe and phosphate because of high affinity toward each other. Besides, owing that the increment of coexisting anion concentrations barely affected the process of the phosphate uptake by magnetic CNF (Fe₃O₄-CNF₂) sorbents, it was speculated that inner-sphere complexes formed between phosphate and Fe₃O₄-CNFs₂ via ligand exchange. The influence of phosphate removal efficiency made the difference with the increase in ion strength (Nakarmi et al., 2020). Overall, such a preferable selectivity of the concerned magnetic CNF (Fe₃O₄-CNF₂) sorbent could further favor its practical utilization for urban sewage treatment and eutrophic river.

FIG. 10.

Effects of coexisting anions. Experimental conditions for 2 mg P/L, 0.5 g/L, pH = 7 ± 0.1, 24 h.

Desorption and regeneration of magnetic CNFs (Fe₃O₄-CNFs₂)

From a practical viewpoint, it is of crucial importance for sorbent to determine the lifetime. The duration can be increased by reusing it multiple times. The reusability of magnetic CNF (Fe₃O₄-CNF₂) sorbents can be explored by five successive adsorption cycles. From Fig. 11 it can be seen that, compared with fresh sorbent, nearly 80% of phosphate sorption capacity was still held after five desorption cycles. This consequence showed that this adsorbent had a strong reusability. Some irreversible and useless sorption sites were the main reason for the declining sorption capacity because a strong chemical bond between phosphate and Fe was formed. Nevertheless, the magnetic CNF (Fe₃O₄-CNF₂) sorbent still showed favorable application potential in terms of effective removal and recovery of phosphate in aqueous solution.

FIG. 11.

The reusability of magnetic CNF (Fe₃O₄-CNF₂) sorbents under five consecutive sorption cycles.

Application of Fe₃O₄-CNFs₂ in real wastewater

A secondary wastewater effluent with a phosphate concentration of 0.95 mg/L was applied to probe the application of Fe₃O₄-CNFs₂ in real wastewater. At a quantity of 0.5 g/L Fe₃O₄-CNFs₂, the phosphorus removal efficiency reached 90.6% within 100 min and the phosphorus concentration could effectively decrease from 0.95 to 0.05 mg P/L within 150 min, which was far below the emission level stipulated by China (Shepherd et al., 2016).

Mechanism of adsorption

To clarify possible sorption mechanisms in the process of phosphate uptake by magnetic CNF (Fe₃O₄-CNF₂) sorbent, a series of complementary analyses were performed. On the one hand, surface precipitation played a significant role in the adsorption process. The exhausted magnetic CNF (Fe₃O₄-CNF₂) adsorbent was characterized by SEM coupling with EDS spectra. As displayed in Fig. 2c, sorbents after adsorption maintained the original fibrous-like morphology and obvious structural damage was not observed. Instead, the surface of adsorbent was filled with a large number of compact microstructures, making the surface smooth, which might be attributed to mass formation of FePO₄. In Fig. 3b, the phosphorus element clearly appeared in the EDS analysis after adsorption. On the other hand, electrostatic attraction was also the adsorption mechanism to dominate the adsorption process. The surface charge on the magnetic CNF (Fe₃O₄-CNF₂) sorbent was positive because pH_pzc was 7.06. Given the anionic character of phosphate, dissimilar charges promoted electrostatic attraction of the negatively charged phosphate ion.

XRD and FTIR analysis of Fe₃O₄-CNFs₂ was carried out to have a deep exploration into the mechanisms during the phosphate uptake process. For Fe₃O₄-CNF₂ adsorbents, almost all the peak positions were consistent with the standard data for Fe₃O₄ (JCPDS No. 19-0629) (Fig. 12a), indicating that magnetic CNFs were successfully synthesized (Mi et al., 2017). In Fig. 12b, the peak at 2,930 cm⁻¹ was related to the C-H symmetric stretching vibration absorption peak. The peak positioned at 1,625 cm⁻¹ was assigned to C-C vibration. Two new peaks in the region of 522–658 cm⁻¹ represented the Fe-O stretching mode of the Fe₃O₄ (Lopez et al., 2010). After phosphate sorption, the peak observed at 1,052 cm⁻¹ belonged to the symmetric stretching vibration of P-O in HPO₄²⁻ or H₂PO₄⁻ groups.

FIG. 12.

(a) XRD pattern of Fe₃O₄-CNFs₂; (b) FTIR spectra of Fe₃O₄-CNFs₂ before and after phosphate uptake. FTIR, Fourier transform infrared spectroscopy; XRD, X-ray diffraction.

XPS spectra of Fe₃O₄-CNFs₂ before and after adsorption were applied to analyze the component further, and Fig. 13a presents the full scan spectrum. Figure 13a reveals that C, O, and Fe were the main composition elements of the materials and proved the successful adsorption of phosphate by Fe₃O₄-CNFs₂. High-resolution Fe 2p XPS spectrum in Fig. 13b showed that peaks at 724.2 and 710.4 eV shifted toward higher binding energy to 725.7 and 711.2 eV. This phenomenon illustrated that electron transfer occurred in the bond of Fe 2p, and inner-sphere surface formed Fe-O-P complexation after phosphate uptake (Liu et al., 201). As for P 2p (Fig. 13c), its binding energy (133.28 eV) was different from pure KH₂PO₄ (134.0 eV), suggesting that strong chemical bonds between phosphate and Fe₃O₄-CNFs₂ existed. High-resolution spectra of C 1s (Fig. 13d) showed its binding energy was nearly 284.58 eV, which was due to the reference carbon of the hydrocarbon (Rodrigues-Filho et al., 1996). O 1s XPS in Fig. 13e showed that two peaks at 530.18 and 532.58 eV were observed in the spectrum of Fe₃O₄-CNFs₂. It is speculated that two specific peaks were attributed to Fe-O-Fe and C-O, respectively.

FIG. 13.

XPS spectra of Fe₃O₄-CNF₂ sorbent before and after phosphate uptake. (a) Wide-scan spectra; (b) Fe 2p; (c) P 2p; (d) C 1s; (e) O 1s. XPS, X-ray photoelectron spectroscopy.

Conclusion

In this work, magnetic CNFs (Fe₃O₄-CNFs₂) with various Fe-to-CNF mass ratios were fabricated by the coprecipitation method and calcination process to remove and recover phosphate from wastewater efficiently. Experimental results illustrated that magnetic CNF (Fe₃O₄-CNF₂) adsorbent showed a sorption capacity of 7.26 mg P/g and fast sorption kinetics. Phosphate adsorption could reach equilibrium at 300 min when the initial phosphate concentration was 2 mg P/L and Fe₃O₄-CNF₂ sorbent dosage was 0.5 g/L. Moreover, the concerned sorbent showed a good sorption selectivity when various interfering anions existed and Fe₃O₄-CNFs₂ made a difference efficiently at a wide pH, which ranged from 3 to 6. Fe₃O₄-CNFs₂ owned large amounts of merits, such as high magnetic separation, good reusability, and high economy. Phosphate desorption tests indicated that NaOH under thermal treatment could regenerate exhausted sorbent validly and recovering high amounts of phosphate in eluate was economically viable. Characterization results revealed that phosphate adsorption mechanism by magnetic CNFs (Fe₃O₄-CNFs₂) could be ascribed to surface precipitation and electrostatic attraction. In summary, magnetic CNF (Fe₃O₄-CNF₂) sorbent could be considered a conceivable sorbent for water purifications (i.e., weakening eutrophication and wastewater treatment) and phosphate recovery owing to the benefits of good adsorption performance and easy separation.

Footnotes

Ethical Approval

Compliance with ethical standards. All authors agreed to participate in this experiment. All authors agreed to publish the article.

Data Availability Statement

All data generated or analyzed during this study are included in this published article.

Acknowledgments

Donghua University provided all the experimental conditions. Authors thank for the technical support provided by Fenfen Zhang, the administrator of ICP-AES, Analysis and Testing Center, Donghua University.

Authors' Contributions

T.Z. conducted feasibility analysis, designed the experiments, processed most of the experimental data, drew charts, and was a major contributor in writing the article. Z.Y. provided innovative ideas and reviewed and edited the first draft of the article. W.S. helped prepare the experimental materials and did part of the experiment. Z.X. provided technical support in chart making and helped prepare the experimental materials. All authors read and approved the final article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was funded by the Shanghai Science and Technology Development Foundation (033919457).

References

Abdelrahman

E.A.

, and Hegazey

R.M.

(2019). Exploitation of Egyptian insecticide cans in the fabrication of Si/Fe nanostructures and their chitosan polymer composites for the removal of Ni(II), Cu(II), and Zn(II) ions from aqueous solutions. Compos. Part B Eng. 166, 382.

Abdelrahman

E.A.

, Hegazey

R.M.

, Kotp

Y.H.

, and Alharbi

(2019). Facile synthesis of Fe₂O₃ nanoparticles from Egyptian insecticide cans for efficient photocatalytic degradation of methylene blue and crystal violet dyes. Spectrochim. Acta A Mol. Biomol. Spectrosc. 222, 117195.

Ajmal

, Muhmood

, Usman

, Kizito

, Lu

, Dong

, and Wu

(2018). Phosphate removal from aqueous solution using iron oxides: Adsorption, desorption and regeneration characteristics. J. Colloid. Interf. Sci. 528, 145.

Aryee

A.A.

, Dovi

, Shi

, Han

, Li

, and Qu

(2021). Zirconium and iminodiacetic acid modified magnetic peanut husk as a novel adsorbent for the sequestration of phosphates from solution: Characterization, equilibrium and kinetic study. Colloid. Surf. A, 615, 126260.

Awual

M.R.

(2019). Efficient phosphate removal from water for controlling eutrophication using novel composite adsorbent. J. Clean. Prod. 228, 1311.

Feng

, Jiang

, Yuan

, Zhang

, and Zhang

(2020). Enhanced removal of aqueous phosphorus by Zr-Fe-, Mn-Fe-, and Mn-Zr-Fe-modified natural zeolites: Comparison studies and adsorption mechanism. Environ. Eng. Sci. 37, 572.

Hao

, Zhou

, Li

, and Wang

(2018). A 3DBER-S-EC process for simultaneous nitrogen and phosphorus removal from wastewater with low organic carbon content. J. Environ. Manage. 209, 57.

Harijan

D.K.L.

, and Chandra

(2017). Akaganeite nanorods decorated graphene oxide sheets for removal and recovery of aqueous phosphate. J. Water Process Eng. 19, 120.

, Lin

, Dong

, and Wang

(2017). Preferable adsorption of phosphate using lanthanum-incorporated porous zeolite: Characteristics and mechanism. Appl. Surf. Sci. 426, 995.

10.

Kajjumba

G.W.

, Yıldırım

, Aydın

, Emik

, Ağun

, Osra

, and Wasswa

(2019). A facile polymerisation of magnetic coal to enhanced phosphate removal from solution. J. Environ. Manage. 247, 356.

11.

Khalil

A.M.E.

, Eljamal

, Amen

T.W.M.

, Sugihara

, and Matsunaga

(2017). Optimized nano-scale zero-valent iron supported on treated activated carbon for enhanced nitrate and phosphate removal from water. Chem. Eng. J. 309, 349.

12.

, Wang

J.J.

, Zhou

, Awasthi

M.K.

, Ali

, Zhang

, Lahori

A.H.

, and Mahar

(2016). Recovery of phosphate from aqueous solution by magnesium oxide decorated magnetic biochar and its potential as phosphate-based fertilizer substitute. Bioresour. Technol. 215, 209.

13.

, Lü

, Wang

, Huang

, Yan

, and Liu

(2021). Lignin-based nanoparticles for recovery and separation of phosphate and reused as renewable magnetic fertilizers. Sci. Total Environ. 765, 142745.

14.

Liao

, Li

, Su

, Yu

, Song

, Zhu

, and Zhang

(2018). La(OH)₃-modified magnetic pineapple biochar as novel adsorbents for efficient phosphate removal. Bioresour. Technol. 263, 207.

15.

Liu

, Borrell

P.F.

, Bozic

, Kokol

, Oksman

, and Mathew

A.P.

(2015). Nanocelluloses and their phosphorylated derivatives for selective adsorption of Ag⁺, Cu²⁺ and Fe³⁺ from industrial effluents. J. Hazard. Mater. 294, 177.

16.

Liu

, Chi

, Wang

, Sui

, Xie

, and Arandiyan

(2019). Effective and selective adsorption of phosphate from aqueous solution via trivalent-metals-based amino-MIL-101 MOFs. Chem. Eng. J. 357, 159.

17.

Loganathan

, Vigneswaran

, Kandasamy

, and Bolan

N.S.

(2014). Removal and recovery of phosphate from water using sorption. Crit. Rev. Env. Sci. Tec. 44, 847.

18.

Lopez

J.A.

, González

, Bonilla

F.A.

, Zambrano

, and Gómez

M.E.

(2010). Synthesis and characterization of Fe₃O₄ magnetic nanofluid. Rev. Latinoam. Metal. Mater. 30, 60.

19.

Luo

, Liu

, Chen

, Wang

, and Zhang

(2019). Preparation and regeneration of iron-modified nanofibres for low-concentration phosphorus-containing wastewater treatment. R. Soc. Open Sci. 6, 190764.

20.

, Li

, Yue

, Gao

, Li

, Xu

, and Zhong

(2011). Adsorption removal of ammonium and phosphate from water by fertilizer controlled release agent prepared from wheat straw. Chem. Eng. J. 171, 1209.

21.

, Wang

, Zhou

, Chai

, Wang

, Zhang

, Meng

, Shang

, Zhao

, and Li

(2017). Preparation of La-modified magnetic composite for enhanced adsorptive removal of tetracycline. Environ. Sci. Pollut. Res. 24, 17127.

22.

Nakarmi

, Bourdo

S.E.

, Ruhl

, Kanel

, Nadagouda

, Kumar Alla

, Pavel

, and Viswanathan

(2020). Benign zinc oxide betaine-modified biochar nanocomposites for phosphate removal from aqueous solutions. J. Environ. Manage. 272, 111048.

23.

Nasiruddin Khan

, and Sarwar

(2007). Determination of points of zero charge of natural and treated adsorbents. Surf. Rev. Lett. 14, 461.

24.

Ning

, Bart

H.-J.

, Li

, Lu

, and Zhang

(2008). Phosphate removal from wastewater by model-La(III) zeolite adsorbents. J. Environ. Sci. 20, 670.

25.

Ozcan

A.S.

, Gok

, and Ozcan

(2009). Adsorption of lead(II) ions onto 8-hydroxy quinoline-immobilized bentonite. J. Hazard. Mater. 161, 499.

26.

Paltrinieri

, Wang

, Sachdeva

, Besseling

N.A.M.

, Sudhölter

E.J.R.

, and De Smet

L.C.P

.M. (2017). Fe₃O₄ nanoparticles coated with a guanidinium-functionalized polyelectrolyte extend the pH range for phosphate binding. J. Mater. Chem. A, 5, 18476.

27.

Rodrigues-Filho

U.P.

, Gushikem

, Do Carmo Gonçalves

, Cachichi

R.C.

, and De Castro

S.C.

(1996). Composite membranes of cellulose acetate and zirconium dioxide: Preparation and study of physicochemical characteristics. Chem. Mater. 8, 1375.

28.

Shepherd

J.G.

, Sohi

S.P.

, and Heal

K.V.

(2016). Optimising the recovery and reuse of phosphorus from wastewater effluent for sustainable fertiliser development. Water Res. 94, 155.

29.

Tran

H.N.

, You

S.-J.

, Hosseini-Bandegharaei

, and Chao

H.-P.

(2017). Mistakes and inconsistencies regarding adsorption of contaminants from aqueous solutions: A critical review. Water Res. 120, 88.

30.

Wang

, Shao

, Liu

, Zhang

, Ma

, and Tian

(2020). Insight into the effect of structural characteristics of magnetic ZrO₂/Fe₃O₄ nanocomposites on phosphate removal in water. Mater. Chem. Phys. 249, 123024.

31.

Yao

, Gao

, Inyang

, Zimmerman

A.R.

, Cao

, Pullammanappallil

, and Yang

(2011). Removal of phosphate from aqueous solution by biochar derived from anaerobically digested sugar beet tailings. J. Hazard. Mater. 190, 501.

32.

Yoon

S.-Y.

, Lee

C.-G.

, Park

J.-A.

, Kim

J.-H.

, Kim

S.-B.

, Lee

S.-H.

, and Choi

J.-W.

(2014). Kinetic, equilibrium and thermodynamic studies for phosphate adsorption to magnetic iron oxide nanoparticles. Chem. Eng. J. 236, 341.

33.

Zeng

, Xu

, Wang

, Tao

, and Hua

(2009). Ferromagnetic behavior of copper oxide-nanowire-covered carbon fibre synthesized by thermal oxidation. Mater. Charact. 60, 1068.

34.

Zheng

, Sun

, Han

, Song

, Hu

, Fan

, and Lv

(2014). Inhibitory factors affecting the process of enhanced biological phosphorus removal (EBPR)—A mini-review. Process Biochem. 49, 2207.

Magnetic Carbon Nanofibers as Potent Adsorbents for Phosphate Removal and Regeneration

Abstract

Introduction

Materials and Methods

Materials

Preparation of the adsorbents of magnetic CNFs (Fe3O4-CNFs)

Characterization of adsorbents

Batch adsorption experiments by magnetic CNFs (Fe3O4-CNFs2)

Adsorption kinetics

Adsorption isotherms

Effects of initial solution pH

Effects of coexisting anions

Adsorbent regeneration

Results and Discussion

Effects of Fe3O4-to-CNF mass ratio on phosphate uptake

Characterization of magnetic CNFs (Fe3O4-CNFs)

Phosphate sorption by magnetic CNFs (Fe3O4-CNFs2)

Adsorption kinetics

Adsorption isotherms

Effects of initial solution pH

Effects of coexisting anions

Desorption and regeneration of magnetic CNFs (Fe3O4-CNFs2)

Application of Fe3O4-CNFs2 in real wastewater

Mechanism of adsorption

Conclusion

Footnotes

Ethical Approval

Data Availability Statement

Acknowledgments

Authors' Contributions

Author Disclosure Statement

Funding Information

References

Preparation of the adsorbents of magnetic CNFs (Fe₃O₄-CNFs)

Batch adsorption experiments by magnetic CNFs (Fe₃O₄-CNFs₂)

Effects of Fe₃O₄-to-CNF mass ratio on phosphate uptake

Characterization of magnetic CNFs (Fe₃O₄-CNFs)

Phosphate sorption by magnetic CNFs (Fe₃O₄-CNFs₂)

Desorption and regeneration of magnetic CNFs (Fe₃O₄-CNFs₂)

Application of Fe₃O₄-CNFs₂ in real wastewater